diff options
Diffstat (limited to 'drivers/net/ethernet/mellanox/mlxsw')
80 files changed, 85504 insertions, 0 deletions
diff --git a/drivers/net/ethernet/mellanox/mlxsw/Kconfig b/drivers/net/ethernet/mellanox/mlxsw/Kconfig new file mode 100644 index 000000000..a510bf2cf --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/Kconfig @@ -0,0 +1,92 @@ +# SPDX-License-Identifier: GPL-2.0-only +# +# Mellanox switch drivers configuration +# + +config MLXSW_CORE + tristate "Mellanox Technologies Switch ASICs support" + select NET_DEVLINK + select MLXFW + select AUXILIARY_BUS + help + This driver supports Mellanox Technologies Switch ASICs family. + + To compile this driver as a module, choose M here: the + module will be called mlxsw_core. + +config MLXSW_CORE_HWMON + bool "HWMON support for Mellanox Technologies Switch ASICs" + depends on MLXSW_CORE && HWMON + depends on !(MLXSW_CORE=y && HWMON=m) + default y + help + Say Y here if you want to expose HWMON interface on mlxsw devices. + +config MLXSW_CORE_THERMAL + bool "Thermal zone support for Mellanox Technologies Switch ASICs" + depends on MLXSW_CORE && THERMAL + default y + help + Say Y here if you want to automatically control fans speed according + ambient temperature reported by ASIC. + +config MLXSW_PCI + tristate "PCI bus implementation for Mellanox Technologies Switch ASICs" + depends on PCI && HAS_IOMEM && MLXSW_CORE + default m + help + This is PCI bus implementation for Mellanox Technologies Switch ASICs. + + To compile this driver as a module, choose M here: the + module will be called mlxsw_pci. + +config MLXSW_I2C + tristate "I2C bus implementation for Mellanox Technologies Switch ASICs" + depends on I2C && MLXSW_CORE + default m + help + This is I2C bus implementation for Mellanox Technologies Switch ASICs. + + To compile this driver as a module, choose M here: the + module will be called mlxsw_i2c. + +config MLXSW_SPECTRUM + tristate "Mellanox Technologies Spectrum family support" + depends on MLXSW_CORE && MLXSW_PCI && NET_SWITCHDEV && VLAN_8021Q + depends on PSAMPLE || PSAMPLE=n + depends on BRIDGE || BRIDGE=n + depends on IPV6 || IPV6=n + depends on NET_IPGRE || NET_IPGRE=n + depends on IPV6_GRE || IPV6_GRE=n + depends on VXLAN || VXLAN=n + depends on PTP_1588_CLOCK_OPTIONAL + select GENERIC_ALLOCATOR + select PARMAN + select OBJAGG + select NET_PTP_CLASSIFY if PTP_1588_CLOCK + default m + help + This driver supports Mellanox Technologies + Spectrum/Spectrum-2/Spectrum-3/Spectrum-4 Ethernet Switch ASICs. + + To compile this driver as a module, choose M here: the + module will be called mlxsw_spectrum. + +config MLXSW_SPECTRUM_DCB + bool "Data Center Bridging (DCB) support" + depends on MLXSW_SPECTRUM && DCB + default y + help + Say Y here if you want to use Data Center Bridging (DCB) in the + driver. + +config MLXSW_MINIMAL + tristate "Mellanox Technologies minimal I2C support" + depends on MLXSW_CORE && MLXSW_I2C + default m + help + This driver supports I2C access for Mellanox Technologies Switch + ASICs. + + To compile this driver as a module, choose M here: the + module will be called mlxsw_minimal. diff --git a/drivers/net/ethernet/mellanox/mlxsw/Makefile b/drivers/net/ethernet/mellanox/mlxsw/Makefile new file mode 100644 index 000000000..3ca9fce75 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/Makefile @@ -0,0 +1,36 @@ +# SPDX-License-Identifier: GPL-2.0 +obj-$(CONFIG_MLXSW_CORE) += mlxsw_core.o +mlxsw_core-objs := core.o core_acl_flex_keys.o \ + core_acl_flex_actions.o core_env.o \ + core_linecards.o core_linecard_dev.o +mlxsw_core-$(CONFIG_MLXSW_CORE_HWMON) += core_hwmon.o +mlxsw_core-$(CONFIG_MLXSW_CORE_THERMAL) += core_thermal.o +obj-$(CONFIG_MLXSW_PCI) += mlxsw_pci.o +mlxsw_pci-objs := pci.o +obj-$(CONFIG_MLXSW_I2C) += mlxsw_i2c.o +mlxsw_i2c-objs := i2c.o +obj-$(CONFIG_MLXSW_SPECTRUM) += mlxsw_spectrum.o +mlxsw_spectrum-objs := spectrum.o spectrum_buffers.o \ + spectrum_switchdev.o spectrum_router.o \ + spectrum1_kvdl.o spectrum2_kvdl.o \ + spectrum_kvdl.o \ + spectrum_acl_tcam.o spectrum_acl_ctcam.o \ + spectrum_acl_atcam.o spectrum_acl_erp.o \ + spectrum1_acl_tcam.o spectrum2_acl_tcam.o \ + spectrum_acl_bloom_filter.o spectrum_acl.o \ + spectrum_flow.o spectrum_matchall.o \ + spectrum_flower.o spectrum_cnt.o \ + spectrum_fid.o spectrum_ipip.o \ + spectrum_acl_flex_actions.o \ + spectrum_acl_flex_keys.o \ + spectrum1_mr_tcam.o spectrum2_mr_tcam.o \ + spectrum_mr_tcam.o spectrum_mr.o \ + spectrum_qdisc.o spectrum_span.o \ + spectrum_nve.o spectrum_nve_vxlan.o \ + spectrum_dpipe.o spectrum_trap.o \ + spectrum_ethtool.o spectrum_policer.o \ + spectrum_pgt.o +mlxsw_spectrum-$(CONFIG_MLXSW_SPECTRUM_DCB) += spectrum_dcb.o +mlxsw_spectrum-$(CONFIG_PTP_1588_CLOCK) += spectrum_ptp.o +obj-$(CONFIG_MLXSW_MINIMAL) += mlxsw_minimal.o +mlxsw_minimal-objs := minimal.o diff --git a/drivers/net/ethernet/mellanox/mlxsw/cmd.h b/drivers/net/ethernet/mellanox/mlxsw/cmd.h new file mode 100644 index 000000000..09bef04b1 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/cmd.h @@ -0,0 +1,1281 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_CMD_H +#define _MLXSW_CMD_H + +#include "item.h" + +#define MLXSW_CMD_MBOX_SIZE 4096 + +static inline char *mlxsw_cmd_mbox_alloc(void) +{ + return kzalloc(MLXSW_CMD_MBOX_SIZE, GFP_KERNEL); +} + +static inline void mlxsw_cmd_mbox_free(char *mbox) +{ + kfree(mbox); +} + +static inline void mlxsw_cmd_mbox_zero(char *mbox) +{ + memset(mbox, 0, MLXSW_CMD_MBOX_SIZE); +} + +struct mlxsw_core; + +int mlxsw_cmd_exec(struct mlxsw_core *mlxsw_core, u16 opcode, u8 opcode_mod, + u32 in_mod, bool out_mbox_direct, bool reset_ok, + char *in_mbox, size_t in_mbox_size, + char *out_mbox, size_t out_mbox_size); + +static inline int mlxsw_cmd_exec_in(struct mlxsw_core *mlxsw_core, u16 opcode, + u8 opcode_mod, u32 in_mod, char *in_mbox, + size_t in_mbox_size) +{ + return mlxsw_cmd_exec(mlxsw_core, opcode, opcode_mod, in_mod, false, + false, in_mbox, in_mbox_size, NULL, 0); +} + +static inline int mlxsw_cmd_exec_out(struct mlxsw_core *mlxsw_core, u16 opcode, + u8 opcode_mod, u32 in_mod, + bool out_mbox_direct, + char *out_mbox, size_t out_mbox_size) +{ + return mlxsw_cmd_exec(mlxsw_core, opcode, opcode_mod, in_mod, + out_mbox_direct, false, NULL, 0, + out_mbox, out_mbox_size); +} + +static inline int mlxsw_cmd_exec_none(struct mlxsw_core *mlxsw_core, u16 opcode, + u8 opcode_mod, u32 in_mod) +{ + return mlxsw_cmd_exec(mlxsw_core, opcode, opcode_mod, in_mod, false, + false, NULL, 0, NULL, 0); +} + +enum mlxsw_cmd_opcode { + MLXSW_CMD_OPCODE_QUERY_FW = 0x004, + MLXSW_CMD_OPCODE_QUERY_BOARDINFO = 0x006, + MLXSW_CMD_OPCODE_QUERY_AQ_CAP = 0x003, + MLXSW_CMD_OPCODE_MAP_FA = 0xFFF, + MLXSW_CMD_OPCODE_UNMAP_FA = 0xFFE, + MLXSW_CMD_OPCODE_CONFIG_PROFILE = 0x100, + MLXSW_CMD_OPCODE_ACCESS_REG = 0x040, + MLXSW_CMD_OPCODE_SW2HW_DQ = 0x201, + MLXSW_CMD_OPCODE_HW2SW_DQ = 0x202, + MLXSW_CMD_OPCODE_2ERR_DQ = 0x01E, + MLXSW_CMD_OPCODE_QUERY_DQ = 0x022, + MLXSW_CMD_OPCODE_SW2HW_CQ = 0x016, + MLXSW_CMD_OPCODE_HW2SW_CQ = 0x017, + MLXSW_CMD_OPCODE_QUERY_CQ = 0x018, + MLXSW_CMD_OPCODE_SW2HW_EQ = 0x013, + MLXSW_CMD_OPCODE_HW2SW_EQ = 0x014, + MLXSW_CMD_OPCODE_QUERY_EQ = 0x015, + MLXSW_CMD_OPCODE_QUERY_RESOURCES = 0x101, +}; + +static inline const char *mlxsw_cmd_opcode_str(u16 opcode) +{ + switch (opcode) { + case MLXSW_CMD_OPCODE_QUERY_FW: + return "QUERY_FW"; + case MLXSW_CMD_OPCODE_QUERY_BOARDINFO: + return "QUERY_BOARDINFO"; + case MLXSW_CMD_OPCODE_QUERY_AQ_CAP: + return "QUERY_AQ_CAP"; + case MLXSW_CMD_OPCODE_MAP_FA: + return "MAP_FA"; + case MLXSW_CMD_OPCODE_UNMAP_FA: + return "UNMAP_FA"; + case MLXSW_CMD_OPCODE_CONFIG_PROFILE: + return "CONFIG_PROFILE"; + case MLXSW_CMD_OPCODE_ACCESS_REG: + return "ACCESS_REG"; + case MLXSW_CMD_OPCODE_SW2HW_DQ: + return "SW2HW_DQ"; + case MLXSW_CMD_OPCODE_HW2SW_DQ: + return "HW2SW_DQ"; + case MLXSW_CMD_OPCODE_2ERR_DQ: + return "2ERR_DQ"; + case MLXSW_CMD_OPCODE_QUERY_DQ: + return "QUERY_DQ"; + case MLXSW_CMD_OPCODE_SW2HW_CQ: + return "SW2HW_CQ"; + case MLXSW_CMD_OPCODE_HW2SW_CQ: + return "HW2SW_CQ"; + case MLXSW_CMD_OPCODE_QUERY_CQ: + return "QUERY_CQ"; + case MLXSW_CMD_OPCODE_SW2HW_EQ: + return "SW2HW_EQ"; + case MLXSW_CMD_OPCODE_HW2SW_EQ: + return "HW2SW_EQ"; + case MLXSW_CMD_OPCODE_QUERY_EQ: + return "QUERY_EQ"; + case MLXSW_CMD_OPCODE_QUERY_RESOURCES: + return "QUERY_RESOURCES"; + default: + return "*UNKNOWN*"; + } +} + +enum mlxsw_cmd_status { + /* Command execution succeeded. */ + MLXSW_CMD_STATUS_OK = 0x00, + /* Internal error (e.g. bus error) occurred while processing command. */ + MLXSW_CMD_STATUS_INTERNAL_ERR = 0x01, + /* Operation/command not supported or opcode modifier not supported. */ + MLXSW_CMD_STATUS_BAD_OP = 0x02, + /* Parameter not supported, parameter out of range. */ + MLXSW_CMD_STATUS_BAD_PARAM = 0x03, + /* System was not enabled or bad system state. */ + MLXSW_CMD_STATUS_BAD_SYS_STATE = 0x04, + /* Attempt to access reserved or unallocated resource, or resource in + * inappropriate ownership. + */ + MLXSW_CMD_STATUS_BAD_RESOURCE = 0x05, + /* Requested resource is currently executing a command. */ + MLXSW_CMD_STATUS_RESOURCE_BUSY = 0x06, + /* Required capability exceeds device limits. */ + MLXSW_CMD_STATUS_EXCEED_LIM = 0x08, + /* Resource is not in the appropriate state or ownership. */ + MLXSW_CMD_STATUS_BAD_RES_STATE = 0x09, + /* Index out of range (might be beyond table size or attempt to + * access a reserved resource). + */ + MLXSW_CMD_STATUS_BAD_INDEX = 0x0A, + /* NVMEM checksum/CRC failed. */ + MLXSW_CMD_STATUS_BAD_NVMEM = 0x0B, + /* Device is currently running reset */ + MLXSW_CMD_STATUS_RUNNING_RESET = 0x26, + /* Bad management packet (silently discarded). */ + MLXSW_CMD_STATUS_BAD_PKT = 0x30, +}; + +static inline const char *mlxsw_cmd_status_str(u8 status) +{ + switch (status) { + case MLXSW_CMD_STATUS_OK: + return "OK"; + case MLXSW_CMD_STATUS_INTERNAL_ERR: + return "INTERNAL_ERR"; + case MLXSW_CMD_STATUS_BAD_OP: + return "BAD_OP"; + case MLXSW_CMD_STATUS_BAD_PARAM: + return "BAD_PARAM"; + case MLXSW_CMD_STATUS_BAD_SYS_STATE: + return "BAD_SYS_STATE"; + case MLXSW_CMD_STATUS_BAD_RESOURCE: + return "BAD_RESOURCE"; + case MLXSW_CMD_STATUS_RESOURCE_BUSY: + return "RESOURCE_BUSY"; + case MLXSW_CMD_STATUS_EXCEED_LIM: + return "EXCEED_LIM"; + case MLXSW_CMD_STATUS_BAD_RES_STATE: + return "BAD_RES_STATE"; + case MLXSW_CMD_STATUS_BAD_INDEX: + return "BAD_INDEX"; + case MLXSW_CMD_STATUS_BAD_NVMEM: + return "BAD_NVMEM"; + case MLXSW_CMD_STATUS_RUNNING_RESET: + return "RUNNING_RESET"; + case MLXSW_CMD_STATUS_BAD_PKT: + return "BAD_PKT"; + default: + return "*UNKNOWN*"; + } +} + +/* QUERY_FW - Query Firmware + * ------------------------- + * OpMod == 0, INMmod == 0 + * ----------------------- + * The QUERY_FW command retrieves information related to firmware, command + * interface version and the amount of resources that should be allocated to + * the firmware. + */ + +static inline int mlxsw_cmd_query_fw(struct mlxsw_core *mlxsw_core, + char *out_mbox) +{ + return mlxsw_cmd_exec_out(mlxsw_core, MLXSW_CMD_OPCODE_QUERY_FW, + 0, 0, false, out_mbox, MLXSW_CMD_MBOX_SIZE); +} + +/* cmd_mbox_query_fw_fw_pages + * Amount of physical memory to be allocatedfor firmware usage in 4KB pages. + */ +MLXSW_ITEM32(cmd_mbox, query_fw, fw_pages, 0x00, 16, 16); + +/* cmd_mbox_query_fw_fw_rev_major + * Firmware Revision - Major + */ +MLXSW_ITEM32(cmd_mbox, query_fw, fw_rev_major, 0x00, 0, 16); + +/* cmd_mbox_query_fw_fw_rev_subminor + * Firmware Sub-minor version (Patch level) + */ +MLXSW_ITEM32(cmd_mbox, query_fw, fw_rev_subminor, 0x04, 16, 16); + +/* cmd_mbox_query_fw_fw_rev_minor + * Firmware Revision - Minor + */ +MLXSW_ITEM32(cmd_mbox, query_fw, fw_rev_minor, 0x04, 0, 16); + +/* cmd_mbox_query_fw_core_clk + * Internal Clock Frequency (in MHz) + */ +MLXSW_ITEM32(cmd_mbox, query_fw, core_clk, 0x08, 16, 16); + +/* cmd_mbox_query_fw_cmd_interface_rev + * Command Interface Interpreter Revision ID. This number is bumped up + * every time a non-backward-compatible change is done for the command + * interface. The current cmd_interface_rev is 1. + */ +MLXSW_ITEM32(cmd_mbox, query_fw, cmd_interface_rev, 0x08, 0, 16); + +/* cmd_mbox_query_fw_dt + * If set, Debug Trace is supported + */ +MLXSW_ITEM32(cmd_mbox, query_fw, dt, 0x0C, 31, 1); + +/* cmd_mbox_query_fw_api_version + * Indicates the version of the API, to enable software querying + * for compatibility. The current api_version is 1. + */ +MLXSW_ITEM32(cmd_mbox, query_fw, api_version, 0x0C, 0, 16); + +/* cmd_mbox_query_fw_fw_hour + * Firmware timestamp - hour + */ +MLXSW_ITEM32(cmd_mbox, query_fw, fw_hour, 0x10, 24, 8); + +/* cmd_mbox_query_fw_fw_minutes + * Firmware timestamp - minutes + */ +MLXSW_ITEM32(cmd_mbox, query_fw, fw_minutes, 0x10, 16, 8); + +/* cmd_mbox_query_fw_fw_seconds + * Firmware timestamp - seconds + */ +MLXSW_ITEM32(cmd_mbox, query_fw, fw_seconds, 0x10, 8, 8); + +/* cmd_mbox_query_fw_fw_year + * Firmware timestamp - year + */ +MLXSW_ITEM32(cmd_mbox, query_fw, fw_year, 0x14, 16, 16); + +/* cmd_mbox_query_fw_fw_month + * Firmware timestamp - month + */ +MLXSW_ITEM32(cmd_mbox, query_fw, fw_month, 0x14, 8, 8); + +/* cmd_mbox_query_fw_fw_day + * Firmware timestamp - day + */ +MLXSW_ITEM32(cmd_mbox, query_fw, fw_day, 0x14, 0, 8); + +/* cmd_mbox_query_fw_clr_int_base_offset + * Clear Interrupt register's offset from clr_int_bar register + * in PCI address space. + */ +MLXSW_ITEM64(cmd_mbox, query_fw, clr_int_base_offset, 0x20, 0, 64); + +/* cmd_mbox_query_fw_clr_int_bar + * PCI base address register (BAR) where clr_int register is located. + * 00 - BAR 0-1 (64 bit BAR) + */ +MLXSW_ITEM32(cmd_mbox, query_fw, clr_int_bar, 0x28, 30, 2); + +/* cmd_mbox_query_fw_error_buf_offset + * Read Only buffer for internal error reports of offset + * from error_buf_bar register in PCI address space). + */ +MLXSW_ITEM64(cmd_mbox, query_fw, error_buf_offset, 0x30, 0, 64); + +/* cmd_mbox_query_fw_error_buf_size + * Internal error buffer size in DWORDs + */ +MLXSW_ITEM32(cmd_mbox, query_fw, error_buf_size, 0x38, 0, 32); + +/* cmd_mbox_query_fw_error_int_bar + * PCI base address register (BAR) where error buffer + * register is located. + * 00 - BAR 0-1 (64 bit BAR) + */ +MLXSW_ITEM32(cmd_mbox, query_fw, error_int_bar, 0x3C, 30, 2); + +/* cmd_mbox_query_fw_doorbell_page_offset + * Offset of the doorbell page + */ +MLXSW_ITEM64(cmd_mbox, query_fw, doorbell_page_offset, 0x40, 0, 64); + +/* cmd_mbox_query_fw_doorbell_page_bar + * PCI base address register (BAR) of the doorbell page + * 00 - BAR 0-1 (64 bit BAR) + */ +MLXSW_ITEM32(cmd_mbox, query_fw, doorbell_page_bar, 0x48, 30, 2); + +/* cmd_mbox_query_fw_free_running_clock_offset + * The offset of the free running clock page + */ +MLXSW_ITEM64(cmd_mbox, query_fw, free_running_clock_offset, 0x50, 0, 64); + +/* cmd_mbox_query_fw_fr_rn_clk_bar + * PCI base address register (BAR) of the free running clock page + * 0: BAR 0 + * 1: 64 bit BAR + */ +MLXSW_ITEM32(cmd_mbox, query_fw, fr_rn_clk_bar, 0x58, 30, 2); + +/* cmd_mbox_query_fw_utc_sec_offset + * The offset of the UTC_Sec page + */ +MLXSW_ITEM64(cmd_mbox, query_fw, utc_sec_offset, 0x70, 0, 64); + +/* cmd_mbox_query_fw_utc_sec_bar + * PCI base address register (BAR) of the UTC_Sec page + * 0: BAR 0 + * 1: 64 bit BAR + * Reserved on SwitchX/-2, Switch-IB/2, Spectrum-1 + */ +MLXSW_ITEM32(cmd_mbox, query_fw, utc_sec_bar, 0x78, 30, 2); + +/* cmd_mbox_query_fw_utc_nsec_offset + * The offset of the UTC_nSec page + */ +MLXSW_ITEM64(cmd_mbox, query_fw, utc_nsec_offset, 0x80, 0, 64); + +/* cmd_mbox_query_fw_utc_nsec_bar + * PCI base address register (BAR) of the UTC_nSec page + * 0: BAR 0 + * 1: 64 bit BAR + * Reserved on SwitchX/-2, Switch-IB/2, Spectrum-1 + */ +MLXSW_ITEM32(cmd_mbox, query_fw, utc_nsec_bar, 0x88, 30, 2); + +/* QUERY_BOARDINFO - Query Board Information + * ----------------------------------------- + * OpMod == 0 (N/A), INMmod == 0 (N/A) + * ----------------------------------- + * The QUERY_BOARDINFO command retrieves adapter specific parameters. + */ + +static inline int mlxsw_cmd_boardinfo(struct mlxsw_core *mlxsw_core, + char *out_mbox) +{ + return mlxsw_cmd_exec_out(mlxsw_core, MLXSW_CMD_OPCODE_QUERY_BOARDINFO, + 0, 0, false, out_mbox, MLXSW_CMD_MBOX_SIZE); +} + +/* cmd_mbox_boardinfo_intapin + * When PCIe interrupt messages are being used, this value is used for clearing + * an interrupt. When using MSI-X, this register is not used. + */ +MLXSW_ITEM32(cmd_mbox, boardinfo, intapin, 0x10, 24, 8); + +/* cmd_mbox_boardinfo_vsd_vendor_id + * PCISIG Vendor ID (www.pcisig.com/membership/vid_search) of the vendor + * specifying/formatting the VSD. The vsd_vendor_id identifies the management + * domain of the VSD/PSID data. Different vendors may choose different VSD/PSID + * format and encoding as long as they use their assigned vsd_vendor_id. + */ +MLXSW_ITEM32(cmd_mbox, boardinfo, vsd_vendor_id, 0x1C, 0, 16); + +/* cmd_mbox_boardinfo_vsd + * Vendor Specific Data. The VSD string that is burnt to the Flash + * with the firmware. + */ +#define MLXSW_CMD_BOARDINFO_VSD_LEN 208 +MLXSW_ITEM_BUF(cmd_mbox, boardinfo, vsd, 0x20, MLXSW_CMD_BOARDINFO_VSD_LEN); + +/* cmd_mbox_boardinfo_psid + * The PSID field is a 16-ascii (byte) character string which acts as + * the board ID. The PSID format is used in conjunction with + * Mellanox vsd_vendor_id (15B3h). + */ +#define MLXSW_CMD_BOARDINFO_PSID_LEN 16 +MLXSW_ITEM_BUF(cmd_mbox, boardinfo, psid, 0xF0, MLXSW_CMD_BOARDINFO_PSID_LEN); + +/* QUERY_AQ_CAP - Query Asynchronous Queues Capabilities + * ----------------------------------------------------- + * OpMod == 0 (N/A), INMmod == 0 (N/A) + * ----------------------------------- + * The QUERY_AQ_CAP command returns the device asynchronous queues + * capabilities supported. + */ + +static inline int mlxsw_cmd_query_aq_cap(struct mlxsw_core *mlxsw_core, + char *out_mbox) +{ + return mlxsw_cmd_exec_out(mlxsw_core, MLXSW_CMD_OPCODE_QUERY_AQ_CAP, + 0, 0, false, out_mbox, MLXSW_CMD_MBOX_SIZE); +} + +/* cmd_mbox_query_aq_cap_log_max_sdq_sz + * Log (base 2) of max WQEs allowed on SDQ. + */ +MLXSW_ITEM32(cmd_mbox, query_aq_cap, log_max_sdq_sz, 0x00, 24, 8); + +/* cmd_mbox_query_aq_cap_max_num_sdqs + * Maximum number of SDQs. + */ +MLXSW_ITEM32(cmd_mbox, query_aq_cap, max_num_sdqs, 0x00, 0, 8); + +/* cmd_mbox_query_aq_cap_log_max_rdq_sz + * Log (base 2) of max WQEs allowed on RDQ. + */ +MLXSW_ITEM32(cmd_mbox, query_aq_cap, log_max_rdq_sz, 0x04, 24, 8); + +/* cmd_mbox_query_aq_cap_max_num_rdqs + * Maximum number of RDQs. + */ +MLXSW_ITEM32(cmd_mbox, query_aq_cap, max_num_rdqs, 0x04, 0, 8); + +/* cmd_mbox_query_aq_cap_log_max_cq_sz + * Log (base 2) of the Maximum CQEs allowed in a CQ for CQEv0 and CQEv1. + */ +MLXSW_ITEM32(cmd_mbox, query_aq_cap, log_max_cq_sz, 0x08, 24, 8); + +/* cmd_mbox_query_aq_cap_log_max_cqv2_sz + * Log (base 2) of the Maximum CQEs allowed in a CQ for CQEv2. + */ +MLXSW_ITEM32(cmd_mbox, query_aq_cap, log_max_cqv2_sz, 0x08, 16, 8); + +/* cmd_mbox_query_aq_cap_max_num_cqs + * Maximum number of CQs. + */ +MLXSW_ITEM32(cmd_mbox, query_aq_cap, max_num_cqs, 0x08, 0, 8); + +/* cmd_mbox_query_aq_cap_log_max_eq_sz + * Log (base 2) of max EQEs allowed on EQ. + */ +MLXSW_ITEM32(cmd_mbox, query_aq_cap, log_max_eq_sz, 0x0C, 24, 8); + +/* cmd_mbox_query_aq_cap_max_num_eqs + * Maximum number of EQs. + */ +MLXSW_ITEM32(cmd_mbox, query_aq_cap, max_num_eqs, 0x0C, 0, 8); + +/* cmd_mbox_query_aq_cap_max_sg_sq + * The maximum S/G list elements in an DSQ. DSQ must not contain + * more S/G entries than indicated here. + */ +MLXSW_ITEM32(cmd_mbox, query_aq_cap, max_sg_sq, 0x10, 8, 8); + +/* cmd_mbox_query_aq_cap_ + * The maximum S/G list elements in an DRQ. DRQ must not contain + * more S/G entries than indicated here. + */ +MLXSW_ITEM32(cmd_mbox, query_aq_cap, max_sg_rq, 0x10, 0, 8); + +/* MAP_FA - Map Firmware Area + * -------------------------- + * OpMod == 0 (N/A), INMmod == Number of VPM entries + * ------------------------------------------------- + * The MAP_FA command passes physical pages to the switch. These pages + * are used to store the device firmware. MAP_FA can be executed multiple + * times until all the firmware area is mapped (the size that should be + * mapped is retrieved through the QUERY_FW command). All required pages + * must be mapped to finish the initialization phase. Physical memory + * passed in this command must be pinned. + */ + +#define MLXSW_CMD_MAP_FA_VPM_ENTRIES_MAX 32 + +static inline int mlxsw_cmd_map_fa(struct mlxsw_core *mlxsw_core, + char *in_mbox, u32 vpm_entries_count) +{ + return mlxsw_cmd_exec_in(mlxsw_core, MLXSW_CMD_OPCODE_MAP_FA, + 0, vpm_entries_count, + in_mbox, MLXSW_CMD_MBOX_SIZE); +} + +/* cmd_mbox_map_fa_pa + * Physical Address. + */ +MLXSW_ITEM64_INDEXED(cmd_mbox, map_fa, pa, 0x00, 12, 52, 0x08, 0x00, true); + +/* cmd_mbox_map_fa_log2size + * Log (base 2) of the size in 4KB pages of the physical and contiguous memory + * that starts at PA_L/H. + */ +MLXSW_ITEM32_INDEXED(cmd_mbox, map_fa, log2size, 0x00, 0, 5, 0x08, 0x04, false); + +/* UNMAP_FA - Unmap Firmware Area + * ------------------------------ + * OpMod == 0 (N/A), INMmod == 0 (N/A) + * ----------------------------------- + * The UNMAP_FA command unload the firmware and unmaps all the + * firmware area. After this command is completed the device will not access + * the pages that were mapped to the firmware area. After executing UNMAP_FA + * command, software reset must be done prior to execution of MAP_FW command. + */ + +static inline int mlxsw_cmd_unmap_fa(struct mlxsw_core *mlxsw_core) +{ + return mlxsw_cmd_exec_none(mlxsw_core, MLXSW_CMD_OPCODE_UNMAP_FA, 0, 0); +} + +/* QUERY_RESOURCES - Query chip resources + * -------------------------------------- + * OpMod == 0 (N/A) , INMmod is index + * ---------------------------------- + * The QUERY_RESOURCES command retrieves information related to chip resources + * by resource ID. Every command returns 32 entries. INmod is being use as base. + * for example, index 1 will return entries 32-63. When the tables end and there + * are no more sources in the table, will return resource id 0xFFF to indicate + * it. + */ + +#define MLXSW_CMD_QUERY_RESOURCES_TABLE_END_ID 0xffff +#define MLXSW_CMD_QUERY_RESOURCES_MAX_QUERIES 100 +#define MLXSW_CMD_QUERY_RESOURCES_PER_QUERY 32 + +static inline int mlxsw_cmd_query_resources(struct mlxsw_core *mlxsw_core, + char *out_mbox, int index) +{ + return mlxsw_cmd_exec_out(mlxsw_core, MLXSW_CMD_OPCODE_QUERY_RESOURCES, + 0, index, false, out_mbox, + MLXSW_CMD_MBOX_SIZE); +} + +/* cmd_mbox_query_resource_id + * The resource id. 0xFFFF indicates table's end. + */ +MLXSW_ITEM32_INDEXED(cmd_mbox, query_resource, id, 0x00, 16, 16, 0x8, 0, false); + +/* cmd_mbox_query_resource_data + * The resource + */ +MLXSW_ITEM64_INDEXED(cmd_mbox, query_resource, data, + 0x00, 0, 40, 0x8, 0, false); + +/* CONFIG_PROFILE (Set) - Configure Switch Profile + * ------------------------------ + * OpMod == 1 (Set), INMmod == 0 (N/A) + * ----------------------------------- + * The CONFIG_PROFILE command sets the switch profile. The command can be + * executed on the device only once at startup in order to allocate and + * configure all switch resources and prepare it for operational mode. + * It is not possible to change the device profile after the chip is + * in operational mode. + * Failure of the CONFIG_PROFILE command leaves the hardware in an indeterminate + * state therefore it is required to perform software reset to the device + * following an unsuccessful completion of the command. It is required + * to perform software reset to the device to change an existing profile. + */ + +static inline int mlxsw_cmd_config_profile_set(struct mlxsw_core *mlxsw_core, + char *in_mbox) +{ + return mlxsw_cmd_exec_in(mlxsw_core, MLXSW_CMD_OPCODE_CONFIG_PROFILE, + 1, 0, in_mbox, MLXSW_CMD_MBOX_SIZE); +} + +/* cmd_mbox_config_profile_set_max_vepa_channels + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_max_vepa_channels, 0x0C, 0, 1); + +/* cmd_mbox_config_profile_set_max_lag + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_max_lag, 0x0C, 1, 1); + +/* cmd_mbox_config_profile_set_max_port_per_lag + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_max_port_per_lag, 0x0C, 2, 1); + +/* cmd_mbox_config_profile_set_max_mid + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_max_mid, 0x0C, 3, 1); + +/* cmd_mbox_config_profile_set_max_pgt + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_max_pgt, 0x0C, 4, 1); + +/* cmd_mbox_config_profile_set_max_system_port + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_max_system_port, 0x0C, 5, 1); + +/* cmd_mbox_config_profile_set_max_vlan_groups + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_max_vlan_groups, 0x0C, 6, 1); + +/* cmd_mbox_config_profile_set_max_regions + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_max_regions, 0x0C, 7, 1); + +/* cmd_mbox_config_profile_set_flood_mode + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_flood_mode, 0x0C, 8, 1); + +/* cmd_mbox_config_profile_set_max_flood_tables + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_flood_tables, 0x0C, 9, 1); + +/* cmd_mbox_config_profile_set_max_ib_mc + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_max_ib_mc, 0x0C, 12, 1); + +/* cmd_mbox_config_profile_set_max_pkey + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_max_pkey, 0x0C, 13, 1); + +/* cmd_mbox_config_profile_set_adaptive_routing_group_cap + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, + set_adaptive_routing_group_cap, 0x0C, 14, 1); + +/* cmd_mbox_config_profile_set_ar_sec + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_ar_sec, 0x0C, 15, 1); + +/* cmd_mbox_config_set_ubridge + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_ubridge, 0x0C, 22, 1); + +/* cmd_mbox_config_set_kvd_linear_size + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_kvd_linear_size, 0x0C, 24, 1); + +/* cmd_mbox_config_set_kvd_hash_single_size + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_kvd_hash_single_size, 0x0C, 25, 1); + +/* cmd_mbox_config_set_kvd_hash_double_size + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_kvd_hash_double_size, 0x0C, 26, 1); + +/* cmd_mbox_config_set_cqe_version + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_cqe_version, 0x08, 0, 1); + +/* cmd_mbox_config_set_cqe_time_stamp_type + * Capability bit. Setting a bit to 1 configures the profile + * according to the mailbox contents. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, set_cqe_time_stamp_type, 0x08, 2, 1); + +/* cmd_mbox_config_profile_max_vepa_channels + * Maximum number of VEPA channels per port (0 through 16) + * 0 - multi-channel VEPA is disabled + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_vepa_channels, 0x10, 0, 8); + +/* cmd_mbox_config_profile_max_lag + * Maximum number of LAG IDs requested. + * Reserved when Spectrum-1/2/3, supported from Spectrum-4 and above. + * For Spectrum-4, firmware sets 128 for values between 1-128 and 256 for values + * between 129-256. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_lag, 0x14, 0, 16); + +/* cmd_mbox_config_profile_max_port_per_lag + * Maximum number of ports per LAG requested. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_port_per_lag, 0x18, 0, 16); + +/* cmd_mbox_config_profile_max_mid + * Maximum Multicast IDs. + * Multicast IDs are allocated from 0 to max_mid-1 + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_mid, 0x1C, 0, 16); + +/* cmd_mbox_config_profile_max_pgt + * Maximum records in the Port Group Table per Switch Partition. + * Port Group Table indexes are from 0 to max_pgt-1 + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_pgt, 0x20, 0, 16); + +/* cmd_mbox_config_profile_max_system_port + * The maximum number of system ports that can be allocated. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_system_port, 0x24, 0, 16); + +/* cmd_mbox_config_profile_max_vlan_groups + * Maximum number VLAN Groups for VLAN binding. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_vlan_groups, 0x28, 0, 12); + +/* cmd_mbox_config_profile_max_regions + * Maximum number of TCAM Regions. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_regions, 0x2C, 0, 16); + +/* cmd_mbox_config_profile_max_flood_tables + * Maximum number of single-entry flooding tables. Different flooding tables + * can be associated with different packet types. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_flood_tables, 0x30, 16, 4); + +/* cmd_mbox_config_profile_max_vid_flood_tables + * Maximum number of per-vid flooding tables. Flooding tables are associated + * to the different packet types for the different switch partitions. + * Table size is 4K entries covering all VID space. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_vid_flood_tables, 0x30, 8, 4); + +enum mlxsw_cmd_mbox_config_profile_flood_mode { + /* Mixed mode, where: + * max_flood_tables indicates the number of single-entry tables. + * max_vid_flood_tables indicates the number of per-VID tables. + * max_fid_offset_flood_tables indicates the number of FID-offset + * tables. max_fid_flood_tables indicates the number of per-FID tables. + * Reserved when unified bridge model is used. + */ + MLXSW_CMD_MBOX_CONFIG_PROFILE_FLOOD_MODE_MIXED = 3, + /* Controlled flood tables. Reserved when legacy bridge model is + * used. + */ + MLXSW_CMD_MBOX_CONFIG_PROFILE_FLOOD_MODE_CONTROLLED = 4, +}; + +/* cmd_mbox_config_profile_flood_mode + * Flooding mode to use. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, flood_mode, 0x30, 0, 3); + +/* cmd_mbox_config_profile_max_fid_offset_flood_tables + * Maximum number of FID-offset flooding tables. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, + max_fid_offset_flood_tables, 0x34, 24, 4); + +/* cmd_mbox_config_profile_fid_offset_flood_table_size + * The size (number of entries) of each FID-offset flood table. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, + fid_offset_flood_table_size, 0x34, 0, 16); + +/* cmd_mbox_config_profile_max_fid_flood_tables + * Maximum number of per-FID flooding tables. + * + * Note: This flooding tables cover special FIDs only (vFIDs), starting at + * FID value 4K and higher. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_fid_flood_tables, 0x38, 24, 4); + +/* cmd_mbox_config_profile_fid_flood_table_size + * The size (number of entries) of each per-FID table. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, fid_flood_table_size, 0x38, 0, 16); + +/* cmd_mbox_config_profile_max_ib_mc + * Maximum number of multicast FDB records for InfiniBand + * FDB (in 512 chunks) per InfiniBand switch partition. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_ib_mc, 0x40, 0, 15); + +/* cmd_mbox_config_profile_max_pkey + * Maximum per port PKEY table size (for PKEY enforcement) + */ +MLXSW_ITEM32(cmd_mbox, config_profile, max_pkey, 0x44, 0, 15); + +/* cmd_mbox_config_profile_ar_sec + * Primary/secondary capability + * Describes the number of adaptive routing sub-groups + * 0 - disable primary/secondary (single group) + * 1 - enable primary/secondary (2 sub-groups) + * 2 - 3 sub-groups: Not supported in SwitchX, SwitchX-2 + * 3 - 4 sub-groups: Not supported in SwitchX, SwitchX-2 + */ +MLXSW_ITEM32(cmd_mbox, config_profile, ar_sec, 0x4C, 24, 2); + +/* cmd_mbox_config_profile_adaptive_routing_group_cap + * Adaptive Routing Group Capability. Indicates the number of AR groups + * supported. Note that when Primary/secondary is enabled, each + * primary/secondary couple consumes 2 adaptive routing entries. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, adaptive_routing_group_cap, 0x4C, 0, 16); + +/* cmd_mbox_config_profile_arn + * Adaptive Routing Notification Enable + * Not supported in SwitchX, SwitchX-2 + */ +MLXSW_ITEM32(cmd_mbox, config_profile, arn, 0x50, 31, 1); + +/* cmd_mbox_config_profile_ubridge + * Unified Bridge + * 0 - non unified bridge + * 1 - unified bridge + */ +MLXSW_ITEM32(cmd_mbox, config_profile, ubridge, 0x50, 4, 1); + +/* cmd_mbox_config_kvd_linear_size + * KVD Linear Size + * Valid for Spectrum only + * Allowed values are 128*N where N=0 or higher + */ +MLXSW_ITEM32(cmd_mbox, config_profile, kvd_linear_size, 0x54, 0, 24); + +/* cmd_mbox_config_kvd_hash_single_size + * KVD Hash single-entries size + * Valid for Spectrum only + * Allowed values are 128*N where N=0 or higher + * Must be greater or equal to cap_min_kvd_hash_single_size + * Must be smaller or equal to cap_kvd_size - kvd_linear_size + */ +MLXSW_ITEM32(cmd_mbox, config_profile, kvd_hash_single_size, 0x58, 0, 24); + +/* cmd_mbox_config_kvd_hash_double_size + * KVD Hash double-entries size (units of single-size entries) + * Valid for Spectrum only + * Allowed values are 128*N where N=0 or higher + * Must be either 0 or greater or equal to cap_min_kvd_hash_double_size + * Must be smaller or equal to cap_kvd_size - kvd_linear_size + */ +MLXSW_ITEM32(cmd_mbox, config_profile, kvd_hash_double_size, 0x5C, 0, 24); + +/* cmd_mbox_config_profile_swid_config_mask + * Modify Switch Partition Configuration mask. When set, the configu- + * ration value for the Switch Partition are taken from the mailbox. + * When clear, the current configuration values are used. + * Bit 0 - set type + * Bit 1 - properties + * Other - reserved + */ +MLXSW_ITEM32_INDEXED(cmd_mbox, config_profile, swid_config_mask, + 0x60, 24, 8, 0x08, 0x00, false); + +/* cmd_mbox_config_profile_swid_config_type + * Switch Partition type. + * 0000 - disabled (Switch Partition does not exist) + * 0001 - InfiniBand + * 0010 - Ethernet + * 1000 - router port (SwitchX-2 only) + * Other - reserved + */ +MLXSW_ITEM32_INDEXED(cmd_mbox, config_profile, swid_config_type, + 0x60, 20, 4, 0x08, 0x00, false); + +/* cmd_mbox_config_profile_swid_config_properties + * Switch Partition properties. + */ +MLXSW_ITEM32_INDEXED(cmd_mbox, config_profile, swid_config_properties, + 0x60, 0, 8, 0x08, 0x00, false); + +enum mlxsw_cmd_mbox_config_profile_cqe_time_stamp_type { + /* uSec - 1.024uSec (default). Only bits 15:0 are valid. */ + MLXSW_CMD_MBOX_CONFIG_PROFILE_CQE_TIME_STAMP_TYPE_USEC, + /* FRC - Free Running Clock, units of 1nSec. + * Reserved when SwitchX/-2, Switch-IB/2 and Spectrum-1. + */ + MLXSW_CMD_MBOX_CONFIG_PROFILE_CQE_TIME_STAMP_TYPE_FRC, + /* UTC. time_stamp[37:30] = Sec, time_stamp[29:0] = nSec. + * Reserved when SwitchX/2, Switch-IB/2 and Spectrum-1. + */ + MLXSW_CMD_MBOX_CONFIG_PROFILE_CQE_TIME_STAMP_TYPE_UTC, +}; + +/* cmd_mbox_config_profile_cqe_time_stamp_type + * CQE time_stamp_type for non-mirror-packets. + * Configured if set_cqe_time_stamp_type is set. + * Reserved when SwitchX/-2, Switch-IB/2 and Spectrum-1. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, cqe_time_stamp_type, 0xB0, 8, 2); + +/* cmd_mbox_config_profile_cqe_version + * CQE version: + * 0: CQE version is 0 + * 1: CQE version is either 1 or 2 + * CQE ver 1 or 2 is configured by Completion Queue Context field cqe_ver. + */ +MLXSW_ITEM32(cmd_mbox, config_profile, cqe_version, 0xB0, 0, 8); + +/* ACCESS_REG - Access EMAD Supported Register + * ---------------------------------- + * OpMod == 0 (N/A), INMmod == 0 (N/A) + * ------------------------------------- + * The ACCESS_REG command supports accessing device registers. This access + * is mainly used for bootstrapping. + */ + +static inline int mlxsw_cmd_access_reg(struct mlxsw_core *mlxsw_core, + bool reset_ok, + char *in_mbox, char *out_mbox) +{ + return mlxsw_cmd_exec(mlxsw_core, MLXSW_CMD_OPCODE_ACCESS_REG, + 0, 0, false, reset_ok, + in_mbox, MLXSW_CMD_MBOX_SIZE, + out_mbox, MLXSW_CMD_MBOX_SIZE); +} + +/* SW2HW_DQ - Software to Hardware DQ + * ---------------------------------- + * OpMod == 0 (send DQ) / OpMod == 1 (receive DQ) + * INMmod == DQ number + * ---------------------------------------------- + * The SW2HW_DQ command transitions a descriptor queue from software to + * hardware ownership. The command enables posting WQEs and ringing DoorBells + * on the descriptor queue. + */ + +static inline int __mlxsw_cmd_sw2hw_dq(struct mlxsw_core *mlxsw_core, + char *in_mbox, u32 dq_number, + u8 opcode_mod) +{ + return mlxsw_cmd_exec_in(mlxsw_core, MLXSW_CMD_OPCODE_SW2HW_DQ, + opcode_mod, dq_number, + in_mbox, MLXSW_CMD_MBOX_SIZE); +} + +enum { + MLXSW_CMD_OPCODE_MOD_SDQ = 0, + MLXSW_CMD_OPCODE_MOD_RDQ = 1, +}; + +static inline int mlxsw_cmd_sw2hw_sdq(struct mlxsw_core *mlxsw_core, + char *in_mbox, u32 dq_number) +{ + return __mlxsw_cmd_sw2hw_dq(mlxsw_core, in_mbox, dq_number, + MLXSW_CMD_OPCODE_MOD_SDQ); +} + +static inline int mlxsw_cmd_sw2hw_rdq(struct mlxsw_core *mlxsw_core, + char *in_mbox, u32 dq_number) +{ + return __mlxsw_cmd_sw2hw_dq(mlxsw_core, in_mbox, dq_number, + MLXSW_CMD_OPCODE_MOD_RDQ); +} + +/* cmd_mbox_sw2hw_dq_cq + * Number of the CQ that this Descriptor Queue reports completions to. + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_dq, cq, 0x00, 24, 8); + +enum mlxsw_cmd_mbox_sw2hw_dq_sdq_lp { + MLXSW_CMD_MBOX_SW2HW_DQ_SDQ_LP_WQE, + MLXSW_CMD_MBOX_SW2HW_DQ_SDQ_LP_IGNORE_WQE, +}; + +/* cmd_mbox_sw2hw_dq_sdq_lp + * SDQ local Processing + * 0: local processing by wqe.lp + * 1: local processing (ignoring wqe.lp) + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_dq, sdq_lp, 0x00, 23, 1); + +/* cmd_mbox_sw2hw_dq_sdq_tclass + * SDQ: CPU Egress TClass + * RDQ: Reserved + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_dq, sdq_tclass, 0x00, 16, 6); + +/* cmd_mbox_sw2hw_dq_log2_dq_sz + * Log (base 2) of the Descriptor Queue size in 4KB pages. + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_dq, log2_dq_sz, 0x00, 0, 6); + +/* cmd_mbox_sw2hw_dq_pa + * Physical Address. + */ +MLXSW_ITEM64_INDEXED(cmd_mbox, sw2hw_dq, pa, 0x10, 12, 52, 0x08, 0x00, true); + +/* HW2SW_DQ - Hardware to Software DQ + * ---------------------------------- + * OpMod == 0 (send DQ) / OpMod == 1 (receive DQ) + * INMmod == DQ number + * ---------------------------------------------- + * The HW2SW_DQ command transitions a descriptor queue from hardware to + * software ownership. Incoming packets on the DQ are silently discarded, + * SW should not post descriptors on nonoperational DQs. + */ + +static inline int __mlxsw_cmd_hw2sw_dq(struct mlxsw_core *mlxsw_core, + u32 dq_number, u8 opcode_mod) +{ + return mlxsw_cmd_exec_none(mlxsw_core, MLXSW_CMD_OPCODE_HW2SW_DQ, + opcode_mod, dq_number); +} + +static inline int mlxsw_cmd_hw2sw_sdq(struct mlxsw_core *mlxsw_core, + u32 dq_number) +{ + return __mlxsw_cmd_hw2sw_dq(mlxsw_core, dq_number, + MLXSW_CMD_OPCODE_MOD_SDQ); +} + +static inline int mlxsw_cmd_hw2sw_rdq(struct mlxsw_core *mlxsw_core, + u32 dq_number) +{ + return __mlxsw_cmd_hw2sw_dq(mlxsw_core, dq_number, + MLXSW_CMD_OPCODE_MOD_RDQ); +} + +/* 2ERR_DQ - To Error DQ + * --------------------- + * OpMod == 0 (send DQ) / OpMod == 1 (receive DQ) + * INMmod == DQ number + * ---------------------------------------------- + * The 2ERR_DQ command transitions the DQ into the error state from the state + * in which it has been. While the command is executed, some in-process + * descriptors may complete. Once the DQ transitions into the error state, + * if there are posted descriptors on the RDQ/SDQ, the hardware writes + * a completion with error (flushed) for all descriptors posted in the RDQ/SDQ. + * When the command is completed successfully, the DQ is already in + * the error state. + */ + +static inline int __mlxsw_cmd_2err_dq(struct mlxsw_core *mlxsw_core, + u32 dq_number, u8 opcode_mod) +{ + return mlxsw_cmd_exec_none(mlxsw_core, MLXSW_CMD_OPCODE_2ERR_DQ, + opcode_mod, dq_number); +} + +static inline int mlxsw_cmd_2err_sdq(struct mlxsw_core *mlxsw_core, + u32 dq_number) +{ + return __mlxsw_cmd_2err_dq(mlxsw_core, dq_number, + MLXSW_CMD_OPCODE_MOD_SDQ); +} + +static inline int mlxsw_cmd_2err_rdq(struct mlxsw_core *mlxsw_core, + u32 dq_number) +{ + return __mlxsw_cmd_2err_dq(mlxsw_core, dq_number, + MLXSW_CMD_OPCODE_MOD_RDQ); +} + +/* QUERY_DQ - Query DQ + * --------------------- + * OpMod == 0 (send DQ) / OpMod == 1 (receive DQ) + * INMmod == DQ number + * ---------------------------------------------- + * The QUERY_DQ command retrieves a snapshot of DQ parameters from the hardware. + * + * Note: Output mailbox has the same format as SW2HW_DQ. + */ + +static inline int __mlxsw_cmd_query_dq(struct mlxsw_core *mlxsw_core, + char *out_mbox, u32 dq_number, + u8 opcode_mod) +{ + return mlxsw_cmd_exec_out(mlxsw_core, MLXSW_CMD_OPCODE_2ERR_DQ, + opcode_mod, dq_number, false, + out_mbox, MLXSW_CMD_MBOX_SIZE); +} + +static inline int mlxsw_cmd_query_sdq(struct mlxsw_core *mlxsw_core, + char *out_mbox, u32 dq_number) +{ + return __mlxsw_cmd_query_dq(mlxsw_core, out_mbox, dq_number, + MLXSW_CMD_OPCODE_MOD_SDQ); +} + +static inline int mlxsw_cmd_query_rdq(struct mlxsw_core *mlxsw_core, + char *out_mbox, u32 dq_number) +{ + return __mlxsw_cmd_query_dq(mlxsw_core, out_mbox, dq_number, + MLXSW_CMD_OPCODE_MOD_RDQ); +} + +/* SW2HW_CQ - Software to Hardware CQ + * ---------------------------------- + * OpMod == 0 (N/A), INMmod == CQ number + * ------------------------------------- + * The SW2HW_CQ command transfers ownership of a CQ context entry from software + * to hardware. The command takes the CQ context entry from the input mailbox + * and stores it in the CQC in the ownership of the hardware. The command fails + * if the requested CQC entry is already in the ownership of the hardware. + */ + +static inline int mlxsw_cmd_sw2hw_cq(struct mlxsw_core *mlxsw_core, + char *in_mbox, u32 cq_number) +{ + return mlxsw_cmd_exec_in(mlxsw_core, MLXSW_CMD_OPCODE_SW2HW_CQ, + 0, cq_number, in_mbox, MLXSW_CMD_MBOX_SIZE); +} + +enum mlxsw_cmd_mbox_sw2hw_cq_cqe_ver { + MLXSW_CMD_MBOX_SW2HW_CQ_CQE_VER_1, + MLXSW_CMD_MBOX_SW2HW_CQ_CQE_VER_2, +}; + +/* cmd_mbox_sw2hw_cq_cqe_ver + * CQE Version. + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_cq, cqe_ver, 0x00, 28, 4); + +/* cmd_mbox_sw2hw_cq_c_eqn + * Event Queue this CQ reports completion events to. + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_cq, c_eqn, 0x00, 24, 1); + +/* cmd_mbox_sw2hw_cq_st + * Event delivery state machine + * 0x0 - FIRED + * 0x1 - ARMED (Request for Notification) + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_cq, st, 0x00, 8, 1); + +/* cmd_mbox_sw2hw_cq_log_cq_size + * Log (base 2) of the CQ size (in entries). + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_cq, log_cq_size, 0x00, 0, 4); + +/* cmd_mbox_sw2hw_cq_producer_counter + * Producer Counter. The counter is incremented for each CQE that is + * written by the HW to the CQ. + * Maintained by HW (valid for the QUERY_CQ command only) + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_cq, producer_counter, 0x04, 0, 16); + +/* cmd_mbox_sw2hw_cq_pa + * Physical Address. + */ +MLXSW_ITEM64_INDEXED(cmd_mbox, sw2hw_cq, pa, 0x10, 11, 53, 0x08, 0x00, true); + +/* HW2SW_CQ - Hardware to Software CQ + * ---------------------------------- + * OpMod == 0 (N/A), INMmod == CQ number + * ------------------------------------- + * The HW2SW_CQ command transfers ownership of a CQ context entry from hardware + * to software. The CQC entry is invalidated as a result of this command. + */ + +static inline int mlxsw_cmd_hw2sw_cq(struct mlxsw_core *mlxsw_core, + u32 cq_number) +{ + return mlxsw_cmd_exec_none(mlxsw_core, MLXSW_CMD_OPCODE_HW2SW_CQ, + 0, cq_number); +} + +/* QUERY_CQ - Query CQ + * ---------------------------------- + * OpMod == 0 (N/A), INMmod == CQ number + * ------------------------------------- + * The QUERY_CQ command retrieves a snapshot of the current CQ context entry. + * The command stores the snapshot in the output mailbox in the software format. + * Note that the CQ context state and values are not affected by the QUERY_CQ + * command. The QUERY_CQ command is for debug purposes only. + * + * Note: Output mailbox has the same format as SW2HW_CQ. + */ + +static inline int mlxsw_cmd_query_cq(struct mlxsw_core *mlxsw_core, + char *out_mbox, u32 cq_number) +{ + return mlxsw_cmd_exec_out(mlxsw_core, MLXSW_CMD_OPCODE_QUERY_CQ, + 0, cq_number, false, + out_mbox, MLXSW_CMD_MBOX_SIZE); +} + +/* SW2HW_EQ - Software to Hardware EQ + * ---------------------------------- + * OpMod == 0 (N/A), INMmod == EQ number + * ------------------------------------- + * The SW2HW_EQ command transfers ownership of an EQ context entry from software + * to hardware. The command takes the EQ context entry from the input mailbox + * and stores it in the EQC in the ownership of the hardware. The command fails + * if the requested EQC entry is already in the ownership of the hardware. + */ + +static inline int mlxsw_cmd_sw2hw_eq(struct mlxsw_core *mlxsw_core, + char *in_mbox, u32 eq_number) +{ + return mlxsw_cmd_exec_in(mlxsw_core, MLXSW_CMD_OPCODE_SW2HW_EQ, + 0, eq_number, in_mbox, MLXSW_CMD_MBOX_SIZE); +} + +/* cmd_mbox_sw2hw_eq_int_msix + * When set, MSI-X cycles will be generated by this EQ. + * When cleared, an interrupt will be generated by this EQ. + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_eq, int_msix, 0x00, 24, 1); + +/* cmd_mbox_sw2hw_eq_st + * Event delivery state machine + * 0x0 - FIRED + * 0x1 - ARMED (Request for Notification) + * 0x11 - Always ARMED + * other - reserved + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_eq, st, 0x00, 8, 2); + +/* cmd_mbox_sw2hw_eq_log_eq_size + * Log (base 2) of the EQ size (in entries). + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_eq, log_eq_size, 0x00, 0, 4); + +/* cmd_mbox_sw2hw_eq_producer_counter + * Producer Counter. The counter is incremented for each EQE that is written + * by the HW to the EQ. + * Maintained by HW (valid for the QUERY_EQ command only) + */ +MLXSW_ITEM32(cmd_mbox, sw2hw_eq, producer_counter, 0x04, 0, 16); + +/* cmd_mbox_sw2hw_eq_pa + * Physical Address. + */ +MLXSW_ITEM64_INDEXED(cmd_mbox, sw2hw_eq, pa, 0x10, 11, 53, 0x08, 0x00, true); + +/* HW2SW_EQ - Hardware to Software EQ + * ---------------------------------- + * OpMod == 0 (N/A), INMmod == EQ number + * ------------------------------------- + */ + +static inline int mlxsw_cmd_hw2sw_eq(struct mlxsw_core *mlxsw_core, + u32 eq_number) +{ + return mlxsw_cmd_exec_none(mlxsw_core, MLXSW_CMD_OPCODE_HW2SW_EQ, + 0, eq_number); +} + +/* QUERY_EQ - Query EQ + * ---------------------------------- + * OpMod == 0 (N/A), INMmod == EQ number + * ------------------------------------- + * + * Note: Output mailbox has the same format as SW2HW_EQ. + */ + +static inline int mlxsw_cmd_query_eq(struct mlxsw_core *mlxsw_core, + char *out_mbox, u32 eq_number) +{ + return mlxsw_cmd_exec_out(mlxsw_core, MLXSW_CMD_OPCODE_QUERY_EQ, + 0, eq_number, false, + out_mbox, MLXSW_CMD_MBOX_SIZE); +} + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/core.c b/drivers/net/ethernet/mellanox/mlxsw/core.c new file mode 100644 index 000000000..e2a985ec2 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/core.c @@ -0,0 +1,3443 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/device.h> +#include <linux/export.h> +#include <linux/err.h> +#include <linux/if_link.h> +#include <linux/netdevice.h> +#include <linux/completion.h> +#include <linux/skbuff.h> +#include <linux/etherdevice.h> +#include <linux/types.h> +#include <linux/string.h> +#include <linux/gfp.h> +#include <linux/random.h> +#include <linux/jiffies.h> +#include <linux/mutex.h> +#include <linux/rcupdate.h> +#include <linux/slab.h> +#include <linux/workqueue.h> +#include <linux/firmware.h> +#include <asm/byteorder.h> +#include <net/devlink.h> +#include <trace/events/devlink.h> + +#include "core.h" +#include "core_env.h" +#include "item.h" +#include "cmd.h" +#include "port.h" +#include "trap.h" +#include "emad.h" +#include "reg.h" +#include "resources.h" +#include "../mlxfw/mlxfw.h" + +static LIST_HEAD(mlxsw_core_driver_list); +static DEFINE_SPINLOCK(mlxsw_core_driver_list_lock); + +static const char mlxsw_core_driver_name[] = "mlxsw_core"; + +static struct workqueue_struct *mlxsw_wq; +static struct workqueue_struct *mlxsw_owq; + +struct mlxsw_core_port { + struct devlink_port devlink_port; + void *port_driver_priv; + u16 local_port; + struct mlxsw_linecard *linecard; +}; + +void *mlxsw_core_port_driver_priv(struct mlxsw_core_port *mlxsw_core_port) +{ + return mlxsw_core_port->port_driver_priv; +} +EXPORT_SYMBOL(mlxsw_core_port_driver_priv); + +static bool mlxsw_core_port_check(struct mlxsw_core_port *mlxsw_core_port) +{ + return mlxsw_core_port->port_driver_priv != NULL; +} + +struct mlxsw_core { + struct mlxsw_driver *driver; + const struct mlxsw_bus *bus; + void *bus_priv; + const struct mlxsw_bus_info *bus_info; + struct workqueue_struct *emad_wq; + struct list_head rx_listener_list; + struct list_head event_listener_list; + struct list_head irq_event_handler_list; + struct mutex irq_event_handler_lock; /* Locks access to handlers list */ + struct { + atomic64_t tid; + struct list_head trans_list; + spinlock_t trans_list_lock; /* protects trans_list writes */ + bool use_emad; + bool enable_string_tlv; + } emad; + struct { + u16 *mapping; /* lag_id+port_index to local_port mapping */ + } lag; + struct mlxsw_res res; + struct mlxsw_hwmon *hwmon; + struct mlxsw_thermal *thermal; + struct mlxsw_linecards *linecards; + struct mlxsw_core_port *ports; + unsigned int max_ports; + atomic_t active_ports_count; + bool fw_flash_in_progress; + struct { + struct devlink_health_reporter *fw_fatal; + } health; + struct mlxsw_env *env; + unsigned long driver_priv[]; + /* driver_priv has to be always the last item */ +}; + +struct mlxsw_linecards *mlxsw_core_linecards(struct mlxsw_core *mlxsw_core) +{ + return mlxsw_core->linecards; +} + +void mlxsw_core_linecards_set(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecards *linecards) +{ + mlxsw_core->linecards = linecards; +} + +#define MLXSW_PORT_MAX_PORTS_DEFAULT 0x40 + +static u64 mlxsw_ports_occ_get(void *priv) +{ + struct mlxsw_core *mlxsw_core = priv; + + return atomic_read(&mlxsw_core->active_ports_count); +} + +static int mlxsw_core_resources_ports_register(struct mlxsw_core *mlxsw_core) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + struct devlink_resource_size_params ports_num_params; + u32 max_ports; + + max_ports = mlxsw_core->max_ports - 1; + devlink_resource_size_params_init(&ports_num_params, max_ports, + max_ports, 1, + DEVLINK_RESOURCE_UNIT_ENTRY); + + return devl_resource_register(devlink, + DEVLINK_RESOURCE_GENERIC_NAME_PORTS, + max_ports, MLXSW_CORE_RESOURCE_PORTS, + DEVLINK_RESOURCE_ID_PARENT_TOP, + &ports_num_params); +} + +static int mlxsw_ports_init(struct mlxsw_core *mlxsw_core, bool reload) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + int err; + + /* Switch ports are numbered from 1 to queried value */ + if (MLXSW_CORE_RES_VALID(mlxsw_core, MAX_SYSTEM_PORT)) + mlxsw_core->max_ports = MLXSW_CORE_RES_GET(mlxsw_core, + MAX_SYSTEM_PORT) + 1; + else + mlxsw_core->max_ports = MLXSW_PORT_MAX_PORTS_DEFAULT + 1; + + mlxsw_core->ports = kcalloc(mlxsw_core->max_ports, + sizeof(struct mlxsw_core_port), GFP_KERNEL); + if (!mlxsw_core->ports) + return -ENOMEM; + + if (!reload) { + err = mlxsw_core_resources_ports_register(mlxsw_core); + if (err) + goto err_resources_ports_register; + } + atomic_set(&mlxsw_core->active_ports_count, 0); + devl_resource_occ_get_register(devlink, MLXSW_CORE_RESOURCE_PORTS, + mlxsw_ports_occ_get, mlxsw_core); + + return 0; + +err_resources_ports_register: + kfree(mlxsw_core->ports); + return err; +} + +static void mlxsw_ports_fini(struct mlxsw_core *mlxsw_core, bool reload) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + + devl_resource_occ_get_unregister(devlink, MLXSW_CORE_RESOURCE_PORTS); + if (!reload) + devl_resources_unregister(priv_to_devlink(mlxsw_core)); + + kfree(mlxsw_core->ports); +} + +unsigned int mlxsw_core_max_ports(const struct mlxsw_core *mlxsw_core) +{ + return mlxsw_core->max_ports; +} +EXPORT_SYMBOL(mlxsw_core_max_ports); + +int mlxsw_core_max_lag(struct mlxsw_core *mlxsw_core, u16 *p_max_lag) +{ + struct mlxsw_driver *driver = mlxsw_core->driver; + + if (driver->profile->used_max_lag) { + *p_max_lag = driver->profile->max_lag; + return 0; + } + + if (!MLXSW_CORE_RES_VALID(mlxsw_core, MAX_LAG)) + return -EIO; + + *p_max_lag = MLXSW_CORE_RES_GET(mlxsw_core, MAX_LAG); + return 0; +} +EXPORT_SYMBOL(mlxsw_core_max_lag); + +void *mlxsw_core_driver_priv(struct mlxsw_core *mlxsw_core) +{ + return mlxsw_core->driver_priv; +} +EXPORT_SYMBOL(mlxsw_core_driver_priv); + +bool +mlxsw_core_fw_rev_minor_subminor_validate(const struct mlxsw_fw_rev *rev, + const struct mlxsw_fw_rev *req_rev) +{ + return rev->minor > req_rev->minor || + (rev->minor == req_rev->minor && + rev->subminor >= req_rev->subminor); +} +EXPORT_SYMBOL(mlxsw_core_fw_rev_minor_subminor_validate); + +struct mlxsw_rx_listener_item { + struct list_head list; + struct mlxsw_rx_listener rxl; + void *priv; + bool enabled; +}; + +struct mlxsw_event_listener_item { + struct list_head list; + struct mlxsw_core *mlxsw_core; + struct mlxsw_event_listener el; + void *priv; +}; + +static const u8 mlxsw_core_trap_groups[] = { + MLXSW_REG_HTGT_TRAP_GROUP_EMAD, + MLXSW_REG_HTGT_TRAP_GROUP_CORE_EVENT, +}; + +static int mlxsw_core_trap_groups_set(struct mlxsw_core *mlxsw_core) +{ + char htgt_pl[MLXSW_REG_HTGT_LEN]; + int err; + int i; + + if (!(mlxsw_core->bus->features & MLXSW_BUS_F_TXRX)) + return 0; + + for (i = 0; i < ARRAY_SIZE(mlxsw_core_trap_groups); i++) { + mlxsw_reg_htgt_pack(htgt_pl, mlxsw_core_trap_groups[i], + MLXSW_REG_HTGT_INVALID_POLICER, + MLXSW_REG_HTGT_DEFAULT_PRIORITY, + MLXSW_REG_HTGT_DEFAULT_TC); + err = mlxsw_reg_write(mlxsw_core, MLXSW_REG(htgt), htgt_pl); + if (err) + return err; + } + return 0; +} + +/****************** + * EMAD processing + ******************/ + +/* emad_eth_hdr_dmac + * Destination MAC in EMAD's Ethernet header. + * Must be set to 01:02:c9:00:00:01 + */ +MLXSW_ITEM_BUF(emad, eth_hdr, dmac, 0x00, 6); + +/* emad_eth_hdr_smac + * Source MAC in EMAD's Ethernet header. + * Must be set to 00:02:c9:01:02:03 + */ +MLXSW_ITEM_BUF(emad, eth_hdr, smac, 0x06, 6); + +/* emad_eth_hdr_ethertype + * Ethertype in EMAD's Ethernet header. + * Must be set to 0x8932 + */ +MLXSW_ITEM32(emad, eth_hdr, ethertype, 0x0C, 16, 16); + +/* emad_eth_hdr_mlx_proto + * Mellanox protocol. + * Must be set to 0x0. + */ +MLXSW_ITEM32(emad, eth_hdr, mlx_proto, 0x0C, 8, 8); + +/* emad_eth_hdr_ver + * Mellanox protocol version. + * Must be set to 0x0. + */ +MLXSW_ITEM32(emad, eth_hdr, ver, 0x0C, 4, 4); + +/* emad_op_tlv_type + * Type of the TLV. + * Must be set to 0x1 (operation TLV). + */ +MLXSW_ITEM32(emad, op_tlv, type, 0x00, 27, 5); + +/* emad_op_tlv_len + * Length of the operation TLV in u32. + * Must be set to 0x4. + */ +MLXSW_ITEM32(emad, op_tlv, len, 0x00, 16, 11); + +/* emad_op_tlv_dr + * Direct route bit. Setting to 1 indicates the EMAD is a direct route + * EMAD. DR TLV must follow. + * + * Note: Currently not supported and must not be set. + */ +MLXSW_ITEM32(emad, op_tlv, dr, 0x00, 15, 1); + +/* emad_op_tlv_status + * Returned status in case of EMAD response. Must be set to 0 in case + * of EMAD request. + * 0x0 - success + * 0x1 - device is busy. Requester should retry + * 0x2 - Mellanox protocol version not supported + * 0x3 - unknown TLV + * 0x4 - register not supported + * 0x5 - operation class not supported + * 0x6 - EMAD method not supported + * 0x7 - bad parameter (e.g. port out of range) + * 0x8 - resource not available + * 0x9 - message receipt acknowledgment. Requester should retry + * 0x70 - internal error + */ +MLXSW_ITEM32(emad, op_tlv, status, 0x00, 8, 7); + +/* emad_op_tlv_register_id + * Register ID of register within register TLV. + */ +MLXSW_ITEM32(emad, op_tlv, register_id, 0x04, 16, 16); + +/* emad_op_tlv_r + * Response bit. Setting to 1 indicates Response, otherwise request. + */ +MLXSW_ITEM32(emad, op_tlv, r, 0x04, 15, 1); + +/* emad_op_tlv_method + * EMAD method type. + * 0x1 - query + * 0x2 - write + * 0x3 - send (currently not supported) + * 0x4 - event + */ +MLXSW_ITEM32(emad, op_tlv, method, 0x04, 8, 7); + +/* emad_op_tlv_class + * EMAD operation class. Must be set to 0x1 (REG_ACCESS). + */ +MLXSW_ITEM32(emad, op_tlv, class, 0x04, 0, 8); + +/* emad_op_tlv_tid + * EMAD transaction ID. Used for pairing request and response EMADs. + */ +MLXSW_ITEM64(emad, op_tlv, tid, 0x08, 0, 64); + +/* emad_string_tlv_type + * Type of the TLV. + * Must be set to 0x2 (string TLV). + */ +MLXSW_ITEM32(emad, string_tlv, type, 0x00, 27, 5); + +/* emad_string_tlv_len + * Length of the string TLV in u32. + */ +MLXSW_ITEM32(emad, string_tlv, len, 0x00, 16, 11); + +#define MLXSW_EMAD_STRING_TLV_STRING_LEN 128 + +/* emad_string_tlv_string + * String provided by the device's firmware in case of erroneous register access + */ +MLXSW_ITEM_BUF(emad, string_tlv, string, 0x04, + MLXSW_EMAD_STRING_TLV_STRING_LEN); + +/* emad_reg_tlv_type + * Type of the TLV. + * Must be set to 0x3 (register TLV). + */ +MLXSW_ITEM32(emad, reg_tlv, type, 0x00, 27, 5); + +/* emad_reg_tlv_len + * Length of the operation TLV in u32. + */ +MLXSW_ITEM32(emad, reg_tlv, len, 0x00, 16, 11); + +/* emad_end_tlv_type + * Type of the TLV. + * Must be set to 0x0 (end TLV). + */ +MLXSW_ITEM32(emad, end_tlv, type, 0x00, 27, 5); + +/* emad_end_tlv_len + * Length of the end TLV in u32. + * Must be set to 1. + */ +MLXSW_ITEM32(emad, end_tlv, len, 0x00, 16, 11); + +enum mlxsw_core_reg_access_type { + MLXSW_CORE_REG_ACCESS_TYPE_QUERY, + MLXSW_CORE_REG_ACCESS_TYPE_WRITE, +}; + +static inline const char * +mlxsw_core_reg_access_type_str(enum mlxsw_core_reg_access_type type) +{ + switch (type) { + case MLXSW_CORE_REG_ACCESS_TYPE_QUERY: + return "query"; + case MLXSW_CORE_REG_ACCESS_TYPE_WRITE: + return "write"; + } + BUG(); +} + +static void mlxsw_emad_pack_end_tlv(char *end_tlv) +{ + mlxsw_emad_end_tlv_type_set(end_tlv, MLXSW_EMAD_TLV_TYPE_END); + mlxsw_emad_end_tlv_len_set(end_tlv, MLXSW_EMAD_END_TLV_LEN); +} + +static void mlxsw_emad_pack_reg_tlv(char *reg_tlv, + const struct mlxsw_reg_info *reg, + char *payload) +{ + mlxsw_emad_reg_tlv_type_set(reg_tlv, MLXSW_EMAD_TLV_TYPE_REG); + mlxsw_emad_reg_tlv_len_set(reg_tlv, reg->len / sizeof(u32) + 1); + memcpy(reg_tlv + sizeof(u32), payload, reg->len); +} + +static void mlxsw_emad_pack_string_tlv(char *string_tlv) +{ + mlxsw_emad_string_tlv_type_set(string_tlv, MLXSW_EMAD_TLV_TYPE_STRING); + mlxsw_emad_string_tlv_len_set(string_tlv, MLXSW_EMAD_STRING_TLV_LEN); +} + +static void mlxsw_emad_pack_op_tlv(char *op_tlv, + const struct mlxsw_reg_info *reg, + enum mlxsw_core_reg_access_type type, + u64 tid) +{ + mlxsw_emad_op_tlv_type_set(op_tlv, MLXSW_EMAD_TLV_TYPE_OP); + mlxsw_emad_op_tlv_len_set(op_tlv, MLXSW_EMAD_OP_TLV_LEN); + mlxsw_emad_op_tlv_dr_set(op_tlv, 0); + mlxsw_emad_op_tlv_status_set(op_tlv, 0); + mlxsw_emad_op_tlv_register_id_set(op_tlv, reg->id); + mlxsw_emad_op_tlv_r_set(op_tlv, MLXSW_EMAD_OP_TLV_REQUEST); + if (type == MLXSW_CORE_REG_ACCESS_TYPE_QUERY) + mlxsw_emad_op_tlv_method_set(op_tlv, + MLXSW_EMAD_OP_TLV_METHOD_QUERY); + else + mlxsw_emad_op_tlv_method_set(op_tlv, + MLXSW_EMAD_OP_TLV_METHOD_WRITE); + mlxsw_emad_op_tlv_class_set(op_tlv, + MLXSW_EMAD_OP_TLV_CLASS_REG_ACCESS); + mlxsw_emad_op_tlv_tid_set(op_tlv, tid); +} + +static int mlxsw_emad_construct_eth_hdr(struct sk_buff *skb) +{ + char *eth_hdr = skb_push(skb, MLXSW_EMAD_ETH_HDR_LEN); + + mlxsw_emad_eth_hdr_dmac_memcpy_to(eth_hdr, MLXSW_EMAD_EH_DMAC); + mlxsw_emad_eth_hdr_smac_memcpy_to(eth_hdr, MLXSW_EMAD_EH_SMAC); + mlxsw_emad_eth_hdr_ethertype_set(eth_hdr, MLXSW_EMAD_EH_ETHERTYPE); + mlxsw_emad_eth_hdr_mlx_proto_set(eth_hdr, MLXSW_EMAD_EH_MLX_PROTO); + mlxsw_emad_eth_hdr_ver_set(eth_hdr, MLXSW_EMAD_EH_PROTO_VERSION); + + skb_reset_mac_header(skb); + + return 0; +} + +static void mlxsw_emad_construct(struct sk_buff *skb, + const struct mlxsw_reg_info *reg, + char *payload, + enum mlxsw_core_reg_access_type type, + u64 tid, bool enable_string_tlv) +{ + char *buf; + + buf = skb_push(skb, MLXSW_EMAD_END_TLV_LEN * sizeof(u32)); + mlxsw_emad_pack_end_tlv(buf); + + buf = skb_push(skb, reg->len + sizeof(u32)); + mlxsw_emad_pack_reg_tlv(buf, reg, payload); + + if (enable_string_tlv) { + buf = skb_push(skb, MLXSW_EMAD_STRING_TLV_LEN * sizeof(u32)); + mlxsw_emad_pack_string_tlv(buf); + } + + buf = skb_push(skb, MLXSW_EMAD_OP_TLV_LEN * sizeof(u32)); + mlxsw_emad_pack_op_tlv(buf, reg, type, tid); + + mlxsw_emad_construct_eth_hdr(skb); +} + +struct mlxsw_emad_tlv_offsets { + u16 op_tlv; + u16 string_tlv; + u16 reg_tlv; +}; + +static bool mlxsw_emad_tlv_is_string_tlv(const char *tlv) +{ + u8 tlv_type = mlxsw_emad_string_tlv_type_get(tlv); + + return tlv_type == MLXSW_EMAD_TLV_TYPE_STRING; +} + +static void mlxsw_emad_tlv_parse(struct sk_buff *skb) +{ + struct mlxsw_emad_tlv_offsets *offsets = + (struct mlxsw_emad_tlv_offsets *) skb->cb; + + offsets->op_tlv = MLXSW_EMAD_ETH_HDR_LEN; + offsets->string_tlv = 0; + offsets->reg_tlv = MLXSW_EMAD_ETH_HDR_LEN + + MLXSW_EMAD_OP_TLV_LEN * sizeof(u32); + + /* If string TLV is present, it must come after the operation TLV. */ + if (mlxsw_emad_tlv_is_string_tlv(skb->data + offsets->reg_tlv)) { + offsets->string_tlv = offsets->reg_tlv; + offsets->reg_tlv += MLXSW_EMAD_STRING_TLV_LEN * sizeof(u32); + } +} + +static char *mlxsw_emad_op_tlv(const struct sk_buff *skb) +{ + struct mlxsw_emad_tlv_offsets *offsets = + (struct mlxsw_emad_tlv_offsets *) skb->cb; + + return ((char *) (skb->data + offsets->op_tlv)); +} + +static char *mlxsw_emad_string_tlv(const struct sk_buff *skb) +{ + struct mlxsw_emad_tlv_offsets *offsets = + (struct mlxsw_emad_tlv_offsets *) skb->cb; + + if (!offsets->string_tlv) + return NULL; + + return ((char *) (skb->data + offsets->string_tlv)); +} + +static char *mlxsw_emad_reg_tlv(const struct sk_buff *skb) +{ + struct mlxsw_emad_tlv_offsets *offsets = + (struct mlxsw_emad_tlv_offsets *) skb->cb; + + return ((char *) (skb->data + offsets->reg_tlv)); +} + +static char *mlxsw_emad_reg_payload(const char *reg_tlv) +{ + return ((char *) (reg_tlv + sizeof(u32))); +} + +static char *mlxsw_emad_reg_payload_cmd(const char *mbox) +{ + return ((char *) (mbox + (MLXSW_EMAD_OP_TLV_LEN + 1) * sizeof(u32))); +} + +static u64 mlxsw_emad_get_tid(const struct sk_buff *skb) +{ + char *op_tlv; + + op_tlv = mlxsw_emad_op_tlv(skb); + return mlxsw_emad_op_tlv_tid_get(op_tlv); +} + +static bool mlxsw_emad_is_resp(const struct sk_buff *skb) +{ + char *op_tlv; + + op_tlv = mlxsw_emad_op_tlv(skb); + return (mlxsw_emad_op_tlv_r_get(op_tlv) == MLXSW_EMAD_OP_TLV_RESPONSE); +} + +static int mlxsw_emad_process_status(char *op_tlv, + enum mlxsw_emad_op_tlv_status *p_status) +{ + *p_status = mlxsw_emad_op_tlv_status_get(op_tlv); + + switch (*p_status) { + case MLXSW_EMAD_OP_TLV_STATUS_SUCCESS: + return 0; + case MLXSW_EMAD_OP_TLV_STATUS_BUSY: + case MLXSW_EMAD_OP_TLV_STATUS_MESSAGE_RECEIPT_ACK: + return -EAGAIN; + case MLXSW_EMAD_OP_TLV_STATUS_VERSION_NOT_SUPPORTED: + case MLXSW_EMAD_OP_TLV_STATUS_UNKNOWN_TLV: + case MLXSW_EMAD_OP_TLV_STATUS_REGISTER_NOT_SUPPORTED: + case MLXSW_EMAD_OP_TLV_STATUS_CLASS_NOT_SUPPORTED: + case MLXSW_EMAD_OP_TLV_STATUS_METHOD_NOT_SUPPORTED: + case MLXSW_EMAD_OP_TLV_STATUS_BAD_PARAMETER: + case MLXSW_EMAD_OP_TLV_STATUS_RESOURCE_NOT_AVAILABLE: + case MLXSW_EMAD_OP_TLV_STATUS_INTERNAL_ERROR: + default: + return -EIO; + } +} + +static int +mlxsw_emad_process_status_skb(struct sk_buff *skb, + enum mlxsw_emad_op_tlv_status *p_status) +{ + return mlxsw_emad_process_status(mlxsw_emad_op_tlv(skb), p_status); +} + +struct mlxsw_reg_trans { + struct list_head list; + struct list_head bulk_list; + struct mlxsw_core *core; + struct sk_buff *tx_skb; + struct mlxsw_tx_info tx_info; + struct delayed_work timeout_dw; + unsigned int retries; + u64 tid; + struct completion completion; + atomic_t active; + mlxsw_reg_trans_cb_t *cb; + unsigned long cb_priv; + const struct mlxsw_reg_info *reg; + enum mlxsw_core_reg_access_type type; + int err; + char *emad_err_string; + enum mlxsw_emad_op_tlv_status emad_status; + struct rcu_head rcu; +}; + +static void mlxsw_emad_process_string_tlv(const struct sk_buff *skb, + struct mlxsw_reg_trans *trans) +{ + char *string_tlv; + char *string; + + string_tlv = mlxsw_emad_string_tlv(skb); + if (!string_tlv) + return; + + trans->emad_err_string = kzalloc(MLXSW_EMAD_STRING_TLV_STRING_LEN, + GFP_ATOMIC); + if (!trans->emad_err_string) + return; + + string = mlxsw_emad_string_tlv_string_data(string_tlv); + strscpy(trans->emad_err_string, string, + MLXSW_EMAD_STRING_TLV_STRING_LEN); +} + +#define MLXSW_EMAD_TIMEOUT_DURING_FW_FLASH_MS 3000 +#define MLXSW_EMAD_TIMEOUT_MS 200 + +static void mlxsw_emad_trans_timeout_schedule(struct mlxsw_reg_trans *trans) +{ + unsigned long timeout = msecs_to_jiffies(MLXSW_EMAD_TIMEOUT_MS); + + if (trans->core->fw_flash_in_progress) + timeout = msecs_to_jiffies(MLXSW_EMAD_TIMEOUT_DURING_FW_FLASH_MS); + + queue_delayed_work(trans->core->emad_wq, &trans->timeout_dw, + timeout << trans->retries); +} + +static int mlxsw_emad_transmit(struct mlxsw_core *mlxsw_core, + struct mlxsw_reg_trans *trans) +{ + struct sk_buff *skb; + int err; + + skb = skb_clone(trans->tx_skb, GFP_KERNEL); + if (!skb) + return -ENOMEM; + + trace_devlink_hwmsg(priv_to_devlink(mlxsw_core), false, 0, + skb->data + mlxsw_core->driver->txhdr_len, + skb->len - mlxsw_core->driver->txhdr_len); + + atomic_set(&trans->active, 1); + err = mlxsw_core_skb_transmit(mlxsw_core, skb, &trans->tx_info); + if (err) { + dev_kfree_skb(skb); + return err; + } + mlxsw_emad_trans_timeout_schedule(trans); + return 0; +} + +static void mlxsw_emad_trans_finish(struct mlxsw_reg_trans *trans, int err) +{ + struct mlxsw_core *mlxsw_core = trans->core; + + dev_kfree_skb(trans->tx_skb); + spin_lock_bh(&mlxsw_core->emad.trans_list_lock); + list_del_rcu(&trans->list); + spin_unlock_bh(&mlxsw_core->emad.trans_list_lock); + trans->err = err; + complete(&trans->completion); +} + +static void mlxsw_emad_transmit_retry(struct mlxsw_core *mlxsw_core, + struct mlxsw_reg_trans *trans) +{ + int err; + + if (trans->retries < MLXSW_EMAD_MAX_RETRY) { + trans->retries++; + err = mlxsw_emad_transmit(trans->core, trans); + if (err == 0) + return; + + if (!atomic_dec_and_test(&trans->active)) + return; + } else { + err = -EIO; + } + mlxsw_emad_trans_finish(trans, err); +} + +static void mlxsw_emad_trans_timeout_work(struct work_struct *work) +{ + struct mlxsw_reg_trans *trans = container_of(work, + struct mlxsw_reg_trans, + timeout_dw.work); + + if (!atomic_dec_and_test(&trans->active)) + return; + + mlxsw_emad_transmit_retry(trans->core, trans); +} + +static void mlxsw_emad_process_response(struct mlxsw_core *mlxsw_core, + struct mlxsw_reg_trans *trans, + struct sk_buff *skb) +{ + int err; + + if (!atomic_dec_and_test(&trans->active)) + return; + + err = mlxsw_emad_process_status_skb(skb, &trans->emad_status); + if (err == -EAGAIN) { + mlxsw_emad_transmit_retry(mlxsw_core, trans); + } else { + if (err == 0) { + char *reg_tlv = mlxsw_emad_reg_tlv(skb); + + if (trans->cb) + trans->cb(mlxsw_core, + mlxsw_emad_reg_payload(reg_tlv), + trans->reg->len, trans->cb_priv); + } else { + mlxsw_emad_process_string_tlv(skb, trans); + } + mlxsw_emad_trans_finish(trans, err); + } +} + +/* called with rcu read lock held */ +static void mlxsw_emad_rx_listener_func(struct sk_buff *skb, u16 local_port, + void *priv) +{ + struct mlxsw_core *mlxsw_core = priv; + struct mlxsw_reg_trans *trans; + + trace_devlink_hwmsg(priv_to_devlink(mlxsw_core), true, 0, + skb->data, skb->len); + + mlxsw_emad_tlv_parse(skb); + + if (!mlxsw_emad_is_resp(skb)) + goto free_skb; + + list_for_each_entry_rcu(trans, &mlxsw_core->emad.trans_list, list) { + if (mlxsw_emad_get_tid(skb) == trans->tid) { + mlxsw_emad_process_response(mlxsw_core, trans, skb); + break; + } + } + +free_skb: + dev_kfree_skb(skb); +} + +static const struct mlxsw_listener mlxsw_emad_rx_listener = + MLXSW_RXL(mlxsw_emad_rx_listener_func, ETHEMAD, TRAP_TO_CPU, false, + EMAD, DISCARD); + +static int mlxsw_emad_init(struct mlxsw_core *mlxsw_core) +{ + struct workqueue_struct *emad_wq; + u64 tid; + int err; + + if (!(mlxsw_core->bus->features & MLXSW_BUS_F_TXRX)) + return 0; + + emad_wq = alloc_workqueue("mlxsw_core_emad", 0, 0); + if (!emad_wq) + return -ENOMEM; + mlxsw_core->emad_wq = emad_wq; + + /* Set the upper 32 bits of the transaction ID field to a random + * number. This allows us to discard EMADs addressed to other + * devices. + */ + get_random_bytes(&tid, 4); + tid <<= 32; + atomic64_set(&mlxsw_core->emad.tid, tid); + + INIT_LIST_HEAD(&mlxsw_core->emad.trans_list); + spin_lock_init(&mlxsw_core->emad.trans_list_lock); + + err = mlxsw_core_trap_register(mlxsw_core, &mlxsw_emad_rx_listener, + mlxsw_core); + if (err) + goto err_trap_register; + + mlxsw_core->emad.use_emad = true; + + return 0; + +err_trap_register: + destroy_workqueue(mlxsw_core->emad_wq); + return err; +} + +static void mlxsw_emad_fini(struct mlxsw_core *mlxsw_core) +{ + + if (!(mlxsw_core->bus->features & MLXSW_BUS_F_TXRX)) + return; + + mlxsw_core->emad.use_emad = false; + mlxsw_core_trap_unregister(mlxsw_core, &mlxsw_emad_rx_listener, + mlxsw_core); + destroy_workqueue(mlxsw_core->emad_wq); +} + +static struct sk_buff *mlxsw_emad_alloc(const struct mlxsw_core *mlxsw_core, + u16 reg_len, bool enable_string_tlv) +{ + struct sk_buff *skb; + u16 emad_len; + + emad_len = (reg_len + sizeof(u32) + MLXSW_EMAD_ETH_HDR_LEN + + (MLXSW_EMAD_OP_TLV_LEN + MLXSW_EMAD_END_TLV_LEN) * + sizeof(u32) + mlxsw_core->driver->txhdr_len); + if (enable_string_tlv) + emad_len += MLXSW_EMAD_STRING_TLV_LEN * sizeof(u32); + if (emad_len > MLXSW_EMAD_MAX_FRAME_LEN) + return NULL; + + skb = netdev_alloc_skb(NULL, emad_len); + if (!skb) + return NULL; + memset(skb->data, 0, emad_len); + skb_reserve(skb, emad_len); + + return skb; +} + +static int mlxsw_emad_reg_access(struct mlxsw_core *mlxsw_core, + const struct mlxsw_reg_info *reg, + char *payload, + enum mlxsw_core_reg_access_type type, + struct mlxsw_reg_trans *trans, + struct list_head *bulk_list, + mlxsw_reg_trans_cb_t *cb, + unsigned long cb_priv, u64 tid) +{ + bool enable_string_tlv; + struct sk_buff *skb; + int err; + + dev_dbg(mlxsw_core->bus_info->dev, "EMAD reg access (tid=%llx,reg_id=%x(%s),type=%s)\n", + tid, reg->id, mlxsw_reg_id_str(reg->id), + mlxsw_core_reg_access_type_str(type)); + + /* Since this can be changed during emad_reg_access, read it once and + * use the value all the way. + */ + enable_string_tlv = mlxsw_core->emad.enable_string_tlv; + + skb = mlxsw_emad_alloc(mlxsw_core, reg->len, enable_string_tlv); + if (!skb) + return -ENOMEM; + + list_add_tail(&trans->bulk_list, bulk_list); + trans->core = mlxsw_core; + trans->tx_skb = skb; + trans->tx_info.local_port = MLXSW_PORT_CPU_PORT; + trans->tx_info.is_emad = true; + INIT_DELAYED_WORK(&trans->timeout_dw, mlxsw_emad_trans_timeout_work); + trans->tid = tid; + init_completion(&trans->completion); + trans->cb = cb; + trans->cb_priv = cb_priv; + trans->reg = reg; + trans->type = type; + + mlxsw_emad_construct(skb, reg, payload, type, trans->tid, + enable_string_tlv); + mlxsw_core->driver->txhdr_construct(skb, &trans->tx_info); + + spin_lock_bh(&mlxsw_core->emad.trans_list_lock); + list_add_tail_rcu(&trans->list, &mlxsw_core->emad.trans_list); + spin_unlock_bh(&mlxsw_core->emad.trans_list_lock); + err = mlxsw_emad_transmit(mlxsw_core, trans); + if (err) + goto err_out; + return 0; + +err_out: + spin_lock_bh(&mlxsw_core->emad.trans_list_lock); + list_del_rcu(&trans->list); + spin_unlock_bh(&mlxsw_core->emad.trans_list_lock); + list_del(&trans->bulk_list); + dev_kfree_skb(trans->tx_skb); + return err; +} + +/***************** + * Core functions + *****************/ + +int mlxsw_core_driver_register(struct mlxsw_driver *mlxsw_driver) +{ + spin_lock(&mlxsw_core_driver_list_lock); + list_add_tail(&mlxsw_driver->list, &mlxsw_core_driver_list); + spin_unlock(&mlxsw_core_driver_list_lock); + return 0; +} +EXPORT_SYMBOL(mlxsw_core_driver_register); + +void mlxsw_core_driver_unregister(struct mlxsw_driver *mlxsw_driver) +{ + spin_lock(&mlxsw_core_driver_list_lock); + list_del(&mlxsw_driver->list); + spin_unlock(&mlxsw_core_driver_list_lock); +} +EXPORT_SYMBOL(mlxsw_core_driver_unregister); + +static struct mlxsw_driver *__driver_find(const char *kind) +{ + struct mlxsw_driver *mlxsw_driver; + + list_for_each_entry(mlxsw_driver, &mlxsw_core_driver_list, list) { + if (strcmp(mlxsw_driver->kind, kind) == 0) + return mlxsw_driver; + } + return NULL; +} + +static struct mlxsw_driver *mlxsw_core_driver_get(const char *kind) +{ + struct mlxsw_driver *mlxsw_driver; + + spin_lock(&mlxsw_core_driver_list_lock); + mlxsw_driver = __driver_find(kind); + spin_unlock(&mlxsw_core_driver_list_lock); + return mlxsw_driver; +} + +int mlxsw_core_fw_flash(struct mlxsw_core *mlxsw_core, + struct mlxfw_dev *mlxfw_dev, + const struct firmware *firmware, + struct netlink_ext_ack *extack) +{ + int err; + + mlxsw_core->fw_flash_in_progress = true; + err = mlxfw_firmware_flash(mlxfw_dev, firmware, extack); + mlxsw_core->fw_flash_in_progress = false; + + return err; +} + +struct mlxsw_core_fw_info { + struct mlxfw_dev mlxfw_dev; + struct mlxsw_core *mlxsw_core; +}; + +static int mlxsw_core_fw_component_query(struct mlxfw_dev *mlxfw_dev, + u16 component_index, u32 *p_max_size, + u8 *p_align_bits, u16 *p_max_write_size) +{ + struct mlxsw_core_fw_info *mlxsw_core_fw_info = + container_of(mlxfw_dev, struct mlxsw_core_fw_info, mlxfw_dev); + struct mlxsw_core *mlxsw_core = mlxsw_core_fw_info->mlxsw_core; + char mcqi_pl[MLXSW_REG_MCQI_LEN]; + int err; + + mlxsw_reg_mcqi_pack(mcqi_pl, component_index); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mcqi), mcqi_pl); + if (err) + return err; + mlxsw_reg_mcqi_unpack(mcqi_pl, p_max_size, p_align_bits, p_max_write_size); + + *p_align_bits = max_t(u8, *p_align_bits, 2); + *p_max_write_size = min_t(u16, *p_max_write_size, MLXSW_REG_MCDA_MAX_DATA_LEN); + return 0; +} + +static int mlxsw_core_fw_fsm_lock(struct mlxfw_dev *mlxfw_dev, u32 *fwhandle) +{ + struct mlxsw_core_fw_info *mlxsw_core_fw_info = + container_of(mlxfw_dev, struct mlxsw_core_fw_info, mlxfw_dev); + struct mlxsw_core *mlxsw_core = mlxsw_core_fw_info->mlxsw_core; + char mcc_pl[MLXSW_REG_MCC_LEN]; + u8 control_state; + int err; + + mlxsw_reg_mcc_pack(mcc_pl, 0, 0, 0, 0); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mcc), mcc_pl); + if (err) + return err; + + mlxsw_reg_mcc_unpack(mcc_pl, fwhandle, NULL, &control_state); + if (control_state != MLXFW_FSM_STATE_IDLE) + return -EBUSY; + + mlxsw_reg_mcc_pack(mcc_pl, MLXSW_REG_MCC_INSTRUCTION_LOCK_UPDATE_HANDLE, 0, *fwhandle, 0); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mcc), mcc_pl); +} + +static int mlxsw_core_fw_fsm_component_update(struct mlxfw_dev *mlxfw_dev, u32 fwhandle, + u16 component_index, u32 component_size) +{ + struct mlxsw_core_fw_info *mlxsw_core_fw_info = + container_of(mlxfw_dev, struct mlxsw_core_fw_info, mlxfw_dev); + struct mlxsw_core *mlxsw_core = mlxsw_core_fw_info->mlxsw_core; + char mcc_pl[MLXSW_REG_MCC_LEN]; + + mlxsw_reg_mcc_pack(mcc_pl, MLXSW_REG_MCC_INSTRUCTION_UPDATE_COMPONENT, + component_index, fwhandle, component_size); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mcc), mcc_pl); +} + +static int mlxsw_core_fw_fsm_block_download(struct mlxfw_dev *mlxfw_dev, u32 fwhandle, + u8 *data, u16 size, u32 offset) +{ + struct mlxsw_core_fw_info *mlxsw_core_fw_info = + container_of(mlxfw_dev, struct mlxsw_core_fw_info, mlxfw_dev); + struct mlxsw_core *mlxsw_core = mlxsw_core_fw_info->mlxsw_core; + char mcda_pl[MLXSW_REG_MCDA_LEN]; + + mlxsw_reg_mcda_pack(mcda_pl, fwhandle, offset, size, data); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mcda), mcda_pl); +} + +static int mlxsw_core_fw_fsm_component_verify(struct mlxfw_dev *mlxfw_dev, u32 fwhandle, + u16 component_index) +{ + struct mlxsw_core_fw_info *mlxsw_core_fw_info = + container_of(mlxfw_dev, struct mlxsw_core_fw_info, mlxfw_dev); + struct mlxsw_core *mlxsw_core = mlxsw_core_fw_info->mlxsw_core; + char mcc_pl[MLXSW_REG_MCC_LEN]; + + mlxsw_reg_mcc_pack(mcc_pl, MLXSW_REG_MCC_INSTRUCTION_VERIFY_COMPONENT, + component_index, fwhandle, 0); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mcc), mcc_pl); +} + +static int mlxsw_core_fw_fsm_activate(struct mlxfw_dev *mlxfw_dev, u32 fwhandle) +{ + struct mlxsw_core_fw_info *mlxsw_core_fw_info = + container_of(mlxfw_dev, struct mlxsw_core_fw_info, mlxfw_dev); + struct mlxsw_core *mlxsw_core = mlxsw_core_fw_info->mlxsw_core; + char mcc_pl[MLXSW_REG_MCC_LEN]; + + mlxsw_reg_mcc_pack(mcc_pl, MLXSW_REG_MCC_INSTRUCTION_ACTIVATE, 0, fwhandle, 0); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mcc), mcc_pl); +} + +static int mlxsw_core_fw_fsm_query_state(struct mlxfw_dev *mlxfw_dev, u32 fwhandle, + enum mlxfw_fsm_state *fsm_state, + enum mlxfw_fsm_state_err *fsm_state_err) +{ + struct mlxsw_core_fw_info *mlxsw_core_fw_info = + container_of(mlxfw_dev, struct mlxsw_core_fw_info, mlxfw_dev); + struct mlxsw_core *mlxsw_core = mlxsw_core_fw_info->mlxsw_core; + char mcc_pl[MLXSW_REG_MCC_LEN]; + u8 control_state; + u8 error_code; + int err; + + mlxsw_reg_mcc_pack(mcc_pl, 0, 0, fwhandle, 0); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mcc), mcc_pl); + if (err) + return err; + + mlxsw_reg_mcc_unpack(mcc_pl, NULL, &error_code, &control_state); + *fsm_state = control_state; + *fsm_state_err = min_t(enum mlxfw_fsm_state_err, error_code, MLXFW_FSM_STATE_ERR_MAX); + return 0; +} + +static void mlxsw_core_fw_fsm_cancel(struct mlxfw_dev *mlxfw_dev, u32 fwhandle) +{ + struct mlxsw_core_fw_info *mlxsw_core_fw_info = + container_of(mlxfw_dev, struct mlxsw_core_fw_info, mlxfw_dev); + struct mlxsw_core *mlxsw_core = mlxsw_core_fw_info->mlxsw_core; + char mcc_pl[MLXSW_REG_MCC_LEN]; + + mlxsw_reg_mcc_pack(mcc_pl, MLXSW_REG_MCC_INSTRUCTION_CANCEL, 0, fwhandle, 0); + mlxsw_reg_write(mlxsw_core, MLXSW_REG(mcc), mcc_pl); +} + +static void mlxsw_core_fw_fsm_release(struct mlxfw_dev *mlxfw_dev, u32 fwhandle) +{ + struct mlxsw_core_fw_info *mlxsw_core_fw_info = + container_of(mlxfw_dev, struct mlxsw_core_fw_info, mlxfw_dev); + struct mlxsw_core *mlxsw_core = mlxsw_core_fw_info->mlxsw_core; + char mcc_pl[MLXSW_REG_MCC_LEN]; + + mlxsw_reg_mcc_pack(mcc_pl, MLXSW_REG_MCC_INSTRUCTION_RELEASE_UPDATE_HANDLE, 0, fwhandle, 0); + mlxsw_reg_write(mlxsw_core, MLXSW_REG(mcc), mcc_pl); +} + +static const struct mlxfw_dev_ops mlxsw_core_fw_mlxsw_dev_ops = { + .component_query = mlxsw_core_fw_component_query, + .fsm_lock = mlxsw_core_fw_fsm_lock, + .fsm_component_update = mlxsw_core_fw_fsm_component_update, + .fsm_block_download = mlxsw_core_fw_fsm_block_download, + .fsm_component_verify = mlxsw_core_fw_fsm_component_verify, + .fsm_activate = mlxsw_core_fw_fsm_activate, + .fsm_query_state = mlxsw_core_fw_fsm_query_state, + .fsm_cancel = mlxsw_core_fw_fsm_cancel, + .fsm_release = mlxsw_core_fw_fsm_release, +}; + +static int mlxsw_core_dev_fw_flash(struct mlxsw_core *mlxsw_core, + const struct firmware *firmware, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core_fw_info mlxsw_core_fw_info = { + .mlxfw_dev = { + .ops = &mlxsw_core_fw_mlxsw_dev_ops, + .psid = mlxsw_core->bus_info->psid, + .psid_size = strlen(mlxsw_core->bus_info->psid), + .devlink = priv_to_devlink(mlxsw_core), + }, + .mlxsw_core = mlxsw_core + }; + + return mlxsw_core_fw_flash(mlxsw_core, &mlxsw_core_fw_info.mlxfw_dev, + firmware, extack); +} + +static int mlxsw_core_fw_rev_validate(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + const struct mlxsw_fw_rev *req_rev, + const char *filename) +{ + const struct mlxsw_fw_rev *rev = &mlxsw_bus_info->fw_rev; + union devlink_param_value value; + const struct firmware *firmware; + int err; + + /* Don't check if driver does not require it */ + if (!req_rev || !filename) + return 0; + + /* Don't check if devlink 'fw_load_policy' param is 'flash' */ + err = devlink_param_driverinit_value_get(priv_to_devlink(mlxsw_core), + DEVLINK_PARAM_GENERIC_ID_FW_LOAD_POLICY, + &value); + if (err) + return err; + if (value.vu8 == DEVLINK_PARAM_FW_LOAD_POLICY_VALUE_FLASH) + return 0; + + /* Validate driver & FW are compatible */ + if (rev->major != req_rev->major) { + WARN(1, "Mismatch in major FW version [%d:%d] is never expected; Please contact support\n", + rev->major, req_rev->major); + return -EINVAL; + } + if (mlxsw_core_fw_rev_minor_subminor_validate(rev, req_rev)) + return 0; + + dev_err(mlxsw_bus_info->dev, "The firmware version %d.%d.%d is incompatible with the driver (required >= %d.%d.%d)\n", + rev->major, rev->minor, rev->subminor, req_rev->major, + req_rev->minor, req_rev->subminor); + dev_info(mlxsw_bus_info->dev, "Flashing firmware using file %s\n", filename); + + err = request_firmware_direct(&firmware, filename, mlxsw_bus_info->dev); + if (err) { + dev_err(mlxsw_bus_info->dev, "Could not request firmware file %s\n", filename); + return err; + } + + err = mlxsw_core_dev_fw_flash(mlxsw_core, firmware, NULL); + release_firmware(firmware); + if (err) + dev_err(mlxsw_bus_info->dev, "Could not upgrade firmware\n"); + + /* On FW flash success, tell the caller FW reset is needed + * if current FW supports it. + */ + if (rev->minor >= req_rev->can_reset_minor) + return err ? err : -EAGAIN; + else + return 0; +} + +static int mlxsw_core_fw_flash_update(struct mlxsw_core *mlxsw_core, + struct devlink_flash_update_params *params, + struct netlink_ext_ack *extack) +{ + return mlxsw_core_dev_fw_flash(mlxsw_core, params->fw, extack); +} + +static int mlxsw_core_devlink_param_fw_load_policy_validate(struct devlink *devlink, u32 id, + union devlink_param_value val, + struct netlink_ext_ack *extack) +{ + if (val.vu8 != DEVLINK_PARAM_FW_LOAD_POLICY_VALUE_DRIVER && + val.vu8 != DEVLINK_PARAM_FW_LOAD_POLICY_VALUE_FLASH) { + NL_SET_ERR_MSG_MOD(extack, "'fw_load_policy' must be 'driver' or 'flash'"); + return -EINVAL; + } + + return 0; +} + +static const struct devlink_param mlxsw_core_fw_devlink_params[] = { + DEVLINK_PARAM_GENERIC(FW_LOAD_POLICY, BIT(DEVLINK_PARAM_CMODE_DRIVERINIT), NULL, NULL, + mlxsw_core_devlink_param_fw_load_policy_validate), +}; + +static int mlxsw_core_fw_params_register(struct mlxsw_core *mlxsw_core) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + union devlink_param_value value; + int err; + + err = devlink_params_register(devlink, mlxsw_core_fw_devlink_params, + ARRAY_SIZE(mlxsw_core_fw_devlink_params)); + if (err) + return err; + + value.vu8 = DEVLINK_PARAM_FW_LOAD_POLICY_VALUE_DRIVER; + devlink_param_driverinit_value_set(devlink, DEVLINK_PARAM_GENERIC_ID_FW_LOAD_POLICY, value); + return 0; +} + +static void mlxsw_core_fw_params_unregister(struct mlxsw_core *mlxsw_core) +{ + devlink_params_unregister(priv_to_devlink(mlxsw_core), mlxsw_core_fw_devlink_params, + ARRAY_SIZE(mlxsw_core_fw_devlink_params)); +} + +static void *__dl_port(struct devlink_port *devlink_port) +{ + return container_of(devlink_port, struct mlxsw_core_port, devlink_port); +} + +static int mlxsw_devlink_port_split(struct devlink *devlink, + struct devlink_port *port, + unsigned int count, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core_port *mlxsw_core_port = __dl_port(port); + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + + if (!mlxsw_core->driver->port_split) + return -EOPNOTSUPP; + return mlxsw_core->driver->port_split(mlxsw_core, + mlxsw_core_port->local_port, + count, extack); +} + +static int mlxsw_devlink_port_unsplit(struct devlink *devlink, + struct devlink_port *port, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core_port *mlxsw_core_port = __dl_port(port); + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + + if (!mlxsw_core->driver->port_unsplit) + return -EOPNOTSUPP; + return mlxsw_core->driver->port_unsplit(mlxsw_core, + mlxsw_core_port->local_port, + extack); +} + +static int +mlxsw_devlink_sb_pool_get(struct devlink *devlink, + unsigned int sb_index, u16 pool_index, + struct devlink_sb_pool_info *pool_info) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->sb_pool_get) + return -EOPNOTSUPP; + return mlxsw_driver->sb_pool_get(mlxsw_core, sb_index, + pool_index, pool_info); +} + +static int +mlxsw_devlink_sb_pool_set(struct devlink *devlink, + unsigned int sb_index, u16 pool_index, u32 size, + enum devlink_sb_threshold_type threshold_type, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->sb_pool_set) + return -EOPNOTSUPP; + return mlxsw_driver->sb_pool_set(mlxsw_core, sb_index, + pool_index, size, threshold_type, + extack); +} + +static int mlxsw_devlink_sb_port_pool_get(struct devlink_port *devlink_port, + unsigned int sb_index, u16 pool_index, + u32 *p_threshold) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink_port->devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + struct mlxsw_core_port *mlxsw_core_port = __dl_port(devlink_port); + + if (!mlxsw_driver->sb_port_pool_get || + !mlxsw_core_port_check(mlxsw_core_port)) + return -EOPNOTSUPP; + return mlxsw_driver->sb_port_pool_get(mlxsw_core_port, sb_index, + pool_index, p_threshold); +} + +static int mlxsw_devlink_sb_port_pool_set(struct devlink_port *devlink_port, + unsigned int sb_index, u16 pool_index, + u32 threshold, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink_port->devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + struct mlxsw_core_port *mlxsw_core_port = __dl_port(devlink_port); + + if (!mlxsw_driver->sb_port_pool_set || + !mlxsw_core_port_check(mlxsw_core_port)) + return -EOPNOTSUPP; + return mlxsw_driver->sb_port_pool_set(mlxsw_core_port, sb_index, + pool_index, threshold, extack); +} + +static int +mlxsw_devlink_sb_tc_pool_bind_get(struct devlink_port *devlink_port, + unsigned int sb_index, u16 tc_index, + enum devlink_sb_pool_type pool_type, + u16 *p_pool_index, u32 *p_threshold) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink_port->devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + struct mlxsw_core_port *mlxsw_core_port = __dl_port(devlink_port); + + if (!mlxsw_driver->sb_tc_pool_bind_get || + !mlxsw_core_port_check(mlxsw_core_port)) + return -EOPNOTSUPP; + return mlxsw_driver->sb_tc_pool_bind_get(mlxsw_core_port, sb_index, + tc_index, pool_type, + p_pool_index, p_threshold); +} + +static int +mlxsw_devlink_sb_tc_pool_bind_set(struct devlink_port *devlink_port, + unsigned int sb_index, u16 tc_index, + enum devlink_sb_pool_type pool_type, + u16 pool_index, u32 threshold, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink_port->devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + struct mlxsw_core_port *mlxsw_core_port = __dl_port(devlink_port); + + if (!mlxsw_driver->sb_tc_pool_bind_set || + !mlxsw_core_port_check(mlxsw_core_port)) + return -EOPNOTSUPP; + return mlxsw_driver->sb_tc_pool_bind_set(mlxsw_core_port, sb_index, + tc_index, pool_type, + pool_index, threshold, extack); +} + +static int mlxsw_devlink_sb_occ_snapshot(struct devlink *devlink, + unsigned int sb_index) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->sb_occ_snapshot) + return -EOPNOTSUPP; + return mlxsw_driver->sb_occ_snapshot(mlxsw_core, sb_index); +} + +static int mlxsw_devlink_sb_occ_max_clear(struct devlink *devlink, + unsigned int sb_index) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->sb_occ_max_clear) + return -EOPNOTSUPP; + return mlxsw_driver->sb_occ_max_clear(mlxsw_core, sb_index); +} + +static int +mlxsw_devlink_sb_occ_port_pool_get(struct devlink_port *devlink_port, + unsigned int sb_index, u16 pool_index, + u32 *p_cur, u32 *p_max) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink_port->devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + struct mlxsw_core_port *mlxsw_core_port = __dl_port(devlink_port); + + if (!mlxsw_driver->sb_occ_port_pool_get || + !mlxsw_core_port_check(mlxsw_core_port)) + return -EOPNOTSUPP; + return mlxsw_driver->sb_occ_port_pool_get(mlxsw_core_port, sb_index, + pool_index, p_cur, p_max); +} + +static int +mlxsw_devlink_sb_occ_tc_port_bind_get(struct devlink_port *devlink_port, + unsigned int sb_index, u16 tc_index, + enum devlink_sb_pool_type pool_type, + u32 *p_cur, u32 *p_max) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink_port->devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + struct mlxsw_core_port *mlxsw_core_port = __dl_port(devlink_port); + + if (!mlxsw_driver->sb_occ_tc_port_bind_get || + !mlxsw_core_port_check(mlxsw_core_port)) + return -EOPNOTSUPP; + return mlxsw_driver->sb_occ_tc_port_bind_get(mlxsw_core_port, + sb_index, tc_index, + pool_type, p_cur, p_max); +} + +static int +mlxsw_devlink_info_get(struct devlink *devlink, struct devlink_info_req *req, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + char fw_info_psid[MLXSW_REG_MGIR_FW_INFO_PSID_SIZE]; + u32 hw_rev, fw_major, fw_minor, fw_sub_minor; + char mgir_pl[MLXSW_REG_MGIR_LEN]; + char buf[32]; + int err; + + err = devlink_info_driver_name_put(req, + mlxsw_core->bus_info->device_kind); + if (err) + return err; + + mlxsw_reg_mgir_pack(mgir_pl); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mgir), mgir_pl); + if (err) + return err; + mlxsw_reg_mgir_unpack(mgir_pl, &hw_rev, fw_info_psid, &fw_major, + &fw_minor, &fw_sub_minor); + + sprintf(buf, "%X", hw_rev); + err = devlink_info_version_fixed_put(req, "hw.revision", buf); + if (err) + return err; + + err = devlink_info_version_fixed_put(req, + DEVLINK_INFO_VERSION_GENERIC_FW_PSID, + fw_info_psid); + if (err) + return err; + + sprintf(buf, "%d.%d.%d", fw_major, fw_minor, fw_sub_minor); + err = devlink_info_version_running_put(req, "fw.version", buf); + if (err) + return err; + + return devlink_info_version_running_put(req, + DEVLINK_INFO_VERSION_GENERIC_FW, + buf); +} + +static int +mlxsw_devlink_core_bus_device_reload_down(struct devlink *devlink, + bool netns_change, enum devlink_reload_action action, + enum devlink_reload_limit limit, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + + if (!(mlxsw_core->bus->features & MLXSW_BUS_F_RESET)) + return -EOPNOTSUPP; + + mlxsw_core_bus_device_unregister(mlxsw_core, true); + return 0; +} + +static int +mlxsw_devlink_core_bus_device_reload_up(struct devlink *devlink, enum devlink_reload_action action, + enum devlink_reload_limit limit, u32 *actions_performed, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + int err; + + *actions_performed = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT) | + BIT(DEVLINK_RELOAD_ACTION_FW_ACTIVATE); + err = mlxsw_core_bus_device_register(mlxsw_core->bus_info, + mlxsw_core->bus, + mlxsw_core->bus_priv, true, + devlink, extack); + return err; +} + +static int mlxsw_devlink_flash_update(struct devlink *devlink, + struct devlink_flash_update_params *params, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + + return mlxsw_core_fw_flash_update(mlxsw_core, params, extack); +} + +static int mlxsw_devlink_trap_init(struct devlink *devlink, + const struct devlink_trap *trap, + void *trap_ctx) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->trap_init) + return -EOPNOTSUPP; + return mlxsw_driver->trap_init(mlxsw_core, trap, trap_ctx); +} + +static void mlxsw_devlink_trap_fini(struct devlink *devlink, + const struct devlink_trap *trap, + void *trap_ctx) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->trap_fini) + return; + mlxsw_driver->trap_fini(mlxsw_core, trap, trap_ctx); +} + +static int mlxsw_devlink_trap_action_set(struct devlink *devlink, + const struct devlink_trap *trap, + enum devlink_trap_action action, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->trap_action_set) + return -EOPNOTSUPP; + return mlxsw_driver->trap_action_set(mlxsw_core, trap, action, extack); +} + +static int +mlxsw_devlink_trap_group_init(struct devlink *devlink, + const struct devlink_trap_group *group) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->trap_group_init) + return -EOPNOTSUPP; + return mlxsw_driver->trap_group_init(mlxsw_core, group); +} + +static int +mlxsw_devlink_trap_group_set(struct devlink *devlink, + const struct devlink_trap_group *group, + const struct devlink_trap_policer *policer, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->trap_group_set) + return -EOPNOTSUPP; + return mlxsw_driver->trap_group_set(mlxsw_core, group, policer, extack); +} + +static int +mlxsw_devlink_trap_policer_init(struct devlink *devlink, + const struct devlink_trap_policer *policer) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->trap_policer_init) + return -EOPNOTSUPP; + return mlxsw_driver->trap_policer_init(mlxsw_core, policer); +} + +static void +mlxsw_devlink_trap_policer_fini(struct devlink *devlink, + const struct devlink_trap_policer *policer) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->trap_policer_fini) + return; + mlxsw_driver->trap_policer_fini(mlxsw_core, policer); +} + +static int +mlxsw_devlink_trap_policer_set(struct devlink *devlink, + const struct devlink_trap_policer *policer, + u64 rate, u64 burst, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->trap_policer_set) + return -EOPNOTSUPP; + return mlxsw_driver->trap_policer_set(mlxsw_core, policer, rate, burst, + extack); +} + +static int +mlxsw_devlink_trap_policer_counter_get(struct devlink *devlink, + const struct devlink_trap_policer *policer, + u64 *p_drops) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_driver *mlxsw_driver = mlxsw_core->driver; + + if (!mlxsw_driver->trap_policer_counter_get) + return -EOPNOTSUPP; + return mlxsw_driver->trap_policer_counter_get(mlxsw_core, policer, + p_drops); +} + +static const struct devlink_ops mlxsw_devlink_ops = { + .reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT) | + BIT(DEVLINK_RELOAD_ACTION_FW_ACTIVATE), + .reload_down = mlxsw_devlink_core_bus_device_reload_down, + .reload_up = mlxsw_devlink_core_bus_device_reload_up, + .port_split = mlxsw_devlink_port_split, + .port_unsplit = mlxsw_devlink_port_unsplit, + .sb_pool_get = mlxsw_devlink_sb_pool_get, + .sb_pool_set = mlxsw_devlink_sb_pool_set, + .sb_port_pool_get = mlxsw_devlink_sb_port_pool_get, + .sb_port_pool_set = mlxsw_devlink_sb_port_pool_set, + .sb_tc_pool_bind_get = mlxsw_devlink_sb_tc_pool_bind_get, + .sb_tc_pool_bind_set = mlxsw_devlink_sb_tc_pool_bind_set, + .sb_occ_snapshot = mlxsw_devlink_sb_occ_snapshot, + .sb_occ_max_clear = mlxsw_devlink_sb_occ_max_clear, + .sb_occ_port_pool_get = mlxsw_devlink_sb_occ_port_pool_get, + .sb_occ_tc_port_bind_get = mlxsw_devlink_sb_occ_tc_port_bind_get, + .info_get = mlxsw_devlink_info_get, + .flash_update = mlxsw_devlink_flash_update, + .trap_init = mlxsw_devlink_trap_init, + .trap_fini = mlxsw_devlink_trap_fini, + .trap_action_set = mlxsw_devlink_trap_action_set, + .trap_group_init = mlxsw_devlink_trap_group_init, + .trap_group_set = mlxsw_devlink_trap_group_set, + .trap_policer_init = mlxsw_devlink_trap_policer_init, + .trap_policer_fini = mlxsw_devlink_trap_policer_fini, + .trap_policer_set = mlxsw_devlink_trap_policer_set, + .trap_policer_counter_get = mlxsw_devlink_trap_policer_counter_get, +}; + +static int mlxsw_core_params_register(struct mlxsw_core *mlxsw_core) +{ + int err; + + err = mlxsw_core_fw_params_register(mlxsw_core); + if (err) + return err; + + if (mlxsw_core->driver->params_register) { + err = mlxsw_core->driver->params_register(mlxsw_core); + if (err) + goto err_params_register; + } + return 0; + +err_params_register: + mlxsw_core_fw_params_unregister(mlxsw_core); + return err; +} + +static void mlxsw_core_params_unregister(struct mlxsw_core *mlxsw_core) +{ + mlxsw_core_fw_params_unregister(mlxsw_core); + if (mlxsw_core->driver->params_register) + mlxsw_core->driver->params_unregister(mlxsw_core); +} + +struct mlxsw_core_health_event { + struct mlxsw_core *mlxsw_core; + char mfde_pl[MLXSW_REG_MFDE_LEN]; + struct work_struct work; +}; + +static void mlxsw_core_health_event_work(struct work_struct *work) +{ + struct mlxsw_core_health_event *event; + struct mlxsw_core *mlxsw_core; + + event = container_of(work, struct mlxsw_core_health_event, work); + mlxsw_core = event->mlxsw_core; + devlink_health_report(mlxsw_core->health.fw_fatal, "FW fatal event occurred", + event->mfde_pl); + kfree(event); +} + +static void mlxsw_core_health_listener_func(const struct mlxsw_reg_info *reg, + char *mfde_pl, void *priv) +{ + struct mlxsw_core_health_event *event; + struct mlxsw_core *mlxsw_core = priv; + + event = kmalloc(sizeof(*event), GFP_ATOMIC); + if (!event) + return; + event->mlxsw_core = mlxsw_core; + memcpy(event->mfde_pl, mfde_pl, sizeof(event->mfde_pl)); + INIT_WORK(&event->work, mlxsw_core_health_event_work); + mlxsw_core_schedule_work(&event->work); +} + +static const struct mlxsw_listener mlxsw_core_health_listener = + MLXSW_CORE_EVENTL(mlxsw_core_health_listener_func, MFDE); + +static int +mlxsw_core_health_fw_fatal_dump_fatal_cause(const char *mfde_pl, + struct devlink_fmsg *fmsg) +{ + u32 val, tile_v; + int err; + + val = mlxsw_reg_mfde_fatal_cause_id_get(mfde_pl); + err = devlink_fmsg_u32_pair_put(fmsg, "cause_id", val); + if (err) + return err; + tile_v = mlxsw_reg_mfde_fatal_cause_tile_v_get(mfde_pl); + if (tile_v) { + val = mlxsw_reg_mfde_fatal_cause_tile_index_get(mfde_pl); + err = devlink_fmsg_u8_pair_put(fmsg, "tile_index", val); + if (err) + return err; + } + + return 0; +} + +static int +mlxsw_core_health_fw_fatal_dump_fw_assert(const char *mfde_pl, + struct devlink_fmsg *fmsg) +{ + u32 val, tile_v; + int err; + + val = mlxsw_reg_mfde_fw_assert_var0_get(mfde_pl); + err = devlink_fmsg_u32_pair_put(fmsg, "var0", val); + if (err) + return err; + val = mlxsw_reg_mfde_fw_assert_var1_get(mfde_pl); + err = devlink_fmsg_u32_pair_put(fmsg, "var1", val); + if (err) + return err; + val = mlxsw_reg_mfde_fw_assert_var2_get(mfde_pl); + err = devlink_fmsg_u32_pair_put(fmsg, "var2", val); + if (err) + return err; + val = mlxsw_reg_mfde_fw_assert_var3_get(mfde_pl); + err = devlink_fmsg_u32_pair_put(fmsg, "var3", val); + if (err) + return err; + val = mlxsw_reg_mfde_fw_assert_var4_get(mfde_pl); + err = devlink_fmsg_u32_pair_put(fmsg, "var4", val); + if (err) + return err; + val = mlxsw_reg_mfde_fw_assert_existptr_get(mfde_pl); + err = devlink_fmsg_u32_pair_put(fmsg, "existptr", val); + if (err) + return err; + val = mlxsw_reg_mfde_fw_assert_callra_get(mfde_pl); + err = devlink_fmsg_u32_pair_put(fmsg, "callra", val); + if (err) + return err; + val = mlxsw_reg_mfde_fw_assert_oe_get(mfde_pl); + err = devlink_fmsg_bool_pair_put(fmsg, "old_event", val); + if (err) + return err; + tile_v = mlxsw_reg_mfde_fw_assert_tile_v_get(mfde_pl); + if (tile_v) { + val = mlxsw_reg_mfde_fw_assert_tile_index_get(mfde_pl); + err = devlink_fmsg_u8_pair_put(fmsg, "tile_index", val); + if (err) + return err; + } + val = mlxsw_reg_mfde_fw_assert_ext_synd_get(mfde_pl); + err = devlink_fmsg_u32_pair_put(fmsg, "ext_synd", val); + if (err) + return err; + + return 0; +} + +static int +mlxsw_core_health_fw_fatal_dump_kvd_im_stop(const char *mfde_pl, + struct devlink_fmsg *fmsg) +{ + u32 val; + int err; + + val = mlxsw_reg_mfde_kvd_im_stop_oe_get(mfde_pl); + err = devlink_fmsg_bool_pair_put(fmsg, "old_event", val); + if (err) + return err; + val = mlxsw_reg_mfde_kvd_im_stop_pipes_mask_get(mfde_pl); + return devlink_fmsg_u32_pair_put(fmsg, "pipes_mask", val); +} + +static int +mlxsw_core_health_fw_fatal_dump_crspace_to(const char *mfde_pl, + struct devlink_fmsg *fmsg) +{ + u32 val; + int err; + + val = mlxsw_reg_mfde_crspace_to_log_address_get(mfde_pl); + err = devlink_fmsg_u32_pair_put(fmsg, "log_address", val); + if (err) + return err; + val = mlxsw_reg_mfde_crspace_to_oe_get(mfde_pl); + err = devlink_fmsg_bool_pair_put(fmsg, "old_event", val); + if (err) + return err; + val = mlxsw_reg_mfde_crspace_to_log_id_get(mfde_pl); + err = devlink_fmsg_u8_pair_put(fmsg, "log_irisc_id", val); + if (err) + return err; + val = mlxsw_reg_mfde_crspace_to_log_ip_get(mfde_pl); + err = devlink_fmsg_u64_pair_put(fmsg, "log_ip", val); + if (err) + return err; + + return 0; +} + +static int mlxsw_core_health_fw_fatal_dump(struct devlink_health_reporter *reporter, + struct devlink_fmsg *fmsg, void *priv_ctx, + struct netlink_ext_ack *extack) +{ + char *mfde_pl = priv_ctx; + char *val_str; + u8 event_id; + u32 val; + int err; + + if (!priv_ctx) + /* User-triggered dumps are not possible */ + return -EOPNOTSUPP; + + val = mlxsw_reg_mfde_irisc_id_get(mfde_pl); + err = devlink_fmsg_u8_pair_put(fmsg, "irisc_id", val); + if (err) + return err; + err = devlink_fmsg_arr_pair_nest_start(fmsg, "event"); + if (err) + return err; + + event_id = mlxsw_reg_mfde_event_id_get(mfde_pl); + err = devlink_fmsg_u32_pair_put(fmsg, "id", event_id); + if (err) + return err; + switch (event_id) { + case MLXSW_REG_MFDE_EVENT_ID_CRSPACE_TO: + val_str = "CR space timeout"; + break; + case MLXSW_REG_MFDE_EVENT_ID_KVD_IM_STOP: + val_str = "KVD insertion machine stopped"; + break; + case MLXSW_REG_MFDE_EVENT_ID_TEST: + val_str = "Test"; + break; + case MLXSW_REG_MFDE_EVENT_ID_FW_ASSERT: + val_str = "FW assert"; + break; + case MLXSW_REG_MFDE_EVENT_ID_FATAL_CAUSE: + val_str = "Fatal cause"; + break; + default: + val_str = NULL; + } + if (val_str) { + err = devlink_fmsg_string_pair_put(fmsg, "desc", val_str); + if (err) + return err; + } + + err = devlink_fmsg_arr_pair_nest_end(fmsg); + if (err) + return err; + + err = devlink_fmsg_arr_pair_nest_start(fmsg, "severity"); + if (err) + return err; + + val = mlxsw_reg_mfde_severity_get(mfde_pl); + err = devlink_fmsg_u8_pair_put(fmsg, "id", val); + if (err) + return err; + switch (val) { + case MLXSW_REG_MFDE_SEVERITY_FATL: + val_str = "Fatal"; + break; + case MLXSW_REG_MFDE_SEVERITY_NRML: + val_str = "Normal"; + break; + case MLXSW_REG_MFDE_SEVERITY_INTR: + val_str = "Debug"; + break; + default: + val_str = NULL; + } + if (val_str) { + err = devlink_fmsg_string_pair_put(fmsg, "desc", val_str); + if (err) + return err; + } + + err = devlink_fmsg_arr_pair_nest_end(fmsg); + if (err) + return err; + + val = mlxsw_reg_mfde_method_get(mfde_pl); + switch (val) { + case MLXSW_REG_MFDE_METHOD_QUERY: + val_str = "query"; + break; + case MLXSW_REG_MFDE_METHOD_WRITE: + val_str = "write"; + break; + default: + val_str = NULL; + } + if (val_str) { + err = devlink_fmsg_string_pair_put(fmsg, "method", val_str); + if (err) + return err; + } + + val = mlxsw_reg_mfde_long_process_get(mfde_pl); + err = devlink_fmsg_bool_pair_put(fmsg, "long_process", val); + if (err) + return err; + + val = mlxsw_reg_mfde_command_type_get(mfde_pl); + switch (val) { + case MLXSW_REG_MFDE_COMMAND_TYPE_MAD: + val_str = "mad"; + break; + case MLXSW_REG_MFDE_COMMAND_TYPE_EMAD: + val_str = "emad"; + break; + case MLXSW_REG_MFDE_COMMAND_TYPE_CMDIF: + val_str = "cmdif"; + break; + default: + val_str = NULL; + } + if (val_str) { + err = devlink_fmsg_string_pair_put(fmsg, "command_type", val_str); + if (err) + return err; + } + + val = mlxsw_reg_mfde_reg_attr_id_get(mfde_pl); + err = devlink_fmsg_u32_pair_put(fmsg, "reg_attr_id", val); + if (err) + return err; + + switch (event_id) { + case MLXSW_REG_MFDE_EVENT_ID_CRSPACE_TO: + return mlxsw_core_health_fw_fatal_dump_crspace_to(mfde_pl, + fmsg); + case MLXSW_REG_MFDE_EVENT_ID_KVD_IM_STOP: + return mlxsw_core_health_fw_fatal_dump_kvd_im_stop(mfde_pl, + fmsg); + case MLXSW_REG_MFDE_EVENT_ID_FW_ASSERT: + return mlxsw_core_health_fw_fatal_dump_fw_assert(mfde_pl, fmsg); + case MLXSW_REG_MFDE_EVENT_ID_FATAL_CAUSE: + return mlxsw_core_health_fw_fatal_dump_fatal_cause(mfde_pl, + fmsg); + } + + return 0; +} + +static int +mlxsw_core_health_fw_fatal_test(struct devlink_health_reporter *reporter, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_health_reporter_priv(reporter); + char mfgd_pl[MLXSW_REG_MFGD_LEN]; + int err; + + /* Read the register first to make sure no other bits are changed. */ + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mfgd), mfgd_pl); + if (err) + return err; + mlxsw_reg_mfgd_trigger_test_set(mfgd_pl, true); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mfgd), mfgd_pl); +} + +static const struct devlink_health_reporter_ops +mlxsw_core_health_fw_fatal_ops = { + .name = "fw_fatal", + .dump = mlxsw_core_health_fw_fatal_dump, + .test = mlxsw_core_health_fw_fatal_test, +}; + +static int mlxsw_core_health_fw_fatal_config(struct mlxsw_core *mlxsw_core, + bool enable) +{ + char mfgd_pl[MLXSW_REG_MFGD_LEN]; + int err; + + /* Read the register first to make sure no other bits are changed. */ + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mfgd), mfgd_pl); + if (err) + return err; + mlxsw_reg_mfgd_fatal_event_mode_set(mfgd_pl, enable); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mfgd), mfgd_pl); +} + +static int mlxsw_core_health_init(struct mlxsw_core *mlxsw_core) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + struct devlink_health_reporter *fw_fatal; + int err; + + if (!(mlxsw_core->bus->features & MLXSW_BUS_F_TXRX)) + return 0; + + fw_fatal = devlink_health_reporter_create(devlink, &mlxsw_core_health_fw_fatal_ops, + 0, mlxsw_core); + if (IS_ERR(fw_fatal)) { + dev_err(mlxsw_core->bus_info->dev, "Failed to create fw fatal reporter"); + return PTR_ERR(fw_fatal); + } + mlxsw_core->health.fw_fatal = fw_fatal; + + err = mlxsw_core_trap_register(mlxsw_core, &mlxsw_core_health_listener, mlxsw_core); + if (err) + goto err_trap_register; + + err = mlxsw_core_health_fw_fatal_config(mlxsw_core, true); + if (err) + goto err_fw_fatal_config; + + return 0; + +err_fw_fatal_config: + mlxsw_core_trap_unregister(mlxsw_core, &mlxsw_core_health_listener, mlxsw_core); +err_trap_register: + devlink_health_reporter_destroy(mlxsw_core->health.fw_fatal); + return err; +} + +static void mlxsw_core_health_fini(struct mlxsw_core *mlxsw_core) +{ + if (!(mlxsw_core->bus->features & MLXSW_BUS_F_TXRX)) + return; + + mlxsw_core_health_fw_fatal_config(mlxsw_core, false); + mlxsw_core_trap_unregister(mlxsw_core, &mlxsw_core_health_listener, mlxsw_core); + /* Make sure there is no more event work scheduled */ + mlxsw_core_flush_owq(); + devlink_health_reporter_destroy(mlxsw_core->health.fw_fatal); +} + +static void mlxsw_core_irq_event_handler_init(struct mlxsw_core *mlxsw_core) +{ + INIT_LIST_HEAD(&mlxsw_core->irq_event_handler_list); + mutex_init(&mlxsw_core->irq_event_handler_lock); +} + +static void mlxsw_core_irq_event_handler_fini(struct mlxsw_core *mlxsw_core) +{ + mutex_destroy(&mlxsw_core->irq_event_handler_lock); + WARN_ON(!list_empty(&mlxsw_core->irq_event_handler_list)); +} + +static int +__mlxsw_core_bus_device_register(const struct mlxsw_bus_info *mlxsw_bus_info, + const struct mlxsw_bus *mlxsw_bus, + void *bus_priv, bool reload, + struct devlink *devlink, + struct netlink_ext_ack *extack) +{ + const char *device_kind = mlxsw_bus_info->device_kind; + struct mlxsw_core *mlxsw_core; + struct mlxsw_driver *mlxsw_driver; + size_t alloc_size; + u16 max_lag; + int err; + + mlxsw_driver = mlxsw_core_driver_get(device_kind); + if (!mlxsw_driver) + return -EINVAL; + + if (!reload) { + alloc_size = sizeof(*mlxsw_core) + mlxsw_driver->priv_size; + devlink = devlink_alloc(&mlxsw_devlink_ops, alloc_size, + mlxsw_bus_info->dev); + if (!devlink) { + err = -ENOMEM; + goto err_devlink_alloc; + } + devl_lock(devlink); + } + + mlxsw_core = devlink_priv(devlink); + INIT_LIST_HEAD(&mlxsw_core->rx_listener_list); + INIT_LIST_HEAD(&mlxsw_core->event_listener_list); + mlxsw_core->driver = mlxsw_driver; + mlxsw_core->bus = mlxsw_bus; + mlxsw_core->bus_priv = bus_priv; + mlxsw_core->bus_info = mlxsw_bus_info; + mlxsw_core_irq_event_handler_init(mlxsw_core); + + err = mlxsw_bus->init(bus_priv, mlxsw_core, mlxsw_driver->profile, + &mlxsw_core->res); + if (err) + goto err_bus_init; + + if (mlxsw_driver->resources_register && !reload) { + err = mlxsw_driver->resources_register(mlxsw_core); + if (err) + goto err_register_resources; + } + + err = mlxsw_ports_init(mlxsw_core, reload); + if (err) + goto err_ports_init; + + err = mlxsw_core_max_lag(mlxsw_core, &max_lag); + if (!err && MLXSW_CORE_RES_VALID(mlxsw_core, MAX_LAG_MEMBERS)) { + alloc_size = sizeof(*mlxsw_core->lag.mapping) * max_lag * + MLXSW_CORE_RES_GET(mlxsw_core, MAX_LAG_MEMBERS); + mlxsw_core->lag.mapping = kzalloc(alloc_size, GFP_KERNEL); + if (!mlxsw_core->lag.mapping) { + err = -ENOMEM; + goto err_alloc_lag_mapping; + } + } + + err = mlxsw_core_trap_groups_set(mlxsw_core); + if (err) + goto err_trap_groups_set; + + err = mlxsw_emad_init(mlxsw_core); + if (err) + goto err_emad_init; + + if (!reload) { + err = mlxsw_core_params_register(mlxsw_core); + if (err) + goto err_register_params; + } + + err = mlxsw_core_fw_rev_validate(mlxsw_core, mlxsw_bus_info, mlxsw_driver->fw_req_rev, + mlxsw_driver->fw_filename); + if (err) + goto err_fw_rev_validate; + + err = mlxsw_linecards_init(mlxsw_core, mlxsw_bus_info); + if (err) + goto err_linecards_init; + + err = mlxsw_core_health_init(mlxsw_core); + if (err) + goto err_health_init; + + err = mlxsw_hwmon_init(mlxsw_core, mlxsw_bus_info, &mlxsw_core->hwmon); + if (err) + goto err_hwmon_init; + + err = mlxsw_thermal_init(mlxsw_core, mlxsw_bus_info, + &mlxsw_core->thermal); + if (err) + goto err_thermal_init; + + err = mlxsw_env_init(mlxsw_core, mlxsw_bus_info, &mlxsw_core->env); + if (err) + goto err_env_init; + + if (mlxsw_driver->init) { + err = mlxsw_driver->init(mlxsw_core, mlxsw_bus_info, extack); + if (err) + goto err_driver_init; + } + + if (!reload) { + devlink_set_features(devlink, DEVLINK_F_RELOAD); + devl_unlock(devlink); + devlink_register(devlink); + } + return 0; + +err_driver_init: + mlxsw_env_fini(mlxsw_core->env); +err_env_init: + mlxsw_thermal_fini(mlxsw_core->thermal); +err_thermal_init: + mlxsw_hwmon_fini(mlxsw_core->hwmon); +err_hwmon_init: + mlxsw_core_health_fini(mlxsw_core); +err_health_init: + mlxsw_linecards_fini(mlxsw_core); +err_linecards_init: +err_fw_rev_validate: + if (!reload) + mlxsw_core_params_unregister(mlxsw_core); +err_register_params: + mlxsw_emad_fini(mlxsw_core); +err_emad_init: +err_trap_groups_set: + kfree(mlxsw_core->lag.mapping); +err_alloc_lag_mapping: + mlxsw_ports_fini(mlxsw_core, reload); +err_ports_init: + if (!reload) + devl_resources_unregister(devlink); +err_register_resources: + mlxsw_bus->fini(bus_priv); +err_bus_init: + mlxsw_core_irq_event_handler_fini(mlxsw_core); + if (!reload) { + devl_unlock(devlink); + devlink_free(devlink); + } +err_devlink_alloc: + return err; +} + +int mlxsw_core_bus_device_register(const struct mlxsw_bus_info *mlxsw_bus_info, + const struct mlxsw_bus *mlxsw_bus, + void *bus_priv, bool reload, + struct devlink *devlink, + struct netlink_ext_ack *extack) +{ + bool called_again = false; + int err; + +again: + err = __mlxsw_core_bus_device_register(mlxsw_bus_info, mlxsw_bus, + bus_priv, reload, + devlink, extack); + /* -EAGAIN is returned in case the FW was updated. FW needs + * a reset, so lets try to call __mlxsw_core_bus_device_register() + * again. + */ + if (err == -EAGAIN && !called_again) { + called_again = true; + goto again; + } + + return err; +} +EXPORT_SYMBOL(mlxsw_core_bus_device_register); + +void mlxsw_core_bus_device_unregister(struct mlxsw_core *mlxsw_core, + bool reload) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + + if (!reload) { + devlink_unregister(devlink); + devl_lock(devlink); + } + + if (devlink_is_reload_failed(devlink)) { + if (!reload) + /* Only the parts that were not de-initialized in the + * failed reload attempt need to be de-initialized. + */ + goto reload_fail_deinit; + else + return; + } + + if (mlxsw_core->driver->fini) + mlxsw_core->driver->fini(mlxsw_core); + mlxsw_env_fini(mlxsw_core->env); + mlxsw_thermal_fini(mlxsw_core->thermal); + mlxsw_hwmon_fini(mlxsw_core->hwmon); + mlxsw_core_health_fini(mlxsw_core); + mlxsw_linecards_fini(mlxsw_core); + if (!reload) + mlxsw_core_params_unregister(mlxsw_core); + mlxsw_emad_fini(mlxsw_core); + kfree(mlxsw_core->lag.mapping); + mlxsw_ports_fini(mlxsw_core, reload); + if (!reload) + devl_resources_unregister(devlink); + mlxsw_core->bus->fini(mlxsw_core->bus_priv); + mlxsw_core_irq_event_handler_fini(mlxsw_core); + if (!reload) { + devl_unlock(devlink); + devlink_free(devlink); + } + + return; + +reload_fail_deinit: + mlxsw_core_params_unregister(mlxsw_core); + devl_resources_unregister(devlink); + devl_unlock(devlink); + devlink_free(devlink); +} +EXPORT_SYMBOL(mlxsw_core_bus_device_unregister); + +bool mlxsw_core_skb_transmit_busy(struct mlxsw_core *mlxsw_core, + const struct mlxsw_tx_info *tx_info) +{ + return mlxsw_core->bus->skb_transmit_busy(mlxsw_core->bus_priv, + tx_info); +} +EXPORT_SYMBOL(mlxsw_core_skb_transmit_busy); + +int mlxsw_core_skb_transmit(struct mlxsw_core *mlxsw_core, struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info) +{ + return mlxsw_core->bus->skb_transmit(mlxsw_core->bus_priv, skb, + tx_info); +} +EXPORT_SYMBOL(mlxsw_core_skb_transmit); + +void mlxsw_core_ptp_transmitted(struct mlxsw_core *mlxsw_core, + struct sk_buff *skb, u16 local_port) +{ + if (mlxsw_core->driver->ptp_transmitted) + mlxsw_core->driver->ptp_transmitted(mlxsw_core, skb, + local_port); +} +EXPORT_SYMBOL(mlxsw_core_ptp_transmitted); + +static bool __is_rx_listener_equal(const struct mlxsw_rx_listener *rxl_a, + const struct mlxsw_rx_listener *rxl_b) +{ + return (rxl_a->func == rxl_b->func && + rxl_a->local_port == rxl_b->local_port && + rxl_a->trap_id == rxl_b->trap_id && + rxl_a->mirror_reason == rxl_b->mirror_reason); +} + +static struct mlxsw_rx_listener_item * +__find_rx_listener_item(struct mlxsw_core *mlxsw_core, + const struct mlxsw_rx_listener *rxl) +{ + struct mlxsw_rx_listener_item *rxl_item; + + list_for_each_entry(rxl_item, &mlxsw_core->rx_listener_list, list) { + if (__is_rx_listener_equal(&rxl_item->rxl, rxl)) + return rxl_item; + } + return NULL; +} + +int mlxsw_core_rx_listener_register(struct mlxsw_core *mlxsw_core, + const struct mlxsw_rx_listener *rxl, + void *priv, bool enabled) +{ + struct mlxsw_rx_listener_item *rxl_item; + + rxl_item = __find_rx_listener_item(mlxsw_core, rxl); + if (rxl_item) + return -EEXIST; + rxl_item = kmalloc(sizeof(*rxl_item), GFP_KERNEL); + if (!rxl_item) + return -ENOMEM; + rxl_item->rxl = *rxl; + rxl_item->priv = priv; + rxl_item->enabled = enabled; + + list_add_rcu(&rxl_item->list, &mlxsw_core->rx_listener_list); + return 0; +} +EXPORT_SYMBOL(mlxsw_core_rx_listener_register); + +void mlxsw_core_rx_listener_unregister(struct mlxsw_core *mlxsw_core, + const struct mlxsw_rx_listener *rxl) +{ + struct mlxsw_rx_listener_item *rxl_item; + + rxl_item = __find_rx_listener_item(mlxsw_core, rxl); + if (!rxl_item) + return; + list_del_rcu(&rxl_item->list); + synchronize_rcu(); + kfree(rxl_item); +} +EXPORT_SYMBOL(mlxsw_core_rx_listener_unregister); + +static void +mlxsw_core_rx_listener_state_set(struct mlxsw_core *mlxsw_core, + const struct mlxsw_rx_listener *rxl, + bool enabled) +{ + struct mlxsw_rx_listener_item *rxl_item; + + rxl_item = __find_rx_listener_item(mlxsw_core, rxl); + if (WARN_ON(!rxl_item)) + return; + rxl_item->enabled = enabled; +} + +static void mlxsw_core_event_listener_func(struct sk_buff *skb, u16 local_port, + void *priv) +{ + struct mlxsw_event_listener_item *event_listener_item = priv; + struct mlxsw_core *mlxsw_core; + struct mlxsw_reg_info reg; + char *payload; + char *reg_tlv; + char *op_tlv; + + mlxsw_core = event_listener_item->mlxsw_core; + trace_devlink_hwmsg(priv_to_devlink(mlxsw_core), true, 0, + skb->data, skb->len); + + mlxsw_emad_tlv_parse(skb); + op_tlv = mlxsw_emad_op_tlv(skb); + reg_tlv = mlxsw_emad_reg_tlv(skb); + + reg.id = mlxsw_emad_op_tlv_register_id_get(op_tlv); + reg.len = (mlxsw_emad_reg_tlv_len_get(reg_tlv) - 1) * sizeof(u32); + payload = mlxsw_emad_reg_payload(reg_tlv); + event_listener_item->el.func(®, payload, event_listener_item->priv); + dev_kfree_skb(skb); +} + +static bool __is_event_listener_equal(const struct mlxsw_event_listener *el_a, + const struct mlxsw_event_listener *el_b) +{ + return (el_a->func == el_b->func && + el_a->trap_id == el_b->trap_id); +} + +static struct mlxsw_event_listener_item * +__find_event_listener_item(struct mlxsw_core *mlxsw_core, + const struct mlxsw_event_listener *el) +{ + struct mlxsw_event_listener_item *el_item; + + list_for_each_entry(el_item, &mlxsw_core->event_listener_list, list) { + if (__is_event_listener_equal(&el_item->el, el)) + return el_item; + } + return NULL; +} + +int mlxsw_core_event_listener_register(struct mlxsw_core *mlxsw_core, + const struct mlxsw_event_listener *el, + void *priv) +{ + int err; + struct mlxsw_event_listener_item *el_item; + const struct mlxsw_rx_listener rxl = { + .func = mlxsw_core_event_listener_func, + .local_port = MLXSW_PORT_DONT_CARE, + .trap_id = el->trap_id, + }; + + el_item = __find_event_listener_item(mlxsw_core, el); + if (el_item) + return -EEXIST; + el_item = kmalloc(sizeof(*el_item), GFP_KERNEL); + if (!el_item) + return -ENOMEM; + el_item->mlxsw_core = mlxsw_core; + el_item->el = *el; + el_item->priv = priv; + + err = mlxsw_core_rx_listener_register(mlxsw_core, &rxl, el_item, true); + if (err) + goto err_rx_listener_register; + + /* No reason to save item if we did not manage to register an RX + * listener for it. + */ + list_add_rcu(&el_item->list, &mlxsw_core->event_listener_list); + + return 0; + +err_rx_listener_register: + kfree(el_item); + return err; +} +EXPORT_SYMBOL(mlxsw_core_event_listener_register); + +void mlxsw_core_event_listener_unregister(struct mlxsw_core *mlxsw_core, + const struct mlxsw_event_listener *el) +{ + struct mlxsw_event_listener_item *el_item; + const struct mlxsw_rx_listener rxl = { + .func = mlxsw_core_event_listener_func, + .local_port = MLXSW_PORT_DONT_CARE, + .trap_id = el->trap_id, + }; + + el_item = __find_event_listener_item(mlxsw_core, el); + if (!el_item) + return; + mlxsw_core_rx_listener_unregister(mlxsw_core, &rxl); + list_del(&el_item->list); + kfree(el_item); +} +EXPORT_SYMBOL(mlxsw_core_event_listener_unregister); + +static int mlxsw_core_listener_register(struct mlxsw_core *mlxsw_core, + const struct mlxsw_listener *listener, + void *priv, bool enabled) +{ + if (listener->is_event) { + WARN_ON(!enabled); + return mlxsw_core_event_listener_register(mlxsw_core, + &listener->event_listener, + priv); + } else { + return mlxsw_core_rx_listener_register(mlxsw_core, + &listener->rx_listener, + priv, enabled); + } +} + +static void mlxsw_core_listener_unregister(struct mlxsw_core *mlxsw_core, + const struct mlxsw_listener *listener, + void *priv) +{ + if (listener->is_event) + mlxsw_core_event_listener_unregister(mlxsw_core, + &listener->event_listener); + else + mlxsw_core_rx_listener_unregister(mlxsw_core, + &listener->rx_listener); +} + +int mlxsw_core_trap_register(struct mlxsw_core *mlxsw_core, + const struct mlxsw_listener *listener, void *priv) +{ + enum mlxsw_reg_htgt_trap_group trap_group; + enum mlxsw_reg_hpkt_action action; + char hpkt_pl[MLXSW_REG_HPKT_LEN]; + int err; + + if (!(mlxsw_core->bus->features & MLXSW_BUS_F_TXRX)) + return 0; + + err = mlxsw_core_listener_register(mlxsw_core, listener, priv, + listener->enabled_on_register); + if (err) + return err; + + action = listener->enabled_on_register ? listener->en_action : + listener->dis_action; + trap_group = listener->enabled_on_register ? listener->en_trap_group : + listener->dis_trap_group; + mlxsw_reg_hpkt_pack(hpkt_pl, action, listener->trap_id, + trap_group, listener->is_ctrl); + err = mlxsw_reg_write(mlxsw_core, MLXSW_REG(hpkt), hpkt_pl); + if (err) + goto err_trap_set; + + return 0; + +err_trap_set: + mlxsw_core_listener_unregister(mlxsw_core, listener, priv); + return err; +} +EXPORT_SYMBOL(mlxsw_core_trap_register); + +void mlxsw_core_trap_unregister(struct mlxsw_core *mlxsw_core, + const struct mlxsw_listener *listener, + void *priv) +{ + char hpkt_pl[MLXSW_REG_HPKT_LEN]; + + if (!(mlxsw_core->bus->features & MLXSW_BUS_F_TXRX)) + return; + + if (!listener->is_event) { + mlxsw_reg_hpkt_pack(hpkt_pl, listener->dis_action, + listener->trap_id, listener->dis_trap_group, + listener->is_ctrl); + mlxsw_reg_write(mlxsw_core, MLXSW_REG(hpkt), hpkt_pl); + } + + mlxsw_core_listener_unregister(mlxsw_core, listener, priv); +} +EXPORT_SYMBOL(mlxsw_core_trap_unregister); + +int mlxsw_core_traps_register(struct mlxsw_core *mlxsw_core, + const struct mlxsw_listener *listeners, + size_t listeners_count, void *priv) +{ + int i, err; + + for (i = 0; i < listeners_count; i++) { + err = mlxsw_core_trap_register(mlxsw_core, + &listeners[i], + priv); + if (err) + goto err_listener_register; + } + return 0; + +err_listener_register: + for (i--; i >= 0; i--) { + mlxsw_core_trap_unregister(mlxsw_core, + &listeners[i], + priv); + } + return err; +} +EXPORT_SYMBOL(mlxsw_core_traps_register); + +void mlxsw_core_traps_unregister(struct mlxsw_core *mlxsw_core, + const struct mlxsw_listener *listeners, + size_t listeners_count, void *priv) +{ + int i; + + for (i = 0; i < listeners_count; i++) { + mlxsw_core_trap_unregister(mlxsw_core, + &listeners[i], + priv); + } +} +EXPORT_SYMBOL(mlxsw_core_traps_unregister); + +int mlxsw_core_trap_state_set(struct mlxsw_core *mlxsw_core, + const struct mlxsw_listener *listener, + bool enabled) +{ + enum mlxsw_reg_htgt_trap_group trap_group; + enum mlxsw_reg_hpkt_action action; + char hpkt_pl[MLXSW_REG_HPKT_LEN]; + int err; + + /* Not supported for event listener */ + if (WARN_ON(listener->is_event)) + return -EINVAL; + + action = enabled ? listener->en_action : listener->dis_action; + trap_group = enabled ? listener->en_trap_group : + listener->dis_trap_group; + mlxsw_reg_hpkt_pack(hpkt_pl, action, listener->trap_id, + trap_group, listener->is_ctrl); + err = mlxsw_reg_write(mlxsw_core, MLXSW_REG(hpkt), hpkt_pl); + if (err) + return err; + + mlxsw_core_rx_listener_state_set(mlxsw_core, &listener->rx_listener, + enabled); + return 0; +} +EXPORT_SYMBOL(mlxsw_core_trap_state_set); + +static u64 mlxsw_core_tid_get(struct mlxsw_core *mlxsw_core) +{ + return atomic64_inc_return(&mlxsw_core->emad.tid); +} + +static int mlxsw_core_reg_access_emad(struct mlxsw_core *mlxsw_core, + const struct mlxsw_reg_info *reg, + char *payload, + enum mlxsw_core_reg_access_type type, + struct list_head *bulk_list, + mlxsw_reg_trans_cb_t *cb, + unsigned long cb_priv) +{ + u64 tid = mlxsw_core_tid_get(mlxsw_core); + struct mlxsw_reg_trans *trans; + int err; + + trans = kzalloc(sizeof(*trans), GFP_KERNEL); + if (!trans) + return -ENOMEM; + + err = mlxsw_emad_reg_access(mlxsw_core, reg, payload, type, trans, + bulk_list, cb, cb_priv, tid); + if (err) { + kfree_rcu(trans, rcu); + return err; + } + return 0; +} + +int mlxsw_reg_trans_query(struct mlxsw_core *mlxsw_core, + const struct mlxsw_reg_info *reg, char *payload, + struct list_head *bulk_list, + mlxsw_reg_trans_cb_t *cb, unsigned long cb_priv) +{ + return mlxsw_core_reg_access_emad(mlxsw_core, reg, payload, + MLXSW_CORE_REG_ACCESS_TYPE_QUERY, + bulk_list, cb, cb_priv); +} +EXPORT_SYMBOL(mlxsw_reg_trans_query); + +int mlxsw_reg_trans_write(struct mlxsw_core *mlxsw_core, + const struct mlxsw_reg_info *reg, char *payload, + struct list_head *bulk_list, + mlxsw_reg_trans_cb_t *cb, unsigned long cb_priv) +{ + return mlxsw_core_reg_access_emad(mlxsw_core, reg, payload, + MLXSW_CORE_REG_ACCESS_TYPE_WRITE, + bulk_list, cb, cb_priv); +} +EXPORT_SYMBOL(mlxsw_reg_trans_write); + +#define MLXSW_REG_TRANS_ERR_STRING_SIZE 256 + +static int mlxsw_reg_trans_wait(struct mlxsw_reg_trans *trans) +{ + char err_string[MLXSW_REG_TRANS_ERR_STRING_SIZE]; + struct mlxsw_core *mlxsw_core = trans->core; + int err; + + wait_for_completion(&trans->completion); + cancel_delayed_work_sync(&trans->timeout_dw); + err = trans->err; + + if (trans->retries) + dev_warn(mlxsw_core->bus_info->dev, "EMAD retries (%d/%d) (tid=%llx)\n", + trans->retries, MLXSW_EMAD_MAX_RETRY, trans->tid); + if (err) { + dev_err(mlxsw_core->bus_info->dev, "EMAD reg access failed (tid=%llx,reg_id=%x(%s),type=%s,status=%x(%s))\n", + trans->tid, trans->reg->id, + mlxsw_reg_id_str(trans->reg->id), + mlxsw_core_reg_access_type_str(trans->type), + trans->emad_status, + mlxsw_emad_op_tlv_status_str(trans->emad_status)); + + snprintf(err_string, MLXSW_REG_TRANS_ERR_STRING_SIZE, + "(tid=%llx,reg_id=%x(%s)) %s (%s)\n", trans->tid, + trans->reg->id, mlxsw_reg_id_str(trans->reg->id), + mlxsw_emad_op_tlv_status_str(trans->emad_status), + trans->emad_err_string ? trans->emad_err_string : ""); + + trace_devlink_hwerr(priv_to_devlink(mlxsw_core), + trans->emad_status, err_string); + + kfree(trans->emad_err_string); + } + + list_del(&trans->bulk_list); + kfree_rcu(trans, rcu); + return err; +} + +int mlxsw_reg_trans_bulk_wait(struct list_head *bulk_list) +{ + struct mlxsw_reg_trans *trans; + struct mlxsw_reg_trans *tmp; + int sum_err = 0; + int err; + + list_for_each_entry_safe(trans, tmp, bulk_list, bulk_list) { + err = mlxsw_reg_trans_wait(trans); + if (err && sum_err == 0) + sum_err = err; /* first error to be returned */ + } + return sum_err; +} +EXPORT_SYMBOL(mlxsw_reg_trans_bulk_wait); + +struct mlxsw_core_irq_event_handler_item { + struct list_head list; + void (*cb)(struct mlxsw_core *mlxsw_core); +}; + +int mlxsw_core_irq_event_handler_register(struct mlxsw_core *mlxsw_core, + mlxsw_irq_event_cb_t cb) +{ + struct mlxsw_core_irq_event_handler_item *item; + + item = kzalloc(sizeof(*item), GFP_KERNEL); + if (!item) + return -ENOMEM; + item->cb = cb; + mutex_lock(&mlxsw_core->irq_event_handler_lock); + list_add_tail(&item->list, &mlxsw_core->irq_event_handler_list); + mutex_unlock(&mlxsw_core->irq_event_handler_lock); + return 0; +} +EXPORT_SYMBOL(mlxsw_core_irq_event_handler_register); + +void mlxsw_core_irq_event_handler_unregister(struct mlxsw_core *mlxsw_core, + mlxsw_irq_event_cb_t cb) +{ + struct mlxsw_core_irq_event_handler_item *item, *tmp; + + mutex_lock(&mlxsw_core->irq_event_handler_lock); + list_for_each_entry_safe(item, tmp, + &mlxsw_core->irq_event_handler_list, list) { + if (item->cb == cb) { + list_del(&item->list); + kfree(item); + } + } + mutex_unlock(&mlxsw_core->irq_event_handler_lock); +} +EXPORT_SYMBOL(mlxsw_core_irq_event_handler_unregister); + +void mlxsw_core_irq_event_handlers_call(struct mlxsw_core *mlxsw_core) +{ + struct mlxsw_core_irq_event_handler_item *item; + + mutex_lock(&mlxsw_core->irq_event_handler_lock); + list_for_each_entry(item, &mlxsw_core->irq_event_handler_list, list) { + if (item->cb) + item->cb(mlxsw_core); + } + mutex_unlock(&mlxsw_core->irq_event_handler_lock); +} +EXPORT_SYMBOL(mlxsw_core_irq_event_handlers_call); + +static int mlxsw_core_reg_access_cmd(struct mlxsw_core *mlxsw_core, + const struct mlxsw_reg_info *reg, + char *payload, + enum mlxsw_core_reg_access_type type) +{ + enum mlxsw_emad_op_tlv_status status; + int err, n_retry; + bool reset_ok; + char *in_mbox, *out_mbox, *tmp; + + dev_dbg(mlxsw_core->bus_info->dev, "Reg cmd access (reg_id=%x(%s),type=%s)\n", + reg->id, mlxsw_reg_id_str(reg->id), + mlxsw_core_reg_access_type_str(type)); + + in_mbox = mlxsw_cmd_mbox_alloc(); + if (!in_mbox) + return -ENOMEM; + + out_mbox = mlxsw_cmd_mbox_alloc(); + if (!out_mbox) { + err = -ENOMEM; + goto free_in_mbox; + } + + mlxsw_emad_pack_op_tlv(in_mbox, reg, type, + mlxsw_core_tid_get(mlxsw_core)); + tmp = in_mbox + MLXSW_EMAD_OP_TLV_LEN * sizeof(u32); + mlxsw_emad_pack_reg_tlv(tmp, reg, payload); + + /* There is a special treatment needed for MRSR (reset) register. + * The command interface will return error after the command + * is executed, so tell the lower layer to expect it + * and cope accordingly. + */ + reset_ok = reg->id == MLXSW_REG_MRSR_ID; + + n_retry = 0; +retry: + err = mlxsw_cmd_access_reg(mlxsw_core, reset_ok, in_mbox, out_mbox); + if (!err) { + err = mlxsw_emad_process_status(out_mbox, &status); + if (err) { + if (err == -EAGAIN && n_retry++ < MLXSW_EMAD_MAX_RETRY) + goto retry; + dev_err(mlxsw_core->bus_info->dev, "Reg cmd access status failed (status=%x(%s))\n", + status, mlxsw_emad_op_tlv_status_str(status)); + } + } + + if (!err) + memcpy(payload, mlxsw_emad_reg_payload_cmd(out_mbox), + reg->len); + + mlxsw_cmd_mbox_free(out_mbox); +free_in_mbox: + mlxsw_cmd_mbox_free(in_mbox); + if (err) + dev_err(mlxsw_core->bus_info->dev, "Reg cmd access failed (reg_id=%x(%s),type=%s)\n", + reg->id, mlxsw_reg_id_str(reg->id), + mlxsw_core_reg_access_type_str(type)); + return err; +} + +static void mlxsw_core_reg_access_cb(struct mlxsw_core *mlxsw_core, + char *payload, size_t payload_len, + unsigned long cb_priv) +{ + char *orig_payload = (char *) cb_priv; + + memcpy(orig_payload, payload, payload_len); +} + +static int mlxsw_core_reg_access(struct mlxsw_core *mlxsw_core, + const struct mlxsw_reg_info *reg, + char *payload, + enum mlxsw_core_reg_access_type type) +{ + LIST_HEAD(bulk_list); + int err; + + /* During initialization EMAD interface is not available to us, + * so we default to command interface. We switch to EMAD interface + * after setting the appropriate traps. + */ + if (!mlxsw_core->emad.use_emad) + return mlxsw_core_reg_access_cmd(mlxsw_core, reg, + payload, type); + + err = mlxsw_core_reg_access_emad(mlxsw_core, reg, + payload, type, &bulk_list, + mlxsw_core_reg_access_cb, + (unsigned long) payload); + if (err) + return err; + return mlxsw_reg_trans_bulk_wait(&bulk_list); +} + +int mlxsw_reg_query(struct mlxsw_core *mlxsw_core, + const struct mlxsw_reg_info *reg, char *payload) +{ + return mlxsw_core_reg_access(mlxsw_core, reg, payload, + MLXSW_CORE_REG_ACCESS_TYPE_QUERY); +} +EXPORT_SYMBOL(mlxsw_reg_query); + +int mlxsw_reg_write(struct mlxsw_core *mlxsw_core, + const struct mlxsw_reg_info *reg, char *payload) +{ + return mlxsw_core_reg_access(mlxsw_core, reg, payload, + MLXSW_CORE_REG_ACCESS_TYPE_WRITE); +} +EXPORT_SYMBOL(mlxsw_reg_write); + +void mlxsw_core_skb_receive(struct mlxsw_core *mlxsw_core, struct sk_buff *skb, + struct mlxsw_rx_info *rx_info) +{ + struct mlxsw_rx_listener_item *rxl_item; + const struct mlxsw_rx_listener *rxl; + u16 local_port; + bool found = false; + + if (rx_info->is_lag) { + dev_dbg_ratelimited(mlxsw_core->bus_info->dev, "%s: lag_id = %d, lag_port_index = 0x%x\n", + __func__, rx_info->u.lag_id, + rx_info->trap_id); + /* Upper layer does not care if the skb came from LAG or not, + * so just get the local_port for the lag port and push it up. + */ + local_port = mlxsw_core_lag_mapping_get(mlxsw_core, + rx_info->u.lag_id, + rx_info->lag_port_index); + } else { + local_port = rx_info->u.sys_port; + } + + dev_dbg_ratelimited(mlxsw_core->bus_info->dev, "%s: local_port = %d, trap_id = 0x%x\n", + __func__, local_port, rx_info->trap_id); + + if ((rx_info->trap_id >= MLXSW_TRAP_ID_MAX) || + (local_port >= mlxsw_core->max_ports)) + goto drop; + + rcu_read_lock(); + list_for_each_entry_rcu(rxl_item, &mlxsw_core->rx_listener_list, list) { + rxl = &rxl_item->rxl; + if ((rxl->local_port == MLXSW_PORT_DONT_CARE || + rxl->local_port == local_port) && + rxl->trap_id == rx_info->trap_id && + rxl->mirror_reason == rx_info->mirror_reason) { + if (rxl_item->enabled) + found = true; + break; + } + } + if (!found) { + rcu_read_unlock(); + goto drop; + } + + rxl->func(skb, local_port, rxl_item->priv); + rcu_read_unlock(); + return; + +drop: + dev_kfree_skb(skb); +} +EXPORT_SYMBOL(mlxsw_core_skb_receive); + +static int mlxsw_core_lag_mapping_index(struct mlxsw_core *mlxsw_core, + u16 lag_id, u8 port_index) +{ + return MLXSW_CORE_RES_GET(mlxsw_core, MAX_LAG_MEMBERS) * lag_id + + port_index; +} + +void mlxsw_core_lag_mapping_set(struct mlxsw_core *mlxsw_core, + u16 lag_id, u8 port_index, u16 local_port) +{ + int index = mlxsw_core_lag_mapping_index(mlxsw_core, + lag_id, port_index); + + mlxsw_core->lag.mapping[index] = local_port; +} +EXPORT_SYMBOL(mlxsw_core_lag_mapping_set); + +u16 mlxsw_core_lag_mapping_get(struct mlxsw_core *mlxsw_core, + u16 lag_id, u8 port_index) +{ + int index = mlxsw_core_lag_mapping_index(mlxsw_core, + lag_id, port_index); + + return mlxsw_core->lag.mapping[index]; +} +EXPORT_SYMBOL(mlxsw_core_lag_mapping_get); + +void mlxsw_core_lag_mapping_clear(struct mlxsw_core *mlxsw_core, + u16 lag_id, u16 local_port) +{ + int i; + + for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_core, MAX_LAG_MEMBERS); i++) { + int index = mlxsw_core_lag_mapping_index(mlxsw_core, + lag_id, i); + + if (mlxsw_core->lag.mapping[index] == local_port) + mlxsw_core->lag.mapping[index] = 0; + } +} +EXPORT_SYMBOL(mlxsw_core_lag_mapping_clear); + +bool mlxsw_core_res_valid(struct mlxsw_core *mlxsw_core, + enum mlxsw_res_id res_id) +{ + return mlxsw_res_valid(&mlxsw_core->res, res_id); +} +EXPORT_SYMBOL(mlxsw_core_res_valid); + +u64 mlxsw_core_res_get(struct mlxsw_core *mlxsw_core, + enum mlxsw_res_id res_id) +{ + return mlxsw_res_get(&mlxsw_core->res, res_id); +} +EXPORT_SYMBOL(mlxsw_core_res_get); + +static int __mlxsw_core_port_init(struct mlxsw_core *mlxsw_core, u16 local_port, + enum devlink_port_flavour flavour, + u8 slot_index, u32 port_number, bool split, + u32 split_port_subnumber, + bool splittable, u32 lanes, + const unsigned char *switch_id, + unsigned char switch_id_len) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + struct mlxsw_core_port *mlxsw_core_port = + &mlxsw_core->ports[local_port]; + struct devlink_port *devlink_port = &mlxsw_core_port->devlink_port; + struct devlink_port_attrs attrs = {}; + int err; + + attrs.split = split; + attrs.lanes = lanes; + attrs.splittable = splittable; + attrs.flavour = flavour; + attrs.phys.port_number = port_number; + attrs.phys.split_subport_number = split_port_subnumber; + memcpy(attrs.switch_id.id, switch_id, switch_id_len); + attrs.switch_id.id_len = switch_id_len; + mlxsw_core_port->local_port = local_port; + devlink_port_attrs_set(devlink_port, &attrs); + if (slot_index) { + struct mlxsw_linecard *linecard; + + linecard = mlxsw_linecard_get(mlxsw_core->linecards, + slot_index); + mlxsw_core_port->linecard = linecard; + devlink_port_linecard_set(devlink_port, + linecard->devlink_linecard); + } + err = devl_port_register(devlink, devlink_port, local_port); + if (err) + memset(mlxsw_core_port, 0, sizeof(*mlxsw_core_port)); + return err; +} + +static void __mlxsw_core_port_fini(struct mlxsw_core *mlxsw_core, u16 local_port) +{ + struct mlxsw_core_port *mlxsw_core_port = + &mlxsw_core->ports[local_port]; + struct devlink_port *devlink_port = &mlxsw_core_port->devlink_port; + + devl_port_unregister(devlink_port); + memset(mlxsw_core_port, 0, sizeof(*mlxsw_core_port)); +} + +int mlxsw_core_port_init(struct mlxsw_core *mlxsw_core, u16 local_port, + u8 slot_index, u32 port_number, bool split, + u32 split_port_subnumber, + bool splittable, u32 lanes, + const unsigned char *switch_id, + unsigned char switch_id_len) +{ + int err; + + err = __mlxsw_core_port_init(mlxsw_core, local_port, + DEVLINK_PORT_FLAVOUR_PHYSICAL, slot_index, + port_number, split, split_port_subnumber, + splittable, lanes, + switch_id, switch_id_len); + if (err) + return err; + + atomic_inc(&mlxsw_core->active_ports_count); + return 0; +} +EXPORT_SYMBOL(mlxsw_core_port_init); + +void mlxsw_core_port_fini(struct mlxsw_core *mlxsw_core, u16 local_port) +{ + atomic_dec(&mlxsw_core->active_ports_count); + + __mlxsw_core_port_fini(mlxsw_core, local_port); +} +EXPORT_SYMBOL(mlxsw_core_port_fini); + +int mlxsw_core_cpu_port_init(struct mlxsw_core *mlxsw_core, + void *port_driver_priv, + const unsigned char *switch_id, + unsigned char switch_id_len) +{ + struct mlxsw_core_port *mlxsw_core_port = + &mlxsw_core->ports[MLXSW_PORT_CPU_PORT]; + int err; + + err = __mlxsw_core_port_init(mlxsw_core, MLXSW_PORT_CPU_PORT, + DEVLINK_PORT_FLAVOUR_CPU, + 0, 0, false, 0, false, 0, + switch_id, switch_id_len); + if (err) + return err; + + mlxsw_core_port->port_driver_priv = port_driver_priv; + return 0; +} +EXPORT_SYMBOL(mlxsw_core_cpu_port_init); + +void mlxsw_core_cpu_port_fini(struct mlxsw_core *mlxsw_core) +{ + __mlxsw_core_port_fini(mlxsw_core, MLXSW_PORT_CPU_PORT); +} +EXPORT_SYMBOL(mlxsw_core_cpu_port_fini); + +void mlxsw_core_port_eth_set(struct mlxsw_core *mlxsw_core, u16 local_port, + void *port_driver_priv, struct net_device *dev) +{ + struct mlxsw_core_port *mlxsw_core_port = + &mlxsw_core->ports[local_port]; + struct devlink_port *devlink_port = &mlxsw_core_port->devlink_port; + + mlxsw_core_port->port_driver_priv = port_driver_priv; + devlink_port_type_eth_set(devlink_port, dev); +} +EXPORT_SYMBOL(mlxsw_core_port_eth_set); + +void mlxsw_core_port_clear(struct mlxsw_core *mlxsw_core, u16 local_port, + void *port_driver_priv) +{ + struct mlxsw_core_port *mlxsw_core_port = + &mlxsw_core->ports[local_port]; + struct devlink_port *devlink_port = &mlxsw_core_port->devlink_port; + + mlxsw_core_port->port_driver_priv = port_driver_priv; + devlink_port_type_clear(devlink_port); +} +EXPORT_SYMBOL(mlxsw_core_port_clear); + +struct devlink_port * +mlxsw_core_port_devlink_port_get(struct mlxsw_core *mlxsw_core, + u16 local_port) +{ + struct mlxsw_core_port *mlxsw_core_port = + &mlxsw_core->ports[local_port]; + struct devlink_port *devlink_port = &mlxsw_core_port->devlink_port; + + return devlink_port; +} +EXPORT_SYMBOL(mlxsw_core_port_devlink_port_get); + +struct mlxsw_linecard * +mlxsw_core_port_linecard_get(struct mlxsw_core *mlxsw_core, + u16 local_port) +{ + struct mlxsw_core_port *mlxsw_core_port = + &mlxsw_core->ports[local_port]; + + return mlxsw_core_port->linecard; +} + +void mlxsw_core_ports_remove_selected(struct mlxsw_core *mlxsw_core, + bool (*selector)(void *priv, u16 local_port), + void *priv) +{ + if (WARN_ON_ONCE(!mlxsw_core->driver->ports_remove_selected)) + return; + mlxsw_core->driver->ports_remove_selected(mlxsw_core, selector, priv); +} + +struct mlxsw_env *mlxsw_core_env(const struct mlxsw_core *mlxsw_core) +{ + return mlxsw_core->env; +} + +static void mlxsw_core_buf_dump_dbg(struct mlxsw_core *mlxsw_core, + const char *buf, size_t size) +{ + __be32 *m = (__be32 *) buf; + int i; + int count = size / sizeof(__be32); + + for (i = count - 1; i >= 0; i--) + if (m[i]) + break; + i++; + count = i ? i : 1; + for (i = 0; i < count; i += 4) + dev_dbg(mlxsw_core->bus_info->dev, "%04x - %08x %08x %08x %08x\n", + i * 4, be32_to_cpu(m[i]), be32_to_cpu(m[i + 1]), + be32_to_cpu(m[i + 2]), be32_to_cpu(m[i + 3])); +} + +int mlxsw_cmd_exec(struct mlxsw_core *mlxsw_core, u16 opcode, u8 opcode_mod, + u32 in_mod, bool out_mbox_direct, bool reset_ok, + char *in_mbox, size_t in_mbox_size, + char *out_mbox, size_t out_mbox_size) +{ + u8 status; + int err; + + BUG_ON(in_mbox_size % sizeof(u32) || out_mbox_size % sizeof(u32)); + if (!mlxsw_core->bus->cmd_exec) + return -EOPNOTSUPP; + + dev_dbg(mlxsw_core->bus_info->dev, "Cmd exec (opcode=%x(%s),opcode_mod=%x,in_mod=%x)\n", + opcode, mlxsw_cmd_opcode_str(opcode), opcode_mod, in_mod); + if (in_mbox) { + dev_dbg(mlxsw_core->bus_info->dev, "Input mailbox:\n"); + mlxsw_core_buf_dump_dbg(mlxsw_core, in_mbox, in_mbox_size); + } + + err = mlxsw_core->bus->cmd_exec(mlxsw_core->bus_priv, opcode, + opcode_mod, in_mod, out_mbox_direct, + in_mbox, in_mbox_size, + out_mbox, out_mbox_size, &status); + + if (!err && out_mbox) { + dev_dbg(mlxsw_core->bus_info->dev, "Output mailbox:\n"); + mlxsw_core_buf_dump_dbg(mlxsw_core, out_mbox, out_mbox_size); + } + + if (reset_ok && err == -EIO && + status == MLXSW_CMD_STATUS_RUNNING_RESET) { + err = 0; + } else if (err == -EIO && status != MLXSW_CMD_STATUS_OK) { + dev_err(mlxsw_core->bus_info->dev, "Cmd exec failed (opcode=%x(%s),opcode_mod=%x,in_mod=%x,status=%x(%s))\n", + opcode, mlxsw_cmd_opcode_str(opcode), opcode_mod, + in_mod, status, mlxsw_cmd_status_str(status)); + } else if (err == -ETIMEDOUT) { + dev_err(mlxsw_core->bus_info->dev, "Cmd exec timed-out (opcode=%x(%s),opcode_mod=%x,in_mod=%x)\n", + opcode, mlxsw_cmd_opcode_str(opcode), opcode_mod, + in_mod); + } + + return err; +} +EXPORT_SYMBOL(mlxsw_cmd_exec); + +int mlxsw_core_schedule_dw(struct delayed_work *dwork, unsigned long delay) +{ + return queue_delayed_work(mlxsw_wq, dwork, delay); +} +EXPORT_SYMBOL(mlxsw_core_schedule_dw); + +bool mlxsw_core_schedule_work(struct work_struct *work) +{ + return queue_work(mlxsw_owq, work); +} +EXPORT_SYMBOL(mlxsw_core_schedule_work); + +void mlxsw_core_flush_owq(void) +{ + flush_workqueue(mlxsw_owq); +} +EXPORT_SYMBOL(mlxsw_core_flush_owq); + +int mlxsw_core_kvd_sizes_get(struct mlxsw_core *mlxsw_core, + const struct mlxsw_config_profile *profile, + u64 *p_single_size, u64 *p_double_size, + u64 *p_linear_size) +{ + struct mlxsw_driver *driver = mlxsw_core->driver; + + if (!driver->kvd_sizes_get) + return -EINVAL; + + return driver->kvd_sizes_get(mlxsw_core, profile, + p_single_size, p_double_size, + p_linear_size); +} +EXPORT_SYMBOL(mlxsw_core_kvd_sizes_get); + +int mlxsw_core_resources_query(struct mlxsw_core *mlxsw_core, char *mbox, + struct mlxsw_res *res) +{ + int index, i; + u64 data; + u16 id; + int err; + + mlxsw_cmd_mbox_zero(mbox); + + for (index = 0; index < MLXSW_CMD_QUERY_RESOURCES_MAX_QUERIES; + index++) { + err = mlxsw_cmd_query_resources(mlxsw_core, mbox, index); + if (err) + return err; + + for (i = 0; i < MLXSW_CMD_QUERY_RESOURCES_PER_QUERY; i++) { + id = mlxsw_cmd_mbox_query_resource_id_get(mbox, i); + data = mlxsw_cmd_mbox_query_resource_data_get(mbox, i); + + if (id == MLXSW_CMD_QUERY_RESOURCES_TABLE_END_ID) + return 0; + + mlxsw_res_parse(res, id, data); + } + } + + /* If after MLXSW_RESOURCES_QUERY_MAX_QUERIES we still didn't get + * MLXSW_RESOURCES_TABLE_END_ID, something went bad in the FW. + */ + return -EIO; +} +EXPORT_SYMBOL(mlxsw_core_resources_query); + +u32 mlxsw_core_read_frc_h(struct mlxsw_core *mlxsw_core) +{ + return mlxsw_core->bus->read_frc_h(mlxsw_core->bus_priv); +} +EXPORT_SYMBOL(mlxsw_core_read_frc_h); + +u32 mlxsw_core_read_frc_l(struct mlxsw_core *mlxsw_core) +{ + return mlxsw_core->bus->read_frc_l(mlxsw_core->bus_priv); +} +EXPORT_SYMBOL(mlxsw_core_read_frc_l); + +u32 mlxsw_core_read_utc_sec(struct mlxsw_core *mlxsw_core) +{ + return mlxsw_core->bus->read_utc_sec(mlxsw_core->bus_priv); +} +EXPORT_SYMBOL(mlxsw_core_read_utc_sec); + +u32 mlxsw_core_read_utc_nsec(struct mlxsw_core *mlxsw_core) +{ + return mlxsw_core->bus->read_utc_nsec(mlxsw_core->bus_priv); +} +EXPORT_SYMBOL(mlxsw_core_read_utc_nsec); + +bool mlxsw_core_sdq_supports_cqe_v2(struct mlxsw_core *mlxsw_core) +{ + return mlxsw_core->driver->sdq_supports_cqe_v2; +} +EXPORT_SYMBOL(mlxsw_core_sdq_supports_cqe_v2); + +void mlxsw_core_emad_string_tlv_enable(struct mlxsw_core *mlxsw_core) +{ + mlxsw_core->emad.enable_string_tlv = true; +} +EXPORT_SYMBOL(mlxsw_core_emad_string_tlv_enable); + +static int __init mlxsw_core_module_init(void) +{ + int err; + + err = mlxsw_linecard_driver_register(); + if (err) + return err; + + mlxsw_wq = alloc_workqueue(mlxsw_core_driver_name, 0, 0); + if (!mlxsw_wq) { + err = -ENOMEM; + goto err_alloc_workqueue; + } + mlxsw_owq = alloc_ordered_workqueue("%s_ordered", 0, + mlxsw_core_driver_name); + if (!mlxsw_owq) { + err = -ENOMEM; + goto err_alloc_ordered_workqueue; + } + return 0; + +err_alloc_ordered_workqueue: + destroy_workqueue(mlxsw_wq); +err_alloc_workqueue: + mlxsw_linecard_driver_unregister(); + return err; +} + +static void __exit mlxsw_core_module_exit(void) +{ + destroy_workqueue(mlxsw_owq); + destroy_workqueue(mlxsw_wq); + mlxsw_linecard_driver_unregister(); +} + +module_init(mlxsw_core_module_init); +module_exit(mlxsw_core_module_exit); + +MODULE_LICENSE("Dual BSD/GPL"); +MODULE_AUTHOR("Jiri Pirko <jiri@mellanox.com>"); +MODULE_DESCRIPTION("Mellanox switch device core driver"); diff --git a/drivers/net/ethernet/mellanox/mlxsw/core.h b/drivers/net/ethernet/mellanox/mlxsw/core.h new file mode 100644 index 000000000..ca0c3d2be --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/core.h @@ -0,0 +1,674 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_CORE_H +#define _MLXSW_CORE_H + +#include <linux/module.h> +#include <linux/device.h> +#include <linux/slab.h> +#include <linux/gfp.h> +#include <linux/types.h> +#include <linux/skbuff.h> +#include <linux/workqueue.h> +#include <linux/net_namespace.h> +#include <linux/auxiliary_bus.h> +#include <net/devlink.h> + +#include "trap.h" +#include "reg.h" +#include "cmd.h" +#include "resources.h" +#include "../mlxfw/mlxfw.h" + +enum mlxsw_core_resource_id { + MLXSW_CORE_RESOURCE_PORTS = 1, + MLXSW_CORE_RESOURCE_MAX, +}; + +struct mlxsw_core; +struct mlxsw_core_port; +struct mlxsw_driver; +struct mlxsw_bus; +struct mlxsw_bus_info; +struct mlxsw_fw_rev; + +unsigned int mlxsw_core_max_ports(const struct mlxsw_core *mlxsw_core); + +int mlxsw_core_max_lag(struct mlxsw_core *mlxsw_core, u16 *p_max_lag); + +void *mlxsw_core_driver_priv(struct mlxsw_core *mlxsw_core); + +struct mlxsw_linecards *mlxsw_core_linecards(struct mlxsw_core *mlxsw_core); + +void mlxsw_core_linecards_set(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecards *linecard); + +bool +mlxsw_core_fw_rev_minor_subminor_validate(const struct mlxsw_fw_rev *rev, + const struct mlxsw_fw_rev *req_rev); + +int mlxsw_core_driver_register(struct mlxsw_driver *mlxsw_driver); +void mlxsw_core_driver_unregister(struct mlxsw_driver *mlxsw_driver); + +int mlxsw_core_fw_flash(struct mlxsw_core *mlxsw_core, + struct mlxfw_dev *mlxfw_dev, + const struct firmware *firmware, + struct netlink_ext_ack *extack); + +int mlxsw_core_bus_device_register(const struct mlxsw_bus_info *mlxsw_bus_info, + const struct mlxsw_bus *mlxsw_bus, + void *bus_priv, bool reload, + struct devlink *devlink, + struct netlink_ext_ack *extack); +void mlxsw_core_bus_device_unregister(struct mlxsw_core *mlxsw_core, bool reload); + +struct mlxsw_tx_info { + u16 local_port; + bool is_emad; +}; + +struct mlxsw_rx_md_info { + u32 cookie_index; + u32 latency; + u32 tx_congestion; + union { + /* Valid when 'tx_port_valid' is set. */ + u16 tx_sys_port; + u16 tx_lag_id; + }; + u16 tx_lag_port_index; /* Valid when 'tx_port_is_lag' is set. */ + u8 tx_tc; + u8 latency_valid:1, + tx_congestion_valid:1, + tx_tc_valid:1, + tx_port_valid:1, + tx_port_is_lag:1, + unused:3; +}; + +bool mlxsw_core_skb_transmit_busy(struct mlxsw_core *mlxsw_core, + const struct mlxsw_tx_info *tx_info); +int mlxsw_core_skb_transmit(struct mlxsw_core *mlxsw_core, struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info); +void mlxsw_core_ptp_transmitted(struct mlxsw_core *mlxsw_core, + struct sk_buff *skb, u16 local_port); + +struct mlxsw_rx_listener { + void (*func)(struct sk_buff *skb, u16 local_port, void *priv); + u16 local_port; + u8 mirror_reason; + u16 trap_id; +}; + +struct mlxsw_event_listener { + void (*func)(const struct mlxsw_reg_info *reg, + char *payload, void *priv); + enum mlxsw_event_trap_id trap_id; +}; + +struct mlxsw_listener { + u16 trap_id; + union { + struct mlxsw_rx_listener rx_listener; + struct mlxsw_event_listener event_listener; + }; + enum mlxsw_reg_hpkt_action en_action; /* Action when enabled */ + enum mlxsw_reg_hpkt_action dis_action; /* Action when disabled */ + u8 en_trap_group; /* Trap group when enabled */ + u8 dis_trap_group; /* Trap group when disabled */ + u8 is_ctrl:1, /* should go via control buffer or not */ + is_event:1, + enabled_on_register:1; /* Trap should be enabled when listener + * is registered. + */ +}; + +#define __MLXSW_RXL(_func, _trap_id, _en_action, _is_ctrl, _en_trap_group, \ + _dis_action, _enabled_on_register, _dis_trap_group, \ + _mirror_reason) \ + { \ + .trap_id = MLXSW_TRAP_ID_##_trap_id, \ + .rx_listener = \ + { \ + .func = _func, \ + .local_port = MLXSW_PORT_DONT_CARE, \ + .mirror_reason = _mirror_reason, \ + .trap_id = MLXSW_TRAP_ID_##_trap_id, \ + }, \ + .en_action = MLXSW_REG_HPKT_ACTION_##_en_action, \ + .dis_action = MLXSW_REG_HPKT_ACTION_##_dis_action, \ + .en_trap_group = MLXSW_REG_HTGT_TRAP_GROUP_##_en_trap_group, \ + .dis_trap_group = MLXSW_REG_HTGT_TRAP_GROUP_##_dis_trap_group, \ + .is_ctrl = _is_ctrl, \ + .enabled_on_register = _enabled_on_register, \ + } + +#define MLXSW_RXL(_func, _trap_id, _en_action, _is_ctrl, _trap_group, \ + _dis_action) \ + __MLXSW_RXL(_func, _trap_id, _en_action, _is_ctrl, _trap_group, \ + _dis_action, true, _trap_group, 0) + +#define MLXSW_RXL_DIS(_func, _trap_id, _en_action, _is_ctrl, _en_trap_group, \ + _dis_action, _dis_trap_group) \ + __MLXSW_RXL(_func, _trap_id, _en_action, _is_ctrl, _en_trap_group, \ + _dis_action, false, _dis_trap_group, 0) + +#define MLXSW_RXL_MIRROR(_func, _session_id, _trap_group, _mirror_reason) \ + __MLXSW_RXL(_func, MIRROR_SESSION##_session_id, TRAP_TO_CPU, false, \ + _trap_group, TRAP_TO_CPU, true, _trap_group, \ + _mirror_reason) + +#define MLXSW_EVENTL(_func, _trap_id, _trap_group) \ + { \ + .trap_id = MLXSW_TRAP_ID_##_trap_id, \ + .event_listener = \ + { \ + .func = _func, \ + .trap_id = MLXSW_TRAP_ID_##_trap_id, \ + }, \ + .en_action = MLXSW_REG_HPKT_ACTION_TRAP_TO_CPU, \ + .en_trap_group = MLXSW_REG_HTGT_TRAP_GROUP_##_trap_group, \ + .is_event = true, \ + .enabled_on_register = true, \ + } + +#define MLXSW_CORE_EVENTL(_func, _trap_id) \ + MLXSW_EVENTL(_func, _trap_id, CORE_EVENT) + +int mlxsw_core_rx_listener_register(struct mlxsw_core *mlxsw_core, + const struct mlxsw_rx_listener *rxl, + void *priv, bool enabled); +void mlxsw_core_rx_listener_unregister(struct mlxsw_core *mlxsw_core, + const struct mlxsw_rx_listener *rxl); + +int mlxsw_core_event_listener_register(struct mlxsw_core *mlxsw_core, + const struct mlxsw_event_listener *el, + void *priv); +void mlxsw_core_event_listener_unregister(struct mlxsw_core *mlxsw_core, + const struct mlxsw_event_listener *el); + +int mlxsw_core_trap_register(struct mlxsw_core *mlxsw_core, + const struct mlxsw_listener *listener, + void *priv); +void mlxsw_core_trap_unregister(struct mlxsw_core *mlxsw_core, + const struct mlxsw_listener *listener, + void *priv); +int mlxsw_core_traps_register(struct mlxsw_core *mlxsw_core, + const struct mlxsw_listener *listeners, + size_t listeners_count, void *priv); +void mlxsw_core_traps_unregister(struct mlxsw_core *mlxsw_core, + const struct mlxsw_listener *listeners, + size_t listeners_count, void *priv); +int mlxsw_core_trap_state_set(struct mlxsw_core *mlxsw_core, + const struct mlxsw_listener *listener, + bool enabled); + +typedef void mlxsw_reg_trans_cb_t(struct mlxsw_core *mlxsw_core, char *payload, + size_t payload_len, unsigned long cb_priv); + +int mlxsw_reg_trans_query(struct mlxsw_core *mlxsw_core, + const struct mlxsw_reg_info *reg, char *payload, + struct list_head *bulk_list, + mlxsw_reg_trans_cb_t *cb, unsigned long cb_priv); +int mlxsw_reg_trans_write(struct mlxsw_core *mlxsw_core, + const struct mlxsw_reg_info *reg, char *payload, + struct list_head *bulk_list, + mlxsw_reg_trans_cb_t *cb, unsigned long cb_priv); +int mlxsw_reg_trans_bulk_wait(struct list_head *bulk_list); + +typedef void mlxsw_irq_event_cb_t(struct mlxsw_core *mlxsw_core); + +int mlxsw_core_irq_event_handler_register(struct mlxsw_core *mlxsw_core, + mlxsw_irq_event_cb_t cb); +void mlxsw_core_irq_event_handler_unregister(struct mlxsw_core *mlxsw_core, + mlxsw_irq_event_cb_t cb); +void mlxsw_core_irq_event_handlers_call(struct mlxsw_core *mlxsw_core); + +int mlxsw_reg_query(struct mlxsw_core *mlxsw_core, + const struct mlxsw_reg_info *reg, char *payload); +int mlxsw_reg_write(struct mlxsw_core *mlxsw_core, + const struct mlxsw_reg_info *reg, char *payload); + +struct mlxsw_rx_info { + bool is_lag; + union { + u16 sys_port; + u16 lag_id; + } u; + u16 lag_port_index; + u8 mirror_reason; + int trap_id; +}; + +void mlxsw_core_skb_receive(struct mlxsw_core *mlxsw_core, struct sk_buff *skb, + struct mlxsw_rx_info *rx_info); + +void mlxsw_core_lag_mapping_set(struct mlxsw_core *mlxsw_core, + u16 lag_id, u8 port_index, u16 local_port); +u16 mlxsw_core_lag_mapping_get(struct mlxsw_core *mlxsw_core, + u16 lag_id, u8 port_index); +void mlxsw_core_lag_mapping_clear(struct mlxsw_core *mlxsw_core, + u16 lag_id, u16 local_port); + +void *mlxsw_core_port_driver_priv(struct mlxsw_core_port *mlxsw_core_port); +int mlxsw_core_port_init(struct mlxsw_core *mlxsw_core, u16 local_port, + u8 slot_index, u32 port_number, bool split, + u32 split_port_subnumber, + bool splittable, u32 lanes, + const unsigned char *switch_id, + unsigned char switch_id_len); +void mlxsw_core_port_fini(struct mlxsw_core *mlxsw_core, u16 local_port); +int mlxsw_core_cpu_port_init(struct mlxsw_core *mlxsw_core, + void *port_driver_priv, + const unsigned char *switch_id, + unsigned char switch_id_len); +void mlxsw_core_cpu_port_fini(struct mlxsw_core *mlxsw_core); +void mlxsw_core_port_eth_set(struct mlxsw_core *mlxsw_core, u16 local_port, + void *port_driver_priv, struct net_device *dev); +void mlxsw_core_port_clear(struct mlxsw_core *mlxsw_core, u16 local_port, + void *port_driver_priv); +struct devlink_port * +mlxsw_core_port_devlink_port_get(struct mlxsw_core *mlxsw_core, + u16 local_port); +struct mlxsw_linecard * +mlxsw_core_port_linecard_get(struct mlxsw_core *mlxsw_core, + u16 local_port); +void mlxsw_core_ports_remove_selected(struct mlxsw_core *mlxsw_core, + bool (*selector)(void *priv, + u16 local_port), + void *priv); +struct mlxsw_env *mlxsw_core_env(const struct mlxsw_core *mlxsw_core); + +int mlxsw_core_schedule_dw(struct delayed_work *dwork, unsigned long delay); +bool mlxsw_core_schedule_work(struct work_struct *work); +void mlxsw_core_flush_owq(void); +int mlxsw_core_resources_query(struct mlxsw_core *mlxsw_core, char *mbox, + struct mlxsw_res *res); + +#define MLXSW_CONFIG_PROFILE_SWID_COUNT 8 + +struct mlxsw_swid_config { + u8 used_type:1, + used_properties:1; + u8 type; + u8 properties; +}; + +struct mlxsw_config_profile { + u16 used_max_vepa_channels:1, + used_max_lag:1, + used_max_mid:1, + used_max_pgt:1, + used_max_system_port:1, + used_max_vlan_groups:1, + used_max_regions:1, + used_flood_tables:1, + used_flood_mode:1, + used_max_ib_mc:1, + used_max_pkey:1, + used_ar_sec:1, + used_adaptive_routing_group_cap:1, + used_ubridge:1, + used_kvd_sizes:1, + used_cqe_time_stamp_type:1; + u8 max_vepa_channels; + u16 max_lag; + u16 max_mid; + u16 max_pgt; + u16 max_system_port; + u16 max_vlan_groups; + u16 max_regions; + u8 max_flood_tables; + u8 max_vid_flood_tables; + u8 flood_mode; + u8 max_fid_offset_flood_tables; + u16 fid_offset_flood_table_size; + u8 max_fid_flood_tables; + u16 fid_flood_table_size; + u16 max_ib_mc; + u16 max_pkey; + u8 ar_sec; + u16 adaptive_routing_group_cap; + u8 arn; + u8 ubridge; + u32 kvd_linear_size; + u8 kvd_hash_single_parts; + u8 kvd_hash_double_parts; + u8 cqe_time_stamp_type; + struct mlxsw_swid_config swid_config[MLXSW_CONFIG_PROFILE_SWID_COUNT]; +}; + +struct mlxsw_driver { + struct list_head list; + const char *kind; + size_t priv_size; + const struct mlxsw_fw_rev *fw_req_rev; + const char *fw_filename; + int (*init)(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct netlink_ext_ack *extack); + void (*fini)(struct mlxsw_core *mlxsw_core); + int (*port_split)(struct mlxsw_core *mlxsw_core, u16 local_port, + unsigned int count, struct netlink_ext_ack *extack); + int (*port_unsplit)(struct mlxsw_core *mlxsw_core, u16 local_port, + struct netlink_ext_ack *extack); + void (*ports_remove_selected)(struct mlxsw_core *mlxsw_core, + bool (*selector)(void *priv, + u16 local_port), + void *priv); + int (*sb_pool_get)(struct mlxsw_core *mlxsw_core, + unsigned int sb_index, u16 pool_index, + struct devlink_sb_pool_info *pool_info); + int (*sb_pool_set)(struct mlxsw_core *mlxsw_core, + unsigned int sb_index, u16 pool_index, u32 size, + enum devlink_sb_threshold_type threshold_type, + struct netlink_ext_ack *extack); + int (*sb_port_pool_get)(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 pool_index, + u32 *p_threshold); + int (*sb_port_pool_set)(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 pool_index, + u32 threshold, struct netlink_ext_ack *extack); + int (*sb_tc_pool_bind_get)(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 tc_index, + enum devlink_sb_pool_type pool_type, + u16 *p_pool_index, u32 *p_threshold); + int (*sb_tc_pool_bind_set)(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 tc_index, + enum devlink_sb_pool_type pool_type, + u16 pool_index, u32 threshold, + struct netlink_ext_ack *extack); + int (*sb_occ_snapshot)(struct mlxsw_core *mlxsw_core, + unsigned int sb_index); + int (*sb_occ_max_clear)(struct mlxsw_core *mlxsw_core, + unsigned int sb_index); + int (*sb_occ_port_pool_get)(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 pool_index, + u32 *p_cur, u32 *p_max); + int (*sb_occ_tc_port_bind_get)(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 tc_index, + enum devlink_sb_pool_type pool_type, + u32 *p_cur, u32 *p_max); + int (*trap_init)(struct mlxsw_core *mlxsw_core, + const struct devlink_trap *trap, void *trap_ctx); + void (*trap_fini)(struct mlxsw_core *mlxsw_core, + const struct devlink_trap *trap, void *trap_ctx); + int (*trap_action_set)(struct mlxsw_core *mlxsw_core, + const struct devlink_trap *trap, + enum devlink_trap_action action, + struct netlink_ext_ack *extack); + int (*trap_group_init)(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_group *group); + int (*trap_group_set)(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_group *group, + const struct devlink_trap_policer *policer, + struct netlink_ext_ack *extack); + int (*trap_policer_init)(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_policer *policer); + void (*trap_policer_fini)(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_policer *policer); + int (*trap_policer_set)(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_policer *policer, + u64 rate, u64 burst, + struct netlink_ext_ack *extack); + int (*trap_policer_counter_get)(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_policer *policer, + u64 *p_drops); + void (*txhdr_construct)(struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info); + int (*resources_register)(struct mlxsw_core *mlxsw_core); + int (*kvd_sizes_get)(struct mlxsw_core *mlxsw_core, + const struct mlxsw_config_profile *profile, + u64 *p_single_size, u64 *p_double_size, + u64 *p_linear_size); + int (*params_register)(struct mlxsw_core *mlxsw_core); + void (*params_unregister)(struct mlxsw_core *mlxsw_core); + + /* Notify a driver that a timestamped packet was transmitted. Driver + * is responsible for freeing the passed-in SKB. + */ + void (*ptp_transmitted)(struct mlxsw_core *mlxsw_core, + struct sk_buff *skb, u16 local_port); + + u8 txhdr_len; + const struct mlxsw_config_profile *profile; + bool sdq_supports_cqe_v2; +}; + +int mlxsw_core_kvd_sizes_get(struct mlxsw_core *mlxsw_core, + const struct mlxsw_config_profile *profile, + u64 *p_single_size, u64 *p_double_size, + u64 *p_linear_size); + +u32 mlxsw_core_read_frc_h(struct mlxsw_core *mlxsw_core); +u32 mlxsw_core_read_frc_l(struct mlxsw_core *mlxsw_core); + +u32 mlxsw_core_read_utc_sec(struct mlxsw_core *mlxsw_core); +u32 mlxsw_core_read_utc_nsec(struct mlxsw_core *mlxsw_core); + +bool mlxsw_core_sdq_supports_cqe_v2(struct mlxsw_core *mlxsw_core); + +void mlxsw_core_emad_string_tlv_enable(struct mlxsw_core *mlxsw_core); + +bool mlxsw_core_res_valid(struct mlxsw_core *mlxsw_core, + enum mlxsw_res_id res_id); + +#define MLXSW_CORE_RES_VALID(mlxsw_core, short_res_id) \ + mlxsw_core_res_valid(mlxsw_core, MLXSW_RES_ID_##short_res_id) + +u64 mlxsw_core_res_get(struct mlxsw_core *mlxsw_core, + enum mlxsw_res_id res_id); + +#define MLXSW_CORE_RES_GET(mlxsw_core, short_res_id) \ + mlxsw_core_res_get(mlxsw_core, MLXSW_RES_ID_##short_res_id) + +static inline struct net *mlxsw_core_net(struct mlxsw_core *mlxsw_core) +{ + return devlink_net(priv_to_devlink(mlxsw_core)); +} + +#define MLXSW_BUS_F_TXRX BIT(0) +#define MLXSW_BUS_F_RESET BIT(1) + +struct mlxsw_bus { + const char *kind; + int (*init)(void *bus_priv, struct mlxsw_core *mlxsw_core, + const struct mlxsw_config_profile *profile, + struct mlxsw_res *res); + void (*fini)(void *bus_priv); + bool (*skb_transmit_busy)(void *bus_priv, + const struct mlxsw_tx_info *tx_info); + int (*skb_transmit)(void *bus_priv, struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info); + int (*cmd_exec)(void *bus_priv, u16 opcode, u8 opcode_mod, + u32 in_mod, bool out_mbox_direct, + char *in_mbox, size_t in_mbox_size, + char *out_mbox, size_t out_mbox_size, + u8 *p_status); + u32 (*read_frc_h)(void *bus_priv); + u32 (*read_frc_l)(void *bus_priv); + u32 (*read_utc_sec)(void *bus_priv); + u32 (*read_utc_nsec)(void *bus_priv); + u8 features; +}; + +struct mlxsw_fw_rev { + u16 major; + u16 minor; + u16 subminor; + u16 can_reset_minor; +}; + +struct mlxsw_bus_info { + const char *device_kind; + const char *device_name; + struct device *dev; + struct mlxsw_fw_rev fw_rev; + u8 vsd[MLXSW_CMD_BOARDINFO_VSD_LEN]; + u8 psid[MLXSW_CMD_BOARDINFO_PSID_LEN]; + u8 low_frequency:1, + read_clock_capable:1; +}; + +struct mlxsw_hwmon; + +#ifdef CONFIG_MLXSW_CORE_HWMON + +int mlxsw_hwmon_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct mlxsw_hwmon **p_hwmon); +void mlxsw_hwmon_fini(struct mlxsw_hwmon *mlxsw_hwmon); + +#else + +static inline int mlxsw_hwmon_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct mlxsw_hwmon **p_hwmon) +{ + return 0; +} + +static inline void mlxsw_hwmon_fini(struct mlxsw_hwmon *mlxsw_hwmon) +{ +} + +#endif + +struct mlxsw_thermal; + +#ifdef CONFIG_MLXSW_CORE_THERMAL + +int mlxsw_thermal_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct mlxsw_thermal **p_thermal); +void mlxsw_thermal_fini(struct mlxsw_thermal *thermal); + +#else + +static inline int mlxsw_thermal_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct mlxsw_thermal **p_thermal) +{ + return 0; +} + +static inline void mlxsw_thermal_fini(struct mlxsw_thermal *thermal) +{ +} + +#endif + +enum mlxsw_devlink_param_id { + MLXSW_DEVLINK_PARAM_ID_BASE = DEVLINK_PARAM_GENERIC_ID_MAX, + MLXSW_DEVLINK_PARAM_ID_ACL_REGION_REHASH_INTERVAL, +}; + +struct mlxsw_cqe_ts { + u8 sec; + u32 nsec; +}; + +struct mlxsw_skb_cb { + union { + struct mlxsw_tx_info tx_info; + struct mlxsw_rx_md_info rx_md_info; + }; + struct mlxsw_cqe_ts cqe_ts; +}; + +static inline struct mlxsw_skb_cb *mlxsw_skb_cb(struct sk_buff *skb) +{ + BUILD_BUG_ON(sizeof(mlxsw_skb_cb) > sizeof(skb->cb)); + return (struct mlxsw_skb_cb *) skb->cb; +} + +struct mlxsw_linecards; + +enum mlxsw_linecard_status_event_type { + MLXSW_LINECARD_STATUS_EVENT_TYPE_PROVISION, + MLXSW_LINECARD_STATUS_EVENT_TYPE_UNPROVISION, +}; + +struct mlxsw_linecard_bdev; + +struct mlxsw_linecard_device_info { + u16 fw_major; + u16 fw_minor; + u16 fw_sub_minor; + char psid[MLXSW_REG_MGIR_FW_INFO_PSID_SIZE]; +}; + +struct mlxsw_linecard { + u8 slot_index; + struct mlxsw_linecards *linecards; + struct devlink_linecard *devlink_linecard; + struct mutex lock; /* Locks accesses to the linecard structure */ + char name[MLXSW_REG_MDDQ_SLOT_ASCII_NAME_LEN]; + char mbct_pl[MLXSW_REG_MBCT_LEN]; /* Too big for stack */ + enum mlxsw_linecard_status_event_type status_event_type_to; + struct delayed_work status_event_to_dw; + u8 provisioned:1, + ready:1, + active:1; + u16 hw_revision; + u16 ini_version; + struct mlxsw_linecard_bdev *bdev; + struct { + struct mlxsw_linecard_device_info info; + u8 index; + } device; +}; + +struct mlxsw_linecard_types_info; + +struct mlxsw_linecards { + struct mlxsw_core *mlxsw_core; + const struct mlxsw_bus_info *bus_info; + u8 count; + struct mlxsw_linecard_types_info *types_info; + struct list_head event_ops_list; + struct mutex event_ops_list_lock; /* Locks accesses to event ops list */ + struct mlxsw_linecard linecards[]; +}; + +static inline struct mlxsw_linecard * +mlxsw_linecard_get(struct mlxsw_linecards *linecards, u8 slot_index) +{ + return &linecards->linecards[slot_index - 1]; +} + +int mlxsw_linecard_devlink_info_get(struct mlxsw_linecard *linecard, + struct devlink_info_req *req, + struct netlink_ext_ack *extack); +int mlxsw_linecard_flash_update(struct devlink *linecard_devlink, + struct mlxsw_linecard *linecard, + const struct firmware *firmware, + struct netlink_ext_ack *extack); + +int mlxsw_linecards_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *bus_info); +void mlxsw_linecards_fini(struct mlxsw_core *mlxsw_core); + +typedef void mlxsw_linecards_event_op_t(struct mlxsw_core *mlxsw_core, + u8 slot_index, void *priv); + +struct mlxsw_linecards_event_ops { + mlxsw_linecards_event_op_t *got_active; + mlxsw_linecards_event_op_t *got_inactive; +}; + +int mlxsw_linecards_event_ops_register(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecards_event_ops *ops, + void *priv); +void mlxsw_linecards_event_ops_unregister(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecards_event_ops *ops, + void *priv); + +int mlxsw_linecard_bdev_add(struct mlxsw_linecard *linecard); +void mlxsw_linecard_bdev_del(struct mlxsw_linecard *linecard); + +int mlxsw_linecard_driver_register(void); +void mlxsw_linecard_driver_unregister(void); + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.c b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.c new file mode 100644 index 000000000..9dfe71481 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.c @@ -0,0 +1,2218 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/types.h> +#include <linux/slab.h> +#include <linux/errno.h> +#include <linux/rhashtable.h> +#include <linux/list.h> +#include <linux/idr.h> +#include <linux/refcount.h> +#include <net/flow_offload.h> + +#include "item.h" +#include "trap.h" +#include "core_acl_flex_actions.h" + +enum mlxsw_afa_set_type { + MLXSW_AFA_SET_TYPE_NEXT, + MLXSW_AFA_SET_TYPE_GOTO, +}; + +/* afa_set_type + * Type of the record at the end of the action set. + */ +MLXSW_ITEM32(afa, set, type, 0xA0, 28, 4); + +/* afa_set_next_action_set_ptr + * A pointer to the next action set in the KVD Centralized database. + */ +MLXSW_ITEM32(afa, set, next_action_set_ptr, 0xA4, 0, 24); + +/* afa_set_goto_g + * group - When set, the binding is of an ACL group. When cleared, + * the binding is of an ACL. + * Must be set to 1 for Spectrum. + */ +MLXSW_ITEM32(afa, set, goto_g, 0xA4, 29, 1); + +enum mlxsw_afa_set_goto_binding_cmd { + /* continue go the next binding point */ + MLXSW_AFA_SET_GOTO_BINDING_CMD_NONE, + /* jump to the next binding point no return */ + MLXSW_AFA_SET_GOTO_BINDING_CMD_JUMP, + /* terminate the acl binding */ + MLXSW_AFA_SET_GOTO_BINDING_CMD_TERM = 4, +}; + +/* afa_set_goto_binding_cmd */ +MLXSW_ITEM32(afa, set, goto_binding_cmd, 0xA4, 24, 3); + +/* afa_set_goto_next_binding + * ACL/ACL group identifier. If the g bit is set, this field should hold + * the acl_group_id, else it should hold the acl_id. + */ +MLXSW_ITEM32(afa, set, goto_next_binding, 0xA4, 0, 16); + +/* afa_all_action_type + * Action Type. + */ +MLXSW_ITEM32(afa, all, action_type, 0x00, 24, 6); + +struct mlxsw_afa { + unsigned int max_acts_per_set; + const struct mlxsw_afa_ops *ops; + void *ops_priv; + struct rhashtable set_ht; + struct rhashtable fwd_entry_ht; + struct rhashtable cookie_ht; + struct rhashtable policer_ht; + struct idr cookie_idr; + struct list_head policer_list; +}; + +#define MLXSW_AFA_SET_LEN 0xA8 + +struct mlxsw_afa_set_ht_key { + char enc_actions[MLXSW_AFA_SET_LEN]; /* Encoded set */ + bool is_first; +}; + +/* Set structure holds one action set record. It contains up to three + * actions (depends on size of particular actions). The set is either + * put directly to a rule, or it is stored in KVD linear area. + * To prevent duplicate entries in KVD linear area, a hashtable is + * used to track sets that were previously inserted and may be shared. + */ + +struct mlxsw_afa_set { + struct rhash_head ht_node; + struct mlxsw_afa_set_ht_key ht_key; + u32 kvdl_index; + u8 shared:1, /* Inserted in hashtable (doesn't mean that + * kvdl_index is valid). + */ + has_trap:1, + has_police:1; + unsigned int ref_count; + struct mlxsw_afa_set *next; /* Pointer to the next set. */ + struct mlxsw_afa_set *prev; /* Pointer to the previous set, + * note that set may have multiple + * sets from multiple blocks + * pointing at it. This is only + * usable until commit. + */ +}; + +static const struct rhashtable_params mlxsw_afa_set_ht_params = { + .key_len = sizeof(struct mlxsw_afa_set_ht_key), + .key_offset = offsetof(struct mlxsw_afa_set, ht_key), + .head_offset = offsetof(struct mlxsw_afa_set, ht_node), + .automatic_shrinking = true, +}; + +struct mlxsw_afa_fwd_entry_ht_key { + u16 local_port; +}; + +struct mlxsw_afa_fwd_entry { + struct rhash_head ht_node; + struct mlxsw_afa_fwd_entry_ht_key ht_key; + u32 kvdl_index; + unsigned int ref_count; +}; + +static const struct rhashtable_params mlxsw_afa_fwd_entry_ht_params = { + .key_len = sizeof(struct mlxsw_afa_fwd_entry_ht_key), + .key_offset = offsetof(struct mlxsw_afa_fwd_entry, ht_key), + .head_offset = offsetof(struct mlxsw_afa_fwd_entry, ht_node), + .automatic_shrinking = true, +}; + +struct mlxsw_afa_cookie { + struct rhash_head ht_node; + refcount_t ref_count; + struct rcu_head rcu; + u32 cookie_index; + struct flow_action_cookie fa_cookie; +}; + +static u32 mlxsw_afa_cookie_hash(const struct flow_action_cookie *fa_cookie, + u32 seed) +{ + return jhash2((u32 *) fa_cookie->cookie, + fa_cookie->cookie_len / sizeof(u32), seed); +} + +static u32 mlxsw_afa_cookie_key_hashfn(const void *data, u32 len, u32 seed) +{ + const struct flow_action_cookie *fa_cookie = data; + + return mlxsw_afa_cookie_hash(fa_cookie, seed); +} + +static u32 mlxsw_afa_cookie_obj_hashfn(const void *data, u32 len, u32 seed) +{ + const struct mlxsw_afa_cookie *cookie = data; + + return mlxsw_afa_cookie_hash(&cookie->fa_cookie, seed); +} + +static int mlxsw_afa_cookie_obj_cmpfn(struct rhashtable_compare_arg *arg, + const void *obj) +{ + const struct flow_action_cookie *fa_cookie = arg->key; + const struct mlxsw_afa_cookie *cookie = obj; + + if (cookie->fa_cookie.cookie_len == fa_cookie->cookie_len) + return memcmp(cookie->fa_cookie.cookie, fa_cookie->cookie, + fa_cookie->cookie_len); + return 1; +} + +static const struct rhashtable_params mlxsw_afa_cookie_ht_params = { + .head_offset = offsetof(struct mlxsw_afa_cookie, ht_node), + .hashfn = mlxsw_afa_cookie_key_hashfn, + .obj_hashfn = mlxsw_afa_cookie_obj_hashfn, + .obj_cmpfn = mlxsw_afa_cookie_obj_cmpfn, + .automatic_shrinking = true, +}; + +struct mlxsw_afa_policer { + struct rhash_head ht_node; + struct list_head list; /* Member of policer_list */ + refcount_t ref_count; + u32 fa_index; + u16 policer_index; +}; + +static const struct rhashtable_params mlxsw_afa_policer_ht_params = { + .key_len = sizeof(u32), + .key_offset = offsetof(struct mlxsw_afa_policer, fa_index), + .head_offset = offsetof(struct mlxsw_afa_policer, ht_node), + .automatic_shrinking = true, +}; + +struct mlxsw_afa *mlxsw_afa_create(unsigned int max_acts_per_set, + const struct mlxsw_afa_ops *ops, + void *ops_priv) +{ + struct mlxsw_afa *mlxsw_afa; + int err; + + mlxsw_afa = kzalloc(sizeof(*mlxsw_afa), GFP_KERNEL); + if (!mlxsw_afa) + return ERR_PTR(-ENOMEM); + err = rhashtable_init(&mlxsw_afa->set_ht, &mlxsw_afa_set_ht_params); + if (err) + goto err_set_rhashtable_init; + err = rhashtable_init(&mlxsw_afa->fwd_entry_ht, + &mlxsw_afa_fwd_entry_ht_params); + if (err) + goto err_fwd_entry_rhashtable_init; + err = rhashtable_init(&mlxsw_afa->cookie_ht, + &mlxsw_afa_cookie_ht_params); + if (err) + goto err_cookie_rhashtable_init; + err = rhashtable_init(&mlxsw_afa->policer_ht, + &mlxsw_afa_policer_ht_params); + if (err) + goto err_policer_rhashtable_init; + idr_init(&mlxsw_afa->cookie_idr); + INIT_LIST_HEAD(&mlxsw_afa->policer_list); + mlxsw_afa->max_acts_per_set = max_acts_per_set; + mlxsw_afa->ops = ops; + mlxsw_afa->ops_priv = ops_priv; + return mlxsw_afa; + +err_policer_rhashtable_init: + rhashtable_destroy(&mlxsw_afa->cookie_ht); +err_cookie_rhashtable_init: + rhashtable_destroy(&mlxsw_afa->fwd_entry_ht); +err_fwd_entry_rhashtable_init: + rhashtable_destroy(&mlxsw_afa->set_ht); +err_set_rhashtable_init: + kfree(mlxsw_afa); + return ERR_PTR(err); +} +EXPORT_SYMBOL(mlxsw_afa_create); + +void mlxsw_afa_destroy(struct mlxsw_afa *mlxsw_afa) +{ + WARN_ON(!list_empty(&mlxsw_afa->policer_list)); + WARN_ON(!idr_is_empty(&mlxsw_afa->cookie_idr)); + idr_destroy(&mlxsw_afa->cookie_idr); + rhashtable_destroy(&mlxsw_afa->policer_ht); + rhashtable_destroy(&mlxsw_afa->cookie_ht); + rhashtable_destroy(&mlxsw_afa->fwd_entry_ht); + rhashtable_destroy(&mlxsw_afa->set_ht); + kfree(mlxsw_afa); +} +EXPORT_SYMBOL(mlxsw_afa_destroy); + +static void mlxsw_afa_set_goto_set(struct mlxsw_afa_set *set, + enum mlxsw_afa_set_goto_binding_cmd cmd, + u16 group_id) +{ + char *actions = set->ht_key.enc_actions; + + mlxsw_afa_set_type_set(actions, MLXSW_AFA_SET_TYPE_GOTO); + mlxsw_afa_set_goto_g_set(actions, true); + mlxsw_afa_set_goto_binding_cmd_set(actions, cmd); + mlxsw_afa_set_goto_next_binding_set(actions, group_id); +} + +static void mlxsw_afa_set_next_set(struct mlxsw_afa_set *set, + u32 next_set_kvdl_index) +{ + char *actions = set->ht_key.enc_actions; + + mlxsw_afa_set_type_set(actions, MLXSW_AFA_SET_TYPE_NEXT); + mlxsw_afa_set_next_action_set_ptr_set(actions, next_set_kvdl_index); +} + +static struct mlxsw_afa_set *mlxsw_afa_set_create(bool is_first) +{ + struct mlxsw_afa_set *set; + + set = kzalloc(sizeof(*set), GFP_KERNEL); + if (!set) + return NULL; + /* Need to initialize the set to pass by default */ + mlxsw_afa_set_goto_set(set, MLXSW_AFA_SET_GOTO_BINDING_CMD_TERM, 0); + set->ht_key.is_first = is_first; + set->ref_count = 1; + return set; +} + +static void mlxsw_afa_set_destroy(struct mlxsw_afa_set *set) +{ + kfree(set); +} + +static int mlxsw_afa_set_share(struct mlxsw_afa *mlxsw_afa, + struct mlxsw_afa_set *set) +{ + int err; + + err = rhashtable_insert_fast(&mlxsw_afa->set_ht, &set->ht_node, + mlxsw_afa_set_ht_params); + if (err) + return err; + err = mlxsw_afa->ops->kvdl_set_add(mlxsw_afa->ops_priv, + &set->kvdl_index, + set->ht_key.enc_actions, + set->ht_key.is_first); + if (err) + goto err_kvdl_set_add; + set->shared = true; + set->prev = NULL; + return 0; + +err_kvdl_set_add: + rhashtable_remove_fast(&mlxsw_afa->set_ht, &set->ht_node, + mlxsw_afa_set_ht_params); + return err; +} + +static void mlxsw_afa_set_unshare(struct mlxsw_afa *mlxsw_afa, + struct mlxsw_afa_set *set) +{ + mlxsw_afa->ops->kvdl_set_del(mlxsw_afa->ops_priv, + set->kvdl_index, + set->ht_key.is_first); + rhashtable_remove_fast(&mlxsw_afa->set_ht, &set->ht_node, + mlxsw_afa_set_ht_params); + set->shared = false; +} + +static void mlxsw_afa_set_put(struct mlxsw_afa *mlxsw_afa, + struct mlxsw_afa_set *set) +{ + if (--set->ref_count) + return; + if (set->shared) + mlxsw_afa_set_unshare(mlxsw_afa, set); + mlxsw_afa_set_destroy(set); +} + +static struct mlxsw_afa_set *mlxsw_afa_set_get(struct mlxsw_afa *mlxsw_afa, + struct mlxsw_afa_set *orig_set) +{ + struct mlxsw_afa_set *set; + int err; + + /* There is a hashtable of sets maintained. If a set with the exact + * same encoding exists, we reuse it. Otherwise, the current set + * is shared by making it available to others using the hash table. + */ + set = rhashtable_lookup_fast(&mlxsw_afa->set_ht, &orig_set->ht_key, + mlxsw_afa_set_ht_params); + if (set) { + set->ref_count++; + mlxsw_afa_set_put(mlxsw_afa, orig_set); + } else { + set = orig_set; + err = mlxsw_afa_set_share(mlxsw_afa, set); + if (err) + return ERR_PTR(err); + } + return set; +} + +/* Block structure holds a list of action sets. One action block + * represents one chain of actions executed upon match of a rule. + */ + +struct mlxsw_afa_block { + struct mlxsw_afa *afa; + bool finished; + struct mlxsw_afa_set *first_set; + struct mlxsw_afa_set *cur_set; + unsigned int cur_act_index; /* In current set. */ + struct list_head resource_list; /* List of resources held by actions + * in this block. + */ +}; + +struct mlxsw_afa_resource { + struct list_head list; + void (*destructor)(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource); +}; + +static void mlxsw_afa_resource_add(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource) +{ + list_add(&resource->list, &block->resource_list); +} + +static void mlxsw_afa_resource_del(struct mlxsw_afa_resource *resource) +{ + list_del(&resource->list); +} + +static void mlxsw_afa_resources_destroy(struct mlxsw_afa_block *block) +{ + struct mlxsw_afa_resource *resource, *tmp; + + list_for_each_entry_safe(resource, tmp, &block->resource_list, list) { + resource->destructor(block, resource); + } +} + +struct mlxsw_afa_block *mlxsw_afa_block_create(struct mlxsw_afa *mlxsw_afa) +{ + struct mlxsw_afa_block *block; + + block = kzalloc(sizeof(*block), GFP_KERNEL); + if (!block) + return ERR_PTR(-ENOMEM); + INIT_LIST_HEAD(&block->resource_list); + block->afa = mlxsw_afa; + + /* At least one action set is always present, so just create it here */ + block->first_set = mlxsw_afa_set_create(true); + if (!block->first_set) + goto err_first_set_create; + + /* In case user instructs to have dummy first set, we leave it + * empty here and create another, real, set right away. + */ + if (mlxsw_afa->ops->dummy_first_set) { + block->cur_set = mlxsw_afa_set_create(false); + if (!block->cur_set) + goto err_second_set_create; + block->cur_set->prev = block->first_set; + block->first_set->next = block->cur_set; + } else { + block->cur_set = block->first_set; + } + + return block; + +err_second_set_create: + mlxsw_afa_set_destroy(block->first_set); +err_first_set_create: + kfree(block); + return ERR_PTR(-ENOMEM); +} +EXPORT_SYMBOL(mlxsw_afa_block_create); + +void mlxsw_afa_block_destroy(struct mlxsw_afa_block *block) +{ + struct mlxsw_afa_set *set = block->first_set; + struct mlxsw_afa_set *next_set; + + do { + next_set = set->next; + mlxsw_afa_set_put(block->afa, set); + set = next_set; + } while (set); + mlxsw_afa_resources_destroy(block); + kfree(block); +} +EXPORT_SYMBOL(mlxsw_afa_block_destroy); + +int mlxsw_afa_block_commit(struct mlxsw_afa_block *block) +{ + struct mlxsw_afa_set *set = block->cur_set; + struct mlxsw_afa_set *prev_set; + + block->cur_set = NULL; + block->finished = true; + + /* Go over all linked sets starting from last + * and try to find existing set in the hash table. + * In case it is not there, assign a KVD linear index + * and insert it. + */ + do { + prev_set = set->prev; + set = mlxsw_afa_set_get(block->afa, set); + if (IS_ERR(set)) + /* No rollback is needed since the chain is + * in consistent state and mlxsw_afa_block_destroy + * will take care of putting it away. + */ + return PTR_ERR(set); + if (prev_set) { + prev_set->next = set; + mlxsw_afa_set_next_set(prev_set, set->kvdl_index); + set = prev_set; + } + } while (prev_set); + + block->first_set = set; + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_commit); + +char *mlxsw_afa_block_first_set(struct mlxsw_afa_block *block) +{ + return block->first_set->ht_key.enc_actions; +} +EXPORT_SYMBOL(mlxsw_afa_block_first_set); + +char *mlxsw_afa_block_cur_set(struct mlxsw_afa_block *block) +{ + return block->cur_set->ht_key.enc_actions; +} +EXPORT_SYMBOL(mlxsw_afa_block_cur_set); + +u32 mlxsw_afa_block_first_kvdl_index(struct mlxsw_afa_block *block) +{ + /* First set is never in KVD linear. So the first set + * with valid KVD linear index is always the second one. + */ + if (WARN_ON(!block->first_set->next)) + return 0; + return block->first_set->next->kvdl_index; +} +EXPORT_SYMBOL(mlxsw_afa_block_first_kvdl_index); + +int mlxsw_afa_block_activity_get(struct mlxsw_afa_block *block, bool *activity) +{ + u32 kvdl_index = mlxsw_afa_block_first_kvdl_index(block); + + return block->afa->ops->kvdl_set_activity_get(block->afa->ops_priv, + kvdl_index, activity); +} +EXPORT_SYMBOL(mlxsw_afa_block_activity_get); + +int mlxsw_afa_block_continue(struct mlxsw_afa_block *block) +{ + if (block->finished) + return -EINVAL; + mlxsw_afa_set_goto_set(block->cur_set, + MLXSW_AFA_SET_GOTO_BINDING_CMD_NONE, 0); + block->finished = true; + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_continue); + +int mlxsw_afa_block_jump(struct mlxsw_afa_block *block, u16 group_id) +{ + if (block->finished) + return -EINVAL; + mlxsw_afa_set_goto_set(block->cur_set, + MLXSW_AFA_SET_GOTO_BINDING_CMD_JUMP, group_id); + block->finished = true; + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_jump); + +int mlxsw_afa_block_terminate(struct mlxsw_afa_block *block) +{ + if (block->finished) + return -EINVAL; + mlxsw_afa_set_goto_set(block->cur_set, + MLXSW_AFA_SET_GOTO_BINDING_CMD_TERM, 0); + block->finished = true; + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_terminate); + +static struct mlxsw_afa_fwd_entry * +mlxsw_afa_fwd_entry_create(struct mlxsw_afa *mlxsw_afa, u16 local_port) +{ + struct mlxsw_afa_fwd_entry *fwd_entry; + int err; + + fwd_entry = kzalloc(sizeof(*fwd_entry), GFP_KERNEL); + if (!fwd_entry) + return ERR_PTR(-ENOMEM); + fwd_entry->ht_key.local_port = local_port; + fwd_entry->ref_count = 1; + + err = rhashtable_insert_fast(&mlxsw_afa->fwd_entry_ht, + &fwd_entry->ht_node, + mlxsw_afa_fwd_entry_ht_params); + if (err) + goto err_rhashtable_insert; + + err = mlxsw_afa->ops->kvdl_fwd_entry_add(mlxsw_afa->ops_priv, + &fwd_entry->kvdl_index, + local_port); + if (err) + goto err_kvdl_fwd_entry_add; + return fwd_entry; + +err_kvdl_fwd_entry_add: + rhashtable_remove_fast(&mlxsw_afa->fwd_entry_ht, &fwd_entry->ht_node, + mlxsw_afa_fwd_entry_ht_params); +err_rhashtable_insert: + kfree(fwd_entry); + return ERR_PTR(err); +} + +static void mlxsw_afa_fwd_entry_destroy(struct mlxsw_afa *mlxsw_afa, + struct mlxsw_afa_fwd_entry *fwd_entry) +{ + mlxsw_afa->ops->kvdl_fwd_entry_del(mlxsw_afa->ops_priv, + fwd_entry->kvdl_index); + rhashtable_remove_fast(&mlxsw_afa->fwd_entry_ht, &fwd_entry->ht_node, + mlxsw_afa_fwd_entry_ht_params); + kfree(fwd_entry); +} + +static struct mlxsw_afa_fwd_entry * +mlxsw_afa_fwd_entry_get(struct mlxsw_afa *mlxsw_afa, u16 local_port) +{ + struct mlxsw_afa_fwd_entry_ht_key ht_key = {0}; + struct mlxsw_afa_fwd_entry *fwd_entry; + + ht_key.local_port = local_port; + fwd_entry = rhashtable_lookup_fast(&mlxsw_afa->fwd_entry_ht, &ht_key, + mlxsw_afa_fwd_entry_ht_params); + if (fwd_entry) { + fwd_entry->ref_count++; + return fwd_entry; + } + return mlxsw_afa_fwd_entry_create(mlxsw_afa, local_port); +} + +static void mlxsw_afa_fwd_entry_put(struct mlxsw_afa *mlxsw_afa, + struct mlxsw_afa_fwd_entry *fwd_entry) +{ + if (--fwd_entry->ref_count) + return; + mlxsw_afa_fwd_entry_destroy(mlxsw_afa, fwd_entry); +} + +struct mlxsw_afa_fwd_entry_ref { + struct mlxsw_afa_resource resource; + struct mlxsw_afa_fwd_entry *fwd_entry; +}; + +static void +mlxsw_afa_fwd_entry_ref_destroy(struct mlxsw_afa_block *block, + struct mlxsw_afa_fwd_entry_ref *fwd_entry_ref) +{ + mlxsw_afa_resource_del(&fwd_entry_ref->resource); + mlxsw_afa_fwd_entry_put(block->afa, fwd_entry_ref->fwd_entry); + kfree(fwd_entry_ref); +} + +static void +mlxsw_afa_fwd_entry_ref_destructor(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource) +{ + struct mlxsw_afa_fwd_entry_ref *fwd_entry_ref; + + fwd_entry_ref = container_of(resource, struct mlxsw_afa_fwd_entry_ref, + resource); + mlxsw_afa_fwd_entry_ref_destroy(block, fwd_entry_ref); +} + +static struct mlxsw_afa_fwd_entry_ref * +mlxsw_afa_fwd_entry_ref_create(struct mlxsw_afa_block *block, u16 local_port) +{ + struct mlxsw_afa_fwd_entry_ref *fwd_entry_ref; + struct mlxsw_afa_fwd_entry *fwd_entry; + int err; + + fwd_entry_ref = kzalloc(sizeof(*fwd_entry_ref), GFP_KERNEL); + if (!fwd_entry_ref) + return ERR_PTR(-ENOMEM); + fwd_entry = mlxsw_afa_fwd_entry_get(block->afa, local_port); + if (IS_ERR(fwd_entry)) { + err = PTR_ERR(fwd_entry); + goto err_fwd_entry_get; + } + fwd_entry_ref->fwd_entry = fwd_entry; + fwd_entry_ref->resource.destructor = mlxsw_afa_fwd_entry_ref_destructor; + mlxsw_afa_resource_add(block, &fwd_entry_ref->resource); + return fwd_entry_ref; + +err_fwd_entry_get: + kfree(fwd_entry_ref); + return ERR_PTR(err); +} + +struct mlxsw_afa_counter { + struct mlxsw_afa_resource resource; + u32 counter_index; +}; + +static void +mlxsw_afa_counter_destroy(struct mlxsw_afa_block *block, + struct mlxsw_afa_counter *counter) +{ + mlxsw_afa_resource_del(&counter->resource); + block->afa->ops->counter_index_put(block->afa->ops_priv, + counter->counter_index); + kfree(counter); +} + +static void +mlxsw_afa_counter_destructor(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource) +{ + struct mlxsw_afa_counter *counter; + + counter = container_of(resource, struct mlxsw_afa_counter, resource); + mlxsw_afa_counter_destroy(block, counter); +} + +static struct mlxsw_afa_counter * +mlxsw_afa_counter_create(struct mlxsw_afa_block *block) +{ + struct mlxsw_afa_counter *counter; + int err; + + counter = kzalloc(sizeof(*counter), GFP_KERNEL); + if (!counter) + return ERR_PTR(-ENOMEM); + + err = block->afa->ops->counter_index_get(block->afa->ops_priv, + &counter->counter_index); + if (err) + goto err_counter_index_get; + counter->resource.destructor = mlxsw_afa_counter_destructor; + mlxsw_afa_resource_add(block, &counter->resource); + return counter; + +err_counter_index_get: + kfree(counter); + return ERR_PTR(err); +} + +/* 20 bits is a maximum that hardware can handle in trap with userdef action + * and carry along with the trapped packet. + */ +#define MLXSW_AFA_COOKIE_INDEX_BITS 20 +#define MLXSW_AFA_COOKIE_INDEX_MAX ((1 << MLXSW_AFA_COOKIE_INDEX_BITS) - 1) + +static struct mlxsw_afa_cookie * +mlxsw_afa_cookie_create(struct mlxsw_afa *mlxsw_afa, + const struct flow_action_cookie *fa_cookie) +{ + struct mlxsw_afa_cookie *cookie; + u32 cookie_index; + int err; + + cookie = kzalloc(sizeof(*cookie) + fa_cookie->cookie_len, GFP_KERNEL); + if (!cookie) + return ERR_PTR(-ENOMEM); + refcount_set(&cookie->ref_count, 1); + cookie->fa_cookie = *fa_cookie; + memcpy(cookie->fa_cookie.cookie, fa_cookie->cookie, + fa_cookie->cookie_len); + + err = rhashtable_insert_fast(&mlxsw_afa->cookie_ht, &cookie->ht_node, + mlxsw_afa_cookie_ht_params); + if (err) + goto err_rhashtable_insert; + + /* Start cookie indexes with 1. Leave the 0 index unused. Packets + * that come from the HW which are not dropped by drop-with-cookie + * action are going to pass cookie_index 0 to lookup. + */ + cookie_index = 1; + err = idr_alloc_u32(&mlxsw_afa->cookie_idr, cookie, &cookie_index, + MLXSW_AFA_COOKIE_INDEX_MAX, GFP_KERNEL); + if (err) + goto err_idr_alloc; + cookie->cookie_index = cookie_index; + return cookie; + +err_idr_alloc: + rhashtable_remove_fast(&mlxsw_afa->cookie_ht, &cookie->ht_node, + mlxsw_afa_cookie_ht_params); +err_rhashtable_insert: + kfree(cookie); + return ERR_PTR(err); +} + +static void mlxsw_afa_cookie_destroy(struct mlxsw_afa *mlxsw_afa, + struct mlxsw_afa_cookie *cookie) +{ + idr_remove(&mlxsw_afa->cookie_idr, cookie->cookie_index); + rhashtable_remove_fast(&mlxsw_afa->cookie_ht, &cookie->ht_node, + mlxsw_afa_cookie_ht_params); + kfree_rcu(cookie, rcu); +} + +static struct mlxsw_afa_cookie * +mlxsw_afa_cookie_get(struct mlxsw_afa *mlxsw_afa, + const struct flow_action_cookie *fa_cookie) +{ + struct mlxsw_afa_cookie *cookie; + + cookie = rhashtable_lookup_fast(&mlxsw_afa->cookie_ht, fa_cookie, + mlxsw_afa_cookie_ht_params); + if (cookie) { + refcount_inc(&cookie->ref_count); + return cookie; + } + return mlxsw_afa_cookie_create(mlxsw_afa, fa_cookie); +} + +static void mlxsw_afa_cookie_put(struct mlxsw_afa *mlxsw_afa, + struct mlxsw_afa_cookie *cookie) +{ + if (!refcount_dec_and_test(&cookie->ref_count)) + return; + mlxsw_afa_cookie_destroy(mlxsw_afa, cookie); +} + +/* RCU read lock must be held */ +const struct flow_action_cookie * +mlxsw_afa_cookie_lookup(struct mlxsw_afa *mlxsw_afa, u32 cookie_index) +{ + struct mlxsw_afa_cookie *cookie; + + /* 0 index means no cookie */ + if (!cookie_index) + return NULL; + cookie = idr_find(&mlxsw_afa->cookie_idr, cookie_index); + if (!cookie) + return NULL; + return &cookie->fa_cookie; +} +EXPORT_SYMBOL(mlxsw_afa_cookie_lookup); + +struct mlxsw_afa_cookie_ref { + struct mlxsw_afa_resource resource; + struct mlxsw_afa_cookie *cookie; +}; + +static void +mlxsw_afa_cookie_ref_destroy(struct mlxsw_afa_block *block, + struct mlxsw_afa_cookie_ref *cookie_ref) +{ + mlxsw_afa_resource_del(&cookie_ref->resource); + mlxsw_afa_cookie_put(block->afa, cookie_ref->cookie); + kfree(cookie_ref); +} + +static void +mlxsw_afa_cookie_ref_destructor(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource) +{ + struct mlxsw_afa_cookie_ref *cookie_ref; + + cookie_ref = container_of(resource, struct mlxsw_afa_cookie_ref, + resource); + mlxsw_afa_cookie_ref_destroy(block, cookie_ref); +} + +static struct mlxsw_afa_cookie_ref * +mlxsw_afa_cookie_ref_create(struct mlxsw_afa_block *block, + const struct flow_action_cookie *fa_cookie) +{ + struct mlxsw_afa_cookie_ref *cookie_ref; + struct mlxsw_afa_cookie *cookie; + int err; + + cookie_ref = kzalloc(sizeof(*cookie_ref), GFP_KERNEL); + if (!cookie_ref) + return ERR_PTR(-ENOMEM); + cookie = mlxsw_afa_cookie_get(block->afa, fa_cookie); + if (IS_ERR(cookie)) { + err = PTR_ERR(cookie); + goto err_cookie_get; + } + cookie_ref->cookie = cookie; + cookie_ref->resource.destructor = mlxsw_afa_cookie_ref_destructor; + mlxsw_afa_resource_add(block, &cookie_ref->resource); + return cookie_ref; + +err_cookie_get: + kfree(cookie_ref); + return ERR_PTR(err); +} + +static struct mlxsw_afa_policer * +mlxsw_afa_policer_create(struct mlxsw_afa *mlxsw_afa, u32 fa_index, + u64 rate_bytes_ps, u32 burst, + struct netlink_ext_ack *extack) +{ + struct mlxsw_afa_policer *policer; + int err; + + policer = kzalloc(sizeof(*policer), GFP_KERNEL); + if (!policer) + return ERR_PTR(-ENOMEM); + + err = mlxsw_afa->ops->policer_add(mlxsw_afa->ops_priv, rate_bytes_ps, + burst, &policer->policer_index, + extack); + if (err) + goto err_policer_add; + + refcount_set(&policer->ref_count, 1); + policer->fa_index = fa_index; + + err = rhashtable_insert_fast(&mlxsw_afa->policer_ht, &policer->ht_node, + mlxsw_afa_policer_ht_params); + if (err) + goto err_rhashtable_insert; + + list_add_tail(&policer->list, &mlxsw_afa->policer_list); + + return policer; + +err_rhashtable_insert: + mlxsw_afa->ops->policer_del(mlxsw_afa->ops_priv, + policer->policer_index); +err_policer_add: + kfree(policer); + return ERR_PTR(err); +} + +static void mlxsw_afa_policer_destroy(struct mlxsw_afa *mlxsw_afa, + struct mlxsw_afa_policer *policer) +{ + list_del(&policer->list); + rhashtable_remove_fast(&mlxsw_afa->policer_ht, &policer->ht_node, + mlxsw_afa_policer_ht_params); + mlxsw_afa->ops->policer_del(mlxsw_afa->ops_priv, + policer->policer_index); + kfree(policer); +} + +static struct mlxsw_afa_policer * +mlxsw_afa_policer_get(struct mlxsw_afa *mlxsw_afa, u32 fa_index, + u64 rate_bytes_ps, u32 burst, + struct netlink_ext_ack *extack) +{ + struct mlxsw_afa_policer *policer; + + policer = rhashtable_lookup_fast(&mlxsw_afa->policer_ht, &fa_index, + mlxsw_afa_policer_ht_params); + if (policer) { + refcount_inc(&policer->ref_count); + return policer; + } + + return mlxsw_afa_policer_create(mlxsw_afa, fa_index, rate_bytes_ps, + burst, extack); +} + +static void mlxsw_afa_policer_put(struct mlxsw_afa *mlxsw_afa, + struct mlxsw_afa_policer *policer) +{ + if (!refcount_dec_and_test(&policer->ref_count)) + return; + mlxsw_afa_policer_destroy(mlxsw_afa, policer); +} + +struct mlxsw_afa_policer_ref { + struct mlxsw_afa_resource resource; + struct mlxsw_afa_policer *policer; +}; + +static void +mlxsw_afa_policer_ref_destroy(struct mlxsw_afa_block *block, + struct mlxsw_afa_policer_ref *policer_ref) +{ + mlxsw_afa_resource_del(&policer_ref->resource); + mlxsw_afa_policer_put(block->afa, policer_ref->policer); + kfree(policer_ref); +} + +static void +mlxsw_afa_policer_ref_destructor(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource) +{ + struct mlxsw_afa_policer_ref *policer_ref; + + policer_ref = container_of(resource, struct mlxsw_afa_policer_ref, + resource); + mlxsw_afa_policer_ref_destroy(block, policer_ref); +} + +static struct mlxsw_afa_policer_ref * +mlxsw_afa_policer_ref_create(struct mlxsw_afa_block *block, u32 fa_index, + u64 rate_bytes_ps, u32 burst, + struct netlink_ext_ack *extack) +{ + struct mlxsw_afa_policer_ref *policer_ref; + struct mlxsw_afa_policer *policer; + int err; + + policer_ref = kzalloc(sizeof(*policer_ref), GFP_KERNEL); + if (!policer_ref) + return ERR_PTR(-ENOMEM); + + policer = mlxsw_afa_policer_get(block->afa, fa_index, rate_bytes_ps, + burst, extack); + if (IS_ERR(policer)) { + err = PTR_ERR(policer); + goto err_policer_get; + } + + policer_ref->policer = policer; + policer_ref->resource.destructor = mlxsw_afa_policer_ref_destructor; + mlxsw_afa_resource_add(block, &policer_ref->resource); + + return policer_ref; + +err_policer_get: + kfree(policer_ref); + return ERR_PTR(err); +} + +#define MLXSW_AFA_ONE_ACTION_LEN 32 +#define MLXSW_AFA_PAYLOAD_OFFSET 4 + +enum mlxsw_afa_action_type { + MLXSW_AFA_ACTION_TYPE_TRAP, + MLXSW_AFA_ACTION_TYPE_POLICE, + MLXSW_AFA_ACTION_TYPE_OTHER, +}; + +static bool +mlxsw_afa_block_need_split(const struct mlxsw_afa_block *block, + enum mlxsw_afa_action_type type) +{ + struct mlxsw_afa_set *cur_set = block->cur_set; + + /* Due to a hardware limitation, police action cannot be in the same + * action set with MLXSW_AFA_TRAP_CODE or MLXSW_AFA_TRAPWU_CODE + * actions. Work around this limitation by creating a new action set + * and place the new action there. + */ + return (cur_set->has_trap && type == MLXSW_AFA_ACTION_TYPE_POLICE) || + (cur_set->has_police && type == MLXSW_AFA_ACTION_TYPE_TRAP); +} + +static char *mlxsw_afa_block_append_action_ext(struct mlxsw_afa_block *block, + u8 action_code, u8 action_size, + enum mlxsw_afa_action_type type) +{ + char *oneact; + char *actions; + + if (block->finished) + return ERR_PTR(-EINVAL); + if (block->cur_act_index + action_size > block->afa->max_acts_per_set || + mlxsw_afa_block_need_split(block, type)) { + struct mlxsw_afa_set *set; + + /* The appended action won't fit into the current action set, + * so create a new set. + */ + set = mlxsw_afa_set_create(false); + if (!set) + return ERR_PTR(-ENOBUFS); + set->prev = block->cur_set; + block->cur_act_index = 0; + block->cur_set->next = set; + block->cur_set = set; + } + + switch (type) { + case MLXSW_AFA_ACTION_TYPE_TRAP: + block->cur_set->has_trap = true; + break; + case MLXSW_AFA_ACTION_TYPE_POLICE: + block->cur_set->has_police = true; + break; + default: + break; + } + + actions = block->cur_set->ht_key.enc_actions; + oneact = actions + block->cur_act_index * MLXSW_AFA_ONE_ACTION_LEN; + block->cur_act_index += action_size; + mlxsw_afa_all_action_type_set(oneact, action_code); + return oneact + MLXSW_AFA_PAYLOAD_OFFSET; +} + +static char *mlxsw_afa_block_append_action(struct mlxsw_afa_block *block, + u8 action_code, u8 action_size) +{ + return mlxsw_afa_block_append_action_ext(block, action_code, + action_size, + MLXSW_AFA_ACTION_TYPE_OTHER); +} + +/* VLAN Action + * ----------- + * VLAN action is used for manipulating VLANs. It can be used to implement QinQ, + * VLAN translation, change of PCP bits of the VLAN tag, push, pop as swap VLANs + * and more. + */ + +#define MLXSW_AFA_VLAN_CODE 0x02 +#define MLXSW_AFA_VLAN_SIZE 1 + +enum mlxsw_afa_vlan_vlan_tag_cmd { + MLXSW_AFA_VLAN_VLAN_TAG_CMD_NOP, + MLXSW_AFA_VLAN_VLAN_TAG_CMD_PUSH_TAG, + MLXSW_AFA_VLAN_VLAN_TAG_CMD_POP_TAG, +}; + +enum mlxsw_afa_vlan_cmd { + MLXSW_AFA_VLAN_CMD_NOP, + MLXSW_AFA_VLAN_CMD_SET_OUTER, + MLXSW_AFA_VLAN_CMD_SET_INNER, + MLXSW_AFA_VLAN_CMD_COPY_OUTER_TO_INNER, + MLXSW_AFA_VLAN_CMD_COPY_INNER_TO_OUTER, + MLXSW_AFA_VLAN_CMD_SWAP, +}; + +/* afa_vlan_vlan_tag_cmd + * Tag command: push, pop, nop VLAN header. + */ +MLXSW_ITEM32(afa, vlan, vlan_tag_cmd, 0x00, 29, 3); + +/* afa_vlan_vid_cmd */ +MLXSW_ITEM32(afa, vlan, vid_cmd, 0x04, 29, 3); + +/* afa_vlan_vid */ +MLXSW_ITEM32(afa, vlan, vid, 0x04, 0, 12); + +/* afa_vlan_ethertype_cmd */ +MLXSW_ITEM32(afa, vlan, ethertype_cmd, 0x08, 29, 3); + +/* afa_vlan_ethertype + * Index to EtherTypes in Switch VLAN EtherType Register (SVER). + */ +MLXSW_ITEM32(afa, vlan, ethertype, 0x08, 24, 3); + +/* afa_vlan_pcp_cmd */ +MLXSW_ITEM32(afa, vlan, pcp_cmd, 0x08, 13, 3); + +/* afa_vlan_pcp */ +MLXSW_ITEM32(afa, vlan, pcp, 0x08, 8, 3); + +static inline void +mlxsw_afa_vlan_pack(char *payload, + enum mlxsw_afa_vlan_vlan_tag_cmd vlan_tag_cmd, + enum mlxsw_afa_vlan_cmd vid_cmd, u16 vid, + enum mlxsw_afa_vlan_cmd pcp_cmd, u8 pcp, + enum mlxsw_afa_vlan_cmd ethertype_cmd, u8 ethertype) +{ + mlxsw_afa_vlan_vlan_tag_cmd_set(payload, vlan_tag_cmd); + mlxsw_afa_vlan_vid_cmd_set(payload, vid_cmd); + mlxsw_afa_vlan_vid_set(payload, vid); + mlxsw_afa_vlan_pcp_cmd_set(payload, pcp_cmd); + mlxsw_afa_vlan_pcp_set(payload, pcp); + mlxsw_afa_vlan_ethertype_cmd_set(payload, ethertype_cmd); + mlxsw_afa_vlan_ethertype_set(payload, ethertype); +} + +int mlxsw_afa_block_append_vlan_modify(struct mlxsw_afa_block *block, + u16 vid, u8 pcp, u8 et, + struct netlink_ext_ack *extack) +{ + char *act = mlxsw_afa_block_append_action(block, + MLXSW_AFA_VLAN_CODE, + MLXSW_AFA_VLAN_SIZE); + + if (IS_ERR(act)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append vlan_modify action"); + return PTR_ERR(act); + } + mlxsw_afa_vlan_pack(act, MLXSW_AFA_VLAN_VLAN_TAG_CMD_NOP, + MLXSW_AFA_VLAN_CMD_SET_OUTER, vid, + MLXSW_AFA_VLAN_CMD_SET_OUTER, pcp, + MLXSW_AFA_VLAN_CMD_SET_OUTER, et); + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_vlan_modify); + +/* Trap Action / Trap With Userdef Action + * -------------------------------------- + * The Trap action enables trapping / mirroring packets to the CPU + * as well as discarding packets. + * The ACL Trap / Discard separates the forward/discard control from CPU + * trap control. In addition, the Trap / Discard action enables activating + * SPAN (port mirroring). + * + * The Trap with userdef action has the same functionality as + * the Trap action with addition of user defined value that can be set + * and used by higher layer applications. + */ + +#define MLXSW_AFA_TRAP_CODE 0x03 +#define MLXSW_AFA_TRAP_SIZE 1 + +#define MLXSW_AFA_TRAPWU_CODE 0x04 +#define MLXSW_AFA_TRAPWU_SIZE 2 + +enum mlxsw_afa_trap_trap_action { + MLXSW_AFA_TRAP_TRAP_ACTION_NOP = 0, + MLXSW_AFA_TRAP_TRAP_ACTION_TRAP = 2, +}; + +/* afa_trap_trap_action + * Trap Action. + */ +MLXSW_ITEM32(afa, trap, trap_action, 0x00, 24, 4); + +enum mlxsw_afa_trap_forward_action { + MLXSW_AFA_TRAP_FORWARD_ACTION_FORWARD = 1, + MLXSW_AFA_TRAP_FORWARD_ACTION_DISCARD = 3, +}; + +/* afa_trap_forward_action + * Forward Action. + */ +MLXSW_ITEM32(afa, trap, forward_action, 0x00, 0, 4); + +/* afa_trap_trap_id + * Trap ID to configure. + */ +MLXSW_ITEM32(afa, trap, trap_id, 0x04, 0, 9); + +/* afa_trap_mirror_agent + * Mirror agent. + */ +MLXSW_ITEM32(afa, trap, mirror_agent, 0x08, 29, 3); + +/* afa_trap_mirror_enable + * Mirror enable. + */ +MLXSW_ITEM32(afa, trap, mirror_enable, 0x08, 24, 1); + +/* user_def_val + * Value for the SW usage. Can be used to pass information of which + * rule has caused a trap. This may be overwritten by later traps. + * This field does a set on the packet's user_def_val only if this + * is the first trap_id or if the trap_id has replaced the previous + * packet's trap_id. + */ +MLXSW_ITEM32(afa, trap, user_def_val, 0x0C, 0, 20); + +static inline void +mlxsw_afa_trap_pack(char *payload, + enum mlxsw_afa_trap_trap_action trap_action, + enum mlxsw_afa_trap_forward_action forward_action, + u16 trap_id) +{ + mlxsw_afa_trap_trap_action_set(payload, trap_action); + mlxsw_afa_trap_forward_action_set(payload, forward_action); + mlxsw_afa_trap_trap_id_set(payload, trap_id); +} + +static inline void +mlxsw_afa_trapwu_pack(char *payload, + enum mlxsw_afa_trap_trap_action trap_action, + enum mlxsw_afa_trap_forward_action forward_action, + u16 trap_id, u32 user_def_val) +{ + mlxsw_afa_trap_pack(payload, trap_action, forward_action, trap_id); + mlxsw_afa_trap_user_def_val_set(payload, user_def_val); +} + +static inline void +mlxsw_afa_trap_mirror_pack(char *payload, bool mirror_enable, + u8 mirror_agent) +{ + mlxsw_afa_trap_mirror_enable_set(payload, mirror_enable); + mlxsw_afa_trap_mirror_agent_set(payload, mirror_agent); +} + +static char *mlxsw_afa_block_append_action_trap(struct mlxsw_afa_block *block, + u8 action_code, u8 action_size) +{ + return mlxsw_afa_block_append_action_ext(block, action_code, + action_size, + MLXSW_AFA_ACTION_TYPE_TRAP); +} + +static int mlxsw_afa_block_append_drop_plain(struct mlxsw_afa_block *block, + bool ingress) +{ + char *act = mlxsw_afa_block_append_action_trap(block, + MLXSW_AFA_TRAP_CODE, + MLXSW_AFA_TRAP_SIZE); + + if (IS_ERR(act)) + return PTR_ERR(act); + mlxsw_afa_trap_pack(act, MLXSW_AFA_TRAP_TRAP_ACTION_TRAP, + MLXSW_AFA_TRAP_FORWARD_ACTION_DISCARD, + ingress ? MLXSW_TRAP_ID_DISCARD_INGRESS_ACL : + MLXSW_TRAP_ID_DISCARD_EGRESS_ACL); + return 0; +} + +static int +mlxsw_afa_block_append_drop_with_cookie(struct mlxsw_afa_block *block, + bool ingress, + const struct flow_action_cookie *fa_cookie, + struct netlink_ext_ack *extack) +{ + struct mlxsw_afa_cookie_ref *cookie_ref; + u32 cookie_index; + char *act; + int err; + + cookie_ref = mlxsw_afa_cookie_ref_create(block, fa_cookie); + if (IS_ERR(cookie_ref)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot create cookie for drop action"); + return PTR_ERR(cookie_ref); + } + cookie_index = cookie_ref->cookie->cookie_index; + + act = mlxsw_afa_block_append_action_trap(block, MLXSW_AFA_TRAPWU_CODE, + MLXSW_AFA_TRAPWU_SIZE); + if (IS_ERR(act)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append drop with cookie action"); + err = PTR_ERR(act); + goto err_append_action; + } + mlxsw_afa_trapwu_pack(act, MLXSW_AFA_TRAP_TRAP_ACTION_TRAP, + MLXSW_AFA_TRAP_FORWARD_ACTION_DISCARD, + ingress ? MLXSW_TRAP_ID_DISCARD_INGRESS_ACL : + MLXSW_TRAP_ID_DISCARD_EGRESS_ACL, + cookie_index); + return 0; + +err_append_action: + mlxsw_afa_cookie_ref_destroy(block, cookie_ref); + return err; +} + +int mlxsw_afa_block_append_drop(struct mlxsw_afa_block *block, bool ingress, + const struct flow_action_cookie *fa_cookie, + struct netlink_ext_ack *extack) +{ + return fa_cookie ? + mlxsw_afa_block_append_drop_with_cookie(block, ingress, + fa_cookie, extack) : + mlxsw_afa_block_append_drop_plain(block, ingress); +} +EXPORT_SYMBOL(mlxsw_afa_block_append_drop); + +int mlxsw_afa_block_append_trap(struct mlxsw_afa_block *block, u16 trap_id) +{ + char *act = mlxsw_afa_block_append_action_trap(block, + MLXSW_AFA_TRAP_CODE, + MLXSW_AFA_TRAP_SIZE); + + if (IS_ERR(act)) + return PTR_ERR(act); + mlxsw_afa_trap_pack(act, MLXSW_AFA_TRAP_TRAP_ACTION_TRAP, + MLXSW_AFA_TRAP_FORWARD_ACTION_DISCARD, trap_id); + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_trap); + +int mlxsw_afa_block_append_trap_and_forward(struct mlxsw_afa_block *block, + u16 trap_id) +{ + char *act = mlxsw_afa_block_append_action_trap(block, + MLXSW_AFA_TRAP_CODE, + MLXSW_AFA_TRAP_SIZE); + + if (IS_ERR(act)) + return PTR_ERR(act); + mlxsw_afa_trap_pack(act, MLXSW_AFA_TRAP_TRAP_ACTION_TRAP, + MLXSW_AFA_TRAP_FORWARD_ACTION_FORWARD, trap_id); + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_trap_and_forward); + +struct mlxsw_afa_mirror { + struct mlxsw_afa_resource resource; + int span_id; + u16 local_in_port; + bool ingress; +}; + +static void +mlxsw_afa_mirror_destroy(struct mlxsw_afa_block *block, + struct mlxsw_afa_mirror *mirror) +{ + mlxsw_afa_resource_del(&mirror->resource); + block->afa->ops->mirror_del(block->afa->ops_priv, + mirror->local_in_port, + mirror->span_id, + mirror->ingress); + kfree(mirror); +} + +static void +mlxsw_afa_mirror_destructor(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource) +{ + struct mlxsw_afa_mirror *mirror; + + mirror = container_of(resource, struct mlxsw_afa_mirror, resource); + mlxsw_afa_mirror_destroy(block, mirror); +} + +static struct mlxsw_afa_mirror * +mlxsw_afa_mirror_create(struct mlxsw_afa_block *block, u16 local_in_port, + const struct net_device *out_dev, bool ingress) +{ + struct mlxsw_afa_mirror *mirror; + int err; + + mirror = kzalloc(sizeof(*mirror), GFP_KERNEL); + if (!mirror) + return ERR_PTR(-ENOMEM); + + err = block->afa->ops->mirror_add(block->afa->ops_priv, + local_in_port, out_dev, + ingress, &mirror->span_id); + if (err) + goto err_mirror_add; + + mirror->ingress = ingress; + mirror->local_in_port = local_in_port; + mirror->resource.destructor = mlxsw_afa_mirror_destructor; + mlxsw_afa_resource_add(block, &mirror->resource); + return mirror; + +err_mirror_add: + kfree(mirror); + return ERR_PTR(err); +} + +static int +mlxsw_afa_block_append_allocated_mirror(struct mlxsw_afa_block *block, + u8 mirror_agent) +{ + char *act = mlxsw_afa_block_append_action_trap(block, + MLXSW_AFA_TRAP_CODE, + MLXSW_AFA_TRAP_SIZE); + + if (IS_ERR(act)) + return PTR_ERR(act); + mlxsw_afa_trap_pack(act, MLXSW_AFA_TRAP_TRAP_ACTION_NOP, + MLXSW_AFA_TRAP_FORWARD_ACTION_FORWARD, 0); + mlxsw_afa_trap_mirror_pack(act, true, mirror_agent); + return 0; +} + +int +mlxsw_afa_block_append_mirror(struct mlxsw_afa_block *block, u16 local_in_port, + const struct net_device *out_dev, bool ingress, + struct netlink_ext_ack *extack) +{ + struct mlxsw_afa_mirror *mirror; + int err; + + mirror = mlxsw_afa_mirror_create(block, local_in_port, out_dev, + ingress); + if (IS_ERR(mirror)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot create mirror action"); + return PTR_ERR(mirror); + } + err = mlxsw_afa_block_append_allocated_mirror(block, mirror->span_id); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append mirror action"); + goto err_append_allocated_mirror; + } + + return 0; + +err_append_allocated_mirror: + mlxsw_afa_mirror_destroy(block, mirror); + return err; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_mirror); + +/* QoS Action + * ---------- + * The QOS_ACTION is used for manipulating the QoS attributes of a packet. It + * can be used to change the DCSP, ECN, Color and Switch Priority of the packet. + * Note that PCP field can be changed using the VLAN action. + */ + +#define MLXSW_AFA_QOS_CODE 0x06 +#define MLXSW_AFA_QOS_SIZE 1 + +enum mlxsw_afa_qos_ecn_cmd { + /* Do nothing */ + MLXSW_AFA_QOS_ECN_CMD_NOP, + /* Set ECN to afa_qos_ecn */ + MLXSW_AFA_QOS_ECN_CMD_SET, +}; + +/* afa_qos_ecn_cmd + */ +MLXSW_ITEM32(afa, qos, ecn_cmd, 0x04, 29, 3); + +/* afa_qos_ecn + * ECN value. + */ +MLXSW_ITEM32(afa, qos, ecn, 0x04, 24, 2); + +enum mlxsw_afa_qos_dscp_cmd { + /* Do nothing */ + MLXSW_AFA_QOS_DSCP_CMD_NOP, + /* Set DSCP 3 LSB bits according to dscp[2:0] */ + MLXSW_AFA_QOS_DSCP_CMD_SET_3LSB, + /* Set DSCP 3 MSB bits according to dscp[5:3] */ + MLXSW_AFA_QOS_DSCP_CMD_SET_3MSB, + /* Set DSCP 6 bits according to dscp[5:0] */ + MLXSW_AFA_QOS_DSCP_CMD_SET_ALL, +}; + +/* afa_qos_dscp_cmd + * DSCP command. + */ +MLXSW_ITEM32(afa, qos, dscp_cmd, 0x04, 14, 2); + +/* afa_qos_dscp + * DSCP value. + */ +MLXSW_ITEM32(afa, qos, dscp, 0x04, 0, 6); + +enum mlxsw_afa_qos_switch_prio_cmd { + /* Do nothing */ + MLXSW_AFA_QOS_SWITCH_PRIO_CMD_NOP, + /* Set Switch Priority to afa_qos_switch_prio */ + MLXSW_AFA_QOS_SWITCH_PRIO_CMD_SET, +}; + +/* afa_qos_switch_prio_cmd + */ +MLXSW_ITEM32(afa, qos, switch_prio_cmd, 0x08, 14, 2); + +/* afa_qos_switch_prio + * Switch Priority. + */ +MLXSW_ITEM32(afa, qos, switch_prio, 0x08, 0, 4); + +enum mlxsw_afa_qos_dscp_rw { + MLXSW_AFA_QOS_DSCP_RW_PRESERVE, + MLXSW_AFA_QOS_DSCP_RW_SET, + MLXSW_AFA_QOS_DSCP_RW_CLEAR, +}; + +/* afa_qos_dscp_rw + * DSCP Re-write Enable. Controlling the rewrite_enable for DSCP. + */ +MLXSW_ITEM32(afa, qos, dscp_rw, 0x0C, 30, 2); + +static inline void +mlxsw_afa_qos_ecn_pack(char *payload, + enum mlxsw_afa_qos_ecn_cmd ecn_cmd, u8 ecn) +{ + mlxsw_afa_qos_ecn_cmd_set(payload, ecn_cmd); + mlxsw_afa_qos_ecn_set(payload, ecn); +} + +static inline void +mlxsw_afa_qos_dscp_pack(char *payload, + enum mlxsw_afa_qos_dscp_cmd dscp_cmd, u8 dscp) +{ + mlxsw_afa_qos_dscp_cmd_set(payload, dscp_cmd); + mlxsw_afa_qos_dscp_set(payload, dscp); +} + +static inline void +mlxsw_afa_qos_switch_prio_pack(char *payload, + enum mlxsw_afa_qos_switch_prio_cmd prio_cmd, + u8 prio) +{ + mlxsw_afa_qos_switch_prio_cmd_set(payload, prio_cmd); + mlxsw_afa_qos_switch_prio_set(payload, prio); +} + +static int __mlxsw_afa_block_append_qos_dsfield(struct mlxsw_afa_block *block, + bool set_dscp, u8 dscp, + bool set_ecn, u8 ecn, + struct netlink_ext_ack *extack) +{ + char *act = mlxsw_afa_block_append_action(block, + MLXSW_AFA_QOS_CODE, + MLXSW_AFA_QOS_SIZE); + + if (IS_ERR(act)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append QOS action"); + return PTR_ERR(act); + } + + if (set_ecn) + mlxsw_afa_qos_ecn_pack(act, MLXSW_AFA_QOS_ECN_CMD_SET, ecn); + if (set_dscp) { + mlxsw_afa_qos_dscp_pack(act, MLXSW_AFA_QOS_DSCP_CMD_SET_ALL, + dscp); + mlxsw_afa_qos_dscp_rw_set(act, MLXSW_AFA_QOS_DSCP_RW_CLEAR); + } + + return 0; +} + +int mlxsw_afa_block_append_qos_dsfield(struct mlxsw_afa_block *block, + u8 dsfield, + struct netlink_ext_ack *extack) +{ + return __mlxsw_afa_block_append_qos_dsfield(block, + true, dsfield >> 2, + true, dsfield & 0x03, + extack); +} +EXPORT_SYMBOL(mlxsw_afa_block_append_qos_dsfield); + +int mlxsw_afa_block_append_qos_dscp(struct mlxsw_afa_block *block, + u8 dscp, struct netlink_ext_ack *extack) +{ + return __mlxsw_afa_block_append_qos_dsfield(block, + true, dscp, + false, 0, + extack); +} +EXPORT_SYMBOL(mlxsw_afa_block_append_qos_dscp); + +int mlxsw_afa_block_append_qos_ecn(struct mlxsw_afa_block *block, + u8 ecn, struct netlink_ext_ack *extack) +{ + return __mlxsw_afa_block_append_qos_dsfield(block, + false, 0, + true, ecn, + extack); +} +EXPORT_SYMBOL(mlxsw_afa_block_append_qos_ecn); + +int mlxsw_afa_block_append_qos_switch_prio(struct mlxsw_afa_block *block, + u8 prio, + struct netlink_ext_ack *extack) +{ + char *act = mlxsw_afa_block_append_action(block, + MLXSW_AFA_QOS_CODE, + MLXSW_AFA_QOS_SIZE); + + if (IS_ERR(act)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append QOS action"); + return PTR_ERR(act); + } + mlxsw_afa_qos_switch_prio_pack(act, MLXSW_AFA_QOS_SWITCH_PRIO_CMD_SET, + prio); + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_qos_switch_prio); + +/* Forwarding Action + * ----------------- + * Forwarding Action can be used to implement Policy Based Switching (PBS) + * as well as OpenFlow related "Output" action. + */ + +#define MLXSW_AFA_FORWARD_CODE 0x07 +#define MLXSW_AFA_FORWARD_SIZE 1 + +enum mlxsw_afa_forward_type { + /* PBS, Policy Based Switching */ + MLXSW_AFA_FORWARD_TYPE_PBS, + /* Output, OpenFlow output type */ + MLXSW_AFA_FORWARD_TYPE_OUTPUT, +}; + +/* afa_forward_type */ +MLXSW_ITEM32(afa, forward, type, 0x00, 24, 2); + +/* afa_forward_pbs_ptr + * A pointer to the PBS entry configured by PPBS register. + * Reserved when in_port is set. + */ +MLXSW_ITEM32(afa, forward, pbs_ptr, 0x08, 0, 24); + +/* afa_forward_in_port + * Packet is forwarded back to the ingress port. + */ +MLXSW_ITEM32(afa, forward, in_port, 0x0C, 0, 1); + +static inline void +mlxsw_afa_forward_pack(char *payload, enum mlxsw_afa_forward_type type, + u32 pbs_ptr, bool in_port) +{ + mlxsw_afa_forward_type_set(payload, type); + mlxsw_afa_forward_pbs_ptr_set(payload, pbs_ptr); + mlxsw_afa_forward_in_port_set(payload, in_port); +} + +int mlxsw_afa_block_append_fwd(struct mlxsw_afa_block *block, + u16 local_port, bool in_port, + struct netlink_ext_ack *extack) +{ + struct mlxsw_afa_fwd_entry_ref *fwd_entry_ref; + u32 kvdl_index; + char *act; + int err; + + if (in_port) { + NL_SET_ERR_MSG_MOD(extack, "Forwarding to ingress port is not supported"); + return -EOPNOTSUPP; + } + fwd_entry_ref = mlxsw_afa_fwd_entry_ref_create(block, local_port); + if (IS_ERR(fwd_entry_ref)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot create forward action"); + return PTR_ERR(fwd_entry_ref); + } + kvdl_index = fwd_entry_ref->fwd_entry->kvdl_index; + + act = mlxsw_afa_block_append_action(block, MLXSW_AFA_FORWARD_CODE, + MLXSW_AFA_FORWARD_SIZE); + if (IS_ERR(act)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append forward action"); + err = PTR_ERR(act); + goto err_append_action; + } + mlxsw_afa_forward_pack(act, MLXSW_AFA_FORWARD_TYPE_PBS, + kvdl_index, in_port); + return 0; + +err_append_action: + mlxsw_afa_fwd_entry_ref_destroy(block, fwd_entry_ref); + return err; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_fwd); + +/* Policing and Counting Action + * ---------------------------- + * Policing and Counting action is used for binding policer and counter + * to ACL rules. + */ + +#define MLXSW_AFA_POLCNT_CODE 0x08 +#define MLXSW_AFA_POLCNT_SIZE 1 + +enum { + MLXSW_AFA_POLCNT_COUNTER, + MLXSW_AFA_POLCNT_POLICER, +}; + +/* afa_polcnt_c_p + * Counter or policer. + * Indicates whether the action binds a policer or a counter to the flow. + * 0: Counter + * 1: Policer + */ +MLXSW_ITEM32(afa, polcnt, c_p, 0x00, 31, 1); + +enum mlxsw_afa_polcnt_counter_set_type { + /* No count */ + MLXSW_AFA_POLCNT_COUNTER_SET_TYPE_NO_COUNT = 0x00, + /* Count packets and bytes */ + MLXSW_AFA_POLCNT_COUNTER_SET_TYPE_PACKETS_BYTES = 0x03, + /* Count only packets */ + MLXSW_AFA_POLCNT_COUNTER_SET_TYPE_PACKETS = 0x05, +}; + +/* afa_polcnt_counter_set_type + * Counter set type for flow counters. + */ +MLXSW_ITEM32(afa, polcnt, counter_set_type, 0x04, 24, 8); + +/* afa_polcnt_counter_index + * Counter index for flow counters. + */ +MLXSW_ITEM32(afa, polcnt, counter_index, 0x04, 0, 24); + +/* afa_polcnt_pid + * Policer ID. + * Reserved when c_p = 0 + */ +MLXSW_ITEM32(afa, polcnt, pid, 0x08, 0, 14); + +static inline void +mlxsw_afa_polcnt_pack(char *payload, + enum mlxsw_afa_polcnt_counter_set_type set_type, + u32 counter_index) +{ + mlxsw_afa_polcnt_c_p_set(payload, MLXSW_AFA_POLCNT_COUNTER); + mlxsw_afa_polcnt_counter_set_type_set(payload, set_type); + mlxsw_afa_polcnt_counter_index_set(payload, counter_index); +} + +static void mlxsw_afa_polcnt_policer_pack(char *payload, u16 policer_index) +{ + mlxsw_afa_polcnt_c_p_set(payload, MLXSW_AFA_POLCNT_POLICER); + mlxsw_afa_polcnt_pid_set(payload, policer_index); +} + +int mlxsw_afa_block_append_allocated_counter(struct mlxsw_afa_block *block, + u32 counter_index) +{ + char *act = mlxsw_afa_block_append_action(block, MLXSW_AFA_POLCNT_CODE, + MLXSW_AFA_POLCNT_SIZE); + if (IS_ERR(act)) + return PTR_ERR(act); + mlxsw_afa_polcnt_pack(act, MLXSW_AFA_POLCNT_COUNTER_SET_TYPE_PACKETS_BYTES, + counter_index); + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_allocated_counter); + +int mlxsw_afa_block_append_counter(struct mlxsw_afa_block *block, + u32 *p_counter_index, + struct netlink_ext_ack *extack) +{ + struct mlxsw_afa_counter *counter; + u32 counter_index; + int err; + + counter = mlxsw_afa_counter_create(block); + if (IS_ERR(counter)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot create count action"); + return PTR_ERR(counter); + } + counter_index = counter->counter_index; + + err = mlxsw_afa_block_append_allocated_counter(block, counter_index); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append count action"); + goto err_append_allocated_counter; + } + if (p_counter_index) + *p_counter_index = counter_index; + return 0; + +err_append_allocated_counter: + mlxsw_afa_counter_destroy(block, counter); + return err; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_counter); + +int mlxsw_afa_block_append_police(struct mlxsw_afa_block *block, + u32 fa_index, u64 rate_bytes_ps, u32 burst, + u16 *p_policer_index, + struct netlink_ext_ack *extack) +{ + struct mlxsw_afa_policer_ref *policer_ref; + char *act; + int err; + + policer_ref = mlxsw_afa_policer_ref_create(block, fa_index, + rate_bytes_ps, + burst, extack); + if (IS_ERR(policer_ref)) + return PTR_ERR(policer_ref); + *p_policer_index = policer_ref->policer->policer_index; + + act = mlxsw_afa_block_append_action_ext(block, MLXSW_AFA_POLCNT_CODE, + MLXSW_AFA_POLCNT_SIZE, + MLXSW_AFA_ACTION_TYPE_POLICE); + if (IS_ERR(act)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append police action"); + err = PTR_ERR(act); + goto err_append_action; + } + mlxsw_afa_polcnt_policer_pack(act, *p_policer_index); + + return 0; + +err_append_action: + mlxsw_afa_policer_ref_destroy(block, policer_ref); + return err; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_police); + +/* Virtual Router and Forwarding Domain Action + * ------------------------------------------- + * Virtual Switch action is used for manipulate the Virtual Router (VR), + * MPLS label space and the Forwarding Identifier (FID). + */ + +#define MLXSW_AFA_VIRFWD_CODE 0x0E +#define MLXSW_AFA_VIRFWD_SIZE 1 + +enum mlxsw_afa_virfwd_fid_cmd { + /* Do nothing */ + MLXSW_AFA_VIRFWD_FID_CMD_NOOP, + /* Set the Forwarding Identifier (FID) to fid */ + MLXSW_AFA_VIRFWD_FID_CMD_SET, +}; + +/* afa_virfwd_fid_cmd */ +MLXSW_ITEM32(afa, virfwd, fid_cmd, 0x08, 29, 3); + +/* afa_virfwd_fid + * The FID value. + */ +MLXSW_ITEM32(afa, virfwd, fid, 0x08, 0, 16); + +static inline void mlxsw_afa_virfwd_pack(char *payload, + enum mlxsw_afa_virfwd_fid_cmd fid_cmd, + u16 fid) +{ + mlxsw_afa_virfwd_fid_cmd_set(payload, fid_cmd); + mlxsw_afa_virfwd_fid_set(payload, fid); +} + +int mlxsw_afa_block_append_fid_set(struct mlxsw_afa_block *block, u16 fid, + struct netlink_ext_ack *extack) +{ + char *act = mlxsw_afa_block_append_action(block, + MLXSW_AFA_VIRFWD_CODE, + MLXSW_AFA_VIRFWD_SIZE); + if (IS_ERR(act)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append fid_set action"); + return PTR_ERR(act); + } + mlxsw_afa_virfwd_pack(act, MLXSW_AFA_VIRFWD_FID_CMD_SET, fid); + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_fid_set); + +/* MC Routing Action + * ----------------- + * The Multicast router action. Can be used by RMFT_V2 - Router Multicast + * Forwarding Table Version 2 Register. + */ + +#define MLXSW_AFA_MCROUTER_CODE 0x10 +#define MLXSW_AFA_MCROUTER_SIZE 2 + +enum mlxsw_afa_mcrouter_rpf_action { + MLXSW_AFA_MCROUTER_RPF_ACTION_NOP, + MLXSW_AFA_MCROUTER_RPF_ACTION_TRAP, + MLXSW_AFA_MCROUTER_RPF_ACTION_DISCARD_ERROR, +}; + +/* afa_mcrouter_rpf_action */ +MLXSW_ITEM32(afa, mcrouter, rpf_action, 0x00, 28, 3); + +/* afa_mcrouter_expected_irif */ +MLXSW_ITEM32(afa, mcrouter, expected_irif, 0x00, 0, 16); + +/* afa_mcrouter_min_mtu */ +MLXSW_ITEM32(afa, mcrouter, min_mtu, 0x08, 0, 16); + +enum mlxsw_afa_mrouter_vrmid { + MLXSW_AFA_MCROUTER_VRMID_INVALID, + MLXSW_AFA_MCROUTER_VRMID_VALID +}; + +/* afa_mcrouter_vrmid + * Valid RMID: rigr_rmid_index is used as RMID + */ +MLXSW_ITEM32(afa, mcrouter, vrmid, 0x0C, 31, 1); + +/* afa_mcrouter_rigr_rmid_index + * When the vrmid field is set to invalid, the field is used as pointer to + * Router Interface Group (RIGR) Table in the KVD linear. + * When the vrmid is set to valid, the field is used as RMID index, ranged + * from 0 to max_mid - 1. The index is to the Port Group Table. + */ +MLXSW_ITEM32(afa, mcrouter, rigr_rmid_index, 0x0C, 0, 24); + +static inline void +mlxsw_afa_mcrouter_pack(char *payload, + enum mlxsw_afa_mcrouter_rpf_action rpf_action, + u16 expected_irif, u16 min_mtu, + enum mlxsw_afa_mrouter_vrmid vrmid, u32 rigr_rmid_index) + +{ + mlxsw_afa_mcrouter_rpf_action_set(payload, rpf_action); + mlxsw_afa_mcrouter_expected_irif_set(payload, expected_irif); + mlxsw_afa_mcrouter_min_mtu_set(payload, min_mtu); + mlxsw_afa_mcrouter_vrmid_set(payload, vrmid); + mlxsw_afa_mcrouter_rigr_rmid_index_set(payload, rigr_rmid_index); +} + +int mlxsw_afa_block_append_mcrouter(struct mlxsw_afa_block *block, + u16 expected_irif, u16 min_mtu, + bool rmid_valid, u32 kvdl_index) +{ + char *act = mlxsw_afa_block_append_action(block, + MLXSW_AFA_MCROUTER_CODE, + MLXSW_AFA_MCROUTER_SIZE); + if (IS_ERR(act)) + return PTR_ERR(act); + mlxsw_afa_mcrouter_pack(act, MLXSW_AFA_MCROUTER_RPF_ACTION_TRAP, + expected_irif, min_mtu, rmid_valid, kvdl_index); + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_mcrouter); + +/* SIP DIP Action + * -------------- + * The SIP_DIP_ACTION is used for modifying the SIP and DIP fields of the + * packet, e.g. for NAT. The L3 checksum is updated. Also, if the L4 is TCP or + * if the L4 is UDP and the checksum field is not zero, then the L4 checksum is + * updated. + */ + +#define MLXSW_AFA_IP_CODE 0x11 +#define MLXSW_AFA_IP_SIZE 2 + +enum mlxsw_afa_ip_s_d { + /* ip refers to dip */ + MLXSW_AFA_IP_S_D_DIP, + /* ip refers to sip */ + MLXSW_AFA_IP_S_D_SIP, +}; + +/* afa_ip_s_d + * Source or destination. + */ +MLXSW_ITEM32(afa, ip, s_d, 0x00, 31, 1); + +enum mlxsw_afa_ip_m_l { + /* LSB: ip[63:0] refers to ip[63:0] */ + MLXSW_AFA_IP_M_L_LSB, + /* MSB: ip[63:0] refers to ip[127:64] */ + MLXSW_AFA_IP_M_L_MSB, +}; + +/* afa_ip_m_l + * MSB or LSB. + */ +MLXSW_ITEM32(afa, ip, m_l, 0x00, 30, 1); + +/* afa_ip_ip_63_32 + * Bits [63:32] in the IP address to change to. + */ +MLXSW_ITEM32(afa, ip, ip_63_32, 0x08, 0, 32); + +/* afa_ip_ip_31_0 + * Bits [31:0] in the IP address to change to. + */ +MLXSW_ITEM32(afa, ip, ip_31_0, 0x0C, 0, 32); + +static void mlxsw_afa_ip_pack(char *payload, enum mlxsw_afa_ip_s_d s_d, + enum mlxsw_afa_ip_m_l m_l, u32 ip_31_0, + u32 ip_63_32) +{ + mlxsw_afa_ip_s_d_set(payload, s_d); + mlxsw_afa_ip_m_l_set(payload, m_l); + mlxsw_afa_ip_ip_31_0_set(payload, ip_31_0); + mlxsw_afa_ip_ip_63_32_set(payload, ip_63_32); +} + +int mlxsw_afa_block_append_ip(struct mlxsw_afa_block *block, bool is_dip, + bool is_lsb, u32 val_31_0, u32 val_63_32, + struct netlink_ext_ack *extack) +{ + enum mlxsw_afa_ip_s_d s_d = is_dip ? MLXSW_AFA_IP_S_D_DIP : + MLXSW_AFA_IP_S_D_SIP; + enum mlxsw_afa_ip_m_l m_l = is_lsb ? MLXSW_AFA_IP_M_L_LSB : + MLXSW_AFA_IP_M_L_MSB; + char *act = mlxsw_afa_block_append_action(block, + MLXSW_AFA_IP_CODE, + MLXSW_AFA_IP_SIZE); + + if (IS_ERR(act)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append IP action"); + return PTR_ERR(act); + } + + mlxsw_afa_ip_pack(act, s_d, m_l, val_31_0, val_63_32); + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_ip); + +/* L4 Port Action + * -------------- + * The L4_PORT_ACTION is used for modifying the sport and dport fields of the packet, e.g. for NAT. + * If (the L4 is TCP) or if (the L4 is UDP and checksum field!=0) then the L4 checksum is updated. + */ + +#define MLXSW_AFA_L4PORT_CODE 0x12 +#define MLXSW_AFA_L4PORT_SIZE 1 + +enum mlxsw_afa_l4port_s_d { + /* configure src_l4_port */ + MLXSW_AFA_L4PORT_S_D_SRC, + /* configure dst_l4_port */ + MLXSW_AFA_L4PORT_S_D_DST, +}; + +/* afa_l4port_s_d + * Source or destination. + */ +MLXSW_ITEM32(afa, l4port, s_d, 0x00, 31, 1); + +/* afa_l4port_l4_port + * Number of port to change to. + */ +MLXSW_ITEM32(afa, l4port, l4_port, 0x08, 0, 16); + +static void mlxsw_afa_l4port_pack(char *payload, enum mlxsw_afa_l4port_s_d s_d, u16 l4_port) +{ + mlxsw_afa_l4port_s_d_set(payload, s_d); + mlxsw_afa_l4port_l4_port_set(payload, l4_port); +} + +int mlxsw_afa_block_append_l4port(struct mlxsw_afa_block *block, bool is_dport, u16 l4_port, + struct netlink_ext_ack *extack) +{ + enum mlxsw_afa_l4port_s_d s_d = is_dport ? MLXSW_AFA_L4PORT_S_D_DST : + MLXSW_AFA_L4PORT_S_D_SRC; + char *act = mlxsw_afa_block_append_action(block, + MLXSW_AFA_L4PORT_CODE, + MLXSW_AFA_L4PORT_SIZE); + + if (IS_ERR(act)) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append L4_PORT action"); + return PTR_ERR(act); + } + + mlxsw_afa_l4port_pack(act, s_d, l4_port); + return 0; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_l4port); + +/* Mirror Sampler Action + * --------------------- + * The SAMPLER_ACTION is used to mirror packets with a probability (sampling). + */ + +#define MLXSW_AFA_SAMPLER_CODE 0x13 +#define MLXSW_AFA_SAMPLER_SIZE 1 + +/* afa_sampler_mirror_agent + * Mirror (SPAN) agent. + */ +MLXSW_ITEM32(afa, sampler, mirror_agent, 0x04, 0, 3); + +#define MLXSW_AFA_SAMPLER_RATE_MAX (BIT(24) - 1) + +/* afa_sampler_mirror_probability_rate + * Mirroring probability. + * Valid values are 1 to 2^24 - 1 + */ +MLXSW_ITEM32(afa, sampler, mirror_probability_rate, 0x08, 0, 24); + +static void mlxsw_afa_sampler_pack(char *payload, u8 mirror_agent, u32 rate) +{ + mlxsw_afa_sampler_mirror_agent_set(payload, mirror_agent); + mlxsw_afa_sampler_mirror_probability_rate_set(payload, rate); +} + +struct mlxsw_afa_sampler { + struct mlxsw_afa_resource resource; + int span_id; + u16 local_port; + bool ingress; +}; + +static void mlxsw_afa_sampler_destroy(struct mlxsw_afa_block *block, + struct mlxsw_afa_sampler *sampler) +{ + mlxsw_afa_resource_del(&sampler->resource); + block->afa->ops->sampler_del(block->afa->ops_priv, sampler->local_port, + sampler->span_id, sampler->ingress); + kfree(sampler); +} + +static void mlxsw_afa_sampler_destructor(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource) +{ + struct mlxsw_afa_sampler *sampler; + + sampler = container_of(resource, struct mlxsw_afa_sampler, resource); + mlxsw_afa_sampler_destroy(block, sampler); +} + +static struct mlxsw_afa_sampler * +mlxsw_afa_sampler_create(struct mlxsw_afa_block *block, u16 local_port, + struct psample_group *psample_group, u32 rate, + u32 trunc_size, bool truncate, bool ingress, + struct netlink_ext_ack *extack) +{ + struct mlxsw_afa_sampler *sampler; + int err; + + sampler = kzalloc(sizeof(*sampler), GFP_KERNEL); + if (!sampler) + return ERR_PTR(-ENOMEM); + + err = block->afa->ops->sampler_add(block->afa->ops_priv, local_port, + psample_group, rate, trunc_size, + truncate, ingress, &sampler->span_id, + extack); + if (err) + goto err_sampler_add; + + sampler->ingress = ingress; + sampler->local_port = local_port; + sampler->resource.destructor = mlxsw_afa_sampler_destructor; + mlxsw_afa_resource_add(block, &sampler->resource); + return sampler; + +err_sampler_add: + kfree(sampler); + return ERR_PTR(err); +} + +static int +mlxsw_afa_block_append_allocated_sampler(struct mlxsw_afa_block *block, + u8 mirror_agent, u32 rate) +{ + char *act = mlxsw_afa_block_append_action(block, MLXSW_AFA_SAMPLER_CODE, + MLXSW_AFA_SAMPLER_SIZE); + + if (IS_ERR(act)) + return PTR_ERR(act); + mlxsw_afa_sampler_pack(act, mirror_agent, rate); + return 0; +} + +int mlxsw_afa_block_append_sampler(struct mlxsw_afa_block *block, u16 local_port, + struct psample_group *psample_group, + u32 rate, u32 trunc_size, bool truncate, + bool ingress, + struct netlink_ext_ack *extack) +{ + struct mlxsw_afa_sampler *sampler; + int err; + + if (rate > MLXSW_AFA_SAMPLER_RATE_MAX) { + NL_SET_ERR_MSG_MOD(extack, "Sampling rate is too high"); + return -EINVAL; + } + + sampler = mlxsw_afa_sampler_create(block, local_port, psample_group, + rate, trunc_size, truncate, ingress, + extack); + if (IS_ERR(sampler)) + return PTR_ERR(sampler); + + err = mlxsw_afa_block_append_allocated_sampler(block, sampler->span_id, + rate); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append sampler action"); + goto err_append_allocated_sampler; + } + + return 0; + +err_append_allocated_sampler: + mlxsw_afa_sampler_destroy(block, sampler); + return err; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_sampler); diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.h b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.h new file mode 100644 index 000000000..db58037be --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.h @@ -0,0 +1,110 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_CORE_ACL_FLEX_ACTIONS_H +#define _MLXSW_CORE_ACL_FLEX_ACTIONS_H + +#include <linux/types.h> +#include <linux/netdevice.h> +#include <net/flow_offload.h> + +struct mlxsw_afa; +struct mlxsw_afa_block; + +struct mlxsw_afa_ops { + int (*kvdl_set_add)(void *priv, u32 *p_kvdl_index, + char *enc_actions, bool is_first); + void (*kvdl_set_del)(void *priv, u32 kvdl_index, bool is_first); + int (*kvdl_set_activity_get)(void *priv, u32 kvdl_index, + bool *activity); + int (*kvdl_fwd_entry_add)(void *priv, u32 *p_kvdl_index, u16 local_port); + void (*kvdl_fwd_entry_del)(void *priv, u32 kvdl_index); + int (*counter_index_get)(void *priv, unsigned int *p_counter_index); + void (*counter_index_put)(void *priv, unsigned int counter_index); + int (*mirror_add)(void *priv, u16 local_in_port, + const struct net_device *out_dev, + bool ingress, int *p_span_id); + void (*mirror_del)(void *priv, u16 local_in_port, int span_id, + bool ingress); + int (*policer_add)(void *priv, u64 rate_bytes_ps, u32 burst, + u16 *p_policer_index, + struct netlink_ext_ack *extack); + void (*policer_del)(void *priv, u16 policer_index); + int (*sampler_add)(void *priv, u16 local_port, + struct psample_group *psample_group, u32 rate, + u32 trunc_size, bool truncate, bool ingress, + int *p_span_id, struct netlink_ext_ack *extack); + void (*sampler_del)(void *priv, u16 local_port, int span_id, + bool ingress); + bool dummy_first_set; +}; + +struct mlxsw_afa *mlxsw_afa_create(unsigned int max_acts_per_set, + const struct mlxsw_afa_ops *ops, + void *ops_priv); +void mlxsw_afa_destroy(struct mlxsw_afa *mlxsw_afa); +struct mlxsw_afa_block *mlxsw_afa_block_create(struct mlxsw_afa *mlxsw_afa); +void mlxsw_afa_block_destroy(struct mlxsw_afa_block *block); +int mlxsw_afa_block_commit(struct mlxsw_afa_block *block); +char *mlxsw_afa_block_first_set(struct mlxsw_afa_block *block); +char *mlxsw_afa_block_cur_set(struct mlxsw_afa_block *block); +u32 mlxsw_afa_block_first_kvdl_index(struct mlxsw_afa_block *block); +int mlxsw_afa_block_activity_get(struct mlxsw_afa_block *block, bool *activity); +int mlxsw_afa_block_continue(struct mlxsw_afa_block *block); +int mlxsw_afa_block_jump(struct mlxsw_afa_block *block, u16 group_id); +int mlxsw_afa_block_terminate(struct mlxsw_afa_block *block); +const struct flow_action_cookie * +mlxsw_afa_cookie_lookup(struct mlxsw_afa *mlxsw_afa, u32 cookie_index); +int mlxsw_afa_block_append_drop(struct mlxsw_afa_block *block, bool ingress, + const struct flow_action_cookie *fa_cookie, + struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_trap(struct mlxsw_afa_block *block, u16 trap_id); +int mlxsw_afa_block_append_trap_and_forward(struct mlxsw_afa_block *block, + u16 trap_id); +int mlxsw_afa_block_append_mirror(struct mlxsw_afa_block *block, + u16 local_in_port, + const struct net_device *out_dev, + bool ingress, + struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_fwd(struct mlxsw_afa_block *block, + u16 local_port, bool in_port, + struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_vlan_modify(struct mlxsw_afa_block *block, + u16 vid, u8 pcp, u8 et, + struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_qos_switch_prio(struct mlxsw_afa_block *block, + u8 prio, + struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_qos_dsfield(struct mlxsw_afa_block *block, + u8 dsfield, + struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_qos_dscp(struct mlxsw_afa_block *block, + u8 dscp, struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_qos_ecn(struct mlxsw_afa_block *block, + u8 ecn, struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_allocated_counter(struct mlxsw_afa_block *block, + u32 counter_index); +int mlxsw_afa_block_append_counter(struct mlxsw_afa_block *block, + u32 *p_counter_index, + struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_fid_set(struct mlxsw_afa_block *block, u16 fid, + struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_mcrouter(struct mlxsw_afa_block *block, + u16 expected_irif, u16 min_mtu, + bool rmid_valid, u32 kvdl_index); +int mlxsw_afa_block_append_ip(struct mlxsw_afa_block *block, bool is_dip, + bool is_lsb, u32 val_31_0, u32 val_63_32, + struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_l4port(struct mlxsw_afa_block *block, bool is_dport, u16 l4_port, + struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_police(struct mlxsw_afa_block *block, + u32 fa_index, u64 rate_bytes_ps, u32 burst, + u16 *p_policer_index, + struct netlink_ext_ack *extack); +int mlxsw_afa_block_append_sampler(struct mlxsw_afa_block *block, u16 local_port, + struct psample_group *psample_group, + u32 rate, u32 trunc_size, bool truncate, + bool ingress, + struct netlink_ext_ack *extack); + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_keys.c b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_keys.c new file mode 100644 index 000000000..f208a237d --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_keys.c @@ -0,0 +1,506 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/slab.h> +#include <linux/list.h> +#include <linux/errno.h> + +#include "item.h" +#include "core_acl_flex_keys.h" + +/* For the purpose of the driver, define an internal storage scratchpad + * that will be used to store key/mask values. For each defined element type + * define an internal storage geometry. + * + * When adding new elements, MLXSW_AFK_ELEMENT_STORAGE_SIZE must be increased + * accordingly. + */ +static const struct mlxsw_afk_element_info mlxsw_afk_element_infos[] = { + MLXSW_AFK_ELEMENT_INFO_U32(SRC_SYS_PORT, 0x00, 16, 16), + MLXSW_AFK_ELEMENT_INFO_BUF(DMAC_32_47, 0x04, 2), + MLXSW_AFK_ELEMENT_INFO_BUF(DMAC_0_31, 0x06, 4), + MLXSW_AFK_ELEMENT_INFO_BUF(SMAC_32_47, 0x0A, 2), + MLXSW_AFK_ELEMENT_INFO_BUF(SMAC_0_31, 0x0C, 4), + MLXSW_AFK_ELEMENT_INFO_U32(ETHERTYPE, 0x00, 0, 16), + MLXSW_AFK_ELEMENT_INFO_U32(IP_PROTO, 0x10, 0, 8), + MLXSW_AFK_ELEMENT_INFO_U32(VID, 0x10, 8, 12), + MLXSW_AFK_ELEMENT_INFO_U32(PCP, 0x10, 20, 3), + MLXSW_AFK_ELEMENT_INFO_U32(TCP_FLAGS, 0x10, 23, 9), + MLXSW_AFK_ELEMENT_INFO_U32(DST_L4_PORT, 0x14, 0, 16), + MLXSW_AFK_ELEMENT_INFO_U32(SRC_L4_PORT, 0x14, 16, 16), + MLXSW_AFK_ELEMENT_INFO_U32(IP_TTL_, 0x18, 0, 8), + MLXSW_AFK_ELEMENT_INFO_U32(IP_ECN, 0x18, 9, 2), + MLXSW_AFK_ELEMENT_INFO_U32(IP_DSCP, 0x18, 11, 6), + MLXSW_AFK_ELEMENT_INFO_U32(VIRT_ROUTER_MSB, 0x18, 17, 4), + MLXSW_AFK_ELEMENT_INFO_U32(VIRT_ROUTER_LSB, 0x18, 21, 8), + MLXSW_AFK_ELEMENT_INFO_BUF(SRC_IP_96_127, 0x20, 4), + MLXSW_AFK_ELEMENT_INFO_BUF(SRC_IP_64_95, 0x24, 4), + MLXSW_AFK_ELEMENT_INFO_BUF(SRC_IP_32_63, 0x28, 4), + MLXSW_AFK_ELEMENT_INFO_BUF(SRC_IP_0_31, 0x2C, 4), + MLXSW_AFK_ELEMENT_INFO_BUF(DST_IP_96_127, 0x30, 4), + MLXSW_AFK_ELEMENT_INFO_BUF(DST_IP_64_95, 0x34, 4), + MLXSW_AFK_ELEMENT_INFO_BUF(DST_IP_32_63, 0x38, 4), + MLXSW_AFK_ELEMENT_INFO_BUF(DST_IP_0_31, 0x3C, 4), +}; + +struct mlxsw_afk { + struct list_head key_info_list; + unsigned int max_blocks; + const struct mlxsw_afk_ops *ops; + const struct mlxsw_afk_block *blocks; + unsigned int blocks_count; +}; + +static bool mlxsw_afk_blocks_check(struct mlxsw_afk *mlxsw_afk) +{ + int i; + int j; + + for (i = 0; i < mlxsw_afk->blocks_count; i++) { + const struct mlxsw_afk_block *block = &mlxsw_afk->blocks[i]; + + for (j = 0; j < block->instances_count; j++) { + const struct mlxsw_afk_element_info *elinfo; + struct mlxsw_afk_element_inst *elinst; + + elinst = &block->instances[j]; + elinfo = &mlxsw_afk_element_infos[elinst->element]; + if (elinst->type != elinfo->type || + (!elinst->avoid_size_check && + elinst->item.size.bits != + elinfo->item.size.bits)) + return false; + } + } + return true; +} + +struct mlxsw_afk *mlxsw_afk_create(unsigned int max_blocks, + const struct mlxsw_afk_ops *ops) +{ + struct mlxsw_afk *mlxsw_afk; + + mlxsw_afk = kzalloc(sizeof(*mlxsw_afk), GFP_KERNEL); + if (!mlxsw_afk) + return NULL; + INIT_LIST_HEAD(&mlxsw_afk->key_info_list); + mlxsw_afk->max_blocks = max_blocks; + mlxsw_afk->ops = ops; + mlxsw_afk->blocks = ops->blocks; + mlxsw_afk->blocks_count = ops->blocks_count; + WARN_ON(!mlxsw_afk_blocks_check(mlxsw_afk)); + return mlxsw_afk; +} +EXPORT_SYMBOL(mlxsw_afk_create); + +void mlxsw_afk_destroy(struct mlxsw_afk *mlxsw_afk) +{ + WARN_ON(!list_empty(&mlxsw_afk->key_info_list)); + kfree(mlxsw_afk); +} +EXPORT_SYMBOL(mlxsw_afk_destroy); + +struct mlxsw_afk_key_info { + struct list_head list; + unsigned int ref_count; + unsigned int blocks_count; + int element_to_block[MLXSW_AFK_ELEMENT_MAX]; /* index is element, value + * is index inside "blocks" + */ + struct mlxsw_afk_element_usage elusage; + const struct mlxsw_afk_block *blocks[]; +}; + +static bool +mlxsw_afk_key_info_elements_eq(struct mlxsw_afk_key_info *key_info, + struct mlxsw_afk_element_usage *elusage) +{ + return memcmp(&key_info->elusage, elusage, sizeof(*elusage)) == 0; +} + +static struct mlxsw_afk_key_info * +mlxsw_afk_key_info_find(struct mlxsw_afk *mlxsw_afk, + struct mlxsw_afk_element_usage *elusage) +{ + struct mlxsw_afk_key_info *key_info; + + list_for_each_entry(key_info, &mlxsw_afk->key_info_list, list) { + if (mlxsw_afk_key_info_elements_eq(key_info, elusage)) + return key_info; + } + return NULL; +} + +struct mlxsw_afk_picker { + DECLARE_BITMAP(element, MLXSW_AFK_ELEMENT_MAX); + unsigned int total; +}; + +static void mlxsw_afk_picker_count_hits(struct mlxsw_afk *mlxsw_afk, + struct mlxsw_afk_picker *picker, + enum mlxsw_afk_element element) +{ + int i; + int j; + + for (i = 0; i < mlxsw_afk->blocks_count; i++) { + const struct mlxsw_afk_block *block = &mlxsw_afk->blocks[i]; + + for (j = 0; j < block->instances_count; j++) { + struct mlxsw_afk_element_inst *elinst; + + elinst = &block->instances[j]; + if (elinst->element == element) { + __set_bit(element, picker[i].element); + picker[i].total++; + } + } + } +} + +static void mlxsw_afk_picker_subtract_hits(struct mlxsw_afk *mlxsw_afk, + struct mlxsw_afk_picker *picker, + int block_index) +{ + DECLARE_BITMAP(hits_element, MLXSW_AFK_ELEMENT_MAX); + int i; + int j; + + memcpy(&hits_element, &picker[block_index].element, + sizeof(hits_element)); + + for (i = 0; i < mlxsw_afk->blocks_count; i++) { + for_each_set_bit(j, hits_element, MLXSW_AFK_ELEMENT_MAX) { + if (__test_and_clear_bit(j, picker[i].element)) + picker[i].total--; + } + } +} + +static int mlxsw_afk_picker_most_hits_get(struct mlxsw_afk *mlxsw_afk, + struct mlxsw_afk_picker *picker) +{ + int most_index = -EINVAL; /* Should never happen to return this */ + int most_hits = 0; + int i; + + for (i = 0; i < mlxsw_afk->blocks_count; i++) { + if (picker[i].total > most_hits) { + most_hits = picker[i].total; + most_index = i; + } + } + return most_index; +} + +static int mlxsw_afk_picker_key_info_add(struct mlxsw_afk *mlxsw_afk, + struct mlxsw_afk_picker *picker, + int block_index, + struct mlxsw_afk_key_info *key_info) +{ + enum mlxsw_afk_element element; + + if (key_info->blocks_count == mlxsw_afk->max_blocks) + return -EINVAL; + + for_each_set_bit(element, picker[block_index].element, + MLXSW_AFK_ELEMENT_MAX) { + key_info->element_to_block[element] = key_info->blocks_count; + mlxsw_afk_element_usage_add(&key_info->elusage, element); + } + + key_info->blocks[key_info->blocks_count] = + &mlxsw_afk->blocks[block_index]; + key_info->blocks_count++; + return 0; +} + +static int mlxsw_afk_picker(struct mlxsw_afk *mlxsw_afk, + struct mlxsw_afk_key_info *key_info, + struct mlxsw_afk_element_usage *elusage) +{ + struct mlxsw_afk_picker *picker; + enum mlxsw_afk_element element; + int err; + + picker = kcalloc(mlxsw_afk->blocks_count, sizeof(*picker), GFP_KERNEL); + if (!picker) + return -ENOMEM; + + /* Since the same elements could be present in multiple blocks, + * we must find out optimal block list in order to make the + * block count as low as possible. + * + * First, we count hits. We go over all available blocks and count + * how many of requested elements are covered by each. + * + * Then in loop, we find block with most hits and add it to + * output key_info. Then we have to subtract this block hits so + * the next iteration will find most suitable block for + * the rest of requested elements. + */ + + mlxsw_afk_element_usage_for_each(element, elusage) + mlxsw_afk_picker_count_hits(mlxsw_afk, picker, element); + + do { + int block_index; + + block_index = mlxsw_afk_picker_most_hits_get(mlxsw_afk, picker); + if (block_index < 0) { + err = block_index; + goto out; + } + err = mlxsw_afk_picker_key_info_add(mlxsw_afk, picker, + block_index, key_info); + if (err) + goto out; + mlxsw_afk_picker_subtract_hits(mlxsw_afk, picker, block_index); + } while (!mlxsw_afk_key_info_elements_eq(key_info, elusage)); + + err = 0; +out: + kfree(picker); + return err; +} + +static struct mlxsw_afk_key_info * +mlxsw_afk_key_info_create(struct mlxsw_afk *mlxsw_afk, + struct mlxsw_afk_element_usage *elusage) +{ + struct mlxsw_afk_key_info *key_info; + int err; + + key_info = kzalloc(struct_size(key_info, blocks, mlxsw_afk->max_blocks), + GFP_KERNEL); + if (!key_info) + return ERR_PTR(-ENOMEM); + err = mlxsw_afk_picker(mlxsw_afk, key_info, elusage); + if (err) + goto err_picker; + list_add(&key_info->list, &mlxsw_afk->key_info_list); + key_info->ref_count = 1; + return key_info; + +err_picker: + kfree(key_info); + return ERR_PTR(err); +} + +static void mlxsw_afk_key_info_destroy(struct mlxsw_afk_key_info *key_info) +{ + list_del(&key_info->list); + kfree(key_info); +} + +struct mlxsw_afk_key_info * +mlxsw_afk_key_info_get(struct mlxsw_afk *mlxsw_afk, + struct mlxsw_afk_element_usage *elusage) +{ + struct mlxsw_afk_key_info *key_info; + + key_info = mlxsw_afk_key_info_find(mlxsw_afk, elusage); + if (key_info) { + key_info->ref_count++; + return key_info; + } + return mlxsw_afk_key_info_create(mlxsw_afk, elusage); +} +EXPORT_SYMBOL(mlxsw_afk_key_info_get); + +void mlxsw_afk_key_info_put(struct mlxsw_afk_key_info *key_info) +{ + if (--key_info->ref_count) + return; + mlxsw_afk_key_info_destroy(key_info); +} +EXPORT_SYMBOL(mlxsw_afk_key_info_put); + +bool mlxsw_afk_key_info_subset(struct mlxsw_afk_key_info *key_info, + struct mlxsw_afk_element_usage *elusage) +{ + return mlxsw_afk_element_usage_subset(elusage, &key_info->elusage); +} +EXPORT_SYMBOL(mlxsw_afk_key_info_subset); + +static const struct mlxsw_afk_element_inst * +mlxsw_afk_block_elinst_get(const struct mlxsw_afk_block *block, + enum mlxsw_afk_element element) +{ + int i; + + for (i = 0; i < block->instances_count; i++) { + struct mlxsw_afk_element_inst *elinst; + + elinst = &block->instances[i]; + if (elinst->element == element) + return elinst; + } + return NULL; +} + +static const struct mlxsw_afk_element_inst * +mlxsw_afk_key_info_elinst_get(struct mlxsw_afk_key_info *key_info, + enum mlxsw_afk_element element, + int *p_block_index) +{ + const struct mlxsw_afk_element_inst *elinst; + const struct mlxsw_afk_block *block; + int block_index; + + if (WARN_ON(!test_bit(element, key_info->elusage.usage))) + return NULL; + block_index = key_info->element_to_block[element]; + block = key_info->blocks[block_index]; + + elinst = mlxsw_afk_block_elinst_get(block, element); + if (WARN_ON(!elinst)) + return NULL; + + *p_block_index = block_index; + return elinst; +} + +u16 +mlxsw_afk_key_info_block_encoding_get(const struct mlxsw_afk_key_info *key_info, + int block_index) +{ + return key_info->blocks[block_index]->encoding; +} +EXPORT_SYMBOL(mlxsw_afk_key_info_block_encoding_get); + +unsigned int +mlxsw_afk_key_info_blocks_count_get(const struct mlxsw_afk_key_info *key_info) +{ + return key_info->blocks_count; +} +EXPORT_SYMBOL(mlxsw_afk_key_info_blocks_count_get); + +void mlxsw_afk_values_add_u32(struct mlxsw_afk_element_values *values, + enum mlxsw_afk_element element, + u32 key_value, u32 mask_value) +{ + const struct mlxsw_afk_element_info *elinfo = + &mlxsw_afk_element_infos[element]; + const struct mlxsw_item *storage_item = &elinfo->item; + + if (!mask_value) + return; + if (WARN_ON(elinfo->type != MLXSW_AFK_ELEMENT_TYPE_U32)) + return; + __mlxsw_item_set32(values->storage.key, storage_item, 0, key_value); + __mlxsw_item_set32(values->storage.mask, storage_item, 0, mask_value); + mlxsw_afk_element_usage_add(&values->elusage, element); +} +EXPORT_SYMBOL(mlxsw_afk_values_add_u32); + +void mlxsw_afk_values_add_buf(struct mlxsw_afk_element_values *values, + enum mlxsw_afk_element element, + const char *key_value, const char *mask_value, + unsigned int len) +{ + const struct mlxsw_afk_element_info *elinfo = + &mlxsw_afk_element_infos[element]; + const struct mlxsw_item *storage_item = &elinfo->item; + + if (!memchr_inv(mask_value, 0, len)) /* If mask is zero */ + return; + if (WARN_ON(elinfo->type != MLXSW_AFK_ELEMENT_TYPE_BUF) || + WARN_ON(elinfo->item.size.bytes != len)) + return; + __mlxsw_item_memcpy_to(values->storage.key, key_value, + storage_item, 0); + __mlxsw_item_memcpy_to(values->storage.mask, mask_value, + storage_item, 0); + mlxsw_afk_element_usage_add(&values->elusage, element); +} +EXPORT_SYMBOL(mlxsw_afk_values_add_buf); + +static void mlxsw_sp_afk_encode_u32(const struct mlxsw_item *storage_item, + const struct mlxsw_item *output_item, + char *storage, char *output, int diff) +{ + u32 value; + + value = __mlxsw_item_get32(storage, storage_item, 0); + __mlxsw_item_set32(output, output_item, 0, value + diff); +} + +static void mlxsw_sp_afk_encode_buf(const struct mlxsw_item *storage_item, + const struct mlxsw_item *output_item, + char *storage, char *output) +{ + char *storage_data = __mlxsw_item_data(storage, storage_item, 0); + char *output_data = __mlxsw_item_data(output, output_item, 0); + size_t len = output_item->size.bytes; + + memcpy(output_data, storage_data, len); +} + +static void +mlxsw_sp_afk_encode_one(const struct mlxsw_afk_element_inst *elinst, + char *output, char *storage, int u32_diff) +{ + const struct mlxsw_item *output_item = &elinst->item; + const struct mlxsw_afk_element_info *elinfo; + const struct mlxsw_item *storage_item; + + elinfo = &mlxsw_afk_element_infos[elinst->element]; + storage_item = &elinfo->item; + if (elinst->type == MLXSW_AFK_ELEMENT_TYPE_U32) + mlxsw_sp_afk_encode_u32(storage_item, output_item, + storage, output, u32_diff); + else if (elinst->type == MLXSW_AFK_ELEMENT_TYPE_BUF) + mlxsw_sp_afk_encode_buf(storage_item, output_item, + storage, output); +} + +#define MLXSW_SP_AFK_KEY_BLOCK_MAX_SIZE 16 + +void mlxsw_afk_encode(struct mlxsw_afk *mlxsw_afk, + struct mlxsw_afk_key_info *key_info, + struct mlxsw_afk_element_values *values, + char *key, char *mask) +{ + unsigned int blocks_count = + mlxsw_afk_key_info_blocks_count_get(key_info); + char block_mask[MLXSW_SP_AFK_KEY_BLOCK_MAX_SIZE]; + char block_key[MLXSW_SP_AFK_KEY_BLOCK_MAX_SIZE]; + const struct mlxsw_afk_element_inst *elinst; + enum mlxsw_afk_element element; + int block_index, i; + + for (i = 0; i < blocks_count; i++) { + memset(block_key, 0, MLXSW_SP_AFK_KEY_BLOCK_MAX_SIZE); + memset(block_mask, 0, MLXSW_SP_AFK_KEY_BLOCK_MAX_SIZE); + + mlxsw_afk_element_usage_for_each(element, &values->elusage) { + elinst = mlxsw_afk_key_info_elinst_get(key_info, + element, + &block_index); + if (!elinst || block_index != i) + continue; + + mlxsw_sp_afk_encode_one(elinst, block_key, + values->storage.key, + elinst->u32_key_diff); + mlxsw_sp_afk_encode_one(elinst, block_mask, + values->storage.mask, 0); + } + + mlxsw_afk->ops->encode_block(key, i, block_key); + mlxsw_afk->ops->encode_block(mask, i, block_mask); + } +} +EXPORT_SYMBOL(mlxsw_afk_encode); + +void mlxsw_afk_clear(struct mlxsw_afk *mlxsw_afk, char *key, + int block_start, int block_end) +{ + int i; + + for (i = block_start; i <= block_end; i++) + mlxsw_afk->ops->clear_block(key, i); +} +EXPORT_SYMBOL(mlxsw_afk_clear); diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_keys.h b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_keys.h new file mode 100644 index 000000000..3a037fe47 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_keys.h @@ -0,0 +1,222 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_CORE_ACL_FLEX_KEYS_H +#define _MLXSW_CORE_ACL_FLEX_KEYS_H + +#include <linux/types.h> +#include <linux/bitmap.h> + +#include "item.h" + +enum mlxsw_afk_element { + MLXSW_AFK_ELEMENT_SRC_SYS_PORT, + MLXSW_AFK_ELEMENT_DMAC_32_47, + MLXSW_AFK_ELEMENT_DMAC_0_31, + MLXSW_AFK_ELEMENT_SMAC_32_47, + MLXSW_AFK_ELEMENT_SMAC_0_31, + MLXSW_AFK_ELEMENT_ETHERTYPE, + MLXSW_AFK_ELEMENT_IP_PROTO, + MLXSW_AFK_ELEMENT_SRC_IP_96_127, + MLXSW_AFK_ELEMENT_SRC_IP_64_95, + MLXSW_AFK_ELEMENT_SRC_IP_32_63, + MLXSW_AFK_ELEMENT_SRC_IP_0_31, + MLXSW_AFK_ELEMENT_DST_IP_96_127, + MLXSW_AFK_ELEMENT_DST_IP_64_95, + MLXSW_AFK_ELEMENT_DST_IP_32_63, + MLXSW_AFK_ELEMENT_DST_IP_0_31, + MLXSW_AFK_ELEMENT_DST_L4_PORT, + MLXSW_AFK_ELEMENT_SRC_L4_PORT, + MLXSW_AFK_ELEMENT_VID, + MLXSW_AFK_ELEMENT_PCP, + MLXSW_AFK_ELEMENT_TCP_FLAGS, + MLXSW_AFK_ELEMENT_IP_TTL_, + MLXSW_AFK_ELEMENT_IP_ECN, + MLXSW_AFK_ELEMENT_IP_DSCP, + MLXSW_AFK_ELEMENT_VIRT_ROUTER_MSB, + MLXSW_AFK_ELEMENT_VIRT_ROUTER_LSB, + MLXSW_AFK_ELEMENT_MAX, +}; + +enum mlxsw_afk_element_type { + MLXSW_AFK_ELEMENT_TYPE_U32, + MLXSW_AFK_ELEMENT_TYPE_BUF, +}; + +struct mlxsw_afk_element_info { + enum mlxsw_afk_element element; /* element ID */ + enum mlxsw_afk_element_type type; + struct mlxsw_item item; /* element geometry in internal storage */ +}; + +#define MLXSW_AFK_ELEMENT_INFO(_type, _element, _offset, _shift, _size) \ + [MLXSW_AFK_ELEMENT_##_element] = { \ + .element = MLXSW_AFK_ELEMENT_##_element, \ + .type = _type, \ + .item = { \ + .offset = _offset, \ + .shift = _shift, \ + .size = {.bits = _size}, \ + .name = #_element, \ + }, \ + } + +#define MLXSW_AFK_ELEMENT_INFO_U32(_element, _offset, _shift, _size) \ + MLXSW_AFK_ELEMENT_INFO(MLXSW_AFK_ELEMENT_TYPE_U32, \ + _element, _offset, _shift, _size) + +#define MLXSW_AFK_ELEMENT_INFO_BUF(_element, _offset, _size) \ + MLXSW_AFK_ELEMENT_INFO(MLXSW_AFK_ELEMENT_TYPE_BUF, \ + _element, _offset, 0, _size) + +#define MLXSW_AFK_ELEMENT_STORAGE_SIZE 0x40 + +struct mlxsw_afk_element_inst { /* element instance in actual block */ + enum mlxsw_afk_element element; + enum mlxsw_afk_element_type type; + struct mlxsw_item item; /* element geometry in block */ + int u32_key_diff; /* in case value needs to be adjusted before write + * this diff is here to handle that + */ + bool avoid_size_check; +}; + +#define MLXSW_AFK_ELEMENT_INST(_type, _element, _offset, \ + _shift, _size, _u32_key_diff, _avoid_size_check) \ + { \ + .element = MLXSW_AFK_ELEMENT_##_element, \ + .type = _type, \ + .item = { \ + .offset = _offset, \ + .shift = _shift, \ + .size = {.bits = _size}, \ + .name = #_element, \ + }, \ + .u32_key_diff = _u32_key_diff, \ + .avoid_size_check = _avoid_size_check, \ + } + +#define MLXSW_AFK_ELEMENT_INST_U32(_element, _offset, _shift, _size) \ + MLXSW_AFK_ELEMENT_INST(MLXSW_AFK_ELEMENT_TYPE_U32, \ + _element, _offset, _shift, _size, 0, false) + +#define MLXSW_AFK_ELEMENT_INST_EXT_U32(_element, _offset, \ + _shift, _size, _key_diff, \ + _avoid_size_check) \ + MLXSW_AFK_ELEMENT_INST(MLXSW_AFK_ELEMENT_TYPE_U32, \ + _element, _offset, _shift, _size, \ + _key_diff, _avoid_size_check) + +#define MLXSW_AFK_ELEMENT_INST_BUF(_element, _offset, _size) \ + MLXSW_AFK_ELEMENT_INST(MLXSW_AFK_ELEMENT_TYPE_BUF, \ + _element, _offset, 0, _size, 0, false) + +struct mlxsw_afk_block { + u16 encoding; /* block ID */ + struct mlxsw_afk_element_inst *instances; + unsigned int instances_count; +}; + +#define MLXSW_AFK_BLOCK(_encoding, _instances) \ + { \ + .encoding = _encoding, \ + .instances = _instances, \ + .instances_count = ARRAY_SIZE(_instances), \ + } + +struct mlxsw_afk_element_usage { + DECLARE_BITMAP(usage, MLXSW_AFK_ELEMENT_MAX); +}; + +#define mlxsw_afk_element_usage_for_each(element, elusage) \ + for_each_set_bit(element, (elusage)->usage, MLXSW_AFK_ELEMENT_MAX) + +static inline void +mlxsw_afk_element_usage_add(struct mlxsw_afk_element_usage *elusage, + enum mlxsw_afk_element element) +{ + __set_bit(element, elusage->usage); +} + +static inline void +mlxsw_afk_element_usage_zero(struct mlxsw_afk_element_usage *elusage) +{ + bitmap_zero(elusage->usage, MLXSW_AFK_ELEMENT_MAX); +} + +static inline void +mlxsw_afk_element_usage_fill(struct mlxsw_afk_element_usage *elusage, + const enum mlxsw_afk_element *elements, + unsigned int elements_count) +{ + int i; + + mlxsw_afk_element_usage_zero(elusage); + for (i = 0; i < elements_count; i++) + mlxsw_afk_element_usage_add(elusage, elements[i]); +} + +static inline bool +mlxsw_afk_element_usage_subset(struct mlxsw_afk_element_usage *elusage_small, + struct mlxsw_afk_element_usage *elusage_big) +{ + int i; + + for (i = 0; i < MLXSW_AFK_ELEMENT_MAX; i++) + if (test_bit(i, elusage_small->usage) && + !test_bit(i, elusage_big->usage)) + return false; + return true; +} + +struct mlxsw_afk; + +struct mlxsw_afk_ops { + const struct mlxsw_afk_block *blocks; + unsigned int blocks_count; + void (*encode_block)(char *output, int block_index, char *block); + void (*clear_block)(char *output, int block_index); +}; + +struct mlxsw_afk *mlxsw_afk_create(unsigned int max_blocks, + const struct mlxsw_afk_ops *ops); +void mlxsw_afk_destroy(struct mlxsw_afk *mlxsw_afk); + +struct mlxsw_afk_key_info; + +struct mlxsw_afk_key_info * +mlxsw_afk_key_info_get(struct mlxsw_afk *mlxsw_afk, + struct mlxsw_afk_element_usage *elusage); +void mlxsw_afk_key_info_put(struct mlxsw_afk_key_info *key_info); +bool mlxsw_afk_key_info_subset(struct mlxsw_afk_key_info *key_info, + struct mlxsw_afk_element_usage *elusage); + +u16 +mlxsw_afk_key_info_block_encoding_get(const struct mlxsw_afk_key_info *key_info, + int block_index); +unsigned int +mlxsw_afk_key_info_blocks_count_get(const struct mlxsw_afk_key_info *key_info); + +struct mlxsw_afk_element_values { + struct mlxsw_afk_element_usage elusage; + struct { + char key[MLXSW_AFK_ELEMENT_STORAGE_SIZE]; + char mask[MLXSW_AFK_ELEMENT_STORAGE_SIZE]; + } storage; +}; + +void mlxsw_afk_values_add_u32(struct mlxsw_afk_element_values *values, + enum mlxsw_afk_element element, + u32 key_value, u32 mask_value); +void mlxsw_afk_values_add_buf(struct mlxsw_afk_element_values *values, + enum mlxsw_afk_element element, + const char *key_value, const char *mask_value, + unsigned int len); +void mlxsw_afk_encode(struct mlxsw_afk *mlxsw_afk, + struct mlxsw_afk_key_info *key_info, + struct mlxsw_afk_element_values *values, + char *key, char *mask); +void mlxsw_afk_clear(struct mlxsw_afk *mlxsw_afk, char *key, + int block_start, int block_end); + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_env.c b/drivers/net/ethernet/mellanox/mlxsw/core_env.c new file mode 100644 index 000000000..0107cbc32 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/core_env.c @@ -0,0 +1,1464 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/err.h> +#include <linux/ethtool.h> +#include <linux/sfp.h> +#include <linux/mutex.h> + +#include "core.h" +#include "core_env.h" +#include "item.h" +#include "reg.h" + +struct mlxsw_env_module_info { + u64 module_overheat_counter; + bool is_overheat; + int num_ports_mapped; + int num_ports_up; + enum ethtool_module_power_mode_policy power_mode_policy; + enum mlxsw_reg_pmtm_module_type type; +}; + +struct mlxsw_env_line_card { + u8 module_count; + bool active; + struct mlxsw_env_module_info module_info[]; +}; + +struct mlxsw_env { + struct mlxsw_core *core; + const struct mlxsw_bus_info *bus_info; + u8 max_module_count; /* Maximum number of modules per-slot. */ + u8 num_of_slots; /* Including the main board. */ + struct mutex line_cards_lock; /* Protects line cards. */ + struct mlxsw_env_line_card *line_cards[]; +}; + +static bool __mlxsw_env_linecard_is_active(struct mlxsw_env *mlxsw_env, + u8 slot_index) +{ + return mlxsw_env->line_cards[slot_index]->active; +} + +static bool mlxsw_env_linecard_is_active(struct mlxsw_env *mlxsw_env, + u8 slot_index) +{ + bool active; + + mutex_lock(&mlxsw_env->line_cards_lock); + active = __mlxsw_env_linecard_is_active(mlxsw_env, slot_index); + mutex_unlock(&mlxsw_env->line_cards_lock); + + return active; +} + +static struct +mlxsw_env_module_info *mlxsw_env_module_info_get(struct mlxsw_core *mlxsw_core, + u8 slot_index, u8 module) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + + return &mlxsw_env->line_cards[slot_index]->module_info[module]; +} + +static int __mlxsw_env_validate_module_type(struct mlxsw_core *core, + u8 slot_index, u8 module) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(core); + struct mlxsw_env_module_info *module_info; + int err; + + if (!__mlxsw_env_linecard_is_active(mlxsw_env, slot_index)) + return 0; + + module_info = mlxsw_env_module_info_get(core, slot_index, module); + switch (module_info->type) { + case MLXSW_REG_PMTM_MODULE_TYPE_TWISTED_PAIR: + err = -EINVAL; + break; + default: + err = 0; + } + + return err; +} + +static int mlxsw_env_validate_module_type(struct mlxsw_core *core, + u8 slot_index, u8 module) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(core); + int err; + + mutex_lock(&mlxsw_env->line_cards_lock); + err = __mlxsw_env_validate_module_type(core, slot_index, module); + mutex_unlock(&mlxsw_env->line_cards_lock); + + return err; +} + +static int +mlxsw_env_validate_cable_ident(struct mlxsw_core *core, u8 slot_index, int id, + bool *qsfp, bool *cmis) +{ + char mcia_pl[MLXSW_REG_MCIA_LEN]; + char *eeprom_tmp; + u8 ident; + int err; + + err = mlxsw_env_validate_module_type(core, slot_index, id); + if (err) + return err; + + mlxsw_reg_mcia_pack(mcia_pl, slot_index, id, 0, + MLXSW_REG_MCIA_PAGE0_LO_OFF, 0, 1, + MLXSW_REG_MCIA_I2C_ADDR_LOW); + err = mlxsw_reg_query(core, MLXSW_REG(mcia), mcia_pl); + if (err) + return err; + eeprom_tmp = mlxsw_reg_mcia_eeprom_data(mcia_pl); + ident = eeprom_tmp[0]; + *cmis = false; + switch (ident) { + case MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_SFP: + *qsfp = false; + break; + case MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP: + case MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP_PLUS: + case MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP28: + *qsfp = true; + break; + case MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP_DD: + case MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_OSFP: + *qsfp = true; + *cmis = true; + break; + default: + return -EINVAL; + } + + return 0; +} + +static int +mlxsw_env_query_module_eeprom(struct mlxsw_core *mlxsw_core, u8 slot_index, + int module, u16 offset, u16 size, void *data, + bool qsfp, unsigned int *p_read_size) +{ + char mcia_pl[MLXSW_REG_MCIA_LEN]; + char *eeprom_tmp; + u16 i2c_addr; + u8 page = 0; + int status; + int err; + + /* MCIA register accepts buffer size <= 48. Page of size 128 should be + * read by chunks of size 48, 48, 32. Align the size of the last chunk + * to avoid reading after the end of the page. + */ + size = min_t(u16, size, MLXSW_REG_MCIA_EEPROM_SIZE); + + if (offset < MLXSW_REG_MCIA_EEPROM_PAGE_LENGTH && + offset + size > MLXSW_REG_MCIA_EEPROM_PAGE_LENGTH) + /* Cross pages read, read until offset 256 in low page */ + size = MLXSW_REG_MCIA_EEPROM_PAGE_LENGTH - offset; + + i2c_addr = MLXSW_REG_MCIA_I2C_ADDR_LOW; + if (offset >= MLXSW_REG_MCIA_EEPROM_PAGE_LENGTH) { + if (qsfp) { + /* When reading upper pages 1, 2 and 3 the offset + * starts at 128. Please refer to "QSFP+ Memory Map" + * figure in SFF-8436 specification and to "CMIS Module + * Memory Map" figure in CMIS specification for + * graphical depiction. + */ + page = MLXSW_REG_MCIA_PAGE_GET(offset); + offset -= MLXSW_REG_MCIA_EEPROM_UP_PAGE_LENGTH * page; + if (offset + size > MLXSW_REG_MCIA_EEPROM_PAGE_LENGTH) + size = MLXSW_REG_MCIA_EEPROM_PAGE_LENGTH - offset; + } else { + /* When reading upper pages 1, 2 and 3 the offset + * starts at 0 and I2C high address is used. Please refer + * to "Memory Organization" figure in SFF-8472 + * specification for graphical depiction. + */ + i2c_addr = MLXSW_REG_MCIA_I2C_ADDR_HIGH; + offset -= MLXSW_REG_MCIA_EEPROM_PAGE_LENGTH; + } + } + + mlxsw_reg_mcia_pack(mcia_pl, slot_index, module, 0, page, offset, size, + i2c_addr); + + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mcia), mcia_pl); + if (err) + return err; + + status = mlxsw_reg_mcia_status_get(mcia_pl); + if (status) + return -EIO; + + eeprom_tmp = mlxsw_reg_mcia_eeprom_data(mcia_pl); + memcpy(data, eeprom_tmp, size); + *p_read_size = size; + + return 0; +} + +int +mlxsw_env_module_temp_thresholds_get(struct mlxsw_core *core, u8 slot_index, + int module, int off, int *temp) +{ + unsigned int module_temp, module_crit, module_emerg; + union { + u8 buf[MLXSW_REG_MCIA_TH_ITEM_SIZE]; + u16 temp; + } temp_thresh; + char mcia_pl[MLXSW_REG_MCIA_LEN] = {0}; + char mtmp_pl[MLXSW_REG_MTMP_LEN]; + char *eeprom_tmp; + bool qsfp, cmis; + int page; + int err; + + mlxsw_reg_mtmp_pack(mtmp_pl, slot_index, + MLXSW_REG_MTMP_MODULE_INDEX_MIN + module, false, + false); + err = mlxsw_reg_query(core, MLXSW_REG(mtmp), mtmp_pl); + if (err) + return err; + mlxsw_reg_mtmp_unpack(mtmp_pl, &module_temp, NULL, &module_crit, + &module_emerg, NULL); + if (!module_temp) { + *temp = 0; + return 0; + } + + /* Validate if threshold reading is available through MTMP register, + * otherwise fallback to read through MCIA. + */ + if (module_emerg) { + *temp = off == SFP_TEMP_HIGH_WARN ? module_crit : module_emerg; + return 0; + } + + /* Read Free Side Device Temperature Thresholds from page 03h + * (MSB at lower byte address). + * Bytes: + * 128-129 - Temp High Alarm (SFP_TEMP_HIGH_ALARM); + * 130-131 - Temp Low Alarm (SFP_TEMP_LOW_ALARM); + * 132-133 - Temp High Warning (SFP_TEMP_HIGH_WARN); + * 134-135 - Temp Low Warning (SFP_TEMP_LOW_WARN); + */ + + /* Validate module identifier value. */ + err = mlxsw_env_validate_cable_ident(core, slot_index, module, &qsfp, + &cmis); + if (err) + return err; + + if (qsfp) { + /* For QSFP/CMIS module-defined thresholds are located in page + * 02h, otherwise in page 03h. + */ + if (cmis) + page = MLXSW_REG_MCIA_TH_PAGE_CMIS_NUM; + else + page = MLXSW_REG_MCIA_TH_PAGE_NUM; + mlxsw_reg_mcia_pack(mcia_pl, slot_index, module, 0, page, + MLXSW_REG_MCIA_TH_PAGE_OFF + off, + MLXSW_REG_MCIA_TH_ITEM_SIZE, + MLXSW_REG_MCIA_I2C_ADDR_LOW); + } else { + mlxsw_reg_mcia_pack(mcia_pl, slot_index, module, 0, + MLXSW_REG_MCIA_PAGE0_LO, + off, MLXSW_REG_MCIA_TH_ITEM_SIZE, + MLXSW_REG_MCIA_I2C_ADDR_HIGH); + } + + err = mlxsw_reg_query(core, MLXSW_REG(mcia), mcia_pl); + if (err) + return err; + + eeprom_tmp = mlxsw_reg_mcia_eeprom_data(mcia_pl); + memcpy(temp_thresh.buf, eeprom_tmp, MLXSW_REG_MCIA_TH_ITEM_SIZE); + *temp = temp_thresh.temp * 1000; + + return 0; +} + +int mlxsw_env_get_module_info(struct net_device *netdev, + struct mlxsw_core *mlxsw_core, u8 slot_index, + int module, struct ethtool_modinfo *modinfo) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + u8 module_info[MLXSW_REG_MCIA_EEPROM_MODULE_INFO_SIZE]; + u16 offset = MLXSW_REG_MCIA_EEPROM_MODULE_INFO_SIZE; + u8 module_rev_id, module_id, diag_mon; + unsigned int read_size; + int err; + + if (!mlxsw_env_linecard_is_active(mlxsw_env, slot_index)) { + netdev_err(netdev, "Cannot read EEPROM of module on an inactive line card\n"); + return -EIO; + } + + err = mlxsw_env_validate_module_type(mlxsw_core, slot_index, module); + if (err) { + netdev_err(netdev, + "EEPROM is not equipped on port module type"); + return err; + } + + err = mlxsw_env_query_module_eeprom(mlxsw_core, slot_index, module, 0, + offset, module_info, false, + &read_size); + if (err) + return err; + + if (read_size < offset) + return -EIO; + + module_rev_id = module_info[MLXSW_REG_MCIA_EEPROM_MODULE_INFO_REV_ID]; + module_id = module_info[MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID]; + + switch (module_id) { + case MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP: + modinfo->type = ETH_MODULE_SFF_8436; + modinfo->eeprom_len = ETH_MODULE_SFF_8436_MAX_LEN; + break; + case MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP_PLUS: + case MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP28: + if (module_id == MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP28 || + module_rev_id >= + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_REV_ID_8636) { + modinfo->type = ETH_MODULE_SFF_8636; + modinfo->eeprom_len = ETH_MODULE_SFF_8636_MAX_LEN; + } else { + modinfo->type = ETH_MODULE_SFF_8436; + modinfo->eeprom_len = ETH_MODULE_SFF_8436_MAX_LEN; + } + break; + case MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_SFP: + /* Verify if transceiver provides diagnostic monitoring page */ + err = mlxsw_env_query_module_eeprom(mlxsw_core, slot_index, + module, SFP_DIAGMON, 1, + &diag_mon, false, + &read_size); + if (err) + return err; + + if (read_size < 1) + return -EIO; + + modinfo->type = ETH_MODULE_SFF_8472; + if (diag_mon) + modinfo->eeprom_len = ETH_MODULE_SFF_8472_LEN; + else + modinfo->eeprom_len = ETH_MODULE_SFF_8472_LEN / 2; + break; + case MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP_DD: + case MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_OSFP: + /* Use SFF_8636 as base type. ethtool should recognize specific + * type through the identifier value. + */ + modinfo->type = ETH_MODULE_SFF_8636; + /* Verify if module EEPROM is a flat memory. In case of flat + * memory only page 00h (0-255 bytes) can be read. Otherwise + * upper pages 01h and 02h can also be read. Upper pages 10h + * and 11h are currently not supported by the driver. + */ + if (module_info[MLXSW_REG_MCIA_EEPROM_MODULE_INFO_TYPE_ID] & + MLXSW_REG_MCIA_EEPROM_CMIS_FLAT_MEMORY) + modinfo->eeprom_len = ETH_MODULE_SFF_8636_LEN; + else + modinfo->eeprom_len = ETH_MODULE_SFF_8472_LEN; + break; + default: + return -EINVAL; + } + + return 0; +} +EXPORT_SYMBOL(mlxsw_env_get_module_info); + +int mlxsw_env_get_module_eeprom(struct net_device *netdev, + struct mlxsw_core *mlxsw_core, u8 slot_index, + int module, struct ethtool_eeprom *ee, + u8 *data) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + int offset = ee->offset; + unsigned int read_size; + bool qsfp, cmis; + int i = 0; + int err; + + if (!ee->len) + return -EINVAL; + + if (!mlxsw_env_linecard_is_active(mlxsw_env, slot_index)) { + netdev_err(netdev, "Cannot read EEPROM of module on an inactive line card\n"); + return -EIO; + } + + memset(data, 0, ee->len); + /* Validate module identifier value. */ + err = mlxsw_env_validate_cable_ident(mlxsw_core, slot_index, module, + &qsfp, &cmis); + if (err) + return err; + + while (i < ee->len) { + err = mlxsw_env_query_module_eeprom(mlxsw_core, slot_index, + module, offset, + ee->len - i, data + i, + qsfp, &read_size); + if (err) { + netdev_err(netdev, "Eeprom query failed\n"); + return err; + } + + i += read_size; + offset += read_size; + } + + return 0; +} +EXPORT_SYMBOL(mlxsw_env_get_module_eeprom); + +static int mlxsw_env_mcia_status_process(const char *mcia_pl, + struct netlink_ext_ack *extack) +{ + u8 status = mlxsw_reg_mcia_status_get(mcia_pl); + + switch (status) { + case MLXSW_REG_MCIA_STATUS_GOOD: + return 0; + case MLXSW_REG_MCIA_STATUS_NO_EEPROM_MODULE: + NL_SET_ERR_MSG_MOD(extack, "No response from module's EEPROM"); + return -EIO; + case MLXSW_REG_MCIA_STATUS_MODULE_NOT_SUPPORTED: + NL_SET_ERR_MSG_MOD(extack, "Module type not supported by the device"); + return -EOPNOTSUPP; + case MLXSW_REG_MCIA_STATUS_MODULE_NOT_CONNECTED: + NL_SET_ERR_MSG_MOD(extack, "No module present indication"); + return -EIO; + case MLXSW_REG_MCIA_STATUS_I2C_ERROR: + NL_SET_ERR_MSG_MOD(extack, "Error occurred while trying to access module's EEPROM using I2C"); + return -EIO; + case MLXSW_REG_MCIA_STATUS_MODULE_DISABLED: + NL_SET_ERR_MSG_MOD(extack, "Module is disabled"); + return -EIO; + default: + NL_SET_ERR_MSG_MOD(extack, "Unknown error"); + return -EIO; + } +} + +int +mlxsw_env_get_module_eeprom_by_page(struct mlxsw_core *mlxsw_core, + u8 slot_index, u8 module, + const struct ethtool_module_eeprom *page, + struct netlink_ext_ack *extack) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + u32 bytes_read = 0; + u16 device_addr; + int err; + + if (!mlxsw_env_linecard_is_active(mlxsw_env, slot_index)) { + NL_SET_ERR_MSG_MOD(extack, + "Cannot read EEPROM of module on an inactive line card"); + return -EIO; + } + + err = mlxsw_env_validate_module_type(mlxsw_core, slot_index, module); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "EEPROM is not equipped on port module type"); + return err; + } + + /* Offset cannot be larger than 2 * ETH_MODULE_EEPROM_PAGE_LEN */ + device_addr = page->offset; + + while (bytes_read < page->length) { + char mcia_pl[MLXSW_REG_MCIA_LEN]; + char *eeprom_tmp; + u8 size; + + size = min_t(u8, page->length - bytes_read, + MLXSW_REG_MCIA_EEPROM_SIZE); + + mlxsw_reg_mcia_pack(mcia_pl, slot_index, module, 0, page->page, + device_addr + bytes_read, size, + page->i2c_address); + mlxsw_reg_mcia_bank_number_set(mcia_pl, page->bank); + + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mcia), mcia_pl); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to access module's EEPROM"); + return err; + } + + err = mlxsw_env_mcia_status_process(mcia_pl, extack); + if (err) + return err; + + eeprom_tmp = mlxsw_reg_mcia_eeprom_data(mcia_pl); + memcpy(page->data + bytes_read, eeprom_tmp, size); + bytes_read += size; + } + + return bytes_read; +} +EXPORT_SYMBOL(mlxsw_env_get_module_eeprom_by_page); + +static int mlxsw_env_module_reset(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module) +{ + char pmaos_pl[MLXSW_REG_PMAOS_LEN]; + + mlxsw_reg_pmaos_pack(pmaos_pl, slot_index, module); + mlxsw_reg_pmaos_rst_set(pmaos_pl, true); + + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(pmaos), pmaos_pl); +} + +int mlxsw_env_reset_module(struct net_device *netdev, + struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module, u32 *flags) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + struct mlxsw_env_module_info *module_info; + u32 req = *flags; + int err; + + if (!(req & ETH_RESET_PHY) && + !(req & (ETH_RESET_PHY << ETH_RESET_SHARED_SHIFT))) + return 0; + + if (!mlxsw_env_linecard_is_active(mlxsw_env, slot_index)) { + netdev_err(netdev, "Cannot reset module on an inactive line card\n"); + return -EIO; + } + + mutex_lock(&mlxsw_env->line_cards_lock); + + err = __mlxsw_env_validate_module_type(mlxsw_core, slot_index, module); + if (err) { + netdev_err(netdev, "Reset module is not supported on port module type\n"); + goto out; + } + + module_info = mlxsw_env_module_info_get(mlxsw_core, slot_index, module); + if (module_info->num_ports_up) { + netdev_err(netdev, "Cannot reset module when ports using it are administratively up\n"); + err = -EINVAL; + goto out; + } + + if (module_info->num_ports_mapped > 1 && + !(req & (ETH_RESET_PHY << ETH_RESET_SHARED_SHIFT))) { + netdev_err(netdev, "Cannot reset module without \"phy-shared\" flag when shared by multiple ports\n"); + err = -EINVAL; + goto out; + } + + err = mlxsw_env_module_reset(mlxsw_core, slot_index, module); + if (err) { + netdev_err(netdev, "Failed to reset module\n"); + goto out; + } + + *flags &= ~(ETH_RESET_PHY | (ETH_RESET_PHY << ETH_RESET_SHARED_SHIFT)); + +out: + mutex_unlock(&mlxsw_env->line_cards_lock); + return err; +} +EXPORT_SYMBOL(mlxsw_env_reset_module); + +int +mlxsw_env_get_module_power_mode(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module, + struct ethtool_module_power_mode_params *params, + struct netlink_ext_ack *extack) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + struct mlxsw_env_module_info *module_info; + char mcion_pl[MLXSW_REG_MCION_LEN]; + u32 status_bits; + int err = 0; + + mutex_lock(&mlxsw_env->line_cards_lock); + + err = __mlxsw_env_validate_module_type(mlxsw_core, slot_index, module); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Power mode is not supported on port module type"); + goto out; + } + + module_info = mlxsw_env_module_info_get(mlxsw_core, slot_index, module); + params->policy = module_info->power_mode_policy; + + /* Avoid accessing an inactive line card, as it will result in an error. */ + if (!__mlxsw_env_linecard_is_active(mlxsw_env, slot_index)) + goto out; + + mlxsw_reg_mcion_pack(mcion_pl, slot_index, module); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mcion), mcion_pl); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to retrieve module's power mode"); + goto out; + } + + status_bits = mlxsw_reg_mcion_module_status_bits_get(mcion_pl); + if (!(status_bits & MLXSW_REG_MCION_MODULE_STATUS_BITS_PRESENT_MASK)) + goto out; + + if (status_bits & MLXSW_REG_MCION_MODULE_STATUS_BITS_LOW_POWER_MASK) + params->mode = ETHTOOL_MODULE_POWER_MODE_LOW; + else + params->mode = ETHTOOL_MODULE_POWER_MODE_HIGH; + +out: + mutex_unlock(&mlxsw_env->line_cards_lock); + return err; +} +EXPORT_SYMBOL(mlxsw_env_get_module_power_mode); + +static int mlxsw_env_module_enable_set(struct mlxsw_core *mlxsw_core, + u8 slot_index, u8 module, bool enable) +{ + enum mlxsw_reg_pmaos_admin_status admin_status; + char pmaos_pl[MLXSW_REG_PMAOS_LEN]; + + mlxsw_reg_pmaos_pack(pmaos_pl, slot_index, module); + admin_status = enable ? MLXSW_REG_PMAOS_ADMIN_STATUS_ENABLED : + MLXSW_REG_PMAOS_ADMIN_STATUS_DISABLED; + mlxsw_reg_pmaos_admin_status_set(pmaos_pl, admin_status); + mlxsw_reg_pmaos_ase_set(pmaos_pl, true); + + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(pmaos), pmaos_pl); +} + +static int mlxsw_env_module_low_power_set(struct mlxsw_core *mlxsw_core, + u8 slot_index, u8 module, + bool low_power) +{ + u16 eeprom_override_mask, eeprom_override; + char pmmp_pl[MLXSW_REG_PMMP_LEN]; + + mlxsw_reg_pmmp_pack(pmmp_pl, slot_index, module); + mlxsw_reg_pmmp_sticky_set(pmmp_pl, true); + /* Mask all the bits except low power mode. */ + eeprom_override_mask = ~MLXSW_REG_PMMP_EEPROM_OVERRIDE_LOW_POWER_MASK; + mlxsw_reg_pmmp_eeprom_override_mask_set(pmmp_pl, eeprom_override_mask); + eeprom_override = low_power ? MLXSW_REG_PMMP_EEPROM_OVERRIDE_LOW_POWER_MASK : + 0; + mlxsw_reg_pmmp_eeprom_override_set(pmmp_pl, eeprom_override); + + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(pmmp), pmmp_pl); +} + +static int __mlxsw_env_set_module_power_mode(struct mlxsw_core *mlxsw_core, + u8 slot_index, u8 module, + bool low_power, + struct netlink_ext_ack *extack) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + int err; + + /* Avoid accessing an inactive line card, as it will result in an error. + * Cached configuration will be applied by mlxsw_env_got_active() when + * line card becomes active. + */ + if (!__mlxsw_env_linecard_is_active(mlxsw_env, slot_index)) + return 0; + + err = mlxsw_env_module_enable_set(mlxsw_core, slot_index, module, false); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to disable module"); + return err; + } + + err = mlxsw_env_module_low_power_set(mlxsw_core, slot_index, module, + low_power); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to set module's power mode"); + goto err_module_low_power_set; + } + + err = mlxsw_env_module_enable_set(mlxsw_core, slot_index, module, true); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to enable module"); + goto err_module_enable_set; + } + + return 0; + +err_module_enable_set: + mlxsw_env_module_low_power_set(mlxsw_core, slot_index, module, + !low_power); +err_module_low_power_set: + mlxsw_env_module_enable_set(mlxsw_core, slot_index, module, true); + return err; +} + +static int +mlxsw_env_set_module_power_mode_apply(struct mlxsw_core *mlxsw_core, + u8 slot_index, u8 module, + enum ethtool_module_power_mode_policy policy, + struct netlink_ext_ack *extack) +{ + struct mlxsw_env_module_info *module_info; + bool low_power; + int err = 0; + + err = __mlxsw_env_validate_module_type(mlxsw_core, slot_index, module); + if (err) { + NL_SET_ERR_MSG_MOD(extack, + "Power mode set is not supported on port module type"); + goto out; + } + + module_info = mlxsw_env_module_info_get(mlxsw_core, slot_index, module); + if (module_info->power_mode_policy == policy) + goto out; + + /* If any ports are up, we are already in high power mode. */ + if (module_info->num_ports_up) + goto out_set_policy; + + low_power = policy == ETHTOOL_MODULE_POWER_MODE_POLICY_AUTO; + err = __mlxsw_env_set_module_power_mode(mlxsw_core, slot_index, module, + low_power, extack); + if (err) + goto out; + +out_set_policy: + module_info->power_mode_policy = policy; +out: + return err; +} + +int +mlxsw_env_set_module_power_mode(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module, + enum ethtool_module_power_mode_policy policy, + struct netlink_ext_ack *extack) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + int err; + + if (policy != ETHTOOL_MODULE_POWER_MODE_POLICY_HIGH && + policy != ETHTOOL_MODULE_POWER_MODE_POLICY_AUTO) { + NL_SET_ERR_MSG_MOD(extack, "Unsupported power mode policy"); + return -EOPNOTSUPP; + } + + mutex_lock(&mlxsw_env->line_cards_lock); + err = mlxsw_env_set_module_power_mode_apply(mlxsw_core, slot_index, + module, policy, extack); + mutex_unlock(&mlxsw_env->line_cards_lock); + + return err; +} +EXPORT_SYMBOL(mlxsw_env_set_module_power_mode); + +static int mlxsw_env_module_has_temp_sensor(struct mlxsw_core *mlxsw_core, + u8 slot_index, u8 module, + bool *p_has_temp_sensor) +{ + char mtbr_pl[MLXSW_REG_MTBR_LEN]; + u16 temp; + int err; + + mlxsw_reg_mtbr_pack(mtbr_pl, slot_index, + MLXSW_REG_MTBR_BASE_MODULE_INDEX + module, 1); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mtbr), mtbr_pl); + if (err) + return err; + + mlxsw_reg_mtbr_temp_unpack(mtbr_pl, 0, &temp, NULL); + + switch (temp) { + case MLXSW_REG_MTBR_BAD_SENS_INFO: + case MLXSW_REG_MTBR_NO_CONN: + case MLXSW_REG_MTBR_NO_TEMP_SENS: + case MLXSW_REG_MTBR_INDEX_NA: + *p_has_temp_sensor = false; + break; + default: + *p_has_temp_sensor = temp ? true : false; + } + return 0; +} + +static int +mlxsw_env_temp_event_set(struct mlxsw_core *mlxsw_core, u8 slot_index, + u16 sensor_index, bool enable) +{ + char mtmp_pl[MLXSW_REG_MTMP_LEN] = {0}; + enum mlxsw_reg_mtmp_tee tee; + int err, threshold_hi; + + mlxsw_reg_mtmp_slot_index_set(mtmp_pl, slot_index); + mlxsw_reg_mtmp_sensor_index_set(mtmp_pl, sensor_index); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mtmp), mtmp_pl); + if (err) + return err; + + if (enable) { + err = mlxsw_env_module_temp_thresholds_get(mlxsw_core, + slot_index, + sensor_index - + MLXSW_REG_MTMP_MODULE_INDEX_MIN, + SFP_TEMP_HIGH_WARN, + &threshold_hi); + /* In case it is not possible to query the module's threshold, + * use the default value. + */ + if (err) + threshold_hi = MLXSW_REG_MTMP_THRESH_HI; + else + /* mlxsw_env_module_temp_thresholds_get() multiplies + * Celsius degrees by 1000 whereas MTMP expects + * temperature in 0.125 Celsius degrees units. + * Convert threshold_hi to correct units. + */ + threshold_hi = threshold_hi / 1000 * 8; + + mlxsw_reg_mtmp_temperature_threshold_hi_set(mtmp_pl, threshold_hi); + mlxsw_reg_mtmp_temperature_threshold_lo_set(mtmp_pl, threshold_hi - + MLXSW_REG_MTMP_HYSTERESIS_TEMP); + } + tee = enable ? MLXSW_REG_MTMP_TEE_GENERATE_EVENT : MLXSW_REG_MTMP_TEE_NO_EVENT; + mlxsw_reg_mtmp_tee_set(mtmp_pl, tee); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mtmp), mtmp_pl); +} + +static int mlxsw_env_module_temp_event_enable(struct mlxsw_core *mlxsw_core, + u8 slot_index) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + int i, err, sensor_index; + bool has_temp_sensor; + + for (i = 0; i < mlxsw_env->line_cards[slot_index]->module_count; i++) { + err = mlxsw_env_module_has_temp_sensor(mlxsw_core, slot_index, + i, &has_temp_sensor); + if (err) + return err; + + if (!has_temp_sensor) + continue; + + sensor_index = i + MLXSW_REG_MTMP_MODULE_INDEX_MIN; + err = mlxsw_env_temp_event_set(mlxsw_core, slot_index, + sensor_index, true); + if (err) + return err; + } + + return 0; +} + +struct mlxsw_env_module_temp_warn_event { + struct mlxsw_env *mlxsw_env; + char mtwe_pl[MLXSW_REG_MTWE_LEN]; + struct work_struct work; +}; + +static void mlxsw_env_mtwe_event_work(struct work_struct *work) +{ + struct mlxsw_env_module_temp_warn_event *event; + struct mlxsw_env_module_info *module_info; + struct mlxsw_env *mlxsw_env; + int i, sensor_warning; + bool is_overheat; + + event = container_of(work, struct mlxsw_env_module_temp_warn_event, + work); + mlxsw_env = event->mlxsw_env; + + for (i = 0; i < mlxsw_env->max_module_count; i++) { + /* 64-127 of sensor_index are mapped to the port modules + * sequentially (module 0 is mapped to sensor_index 64, + * module 1 to sensor_index 65 and so on) + */ + sensor_warning = + mlxsw_reg_mtwe_sensor_warning_get(event->mtwe_pl, + i + MLXSW_REG_MTMP_MODULE_INDEX_MIN); + mutex_lock(&mlxsw_env->line_cards_lock); + /* MTWE only supports main board. */ + module_info = mlxsw_env_module_info_get(mlxsw_env->core, 0, i); + is_overheat = module_info->is_overheat; + + if ((is_overheat && sensor_warning) || + (!is_overheat && !sensor_warning)) { + /* Current state is "warning" and MTWE still reports + * warning OR current state in "no warning" and MTWE + * does not report warning. + */ + mutex_unlock(&mlxsw_env->line_cards_lock); + continue; + } else if (is_overheat && !sensor_warning) { + /* MTWE reports "no warning", turn is_overheat off. + */ + module_info->is_overheat = false; + mutex_unlock(&mlxsw_env->line_cards_lock); + } else { + /* Current state is "no warning" and MTWE reports + * "warning", increase the counter and turn is_overheat + * on. + */ + module_info->is_overheat = true; + module_info->module_overheat_counter++; + mutex_unlock(&mlxsw_env->line_cards_lock); + } + } + + kfree(event); +} + +static void +mlxsw_env_mtwe_listener_func(const struct mlxsw_reg_info *reg, char *mtwe_pl, + void *priv) +{ + struct mlxsw_env_module_temp_warn_event *event; + struct mlxsw_env *mlxsw_env = priv; + + event = kmalloc(sizeof(*event), GFP_ATOMIC); + if (!event) + return; + + event->mlxsw_env = mlxsw_env; + memcpy(event->mtwe_pl, mtwe_pl, MLXSW_REG_MTWE_LEN); + INIT_WORK(&event->work, mlxsw_env_mtwe_event_work); + mlxsw_core_schedule_work(&event->work); +} + +static const struct mlxsw_listener mlxsw_env_temp_warn_listener = + MLXSW_CORE_EVENTL(mlxsw_env_mtwe_listener_func, MTWE); + +static int mlxsw_env_temp_warn_event_register(struct mlxsw_core *mlxsw_core) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + + return mlxsw_core_trap_register(mlxsw_core, + &mlxsw_env_temp_warn_listener, + mlxsw_env); +} + +static void mlxsw_env_temp_warn_event_unregister(struct mlxsw_env *mlxsw_env) +{ + mlxsw_core_trap_unregister(mlxsw_env->core, + &mlxsw_env_temp_warn_listener, mlxsw_env); +} + +struct mlxsw_env_module_plug_unplug_event { + struct mlxsw_env *mlxsw_env; + u8 slot_index; + u8 module; + struct work_struct work; +}; + +static void mlxsw_env_pmpe_event_work(struct work_struct *work) +{ + struct mlxsw_env_module_plug_unplug_event *event; + struct mlxsw_env_module_info *module_info; + struct mlxsw_env *mlxsw_env; + bool has_temp_sensor; + u16 sensor_index; + int err; + + event = container_of(work, struct mlxsw_env_module_plug_unplug_event, + work); + mlxsw_env = event->mlxsw_env; + + mutex_lock(&mlxsw_env->line_cards_lock); + module_info = mlxsw_env_module_info_get(mlxsw_env->core, + event->slot_index, + event->module); + module_info->is_overheat = false; + mutex_unlock(&mlxsw_env->line_cards_lock); + + err = mlxsw_env_module_has_temp_sensor(mlxsw_env->core, + event->slot_index, + event->module, + &has_temp_sensor); + /* Do not disable events on modules without sensors or faulty sensors + * because FW returns errors. + */ + if (err) + goto out; + + if (!has_temp_sensor) + goto out; + + sensor_index = event->module + MLXSW_REG_MTMP_MODULE_INDEX_MIN; + mlxsw_env_temp_event_set(mlxsw_env->core, event->slot_index, + sensor_index, true); + +out: + kfree(event); +} + +static void +mlxsw_env_pmpe_listener_func(const struct mlxsw_reg_info *reg, char *pmpe_pl, + void *priv) +{ + u8 slot_index = mlxsw_reg_pmpe_slot_index_get(pmpe_pl); + struct mlxsw_env_module_plug_unplug_event *event; + enum mlxsw_reg_pmpe_module_status module_status; + u8 module = mlxsw_reg_pmpe_module_get(pmpe_pl); + struct mlxsw_env *mlxsw_env = priv; + + if (WARN_ON_ONCE(module >= mlxsw_env->max_module_count || + slot_index >= mlxsw_env->num_of_slots)) + return; + + module_status = mlxsw_reg_pmpe_module_status_get(pmpe_pl); + if (module_status != MLXSW_REG_PMPE_MODULE_STATUS_PLUGGED_ENABLED) + return; + + event = kmalloc(sizeof(*event), GFP_ATOMIC); + if (!event) + return; + + event->mlxsw_env = mlxsw_env; + event->slot_index = slot_index; + event->module = module; + INIT_WORK(&event->work, mlxsw_env_pmpe_event_work); + mlxsw_core_schedule_work(&event->work); +} + +static const struct mlxsw_listener mlxsw_env_module_plug_listener = + MLXSW_CORE_EVENTL(mlxsw_env_pmpe_listener_func, PMPE); + +static int +mlxsw_env_module_plug_event_register(struct mlxsw_core *mlxsw_core) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + + return mlxsw_core_trap_register(mlxsw_core, + &mlxsw_env_module_plug_listener, + mlxsw_env); +} + +static void +mlxsw_env_module_plug_event_unregister(struct mlxsw_env *mlxsw_env) +{ + mlxsw_core_trap_unregister(mlxsw_env->core, + &mlxsw_env_module_plug_listener, + mlxsw_env); +} + +static int +mlxsw_env_module_oper_state_event_enable(struct mlxsw_core *mlxsw_core, + u8 slot_index) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + int i, err; + + for (i = 0; i < mlxsw_env->line_cards[slot_index]->module_count; i++) { + char pmaos_pl[MLXSW_REG_PMAOS_LEN]; + + mlxsw_reg_pmaos_pack(pmaos_pl, slot_index, i); + mlxsw_reg_pmaos_e_set(pmaos_pl, + MLXSW_REG_PMAOS_E_GENERATE_EVENT); + mlxsw_reg_pmaos_ee_set(pmaos_pl, true); + err = mlxsw_reg_write(mlxsw_core, MLXSW_REG(pmaos), pmaos_pl); + if (err) + return err; + } + return 0; +} + +int +mlxsw_env_module_overheat_counter_get(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module, u64 *p_counter) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + struct mlxsw_env_module_info *module_info; + + mutex_lock(&mlxsw_env->line_cards_lock); + module_info = mlxsw_env_module_info_get(mlxsw_core, slot_index, module); + *p_counter = module_info->module_overheat_counter; + mutex_unlock(&mlxsw_env->line_cards_lock); + + return 0; +} +EXPORT_SYMBOL(mlxsw_env_module_overheat_counter_get); + +void mlxsw_env_module_port_map(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + struct mlxsw_env_module_info *module_info; + + mutex_lock(&mlxsw_env->line_cards_lock); + module_info = mlxsw_env_module_info_get(mlxsw_core, slot_index, module); + module_info->num_ports_mapped++; + mutex_unlock(&mlxsw_env->line_cards_lock); +} +EXPORT_SYMBOL(mlxsw_env_module_port_map); + +void mlxsw_env_module_port_unmap(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + struct mlxsw_env_module_info *module_info; + + mutex_lock(&mlxsw_env->line_cards_lock); + module_info = mlxsw_env_module_info_get(mlxsw_core, slot_index, module); + module_info->num_ports_mapped--; + mutex_unlock(&mlxsw_env->line_cards_lock); +} +EXPORT_SYMBOL(mlxsw_env_module_port_unmap); + +int mlxsw_env_module_port_up(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + struct mlxsw_env_module_info *module_info; + int err = 0; + + mutex_lock(&mlxsw_env->line_cards_lock); + + module_info = mlxsw_env_module_info_get(mlxsw_core, slot_index, module); + if (module_info->power_mode_policy != + ETHTOOL_MODULE_POWER_MODE_POLICY_AUTO) + goto out_inc; + + if (module_info->num_ports_up != 0) + goto out_inc; + + /* Transition to high power mode following first port using the module + * being put administratively up. + */ + err = __mlxsw_env_set_module_power_mode(mlxsw_core, slot_index, module, + false, NULL); + if (err) + goto out_unlock; + +out_inc: + module_info->num_ports_up++; +out_unlock: + mutex_unlock(&mlxsw_env->line_cards_lock); + return err; +} +EXPORT_SYMBOL(mlxsw_env_module_port_up); + +void mlxsw_env_module_port_down(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + struct mlxsw_env_module_info *module_info; + + mutex_lock(&mlxsw_env->line_cards_lock); + + module_info = mlxsw_env_module_info_get(mlxsw_core, slot_index, module); + module_info->num_ports_up--; + + if (module_info->power_mode_policy != + ETHTOOL_MODULE_POWER_MODE_POLICY_AUTO) + goto out_unlock; + + if (module_info->num_ports_up != 0) + goto out_unlock; + + /* Transition to low power mode following last port using the module + * being put administratively down. + */ + __mlxsw_env_set_module_power_mode(mlxsw_core, slot_index, module, true, + NULL); + +out_unlock: + mutex_unlock(&mlxsw_env->line_cards_lock); +} +EXPORT_SYMBOL(mlxsw_env_module_port_down); + +static int mlxsw_env_line_cards_alloc(struct mlxsw_env *env) +{ + struct mlxsw_env_module_info *module_info; + int i, j; + + for (i = 0; i < env->num_of_slots; i++) { + env->line_cards[i] = kzalloc(struct_size(env->line_cards[i], + module_info, + env->max_module_count), + GFP_KERNEL); + if (!env->line_cards[i]) + goto kzalloc_err; + + /* Firmware defaults to high power mode policy where modules + * are transitioned to high power mode following plug-in. + */ + for (j = 0; j < env->max_module_count; j++) { + module_info = &env->line_cards[i]->module_info[j]; + module_info->power_mode_policy = + ETHTOOL_MODULE_POWER_MODE_POLICY_HIGH; + } + } + + return 0; + +kzalloc_err: + for (i--; i >= 0; i--) + kfree(env->line_cards[i]); + return -ENOMEM; +} + +static void mlxsw_env_line_cards_free(struct mlxsw_env *env) +{ + int i = env->num_of_slots; + + for (i--; i >= 0; i--) + kfree(env->line_cards[i]); +} + +static int +mlxsw_env_module_event_enable(struct mlxsw_env *mlxsw_env, u8 slot_index) +{ + int err; + + err = mlxsw_env_module_oper_state_event_enable(mlxsw_env->core, + slot_index); + if (err) + return err; + + err = mlxsw_env_module_temp_event_enable(mlxsw_env->core, slot_index); + if (err) + return err; + + return 0; +} + +static void +mlxsw_env_module_event_disable(struct mlxsw_env *mlxsw_env, u8 slot_index) +{ +} + +static int +mlxsw_env_module_type_set(struct mlxsw_core *mlxsw_core, u8 slot_index) +{ + struct mlxsw_env *mlxsw_env = mlxsw_core_env(mlxsw_core); + int i; + + for (i = 0; i < mlxsw_env->line_cards[slot_index]->module_count; i++) { + struct mlxsw_env_module_info *module_info; + char pmtm_pl[MLXSW_REG_PMTM_LEN]; + int err; + + mlxsw_reg_pmtm_pack(pmtm_pl, slot_index, i); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(pmtm), pmtm_pl); + if (err) + return err; + + module_info = mlxsw_env_module_info_get(mlxsw_core, slot_index, + i); + module_info->type = mlxsw_reg_pmtm_module_type_get(pmtm_pl); + } + + return 0; +} + +static void +mlxsw_env_linecard_modules_power_mode_apply(struct mlxsw_core *mlxsw_core, + struct mlxsw_env *env, + u8 slot_index) +{ + int i; + + for (i = 0; i < env->line_cards[slot_index]->module_count; i++) { + enum ethtool_module_power_mode_policy policy; + struct mlxsw_env_module_info *module_info; + struct netlink_ext_ack extack; + int err; + + module_info = &env->line_cards[slot_index]->module_info[i]; + policy = module_info->power_mode_policy; + err = mlxsw_env_set_module_power_mode_apply(mlxsw_core, + slot_index, i, + policy, &extack); + if (err) + dev_err(env->bus_info->dev, "%s\n", extack._msg); + } +} + +static void +mlxsw_env_got_active(struct mlxsw_core *mlxsw_core, u8 slot_index, void *priv) +{ + struct mlxsw_env *mlxsw_env = priv; + char mgpir_pl[MLXSW_REG_MGPIR_LEN]; + int err; + + mutex_lock(&mlxsw_env->line_cards_lock); + if (__mlxsw_env_linecard_is_active(mlxsw_env, slot_index)) + goto out_unlock; + + mlxsw_reg_mgpir_pack(mgpir_pl, slot_index); + err = mlxsw_reg_query(mlxsw_env->core, MLXSW_REG(mgpir), mgpir_pl); + if (err) + goto out_unlock; + + mlxsw_reg_mgpir_unpack(mgpir_pl, NULL, NULL, NULL, + &mlxsw_env->line_cards[slot_index]->module_count, + NULL); + + err = mlxsw_env_module_event_enable(mlxsw_env, slot_index); + if (err) { + dev_err(mlxsw_env->bus_info->dev, "Failed to enable port module events for line card in slot %d\n", + slot_index); + goto err_mlxsw_env_module_event_enable; + } + err = mlxsw_env_module_type_set(mlxsw_env->core, slot_index); + if (err) { + dev_err(mlxsw_env->bus_info->dev, "Failed to set modules' type for line card in slot %d\n", + slot_index); + goto err_type_set; + } + + mlxsw_env->line_cards[slot_index]->active = true; + /* Apply power mode policy. */ + mlxsw_env_linecard_modules_power_mode_apply(mlxsw_core, mlxsw_env, + slot_index); + mutex_unlock(&mlxsw_env->line_cards_lock); + + return; + +err_type_set: + mlxsw_env_module_event_disable(mlxsw_env, slot_index); +err_mlxsw_env_module_event_enable: +out_unlock: + mutex_unlock(&mlxsw_env->line_cards_lock); +} + +static void +mlxsw_env_got_inactive(struct mlxsw_core *mlxsw_core, u8 slot_index, + void *priv) +{ + struct mlxsw_env *mlxsw_env = priv; + + mutex_lock(&mlxsw_env->line_cards_lock); + if (!__mlxsw_env_linecard_is_active(mlxsw_env, slot_index)) + goto out_unlock; + mlxsw_env->line_cards[slot_index]->active = false; + mlxsw_env_module_event_disable(mlxsw_env, slot_index); + mlxsw_env->line_cards[slot_index]->module_count = 0; +out_unlock: + mutex_unlock(&mlxsw_env->line_cards_lock); +} + +static struct mlxsw_linecards_event_ops mlxsw_env_event_ops = { + .got_active = mlxsw_env_got_active, + .got_inactive = mlxsw_env_got_inactive, +}; + +int mlxsw_env_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *bus_info, + struct mlxsw_env **p_env) +{ + u8 module_count, num_of_slots, max_module_count; + char mgpir_pl[MLXSW_REG_MGPIR_LEN]; + struct mlxsw_env *env; + int err; + + mlxsw_reg_mgpir_pack(mgpir_pl, 0); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mgpir), mgpir_pl); + if (err) + return err; + + mlxsw_reg_mgpir_unpack(mgpir_pl, NULL, NULL, NULL, &module_count, + &num_of_slots); + /* If the system is modular, get the maximum number of modules per-slot. + * Otherwise, get the maximum number of modules on the main board. + */ + max_module_count = num_of_slots ? + mlxsw_reg_mgpir_max_modules_per_slot_get(mgpir_pl) : + module_count; + + env = kzalloc(struct_size(env, line_cards, num_of_slots + 1), + GFP_KERNEL); + if (!env) + return -ENOMEM; + + env->core = mlxsw_core; + env->bus_info = bus_info; + env->num_of_slots = num_of_slots + 1; + env->max_module_count = max_module_count; + err = mlxsw_env_line_cards_alloc(env); + if (err) + goto err_mlxsw_env_line_cards_alloc; + + mutex_init(&env->line_cards_lock); + *p_env = env; + + err = mlxsw_linecards_event_ops_register(env->core, + &mlxsw_env_event_ops, env); + if (err) + goto err_linecards_event_ops_register; + + err = mlxsw_env_temp_warn_event_register(mlxsw_core); + if (err) + goto err_temp_warn_event_register; + + err = mlxsw_env_module_plug_event_register(mlxsw_core); + if (err) + goto err_module_plug_event_register; + + /* Set 'module_count' only for main board. Actual count for line card + * is to be set after line card is activated. + */ + env->line_cards[0]->module_count = num_of_slots ? 0 : module_count; + /* Enable events only for main board. Line card events are to be + * configured only after line card is activated. Before that, access to + * modules on line cards is not allowed. + */ + err = mlxsw_env_module_event_enable(env, 0); + if (err) + goto err_mlxsw_env_module_event_enable; + + err = mlxsw_env_module_type_set(mlxsw_core, 0); + if (err) + goto err_type_set; + + env->line_cards[0]->active = true; + + return 0; + +err_type_set: + mlxsw_env_module_event_disable(env, 0); +err_mlxsw_env_module_event_enable: + mlxsw_env_module_plug_event_unregister(env); +err_module_plug_event_register: + mlxsw_env_temp_warn_event_unregister(env); +err_temp_warn_event_register: + mlxsw_linecards_event_ops_unregister(env->core, + &mlxsw_env_event_ops, env); +err_linecards_event_ops_register: + mutex_destroy(&env->line_cards_lock); + mlxsw_env_line_cards_free(env); +err_mlxsw_env_line_cards_alloc: + kfree(env); + return err; +} + +void mlxsw_env_fini(struct mlxsw_env *env) +{ + env->line_cards[0]->active = false; + mlxsw_env_module_event_disable(env, 0); + mlxsw_env_module_plug_event_unregister(env); + /* Make sure there is no more event work scheduled. */ + mlxsw_core_flush_owq(); + mlxsw_env_temp_warn_event_unregister(env); + mlxsw_linecards_event_ops_unregister(env->core, + &mlxsw_env_event_ops, env); + mutex_destroy(&env->line_cards_lock); + mlxsw_env_line_cards_free(env); + kfree(env); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_env.h b/drivers/net/ethernet/mellanox/mlxsw/core_env.h new file mode 100644 index 000000000..a197e3ae0 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/core_env.h @@ -0,0 +1,68 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_CORE_ENV_H +#define _MLXSW_CORE_ENV_H + +#include <linux/ethtool.h> + +struct ethtool_modinfo; +struct ethtool_eeprom; + +int mlxsw_env_module_temp_thresholds_get(struct mlxsw_core *core, + u8 slot_index, int module, int off, + int *temp); + +int mlxsw_env_get_module_info(struct net_device *netdev, + struct mlxsw_core *mlxsw_core, u8 slot_index, + int module, struct ethtool_modinfo *modinfo); + +int mlxsw_env_get_module_eeprom(struct net_device *netdev, + struct mlxsw_core *mlxsw_core, u8 slot_index, + int module, struct ethtool_eeprom *ee, + u8 *data); + +int +mlxsw_env_get_module_eeprom_by_page(struct mlxsw_core *mlxsw_core, + u8 slot_index, u8 module, + const struct ethtool_module_eeprom *page, + struct netlink_ext_ack *extack); + +int mlxsw_env_reset_module(struct net_device *netdev, + struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module, u32 *flags); + +int +mlxsw_env_get_module_power_mode(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module, + struct ethtool_module_power_mode_params *params, + struct netlink_ext_ack *extack); + +int +mlxsw_env_set_module_power_mode(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module, + enum ethtool_module_power_mode_policy policy, + struct netlink_ext_ack *extack); + +int +mlxsw_env_module_overheat_counter_get(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module, u64 *p_counter); + +void mlxsw_env_module_port_map(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module); + +void mlxsw_env_module_port_unmap(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module); + +int mlxsw_env_module_port_up(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module); + +void mlxsw_env_module_port_down(struct mlxsw_core *mlxsw_core, u8 slot_index, + u8 module); + +int mlxsw_env_init(struct mlxsw_core *core, + const struct mlxsw_bus_info *bus_info, + struct mlxsw_env **p_env); +void mlxsw_env_fini(struct mlxsw_env *env); + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_hwmon.c b/drivers/net/ethernet/mellanox/mlxsw/core_hwmon.c new file mode 100644 index 000000000..0fd290d77 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/core_hwmon.c @@ -0,0 +1,941 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/types.h> +#include <linux/device.h> +#include <linux/sysfs.h> +#include <linux/hwmon.h> +#include <linux/err.h> +#include <linux/sfp.h> + +#include "core.h" +#include "core_env.h" + +#define MLXSW_HWMON_SENSORS_MAX_COUNT 64 +#define MLXSW_HWMON_MODULES_MAX_COUNT 64 +#define MLXSW_HWMON_GEARBOXES_MAX_COUNT 32 + +#define MLXSW_HWMON_ATTR_PER_SENSOR 3 +#define MLXSW_HWMON_ATTR_PER_MODULE 7 +#define MLXSW_HWMON_ATTR_PER_GEARBOX 4 +#define MLXSW_HWMON_DEV_NAME_LEN_MAX 16 + +#define MLXSW_HWMON_ATTR_COUNT (MLXSW_HWMON_SENSORS_MAX_COUNT * MLXSW_HWMON_ATTR_PER_SENSOR + \ + MLXSW_HWMON_MODULES_MAX_COUNT * MLXSW_HWMON_ATTR_PER_MODULE + \ + MLXSW_HWMON_GEARBOXES_MAX_COUNT * MLXSW_HWMON_ATTR_PER_GEARBOX + \ + MLXSW_MFCR_TACHOS_MAX + MLXSW_MFCR_PWMS_MAX) + +struct mlxsw_hwmon_attr { + struct device_attribute dev_attr; + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev; + unsigned int type_index; + char name[32]; +}; + +static int mlxsw_hwmon_get_attr_index(int index, int count) +{ + if (index >= count) + return index % count + MLXSW_REG_MTMP_GBOX_INDEX_MIN; + + return index; +} + +struct mlxsw_hwmon_dev { + char name[MLXSW_HWMON_DEV_NAME_LEN_MAX]; + struct mlxsw_hwmon *hwmon; + struct device *hwmon_dev; + struct attribute_group group; + const struct attribute_group *groups[2]; + struct attribute *attrs[MLXSW_HWMON_ATTR_COUNT + 1]; + struct mlxsw_hwmon_attr hwmon_attrs[MLXSW_HWMON_ATTR_COUNT]; + unsigned int attrs_count; + u8 sensor_count; + u8 module_sensor_max; + u8 slot_index; + bool active; +}; + +struct mlxsw_hwmon { + struct mlxsw_core *core; + const struct mlxsw_bus_info *bus_info; + struct mlxsw_hwmon_dev line_cards[]; +}; + +static ssize_t mlxsw_hwmon_temp_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev = mlxsw_hwmon_attr->mlxsw_hwmon_dev; + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + char mtmp_pl[MLXSW_REG_MTMP_LEN]; + int temp, index; + int err; + + index = mlxsw_hwmon_get_attr_index(mlxsw_hwmon_attr->type_index, + mlxsw_hwmon_dev->module_sensor_max); + mlxsw_reg_mtmp_pack(mtmp_pl, mlxsw_hwmon_dev->slot_index, index, false, + false); + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(mtmp), mtmp_pl); + if (err) { + dev_err(mlxsw_hwmon->bus_info->dev, "Failed to query temp sensor\n"); + return err; + } + mlxsw_reg_mtmp_unpack(mtmp_pl, &temp, NULL, NULL, NULL, NULL); + return sprintf(buf, "%d\n", temp); +} + +static ssize_t mlxsw_hwmon_temp_max_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev = mlxsw_hwmon_attr->mlxsw_hwmon_dev; + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + char mtmp_pl[MLXSW_REG_MTMP_LEN]; + int temp_max, index; + int err; + + index = mlxsw_hwmon_get_attr_index(mlxsw_hwmon_attr->type_index, + mlxsw_hwmon_dev->module_sensor_max); + mlxsw_reg_mtmp_pack(mtmp_pl, mlxsw_hwmon_dev->slot_index, index, false, + false); + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(mtmp), mtmp_pl); + if (err) { + dev_err(mlxsw_hwmon->bus_info->dev, "Failed to query temp sensor\n"); + return err; + } + mlxsw_reg_mtmp_unpack(mtmp_pl, NULL, &temp_max, NULL, NULL, NULL); + return sprintf(buf, "%d\n", temp_max); +} + +static ssize_t mlxsw_hwmon_temp_rst_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t len) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev = mlxsw_hwmon_attr->mlxsw_hwmon_dev; + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + char mtmp_pl[MLXSW_REG_MTMP_LEN]; + unsigned long val; + int index; + int err; + + err = kstrtoul(buf, 10, &val); + if (err) + return err; + if (val != 1) + return -EINVAL; + + index = mlxsw_hwmon_get_attr_index(mlxsw_hwmon_attr->type_index, + mlxsw_hwmon_dev->module_sensor_max); + + mlxsw_reg_mtmp_slot_index_set(mtmp_pl, mlxsw_hwmon_dev->slot_index); + mlxsw_reg_mtmp_sensor_index_set(mtmp_pl, index); + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(mtmp), mtmp_pl); + if (err) + return err; + mlxsw_reg_mtmp_mte_set(mtmp_pl, true); + mlxsw_reg_mtmp_mtr_set(mtmp_pl, true); + err = mlxsw_reg_write(mlxsw_hwmon->core, MLXSW_REG(mtmp), mtmp_pl); + if (err) { + dev_err(mlxsw_hwmon->bus_info->dev, "Failed to reset temp sensor history\n"); + return err; + } + return len; +} + +static ssize_t mlxsw_hwmon_fan_rpm_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev = mlxsw_hwmon_attr->mlxsw_hwmon_dev; + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + char mfsm_pl[MLXSW_REG_MFSM_LEN]; + int err; + + mlxsw_reg_mfsm_pack(mfsm_pl, mlxsw_hwmon_attr->type_index); + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(mfsm), mfsm_pl); + if (err) { + dev_err(mlxsw_hwmon->bus_info->dev, "Failed to query fan\n"); + return err; + } + return sprintf(buf, "%u\n", mlxsw_reg_mfsm_rpm_get(mfsm_pl)); +} + +static ssize_t mlxsw_hwmon_fan_fault_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev = mlxsw_hwmon_attr->mlxsw_hwmon_dev; + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + char fore_pl[MLXSW_REG_FORE_LEN]; + bool fault; + int err; + + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(fore), fore_pl); + if (err) { + dev_err(mlxsw_hwmon->bus_info->dev, "Failed to query fan\n"); + return err; + } + mlxsw_reg_fore_unpack(fore_pl, mlxsw_hwmon_attr->type_index, &fault); + + return sprintf(buf, "%u\n", fault); +} + +static ssize_t mlxsw_hwmon_pwm_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev = mlxsw_hwmon_attr->mlxsw_hwmon_dev; + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + char mfsc_pl[MLXSW_REG_MFSC_LEN]; + int err; + + mlxsw_reg_mfsc_pack(mfsc_pl, mlxsw_hwmon_attr->type_index, 0); + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(mfsc), mfsc_pl); + if (err) { + dev_err(mlxsw_hwmon->bus_info->dev, "Failed to query PWM\n"); + return err; + } + return sprintf(buf, "%u\n", + mlxsw_reg_mfsc_pwm_duty_cycle_get(mfsc_pl)); +} + +static ssize_t mlxsw_hwmon_pwm_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t len) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev = mlxsw_hwmon_attr->mlxsw_hwmon_dev; + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + char mfsc_pl[MLXSW_REG_MFSC_LEN]; + unsigned long val; + int err; + + err = kstrtoul(buf, 10, &val); + if (err) + return err; + if (val > 255) + return -EINVAL; + + mlxsw_reg_mfsc_pack(mfsc_pl, mlxsw_hwmon_attr->type_index, val); + err = mlxsw_reg_write(mlxsw_hwmon->core, MLXSW_REG(mfsc), mfsc_pl); + if (err) { + dev_err(mlxsw_hwmon->bus_info->dev, "Failed to write PWM\n"); + return err; + } + return len; +} + +static int mlxsw_hwmon_module_temp_get(struct device *dev, + struct device_attribute *attr, + int *p_temp) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev = mlxsw_hwmon_attr->mlxsw_hwmon_dev; + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + char mtmp_pl[MLXSW_REG_MTMP_LEN]; + u8 module; + int err; + + module = mlxsw_hwmon_attr->type_index - mlxsw_hwmon_dev->sensor_count; + mlxsw_reg_mtmp_pack(mtmp_pl, mlxsw_hwmon_dev->slot_index, + MLXSW_REG_MTMP_MODULE_INDEX_MIN + module, false, + false); + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(mtmp), mtmp_pl); + if (err) { + dev_err(dev, "Failed to query module temperature\n"); + return err; + } + mlxsw_reg_mtmp_unpack(mtmp_pl, p_temp, NULL, NULL, NULL, NULL); + + return 0; +} + +static ssize_t mlxsw_hwmon_module_temp_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + int err, temp; + + err = mlxsw_hwmon_module_temp_get(dev, attr, &temp); + if (err) + return err; + + return sprintf(buf, "%d\n", temp); +} + +static ssize_t mlxsw_hwmon_module_temp_fault_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev = mlxsw_hwmon_attr->mlxsw_hwmon_dev; + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + char mtbr_pl[MLXSW_REG_MTBR_LEN] = {0}; + u8 module, fault; + u16 temp; + int err; + + module = mlxsw_hwmon_attr->type_index - mlxsw_hwmon_dev->sensor_count; + mlxsw_reg_mtbr_pack(mtbr_pl, mlxsw_hwmon_dev->slot_index, + MLXSW_REG_MTBR_BASE_MODULE_INDEX + module, 1); + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(mtbr), mtbr_pl); + if (err) { + dev_err(dev, "Failed to query module temperature sensor\n"); + return err; + } + + mlxsw_reg_mtbr_temp_unpack(mtbr_pl, 0, &temp, NULL); + + /* Update status and temperature cache. */ + switch (temp) { + case MLXSW_REG_MTBR_BAD_SENS_INFO: + /* Untrusted cable is connected. Reading temperature from its + * sensor is faulty. + */ + fault = 1; + break; + case MLXSW_REG_MTBR_NO_CONN: + case MLXSW_REG_MTBR_NO_TEMP_SENS: + case MLXSW_REG_MTBR_INDEX_NA: + default: + fault = 0; + break; + } + + return sprintf(buf, "%u\n", fault); +} + +static int mlxsw_hwmon_module_temp_critical_get(struct device *dev, + struct device_attribute *attr, + int *p_temp) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev = mlxsw_hwmon_attr->mlxsw_hwmon_dev; + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + u8 module; + int err; + + module = mlxsw_hwmon_attr->type_index - mlxsw_hwmon_dev->sensor_count; + err = mlxsw_env_module_temp_thresholds_get(mlxsw_hwmon->core, + mlxsw_hwmon_dev->slot_index, + module, SFP_TEMP_HIGH_WARN, + p_temp); + if (err) { + dev_err(dev, "Failed to query module temperature thresholds\n"); + return err; + } + + return 0; +} + +static ssize_t +mlxsw_hwmon_module_temp_critical_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + int err, temp; + + err = mlxsw_hwmon_module_temp_critical_get(dev, attr, &temp); + if (err) + return err; + + return sprintf(buf, "%u\n", temp); +} + +static int mlxsw_hwmon_module_temp_emergency_get(struct device *dev, + struct device_attribute *attr, + int *p_temp) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev = mlxsw_hwmon_attr->mlxsw_hwmon_dev; + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + u8 module; + int err; + + module = mlxsw_hwmon_attr->type_index - mlxsw_hwmon_dev->sensor_count; + err = mlxsw_env_module_temp_thresholds_get(mlxsw_hwmon->core, + mlxsw_hwmon_dev->slot_index, + module, SFP_TEMP_HIGH_ALARM, + p_temp); + if (err) { + dev_err(dev, "Failed to query module temperature thresholds\n"); + return err; + } + + return 0; +} + +static ssize_t +mlxsw_hwmon_module_temp_emergency_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + int err, temp; + + err = mlxsw_hwmon_module_temp_emergency_get(dev, attr, &temp); + if (err) + return err; + + return sprintf(buf, "%u\n", temp); +} + +static ssize_t +mlxsw_hwmon_module_temp_label_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + + return sprintf(buf, "front panel %03u\n", + mlxsw_hwmon_attr->type_index + 1 - + mlxsw_hwmon_attr->mlxsw_hwmon_dev->sensor_count); +} + +static ssize_t +mlxsw_hwmon_gbox_temp_label_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr = + container_of(attr, struct mlxsw_hwmon_attr, dev_attr); + struct mlxsw_hwmon_dev *mlxsw_hwmon_dev = mlxsw_hwmon_attr->mlxsw_hwmon_dev; + int index = mlxsw_hwmon_attr->type_index - + mlxsw_hwmon_dev->module_sensor_max + 1; + + return sprintf(buf, "gearbox %03u\n", index); +} + +static ssize_t mlxsw_hwmon_temp_critical_alarm_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + int err, temp, emergency_temp, critic_temp; + + err = mlxsw_hwmon_module_temp_get(dev, attr, &temp); + if (err) + return err; + + if (temp <= 0) + return sprintf(buf, "%d\n", false); + + err = mlxsw_hwmon_module_temp_emergency_get(dev, attr, &emergency_temp); + if (err) + return err; + + if (temp >= emergency_temp) + return sprintf(buf, "%d\n", false); + + err = mlxsw_hwmon_module_temp_critical_get(dev, attr, &critic_temp); + if (err) + return err; + + return sprintf(buf, "%d\n", temp >= critic_temp); +} + +static ssize_t mlxsw_hwmon_temp_emergency_alarm_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + int err, temp, emergency_temp; + + err = mlxsw_hwmon_module_temp_get(dev, attr, &temp); + if (err) + return err; + + if (temp <= 0) + return sprintf(buf, "%d\n", false); + + err = mlxsw_hwmon_module_temp_emergency_get(dev, attr, &emergency_temp); + if (err) + return err; + + return sprintf(buf, "%d\n", temp >= emergency_temp); +} + +enum mlxsw_hwmon_attr_type { + MLXSW_HWMON_ATTR_TYPE_TEMP, + MLXSW_HWMON_ATTR_TYPE_TEMP_MAX, + MLXSW_HWMON_ATTR_TYPE_TEMP_RST, + MLXSW_HWMON_ATTR_TYPE_FAN_RPM, + MLXSW_HWMON_ATTR_TYPE_FAN_FAULT, + MLXSW_HWMON_ATTR_TYPE_PWM, + MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE, + MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE_FAULT, + MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE_CRIT, + MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE_EMERG, + MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE_LABEL, + MLXSW_HWMON_ATTR_TYPE_TEMP_GBOX_LABEL, + MLXSW_HWMON_ATTR_TYPE_TEMP_CRIT_ALARM, + MLXSW_HWMON_ATTR_TYPE_TEMP_EMERGENCY_ALARM, +}; + +static void mlxsw_hwmon_attr_add(struct mlxsw_hwmon_dev *mlxsw_hwmon_dev, + enum mlxsw_hwmon_attr_type attr_type, + unsigned int type_index, unsigned int num) +{ + struct mlxsw_hwmon_attr *mlxsw_hwmon_attr; + unsigned int attr_index; + + attr_index = mlxsw_hwmon_dev->attrs_count; + mlxsw_hwmon_attr = &mlxsw_hwmon_dev->hwmon_attrs[attr_index]; + + switch (attr_type) { + case MLXSW_HWMON_ATTR_TYPE_TEMP: + mlxsw_hwmon_attr->dev_attr.show = mlxsw_hwmon_temp_show; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0444; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "temp%u_input", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_TEMP_MAX: + mlxsw_hwmon_attr->dev_attr.show = mlxsw_hwmon_temp_max_show; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0444; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "temp%u_highest", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_TEMP_RST: + mlxsw_hwmon_attr->dev_attr.store = mlxsw_hwmon_temp_rst_store; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0200; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "temp%u_reset_history", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_FAN_RPM: + mlxsw_hwmon_attr->dev_attr.show = mlxsw_hwmon_fan_rpm_show; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0444; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "fan%u_input", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_FAN_FAULT: + mlxsw_hwmon_attr->dev_attr.show = mlxsw_hwmon_fan_fault_show; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0444; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "fan%u_fault", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_PWM: + mlxsw_hwmon_attr->dev_attr.show = mlxsw_hwmon_pwm_show; + mlxsw_hwmon_attr->dev_attr.store = mlxsw_hwmon_pwm_store; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0644; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "pwm%u", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE: + mlxsw_hwmon_attr->dev_attr.show = mlxsw_hwmon_module_temp_show; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0444; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "temp%u_input", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE_FAULT: + mlxsw_hwmon_attr->dev_attr.show = + mlxsw_hwmon_module_temp_fault_show; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0444; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "temp%u_fault", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE_CRIT: + mlxsw_hwmon_attr->dev_attr.show = + mlxsw_hwmon_module_temp_critical_show; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0444; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "temp%u_crit", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE_EMERG: + mlxsw_hwmon_attr->dev_attr.show = + mlxsw_hwmon_module_temp_emergency_show; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0444; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "temp%u_emergency", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE_LABEL: + mlxsw_hwmon_attr->dev_attr.show = + mlxsw_hwmon_module_temp_label_show; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0444; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "temp%u_label", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_TEMP_GBOX_LABEL: + mlxsw_hwmon_attr->dev_attr.show = + mlxsw_hwmon_gbox_temp_label_show; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0444; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "temp%u_label", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_TEMP_CRIT_ALARM: + mlxsw_hwmon_attr->dev_attr.show = + mlxsw_hwmon_temp_critical_alarm_show; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0444; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "temp%u_crit_alarm", num + 1); + break; + case MLXSW_HWMON_ATTR_TYPE_TEMP_EMERGENCY_ALARM: + mlxsw_hwmon_attr->dev_attr.show = + mlxsw_hwmon_temp_emergency_alarm_show; + mlxsw_hwmon_attr->dev_attr.attr.mode = 0444; + snprintf(mlxsw_hwmon_attr->name, sizeof(mlxsw_hwmon_attr->name), + "temp%u_emergency_alarm", num + 1); + break; + default: + WARN_ON(1); + } + + mlxsw_hwmon_attr->type_index = type_index; + mlxsw_hwmon_attr->mlxsw_hwmon_dev = mlxsw_hwmon_dev; + mlxsw_hwmon_attr->dev_attr.attr.name = mlxsw_hwmon_attr->name; + sysfs_attr_init(&mlxsw_hwmon_attr->dev_attr.attr); + + mlxsw_hwmon_dev->attrs[attr_index] = &mlxsw_hwmon_attr->dev_attr.attr; + mlxsw_hwmon_dev->attrs_count++; +} + +static int mlxsw_hwmon_temp_init(struct mlxsw_hwmon_dev *mlxsw_hwmon_dev) +{ + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + char mtcap_pl[MLXSW_REG_MTCAP_LEN] = {0}; + int i; + int err; + + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(mtcap), mtcap_pl); + if (err) { + dev_err(mlxsw_hwmon->bus_info->dev, "Failed to get number of temp sensors\n"); + return err; + } + mlxsw_hwmon_dev->sensor_count = mlxsw_reg_mtcap_sensor_count_get(mtcap_pl); + for (i = 0; i < mlxsw_hwmon_dev->sensor_count; i++) { + char mtmp_pl[MLXSW_REG_MTMP_LEN] = {0}; + + mlxsw_reg_mtmp_slot_index_set(mtmp_pl, + mlxsw_hwmon_dev->slot_index); + mlxsw_reg_mtmp_sensor_index_set(mtmp_pl, i); + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(mtmp), + mtmp_pl); + if (err) + return err; + mlxsw_reg_mtmp_mte_set(mtmp_pl, true); + mlxsw_reg_mtmp_mtr_set(mtmp_pl, true); + err = mlxsw_reg_write(mlxsw_hwmon->core, + MLXSW_REG(mtmp), mtmp_pl); + if (err) { + dev_err(mlxsw_hwmon->bus_info->dev, "Failed to setup temp sensor number %d\n", + i); + return err; + } + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP, i, i); + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP_MAX, i, i); + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP_RST, i, i); + } + return 0; +} + +static int mlxsw_hwmon_fans_init(struct mlxsw_hwmon_dev *mlxsw_hwmon_dev) +{ + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + char mfcr_pl[MLXSW_REG_MFCR_LEN] = {0}; + enum mlxsw_reg_mfcr_pwm_frequency freq; + unsigned int type_index; + unsigned int num; + u16 tacho_active; + u8 pwm_active; + int err; + + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(mfcr), mfcr_pl); + if (err) { + dev_err(mlxsw_hwmon->bus_info->dev, "Failed to get to probe PWMs and Tachometers\n"); + return err; + } + mlxsw_reg_mfcr_unpack(mfcr_pl, &freq, &tacho_active, &pwm_active); + num = 0; + for (type_index = 0; type_index < MLXSW_MFCR_TACHOS_MAX; type_index++) { + if (tacho_active & BIT(type_index)) { + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_FAN_RPM, + type_index, num); + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_FAN_FAULT, + type_index, num++); + } + } + num = 0; + for (type_index = 0; type_index < MLXSW_MFCR_PWMS_MAX; type_index++) { + if (pwm_active & BIT(type_index)) + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_PWM, + type_index, num++); + } + return 0; +} + +static int mlxsw_hwmon_module_init(struct mlxsw_hwmon_dev *mlxsw_hwmon_dev) +{ + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + char mgpir_pl[MLXSW_REG_MGPIR_LEN]; + u8 module_sensor_max; + int i, err; + + mlxsw_reg_mgpir_pack(mgpir_pl, mlxsw_hwmon_dev->slot_index); + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(mgpir), mgpir_pl); + if (err) + return err; + + mlxsw_reg_mgpir_unpack(mgpir_pl, NULL, NULL, NULL, + &module_sensor_max, NULL); + + /* Add extra attributes for module temperature. Sensor index is + * assigned to sensor_count value, while all indexed before + * sensor_count are already utilized by the sensors connected through + * mtmp register by mlxsw_hwmon_temp_init(). + */ + mlxsw_hwmon_dev->module_sensor_max = mlxsw_hwmon_dev->sensor_count + + module_sensor_max; + for (i = mlxsw_hwmon_dev->sensor_count; + i < mlxsw_hwmon_dev->module_sensor_max; i++) { + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE, i, i); + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE_FAULT, + i, i); + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE_CRIT, i, + i); + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE_EMERG, + i, i); + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP_MODULE_LABEL, + i, i); + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP_CRIT_ALARM, + i, i); + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP_EMERGENCY_ALARM, + i, i); + } + + return 0; +} + +static int mlxsw_hwmon_gearbox_init(struct mlxsw_hwmon_dev *mlxsw_hwmon_dev) +{ + struct mlxsw_hwmon *mlxsw_hwmon = mlxsw_hwmon_dev->hwmon; + enum mlxsw_reg_mgpir_device_type device_type; + int index, max_index, sensor_index; + char mgpir_pl[MLXSW_REG_MGPIR_LEN]; + char mtmp_pl[MLXSW_REG_MTMP_LEN]; + u8 gbox_num; + int err; + + mlxsw_reg_mgpir_pack(mgpir_pl, mlxsw_hwmon_dev->slot_index); + err = mlxsw_reg_query(mlxsw_hwmon->core, MLXSW_REG(mgpir), mgpir_pl); + if (err) + return err; + + mlxsw_reg_mgpir_unpack(mgpir_pl, &gbox_num, &device_type, NULL, NULL, + NULL); + if (device_type != MLXSW_REG_MGPIR_DEVICE_TYPE_GEARBOX_DIE || + !gbox_num) + return 0; + + index = mlxsw_hwmon_dev->module_sensor_max; + max_index = mlxsw_hwmon_dev->module_sensor_max + gbox_num; + while (index < max_index) { + sensor_index = index % mlxsw_hwmon_dev->module_sensor_max + + MLXSW_REG_MTMP_GBOX_INDEX_MIN; + mlxsw_reg_mtmp_pack(mtmp_pl, mlxsw_hwmon_dev->slot_index, + sensor_index, true, true); + err = mlxsw_reg_write(mlxsw_hwmon->core, + MLXSW_REG(mtmp), mtmp_pl); + if (err) { + dev_err(mlxsw_hwmon->bus_info->dev, "Failed to setup temp sensor number %d\n", + sensor_index); + return err; + } + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP, index, index); + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP_MAX, index, + index); + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP_RST, index, + index); + mlxsw_hwmon_attr_add(mlxsw_hwmon_dev, + MLXSW_HWMON_ATTR_TYPE_TEMP_GBOX_LABEL, + index, index); + index++; + } + + return 0; +} + +static void +mlxsw_hwmon_got_active(struct mlxsw_core *mlxsw_core, u8 slot_index, + void *priv) +{ + struct mlxsw_hwmon *hwmon = priv; + struct mlxsw_hwmon_dev *linecard; + struct device *dev; + int err; + + dev = hwmon->bus_info->dev; + linecard = &hwmon->line_cards[slot_index]; + if (linecard->active) + return; + /* For the main board, module sensor indexes start from 1, sensor index + * 0 is used for the ASIC. Use the same numbering for line cards. + */ + linecard->sensor_count = 1; + linecard->slot_index = slot_index; + linecard->hwmon = hwmon; + err = mlxsw_hwmon_module_init(linecard); + if (err) { + dev_err(dev, "Failed to configure hwmon objects for line card modules in slot %d\n", + slot_index); + return; + } + + err = mlxsw_hwmon_gearbox_init(linecard); + if (err) { + dev_err(dev, "Failed to configure hwmon objects for line card gearboxes in slot %d\n", + slot_index); + return; + } + + linecard->groups[0] = &linecard->group; + linecard->group.attrs = linecard->attrs; + sprintf(linecard->name, "%s#%02u", "linecard", slot_index); + linecard->hwmon_dev = + hwmon_device_register_with_groups(dev, linecard->name, + linecard, linecard->groups); + if (IS_ERR(linecard->hwmon_dev)) { + dev_err(dev, "Failed to register hwmon objects for line card in slot %d\n", + slot_index); + return; + } + + linecard->active = true; +} + +static void +mlxsw_hwmon_got_inactive(struct mlxsw_core *mlxsw_core, u8 slot_index, + void *priv) +{ + struct mlxsw_hwmon *hwmon = priv; + struct mlxsw_hwmon_dev *linecard; + + linecard = &hwmon->line_cards[slot_index]; + if (!linecard->active) + return; + linecard->active = false; + hwmon_device_unregister(linecard->hwmon_dev); + /* Reset attributes counter */ + linecard->attrs_count = 0; +} + +static struct mlxsw_linecards_event_ops mlxsw_hwmon_event_ops = { + .got_active = mlxsw_hwmon_got_active, + .got_inactive = mlxsw_hwmon_got_inactive, +}; + +int mlxsw_hwmon_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct mlxsw_hwmon **p_hwmon) +{ + char mgpir_pl[MLXSW_REG_MGPIR_LEN]; + struct mlxsw_hwmon *mlxsw_hwmon; + struct device *hwmon_dev; + u8 num_of_slots; + int err; + + mlxsw_reg_mgpir_pack(mgpir_pl, 0); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mgpir), mgpir_pl); + if (err) + return err; + + mlxsw_reg_mgpir_unpack(mgpir_pl, NULL, NULL, NULL, NULL, + &num_of_slots); + + mlxsw_hwmon = kzalloc(struct_size(mlxsw_hwmon, line_cards, + num_of_slots + 1), GFP_KERNEL); + if (!mlxsw_hwmon) + return -ENOMEM; + + mlxsw_hwmon->core = mlxsw_core; + mlxsw_hwmon->bus_info = mlxsw_bus_info; + mlxsw_hwmon->line_cards[0].hwmon = mlxsw_hwmon; + mlxsw_hwmon->line_cards[0].slot_index = 0; + + err = mlxsw_hwmon_temp_init(&mlxsw_hwmon->line_cards[0]); + if (err) + goto err_temp_init; + + err = mlxsw_hwmon_fans_init(&mlxsw_hwmon->line_cards[0]); + if (err) + goto err_fans_init; + + err = mlxsw_hwmon_module_init(&mlxsw_hwmon->line_cards[0]); + if (err) + goto err_temp_module_init; + + err = mlxsw_hwmon_gearbox_init(&mlxsw_hwmon->line_cards[0]); + if (err) + goto err_temp_gearbox_init; + + mlxsw_hwmon->line_cards[0].groups[0] = &mlxsw_hwmon->line_cards[0].group; + mlxsw_hwmon->line_cards[0].group.attrs = mlxsw_hwmon->line_cards[0].attrs; + + hwmon_dev = hwmon_device_register_with_groups(mlxsw_bus_info->dev, + "mlxsw", + &mlxsw_hwmon->line_cards[0], + mlxsw_hwmon->line_cards[0].groups); + if (IS_ERR(hwmon_dev)) { + err = PTR_ERR(hwmon_dev); + goto err_hwmon_register; + } + + err = mlxsw_linecards_event_ops_register(mlxsw_hwmon->core, + &mlxsw_hwmon_event_ops, + mlxsw_hwmon); + if (err) + goto err_linecards_event_ops_register; + + mlxsw_hwmon->line_cards[0].hwmon_dev = hwmon_dev; + mlxsw_hwmon->line_cards[0].active = true; + *p_hwmon = mlxsw_hwmon; + return 0; + +err_linecards_event_ops_register: + hwmon_device_unregister(mlxsw_hwmon->line_cards[0].hwmon_dev); +err_hwmon_register: +err_temp_gearbox_init: +err_temp_module_init: +err_fans_init: +err_temp_init: + kfree(mlxsw_hwmon); + return err; +} + +void mlxsw_hwmon_fini(struct mlxsw_hwmon *mlxsw_hwmon) +{ + mlxsw_hwmon->line_cards[0].active = false; + mlxsw_linecards_event_ops_unregister(mlxsw_hwmon->core, + &mlxsw_hwmon_event_ops, mlxsw_hwmon); + hwmon_device_unregister(mlxsw_hwmon->line_cards[0].hwmon_dev); + kfree(mlxsw_hwmon); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_linecard_dev.c b/drivers/net/ethernet/mellanox/mlxsw/core_linecard_dev.c new file mode 100644 index 000000000..af37e650a --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/core_linecard_dev.c @@ -0,0 +1,183 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2022 NVIDIA Corporation and Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/err.h> +#include <linux/types.h> +#include <linux/auxiliary_bus.h> +#include <linux/idr.h> +#include <linux/gfp.h> +#include <linux/slab.h> +#include <net/devlink.h> +#include "core.h" + +#define MLXSW_LINECARD_DEV_ID_NAME "lc" + +struct mlxsw_linecard_dev { + struct mlxsw_linecard *linecard; +}; + +struct mlxsw_linecard_bdev { + struct auxiliary_device adev; + struct mlxsw_linecard *linecard; + struct mlxsw_linecard_dev *linecard_dev; +}; + +static DEFINE_IDA(mlxsw_linecard_bdev_ida); + +static int mlxsw_linecard_bdev_id_alloc(void) +{ + return ida_alloc(&mlxsw_linecard_bdev_ida, GFP_KERNEL); +} + +static void mlxsw_linecard_bdev_id_free(int id) +{ + ida_free(&mlxsw_linecard_bdev_ida, id); +} + +static void mlxsw_linecard_bdev_release(struct device *device) +{ + struct auxiliary_device *adev = + container_of(device, struct auxiliary_device, dev); + struct mlxsw_linecard_bdev *linecard_bdev = + container_of(adev, struct mlxsw_linecard_bdev, adev); + + mlxsw_linecard_bdev_id_free(adev->id); + kfree(linecard_bdev); +} + +int mlxsw_linecard_bdev_add(struct mlxsw_linecard *linecard) +{ + struct mlxsw_linecard_bdev *linecard_bdev; + int err; + int id; + + id = mlxsw_linecard_bdev_id_alloc(); + if (id < 0) + return id; + + linecard_bdev = kzalloc(sizeof(*linecard_bdev), GFP_KERNEL); + if (!linecard_bdev) { + mlxsw_linecard_bdev_id_free(id); + return -ENOMEM; + } + linecard_bdev->adev.id = id; + linecard_bdev->adev.name = MLXSW_LINECARD_DEV_ID_NAME; + linecard_bdev->adev.dev.release = mlxsw_linecard_bdev_release; + linecard_bdev->adev.dev.parent = linecard->linecards->bus_info->dev; + linecard_bdev->linecard = linecard; + + err = auxiliary_device_init(&linecard_bdev->adev); + if (err) { + mlxsw_linecard_bdev_id_free(id); + kfree(linecard_bdev); + return err; + } + + err = auxiliary_device_add(&linecard_bdev->adev); + if (err) { + auxiliary_device_uninit(&linecard_bdev->adev); + return err; + } + + linecard->bdev = linecard_bdev; + return 0; +} + +void mlxsw_linecard_bdev_del(struct mlxsw_linecard *linecard) +{ + struct mlxsw_linecard_bdev *linecard_bdev = linecard->bdev; + + if (!linecard_bdev) + /* Unprovisioned line cards do not have an auxiliary device. */ + return; + auxiliary_device_delete(&linecard_bdev->adev); + auxiliary_device_uninit(&linecard_bdev->adev); + linecard->bdev = NULL; +} + +static int mlxsw_linecard_dev_devlink_info_get(struct devlink *devlink, + struct devlink_info_req *req, + struct netlink_ext_ack *extack) +{ + struct mlxsw_linecard_dev *linecard_dev = devlink_priv(devlink); + struct mlxsw_linecard *linecard = linecard_dev->linecard; + + return mlxsw_linecard_devlink_info_get(linecard, req, extack); +} + +static int +mlxsw_linecard_dev_devlink_flash_update(struct devlink *devlink, + struct devlink_flash_update_params *params, + struct netlink_ext_ack *extack) +{ + struct mlxsw_linecard_dev *linecard_dev = devlink_priv(devlink); + struct mlxsw_linecard *linecard = linecard_dev->linecard; + + return mlxsw_linecard_flash_update(devlink, linecard, + params->fw, extack); +} + +static const struct devlink_ops mlxsw_linecard_dev_devlink_ops = { + .info_get = mlxsw_linecard_dev_devlink_info_get, + .flash_update = mlxsw_linecard_dev_devlink_flash_update, +}; + +static int mlxsw_linecard_bdev_probe(struct auxiliary_device *adev, + const struct auxiliary_device_id *id) +{ + struct mlxsw_linecard_bdev *linecard_bdev = + container_of(adev, struct mlxsw_linecard_bdev, adev); + struct mlxsw_linecard *linecard = linecard_bdev->linecard; + struct mlxsw_linecard_dev *linecard_dev; + struct devlink *devlink; + + devlink = devlink_alloc(&mlxsw_linecard_dev_devlink_ops, + sizeof(*linecard_dev), &adev->dev); + if (!devlink) + return -ENOMEM; + linecard_dev = devlink_priv(devlink); + linecard_dev->linecard = linecard_bdev->linecard; + linecard_bdev->linecard_dev = linecard_dev; + + devlink_register(devlink); + devlink_linecard_nested_dl_set(linecard->devlink_linecard, devlink); + return 0; +} + +static void mlxsw_linecard_bdev_remove(struct auxiliary_device *adev) +{ + struct mlxsw_linecard_bdev *linecard_bdev = + container_of(adev, struct mlxsw_linecard_bdev, adev); + struct devlink *devlink = priv_to_devlink(linecard_bdev->linecard_dev); + struct mlxsw_linecard *linecard = linecard_bdev->linecard; + + devlink_linecard_nested_dl_set(linecard->devlink_linecard, NULL); + devlink_unregister(devlink); + devlink_free(devlink); +} + +static const struct auxiliary_device_id mlxsw_linecard_bdev_id_table[] = { + { .name = KBUILD_MODNAME "." MLXSW_LINECARD_DEV_ID_NAME }, + {}, +}; + +MODULE_DEVICE_TABLE(auxiliary, mlxsw_linecard_bdev_id_table); + +static struct auxiliary_driver mlxsw_linecard_driver = { + .name = MLXSW_LINECARD_DEV_ID_NAME, + .probe = mlxsw_linecard_bdev_probe, + .remove = mlxsw_linecard_bdev_remove, + .id_table = mlxsw_linecard_bdev_id_table, +}; + +int mlxsw_linecard_driver_register(void) +{ + return auxiliary_driver_register(&mlxsw_linecard_driver); +} + +void mlxsw_linecard_driver_unregister(void) +{ + auxiliary_driver_unregister(&mlxsw_linecard_driver); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_linecards.c b/drivers/net/ethernet/mellanox/mlxsw/core_linecards.c new file mode 100644 index 000000000..83d2dc91b --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/core_linecards.c @@ -0,0 +1,1601 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2022 NVIDIA Corporation and Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/err.h> +#include <linux/types.h> +#include <linux/string.h> +#include <linux/workqueue.h> +#include <linux/gfp.h> +#include <linux/slab.h> +#include <linux/list.h> +#include <linux/vmalloc.h> + +#include "core.h" +#include "../mlxfw/mlxfw.h" + +struct mlxsw_linecard_ini_file { + __le16 size; + union { + u8 data[0]; + struct { + __be16 hw_revision; + __be16 ini_version; + u8 __dontcare[3]; + u8 type; + u8 name[20]; + } format; + }; +}; + +struct mlxsw_linecard_types_info { + struct mlxsw_linecard_ini_file **ini_files; + unsigned int count; + size_t data_size; + char *data; +}; + +#define MLXSW_LINECARD_STATUS_EVENT_TO (10 * MSEC_PER_SEC) + +static void +mlxsw_linecard_status_event_to_schedule(struct mlxsw_linecard *linecard, + enum mlxsw_linecard_status_event_type status_event_type) +{ + cancel_delayed_work_sync(&linecard->status_event_to_dw); + linecard->status_event_type_to = status_event_type; + mlxsw_core_schedule_dw(&linecard->status_event_to_dw, + msecs_to_jiffies(MLXSW_LINECARD_STATUS_EVENT_TO)); +} + +static void +mlxsw_linecard_status_event_done(struct mlxsw_linecard *linecard, + enum mlxsw_linecard_status_event_type status_event_type) +{ + if (linecard->status_event_type_to == status_event_type) + cancel_delayed_work_sync(&linecard->status_event_to_dw); +} + +static const char * +mlxsw_linecard_types_lookup(struct mlxsw_linecards *linecards, u8 card_type) +{ + struct mlxsw_linecard_types_info *types_info; + struct mlxsw_linecard_ini_file *ini_file; + int i; + + types_info = linecards->types_info; + if (!types_info) + return NULL; + for (i = 0; i < types_info->count; i++) { + ini_file = linecards->types_info->ini_files[i]; + if (ini_file->format.type == card_type) + return ini_file->format.name; + } + return NULL; +} + +static const char *mlxsw_linecard_type_name(struct mlxsw_linecard *linecard) +{ + struct mlxsw_core *mlxsw_core = linecard->linecards->mlxsw_core; + char mddq_pl[MLXSW_REG_MDDQ_LEN]; + int err; + + mlxsw_reg_mddq_slot_name_pack(mddq_pl, linecard->slot_index); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mddq), mddq_pl); + if (err) + return ERR_PTR(err); + mlxsw_reg_mddq_slot_name_unpack(mddq_pl, linecard->name); + return linecard->name; +} + +struct mlxsw_linecard_device_fw_info { + struct mlxfw_dev mlxfw_dev; + struct mlxsw_core *mlxsw_core; + struct mlxsw_linecard *linecard; +}; + +static int mlxsw_linecard_device_fw_component_query(struct mlxfw_dev *mlxfw_dev, + u16 component_index, + u32 *p_max_size, + u8 *p_align_bits, + u16 *p_max_write_size) +{ + struct mlxsw_linecard_device_fw_info *info = + container_of(mlxfw_dev, struct mlxsw_linecard_device_fw_info, + mlxfw_dev); + struct mlxsw_linecard *linecard = info->linecard; + struct mlxsw_core *mlxsw_core = info->mlxsw_core; + char mddt_pl[MLXSW_REG_MDDT_LEN]; + char *mcqi_pl; + int err; + + mlxsw_reg_mddt_pack(mddt_pl, linecard->slot_index, + linecard->device.index, + MLXSW_REG_MDDT_METHOD_QUERY, + MLXSW_REG(mcqi), &mcqi_pl); + + mlxsw_reg_mcqi_pack(mcqi_pl, component_index); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mddt), mddt_pl); + if (err) + return err; + mlxsw_reg_mcqi_unpack(mcqi_pl, p_max_size, p_align_bits, + p_max_write_size); + + *p_align_bits = max_t(u8, *p_align_bits, 2); + *p_max_write_size = min_t(u16, *p_max_write_size, + MLXSW_REG_MCDA_MAX_DATA_LEN); + return 0; +} + +static int mlxsw_linecard_device_fw_fsm_lock(struct mlxfw_dev *mlxfw_dev, + u32 *fwhandle) +{ + struct mlxsw_linecard_device_fw_info *info = + container_of(mlxfw_dev, struct mlxsw_linecard_device_fw_info, + mlxfw_dev); + struct mlxsw_linecard *linecard = info->linecard; + struct mlxsw_core *mlxsw_core = info->mlxsw_core; + char mddt_pl[MLXSW_REG_MDDT_LEN]; + u8 control_state; + char *mcc_pl; + int err; + + mlxsw_reg_mddt_pack(mddt_pl, linecard->slot_index, + linecard->device.index, + MLXSW_REG_MDDT_METHOD_QUERY, + MLXSW_REG(mcc), &mcc_pl); + mlxsw_reg_mcc_pack(mcc_pl, 0, 0, 0, 0); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mddt), mddt_pl); + if (err) + return err; + + mlxsw_reg_mcc_unpack(mcc_pl, fwhandle, NULL, &control_state); + if (control_state != MLXFW_FSM_STATE_IDLE) + return -EBUSY; + + mlxsw_reg_mddt_pack(mddt_pl, linecard->slot_index, + linecard->device.index, + MLXSW_REG_MDDT_METHOD_WRITE, + MLXSW_REG(mcc), &mcc_pl); + mlxsw_reg_mcc_pack(mcc_pl, MLXSW_REG_MCC_INSTRUCTION_LOCK_UPDATE_HANDLE, + 0, *fwhandle, 0); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mddt), mddt_pl); +} + +static int +mlxsw_linecard_device_fw_fsm_component_update(struct mlxfw_dev *mlxfw_dev, + u32 fwhandle, + u16 component_index, + u32 component_size) +{ + struct mlxsw_linecard_device_fw_info *info = + container_of(mlxfw_dev, struct mlxsw_linecard_device_fw_info, + mlxfw_dev); + struct mlxsw_linecard *linecard = info->linecard; + struct mlxsw_core *mlxsw_core = info->mlxsw_core; + char mddt_pl[MLXSW_REG_MDDT_LEN]; + char *mcc_pl; + + mlxsw_reg_mddt_pack(mddt_pl, linecard->slot_index, + linecard->device.index, + MLXSW_REG_MDDT_METHOD_WRITE, + MLXSW_REG(mcc), &mcc_pl); + mlxsw_reg_mcc_pack(mcc_pl, MLXSW_REG_MCC_INSTRUCTION_UPDATE_COMPONENT, + component_index, fwhandle, component_size); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mddt), mddt_pl); +} + +static int +mlxsw_linecard_device_fw_fsm_block_download(struct mlxfw_dev *mlxfw_dev, + u32 fwhandle, u8 *data, + u16 size, u32 offset) +{ + struct mlxsw_linecard_device_fw_info *info = + container_of(mlxfw_dev, struct mlxsw_linecard_device_fw_info, + mlxfw_dev); + struct mlxsw_linecard *linecard = info->linecard; + struct mlxsw_core *mlxsw_core = info->mlxsw_core; + char mddt_pl[MLXSW_REG_MDDT_LEN]; + char *mcda_pl; + + mlxsw_reg_mddt_pack(mddt_pl, linecard->slot_index, + linecard->device.index, + MLXSW_REG_MDDT_METHOD_WRITE, + MLXSW_REG(mcda), &mcda_pl); + mlxsw_reg_mcda_pack(mcda_pl, fwhandle, offset, size, data); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mddt), mddt_pl); +} + +static int +mlxsw_linecard_device_fw_fsm_component_verify(struct mlxfw_dev *mlxfw_dev, + u32 fwhandle, u16 component_index) +{ + struct mlxsw_linecard_device_fw_info *info = + container_of(mlxfw_dev, struct mlxsw_linecard_device_fw_info, + mlxfw_dev); + struct mlxsw_linecard *linecard = info->linecard; + struct mlxsw_core *mlxsw_core = info->mlxsw_core; + char mddt_pl[MLXSW_REG_MDDT_LEN]; + char *mcc_pl; + + mlxsw_reg_mddt_pack(mddt_pl, linecard->slot_index, + linecard->device.index, + MLXSW_REG_MDDT_METHOD_WRITE, + MLXSW_REG(mcc), &mcc_pl); + mlxsw_reg_mcc_pack(mcc_pl, MLXSW_REG_MCC_INSTRUCTION_VERIFY_COMPONENT, + component_index, fwhandle, 0); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mddt), mddt_pl); +} + +static int mlxsw_linecard_device_fw_fsm_activate(struct mlxfw_dev *mlxfw_dev, + u32 fwhandle) +{ + struct mlxsw_linecard_device_fw_info *info = + container_of(mlxfw_dev, struct mlxsw_linecard_device_fw_info, + mlxfw_dev); + struct mlxsw_linecard *linecard = info->linecard; + struct mlxsw_core *mlxsw_core = info->mlxsw_core; + char mddt_pl[MLXSW_REG_MDDT_LEN]; + char *mcc_pl; + + mlxsw_reg_mddt_pack(mddt_pl, linecard->slot_index, + linecard->device.index, + MLXSW_REG_MDDT_METHOD_WRITE, + MLXSW_REG(mcc), &mcc_pl); + mlxsw_reg_mcc_pack(mcc_pl, MLXSW_REG_MCC_INSTRUCTION_ACTIVATE, + 0, fwhandle, 0); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mddt), mddt_pl); +} + +static int +mlxsw_linecard_device_fw_fsm_query_state(struct mlxfw_dev *mlxfw_dev, + u32 fwhandle, + enum mlxfw_fsm_state *fsm_state, + enum mlxfw_fsm_state_err *fsm_state_err) +{ + struct mlxsw_linecard_device_fw_info *info = + container_of(mlxfw_dev, struct mlxsw_linecard_device_fw_info, + mlxfw_dev); + struct mlxsw_linecard *linecard = info->linecard; + struct mlxsw_core *mlxsw_core = info->mlxsw_core; + char mddt_pl[MLXSW_REG_MDDT_LEN]; + u8 control_state; + u8 error_code; + char *mcc_pl; + int err; + + mlxsw_reg_mddt_pack(mddt_pl, linecard->slot_index, + linecard->device.index, + MLXSW_REG_MDDT_METHOD_QUERY, + MLXSW_REG(mcc), &mcc_pl); + mlxsw_reg_mcc_pack(mcc_pl, 0, 0, fwhandle, 0); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mddt), mddt_pl); + if (err) + return err; + + mlxsw_reg_mcc_unpack(mcc_pl, NULL, &error_code, &control_state); + *fsm_state = control_state; + *fsm_state_err = min_t(enum mlxfw_fsm_state_err, error_code, + MLXFW_FSM_STATE_ERR_MAX); + return 0; +} + +static void mlxsw_linecard_device_fw_fsm_cancel(struct mlxfw_dev *mlxfw_dev, + u32 fwhandle) +{ + struct mlxsw_linecard_device_fw_info *info = + container_of(mlxfw_dev, struct mlxsw_linecard_device_fw_info, + mlxfw_dev); + struct mlxsw_linecard *linecard = info->linecard; + struct mlxsw_core *mlxsw_core = info->mlxsw_core; + char mddt_pl[MLXSW_REG_MDDT_LEN]; + char *mcc_pl; + + mlxsw_reg_mddt_pack(mddt_pl, linecard->slot_index, + linecard->device.index, + MLXSW_REG_MDDT_METHOD_WRITE, + MLXSW_REG(mcc), &mcc_pl); + mlxsw_reg_mcc_pack(mcc_pl, MLXSW_REG_MCC_INSTRUCTION_CANCEL, + 0, fwhandle, 0); + mlxsw_reg_write(mlxsw_core, MLXSW_REG(mddt), mddt_pl); +} + +static void mlxsw_linecard_device_fw_fsm_release(struct mlxfw_dev *mlxfw_dev, + u32 fwhandle) +{ + struct mlxsw_linecard_device_fw_info *info = + container_of(mlxfw_dev, struct mlxsw_linecard_device_fw_info, + mlxfw_dev); + struct mlxsw_linecard *linecard = info->linecard; + struct mlxsw_core *mlxsw_core = info->mlxsw_core; + char mddt_pl[MLXSW_REG_MDDT_LEN]; + char *mcc_pl; + + mlxsw_reg_mddt_pack(mddt_pl, linecard->slot_index, + linecard->device.index, + MLXSW_REG_MDDT_METHOD_WRITE, + MLXSW_REG(mcc), &mcc_pl); + mlxsw_reg_mcc_pack(mcc_pl, + MLXSW_REG_MCC_INSTRUCTION_RELEASE_UPDATE_HANDLE, + 0, fwhandle, 0); + mlxsw_reg_write(mlxsw_core, MLXSW_REG(mddt), mddt_pl); +} + +static const struct mlxfw_dev_ops mlxsw_linecard_device_dev_ops = { + .component_query = mlxsw_linecard_device_fw_component_query, + .fsm_lock = mlxsw_linecard_device_fw_fsm_lock, + .fsm_component_update = mlxsw_linecard_device_fw_fsm_component_update, + .fsm_block_download = mlxsw_linecard_device_fw_fsm_block_download, + .fsm_component_verify = mlxsw_linecard_device_fw_fsm_component_verify, + .fsm_activate = mlxsw_linecard_device_fw_fsm_activate, + .fsm_query_state = mlxsw_linecard_device_fw_fsm_query_state, + .fsm_cancel = mlxsw_linecard_device_fw_fsm_cancel, + .fsm_release = mlxsw_linecard_device_fw_fsm_release, +}; + +int mlxsw_linecard_flash_update(struct devlink *linecard_devlink, + struct mlxsw_linecard *linecard, + const struct firmware *firmware, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = linecard->linecards->mlxsw_core; + struct mlxsw_linecard_device_fw_info info = { + .mlxfw_dev = { + .ops = &mlxsw_linecard_device_dev_ops, + .psid = linecard->device.info.psid, + .psid_size = strlen(linecard->device.info.psid), + .devlink = linecard_devlink, + }, + .mlxsw_core = mlxsw_core, + .linecard = linecard, + }; + int err; + + mutex_lock(&linecard->lock); + if (!linecard->active) { + NL_SET_ERR_MSG_MOD(extack, "Only active line cards can be flashed"); + err = -EINVAL; + goto unlock; + } + err = mlxsw_core_fw_flash(mlxsw_core, &info.mlxfw_dev, + firmware, extack); +unlock: + mutex_unlock(&linecard->lock); + return err; +} + +static int mlxsw_linecard_device_psid_get(struct mlxsw_linecard *linecard, + u8 device_index, char *psid) +{ + struct mlxsw_core *mlxsw_core = linecard->linecards->mlxsw_core; + char mddt_pl[MLXSW_REG_MDDT_LEN]; + char *mgir_pl; + int err; + + mlxsw_reg_mddt_pack(mddt_pl, linecard->slot_index, device_index, + MLXSW_REG_MDDT_METHOD_QUERY, + MLXSW_REG(mgir), &mgir_pl); + + mlxsw_reg_mgir_pack(mgir_pl); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mddt), mddt_pl); + if (err) + return err; + + mlxsw_reg_mgir_fw_info_psid_memcpy_from(mgir_pl, psid); + return 0; +} + +static int mlxsw_linecard_device_info_update(struct mlxsw_linecard *linecard) +{ + struct mlxsw_core *mlxsw_core = linecard->linecards->mlxsw_core; + bool flashable_found = false; + u8 msg_seq = 0; + + do { + struct mlxsw_linecard_device_info info; + char mddq_pl[MLXSW_REG_MDDQ_LEN]; + bool flash_owner; + bool data_valid; + u8 device_index; + int err; + + mlxsw_reg_mddq_device_info_pack(mddq_pl, linecard->slot_index, + msg_seq); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mddq), mddq_pl); + if (err) + return err; + mlxsw_reg_mddq_device_info_unpack(mddq_pl, &msg_seq, + &data_valid, &flash_owner, + &device_index, + &info.fw_major, + &info.fw_minor, + &info.fw_sub_minor); + if (!data_valid) + break; + if (!flash_owner) /* We care only about flashable ones. */ + continue; + if (flashable_found) { + dev_warn_once(linecard->linecards->bus_info->dev, "linecard %u: More flashable devices present, exposing only the first one\n", + linecard->slot_index); + return 0; + } + + err = mlxsw_linecard_device_psid_get(linecard, device_index, + info.psid); + if (err) + return err; + + linecard->device.info = info; + linecard->device.index = device_index; + flashable_found = true; + } while (msg_seq); + + return 0; +} + +static void mlxsw_linecard_provision_fail(struct mlxsw_linecard *linecard) +{ + linecard->provisioned = false; + linecard->ready = false; + linecard->active = false; + devlink_linecard_provision_fail(linecard->devlink_linecard); +} + +struct mlxsw_linecards_event_ops_item { + struct list_head list; + const struct mlxsw_linecards_event_ops *event_ops; + void *priv; +}; + +static void +mlxsw_linecard_event_op_call(struct mlxsw_linecard *linecard, + mlxsw_linecards_event_op_t *op, void *priv) +{ + struct mlxsw_core *mlxsw_core = linecard->linecards->mlxsw_core; + + if (!op) + return; + op(mlxsw_core, linecard->slot_index, priv); +} + +static void +mlxsw_linecard_active_ops_call(struct mlxsw_linecard *linecard) +{ + struct mlxsw_linecards *linecards = linecard->linecards; + struct mlxsw_linecards_event_ops_item *item; + + mutex_lock(&linecards->event_ops_list_lock); + list_for_each_entry(item, &linecards->event_ops_list, list) + mlxsw_linecard_event_op_call(linecard, + item->event_ops->got_active, + item->priv); + mutex_unlock(&linecards->event_ops_list_lock); +} + +static void +mlxsw_linecard_inactive_ops_call(struct mlxsw_linecard *linecard) +{ + struct mlxsw_linecards *linecards = linecard->linecards; + struct mlxsw_linecards_event_ops_item *item; + + mutex_lock(&linecards->event_ops_list_lock); + list_for_each_entry(item, &linecards->event_ops_list, list) + mlxsw_linecard_event_op_call(linecard, + item->event_ops->got_inactive, + item->priv); + mutex_unlock(&linecards->event_ops_list_lock); +} + +static void +mlxsw_linecards_event_ops_register_call(struct mlxsw_linecards *linecards, + const struct mlxsw_linecards_event_ops_item *item) +{ + struct mlxsw_linecard *linecard; + int i; + + for (i = 0; i < linecards->count; i++) { + linecard = mlxsw_linecard_get(linecards, i + 1); + mutex_lock(&linecard->lock); + if (linecard->active) + mlxsw_linecard_event_op_call(linecard, + item->event_ops->got_active, + item->priv); + mutex_unlock(&linecard->lock); + } +} + +static void +mlxsw_linecards_event_ops_unregister_call(struct mlxsw_linecards *linecards, + const struct mlxsw_linecards_event_ops_item *item) +{ + struct mlxsw_linecard *linecard; + int i; + + for (i = 0; i < linecards->count; i++) { + linecard = mlxsw_linecard_get(linecards, i + 1); + mutex_lock(&linecard->lock); + if (linecard->active) + mlxsw_linecard_event_op_call(linecard, + item->event_ops->got_inactive, + item->priv); + mutex_unlock(&linecard->lock); + } +} + +int mlxsw_linecards_event_ops_register(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecards_event_ops *ops, + void *priv) +{ + struct mlxsw_linecards *linecards = mlxsw_core_linecards(mlxsw_core); + struct mlxsw_linecards_event_ops_item *item; + + if (!linecards) + return 0; + item = kzalloc(sizeof(*item), GFP_KERNEL); + if (!item) + return -ENOMEM; + item->event_ops = ops; + item->priv = priv; + + mutex_lock(&linecards->event_ops_list_lock); + list_add_tail(&item->list, &linecards->event_ops_list); + mutex_unlock(&linecards->event_ops_list_lock); + mlxsw_linecards_event_ops_register_call(linecards, item); + return 0; +} +EXPORT_SYMBOL(mlxsw_linecards_event_ops_register); + +void mlxsw_linecards_event_ops_unregister(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecards_event_ops *ops, + void *priv) +{ + struct mlxsw_linecards *linecards = mlxsw_core_linecards(mlxsw_core); + struct mlxsw_linecards_event_ops_item *item, *tmp; + bool found = false; + + if (!linecards) + return; + mutex_lock(&linecards->event_ops_list_lock); + list_for_each_entry_safe(item, tmp, &linecards->event_ops_list, list) { + if (item->event_ops == ops && item->priv == priv) { + list_del(&item->list); + found = true; + break; + } + } + mutex_unlock(&linecards->event_ops_list_lock); + + if (!found) + return; + mlxsw_linecards_event_ops_unregister_call(linecards, item); + kfree(item); +} +EXPORT_SYMBOL(mlxsw_linecards_event_ops_unregister); + +int mlxsw_linecard_devlink_info_get(struct mlxsw_linecard *linecard, + struct devlink_info_req *req, + struct netlink_ext_ack *extack) +{ + char buf[32]; + int err; + + mutex_lock(&linecard->lock); + if (WARN_ON(!linecard->provisioned)) { + err = -EOPNOTSUPP; + goto unlock; + } + + sprintf(buf, "%d", linecard->hw_revision); + err = devlink_info_version_fixed_put(req, "hw.revision", buf); + if (err) + goto unlock; + + sprintf(buf, "%d", linecard->ini_version); + err = devlink_info_version_running_put(req, "ini.version", buf); + if (err) + goto unlock; + + if (linecard->active) { + struct mlxsw_linecard_device_info *info = &linecard->device.info; + + err = devlink_info_version_fixed_put(req, + DEVLINK_INFO_VERSION_GENERIC_FW_PSID, + info->psid); + + sprintf(buf, "%u.%u.%u", info->fw_major, info->fw_minor, + info->fw_sub_minor); + err = devlink_info_version_running_put(req, + DEVLINK_INFO_VERSION_GENERIC_FW, + buf); + if (err) + goto unlock; + } + +unlock: + mutex_unlock(&linecard->lock); + return err; +} + +static int +mlxsw_linecard_provision_set(struct mlxsw_linecard *linecard, u8 card_type, + u16 hw_revision, u16 ini_version) +{ + struct mlxsw_linecards *linecards = linecard->linecards; + const char *type; + int err; + + type = mlxsw_linecard_types_lookup(linecards, card_type); + mlxsw_linecard_status_event_done(linecard, + MLXSW_LINECARD_STATUS_EVENT_TYPE_PROVISION); + if (!type) { + /* It is possible for a line card to be provisioned before + * driver initialization. Due to a missing INI bundle file + * or an outdated one, the queried card's type might not + * be recognized by the driver. In this case, try to query + * the card's name from the device. + */ + type = mlxsw_linecard_type_name(linecard); + if (IS_ERR(type)) { + mlxsw_linecard_provision_fail(linecard); + return PTR_ERR(type); + } + } + linecard->provisioned = true; + linecard->hw_revision = hw_revision; + linecard->ini_version = ini_version; + + err = mlxsw_linecard_bdev_add(linecard); + if (err) { + linecard->provisioned = false; + mlxsw_linecard_provision_fail(linecard); + return err; + } + + devlink_linecard_provision_set(linecard->devlink_linecard, type); + return 0; +} + +static void mlxsw_linecard_provision_clear(struct mlxsw_linecard *linecard) +{ + mlxsw_linecard_status_event_done(linecard, + MLXSW_LINECARD_STATUS_EVENT_TYPE_UNPROVISION); + mlxsw_linecard_bdev_del(linecard); + linecard->provisioned = false; + devlink_linecard_provision_clear(linecard->devlink_linecard); +} + +static int mlxsw_linecard_ready_set(struct mlxsw_linecard *linecard) +{ + struct mlxsw_core *mlxsw_core = linecard->linecards->mlxsw_core; + char mddc_pl[MLXSW_REG_MDDC_LEN]; + int err; + + err = mlxsw_linecard_device_info_update(linecard); + if (err) + return err; + + mlxsw_reg_mddc_pack(mddc_pl, linecard->slot_index, false, true); + err = mlxsw_reg_write(mlxsw_core, MLXSW_REG(mddc), mddc_pl); + if (err) + return err; + linecard->ready = true; + return 0; +} + +static int mlxsw_linecard_ready_clear(struct mlxsw_linecard *linecard) +{ + struct mlxsw_core *mlxsw_core = linecard->linecards->mlxsw_core; + char mddc_pl[MLXSW_REG_MDDC_LEN]; + int err; + + mlxsw_reg_mddc_pack(mddc_pl, linecard->slot_index, false, false); + err = mlxsw_reg_write(mlxsw_core, MLXSW_REG(mddc), mddc_pl); + if (err) + return err; + linecard->ready = false; + return 0; +} + +static void mlxsw_linecard_active_set(struct mlxsw_linecard *linecard) +{ + mlxsw_linecard_active_ops_call(linecard); + linecard->active = true; + devlink_linecard_activate(linecard->devlink_linecard); +} + +static void mlxsw_linecard_active_clear(struct mlxsw_linecard *linecard) +{ + mlxsw_linecard_inactive_ops_call(linecard); + linecard->active = false; + devlink_linecard_deactivate(linecard->devlink_linecard); +} + +static int mlxsw_linecard_status_process(struct mlxsw_linecards *linecards, + struct mlxsw_linecard *linecard, + const char *mddq_pl) +{ + enum mlxsw_reg_mddq_slot_info_ready ready; + bool provisioned, sr_valid, active; + u16 ini_version, hw_revision; + u8 slot_index, card_type; + int err = 0; + + mlxsw_reg_mddq_slot_info_unpack(mddq_pl, &slot_index, &provisioned, + &sr_valid, &ready, &active, + &hw_revision, &ini_version, + &card_type); + + if (linecard) { + if (WARN_ON(slot_index != linecard->slot_index)) + return -EINVAL; + } else { + if (WARN_ON(slot_index > linecards->count)) + return -EINVAL; + linecard = mlxsw_linecard_get(linecards, slot_index); + } + + mutex_lock(&linecard->lock); + + if (provisioned && linecard->provisioned != provisioned) { + err = mlxsw_linecard_provision_set(linecard, card_type, + hw_revision, ini_version); + if (err) + goto out; + } + + if (ready == MLXSW_REG_MDDQ_SLOT_INFO_READY_READY && !linecard->ready) { + err = mlxsw_linecard_ready_set(linecard); + if (err) + goto out; + } + + if (active && linecard->active != active) + mlxsw_linecard_active_set(linecard); + + if (!active && linecard->active != active) + mlxsw_linecard_active_clear(linecard); + + if (ready != MLXSW_REG_MDDQ_SLOT_INFO_READY_READY && + linecard->ready) { + err = mlxsw_linecard_ready_clear(linecard); + if (err) + goto out; + } + + if (!provisioned && linecard->provisioned != provisioned) + mlxsw_linecard_provision_clear(linecard); + +out: + mutex_unlock(&linecard->lock); + return err; +} + +static int mlxsw_linecard_status_get_and_process(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecards *linecards, + struct mlxsw_linecard *linecard) +{ + char mddq_pl[MLXSW_REG_MDDQ_LEN]; + int err; + + mlxsw_reg_mddq_slot_info_pack(mddq_pl, linecard->slot_index, false); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mddq), mddq_pl); + if (err) + return err; + + return mlxsw_linecard_status_process(linecards, linecard, mddq_pl); +} + +static void mlxsw_linecards_irq_event_handler(struct mlxsw_core *mlxsw_core) +{ + struct mlxsw_linecards *linecards = mlxsw_core_linecards(mlxsw_core); + int i; + + /* Handle change of line card active state. */ + for (i = 0; i < linecards->count; i++) { + struct mlxsw_linecard *linecard = mlxsw_linecard_get(linecards, + i + 1); + + mlxsw_linecard_status_get_and_process(mlxsw_core, linecards, + linecard); + } +} + +static const char * const mlxsw_linecard_status_event_type_name[] = { + [MLXSW_LINECARD_STATUS_EVENT_TYPE_PROVISION] = "provision", + [MLXSW_LINECARD_STATUS_EVENT_TYPE_UNPROVISION] = "unprovision", +}; + +static void mlxsw_linecard_status_event_to_work(struct work_struct *work) +{ + struct mlxsw_linecard *linecard = + container_of(work, struct mlxsw_linecard, + status_event_to_dw.work); + + mutex_lock(&linecard->lock); + dev_err(linecard->linecards->bus_info->dev, "linecard %u: Timeout reached waiting on %s status event", + linecard->slot_index, + mlxsw_linecard_status_event_type_name[linecard->status_event_type_to]); + mlxsw_linecard_provision_fail(linecard); + mutex_unlock(&linecard->lock); +} + +static int __mlxsw_linecard_fix_fsm_state(struct mlxsw_linecard *linecard) +{ + dev_info(linecard->linecards->bus_info->dev, "linecard %u: Clearing FSM state error", + linecard->slot_index); + mlxsw_reg_mbct_pack(linecard->mbct_pl, linecard->slot_index, + MLXSW_REG_MBCT_OP_CLEAR_ERRORS, false); + return mlxsw_reg_write(linecard->linecards->mlxsw_core, + MLXSW_REG(mbct), linecard->mbct_pl); +} + +static int mlxsw_linecard_fix_fsm_state(struct mlxsw_linecard *linecard, + enum mlxsw_reg_mbct_fsm_state fsm_state) +{ + if (fsm_state != MLXSW_REG_MBCT_FSM_STATE_ERROR) + return 0; + return __mlxsw_linecard_fix_fsm_state(linecard); +} + +static int +mlxsw_linecard_query_ini_status(struct mlxsw_linecard *linecard, + enum mlxsw_reg_mbct_status *status, + enum mlxsw_reg_mbct_fsm_state *fsm_state, + struct netlink_ext_ack *extack) +{ + int err; + + mlxsw_reg_mbct_pack(linecard->mbct_pl, linecard->slot_index, + MLXSW_REG_MBCT_OP_QUERY_STATUS, false); + err = mlxsw_reg_query(linecard->linecards->mlxsw_core, MLXSW_REG(mbct), + linecard->mbct_pl); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to query linecard INI status"); + return err; + } + mlxsw_reg_mbct_unpack(linecard->mbct_pl, NULL, status, fsm_state); + return err; +} + +static int +mlxsw_linecard_ini_transfer(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecard *linecard, + const struct mlxsw_linecard_ini_file *ini_file, + struct netlink_ext_ack *extack) +{ + enum mlxsw_reg_mbct_fsm_state fsm_state; + enum mlxsw_reg_mbct_status status; + size_t size_left; + const u8 *data; + int err; + + size_left = le16_to_cpu(ini_file->size); + data = ini_file->data; + while (size_left) { + size_t data_size = MLXSW_REG_MBCT_DATA_LEN; + bool is_last = false; + + if (size_left <= MLXSW_REG_MBCT_DATA_LEN) { + data_size = size_left; + is_last = true; + } + + mlxsw_reg_mbct_pack(linecard->mbct_pl, linecard->slot_index, + MLXSW_REG_MBCT_OP_DATA_TRANSFER, false); + mlxsw_reg_mbct_dt_pack(linecard->mbct_pl, data_size, + is_last, data); + err = mlxsw_reg_write(mlxsw_core, MLXSW_REG(mbct), + linecard->mbct_pl); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to issue linecard INI data transfer"); + return err; + } + mlxsw_reg_mbct_unpack(linecard->mbct_pl, NULL, + &status, &fsm_state); + if ((!is_last && status != MLXSW_REG_MBCT_STATUS_PART_DATA) || + (is_last && status != MLXSW_REG_MBCT_STATUS_LAST_DATA)) { + NL_SET_ERR_MSG_MOD(extack, "Failed to transfer linecard INI data"); + mlxsw_linecard_fix_fsm_state(linecard, fsm_state); + return -EINVAL; + } + size_left -= data_size; + data += data_size; + } + + return 0; +} + +static int +mlxsw_linecard_ini_erase(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecard *linecard, + struct netlink_ext_ack *extack) +{ + enum mlxsw_reg_mbct_fsm_state fsm_state; + enum mlxsw_reg_mbct_status status; + int err; + + mlxsw_reg_mbct_pack(linecard->mbct_pl, linecard->slot_index, + MLXSW_REG_MBCT_OP_ERASE_INI_IMAGE, false); + err = mlxsw_reg_write(mlxsw_core, MLXSW_REG(mbct), + linecard->mbct_pl); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to issue linecard INI erase"); + return err; + } + mlxsw_reg_mbct_unpack(linecard->mbct_pl, NULL, &status, &fsm_state); + switch (status) { + case MLXSW_REG_MBCT_STATUS_ERASE_COMPLETE: + break; + default: + /* Should not happen */ + fallthrough; + case MLXSW_REG_MBCT_STATUS_ERASE_FAILED: + NL_SET_ERR_MSG_MOD(extack, "Failed to erase linecard INI"); + goto fix_fsm_err_out; + case MLXSW_REG_MBCT_STATUS_ERROR_INI_IN_USE: + NL_SET_ERR_MSG_MOD(extack, "Failed to erase linecard INI while being used"); + goto fix_fsm_err_out; + } + return 0; + +fix_fsm_err_out: + mlxsw_linecard_fix_fsm_state(linecard, fsm_state); + return -EINVAL; +} + +static void mlxsw_linecard_bct_process(struct mlxsw_core *mlxsw_core, + const char *mbct_pl) +{ + struct mlxsw_linecards *linecards = mlxsw_core_linecards(mlxsw_core); + enum mlxsw_reg_mbct_fsm_state fsm_state; + enum mlxsw_reg_mbct_status status; + struct mlxsw_linecard *linecard; + u8 slot_index; + + mlxsw_reg_mbct_unpack(mbct_pl, &slot_index, &status, &fsm_state); + if (WARN_ON(slot_index > linecards->count)) + return; + linecard = mlxsw_linecard_get(linecards, slot_index); + mutex_lock(&linecard->lock); + if (status == MLXSW_REG_MBCT_STATUS_ACTIVATION_FAILED) { + dev_err(linecards->bus_info->dev, "linecard %u: Failed to activate INI", + linecard->slot_index); + goto fix_fsm_out; + } + mutex_unlock(&linecard->lock); + return; + +fix_fsm_out: + mlxsw_linecard_fix_fsm_state(linecard, fsm_state); + mlxsw_linecard_provision_fail(linecard); + mutex_unlock(&linecard->lock); +} + +static int +mlxsw_linecard_ini_activate(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecard *linecard, + struct netlink_ext_ack *extack) +{ + enum mlxsw_reg_mbct_fsm_state fsm_state; + enum mlxsw_reg_mbct_status status; + int err; + + mlxsw_reg_mbct_pack(linecard->mbct_pl, linecard->slot_index, + MLXSW_REG_MBCT_OP_ACTIVATE, true); + err = mlxsw_reg_write(mlxsw_core, MLXSW_REG(mbct), linecard->mbct_pl); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to issue linecard INI activation"); + return err; + } + mlxsw_reg_mbct_unpack(linecard->mbct_pl, NULL, &status, &fsm_state); + if (status == MLXSW_REG_MBCT_STATUS_ACTIVATION_FAILED) { + NL_SET_ERR_MSG_MOD(extack, "Failed to activate linecard INI"); + goto fix_fsm_err_out; + } + + return 0; + +fix_fsm_err_out: + mlxsw_linecard_fix_fsm_state(linecard, fsm_state); + return -EINVAL; +} + +#define MLXSW_LINECARD_INI_WAIT_RETRIES 10 +#define MLXSW_LINECARD_INI_WAIT_MS 500 + +static int +mlxsw_linecard_ini_in_use_wait(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecard *linecard, + struct netlink_ext_ack *extack) +{ + enum mlxsw_reg_mbct_fsm_state fsm_state; + enum mlxsw_reg_mbct_status status; + unsigned int ini_wait_retries = 0; + int err; + +query_ini_status: + err = mlxsw_linecard_query_ini_status(linecard, &status, + &fsm_state, extack); + if (err) + return err; + + switch (fsm_state) { + case MLXSW_REG_MBCT_FSM_STATE_INI_IN_USE: + if (ini_wait_retries++ > MLXSW_LINECARD_INI_WAIT_RETRIES) { + NL_SET_ERR_MSG_MOD(extack, "Failed to wait for linecard INI to be unused"); + return -EINVAL; + } + mdelay(MLXSW_LINECARD_INI_WAIT_MS); + goto query_ini_status; + default: + break; + } + return 0; +} + +static bool mlxsw_linecard_port_selector(void *priv, u16 local_port) +{ + struct mlxsw_linecard *linecard = priv; + struct mlxsw_core *mlxsw_core; + + mlxsw_core = linecard->linecards->mlxsw_core; + return linecard == mlxsw_core_port_linecard_get(mlxsw_core, local_port); +} + +static int mlxsw_linecard_provision(struct devlink_linecard *devlink_linecard, + void *priv, const char *type, + const void *type_priv, + struct netlink_ext_ack *extack) +{ + const struct mlxsw_linecard_ini_file *ini_file = type_priv; + struct mlxsw_linecard *linecard = priv; + struct mlxsw_core *mlxsw_core; + int err; + + mutex_lock(&linecard->lock); + + mlxsw_core = linecard->linecards->mlxsw_core; + + err = mlxsw_linecard_ini_erase(mlxsw_core, linecard, extack); + if (err) + goto err_out; + + err = mlxsw_linecard_ini_transfer(mlxsw_core, linecard, + ini_file, extack); + if (err) + goto err_out; + + mlxsw_linecard_status_event_to_schedule(linecard, + MLXSW_LINECARD_STATUS_EVENT_TYPE_PROVISION); + err = mlxsw_linecard_ini_activate(mlxsw_core, linecard, extack); + if (err) + goto err_out; + + goto out; + +err_out: + mlxsw_linecard_provision_fail(linecard); +out: + mutex_unlock(&linecard->lock); + return err; +} + +static int mlxsw_linecard_unprovision(struct devlink_linecard *devlink_linecard, + void *priv, + struct netlink_ext_ack *extack) +{ + struct mlxsw_linecard *linecard = priv; + struct mlxsw_core *mlxsw_core; + int err; + + mutex_lock(&linecard->lock); + + mlxsw_core = linecard->linecards->mlxsw_core; + + mlxsw_core_ports_remove_selected(mlxsw_core, + mlxsw_linecard_port_selector, + linecard); + + err = mlxsw_linecard_ini_in_use_wait(mlxsw_core, linecard, extack); + if (err) + goto err_out; + + mlxsw_linecard_status_event_to_schedule(linecard, + MLXSW_LINECARD_STATUS_EVENT_TYPE_UNPROVISION); + err = mlxsw_linecard_ini_erase(mlxsw_core, linecard, extack); + if (err) + goto err_out; + + goto out; + +err_out: + mlxsw_linecard_provision_fail(linecard); +out: + mutex_unlock(&linecard->lock); + return err; +} + +static bool mlxsw_linecard_same_provision(struct devlink_linecard *devlink_linecard, + void *priv, const char *type, + const void *type_priv) +{ + const struct mlxsw_linecard_ini_file *ini_file = type_priv; + struct mlxsw_linecard *linecard = priv; + bool ret; + + mutex_lock(&linecard->lock); + ret = linecard->hw_revision == be16_to_cpu(ini_file->format.hw_revision) && + linecard->ini_version == be16_to_cpu(ini_file->format.ini_version); + mutex_unlock(&linecard->lock); + return ret; +} + +static unsigned int +mlxsw_linecard_types_count(struct devlink_linecard *devlink_linecard, + void *priv) +{ + struct mlxsw_linecard *linecard = priv; + + return linecard->linecards->types_info ? + linecard->linecards->types_info->count : 0; +} + +static void mlxsw_linecard_types_get(struct devlink_linecard *devlink_linecard, + void *priv, unsigned int index, + const char **type, const void **type_priv) +{ + struct mlxsw_linecard_types_info *types_info; + struct mlxsw_linecard_ini_file *ini_file; + struct mlxsw_linecard *linecard = priv; + + types_info = linecard->linecards->types_info; + if (WARN_ON_ONCE(!types_info)) + return; + ini_file = types_info->ini_files[index]; + *type = ini_file->format.name; + *type_priv = ini_file; +} + +static const struct devlink_linecard_ops mlxsw_linecard_ops = { + .provision = mlxsw_linecard_provision, + .unprovision = mlxsw_linecard_unprovision, + .same_provision = mlxsw_linecard_same_provision, + .types_count = mlxsw_linecard_types_count, + .types_get = mlxsw_linecard_types_get, +}; + +struct mlxsw_linecard_status_event { + struct mlxsw_core *mlxsw_core; + char mddq_pl[MLXSW_REG_MDDQ_LEN]; + struct work_struct work; +}; + +static void mlxsw_linecard_status_event_work(struct work_struct *work) +{ + struct mlxsw_linecard_status_event *event; + struct mlxsw_linecards *linecards; + struct mlxsw_core *mlxsw_core; + + event = container_of(work, struct mlxsw_linecard_status_event, work); + mlxsw_core = event->mlxsw_core; + linecards = mlxsw_core_linecards(mlxsw_core); + mlxsw_linecard_status_process(linecards, NULL, event->mddq_pl); + kfree(event); +} + +static void +mlxsw_linecard_status_listener_func(const struct mlxsw_reg_info *reg, + char *mddq_pl, void *priv) +{ + struct mlxsw_linecard_status_event *event; + struct mlxsw_core *mlxsw_core = priv; + + event = kmalloc(sizeof(*event), GFP_ATOMIC); + if (!event) + return; + event->mlxsw_core = mlxsw_core; + memcpy(event->mddq_pl, mddq_pl, sizeof(event->mddq_pl)); + INIT_WORK(&event->work, mlxsw_linecard_status_event_work); + mlxsw_core_schedule_work(&event->work); +} + +struct mlxsw_linecard_bct_event { + struct mlxsw_core *mlxsw_core; + char mbct_pl[MLXSW_REG_MBCT_LEN]; + struct work_struct work; +}; + +static void mlxsw_linecard_bct_event_work(struct work_struct *work) +{ + struct mlxsw_linecard_bct_event *event; + struct mlxsw_core *mlxsw_core; + + event = container_of(work, struct mlxsw_linecard_bct_event, work); + mlxsw_core = event->mlxsw_core; + mlxsw_linecard_bct_process(mlxsw_core, event->mbct_pl); + kfree(event); +} + +static void +mlxsw_linecard_bct_listener_func(const struct mlxsw_reg_info *reg, + char *mbct_pl, void *priv) +{ + struct mlxsw_linecard_bct_event *event; + struct mlxsw_core *mlxsw_core = priv; + + event = kmalloc(sizeof(*event), GFP_ATOMIC); + if (!event) + return; + event->mlxsw_core = mlxsw_core; + memcpy(event->mbct_pl, mbct_pl, sizeof(event->mbct_pl)); + INIT_WORK(&event->work, mlxsw_linecard_bct_event_work); + mlxsw_core_schedule_work(&event->work); +} + +static const struct mlxsw_listener mlxsw_linecard_listener[] = { + MLXSW_CORE_EVENTL(mlxsw_linecard_status_listener_func, DSDSC), + MLXSW_CORE_EVENTL(mlxsw_linecard_bct_listener_func, BCTOE), +}; + +static int mlxsw_linecard_event_delivery_set(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecard *linecard, + bool enable) +{ + char mddq_pl[MLXSW_REG_MDDQ_LEN]; + + mlxsw_reg_mddq_slot_info_pack(mddq_pl, linecard->slot_index, enable); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mddq), mddq_pl); +} + +static int mlxsw_linecard_init(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecards *linecards, + u8 slot_index) +{ + struct devlink_linecard *devlink_linecard; + struct mlxsw_linecard *linecard; + + linecard = mlxsw_linecard_get(linecards, slot_index); + linecard->slot_index = slot_index; + linecard->linecards = linecards; + mutex_init(&linecard->lock); + + devlink_linecard = devlink_linecard_create(priv_to_devlink(mlxsw_core), + slot_index, &mlxsw_linecard_ops, + linecard); + if (IS_ERR(devlink_linecard)) + return PTR_ERR(devlink_linecard); + + linecard->devlink_linecard = devlink_linecard; + INIT_DELAYED_WORK(&linecard->status_event_to_dw, + &mlxsw_linecard_status_event_to_work); + + return 0; +} + +static void mlxsw_linecard_fini(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecards *linecards, + u8 slot_index) +{ + struct mlxsw_linecard *linecard; + + linecard = mlxsw_linecard_get(linecards, slot_index); + cancel_delayed_work_sync(&linecard->status_event_to_dw); + /* Make sure all scheduled events are processed */ + mlxsw_core_flush_owq(); + if (linecard->active) + mlxsw_linecard_active_clear(linecard); + mlxsw_linecard_bdev_del(linecard); + devlink_linecard_destroy(linecard->devlink_linecard); + mutex_destroy(&linecard->lock); +} + +static int +mlxsw_linecard_event_delivery_init(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecards *linecards, + u8 slot_index) +{ + struct mlxsw_linecard *linecard; + int err; + + linecard = mlxsw_linecard_get(linecards, slot_index); + err = mlxsw_linecard_event_delivery_set(mlxsw_core, linecard, true); + if (err) + return err; + + err = mlxsw_linecard_status_get_and_process(mlxsw_core, linecards, + linecard); + if (err) + goto err_status_get_and_process; + + return 0; + +err_status_get_and_process: + mlxsw_linecard_event_delivery_set(mlxsw_core, linecard, false); + return err; +} + +static void +mlxsw_linecard_event_delivery_fini(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecards *linecards, + u8 slot_index) +{ + struct mlxsw_linecard *linecard; + + linecard = mlxsw_linecard_get(linecards, slot_index); + mlxsw_linecard_event_delivery_set(mlxsw_core, linecard, false); +} + +/* LINECARDS INI BUNDLE FILE + * +----------------------------------+ + * | MAGIC ("NVLCINI+") | + * +----------------------------------+ +--------------------+ + * | INI 0 +---> | __le16 size | + * +----------------------------------+ | __be16 hw_revision | + * | INI 1 | | __be16 ini_version | + * +----------------------------------+ | u8 __dontcare[3] | + * | ... | | u8 type | + * +----------------------------------+ | u8 name[20] | + * | INI N | | ... | + * +----------------------------------+ +--------------------+ + */ + +#define MLXSW_LINECARDS_INI_BUNDLE_MAGIC "NVLCINI+" + +static int +mlxsw_linecard_types_file_validate(struct mlxsw_linecards *linecards, + struct mlxsw_linecard_types_info *types_info) +{ + size_t magic_size = strlen(MLXSW_LINECARDS_INI_BUNDLE_MAGIC); + struct mlxsw_linecard_ini_file *ini_file; + size_t size = types_info->data_size; + const u8 *data = types_info->data; + unsigned int count = 0; + u16 ini_file_size; + + if (size < magic_size) { + dev_warn(linecards->bus_info->dev, "Invalid linecards INIs file size, smaller than magic size\n"); + return -EINVAL; + } + if (memcmp(data, MLXSW_LINECARDS_INI_BUNDLE_MAGIC, magic_size)) { + dev_warn(linecards->bus_info->dev, "Invalid linecards INIs file magic pattern\n"); + return -EINVAL; + } + + data += magic_size; + size -= magic_size; + + while (size > 0) { + if (size < sizeof(*ini_file)) { + dev_warn(linecards->bus_info->dev, "Linecards INIs file contains INI which is smaller than bare minimum\n"); + return -EINVAL; + } + ini_file = (struct mlxsw_linecard_ini_file *) data; + ini_file_size = le16_to_cpu(ini_file->size); + if (ini_file_size + sizeof(__le16) > size) { + dev_warn(linecards->bus_info->dev, "Linecards INIs file appears to be truncated\n"); + return -EINVAL; + } + if (ini_file_size % 4) { + dev_warn(linecards->bus_info->dev, "Linecards INIs file contains INI with invalid size\n"); + return -EINVAL; + } + data += ini_file_size + sizeof(__le16); + size -= ini_file_size + sizeof(__le16); + count++; + } + if (!count) { + dev_warn(linecards->bus_info->dev, "Linecards INIs file does not contain any INI\n"); + return -EINVAL; + } + types_info->count = count; + return 0; +} + +static void +mlxsw_linecard_types_file_parse(struct mlxsw_linecard_types_info *types_info) +{ + size_t magic_size = strlen(MLXSW_LINECARDS_INI_BUNDLE_MAGIC); + size_t size = types_info->data_size - magic_size; + const u8 *data = types_info->data + magic_size; + struct mlxsw_linecard_ini_file *ini_file; + unsigned int count = 0; + u16 ini_file_size; + int i; + + while (size) { + ini_file = (struct mlxsw_linecard_ini_file *) data; + ini_file_size = le16_to_cpu(ini_file->size); + for (i = 0; i < ini_file_size / 4; i++) { + u32 *val = &((u32 *) ini_file->data)[i]; + + *val = swab32(*val); + } + types_info->ini_files[count] = ini_file; + data += ini_file_size + sizeof(__le16); + size -= ini_file_size + sizeof(__le16); + count++; + } +} + +#define MLXSW_LINECARDS_INI_BUNDLE_FILENAME_FMT \ + "mellanox/lc_ini_bundle_%u_%u.bin" +#define MLXSW_LINECARDS_INI_BUNDLE_FILENAME_LEN \ + (sizeof(MLXSW_LINECARDS_INI_BUNDLE_FILENAME_FMT) + 4) + +static int mlxsw_linecard_types_init(struct mlxsw_core *mlxsw_core, + struct mlxsw_linecards *linecards) +{ + const struct mlxsw_fw_rev *rev = &linecards->bus_info->fw_rev; + char filename[MLXSW_LINECARDS_INI_BUNDLE_FILENAME_LEN]; + struct mlxsw_linecard_types_info *types_info; + const struct firmware *firmware; + int err; + + err = snprintf(filename, sizeof(filename), + MLXSW_LINECARDS_INI_BUNDLE_FILENAME_FMT, + rev->minor, rev->subminor); + WARN_ON(err >= sizeof(filename)); + + err = request_firmware_direct(&firmware, filename, + linecards->bus_info->dev); + if (err) { + dev_warn(linecards->bus_info->dev, "Could not request linecards INI file \"%s\", provisioning will not be possible\n", + filename); + return 0; + } + + types_info = kzalloc(sizeof(*types_info), GFP_KERNEL); + if (!types_info) { + release_firmware(firmware); + return -ENOMEM; + } + linecards->types_info = types_info; + + types_info->data_size = firmware->size; + types_info->data = vmalloc(types_info->data_size); + if (!types_info->data) { + err = -ENOMEM; + release_firmware(firmware); + goto err_data_alloc; + } + memcpy(types_info->data, firmware->data, types_info->data_size); + release_firmware(firmware); + + err = mlxsw_linecard_types_file_validate(linecards, types_info); + if (err) { + err = 0; + goto err_type_file_file_validate; + } + + types_info->ini_files = kmalloc_array(types_info->count, + sizeof(struct mlxsw_linecard_ini_file *), + GFP_KERNEL); + if (!types_info->ini_files) { + err = -ENOMEM; + goto err_ini_files_alloc; + } + + mlxsw_linecard_types_file_parse(types_info); + + return 0; + +err_ini_files_alloc: +err_type_file_file_validate: + vfree(types_info->data); +err_data_alloc: + kfree(types_info); + return err; +} + +static void mlxsw_linecard_types_fini(struct mlxsw_linecards *linecards) +{ + struct mlxsw_linecard_types_info *types_info = linecards->types_info; + + if (!types_info) + return; + kfree(types_info->ini_files); + vfree(types_info->data); + kfree(types_info); +} + +int mlxsw_linecards_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *bus_info) +{ + char mgpir_pl[MLXSW_REG_MGPIR_LEN]; + struct mlxsw_linecards *linecards; + u8 slot_count; + int err; + int i; + + mlxsw_reg_mgpir_pack(mgpir_pl, 0); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(mgpir), mgpir_pl); + if (err) + return err; + + mlxsw_reg_mgpir_unpack(mgpir_pl, NULL, NULL, NULL, + NULL, &slot_count); + if (!slot_count) + return 0; + + linecards = vzalloc(struct_size(linecards, linecards, slot_count)); + if (!linecards) + return -ENOMEM; + linecards->count = slot_count; + linecards->mlxsw_core = mlxsw_core; + linecards->bus_info = bus_info; + INIT_LIST_HEAD(&linecards->event_ops_list); + mutex_init(&linecards->event_ops_list_lock); + + err = mlxsw_linecard_types_init(mlxsw_core, linecards); + if (err) + goto err_types_init; + + err = mlxsw_core_traps_register(mlxsw_core, mlxsw_linecard_listener, + ARRAY_SIZE(mlxsw_linecard_listener), + mlxsw_core); + if (err) + goto err_traps_register; + + err = mlxsw_core_irq_event_handler_register(mlxsw_core, + mlxsw_linecards_irq_event_handler); + if (err) + goto err_irq_event_handler_register; + + mlxsw_core_linecards_set(mlxsw_core, linecards); + + for (i = 0; i < linecards->count; i++) { + err = mlxsw_linecard_init(mlxsw_core, linecards, i + 1); + if (err) + goto err_linecard_init; + } + + for (i = 0; i < linecards->count; i++) { + err = mlxsw_linecard_event_delivery_init(mlxsw_core, linecards, + i + 1); + if (err) + goto err_linecard_event_delivery_init; + } + + return 0; + +err_linecard_event_delivery_init: + for (i--; i >= 0; i--) + mlxsw_linecard_event_delivery_fini(mlxsw_core, linecards, i + 1); + i = linecards->count; +err_linecard_init: + for (i--; i >= 0; i--) + mlxsw_linecard_fini(mlxsw_core, linecards, i + 1); + mlxsw_core_irq_event_handler_unregister(mlxsw_core, + mlxsw_linecards_irq_event_handler); +err_irq_event_handler_register: + mlxsw_core_traps_unregister(mlxsw_core, mlxsw_linecard_listener, + ARRAY_SIZE(mlxsw_linecard_listener), + mlxsw_core); +err_traps_register: + mlxsw_linecard_types_fini(linecards); +err_types_init: + vfree(linecards); + return err; +} + +void mlxsw_linecards_fini(struct mlxsw_core *mlxsw_core) +{ + struct mlxsw_linecards *linecards = mlxsw_core_linecards(mlxsw_core); + int i; + + if (!linecards) + return; + for (i = 0; i < linecards->count; i++) + mlxsw_linecard_event_delivery_fini(mlxsw_core, linecards, i + 1); + for (i = 0; i < linecards->count; i++) + mlxsw_linecard_fini(mlxsw_core, linecards, i + 1); + mlxsw_core_irq_event_handler_unregister(mlxsw_core, + mlxsw_linecards_irq_event_handler); + mlxsw_core_traps_unregister(mlxsw_core, mlxsw_linecard_listener, + ARRAY_SIZE(mlxsw_linecard_listener), + mlxsw_core); + mlxsw_linecard_types_fini(linecards); + mutex_destroy(&linecards->event_ops_list_lock); + WARN_ON(!list_empty(&linecards->event_ops_list)); + vfree(linecards); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_thermal.c b/drivers/net/ethernet/mellanox/mlxsw/core_thermal.c new file mode 100644 index 000000000..ef5e61708 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/core_thermal.c @@ -0,0 +1,1063 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2016-2018 Mellanox Technologies. All rights reserved + * Copyright (c) 2016 Ivan Vecera <cera@cera.cz> + */ + +#include <linux/kernel.h> +#include <linux/types.h> +#include <linux/device.h> +#include <linux/sysfs.h> +#include <linux/thermal.h> +#include <linux/err.h> +#include <linux/sfp.h> + +#include "core.h" +#include "core_env.h" + +#define MLXSW_THERMAL_POLL_INT 1000 /* ms */ +#define MLXSW_THERMAL_SLOW_POLL_INT 20000 /* ms */ +#define MLXSW_THERMAL_ASIC_TEMP_NORM 75000 /* 75C */ +#define MLXSW_THERMAL_ASIC_TEMP_HIGH 85000 /* 85C */ +#define MLXSW_THERMAL_ASIC_TEMP_HOT 105000 /* 105C */ +#define MLXSW_THERMAL_HYSTERESIS_TEMP 5000 /* 5C */ +#define MLXSW_THERMAL_MODULE_TEMP_SHIFT (MLXSW_THERMAL_HYSTERESIS_TEMP * 2) +#define MLXSW_THERMAL_MAX_STATE 10 +#define MLXSW_THERMAL_MIN_STATE 2 +#define MLXSW_THERMAL_MAX_DUTY 255 + +/* External cooling devices, allowed for binding to mlxsw thermal zones. */ +static char * const mlxsw_thermal_external_allowed_cdev[] = { + "mlxreg_fan", +}; + +enum mlxsw_thermal_trips { + MLXSW_THERMAL_TEMP_TRIP_NORM, + MLXSW_THERMAL_TEMP_TRIP_HIGH, + MLXSW_THERMAL_TEMP_TRIP_HOT, +}; + +struct mlxsw_thermal_trip { + int type; + int temp; + int hyst; + int min_state; + int max_state; +}; + +static const struct mlxsw_thermal_trip default_thermal_trips[] = { + { /* In range - 0-40% PWM */ + .type = THERMAL_TRIP_ACTIVE, + .temp = MLXSW_THERMAL_ASIC_TEMP_NORM, + .hyst = MLXSW_THERMAL_HYSTERESIS_TEMP, + .min_state = 0, + .max_state = (4 * MLXSW_THERMAL_MAX_STATE) / 10, + }, + { + /* In range - 40-100% PWM */ + .type = THERMAL_TRIP_ACTIVE, + .temp = MLXSW_THERMAL_ASIC_TEMP_HIGH, + .hyst = MLXSW_THERMAL_HYSTERESIS_TEMP, + .min_state = (4 * MLXSW_THERMAL_MAX_STATE) / 10, + .max_state = MLXSW_THERMAL_MAX_STATE, + }, + { /* Warning */ + .type = THERMAL_TRIP_HOT, + .temp = MLXSW_THERMAL_ASIC_TEMP_HOT, + .min_state = MLXSW_THERMAL_MAX_STATE, + .max_state = MLXSW_THERMAL_MAX_STATE, + }, +}; + +#define MLXSW_THERMAL_NUM_TRIPS ARRAY_SIZE(default_thermal_trips) + +/* Make sure all trips are writable */ +#define MLXSW_THERMAL_TRIP_MASK (BIT(MLXSW_THERMAL_NUM_TRIPS) - 1) + +struct mlxsw_thermal; + +struct mlxsw_thermal_module { + struct mlxsw_thermal *parent; + struct thermal_zone_device *tzdev; + struct mlxsw_thermal_trip trips[MLXSW_THERMAL_NUM_TRIPS]; + int module; /* Module or gearbox number */ + u8 slot_index; +}; + +struct mlxsw_thermal_area { + struct mlxsw_thermal_module *tz_module_arr; + u8 tz_module_num; + struct mlxsw_thermal_module *tz_gearbox_arr; + u8 tz_gearbox_num; + u8 slot_index; + bool active; +}; + +struct mlxsw_thermal { + struct mlxsw_core *core; + const struct mlxsw_bus_info *bus_info; + struct thermal_zone_device *tzdev; + int polling_delay; + struct thermal_cooling_device *cdevs[MLXSW_MFCR_PWMS_MAX]; + struct mlxsw_thermal_trip trips[MLXSW_THERMAL_NUM_TRIPS]; + struct mlxsw_thermal_area line_cards[]; +}; + +static inline u8 mlxsw_state_to_duty(int state) +{ + return DIV_ROUND_CLOSEST(state * MLXSW_THERMAL_MAX_DUTY, + MLXSW_THERMAL_MAX_STATE); +} + +static inline int mlxsw_duty_to_state(u8 duty) +{ + return DIV_ROUND_CLOSEST(duty * MLXSW_THERMAL_MAX_STATE, + MLXSW_THERMAL_MAX_DUTY); +} + +static int mlxsw_get_cooling_device_idx(struct mlxsw_thermal *thermal, + struct thermal_cooling_device *cdev) +{ + int i; + + for (i = 0; i < MLXSW_MFCR_PWMS_MAX; i++) + if (thermal->cdevs[i] == cdev) + return i; + + /* Allow mlxsw thermal zone binding to an external cooling device */ + for (i = 0; i < ARRAY_SIZE(mlxsw_thermal_external_allowed_cdev); i++) { + if (!strcmp(cdev->type, mlxsw_thermal_external_allowed_cdev[i])) + return 0; + } + + return -ENODEV; +} + +static void +mlxsw_thermal_module_trips_reset(struct mlxsw_thermal_module *tz) +{ + tz->trips[MLXSW_THERMAL_TEMP_TRIP_NORM].temp = 0; + tz->trips[MLXSW_THERMAL_TEMP_TRIP_HIGH].temp = 0; + tz->trips[MLXSW_THERMAL_TEMP_TRIP_HOT].temp = 0; +} + +static int +mlxsw_thermal_module_trips_update(struct device *dev, struct mlxsw_core *core, + struct mlxsw_thermal_module *tz, + int crit_temp, int emerg_temp) +{ + int err; + + /* Do not try to query temperature thresholds directly from the module's + * EEPROM if we got valid thresholds from MTMP. + */ + if (!emerg_temp || !crit_temp) { + err = mlxsw_env_module_temp_thresholds_get(core, tz->slot_index, + tz->module, + SFP_TEMP_HIGH_WARN, + &crit_temp); + if (err) + return err; + + err = mlxsw_env_module_temp_thresholds_get(core, tz->slot_index, + tz->module, + SFP_TEMP_HIGH_ALARM, + &emerg_temp); + if (err) + return err; + } + + if (crit_temp > emerg_temp) { + dev_warn(dev, "%s : Critical threshold %d is above emergency threshold %d\n", + tz->tzdev->type, crit_temp, emerg_temp); + return 0; + } + + /* According to the system thermal requirements, the thermal zones are + * defined with three trip points. The critical and emergency + * temperature thresholds, provided by QSFP module are set as "active" + * and "hot" trip points, "normal" trip point is derived from "active" + * by subtracting double hysteresis value. + */ + if (crit_temp >= MLXSW_THERMAL_MODULE_TEMP_SHIFT) + tz->trips[MLXSW_THERMAL_TEMP_TRIP_NORM].temp = crit_temp - + MLXSW_THERMAL_MODULE_TEMP_SHIFT; + else + tz->trips[MLXSW_THERMAL_TEMP_TRIP_NORM].temp = crit_temp; + tz->trips[MLXSW_THERMAL_TEMP_TRIP_HIGH].temp = crit_temp; + tz->trips[MLXSW_THERMAL_TEMP_TRIP_HOT].temp = emerg_temp; + + return 0; +} + +static int mlxsw_thermal_bind(struct thermal_zone_device *tzdev, + struct thermal_cooling_device *cdev) +{ + struct mlxsw_thermal *thermal = tzdev->devdata; + struct device *dev = thermal->bus_info->dev; + int i, err; + + /* If the cooling device is one of ours bind it */ + if (mlxsw_get_cooling_device_idx(thermal, cdev) < 0) + return 0; + + for (i = 0; i < MLXSW_THERMAL_NUM_TRIPS; i++) { + const struct mlxsw_thermal_trip *trip = &thermal->trips[i]; + + err = thermal_zone_bind_cooling_device(tzdev, i, cdev, + trip->max_state, + trip->min_state, + THERMAL_WEIGHT_DEFAULT); + if (err < 0) { + dev_err(dev, "Failed to bind cooling device to trip %d\n", i); + return err; + } + } + return 0; +} + +static int mlxsw_thermal_unbind(struct thermal_zone_device *tzdev, + struct thermal_cooling_device *cdev) +{ + struct mlxsw_thermal *thermal = tzdev->devdata; + struct device *dev = thermal->bus_info->dev; + int i; + int err; + + /* If the cooling device is our one unbind it */ + if (mlxsw_get_cooling_device_idx(thermal, cdev) < 0) + return 0; + + for (i = 0; i < MLXSW_THERMAL_NUM_TRIPS; i++) { + err = thermal_zone_unbind_cooling_device(tzdev, i, cdev); + if (err < 0) { + dev_err(dev, "Failed to unbind cooling device\n"); + return err; + } + } + return 0; +} + +static int mlxsw_thermal_get_temp(struct thermal_zone_device *tzdev, + int *p_temp) +{ + struct mlxsw_thermal *thermal = tzdev->devdata; + struct device *dev = thermal->bus_info->dev; + char mtmp_pl[MLXSW_REG_MTMP_LEN]; + int temp; + int err; + + mlxsw_reg_mtmp_pack(mtmp_pl, 0, 0, false, false); + + err = mlxsw_reg_query(thermal->core, MLXSW_REG(mtmp), mtmp_pl); + if (err) { + dev_err(dev, "Failed to query temp sensor\n"); + return err; + } + mlxsw_reg_mtmp_unpack(mtmp_pl, &temp, NULL, NULL, NULL, NULL); + + *p_temp = temp; + return 0; +} + +static int mlxsw_thermal_get_trip_type(struct thermal_zone_device *tzdev, + int trip, + enum thermal_trip_type *p_type) +{ + struct mlxsw_thermal *thermal = tzdev->devdata; + + if (trip < 0 || trip >= MLXSW_THERMAL_NUM_TRIPS) + return -EINVAL; + + *p_type = thermal->trips[trip].type; + return 0; +} + +static int mlxsw_thermal_get_trip_temp(struct thermal_zone_device *tzdev, + int trip, int *p_temp) +{ + struct mlxsw_thermal *thermal = tzdev->devdata; + + if (trip < 0 || trip >= MLXSW_THERMAL_NUM_TRIPS) + return -EINVAL; + + *p_temp = thermal->trips[trip].temp; + return 0; +} + +static int mlxsw_thermal_set_trip_temp(struct thermal_zone_device *tzdev, + int trip, int temp) +{ + struct mlxsw_thermal *thermal = tzdev->devdata; + + if (trip < 0 || trip >= MLXSW_THERMAL_NUM_TRIPS) + return -EINVAL; + + thermal->trips[trip].temp = temp; + return 0; +} + +static int mlxsw_thermal_get_trip_hyst(struct thermal_zone_device *tzdev, + int trip, int *p_hyst) +{ + struct mlxsw_thermal *thermal = tzdev->devdata; + + *p_hyst = thermal->trips[trip].hyst; + return 0; +} + +static int mlxsw_thermal_set_trip_hyst(struct thermal_zone_device *tzdev, + int trip, int hyst) +{ + struct mlxsw_thermal *thermal = tzdev->devdata; + + thermal->trips[trip].hyst = hyst; + return 0; +} + +static struct thermal_zone_params mlxsw_thermal_params = { + .no_hwmon = true, +}; + +static struct thermal_zone_device_ops mlxsw_thermal_ops = { + .bind = mlxsw_thermal_bind, + .unbind = mlxsw_thermal_unbind, + .get_temp = mlxsw_thermal_get_temp, + .get_trip_type = mlxsw_thermal_get_trip_type, + .get_trip_temp = mlxsw_thermal_get_trip_temp, + .set_trip_temp = mlxsw_thermal_set_trip_temp, + .get_trip_hyst = mlxsw_thermal_get_trip_hyst, + .set_trip_hyst = mlxsw_thermal_set_trip_hyst, +}; + +static int mlxsw_thermal_module_bind(struct thermal_zone_device *tzdev, + struct thermal_cooling_device *cdev) +{ + struct mlxsw_thermal_module *tz = tzdev->devdata; + struct mlxsw_thermal *thermal = tz->parent; + int i, j, err; + + /* If the cooling device is one of ours bind it */ + if (mlxsw_get_cooling_device_idx(thermal, cdev) < 0) + return 0; + + for (i = 0; i < MLXSW_THERMAL_NUM_TRIPS; i++) { + const struct mlxsw_thermal_trip *trip = &tz->trips[i]; + + err = thermal_zone_bind_cooling_device(tzdev, i, cdev, + trip->max_state, + trip->min_state, + THERMAL_WEIGHT_DEFAULT); + if (err < 0) + goto err_thermal_zone_bind_cooling_device; + } + return 0; + +err_thermal_zone_bind_cooling_device: + for (j = i - 1; j >= 0; j--) + thermal_zone_unbind_cooling_device(tzdev, j, cdev); + return err; +} + +static int mlxsw_thermal_module_unbind(struct thermal_zone_device *tzdev, + struct thermal_cooling_device *cdev) +{ + struct mlxsw_thermal_module *tz = tzdev->devdata; + struct mlxsw_thermal *thermal = tz->parent; + int i; + int err; + + /* If the cooling device is one of ours unbind it */ + if (mlxsw_get_cooling_device_idx(thermal, cdev) < 0) + return 0; + + for (i = 0; i < MLXSW_THERMAL_NUM_TRIPS; i++) { + err = thermal_zone_unbind_cooling_device(tzdev, i, cdev); + WARN_ON(err); + } + return err; +} + +static void +mlxsw_thermal_module_temp_and_thresholds_get(struct mlxsw_core *core, + u8 slot_index, u16 sensor_index, + int *p_temp, int *p_crit_temp, + int *p_emerg_temp) +{ + char mtmp_pl[MLXSW_REG_MTMP_LEN]; + int err; + + /* Read module temperature and thresholds. */ + mlxsw_reg_mtmp_pack(mtmp_pl, slot_index, sensor_index, + false, false); + err = mlxsw_reg_query(core, MLXSW_REG(mtmp), mtmp_pl); + if (err) { + /* Set temperature and thresholds to zero to avoid passing + * uninitialized data back to the caller. + */ + *p_temp = 0; + *p_crit_temp = 0; + *p_emerg_temp = 0; + + return; + } + mlxsw_reg_mtmp_unpack(mtmp_pl, p_temp, NULL, p_crit_temp, p_emerg_temp, + NULL); +} + +static int mlxsw_thermal_module_temp_get(struct thermal_zone_device *tzdev, + int *p_temp) +{ + struct mlxsw_thermal_module *tz = tzdev->devdata; + struct mlxsw_thermal *thermal = tz->parent; + int temp, crit_temp, emerg_temp; + struct device *dev; + u16 sensor_index; + + dev = thermal->bus_info->dev; + sensor_index = MLXSW_REG_MTMP_MODULE_INDEX_MIN + tz->module; + + /* Read module temperature and thresholds. */ + mlxsw_thermal_module_temp_and_thresholds_get(thermal->core, + tz->slot_index, + sensor_index, &temp, + &crit_temp, &emerg_temp); + *p_temp = temp; + + if (!temp) + return 0; + + /* Update trip points. */ + mlxsw_thermal_module_trips_update(dev, thermal->core, tz, + crit_temp, emerg_temp); + + return 0; +} + +static int +mlxsw_thermal_module_trip_type_get(struct thermal_zone_device *tzdev, int trip, + enum thermal_trip_type *p_type) +{ + struct mlxsw_thermal_module *tz = tzdev->devdata; + + if (trip < 0 || trip >= MLXSW_THERMAL_NUM_TRIPS) + return -EINVAL; + + *p_type = tz->trips[trip].type; + return 0; +} + +static int +mlxsw_thermal_module_trip_temp_get(struct thermal_zone_device *tzdev, + int trip, int *p_temp) +{ + struct mlxsw_thermal_module *tz = tzdev->devdata; + + if (trip < 0 || trip >= MLXSW_THERMAL_NUM_TRIPS) + return -EINVAL; + + *p_temp = tz->trips[trip].temp; + return 0; +} + +static int +mlxsw_thermal_module_trip_temp_set(struct thermal_zone_device *tzdev, + int trip, int temp) +{ + struct mlxsw_thermal_module *tz = tzdev->devdata; + + if (trip < 0 || trip >= MLXSW_THERMAL_NUM_TRIPS) + return -EINVAL; + + tz->trips[trip].temp = temp; + return 0; +} + +static int +mlxsw_thermal_module_trip_hyst_get(struct thermal_zone_device *tzdev, int trip, + int *p_hyst) +{ + struct mlxsw_thermal_module *tz = tzdev->devdata; + + *p_hyst = tz->trips[trip].hyst; + return 0; +} + +static int +mlxsw_thermal_module_trip_hyst_set(struct thermal_zone_device *tzdev, int trip, + int hyst) +{ + struct mlxsw_thermal_module *tz = tzdev->devdata; + + tz->trips[trip].hyst = hyst; + return 0; +} + +static struct thermal_zone_device_ops mlxsw_thermal_module_ops = { + .bind = mlxsw_thermal_module_bind, + .unbind = mlxsw_thermal_module_unbind, + .get_temp = mlxsw_thermal_module_temp_get, + .get_trip_type = mlxsw_thermal_module_trip_type_get, + .get_trip_temp = mlxsw_thermal_module_trip_temp_get, + .set_trip_temp = mlxsw_thermal_module_trip_temp_set, + .get_trip_hyst = mlxsw_thermal_module_trip_hyst_get, + .set_trip_hyst = mlxsw_thermal_module_trip_hyst_set, +}; + +static int mlxsw_thermal_gearbox_temp_get(struct thermal_zone_device *tzdev, + int *p_temp) +{ + struct mlxsw_thermal_module *tz = tzdev->devdata; + struct mlxsw_thermal *thermal = tz->parent; + char mtmp_pl[MLXSW_REG_MTMP_LEN]; + u16 index; + int temp; + int err; + + index = MLXSW_REG_MTMP_GBOX_INDEX_MIN + tz->module; + mlxsw_reg_mtmp_pack(mtmp_pl, tz->slot_index, index, false, false); + + err = mlxsw_reg_query(thermal->core, MLXSW_REG(mtmp), mtmp_pl); + if (err) + return err; + + mlxsw_reg_mtmp_unpack(mtmp_pl, &temp, NULL, NULL, NULL, NULL); + + *p_temp = temp; + return 0; +} + +static struct thermal_zone_device_ops mlxsw_thermal_gearbox_ops = { + .bind = mlxsw_thermal_module_bind, + .unbind = mlxsw_thermal_module_unbind, + .get_temp = mlxsw_thermal_gearbox_temp_get, + .get_trip_type = mlxsw_thermal_module_trip_type_get, + .get_trip_temp = mlxsw_thermal_module_trip_temp_get, + .set_trip_temp = mlxsw_thermal_module_trip_temp_set, + .get_trip_hyst = mlxsw_thermal_module_trip_hyst_get, + .set_trip_hyst = mlxsw_thermal_module_trip_hyst_set, +}; + +static int mlxsw_thermal_get_max_state(struct thermal_cooling_device *cdev, + unsigned long *p_state) +{ + *p_state = MLXSW_THERMAL_MAX_STATE; + return 0; +} + +static int mlxsw_thermal_get_cur_state(struct thermal_cooling_device *cdev, + unsigned long *p_state) + +{ + struct mlxsw_thermal *thermal = cdev->devdata; + struct device *dev = thermal->bus_info->dev; + char mfsc_pl[MLXSW_REG_MFSC_LEN]; + int err, idx; + u8 duty; + + idx = mlxsw_get_cooling_device_idx(thermal, cdev); + if (idx < 0) + return idx; + + mlxsw_reg_mfsc_pack(mfsc_pl, idx, 0); + err = mlxsw_reg_query(thermal->core, MLXSW_REG(mfsc), mfsc_pl); + if (err) { + dev_err(dev, "Failed to query PWM duty\n"); + return err; + } + + duty = mlxsw_reg_mfsc_pwm_duty_cycle_get(mfsc_pl); + *p_state = mlxsw_duty_to_state(duty); + return 0; +} + +static int mlxsw_thermal_set_cur_state(struct thermal_cooling_device *cdev, + unsigned long state) + +{ + struct mlxsw_thermal *thermal = cdev->devdata; + struct device *dev = thermal->bus_info->dev; + char mfsc_pl[MLXSW_REG_MFSC_LEN]; + int idx; + int err; + + if (state > MLXSW_THERMAL_MAX_STATE) + return -EINVAL; + + idx = mlxsw_get_cooling_device_idx(thermal, cdev); + if (idx < 0) + return idx; + + /* Normalize the state to the valid speed range. */ + state = max_t(unsigned long, MLXSW_THERMAL_MIN_STATE, state); + mlxsw_reg_mfsc_pack(mfsc_pl, idx, mlxsw_state_to_duty(state)); + err = mlxsw_reg_write(thermal->core, MLXSW_REG(mfsc), mfsc_pl); + if (err) { + dev_err(dev, "Failed to write PWM duty\n"); + return err; + } + return 0; +} + +static const struct thermal_cooling_device_ops mlxsw_cooling_ops = { + .get_max_state = mlxsw_thermal_get_max_state, + .get_cur_state = mlxsw_thermal_get_cur_state, + .set_cur_state = mlxsw_thermal_set_cur_state, +}; + +static int +mlxsw_thermal_module_tz_init(struct mlxsw_thermal_module *module_tz) +{ + char tz_name[THERMAL_NAME_LENGTH]; + int err; + + if (module_tz->slot_index) + snprintf(tz_name, sizeof(tz_name), "mlxsw-lc%d-module%d", + module_tz->slot_index, module_tz->module + 1); + else + snprintf(tz_name, sizeof(tz_name), "mlxsw-module%d", + module_tz->module + 1); + module_tz->tzdev = thermal_zone_device_register(tz_name, + MLXSW_THERMAL_NUM_TRIPS, + MLXSW_THERMAL_TRIP_MASK, + module_tz, + &mlxsw_thermal_module_ops, + &mlxsw_thermal_params, + 0, + module_tz->parent->polling_delay); + if (IS_ERR(module_tz->tzdev)) { + err = PTR_ERR(module_tz->tzdev); + return err; + } + + err = thermal_zone_device_enable(module_tz->tzdev); + if (err) + thermal_zone_device_unregister(module_tz->tzdev); + + return err; +} + +static void mlxsw_thermal_module_tz_fini(struct thermal_zone_device *tzdev) +{ + thermal_zone_device_unregister(tzdev); +} + +static int +mlxsw_thermal_module_init(struct device *dev, struct mlxsw_core *core, + struct mlxsw_thermal *thermal, + struct mlxsw_thermal_area *area, u8 module) +{ + struct mlxsw_thermal_module *module_tz; + int dummy_temp, crit_temp, emerg_temp; + u16 sensor_index; + + sensor_index = MLXSW_REG_MTMP_MODULE_INDEX_MIN + module; + module_tz = &area->tz_module_arr[module]; + /* Skip if parent is already set (case of port split). */ + if (module_tz->parent) + return 0; + module_tz->module = module; + module_tz->slot_index = area->slot_index; + module_tz->parent = thermal; + memcpy(module_tz->trips, default_thermal_trips, + sizeof(thermal->trips)); + /* Initialize all trip point. */ + mlxsw_thermal_module_trips_reset(module_tz); + /* Read module temperature and thresholds. */ + mlxsw_thermal_module_temp_and_thresholds_get(core, area->slot_index, + sensor_index, &dummy_temp, + &crit_temp, &emerg_temp); + /* Update trip point according to the module data. */ + return mlxsw_thermal_module_trips_update(dev, core, module_tz, + crit_temp, emerg_temp); +} + +static void mlxsw_thermal_module_fini(struct mlxsw_thermal_module *module_tz) +{ + if (module_tz && module_tz->tzdev) { + mlxsw_thermal_module_tz_fini(module_tz->tzdev); + module_tz->tzdev = NULL; + module_tz->parent = NULL; + } +} + +static int +mlxsw_thermal_modules_init(struct device *dev, struct mlxsw_core *core, + struct mlxsw_thermal *thermal, + struct mlxsw_thermal_area *area) +{ + struct mlxsw_thermal_module *module_tz; + char mgpir_pl[MLXSW_REG_MGPIR_LEN]; + int i, err; + + mlxsw_reg_mgpir_pack(mgpir_pl, area->slot_index); + err = mlxsw_reg_query(core, MLXSW_REG(mgpir), mgpir_pl); + if (err) + return err; + + mlxsw_reg_mgpir_unpack(mgpir_pl, NULL, NULL, NULL, + &area->tz_module_num, NULL); + + /* For modular system module counter could be zero. */ + if (!area->tz_module_num) + return 0; + + area->tz_module_arr = kcalloc(area->tz_module_num, + sizeof(*area->tz_module_arr), + GFP_KERNEL); + if (!area->tz_module_arr) + return -ENOMEM; + + for (i = 0; i < area->tz_module_num; i++) { + err = mlxsw_thermal_module_init(dev, core, thermal, area, i); + if (err) + goto err_thermal_module_init; + } + + for (i = 0; i < area->tz_module_num; i++) { + module_tz = &area->tz_module_arr[i]; + if (!module_tz->parent) + continue; + err = mlxsw_thermal_module_tz_init(module_tz); + if (err) + goto err_thermal_module_tz_init; + } + + return 0; + +err_thermal_module_tz_init: +err_thermal_module_init: + for (i = area->tz_module_num - 1; i >= 0; i--) + mlxsw_thermal_module_fini(&area->tz_module_arr[i]); + kfree(area->tz_module_arr); + return err; +} + +static void +mlxsw_thermal_modules_fini(struct mlxsw_thermal *thermal, + struct mlxsw_thermal_area *area) +{ + int i; + + for (i = area->tz_module_num - 1; i >= 0; i--) + mlxsw_thermal_module_fini(&area->tz_module_arr[i]); + kfree(area->tz_module_arr); +} + +static int +mlxsw_thermal_gearbox_tz_init(struct mlxsw_thermal_module *gearbox_tz) +{ + char tz_name[THERMAL_NAME_LENGTH]; + int ret; + + if (gearbox_tz->slot_index) + snprintf(tz_name, sizeof(tz_name), "mlxsw-lc%d-gearbox%d", + gearbox_tz->slot_index, gearbox_tz->module + 1); + else + snprintf(tz_name, sizeof(tz_name), "mlxsw-gearbox%d", + gearbox_tz->module + 1); + gearbox_tz->tzdev = thermal_zone_device_register(tz_name, + MLXSW_THERMAL_NUM_TRIPS, + MLXSW_THERMAL_TRIP_MASK, + gearbox_tz, + &mlxsw_thermal_gearbox_ops, + &mlxsw_thermal_params, 0, + gearbox_tz->parent->polling_delay); + if (IS_ERR(gearbox_tz->tzdev)) + return PTR_ERR(gearbox_tz->tzdev); + + ret = thermal_zone_device_enable(gearbox_tz->tzdev); + if (ret) + thermal_zone_device_unregister(gearbox_tz->tzdev); + + return ret; +} + +static void +mlxsw_thermal_gearbox_tz_fini(struct mlxsw_thermal_module *gearbox_tz) +{ + thermal_zone_device_unregister(gearbox_tz->tzdev); +} + +static int +mlxsw_thermal_gearboxes_init(struct device *dev, struct mlxsw_core *core, + struct mlxsw_thermal *thermal, + struct mlxsw_thermal_area *area) +{ + enum mlxsw_reg_mgpir_device_type device_type; + struct mlxsw_thermal_module *gearbox_tz; + char mgpir_pl[MLXSW_REG_MGPIR_LEN]; + u8 gbox_num; + int i; + int err; + + mlxsw_reg_mgpir_pack(mgpir_pl, area->slot_index); + err = mlxsw_reg_query(core, MLXSW_REG(mgpir), mgpir_pl); + if (err) + return err; + + mlxsw_reg_mgpir_unpack(mgpir_pl, &gbox_num, &device_type, NULL, + NULL, NULL); + if (device_type != MLXSW_REG_MGPIR_DEVICE_TYPE_GEARBOX_DIE || + !gbox_num) + return 0; + + area->tz_gearbox_num = gbox_num; + area->tz_gearbox_arr = kcalloc(area->tz_gearbox_num, + sizeof(*area->tz_gearbox_arr), + GFP_KERNEL); + if (!area->tz_gearbox_arr) + return -ENOMEM; + + for (i = 0; i < area->tz_gearbox_num; i++) { + gearbox_tz = &area->tz_gearbox_arr[i]; + memcpy(gearbox_tz->trips, default_thermal_trips, + sizeof(thermal->trips)); + gearbox_tz->module = i; + gearbox_tz->parent = thermal; + gearbox_tz->slot_index = area->slot_index; + err = mlxsw_thermal_gearbox_tz_init(gearbox_tz); + if (err) + goto err_thermal_gearbox_tz_init; + } + + return 0; + +err_thermal_gearbox_tz_init: + for (i--; i >= 0; i--) + mlxsw_thermal_gearbox_tz_fini(&area->tz_gearbox_arr[i]); + kfree(area->tz_gearbox_arr); + return err; +} + +static void +mlxsw_thermal_gearboxes_fini(struct mlxsw_thermal *thermal, + struct mlxsw_thermal_area *area) +{ + int i; + + for (i = area->tz_gearbox_num - 1; i >= 0; i--) + mlxsw_thermal_gearbox_tz_fini(&area->tz_gearbox_arr[i]); + kfree(area->tz_gearbox_arr); +} + +static void +mlxsw_thermal_got_active(struct mlxsw_core *mlxsw_core, u8 slot_index, + void *priv) +{ + struct mlxsw_thermal *thermal = priv; + struct mlxsw_thermal_area *linecard; + int err; + + linecard = &thermal->line_cards[slot_index]; + + if (linecard->active) + return; + + linecard->slot_index = slot_index; + err = mlxsw_thermal_modules_init(thermal->bus_info->dev, thermal->core, + thermal, linecard); + if (err) { + dev_err(thermal->bus_info->dev, "Failed to configure thermal objects for line card modules in slot %d\n", + slot_index); + return; + } + + err = mlxsw_thermal_gearboxes_init(thermal->bus_info->dev, + thermal->core, thermal, linecard); + if (err) { + dev_err(thermal->bus_info->dev, "Failed to configure thermal objects for line card gearboxes in slot %d\n", + slot_index); + goto err_thermal_linecard_gearboxes_init; + } + + linecard->active = true; + + return; + +err_thermal_linecard_gearboxes_init: + mlxsw_thermal_modules_fini(thermal, linecard); +} + +static void +mlxsw_thermal_got_inactive(struct mlxsw_core *mlxsw_core, u8 slot_index, + void *priv) +{ + struct mlxsw_thermal *thermal = priv; + struct mlxsw_thermal_area *linecard; + + linecard = &thermal->line_cards[slot_index]; + if (!linecard->active) + return; + linecard->active = false; + mlxsw_thermal_gearboxes_fini(thermal, linecard); + mlxsw_thermal_modules_fini(thermal, linecard); +} + +static struct mlxsw_linecards_event_ops mlxsw_thermal_event_ops = { + .got_active = mlxsw_thermal_got_active, + .got_inactive = mlxsw_thermal_got_inactive, +}; + +int mlxsw_thermal_init(struct mlxsw_core *core, + const struct mlxsw_bus_info *bus_info, + struct mlxsw_thermal **p_thermal) +{ + char mfcr_pl[MLXSW_REG_MFCR_LEN] = { 0 }; + enum mlxsw_reg_mfcr_pwm_frequency freq; + struct device *dev = bus_info->dev; + char mgpir_pl[MLXSW_REG_MGPIR_LEN]; + struct mlxsw_thermal *thermal; + u8 pwm_active, num_of_slots; + u16 tacho_active; + int err, i; + + mlxsw_reg_mgpir_pack(mgpir_pl, 0); + err = mlxsw_reg_query(core, MLXSW_REG(mgpir), mgpir_pl); + if (err) + return err; + + mlxsw_reg_mgpir_unpack(mgpir_pl, NULL, NULL, NULL, NULL, + &num_of_slots); + + thermal = kzalloc(struct_size(thermal, line_cards, num_of_slots + 1), + GFP_KERNEL); + if (!thermal) + return -ENOMEM; + + thermal->core = core; + thermal->bus_info = bus_info; + memcpy(thermal->trips, default_thermal_trips, sizeof(thermal->trips)); + thermal->line_cards[0].slot_index = 0; + + err = mlxsw_reg_query(thermal->core, MLXSW_REG(mfcr), mfcr_pl); + if (err) { + dev_err(dev, "Failed to probe PWMs\n"); + goto err_reg_query; + } + mlxsw_reg_mfcr_unpack(mfcr_pl, &freq, &tacho_active, &pwm_active); + + for (i = 0; i < MLXSW_MFCR_TACHOS_MAX; i++) { + if (tacho_active & BIT(i)) { + char mfsl_pl[MLXSW_REG_MFSL_LEN]; + + mlxsw_reg_mfsl_pack(mfsl_pl, i, 0, 0); + + /* We need to query the register to preserve maximum */ + err = mlxsw_reg_query(thermal->core, MLXSW_REG(mfsl), + mfsl_pl); + if (err) + goto err_reg_query; + + /* set the minimal RPMs to 0 */ + mlxsw_reg_mfsl_tach_min_set(mfsl_pl, 0); + err = mlxsw_reg_write(thermal->core, MLXSW_REG(mfsl), + mfsl_pl); + if (err) + goto err_reg_write; + } + } + for (i = 0; i < MLXSW_MFCR_PWMS_MAX; i++) { + if (pwm_active & BIT(i)) { + struct thermal_cooling_device *cdev; + + cdev = thermal_cooling_device_register("mlxsw_fan", + thermal, + &mlxsw_cooling_ops); + if (IS_ERR(cdev)) { + err = PTR_ERR(cdev); + dev_err(dev, "Failed to register cooling device\n"); + goto err_thermal_cooling_device_register; + } + thermal->cdevs[i] = cdev; + } + } + + thermal->polling_delay = bus_info->low_frequency ? + MLXSW_THERMAL_SLOW_POLL_INT : + MLXSW_THERMAL_POLL_INT; + + thermal->tzdev = thermal_zone_device_register("mlxsw", + MLXSW_THERMAL_NUM_TRIPS, + MLXSW_THERMAL_TRIP_MASK, + thermal, + &mlxsw_thermal_ops, + &mlxsw_thermal_params, 0, + thermal->polling_delay); + if (IS_ERR(thermal->tzdev)) { + err = PTR_ERR(thermal->tzdev); + dev_err(dev, "Failed to register thermal zone\n"); + goto err_thermal_zone_device_register; + } + + err = mlxsw_thermal_modules_init(dev, core, thermal, + &thermal->line_cards[0]); + if (err) + goto err_thermal_modules_init; + + err = mlxsw_thermal_gearboxes_init(dev, core, thermal, + &thermal->line_cards[0]); + if (err) + goto err_thermal_gearboxes_init; + + err = mlxsw_linecards_event_ops_register(core, + &mlxsw_thermal_event_ops, + thermal); + if (err) + goto err_linecards_event_ops_register; + + err = thermal_zone_device_enable(thermal->tzdev); + if (err) + goto err_thermal_zone_device_enable; + + thermal->line_cards[0].active = true; + *p_thermal = thermal; + return 0; + +err_thermal_zone_device_enable: + mlxsw_linecards_event_ops_unregister(thermal->core, + &mlxsw_thermal_event_ops, + thermal); +err_linecards_event_ops_register: + mlxsw_thermal_gearboxes_fini(thermal, &thermal->line_cards[0]); +err_thermal_gearboxes_init: + mlxsw_thermal_modules_fini(thermal, &thermal->line_cards[0]); +err_thermal_modules_init: + if (thermal->tzdev) { + thermal_zone_device_unregister(thermal->tzdev); + thermal->tzdev = NULL; + } +err_thermal_zone_device_register: +err_thermal_cooling_device_register: + for (i = 0; i < MLXSW_MFCR_PWMS_MAX; i++) + if (thermal->cdevs[i]) + thermal_cooling_device_unregister(thermal->cdevs[i]); +err_reg_write: +err_reg_query: + kfree(thermal); + return err; +} + +void mlxsw_thermal_fini(struct mlxsw_thermal *thermal) +{ + int i; + + thermal->line_cards[0].active = false; + mlxsw_linecards_event_ops_unregister(thermal->core, + &mlxsw_thermal_event_ops, + thermal); + mlxsw_thermal_gearboxes_fini(thermal, &thermal->line_cards[0]); + mlxsw_thermal_modules_fini(thermal, &thermal->line_cards[0]); + if (thermal->tzdev) { + thermal_zone_device_unregister(thermal->tzdev); + thermal->tzdev = NULL; + } + + for (i = 0; i < MLXSW_MFCR_PWMS_MAX; i++) { + if (thermal->cdevs[i]) { + thermal_cooling_device_unregister(thermal->cdevs[i]); + thermal->cdevs[i] = NULL; + } + } + + kfree(thermal); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/emad.h b/drivers/net/ethernet/mellanox/mlxsw/emad.h new file mode 100644 index 000000000..acfbbec52 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/emad.h @@ -0,0 +1,96 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_EMAD_H +#define _MLXSW_EMAD_H + +#define MLXSW_EMAD_MAX_FRAME_LEN 1518 /* Length in u8 */ +#define MLXSW_EMAD_MAX_RETRY 5 + +/* EMAD Ethernet header */ +#define MLXSW_EMAD_ETH_HDR_LEN 0x10 /* Length in u8 */ +#define MLXSW_EMAD_EH_DMAC "\x01\x02\xc9\x00\x00\x01" +#define MLXSW_EMAD_EH_SMAC "\x00\x02\xc9\x01\x02\x03" +#define MLXSW_EMAD_EH_ETHERTYPE 0x8932 +#define MLXSW_EMAD_EH_MLX_PROTO 0 +#define MLXSW_EMAD_EH_PROTO_VERSION 0 + +/* EMAD TLV Types */ +enum { + MLXSW_EMAD_TLV_TYPE_END, + MLXSW_EMAD_TLV_TYPE_OP, + MLXSW_EMAD_TLV_TYPE_STRING, + MLXSW_EMAD_TLV_TYPE_REG, +}; + +/* OP TLV */ +#define MLXSW_EMAD_OP_TLV_LEN 4 /* Length in u32 */ + +enum { + MLXSW_EMAD_OP_TLV_CLASS_REG_ACCESS = 1, + MLXSW_EMAD_OP_TLV_CLASS_IPC = 2, +}; + +enum mlxsw_emad_op_tlv_status { + MLXSW_EMAD_OP_TLV_STATUS_SUCCESS, + MLXSW_EMAD_OP_TLV_STATUS_BUSY, + MLXSW_EMAD_OP_TLV_STATUS_VERSION_NOT_SUPPORTED, + MLXSW_EMAD_OP_TLV_STATUS_UNKNOWN_TLV, + MLXSW_EMAD_OP_TLV_STATUS_REGISTER_NOT_SUPPORTED, + MLXSW_EMAD_OP_TLV_STATUS_CLASS_NOT_SUPPORTED, + MLXSW_EMAD_OP_TLV_STATUS_METHOD_NOT_SUPPORTED, + MLXSW_EMAD_OP_TLV_STATUS_BAD_PARAMETER, + MLXSW_EMAD_OP_TLV_STATUS_RESOURCE_NOT_AVAILABLE, + MLXSW_EMAD_OP_TLV_STATUS_MESSAGE_RECEIPT_ACK, + MLXSW_EMAD_OP_TLV_STATUS_INTERNAL_ERROR = 0x70, +}; + +static inline char *mlxsw_emad_op_tlv_status_str(u8 status) +{ + switch (status) { + case MLXSW_EMAD_OP_TLV_STATUS_SUCCESS: + return "operation performed"; + case MLXSW_EMAD_OP_TLV_STATUS_BUSY: + return "device is busy"; + case MLXSW_EMAD_OP_TLV_STATUS_VERSION_NOT_SUPPORTED: + return "version not supported"; + case MLXSW_EMAD_OP_TLV_STATUS_UNKNOWN_TLV: + return "unknown TLV"; + case MLXSW_EMAD_OP_TLV_STATUS_REGISTER_NOT_SUPPORTED: + return "register not supported"; + case MLXSW_EMAD_OP_TLV_STATUS_CLASS_NOT_SUPPORTED: + return "class not supported"; + case MLXSW_EMAD_OP_TLV_STATUS_METHOD_NOT_SUPPORTED: + return "method not supported"; + case MLXSW_EMAD_OP_TLV_STATUS_BAD_PARAMETER: + return "bad parameter"; + case MLXSW_EMAD_OP_TLV_STATUS_RESOURCE_NOT_AVAILABLE: + return "resource not available"; + case MLXSW_EMAD_OP_TLV_STATUS_MESSAGE_RECEIPT_ACK: + return "acknowledged. retransmit"; + case MLXSW_EMAD_OP_TLV_STATUS_INTERNAL_ERROR: + return "internal error"; + default: + return "*UNKNOWN*"; + } +} + +enum { + MLXSW_EMAD_OP_TLV_REQUEST, + MLXSW_EMAD_OP_TLV_RESPONSE +}; + +enum { + MLXSW_EMAD_OP_TLV_METHOD_QUERY = 1, + MLXSW_EMAD_OP_TLV_METHOD_WRITE = 2, + MLXSW_EMAD_OP_TLV_METHOD_SEND = 3, + MLXSW_EMAD_OP_TLV_METHOD_EVENT = 5, +}; + +/* STRING TLV */ +#define MLXSW_EMAD_STRING_TLV_LEN 33 /* Length in u32 */ + +/* END TLV */ +#define MLXSW_EMAD_END_TLV_LEN 1 /* Length in u32 */ + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/i2c.c b/drivers/net/ethernet/mellanox/mlxsw/i2c.c new file mode 100644 index 000000000..3beefc167 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/i2c.c @@ -0,0 +1,769 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2016-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/err.h> +#include <linux/i2c.h> +#include <linux/init.h> +#include <linux/jiffies.h> +#include <linux/kernel.h> +#include <linux/mutex.h> +#include <linux/module.h> +#include <linux/mod_devicetable.h> +#include <linux/platform_data/mlxreg.h> +#include <linux/slab.h> + +#include "cmd.h" +#include "core.h" +#include "i2c.h" +#include "resources.h" + +#define MLXSW_I2C_CIR2_BASE 0x72000 +#define MLXSW_I2C_CIR_STATUS_OFF 0x18 +#define MLXSW_I2C_CIR2_OFF_STATUS (MLXSW_I2C_CIR2_BASE + \ + MLXSW_I2C_CIR_STATUS_OFF) +#define MLXSW_I2C_OPMOD_SHIFT 12 +#define MLXSW_I2C_EVENT_BIT_SHIFT 22 +#define MLXSW_I2C_GO_BIT_SHIFT 23 +#define MLXSW_I2C_CIR_CTRL_STATUS_SHIFT 24 +#define MLXSW_I2C_EVENT_BIT BIT(MLXSW_I2C_EVENT_BIT_SHIFT) +#define MLXSW_I2C_GO_BIT BIT(MLXSW_I2C_GO_BIT_SHIFT) +#define MLXSW_I2C_GO_OPMODE BIT(MLXSW_I2C_OPMOD_SHIFT) +#define MLXSW_I2C_SET_IMM_CMD (MLXSW_I2C_GO_OPMODE | \ + MLXSW_CMD_OPCODE_QUERY_FW) +#define MLXSW_I2C_PUSH_IMM_CMD (MLXSW_I2C_GO_BIT | \ + MLXSW_I2C_SET_IMM_CMD) +#define MLXSW_I2C_SET_CMD (MLXSW_CMD_OPCODE_ACCESS_REG) +#define MLXSW_I2C_PUSH_CMD (MLXSW_I2C_GO_BIT | MLXSW_I2C_SET_CMD) +#define MLXSW_I2C_TLV_HDR_SIZE 0x10 +#define MLXSW_I2C_ADDR_WIDTH 4 +#define MLXSW_I2C_PUSH_CMD_SIZE (MLXSW_I2C_ADDR_WIDTH + 4) +#define MLXSW_I2C_SET_EVENT_CMD (MLXSW_I2C_EVENT_BIT) +#define MLXSW_I2C_PUSH_EVENT_CMD (MLXSW_I2C_GO_BIT | \ + MLXSW_I2C_SET_EVENT_CMD) +#define MLXSW_I2C_READ_SEMA_SIZE 4 +#define MLXSW_I2C_PREP_SIZE (MLXSW_I2C_ADDR_WIDTH + 28) +#define MLXSW_I2C_MBOX_SIZE 20 +#define MLXSW_I2C_MBOX_OUT_PARAM_OFF 12 +#define MLXSW_I2C_MBOX_OFFSET_BITS 20 +#define MLXSW_I2C_MBOX_SIZE_BITS 12 +#define MLXSW_I2C_ADDR_BUF_SIZE 4 +#define MLXSW_I2C_BLK_DEF 32 +#define MLXSW_I2C_BLK_MAX 100 +#define MLXSW_I2C_RETRY 5 +#define MLXSW_I2C_TIMEOUT_MSECS 5000 +#define MLXSW_I2C_MAX_DATA_SIZE 256 + +/* Driver can be initialized by kernel platform driver or from the user + * space. In the first case IRQ line number is passed through the platform + * data, otherwise default IRQ line is to be used. Default IRQ is relevant + * only for specific I2C slave address, allowing 3.4 MHz I2C path to the chip + * (special hardware feature for I2C acceleration). + */ +#define MLXSW_I2C_DEFAULT_IRQ 17 +#define MLXSW_FAST_I2C_SLAVE 0x37 + +/** + * struct mlxsw_i2c - device private data: + * @cmd: command attributes; + * @cmd.mb_size_in: input mailbox size; + * @cmd.mb_off_in: input mailbox offset in register space; + * @cmd.mb_size_out: output mailbox size; + * @cmd.mb_off_out: output mailbox offset in register space; + * @cmd.lock: command execution lock; + * @dev: I2C device; + * @core: switch core pointer; + * @bus_info: bus info block; + * @block_size: maximum block size allowed to pass to under layer; + * @pdata: device platform data; + * @irq_work: interrupts work item; + * @irq: IRQ line number; + */ +struct mlxsw_i2c { + struct { + u32 mb_size_in; + u32 mb_off_in; + u32 mb_size_out; + u32 mb_off_out; + struct mutex lock; + } cmd; + struct device *dev; + struct mlxsw_core *core; + struct mlxsw_bus_info bus_info; + u16 block_size; + struct mlxreg_core_hotplug_platform_data *pdata; + struct work_struct irq_work; + int irq; +}; + +#define MLXSW_I2C_READ_MSG(_client, _addr_buf, _buf, _len) { \ + { .addr = (_client)->addr, \ + .buf = (_addr_buf), \ + .len = MLXSW_I2C_ADDR_BUF_SIZE, \ + .flags = 0 }, \ + { .addr = (_client)->addr, \ + .buf = (_buf), \ + .len = (_len), \ + .flags = I2C_M_RD } } + +#define MLXSW_I2C_WRITE_MSG(_client, _buf, _len) \ + { .addr = (_client)->addr, \ + .buf = (u8 *)(_buf), \ + .len = (_len), \ + .flags = 0 } + +/* Routine converts in and out mail boxes offset and size. */ +static inline void +mlxsw_i2c_convert_mbox(struct mlxsw_i2c *mlxsw_i2c, u8 *buf) +{ + u32 tmp; + + /* Local in/out mailboxes: 20 bits for offset, 12 for size */ + tmp = be32_to_cpup((__be32 *) buf); + mlxsw_i2c->cmd.mb_off_in = tmp & + GENMASK(MLXSW_I2C_MBOX_OFFSET_BITS - 1, 0); + mlxsw_i2c->cmd.mb_size_in = (tmp & GENMASK(31, + MLXSW_I2C_MBOX_OFFSET_BITS)) >> + MLXSW_I2C_MBOX_OFFSET_BITS; + + tmp = be32_to_cpup((__be32 *) (buf + MLXSW_I2C_ADDR_WIDTH)); + mlxsw_i2c->cmd.mb_off_out = tmp & + GENMASK(MLXSW_I2C_MBOX_OFFSET_BITS - 1, 0); + mlxsw_i2c->cmd.mb_size_out = (tmp & GENMASK(31, + MLXSW_I2C_MBOX_OFFSET_BITS)) >> + MLXSW_I2C_MBOX_OFFSET_BITS; +} + +/* Routine obtains register size from mail box buffer. */ +static inline int mlxsw_i2c_get_reg_size(u8 *in_mbox) +{ + u16 tmp = be16_to_cpup((__be16 *) (in_mbox + MLXSW_I2C_TLV_HDR_SIZE)); + + return (tmp & 0x7ff) * 4 + MLXSW_I2C_TLV_HDR_SIZE; +} + +/* Routine sets I2C device internal offset in the transaction buffer. */ +static inline void mlxsw_i2c_set_slave_addr(u8 *buf, u32 off) +{ + __be32 *val = (__be32 *) buf; + + *val = htonl(off); +} + +/* Routine waits until go bit is cleared. */ +static int mlxsw_i2c_wait_go_bit(struct i2c_client *client, + struct mlxsw_i2c *mlxsw_i2c, u8 *p_status) +{ + u8 addr_buf[MLXSW_I2C_ADDR_BUF_SIZE]; + u8 buf[MLXSW_I2C_READ_SEMA_SIZE]; + int len = MLXSW_I2C_READ_SEMA_SIZE; + struct i2c_msg read_sema[] = + MLXSW_I2C_READ_MSG(client, addr_buf, buf, len); + bool wait_done = false; + unsigned long end; + int i = 0, err; + + mlxsw_i2c_set_slave_addr(addr_buf, MLXSW_I2C_CIR2_OFF_STATUS); + + end = jiffies + msecs_to_jiffies(MLXSW_I2C_TIMEOUT_MSECS); + do { + u32 ctrl; + + err = i2c_transfer(client->adapter, read_sema, + ARRAY_SIZE(read_sema)); + + ctrl = be32_to_cpu(*(__be32 *) buf); + if (err == ARRAY_SIZE(read_sema)) { + if (!(ctrl & MLXSW_I2C_GO_BIT)) { + wait_done = true; + *p_status = ctrl >> + MLXSW_I2C_CIR_CTRL_STATUS_SHIFT; + break; + } + } + cond_resched(); + } while ((time_before(jiffies, end)) || (i++ < MLXSW_I2C_RETRY)); + + if (wait_done) { + if (*p_status) + err = -EIO; + } else { + return -ETIMEDOUT; + } + + return err > 0 ? 0 : err; +} + +/* Routine posts a command to ASIC through mail box. */ +static int mlxsw_i2c_write_cmd(struct i2c_client *client, + struct mlxsw_i2c *mlxsw_i2c, + int immediate) +{ + __be32 push_cmd_buf[MLXSW_I2C_PUSH_CMD_SIZE / 4] = { + 0, cpu_to_be32(MLXSW_I2C_PUSH_IMM_CMD) + }; + __be32 prep_cmd_buf[MLXSW_I2C_PREP_SIZE / 4] = { + 0, 0, 0, 0, 0, 0, + cpu_to_be32(client->adapter->nr & 0xffff), + cpu_to_be32(MLXSW_I2C_SET_IMM_CMD) + }; + struct i2c_msg push_cmd = + MLXSW_I2C_WRITE_MSG(client, push_cmd_buf, + MLXSW_I2C_PUSH_CMD_SIZE); + struct i2c_msg prep_cmd = + MLXSW_I2C_WRITE_MSG(client, prep_cmd_buf, MLXSW_I2C_PREP_SIZE); + int err; + + if (!immediate) { + push_cmd_buf[1] = cpu_to_be32(MLXSW_I2C_PUSH_CMD); + prep_cmd_buf[7] = cpu_to_be32(MLXSW_I2C_SET_CMD); + } + mlxsw_i2c_set_slave_addr((u8 *)prep_cmd_buf, + MLXSW_I2C_CIR2_BASE); + mlxsw_i2c_set_slave_addr((u8 *)push_cmd_buf, + MLXSW_I2C_CIR2_OFF_STATUS); + + /* Prepare Command Interface Register for transaction */ + err = i2c_transfer(client->adapter, &prep_cmd, 1); + if (err < 0) + return err; + else if (err != 1) + return -EIO; + + /* Write out Command Interface Register GO bit to push transaction */ + err = i2c_transfer(client->adapter, &push_cmd, 1); + if (err < 0) + return err; + else if (err != 1) + return -EIO; + + return 0; +} + +/* Routine posts initialization command to ASIC through mail box. */ +static int +mlxsw_i2c_write_init_cmd(struct i2c_client *client, + struct mlxsw_i2c *mlxsw_i2c, u16 opcode, u32 in_mod) +{ + __be32 push_cmd_buf[MLXSW_I2C_PUSH_CMD_SIZE / 4] = { + 0, cpu_to_be32(MLXSW_I2C_PUSH_EVENT_CMD) + }; + __be32 prep_cmd_buf[MLXSW_I2C_PREP_SIZE / 4] = { + 0, 0, 0, 0, 0, 0, + cpu_to_be32(client->adapter->nr & 0xffff), + cpu_to_be32(MLXSW_I2C_SET_EVENT_CMD) + }; + struct i2c_msg push_cmd = + MLXSW_I2C_WRITE_MSG(client, push_cmd_buf, + MLXSW_I2C_PUSH_CMD_SIZE); + struct i2c_msg prep_cmd = + MLXSW_I2C_WRITE_MSG(client, prep_cmd_buf, MLXSW_I2C_PREP_SIZE); + u8 status; + int err; + + push_cmd_buf[1] = cpu_to_be32(MLXSW_I2C_PUSH_EVENT_CMD | opcode); + prep_cmd_buf[3] = cpu_to_be32(in_mod); + prep_cmd_buf[7] = cpu_to_be32(MLXSW_I2C_GO_BIT | opcode); + mlxsw_i2c_set_slave_addr((u8 *)prep_cmd_buf, + MLXSW_I2C_CIR2_BASE); + mlxsw_i2c_set_slave_addr((u8 *)push_cmd_buf, + MLXSW_I2C_CIR2_OFF_STATUS); + + /* Prepare Command Interface Register for transaction */ + err = i2c_transfer(client->adapter, &prep_cmd, 1); + if (err < 0) + return err; + else if (err != 1) + return -EIO; + + /* Write out Command Interface Register GO bit to push transaction */ + err = i2c_transfer(client->adapter, &push_cmd, 1); + if (err < 0) + return err; + else if (err != 1) + return -EIO; + + /* Wait until go bit is cleared. */ + err = mlxsw_i2c_wait_go_bit(client, mlxsw_i2c, &status); + if (err) { + dev_err(&client->dev, "HW semaphore is not released"); + return err; + } + + /* Validate transaction completion status. */ + if (status) { + dev_err(&client->dev, "Bad transaction completion status %x\n", + status); + return -EIO; + } + + return 0; +} + +/* Routine obtains mail box offsets from ASIC register space. */ +static int mlxsw_i2c_get_mbox(struct i2c_client *client, + struct mlxsw_i2c *mlxsw_i2c) +{ + u8 addr_buf[MLXSW_I2C_ADDR_BUF_SIZE]; + u8 buf[MLXSW_I2C_MBOX_SIZE]; + struct i2c_msg mbox_cmd[] = + MLXSW_I2C_READ_MSG(client, addr_buf, buf, MLXSW_I2C_MBOX_SIZE); + int err; + + /* Read mail boxes offsets. */ + mlxsw_i2c_set_slave_addr(addr_buf, MLXSW_I2C_CIR2_BASE); + err = i2c_transfer(client->adapter, mbox_cmd, 2); + if (err != 2) { + dev_err(&client->dev, "Could not obtain mail boxes\n"); + if (!err) + return -EIO; + else + return err; + } + + /* Convert mail boxes. */ + mlxsw_i2c_convert_mbox(mlxsw_i2c, &buf[MLXSW_I2C_MBOX_OUT_PARAM_OFF]); + + return err; +} + +/* Routine sends I2C write transaction to ASIC device. */ +static int +mlxsw_i2c_write(struct device *dev, size_t in_mbox_size, u8 *in_mbox, int num, + u8 *p_status) +{ + struct i2c_client *client = to_i2c_client(dev); + struct mlxsw_i2c *mlxsw_i2c = i2c_get_clientdata(client); + unsigned long timeout = msecs_to_jiffies(MLXSW_I2C_TIMEOUT_MSECS); + int off = mlxsw_i2c->cmd.mb_off_in, chunk_size, i, j; + unsigned long end; + u8 *tran_buf; + struct i2c_msg write_tran = + MLXSW_I2C_WRITE_MSG(client, NULL, MLXSW_I2C_PUSH_CMD_SIZE); + int err; + + tran_buf = kmalloc(mlxsw_i2c->block_size + MLXSW_I2C_ADDR_BUF_SIZE, + GFP_KERNEL); + if (!tran_buf) + return -ENOMEM; + + write_tran.buf = tran_buf; + for (i = 0; i < num; i++) { + chunk_size = (in_mbox_size > mlxsw_i2c->block_size) ? + mlxsw_i2c->block_size : in_mbox_size; + write_tran.len = MLXSW_I2C_ADDR_WIDTH + chunk_size; + mlxsw_i2c_set_slave_addr(tran_buf, off); + memcpy(&tran_buf[MLXSW_I2C_ADDR_BUF_SIZE], in_mbox + + mlxsw_i2c->block_size * i, chunk_size); + + j = 0; + end = jiffies + timeout; + do { + err = i2c_transfer(client->adapter, &write_tran, 1); + if (err == 1) + break; + + cond_resched(); + } while ((time_before(jiffies, end)) || + (j++ < MLXSW_I2C_RETRY)); + + if (err != 1) { + if (!err) { + err = -EIO; + goto mlxsw_i2c_write_exit; + } + } + + off += chunk_size; + in_mbox_size -= chunk_size; + } + + /* Prepare and write out Command Interface Register for transaction. */ + err = mlxsw_i2c_write_cmd(client, mlxsw_i2c, 0); + if (err) { + dev_err(&client->dev, "Could not start transaction"); + err = -EIO; + goto mlxsw_i2c_write_exit; + } + + /* Wait until go bit is cleared. */ + err = mlxsw_i2c_wait_go_bit(client, mlxsw_i2c, p_status); + if (err) { + dev_err(&client->dev, "HW semaphore is not released"); + goto mlxsw_i2c_write_exit; + } + + /* Validate transaction completion status. */ + if (*p_status) { + dev_err(&client->dev, "Bad transaction completion status %x\n", + *p_status); + err = -EIO; + } + +mlxsw_i2c_write_exit: + kfree(tran_buf); + return err; +} + +/* Routine executes I2C command. */ +static int +mlxsw_i2c_cmd(struct device *dev, u16 opcode, u32 in_mod, size_t in_mbox_size, + u8 *in_mbox, size_t out_mbox_size, u8 *out_mbox, u8 *status) +{ + struct i2c_client *client = to_i2c_client(dev); + struct mlxsw_i2c *mlxsw_i2c = i2c_get_clientdata(client); + unsigned long timeout = msecs_to_jiffies(MLXSW_I2C_TIMEOUT_MSECS); + u8 tran_buf[MLXSW_I2C_ADDR_BUF_SIZE]; + int num, chunk_size, reg_size, i, j; + int off = mlxsw_i2c->cmd.mb_off_out; + unsigned long end; + struct i2c_msg read_tran[] = + MLXSW_I2C_READ_MSG(client, tran_buf, NULL, 0); + int err; + + WARN_ON(in_mbox_size % sizeof(u32) || out_mbox_size % sizeof(u32)); + + if (in_mbox) { + reg_size = mlxsw_i2c_get_reg_size(in_mbox); + num = reg_size / mlxsw_i2c->block_size; + if (reg_size % mlxsw_i2c->block_size) + num++; + + if (mutex_lock_interruptible(&mlxsw_i2c->cmd.lock) < 0) { + dev_err(&client->dev, "Could not acquire lock"); + return -EINVAL; + } + + err = mlxsw_i2c_write(dev, reg_size, in_mbox, num, status); + if (err) + goto cmd_fail; + + /* No out mailbox is case of write transaction. */ + if (!out_mbox) { + mutex_unlock(&mlxsw_i2c->cmd.lock); + return 0; + } + } else { + /* No input mailbox is case of initialization query command. */ + reg_size = MLXSW_I2C_MAX_DATA_SIZE; + num = DIV_ROUND_UP(reg_size, mlxsw_i2c->block_size); + + if (mutex_lock_interruptible(&mlxsw_i2c->cmd.lock) < 0) { + dev_err(&client->dev, "Could not acquire lock"); + return -EINVAL; + } + + err = mlxsw_i2c_write_init_cmd(client, mlxsw_i2c, opcode, + in_mod); + if (err) + goto cmd_fail; + } + + /* Send read transaction to get output mailbox content. */ + read_tran[1].buf = out_mbox; + for (i = 0; i < num; i++) { + chunk_size = (reg_size > mlxsw_i2c->block_size) ? + mlxsw_i2c->block_size : reg_size; + read_tran[1].len = chunk_size; + mlxsw_i2c_set_slave_addr(tran_buf, off); + + j = 0; + end = jiffies + timeout; + do { + err = i2c_transfer(client->adapter, read_tran, + ARRAY_SIZE(read_tran)); + if (err == ARRAY_SIZE(read_tran)) + break; + + cond_resched(); + } while ((time_before(jiffies, end)) || + (j++ < MLXSW_I2C_RETRY)); + + if (err != ARRAY_SIZE(read_tran)) { + if (!err) + err = -EIO; + + goto cmd_fail; + } + + off += chunk_size; + reg_size -= chunk_size; + read_tran[1].buf += chunk_size; + } + + mutex_unlock(&mlxsw_i2c->cmd.lock); + + return 0; + +cmd_fail: + mutex_unlock(&mlxsw_i2c->cmd.lock); + return err; +} + +static int mlxsw_i2c_cmd_exec(void *bus_priv, u16 opcode, u8 opcode_mod, + u32 in_mod, bool out_mbox_direct, + char *in_mbox, size_t in_mbox_size, + char *out_mbox, size_t out_mbox_size, + u8 *status) +{ + struct mlxsw_i2c *mlxsw_i2c = bus_priv; + + return mlxsw_i2c_cmd(mlxsw_i2c->dev, opcode, in_mod, in_mbox_size, + in_mbox, out_mbox_size, out_mbox, status); +} + +static bool mlxsw_i2c_skb_transmit_busy(void *bus_priv, + const struct mlxsw_tx_info *tx_info) +{ + return false; +} + +static int mlxsw_i2c_skb_transmit(void *bus_priv, struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info) +{ + return 0; +} + +static int +mlxsw_i2c_init(void *bus_priv, struct mlxsw_core *mlxsw_core, + const struct mlxsw_config_profile *profile, + struct mlxsw_res *res) +{ + struct mlxsw_i2c *mlxsw_i2c = bus_priv; + char *mbox; + int err; + + mlxsw_i2c->core = mlxsw_core; + + mbox = mlxsw_cmd_mbox_alloc(); + if (!mbox) + return -ENOMEM; + + err = mlxsw_cmd_query_fw(mlxsw_core, mbox); + if (err) + goto mbox_put; + + mlxsw_i2c->bus_info.fw_rev.major = + mlxsw_cmd_mbox_query_fw_fw_rev_major_get(mbox); + mlxsw_i2c->bus_info.fw_rev.minor = + mlxsw_cmd_mbox_query_fw_fw_rev_minor_get(mbox); + mlxsw_i2c->bus_info.fw_rev.subminor = + mlxsw_cmd_mbox_query_fw_fw_rev_subminor_get(mbox); + + err = mlxsw_core_resources_query(mlxsw_core, mbox, res); + +mbox_put: + mlxsw_cmd_mbox_free(mbox); + return err; +} + +static void mlxsw_i2c_fini(void *bus_priv) +{ + struct mlxsw_i2c *mlxsw_i2c = bus_priv; + + mlxsw_i2c->core = NULL; +} + +static void mlxsw_i2c_work_handler(struct work_struct *work) +{ + struct mlxsw_i2c *mlxsw_i2c; + + mlxsw_i2c = container_of(work, struct mlxsw_i2c, irq_work); + mlxsw_core_irq_event_handlers_call(mlxsw_i2c->core); +} + +static irqreturn_t mlxsw_i2c_irq_handler(int irq, void *dev) +{ + struct mlxsw_i2c *mlxsw_i2c = dev; + + mlxsw_core_schedule_work(&mlxsw_i2c->irq_work); + + /* Interrupt handler shares IRQ line with 'main' interrupt handler. + * Return here IRQ_NONE, while main handler will return IRQ_HANDLED. + */ + return IRQ_NONE; +} + +static int mlxsw_i2c_irq_init(struct mlxsw_i2c *mlxsw_i2c, u8 addr) +{ + int err; + + /* Initialize interrupt handler if system hotplug driver is reachable, + * otherwise interrupt line is not enabled and interrupts will not be + * raised to CPU. Also request_irq() call will be not valid. + */ + if (!IS_REACHABLE(CONFIG_MLXREG_HOTPLUG)) + return 0; + + /* Set default interrupt line. */ + if (mlxsw_i2c->pdata && mlxsw_i2c->pdata->irq) + mlxsw_i2c->irq = mlxsw_i2c->pdata->irq; + else if (addr == MLXSW_FAST_I2C_SLAVE) + mlxsw_i2c->irq = MLXSW_I2C_DEFAULT_IRQ; + + if (!mlxsw_i2c->irq) + return 0; + + INIT_WORK(&mlxsw_i2c->irq_work, mlxsw_i2c_work_handler); + err = request_irq(mlxsw_i2c->irq, mlxsw_i2c_irq_handler, + IRQF_TRIGGER_FALLING | IRQF_SHARED, "mlxsw-i2c", + mlxsw_i2c); + if (err) { + dev_err(mlxsw_i2c->bus_info.dev, "Failed to request irq: %d\n", + err); + return err; + } + + return 0; +} + +static void mlxsw_i2c_irq_fini(struct mlxsw_i2c *mlxsw_i2c) +{ + if (!IS_REACHABLE(CONFIG_MLXREG_HOTPLUG) || !mlxsw_i2c->irq) + return; + cancel_work_sync(&mlxsw_i2c->irq_work); + free_irq(mlxsw_i2c->irq, mlxsw_i2c); +} + +static const struct mlxsw_bus mlxsw_i2c_bus = { + .kind = "i2c", + .init = mlxsw_i2c_init, + .fini = mlxsw_i2c_fini, + .skb_transmit_busy = mlxsw_i2c_skb_transmit_busy, + .skb_transmit = mlxsw_i2c_skb_transmit, + .cmd_exec = mlxsw_i2c_cmd_exec, +}; + +static int mlxsw_i2c_probe(struct i2c_client *client, + const struct i2c_device_id *id) +{ + const struct i2c_adapter_quirks *quirks = client->adapter->quirks; + struct mlxsw_i2c *mlxsw_i2c; + u8 status; + int err; + + mlxsw_i2c = devm_kzalloc(&client->dev, sizeof(*mlxsw_i2c), GFP_KERNEL); + if (!mlxsw_i2c) + return -ENOMEM; + + if (quirks) { + if ((quirks->max_read_len && + quirks->max_read_len < MLXSW_I2C_BLK_DEF) || + (quirks->max_write_len && + quirks->max_write_len < MLXSW_I2C_BLK_DEF)) { + dev_err(&client->dev, "Insufficient transaction buffer length\n"); + return -EOPNOTSUPP; + } + + mlxsw_i2c->block_size = min_t(u16, MLXSW_I2C_BLK_MAX, + min_t(u16, quirks->max_read_len, + quirks->max_write_len)); + } else { + mlxsw_i2c->block_size = MLXSW_I2C_BLK_DEF; + } + + i2c_set_clientdata(client, mlxsw_i2c); + mutex_init(&mlxsw_i2c->cmd.lock); + + /* In order to use mailboxes through the i2c, special area is reserved + * on the i2c address space that can be used for input and output + * mailboxes. Such mailboxes are called local mailboxes. When using a + * local mailbox, software should specify 0 as the Input/Output + * parameters. The location of the Local Mailbox addresses on the i2c + * space can be retrieved through the QUERY_FW command. + * For this purpose QUERY_FW is to be issued with opcode modifier equal + * 0x01. For such command the output parameter is an immediate value. + * Here QUERY_FW command is invoked for ASIC probing and for getting + * local mailboxes addresses from immedate output parameters. + */ + + /* Prepare and write out Command Interface Register for transaction */ + err = mlxsw_i2c_write_cmd(client, mlxsw_i2c, 1); + if (err) { + dev_err(&client->dev, "Could not start transaction"); + goto errout; + } + + /* Wait until go bit is cleared. */ + err = mlxsw_i2c_wait_go_bit(client, mlxsw_i2c, &status); + if (err) { + dev_err(&client->dev, "HW semaphore is not released"); + goto errout; + } + + /* Validate transaction completion status. */ + if (status) { + dev_err(&client->dev, "Bad transaction completion status %x\n", + status); + err = -EIO; + goto errout; + } + + /* Get mailbox offsets. */ + err = mlxsw_i2c_get_mbox(client, mlxsw_i2c); + if (err < 0) { + dev_err(&client->dev, "Fail to get mailboxes\n"); + goto errout; + } + + dev_info(&client->dev, "%s mb size=%x off=0x%08x out mb size=%x off=0x%08x\n", + id->name, mlxsw_i2c->cmd.mb_size_in, + mlxsw_i2c->cmd.mb_off_in, mlxsw_i2c->cmd.mb_size_out, + mlxsw_i2c->cmd.mb_off_out); + + /* Register device bus. */ + mlxsw_i2c->bus_info.device_kind = id->name; + mlxsw_i2c->bus_info.device_name = client->name; + mlxsw_i2c->bus_info.dev = &client->dev; + mlxsw_i2c->bus_info.low_frequency = true; + mlxsw_i2c->dev = &client->dev; + mlxsw_i2c->pdata = client->dev.platform_data; + + err = mlxsw_i2c_irq_init(mlxsw_i2c, client->addr); + if (err) + goto errout; + + err = mlxsw_core_bus_device_register(&mlxsw_i2c->bus_info, + &mlxsw_i2c_bus, mlxsw_i2c, false, + NULL, NULL); + if (err) { + dev_err(&client->dev, "Fail to register core bus\n"); + goto err_bus_device_register; + } + + return 0; + +err_bus_device_register: + mlxsw_i2c_irq_fini(mlxsw_i2c); +errout: + mutex_destroy(&mlxsw_i2c->cmd.lock); + i2c_set_clientdata(client, NULL); + + return err; +} + +static void mlxsw_i2c_remove(struct i2c_client *client) +{ + struct mlxsw_i2c *mlxsw_i2c = i2c_get_clientdata(client); + + mlxsw_core_bus_device_unregister(mlxsw_i2c->core, false); + mlxsw_i2c_irq_fini(mlxsw_i2c); + mutex_destroy(&mlxsw_i2c->cmd.lock); +} + +int mlxsw_i2c_driver_register(struct i2c_driver *i2c_driver) +{ + i2c_driver->probe = mlxsw_i2c_probe; + i2c_driver->remove = mlxsw_i2c_remove; + return i2c_add_driver(i2c_driver); +} +EXPORT_SYMBOL(mlxsw_i2c_driver_register); + +void mlxsw_i2c_driver_unregister(struct i2c_driver *i2c_driver) +{ + i2c_del_driver(i2c_driver); +} +EXPORT_SYMBOL(mlxsw_i2c_driver_unregister); + +MODULE_AUTHOR("Vadim Pasternak <vadimp@mellanox.com>"); +MODULE_DESCRIPTION("Mellanox switch I2C interface driver"); +MODULE_LICENSE("Dual BSD/GPL"); diff --git a/drivers/net/ethernet/mellanox/mlxsw/i2c.h b/drivers/net/ethernet/mellanox/mlxsw/i2c.h new file mode 100644 index 000000000..17e059d47 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/i2c.h @@ -0,0 +1,29 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2016-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_I2C_H +#define _MLXSW_I2C_H + +#include <linux/i2c.h> + +#if IS_ENABLED(CONFIG_MLXSW_I2C) + +int mlxsw_i2c_driver_register(struct i2c_driver *i2c_driver); +void mlxsw_i2c_driver_unregister(struct i2c_driver *i2c_driver); + +#else + +static inline int +mlxsw_i2c_driver_register(struct i2c_driver *i2c_driver) +{ + return -ENODEV; +} + +static inline void +mlxsw_i2c_driver_unregister(struct i2c_driver *i2c_driver) +{ +} + +#endif + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/item.h b/drivers/net/ethernet/mellanox/mlxsw/item.h new file mode 100644 index 000000000..cfafbeb42 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/item.h @@ -0,0 +1,553 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_ITEM_H +#define _MLXSW_ITEM_H + +#include <linux/types.h> +#include <linux/string.h> +#include <linux/bitops.h> + +struct mlxsw_item { + unsigned short offset; /* bytes in container */ + short step; /* step in bytes for indexed items */ + unsigned short in_step_offset; /* offset within one step */ + unsigned char shift; /* shift in bits */ + unsigned char element_size; /* size of element in bit array */ + bool no_real_shift; + union { + unsigned char bits; + unsigned short bytes; + } size; + const char *name; +}; + +static inline unsigned int +__mlxsw_item_offset(const struct mlxsw_item *item, unsigned short index, + size_t typesize) +{ + BUG_ON(index && !item->step); + if (item->offset % typesize != 0 || + item->step % typesize != 0 || + item->in_step_offset % typesize != 0) { + pr_err("mlxsw: item bug (name=%s,offset=%x,step=%x,in_step_offset=%x,typesize=%zx)\n", + item->name, item->offset, item->step, + item->in_step_offset, typesize); + BUG(); + } + + return ((item->offset + item->step * index + item->in_step_offset) / + typesize); +} + +static inline u8 __mlxsw_item_get8(const char *buf, + const struct mlxsw_item *item, + unsigned short index) +{ + unsigned int offset = __mlxsw_item_offset(item, index, sizeof(u8)); + u8 *b = (u8 *) buf; + u8 tmp; + + tmp = b[offset]; + tmp >>= item->shift; + tmp &= GENMASK(item->size.bits - 1, 0); + if (item->no_real_shift) + tmp <<= item->shift; + return tmp; +} + +static inline void __mlxsw_item_set8(char *buf, const struct mlxsw_item *item, + unsigned short index, u8 val) +{ + unsigned int offset = __mlxsw_item_offset(item, index, + sizeof(u8)); + u8 *b = (u8 *) buf; + u8 mask = GENMASK(item->size.bits - 1, 0) << item->shift; + u8 tmp; + + if (!item->no_real_shift) + val <<= item->shift; + val &= mask; + tmp = b[offset]; + tmp &= ~mask; + tmp |= val; + b[offset] = tmp; +} + +static inline u16 __mlxsw_item_get16(const char *buf, + const struct mlxsw_item *item, + unsigned short index) +{ + unsigned int offset = __mlxsw_item_offset(item, index, sizeof(u16)); + __be16 *b = (__be16 *) buf; + u16 tmp; + + tmp = be16_to_cpu(b[offset]); + tmp >>= item->shift; + tmp &= GENMASK(item->size.bits - 1, 0); + if (item->no_real_shift) + tmp <<= item->shift; + return tmp; +} + +static inline void __mlxsw_item_set16(char *buf, const struct mlxsw_item *item, + unsigned short index, u16 val) +{ + unsigned int offset = __mlxsw_item_offset(item, index, + sizeof(u16)); + __be16 *b = (__be16 *) buf; + u16 mask = GENMASK(item->size.bits - 1, 0) << item->shift; + u16 tmp; + + if (!item->no_real_shift) + val <<= item->shift; + val &= mask; + tmp = be16_to_cpu(b[offset]); + tmp &= ~mask; + tmp |= val; + b[offset] = cpu_to_be16(tmp); +} + +static inline u32 __mlxsw_item_get32(const char *buf, + const struct mlxsw_item *item, + unsigned short index) +{ + unsigned int offset = __mlxsw_item_offset(item, index, sizeof(u32)); + __be32 *b = (__be32 *) buf; + u32 tmp; + + tmp = be32_to_cpu(b[offset]); + tmp >>= item->shift; + tmp &= GENMASK(item->size.bits - 1, 0); + if (item->no_real_shift) + tmp <<= item->shift; + return tmp; +} + +static inline void __mlxsw_item_set32(char *buf, const struct mlxsw_item *item, + unsigned short index, u32 val) +{ + unsigned int offset = __mlxsw_item_offset(item, index, + sizeof(u32)); + __be32 *b = (__be32 *) buf; + u32 mask = GENMASK(item->size.bits - 1, 0) << item->shift; + u32 tmp; + + if (!item->no_real_shift) + val <<= item->shift; + val &= mask; + tmp = be32_to_cpu(b[offset]); + tmp &= ~mask; + tmp |= val; + b[offset] = cpu_to_be32(tmp); +} + +static inline u64 __mlxsw_item_get64(const char *buf, + const struct mlxsw_item *item, + unsigned short index) +{ + unsigned int offset = __mlxsw_item_offset(item, index, sizeof(u64)); + __be64 *b = (__be64 *) buf; + u64 tmp; + + tmp = be64_to_cpu(b[offset]); + tmp >>= item->shift; + tmp &= GENMASK_ULL(item->size.bits - 1, 0); + if (item->no_real_shift) + tmp <<= item->shift; + return tmp; +} + +static inline void __mlxsw_item_set64(char *buf, const struct mlxsw_item *item, + unsigned short index, u64 val) +{ + unsigned int offset = __mlxsw_item_offset(item, index, sizeof(u64)); + __be64 *b = (__be64 *) buf; + u64 mask = GENMASK_ULL(item->size.bits - 1, 0) << item->shift; + u64 tmp; + + if (!item->no_real_shift) + val <<= item->shift; + val &= mask; + tmp = be64_to_cpu(b[offset]); + tmp &= ~mask; + tmp |= val; + b[offset] = cpu_to_be64(tmp); +} + +static inline void __mlxsw_item_memcpy_from(const char *buf, char *dst, + const struct mlxsw_item *item, + unsigned short index) +{ + unsigned int offset = __mlxsw_item_offset(item, index, sizeof(char)); + + memcpy(dst, &buf[offset], item->size.bytes); +} + +static inline void __mlxsw_item_memcpy_to(char *buf, const char *src, + const struct mlxsw_item *item, + unsigned short index) +{ + unsigned int offset = __mlxsw_item_offset(item, index, sizeof(char)); + + memcpy(&buf[offset], src, item->size.bytes); +} + +static inline char *__mlxsw_item_data(char *buf, const struct mlxsw_item *item, + unsigned short index) +{ + unsigned int offset = __mlxsw_item_offset(item, index, sizeof(char)); + + return &buf[offset]; +} + +static inline u16 +__mlxsw_item_bit_array_offset(const struct mlxsw_item *item, + u16 index, u8 *shift) +{ + u16 max_index, be_index; + u16 offset; /* byte offset inside the array */ + u8 in_byte_index; + + BUG_ON(index && !item->element_size); + if (item->offset % sizeof(u32) != 0 || + BITS_PER_BYTE % item->element_size != 0) { + pr_err("mlxsw: item bug (name=%s,offset=%x,element_size=%x)\n", + item->name, item->offset, item->element_size); + BUG(); + } + + max_index = (item->size.bytes << 3) / item->element_size - 1; + be_index = max_index - index; + offset = be_index * item->element_size >> 3; + in_byte_index = index % (BITS_PER_BYTE / item->element_size); + *shift = in_byte_index * item->element_size; + + return item->offset + offset; +} + +static inline u8 __mlxsw_item_bit_array_get(const char *buf, + const struct mlxsw_item *item, + u16 index) +{ + u8 shift, tmp; + u16 offset = __mlxsw_item_bit_array_offset(item, index, &shift); + + tmp = buf[offset]; + tmp >>= shift; + tmp &= GENMASK(item->element_size - 1, 0); + return tmp; +} + +static inline void __mlxsw_item_bit_array_set(char *buf, + const struct mlxsw_item *item, + u16 index, u8 val) +{ + u8 shift, tmp; + u16 offset = __mlxsw_item_bit_array_offset(item, index, &shift); + u8 mask = GENMASK(item->element_size - 1, 0) << shift; + + val <<= shift; + val &= mask; + tmp = buf[offset]; + tmp &= ~mask; + tmp |= val; + buf[offset] = tmp; +} + +#define __ITEM_NAME(_type, _cname, _iname) \ + mlxsw_##_type##_##_cname##_##_iname##_item + +/* _type: cmd_mbox, reg, etc. + * _cname: containter name (e.g. command name, register name) + * _iname: item name within the container + */ + +#define MLXSW_ITEM8(_type, _cname, _iname, _offset, _shift, _sizebits) \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, _iname) = { \ + .offset = _offset, \ + .shift = _shift, \ + .size = {.bits = _sizebits,}, \ + .name = #_type "_" #_cname "_" #_iname, \ +}; \ +static inline u8 __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_get(const char *buf) \ +{ \ + return __mlxsw_item_get8(buf, &__ITEM_NAME(_type, _cname, _iname), 0); \ +} \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_set(char *buf, u8 val) \ +{ \ + __mlxsw_item_set8(buf, &__ITEM_NAME(_type, _cname, _iname), 0, val); \ +} + +#define MLXSW_ITEM8_INDEXED(_type, _cname, _iname, _offset, _shift, _sizebits, \ + _step, _instepoffset, _norealshift) \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, _iname) = { \ + .offset = _offset, \ + .step = _step, \ + .in_step_offset = _instepoffset, \ + .shift = _shift, \ + .no_real_shift = _norealshift, \ + .size = {.bits = _sizebits,}, \ + .name = #_type "_" #_cname "_" #_iname, \ +}; \ +static inline u8 __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_get(const char *buf, unsigned short index)\ +{ \ + return __mlxsw_item_get8(buf, &__ITEM_NAME(_type, _cname, _iname), \ + index); \ +} \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_set(char *buf, unsigned short index, \ + u8 val) \ +{ \ + __mlxsw_item_set8(buf, &__ITEM_NAME(_type, _cname, _iname), \ + index, val); \ +} + +#define MLXSW_ITEM16(_type, _cname, _iname, _offset, _shift, _sizebits) \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, _iname) = { \ + .offset = _offset, \ + .shift = _shift, \ + .size = {.bits = _sizebits,}, \ + .name = #_type "_" #_cname "_" #_iname, \ +}; \ +static inline u16 __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_get(const char *buf) \ +{ \ + return __mlxsw_item_get16(buf, &__ITEM_NAME(_type, _cname, _iname), 0); \ +} \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_set(char *buf, u16 val) \ +{ \ + __mlxsw_item_set16(buf, &__ITEM_NAME(_type, _cname, _iname), 0, val); \ +} + +#define MLXSW_ITEM16_INDEXED(_type, _cname, _iname, _offset, _shift, _sizebits, \ + _step, _instepoffset, _norealshift) \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, _iname) = { \ + .offset = _offset, \ + .step = _step, \ + .in_step_offset = _instepoffset, \ + .shift = _shift, \ + .no_real_shift = _norealshift, \ + .size = {.bits = _sizebits,}, \ + .name = #_type "_" #_cname "_" #_iname, \ +}; \ +static inline u16 __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_get(const char *buf, unsigned short index)\ +{ \ + return __mlxsw_item_get16(buf, &__ITEM_NAME(_type, _cname, _iname), \ + index); \ +} \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_set(char *buf, unsigned short index, \ + u16 val) \ +{ \ + __mlxsw_item_set16(buf, &__ITEM_NAME(_type, _cname, _iname), \ + index, val); \ +} + +#define MLXSW_ITEM32(_type, _cname, _iname, _offset, _shift, _sizebits) \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, _iname) = { \ + .offset = _offset, \ + .shift = _shift, \ + .size = {.bits = _sizebits,}, \ + .name = #_type "_" #_cname "_" #_iname, \ +}; \ +static inline u32 __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_get(const char *buf) \ +{ \ + return __mlxsw_item_get32(buf, &__ITEM_NAME(_type, _cname, _iname), 0); \ +} \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_set(char *buf, u32 val) \ +{ \ + __mlxsw_item_set32(buf, &__ITEM_NAME(_type, _cname, _iname), 0, val); \ +} + +#define LOCAL_PORT_LSB_SIZE 8 +#define LOCAL_PORT_MSB_SIZE 2 + +#define MLXSW_ITEM32_LP(_type, _cname, _offset1, _shift1, _offset2, _shift2) \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, local_port) = { \ + .offset = _offset1, \ + .shift = _shift1, \ + .size = {.bits = LOCAL_PORT_LSB_SIZE,}, \ + .name = #_type "_" #_cname "_local_port", \ +}; \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, lp_msb) = { \ + .offset = _offset2, \ + .shift = _shift2, \ + .size = {.bits = LOCAL_PORT_MSB_SIZE,}, \ + .name = #_type "_" #_cname "_lp_msb", \ +}; \ +static inline u32 __maybe_unused \ +mlxsw_##_type##_##_cname##_local_port_get(const char *buf) \ +{ \ + u32 local_port, lp_msb; \ + \ + local_port = __mlxsw_item_get32(buf, &__ITEM_NAME(_type, _cname, \ + local_port), 0); \ + lp_msb = __mlxsw_item_get32(buf, &__ITEM_NAME(_type, _cname, lp_msb), \ + 0); \ + return (lp_msb << LOCAL_PORT_LSB_SIZE) + local_port; \ +} \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_local_port_set(char *buf, u32 val) \ +{ \ + __mlxsw_item_set32(buf, &__ITEM_NAME(_type, _cname, local_port), 0, \ + val & ((1 << LOCAL_PORT_LSB_SIZE) - 1)); \ + __mlxsw_item_set32(buf, &__ITEM_NAME(_type, _cname, lp_msb), 0, \ + val >> LOCAL_PORT_LSB_SIZE); \ +} + +#define MLXSW_ITEM32_INDEXED(_type, _cname, _iname, _offset, _shift, _sizebits, \ + _step, _instepoffset, _norealshift) \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, _iname) = { \ + .offset = _offset, \ + .step = _step, \ + .in_step_offset = _instepoffset, \ + .shift = _shift, \ + .no_real_shift = _norealshift, \ + .size = {.bits = _sizebits,}, \ + .name = #_type "_" #_cname "_" #_iname, \ +}; \ +static inline u32 __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_get(const char *buf, unsigned short index)\ +{ \ + return __mlxsw_item_get32(buf, &__ITEM_NAME(_type, _cname, _iname), \ + index); \ +} \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_set(char *buf, unsigned short index, \ + u32 val) \ +{ \ + __mlxsw_item_set32(buf, &__ITEM_NAME(_type, _cname, _iname), \ + index, val); \ +} + +#define MLXSW_ITEM64(_type, _cname, _iname, _offset, _shift, _sizebits) \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, _iname) = { \ + .offset = _offset, \ + .shift = _shift, \ + .size = {.bits = _sizebits,}, \ + .name = #_type "_" #_cname "_" #_iname, \ +}; \ +static inline u64 __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_get(const char *buf) \ +{ \ + return __mlxsw_item_get64(buf, &__ITEM_NAME(_type, _cname, _iname), 0); \ +} \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_set(char *buf, u64 val) \ +{ \ + __mlxsw_item_set64(buf, &__ITEM_NAME(_type, _cname, _iname), 0, val); \ +} + +#define MLXSW_ITEM64_INDEXED(_type, _cname, _iname, _offset, _shift, \ + _sizebits, _step, _instepoffset, _norealshift) \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, _iname) = { \ + .offset = _offset, \ + .step = _step, \ + .in_step_offset = _instepoffset, \ + .shift = _shift, \ + .no_real_shift = _norealshift, \ + .size = {.bits = _sizebits,}, \ + .name = #_type "_" #_cname "_" #_iname, \ +}; \ +static inline u64 __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_get(const char *buf, unsigned short index)\ +{ \ + return __mlxsw_item_get64(buf, &__ITEM_NAME(_type, _cname, _iname), \ + index); \ +} \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_set(char *buf, unsigned short index, \ + u64 val) \ +{ \ + __mlxsw_item_set64(buf, &__ITEM_NAME(_type, _cname, _iname), \ + index, val); \ +} + +#define MLXSW_ITEM_BUF(_type, _cname, _iname, _offset, _sizebytes) \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, _iname) = { \ + .offset = _offset, \ + .size = {.bytes = _sizebytes,}, \ + .name = #_type "_" #_cname "_" #_iname, \ +}; \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_memcpy_from(const char *buf, char *dst) \ +{ \ + __mlxsw_item_memcpy_from(buf, dst, \ + &__ITEM_NAME(_type, _cname, _iname), 0); \ +} \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_memcpy_to(char *buf, const char *src) \ +{ \ + __mlxsw_item_memcpy_to(buf, src, \ + &__ITEM_NAME(_type, _cname, _iname), 0); \ +} \ +static inline char * __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_data(char *buf) \ +{ \ + return __mlxsw_item_data(buf, &__ITEM_NAME(_type, _cname, _iname), 0); \ +} + +#define MLXSW_ITEM_BUF_INDEXED(_type, _cname, _iname, _offset, _sizebytes, \ + _step, _instepoffset) \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, _iname) = { \ + .offset = _offset, \ + .step = _step, \ + .in_step_offset = _instepoffset, \ + .size = {.bytes = _sizebytes,}, \ + .name = #_type "_" #_cname "_" #_iname, \ +}; \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_memcpy_from(const char *buf, \ + unsigned short index, \ + char *dst) \ +{ \ + __mlxsw_item_memcpy_from(buf, dst, \ + &__ITEM_NAME(_type, _cname, _iname), index); \ +} \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_memcpy_to(char *buf, \ + unsigned short index, \ + const char *src) \ +{ \ + __mlxsw_item_memcpy_to(buf, src, \ + &__ITEM_NAME(_type, _cname, _iname), index); \ +} \ +static inline char * __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_data(char *buf, unsigned short index) \ +{ \ + return __mlxsw_item_data(buf, \ + &__ITEM_NAME(_type, _cname, _iname), index); \ +} + +#define MLXSW_ITEM_BIT_ARRAY(_type, _cname, _iname, _offset, _sizebytes, \ + _element_size) \ +static struct mlxsw_item __ITEM_NAME(_type, _cname, _iname) = { \ + .offset = _offset, \ + .element_size = _element_size, \ + .size = {.bytes = _sizebytes,}, \ + .name = #_type "_" #_cname "_" #_iname, \ +}; \ +static inline u8 __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_get(const char *buf, u16 index) \ +{ \ + return __mlxsw_item_bit_array_get(buf, \ + &__ITEM_NAME(_type, _cname, _iname), \ + index); \ +} \ +static inline void __maybe_unused \ +mlxsw_##_type##_##_cname##_##_iname##_set(char *buf, u16 index, u8 val) \ +{ \ + return __mlxsw_item_bit_array_set(buf, \ + &__ITEM_NAME(_type, _cname, _iname), \ + index, val); \ +} \ + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/minimal.c b/drivers/net/ethernet/mellanox/mlxsw/minimal.c new file mode 100644 index 000000000..15116d930 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/minimal.c @@ -0,0 +1,760 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2016-2019 Mellanox Technologies. All rights reserved */ + +#include <linux/netdevice.h> +#include <linux/etherdevice.h> +#include <linux/ethtool.h> +#include <linux/i2c.h> +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/mod_devicetable.h> +#include <linux/types.h> + +#include "core.h" +#include "core_env.h" +#include "i2c.h" + +static const char mlxsw_m_driver_name[] = "mlxsw_minimal"; + +#define MLXSW_M_FWREV_MINOR 2000 +#define MLXSW_M_FWREV_SUBMINOR 1886 + +static const struct mlxsw_fw_rev mlxsw_m_fw_rev = { + .minor = MLXSW_M_FWREV_MINOR, + .subminor = MLXSW_M_FWREV_SUBMINOR, +}; + +struct mlxsw_m_port; + +struct mlxsw_m_line_card { + bool active; + int module_to_port[]; +}; + +struct mlxsw_m { + struct mlxsw_m_port **ports; + struct mlxsw_core *core; + const struct mlxsw_bus_info *bus_info; + u8 base_mac[ETH_ALEN]; + u8 max_ports; + u8 max_modules_per_slot; /* Maximum number of modules per-slot. */ + u8 num_of_slots; /* Including the main board. */ + struct mlxsw_m_line_card **line_cards; +}; + +struct mlxsw_m_port { + struct net_device *dev; + struct mlxsw_m *mlxsw_m; + u16 local_port; + u8 slot_index; + u8 module; + u8 module_offset; +}; + +static int mlxsw_m_base_mac_get(struct mlxsw_m *mlxsw_m) +{ + char spad_pl[MLXSW_REG_SPAD_LEN] = {0}; + int err; + + err = mlxsw_reg_query(mlxsw_m->core, MLXSW_REG(spad), spad_pl); + if (err) + return err; + mlxsw_reg_spad_base_mac_memcpy_from(spad_pl, mlxsw_m->base_mac); + return 0; +} + +static int mlxsw_m_port_open(struct net_device *dev) +{ + struct mlxsw_m_port *mlxsw_m_port = netdev_priv(dev); + struct mlxsw_m *mlxsw_m = mlxsw_m_port->mlxsw_m; + + return mlxsw_env_module_port_up(mlxsw_m->core, 0, + mlxsw_m_port->module); +} + +static int mlxsw_m_port_stop(struct net_device *dev) +{ + struct mlxsw_m_port *mlxsw_m_port = netdev_priv(dev); + struct mlxsw_m *mlxsw_m = mlxsw_m_port->mlxsw_m; + + mlxsw_env_module_port_down(mlxsw_m->core, 0, mlxsw_m_port->module); + return 0; +} + +static struct devlink_port * +mlxsw_m_port_get_devlink_port(struct net_device *dev) +{ + struct mlxsw_m_port *mlxsw_m_port = netdev_priv(dev); + struct mlxsw_m *mlxsw_m = mlxsw_m_port->mlxsw_m; + + return mlxsw_core_port_devlink_port_get(mlxsw_m->core, + mlxsw_m_port->local_port); +} + +static const struct net_device_ops mlxsw_m_port_netdev_ops = { + .ndo_open = mlxsw_m_port_open, + .ndo_stop = mlxsw_m_port_stop, + .ndo_get_devlink_port = mlxsw_m_port_get_devlink_port, +}; + +static void mlxsw_m_module_get_drvinfo(struct net_device *dev, + struct ethtool_drvinfo *drvinfo) +{ + struct mlxsw_m_port *mlxsw_m_port = netdev_priv(dev); + struct mlxsw_m *mlxsw_m = mlxsw_m_port->mlxsw_m; + + strscpy(drvinfo->driver, mlxsw_m->bus_info->device_kind, + sizeof(drvinfo->driver)); + snprintf(drvinfo->fw_version, sizeof(drvinfo->fw_version), + "%d.%d.%d", + mlxsw_m->bus_info->fw_rev.major, + mlxsw_m->bus_info->fw_rev.minor, + mlxsw_m->bus_info->fw_rev.subminor); + strscpy(drvinfo->bus_info, mlxsw_m->bus_info->device_name, + sizeof(drvinfo->bus_info)); +} + +static int mlxsw_m_get_module_info(struct net_device *netdev, + struct ethtool_modinfo *modinfo) +{ + struct mlxsw_m_port *mlxsw_m_port = netdev_priv(netdev); + struct mlxsw_core *core = mlxsw_m_port->mlxsw_m->core; + + return mlxsw_env_get_module_info(netdev, core, + mlxsw_m_port->slot_index, + mlxsw_m_port->module, modinfo); +} + +static int +mlxsw_m_get_module_eeprom(struct net_device *netdev, struct ethtool_eeprom *ee, + u8 *data) +{ + struct mlxsw_m_port *mlxsw_m_port = netdev_priv(netdev); + struct mlxsw_core *core = mlxsw_m_port->mlxsw_m->core; + + return mlxsw_env_get_module_eeprom(netdev, core, + mlxsw_m_port->slot_index, + mlxsw_m_port->module, ee, data); +} + +static int +mlxsw_m_get_module_eeprom_by_page(struct net_device *netdev, + const struct ethtool_module_eeprom *page, + struct netlink_ext_ack *extack) +{ + struct mlxsw_m_port *mlxsw_m_port = netdev_priv(netdev); + struct mlxsw_core *core = mlxsw_m_port->mlxsw_m->core; + + return mlxsw_env_get_module_eeprom_by_page(core, + mlxsw_m_port->slot_index, + mlxsw_m_port->module, + page, extack); +} + +static int mlxsw_m_reset(struct net_device *netdev, u32 *flags) +{ + struct mlxsw_m_port *mlxsw_m_port = netdev_priv(netdev); + struct mlxsw_core *core = mlxsw_m_port->mlxsw_m->core; + + return mlxsw_env_reset_module(netdev, core, mlxsw_m_port->slot_index, + mlxsw_m_port->module, + flags); +} + +static int +mlxsw_m_get_module_power_mode(struct net_device *netdev, + struct ethtool_module_power_mode_params *params, + struct netlink_ext_ack *extack) +{ + struct mlxsw_m_port *mlxsw_m_port = netdev_priv(netdev); + struct mlxsw_core *core = mlxsw_m_port->mlxsw_m->core; + + return mlxsw_env_get_module_power_mode(core, mlxsw_m_port->slot_index, + mlxsw_m_port->module, + params, extack); +} + +static int +mlxsw_m_set_module_power_mode(struct net_device *netdev, + const struct ethtool_module_power_mode_params *params, + struct netlink_ext_ack *extack) +{ + struct mlxsw_m_port *mlxsw_m_port = netdev_priv(netdev); + struct mlxsw_core *core = mlxsw_m_port->mlxsw_m->core; + + return mlxsw_env_set_module_power_mode(core, mlxsw_m_port->slot_index, + mlxsw_m_port->module, + params->policy, extack); +} + +static const struct ethtool_ops mlxsw_m_port_ethtool_ops = { + .get_drvinfo = mlxsw_m_module_get_drvinfo, + .get_module_info = mlxsw_m_get_module_info, + .get_module_eeprom = mlxsw_m_get_module_eeprom, + .get_module_eeprom_by_page = mlxsw_m_get_module_eeprom_by_page, + .reset = mlxsw_m_reset, + .get_module_power_mode = mlxsw_m_get_module_power_mode, + .set_module_power_mode = mlxsw_m_set_module_power_mode, +}; + +static int +mlxsw_m_port_module_info_get(struct mlxsw_m *mlxsw_m, u16 local_port, + u8 *p_module, u8 *p_width, u8 *p_slot_index) +{ + char pmlp_pl[MLXSW_REG_PMLP_LEN]; + int err; + + mlxsw_reg_pmlp_pack(pmlp_pl, local_port); + err = mlxsw_reg_query(mlxsw_m->core, MLXSW_REG(pmlp), pmlp_pl); + if (err) + return err; + *p_module = mlxsw_reg_pmlp_module_get(pmlp_pl, 0); + *p_width = mlxsw_reg_pmlp_width_get(pmlp_pl); + *p_slot_index = mlxsw_reg_pmlp_slot_index_get(pmlp_pl, 0); + + return 0; +} + +static int +mlxsw_m_port_dev_addr_get(struct mlxsw_m_port *mlxsw_m_port) +{ + struct mlxsw_m *mlxsw_m = mlxsw_m_port->mlxsw_m; + char ppad_pl[MLXSW_REG_PPAD_LEN]; + u8 addr[ETH_ALEN]; + int err; + + mlxsw_reg_ppad_pack(ppad_pl, false, 0); + err = mlxsw_reg_query(mlxsw_m->core, MLXSW_REG(ppad), ppad_pl); + if (err) + return err; + mlxsw_reg_ppad_mac_memcpy_from(ppad_pl, addr); + eth_hw_addr_gen(mlxsw_m_port->dev, addr, mlxsw_m_port->module + 1 + + mlxsw_m_port->module_offset); + return 0; +} + +static bool mlxsw_m_port_created(struct mlxsw_m *mlxsw_m, u16 local_port) +{ + return mlxsw_m->ports[local_port]; +} + +static int +mlxsw_m_port_create(struct mlxsw_m *mlxsw_m, u16 local_port, u8 slot_index, + u8 module) +{ + struct mlxsw_m_port *mlxsw_m_port; + struct net_device *dev; + int err; + + err = mlxsw_core_port_init(mlxsw_m->core, local_port, slot_index, + module + 1, false, 0, false, + 0, mlxsw_m->base_mac, + sizeof(mlxsw_m->base_mac)); + if (err) { + dev_err(mlxsw_m->bus_info->dev, "Port %d: Failed to init core port\n", + local_port); + return err; + } + + dev = alloc_etherdev(sizeof(struct mlxsw_m_port)); + if (!dev) { + err = -ENOMEM; + goto err_alloc_etherdev; + } + + SET_NETDEV_DEV(dev, mlxsw_m->bus_info->dev); + dev_net_set(dev, mlxsw_core_net(mlxsw_m->core)); + mlxsw_m_port = netdev_priv(dev); + mlxsw_m_port->dev = dev; + mlxsw_m_port->mlxsw_m = mlxsw_m; + mlxsw_m_port->local_port = local_port; + mlxsw_m_port->module = module; + mlxsw_m_port->slot_index = slot_index; + /* Add module offset for line card. Offset for main board iz zero. + * For line card in slot #n offset is calculated as (#n - 1) + * multiplied by maximum modules number, which could be found on a line + * card. + */ + mlxsw_m_port->module_offset = mlxsw_m_port->slot_index ? + (mlxsw_m_port->slot_index - 1) * + mlxsw_m->max_modules_per_slot : 0; + + dev->netdev_ops = &mlxsw_m_port_netdev_ops; + dev->ethtool_ops = &mlxsw_m_port_ethtool_ops; + + err = mlxsw_m_port_dev_addr_get(mlxsw_m_port); + if (err) { + dev_err(mlxsw_m->bus_info->dev, "Port %d: Unable to get port mac address\n", + mlxsw_m_port->local_port); + goto err_dev_addr_get; + } + + netif_carrier_off(dev); + mlxsw_m->ports[local_port] = mlxsw_m_port; + err = register_netdev(dev); + if (err) { + dev_err(mlxsw_m->bus_info->dev, "Port %d: Failed to register netdev\n", + mlxsw_m_port->local_port); + goto err_register_netdev; + } + + mlxsw_core_port_eth_set(mlxsw_m->core, mlxsw_m_port->local_port, + mlxsw_m_port, dev); + + return 0; + +err_register_netdev: + mlxsw_m->ports[local_port] = NULL; +err_dev_addr_get: + free_netdev(dev); +err_alloc_etherdev: + mlxsw_core_port_fini(mlxsw_m->core, local_port); + return err; +} + +static void mlxsw_m_port_remove(struct mlxsw_m *mlxsw_m, u16 local_port) +{ + struct mlxsw_m_port *mlxsw_m_port = mlxsw_m->ports[local_port]; + + mlxsw_core_port_clear(mlxsw_m->core, local_port, mlxsw_m); + unregister_netdev(mlxsw_m_port->dev); /* This calls ndo_stop */ + mlxsw_m->ports[local_port] = NULL; + free_netdev(mlxsw_m_port->dev); + mlxsw_core_port_fini(mlxsw_m->core, local_port); +} + +static int* +mlxsw_m_port_mapping_get(struct mlxsw_m *mlxsw_m, u8 slot_index, u8 module) +{ + return &mlxsw_m->line_cards[slot_index]->module_to_port[module]; +} + +static int mlxsw_m_port_module_map(struct mlxsw_m *mlxsw_m, u16 local_port, + u8 *last_module) +{ + unsigned int max_ports = mlxsw_core_max_ports(mlxsw_m->core); + u8 module, width, slot_index; + int *module_to_port; + int err; + + /* Fill out to local port mapping array */ + err = mlxsw_m_port_module_info_get(mlxsw_m, local_port, &module, + &width, &slot_index); + if (err) + return err; + + /* Skip if line card has been already configured */ + if (mlxsw_m->line_cards[slot_index]->active) + return 0; + if (!width) + return 0; + /* Skip, if port belongs to the cluster */ + if (module == *last_module) + return 0; + *last_module = module; + + if (WARN_ON_ONCE(module >= max_ports)) + return -EINVAL; + mlxsw_env_module_port_map(mlxsw_m->core, slot_index, module); + module_to_port = mlxsw_m_port_mapping_get(mlxsw_m, slot_index, module); + *module_to_port = local_port; + + return 0; +} + +static void +mlxsw_m_port_module_unmap(struct mlxsw_m *mlxsw_m, u8 slot_index, u8 module) +{ + int *module_to_port = mlxsw_m_port_mapping_get(mlxsw_m, slot_index, + module); + *module_to_port = -1; + mlxsw_env_module_port_unmap(mlxsw_m->core, slot_index, module); +} + +static int mlxsw_m_linecards_init(struct mlxsw_m *mlxsw_m) +{ + unsigned int max_ports = mlxsw_core_max_ports(mlxsw_m->core); + char mgpir_pl[MLXSW_REG_MGPIR_LEN]; + u8 num_of_modules; + int i, j, err; + + mlxsw_reg_mgpir_pack(mgpir_pl, 0); + err = mlxsw_reg_query(mlxsw_m->core, MLXSW_REG(mgpir), mgpir_pl); + if (err) + return err; + + mlxsw_reg_mgpir_unpack(mgpir_pl, NULL, NULL, NULL, &num_of_modules, + &mlxsw_m->num_of_slots); + /* If the system is modular, get the maximum number of modules per-slot. + * Otherwise, get the maximum number of modules on the main board. + */ + if (mlxsw_m->num_of_slots) + mlxsw_m->max_modules_per_slot = + mlxsw_reg_mgpir_max_modules_per_slot_get(mgpir_pl); + else + mlxsw_m->max_modules_per_slot = num_of_modules; + /* Add slot for main board. */ + mlxsw_m->num_of_slots += 1; + + mlxsw_m->ports = kcalloc(max_ports, sizeof(*mlxsw_m->ports), + GFP_KERNEL); + if (!mlxsw_m->ports) + return -ENOMEM; + + mlxsw_m->line_cards = kcalloc(mlxsw_m->num_of_slots, + sizeof(*mlxsw_m->line_cards), + GFP_KERNEL); + if (!mlxsw_m->line_cards) { + err = -ENOMEM; + goto err_kcalloc; + } + + for (i = 0; i < mlxsw_m->num_of_slots; i++) { + mlxsw_m->line_cards[i] = + kzalloc(struct_size(mlxsw_m->line_cards[i], + module_to_port, + mlxsw_m->max_modules_per_slot), + GFP_KERNEL); + if (!mlxsw_m->line_cards[i]) { + err = -ENOMEM; + goto err_kmalloc_array; + } + + /* Invalidate the entries of module to local port mapping array. */ + for (j = 0; j < mlxsw_m->max_modules_per_slot; j++) + mlxsw_m->line_cards[i]->module_to_port[j] = -1; + } + + return 0; + +err_kmalloc_array: + for (i--; i >= 0; i--) + kfree(mlxsw_m->line_cards[i]); + kfree(mlxsw_m->line_cards); +err_kcalloc: + kfree(mlxsw_m->ports); + return err; +} + +static void mlxsw_m_linecards_fini(struct mlxsw_m *mlxsw_m) +{ + int i = mlxsw_m->num_of_slots; + + for (i--; i >= 0; i--) + kfree(mlxsw_m->line_cards[i]); + kfree(mlxsw_m->line_cards); + kfree(mlxsw_m->ports); +} + +static void +mlxsw_m_linecard_port_module_unmap(struct mlxsw_m *mlxsw_m, u8 slot_index) +{ + int i; + + for (i = mlxsw_m->max_modules_per_slot - 1; i >= 0; i--) { + int *module_to_port; + + module_to_port = mlxsw_m_port_mapping_get(mlxsw_m, slot_index, i); + if (*module_to_port > 0) + mlxsw_m_port_module_unmap(mlxsw_m, slot_index, i); + } +} + +static int +mlxsw_m_linecard_ports_create(struct mlxsw_m *mlxsw_m, u8 slot_index) +{ + int *module_to_port; + int i, err; + + for (i = 0; i < mlxsw_m->max_modules_per_slot; i++) { + module_to_port = mlxsw_m_port_mapping_get(mlxsw_m, slot_index, i); + if (*module_to_port > 0) { + err = mlxsw_m_port_create(mlxsw_m, *module_to_port, + slot_index, i); + if (err) + goto err_port_create; + /* Mark slot as active */ + if (!mlxsw_m->line_cards[slot_index]->active) + mlxsw_m->line_cards[slot_index]->active = true; + } + } + return 0; + +err_port_create: + for (i--; i >= 0; i--) { + module_to_port = mlxsw_m_port_mapping_get(mlxsw_m, slot_index, i); + if (*module_to_port > 0 && + mlxsw_m_port_created(mlxsw_m, *module_to_port)) { + mlxsw_m_port_remove(mlxsw_m, *module_to_port); + /* Mark slot as inactive */ + if (mlxsw_m->line_cards[slot_index]->active) + mlxsw_m->line_cards[slot_index]->active = false; + } + } + return err; +} + +static void +mlxsw_m_linecard_ports_remove(struct mlxsw_m *mlxsw_m, u8 slot_index) +{ + int i; + + for (i = 0; i < mlxsw_m->max_modules_per_slot; i++) { + int *module_to_port = mlxsw_m_port_mapping_get(mlxsw_m, + slot_index, i); + + if (*module_to_port > 0 && + mlxsw_m_port_created(mlxsw_m, *module_to_port)) { + mlxsw_m_port_remove(mlxsw_m, *module_to_port); + mlxsw_m_port_module_unmap(mlxsw_m, slot_index, i); + } + } +} + +static int mlxsw_m_ports_module_map(struct mlxsw_m *mlxsw_m) +{ + unsigned int max_ports = mlxsw_core_max_ports(mlxsw_m->core); + u8 last_module = max_ports; + int i, err; + + for (i = 1; i < max_ports; i++) { + err = mlxsw_m_port_module_map(mlxsw_m, i, &last_module); + if (err) + return err; + } + + return 0; +} + +static int mlxsw_m_ports_create(struct mlxsw_m *mlxsw_m) +{ + int err; + + /* Fill out module to local port mapping array */ + err = mlxsw_m_ports_module_map(mlxsw_m); + if (err) + goto err_ports_module_map; + + /* Create port objects for each valid entry */ + err = mlxsw_m_linecard_ports_create(mlxsw_m, 0); + if (err) + goto err_linecard_ports_create; + + return 0; + +err_linecard_ports_create: +err_ports_module_map: + mlxsw_m_linecard_port_module_unmap(mlxsw_m, 0); + + return err; +} + +static void mlxsw_m_ports_remove(struct mlxsw_m *mlxsw_m) +{ + mlxsw_m_linecard_ports_remove(mlxsw_m, 0); +} + +static void +mlxsw_m_ports_remove_selected(struct mlxsw_core *mlxsw_core, + bool (*selector)(void *priv, u16 local_port), + void *priv) +{ + struct mlxsw_m *mlxsw_m = mlxsw_core_driver_priv(mlxsw_core); + struct mlxsw_linecard *linecard_priv = priv; + struct mlxsw_m_line_card *linecard; + + linecard = mlxsw_m->line_cards[linecard_priv->slot_index]; + + if (WARN_ON(!linecard->active)) + return; + + mlxsw_m_linecard_ports_remove(mlxsw_m, linecard_priv->slot_index); + linecard->active = false; +} + +static int mlxsw_m_fw_rev_validate(struct mlxsw_m *mlxsw_m) +{ + const struct mlxsw_fw_rev *rev = &mlxsw_m->bus_info->fw_rev; + + /* Validate driver and FW are compatible. + * Do not check major version, since it defines chip type, while + * driver is supposed to support any type. + */ + if (mlxsw_core_fw_rev_minor_subminor_validate(rev, &mlxsw_m_fw_rev)) + return 0; + + dev_err(mlxsw_m->bus_info->dev, "The firmware version %d.%d.%d is incompatible with the driver (required >= %d.%d.%d)\n", + rev->major, rev->minor, rev->subminor, rev->major, + mlxsw_m_fw_rev.minor, mlxsw_m_fw_rev.subminor); + + return -EINVAL; +} + +static void +mlxsw_m_got_active(struct mlxsw_core *mlxsw_core, u8 slot_index, void *priv) +{ + struct mlxsw_m_line_card *linecard; + struct mlxsw_m *mlxsw_m = priv; + int err; + + linecard = mlxsw_m->line_cards[slot_index]; + /* Skip if line card has been already configured during init */ + if (linecard->active) + return; + + /* Fill out module to local port mapping array */ + err = mlxsw_m_ports_module_map(mlxsw_m); + if (err) + goto err_ports_module_map; + + /* Create port objects for each valid entry */ + err = mlxsw_m_linecard_ports_create(mlxsw_m, slot_index); + if (err) { + dev_err(mlxsw_m->bus_info->dev, "Failed to create port for line card at slot %d\n", + slot_index); + goto err_linecard_ports_create; + } + + linecard->active = true; + + return; + +err_linecard_ports_create: +err_ports_module_map: + mlxsw_m_linecard_port_module_unmap(mlxsw_m, slot_index); +} + +static void +mlxsw_m_got_inactive(struct mlxsw_core *mlxsw_core, u8 slot_index, void *priv) +{ + struct mlxsw_m_line_card *linecard; + struct mlxsw_m *mlxsw_m = priv; + + linecard = mlxsw_m->line_cards[slot_index]; + + if (WARN_ON(!linecard->active)) + return; + + mlxsw_m_linecard_ports_remove(mlxsw_m, slot_index); + linecard->active = false; +} + +static struct mlxsw_linecards_event_ops mlxsw_m_event_ops = { + .got_active = mlxsw_m_got_active, + .got_inactive = mlxsw_m_got_inactive, +}; + +static int mlxsw_m_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct netlink_ext_ack *extack) +{ + struct mlxsw_m *mlxsw_m = mlxsw_core_driver_priv(mlxsw_core); + int err; + + mlxsw_m->core = mlxsw_core; + mlxsw_m->bus_info = mlxsw_bus_info; + + err = mlxsw_m_fw_rev_validate(mlxsw_m); + if (err) + return err; + + err = mlxsw_m_base_mac_get(mlxsw_m); + if (err) { + dev_err(mlxsw_m->bus_info->dev, "Failed to get base mac\n"); + return err; + } + + err = mlxsw_m_linecards_init(mlxsw_m); + if (err) { + dev_err(mlxsw_m->bus_info->dev, "Failed to create line cards\n"); + return err; + } + + err = mlxsw_linecards_event_ops_register(mlxsw_core, + &mlxsw_m_event_ops, mlxsw_m); + if (err) { + dev_err(mlxsw_m->bus_info->dev, "Failed to register line cards operations\n"); + goto linecards_event_ops_register; + } + + err = mlxsw_m_ports_create(mlxsw_m); + if (err) { + dev_err(mlxsw_m->bus_info->dev, "Failed to create ports\n"); + goto err_ports_create; + } + + return 0; + +err_ports_create: + mlxsw_linecards_event_ops_unregister(mlxsw_core, + &mlxsw_m_event_ops, mlxsw_m); +linecards_event_ops_register: + mlxsw_m_linecards_fini(mlxsw_m); + return err; +} + +static void mlxsw_m_fini(struct mlxsw_core *mlxsw_core) +{ + struct mlxsw_m *mlxsw_m = mlxsw_core_driver_priv(mlxsw_core); + + mlxsw_m_ports_remove(mlxsw_m); + mlxsw_linecards_event_ops_unregister(mlxsw_core, + &mlxsw_m_event_ops, mlxsw_m); + mlxsw_m_linecards_fini(mlxsw_m); +} + +static const struct mlxsw_config_profile mlxsw_m_config_profile; + +static struct mlxsw_driver mlxsw_m_driver = { + .kind = mlxsw_m_driver_name, + .priv_size = sizeof(struct mlxsw_m), + .init = mlxsw_m_init, + .fini = mlxsw_m_fini, + .ports_remove_selected = mlxsw_m_ports_remove_selected, + .profile = &mlxsw_m_config_profile, +}; + +static const struct i2c_device_id mlxsw_m_i2c_id[] = { + { "mlxsw_minimal", 0}, + { }, +}; + +static struct i2c_driver mlxsw_m_i2c_driver = { + .driver.name = "mlxsw_minimal", + .class = I2C_CLASS_HWMON, + .id_table = mlxsw_m_i2c_id, +}; + +static int __init mlxsw_m_module_init(void) +{ + int err; + + err = mlxsw_core_driver_register(&mlxsw_m_driver); + if (err) + return err; + + err = mlxsw_i2c_driver_register(&mlxsw_m_i2c_driver); + if (err) + goto err_i2c_driver_register; + + return 0; + +err_i2c_driver_register: + mlxsw_core_driver_unregister(&mlxsw_m_driver); + + return err; +} + +static void __exit mlxsw_m_module_exit(void) +{ + mlxsw_i2c_driver_unregister(&mlxsw_m_i2c_driver); + mlxsw_core_driver_unregister(&mlxsw_m_driver); +} + +module_init(mlxsw_m_module_init); +module_exit(mlxsw_m_module_exit); + +MODULE_LICENSE("Dual BSD/GPL"); +MODULE_AUTHOR("Vadim Pasternak <vadimp@mellanox.com>"); +MODULE_DESCRIPTION("Mellanox minimal driver"); +MODULE_DEVICE_TABLE(i2c, mlxsw_m_i2c_id); diff --git a/drivers/net/ethernet/mellanox/mlxsw/pci.c b/drivers/net/ethernet/mellanox/mlxsw/pci.c new file mode 100644 index 000000000..51eea1f05 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/pci.c @@ -0,0 +1,2068 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/export.h> +#include <linux/err.h> +#include <linux/device.h> +#include <linux/pci.h> +#include <linux/interrupt.h> +#include <linux/wait.h> +#include <linux/types.h> +#include <linux/skbuff.h> +#include <linux/if_vlan.h> +#include <linux/log2.h> +#include <linux/string.h> + +#include "pci_hw.h" +#include "pci.h" +#include "core.h" +#include "cmd.h" +#include "port.h" +#include "resources.h" + +#define mlxsw_pci_write32(mlxsw_pci, reg, val) \ + iowrite32be(val, (mlxsw_pci)->hw_addr + (MLXSW_PCI_ ## reg)) +#define mlxsw_pci_read32(mlxsw_pci, reg) \ + ioread32be((mlxsw_pci)->hw_addr + (MLXSW_PCI_ ## reg)) + +enum mlxsw_pci_queue_type { + MLXSW_PCI_QUEUE_TYPE_SDQ, + MLXSW_PCI_QUEUE_TYPE_RDQ, + MLXSW_PCI_QUEUE_TYPE_CQ, + MLXSW_PCI_QUEUE_TYPE_EQ, +}; + +#define MLXSW_PCI_QUEUE_TYPE_COUNT 4 + +static const u16 mlxsw_pci_doorbell_type_offset[] = { + MLXSW_PCI_DOORBELL_SDQ_OFFSET, /* for type MLXSW_PCI_QUEUE_TYPE_SDQ */ + MLXSW_PCI_DOORBELL_RDQ_OFFSET, /* for type MLXSW_PCI_QUEUE_TYPE_RDQ */ + MLXSW_PCI_DOORBELL_CQ_OFFSET, /* for type MLXSW_PCI_QUEUE_TYPE_CQ */ + MLXSW_PCI_DOORBELL_EQ_OFFSET, /* for type MLXSW_PCI_QUEUE_TYPE_EQ */ +}; + +static const u16 mlxsw_pci_doorbell_arm_type_offset[] = { + 0, /* unused */ + 0, /* unused */ + MLXSW_PCI_DOORBELL_ARM_CQ_OFFSET, /* for type MLXSW_PCI_QUEUE_TYPE_CQ */ + MLXSW_PCI_DOORBELL_ARM_EQ_OFFSET, /* for type MLXSW_PCI_QUEUE_TYPE_EQ */ +}; + +struct mlxsw_pci_mem_item { + char *buf; + dma_addr_t mapaddr; + size_t size; +}; + +struct mlxsw_pci_queue_elem_info { + char *elem; /* pointer to actual dma mapped element mem chunk */ + union { + struct { + struct sk_buff *skb; + } sdq; + struct { + struct sk_buff *skb; + } rdq; + } u; +}; + +struct mlxsw_pci_queue { + spinlock_t lock; /* for queue accesses */ + struct mlxsw_pci_mem_item mem_item; + struct mlxsw_pci_queue_elem_info *elem_info; + u16 producer_counter; + u16 consumer_counter; + u16 count; /* number of elements in queue */ + u8 num; /* queue number */ + u8 elem_size; /* size of one element */ + enum mlxsw_pci_queue_type type; + struct tasklet_struct tasklet; /* queue processing tasklet */ + struct mlxsw_pci *pci; + union { + struct { + u32 comp_sdq_count; + u32 comp_rdq_count; + enum mlxsw_pci_cqe_v v; + } cq; + struct { + u32 ev_cmd_count; + u32 ev_comp_count; + u32 ev_other_count; + } eq; + } u; +}; + +struct mlxsw_pci_queue_type_group { + struct mlxsw_pci_queue *q; + u8 count; /* number of queues in group */ +}; + +struct mlxsw_pci { + struct pci_dev *pdev; + u8 __iomem *hw_addr; + u64 free_running_clock_offset; + u64 utc_sec_offset; + u64 utc_nsec_offset; + struct mlxsw_pci_queue_type_group queues[MLXSW_PCI_QUEUE_TYPE_COUNT]; + u32 doorbell_offset; + struct mlxsw_core *core; + struct { + struct mlxsw_pci_mem_item *items; + unsigned int count; + } fw_area; + struct { + struct mlxsw_pci_mem_item out_mbox; + struct mlxsw_pci_mem_item in_mbox; + struct mutex lock; /* Lock access to command registers */ + bool nopoll; + wait_queue_head_t wait; + bool wait_done; + struct { + u8 status; + u64 out_param; + } comp; + } cmd; + struct mlxsw_bus_info bus_info; + const struct pci_device_id *id; + enum mlxsw_pci_cqe_v max_cqe_ver; /* Maximal supported CQE version */ + u8 num_sdq_cqs; /* Number of CQs used for SDQs */ +}; + +static void mlxsw_pci_queue_tasklet_schedule(struct mlxsw_pci_queue *q) +{ + tasklet_schedule(&q->tasklet); +} + +static char *__mlxsw_pci_queue_elem_get(struct mlxsw_pci_queue *q, + size_t elem_size, int elem_index) +{ + return q->mem_item.buf + (elem_size * elem_index); +} + +static struct mlxsw_pci_queue_elem_info * +mlxsw_pci_queue_elem_info_get(struct mlxsw_pci_queue *q, int elem_index) +{ + return &q->elem_info[elem_index]; +} + +static struct mlxsw_pci_queue_elem_info * +mlxsw_pci_queue_elem_info_producer_get(struct mlxsw_pci_queue *q) +{ + int index = q->producer_counter & (q->count - 1); + + if ((u16) (q->producer_counter - q->consumer_counter) == q->count) + return NULL; + return mlxsw_pci_queue_elem_info_get(q, index); +} + +static struct mlxsw_pci_queue_elem_info * +mlxsw_pci_queue_elem_info_consumer_get(struct mlxsw_pci_queue *q) +{ + int index = q->consumer_counter & (q->count - 1); + + return mlxsw_pci_queue_elem_info_get(q, index); +} + +static char *mlxsw_pci_queue_elem_get(struct mlxsw_pci_queue *q, int elem_index) +{ + return mlxsw_pci_queue_elem_info_get(q, elem_index)->elem; +} + +static bool mlxsw_pci_elem_hw_owned(struct mlxsw_pci_queue *q, bool owner_bit) +{ + return owner_bit != !!(q->consumer_counter & q->count); +} + +static struct mlxsw_pci_queue_type_group * +mlxsw_pci_queue_type_group_get(struct mlxsw_pci *mlxsw_pci, + enum mlxsw_pci_queue_type q_type) +{ + return &mlxsw_pci->queues[q_type]; +} + +static u8 __mlxsw_pci_queue_count(struct mlxsw_pci *mlxsw_pci, + enum mlxsw_pci_queue_type q_type) +{ + struct mlxsw_pci_queue_type_group *queue_group; + + queue_group = mlxsw_pci_queue_type_group_get(mlxsw_pci, q_type); + return queue_group->count; +} + +static u8 mlxsw_pci_sdq_count(struct mlxsw_pci *mlxsw_pci) +{ + return __mlxsw_pci_queue_count(mlxsw_pci, MLXSW_PCI_QUEUE_TYPE_SDQ); +} + +static u8 mlxsw_pci_cq_count(struct mlxsw_pci *mlxsw_pci) +{ + return __mlxsw_pci_queue_count(mlxsw_pci, MLXSW_PCI_QUEUE_TYPE_CQ); +} + +static struct mlxsw_pci_queue * +__mlxsw_pci_queue_get(struct mlxsw_pci *mlxsw_pci, + enum mlxsw_pci_queue_type q_type, u8 q_num) +{ + return &mlxsw_pci->queues[q_type].q[q_num]; +} + +static struct mlxsw_pci_queue *mlxsw_pci_sdq_get(struct mlxsw_pci *mlxsw_pci, + u8 q_num) +{ + return __mlxsw_pci_queue_get(mlxsw_pci, + MLXSW_PCI_QUEUE_TYPE_SDQ, q_num); +} + +static struct mlxsw_pci_queue *mlxsw_pci_rdq_get(struct mlxsw_pci *mlxsw_pci, + u8 q_num) +{ + return __mlxsw_pci_queue_get(mlxsw_pci, + MLXSW_PCI_QUEUE_TYPE_RDQ, q_num); +} + +static struct mlxsw_pci_queue *mlxsw_pci_cq_get(struct mlxsw_pci *mlxsw_pci, + u8 q_num) +{ + return __mlxsw_pci_queue_get(mlxsw_pci, MLXSW_PCI_QUEUE_TYPE_CQ, q_num); +} + +static struct mlxsw_pci_queue *mlxsw_pci_eq_get(struct mlxsw_pci *mlxsw_pci, + u8 q_num) +{ + return __mlxsw_pci_queue_get(mlxsw_pci, MLXSW_PCI_QUEUE_TYPE_EQ, q_num); +} + +static void __mlxsw_pci_queue_doorbell_set(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q, + u16 val) +{ + mlxsw_pci_write32(mlxsw_pci, + DOORBELL(mlxsw_pci->doorbell_offset, + mlxsw_pci_doorbell_type_offset[q->type], + q->num), val); +} + +static void __mlxsw_pci_queue_doorbell_arm_set(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q, + u16 val) +{ + mlxsw_pci_write32(mlxsw_pci, + DOORBELL(mlxsw_pci->doorbell_offset, + mlxsw_pci_doorbell_arm_type_offset[q->type], + q->num), val); +} + +static void mlxsw_pci_queue_doorbell_producer_ring(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q) +{ + wmb(); /* ensure all writes are done before we ring a bell */ + __mlxsw_pci_queue_doorbell_set(mlxsw_pci, q, q->producer_counter); +} + +static void mlxsw_pci_queue_doorbell_consumer_ring(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q) +{ + wmb(); /* ensure all writes are done before we ring a bell */ + __mlxsw_pci_queue_doorbell_set(mlxsw_pci, q, + q->consumer_counter + q->count); +} + +static void +mlxsw_pci_queue_doorbell_arm_consumer_ring(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q) +{ + wmb(); /* ensure all writes are done before we ring a bell */ + __mlxsw_pci_queue_doorbell_arm_set(mlxsw_pci, q, q->consumer_counter); +} + +static dma_addr_t __mlxsw_pci_queue_page_get(struct mlxsw_pci_queue *q, + int page_index) +{ + return q->mem_item.mapaddr + MLXSW_PCI_PAGE_SIZE * page_index; +} + +static int mlxsw_pci_sdq_init(struct mlxsw_pci *mlxsw_pci, char *mbox, + struct mlxsw_pci_queue *q) +{ + int tclass; + int lp; + int i; + int err; + + q->producer_counter = 0; + q->consumer_counter = 0; + tclass = q->num == MLXSW_PCI_SDQ_EMAD_INDEX ? MLXSW_PCI_SDQ_EMAD_TC : + MLXSW_PCI_SDQ_CTL_TC; + lp = q->num == MLXSW_PCI_SDQ_EMAD_INDEX ? MLXSW_CMD_MBOX_SW2HW_DQ_SDQ_LP_IGNORE_WQE : + MLXSW_CMD_MBOX_SW2HW_DQ_SDQ_LP_WQE; + + /* Set CQ of same number of this SDQ. */ + mlxsw_cmd_mbox_sw2hw_dq_cq_set(mbox, q->num); + mlxsw_cmd_mbox_sw2hw_dq_sdq_lp_set(mbox, lp); + mlxsw_cmd_mbox_sw2hw_dq_sdq_tclass_set(mbox, tclass); + mlxsw_cmd_mbox_sw2hw_dq_log2_dq_sz_set(mbox, 3); /* 8 pages */ + for (i = 0; i < MLXSW_PCI_AQ_PAGES; i++) { + dma_addr_t mapaddr = __mlxsw_pci_queue_page_get(q, i); + + mlxsw_cmd_mbox_sw2hw_dq_pa_set(mbox, i, mapaddr); + } + + err = mlxsw_cmd_sw2hw_sdq(mlxsw_pci->core, mbox, q->num); + if (err) + return err; + mlxsw_pci_queue_doorbell_producer_ring(mlxsw_pci, q); + return 0; +} + +static void mlxsw_pci_sdq_fini(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q) +{ + mlxsw_cmd_hw2sw_sdq(mlxsw_pci->core, q->num); +} + +static int mlxsw_pci_wqe_frag_map(struct mlxsw_pci *mlxsw_pci, char *wqe, + int index, char *frag_data, size_t frag_len, + int direction) +{ + struct pci_dev *pdev = mlxsw_pci->pdev; + dma_addr_t mapaddr; + + mapaddr = dma_map_single(&pdev->dev, frag_data, frag_len, direction); + if (unlikely(dma_mapping_error(&pdev->dev, mapaddr))) { + dev_err_ratelimited(&pdev->dev, "failed to dma map tx frag\n"); + return -EIO; + } + mlxsw_pci_wqe_address_set(wqe, index, mapaddr); + mlxsw_pci_wqe_byte_count_set(wqe, index, frag_len); + return 0; +} + +static void mlxsw_pci_wqe_frag_unmap(struct mlxsw_pci *mlxsw_pci, char *wqe, + int index, int direction) +{ + struct pci_dev *pdev = mlxsw_pci->pdev; + size_t frag_len = mlxsw_pci_wqe_byte_count_get(wqe, index); + dma_addr_t mapaddr = mlxsw_pci_wqe_address_get(wqe, index); + + if (!frag_len) + return; + dma_unmap_single(&pdev->dev, mapaddr, frag_len, direction); +} + +static int mlxsw_pci_rdq_skb_alloc(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue_elem_info *elem_info) +{ + size_t buf_len = MLXSW_PORT_MAX_MTU; + char *wqe = elem_info->elem; + struct sk_buff *skb; + int err; + + skb = netdev_alloc_skb_ip_align(NULL, buf_len); + if (!skb) + return -ENOMEM; + + err = mlxsw_pci_wqe_frag_map(mlxsw_pci, wqe, 0, skb->data, + buf_len, DMA_FROM_DEVICE); + if (err) + goto err_frag_map; + + elem_info->u.rdq.skb = skb; + return 0; + +err_frag_map: + dev_kfree_skb_any(skb); + return err; +} + +static void mlxsw_pci_rdq_skb_free(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue_elem_info *elem_info) +{ + struct sk_buff *skb; + char *wqe; + + skb = elem_info->u.rdq.skb; + wqe = elem_info->elem; + + mlxsw_pci_wqe_frag_unmap(mlxsw_pci, wqe, 0, DMA_FROM_DEVICE); + dev_kfree_skb_any(skb); +} + +static int mlxsw_pci_rdq_init(struct mlxsw_pci *mlxsw_pci, char *mbox, + struct mlxsw_pci_queue *q) +{ + struct mlxsw_pci_queue_elem_info *elem_info; + u8 sdq_count = mlxsw_pci_sdq_count(mlxsw_pci); + int i; + int err; + + q->producer_counter = 0; + q->consumer_counter = 0; + + /* Set CQ of same number of this RDQ with base + * above SDQ count as the lower ones are assigned to SDQs. + */ + mlxsw_cmd_mbox_sw2hw_dq_cq_set(mbox, sdq_count + q->num); + mlxsw_cmd_mbox_sw2hw_dq_log2_dq_sz_set(mbox, 3); /* 8 pages */ + for (i = 0; i < MLXSW_PCI_AQ_PAGES; i++) { + dma_addr_t mapaddr = __mlxsw_pci_queue_page_get(q, i); + + mlxsw_cmd_mbox_sw2hw_dq_pa_set(mbox, i, mapaddr); + } + + err = mlxsw_cmd_sw2hw_rdq(mlxsw_pci->core, mbox, q->num); + if (err) + return err; + + mlxsw_pci_queue_doorbell_producer_ring(mlxsw_pci, q); + + for (i = 0; i < q->count; i++) { + elem_info = mlxsw_pci_queue_elem_info_producer_get(q); + BUG_ON(!elem_info); + err = mlxsw_pci_rdq_skb_alloc(mlxsw_pci, elem_info); + if (err) + goto rollback; + /* Everything is set up, ring doorbell to pass elem to HW */ + q->producer_counter++; + mlxsw_pci_queue_doorbell_producer_ring(mlxsw_pci, q); + } + + return 0; + +rollback: + for (i--; i >= 0; i--) { + elem_info = mlxsw_pci_queue_elem_info_get(q, i); + mlxsw_pci_rdq_skb_free(mlxsw_pci, elem_info); + } + mlxsw_cmd_hw2sw_rdq(mlxsw_pci->core, q->num); + + return err; +} + +static void mlxsw_pci_rdq_fini(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q) +{ + struct mlxsw_pci_queue_elem_info *elem_info; + int i; + + mlxsw_cmd_hw2sw_rdq(mlxsw_pci->core, q->num); + for (i = 0; i < q->count; i++) { + elem_info = mlxsw_pci_queue_elem_info_get(q, i); + mlxsw_pci_rdq_skb_free(mlxsw_pci, elem_info); + } +} + +static void mlxsw_pci_cq_pre_init(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q) +{ + q->u.cq.v = mlxsw_pci->max_cqe_ver; + + if (q->u.cq.v == MLXSW_PCI_CQE_V2 && + q->num < mlxsw_pci->num_sdq_cqs && + !mlxsw_core_sdq_supports_cqe_v2(mlxsw_pci->core)) + q->u.cq.v = MLXSW_PCI_CQE_V1; +} + +static int mlxsw_pci_cq_init(struct mlxsw_pci *mlxsw_pci, char *mbox, + struct mlxsw_pci_queue *q) +{ + int i; + int err; + + q->consumer_counter = 0; + + for (i = 0; i < q->count; i++) { + char *elem = mlxsw_pci_queue_elem_get(q, i); + + mlxsw_pci_cqe_owner_set(q->u.cq.v, elem, 1); + } + + if (q->u.cq.v == MLXSW_PCI_CQE_V1) + mlxsw_cmd_mbox_sw2hw_cq_cqe_ver_set(mbox, + MLXSW_CMD_MBOX_SW2HW_CQ_CQE_VER_1); + else if (q->u.cq.v == MLXSW_PCI_CQE_V2) + mlxsw_cmd_mbox_sw2hw_cq_cqe_ver_set(mbox, + MLXSW_CMD_MBOX_SW2HW_CQ_CQE_VER_2); + + mlxsw_cmd_mbox_sw2hw_cq_c_eqn_set(mbox, MLXSW_PCI_EQ_COMP_NUM); + mlxsw_cmd_mbox_sw2hw_cq_st_set(mbox, 0); + mlxsw_cmd_mbox_sw2hw_cq_log_cq_size_set(mbox, ilog2(q->count)); + for (i = 0; i < MLXSW_PCI_AQ_PAGES; i++) { + dma_addr_t mapaddr = __mlxsw_pci_queue_page_get(q, i); + + mlxsw_cmd_mbox_sw2hw_cq_pa_set(mbox, i, mapaddr); + } + err = mlxsw_cmd_sw2hw_cq(mlxsw_pci->core, mbox, q->num); + if (err) + return err; + mlxsw_pci_queue_doorbell_consumer_ring(mlxsw_pci, q); + mlxsw_pci_queue_doorbell_arm_consumer_ring(mlxsw_pci, q); + return 0; +} + +static void mlxsw_pci_cq_fini(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q) +{ + mlxsw_cmd_hw2sw_cq(mlxsw_pci->core, q->num); +} + +static unsigned int mlxsw_pci_read32_off(struct mlxsw_pci *mlxsw_pci, + ptrdiff_t off) +{ + return ioread32be(mlxsw_pci->hw_addr + off); +} + +static void mlxsw_pci_skb_cb_ts_set(struct mlxsw_pci *mlxsw_pci, + struct sk_buff *skb, + enum mlxsw_pci_cqe_v cqe_v, char *cqe) +{ + u8 ts_type; + + if (cqe_v != MLXSW_PCI_CQE_V2) + return; + + ts_type = mlxsw_pci_cqe2_time_stamp_type_get(cqe); + + if (ts_type != MLXSW_PCI_CQE_TIME_STAMP_TYPE_UTC && + ts_type != MLXSW_PCI_CQE_TIME_STAMP_TYPE_MIRROR_UTC) + return; + + mlxsw_skb_cb(skb)->cqe_ts.sec = mlxsw_pci_cqe2_time_stamp_sec_get(cqe); + mlxsw_skb_cb(skb)->cqe_ts.nsec = + mlxsw_pci_cqe2_time_stamp_nsec_get(cqe); +} + +static void mlxsw_pci_cqe_sdq_handle(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q, + u16 consumer_counter_limit, + enum mlxsw_pci_cqe_v cqe_v, + char *cqe) +{ + struct pci_dev *pdev = mlxsw_pci->pdev; + struct mlxsw_pci_queue_elem_info *elem_info; + struct mlxsw_tx_info tx_info; + char *wqe; + struct sk_buff *skb; + int i; + + spin_lock(&q->lock); + elem_info = mlxsw_pci_queue_elem_info_consumer_get(q); + tx_info = mlxsw_skb_cb(elem_info->u.sdq.skb)->tx_info; + skb = elem_info->u.sdq.skb; + wqe = elem_info->elem; + for (i = 0; i < MLXSW_PCI_WQE_SG_ENTRIES; i++) + mlxsw_pci_wqe_frag_unmap(mlxsw_pci, wqe, i, DMA_TO_DEVICE); + + if (unlikely(!tx_info.is_emad && + skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)) { + mlxsw_pci_skb_cb_ts_set(mlxsw_pci, skb, cqe_v, cqe); + mlxsw_core_ptp_transmitted(mlxsw_pci->core, skb, + tx_info.local_port); + skb = NULL; + } + + if (skb) + dev_kfree_skb_any(skb); + elem_info->u.sdq.skb = NULL; + + if (q->consumer_counter++ != consumer_counter_limit) + dev_dbg_ratelimited(&pdev->dev, "Consumer counter does not match limit in SDQ\n"); + spin_unlock(&q->lock); +} + +static void mlxsw_pci_cqe_rdq_md_tx_port_init(struct sk_buff *skb, + const char *cqe) +{ + struct mlxsw_skb_cb *cb = mlxsw_skb_cb(skb); + + if (mlxsw_pci_cqe2_tx_lag_get(cqe)) { + cb->rx_md_info.tx_port_is_lag = true; + cb->rx_md_info.tx_lag_id = mlxsw_pci_cqe2_tx_lag_id_get(cqe); + cb->rx_md_info.tx_lag_port_index = + mlxsw_pci_cqe2_tx_lag_subport_get(cqe); + } else { + cb->rx_md_info.tx_port_is_lag = false; + cb->rx_md_info.tx_sys_port = + mlxsw_pci_cqe2_tx_system_port_get(cqe); + } + + if (cb->rx_md_info.tx_sys_port != MLXSW_PCI_CQE2_TX_PORT_MULTI_PORT && + cb->rx_md_info.tx_sys_port != MLXSW_PCI_CQE2_TX_PORT_INVALID) + cb->rx_md_info.tx_port_valid = 1; + else + cb->rx_md_info.tx_port_valid = 0; +} + +static void mlxsw_pci_cqe_rdq_md_init(struct sk_buff *skb, const char *cqe) +{ + struct mlxsw_skb_cb *cb = mlxsw_skb_cb(skb); + + cb->rx_md_info.tx_congestion = mlxsw_pci_cqe2_mirror_cong_get(cqe); + if (cb->rx_md_info.tx_congestion != MLXSW_PCI_CQE2_MIRROR_CONG_INVALID) + cb->rx_md_info.tx_congestion_valid = 1; + else + cb->rx_md_info.tx_congestion_valid = 0; + cb->rx_md_info.tx_congestion <<= MLXSW_PCI_CQE2_MIRROR_CONG_SHIFT; + + cb->rx_md_info.latency = mlxsw_pci_cqe2_mirror_latency_get(cqe); + if (cb->rx_md_info.latency != MLXSW_PCI_CQE2_MIRROR_LATENCY_INVALID) + cb->rx_md_info.latency_valid = 1; + else + cb->rx_md_info.latency_valid = 0; + + cb->rx_md_info.tx_tc = mlxsw_pci_cqe2_mirror_tclass_get(cqe); + if (cb->rx_md_info.tx_tc != MLXSW_PCI_CQE2_MIRROR_TCLASS_INVALID) + cb->rx_md_info.tx_tc_valid = 1; + else + cb->rx_md_info.tx_tc_valid = 0; + + mlxsw_pci_cqe_rdq_md_tx_port_init(skb, cqe); +} + +static void mlxsw_pci_cqe_rdq_handle(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q, + u16 consumer_counter_limit, + enum mlxsw_pci_cqe_v cqe_v, char *cqe) +{ + struct pci_dev *pdev = mlxsw_pci->pdev; + struct mlxsw_pci_queue_elem_info *elem_info; + struct mlxsw_rx_info rx_info = {}; + char wqe[MLXSW_PCI_WQE_SIZE]; + struct sk_buff *skb; + u16 byte_count; + int err; + + elem_info = mlxsw_pci_queue_elem_info_consumer_get(q); + skb = elem_info->u.rdq.skb; + memcpy(wqe, elem_info->elem, MLXSW_PCI_WQE_SIZE); + + if (q->consumer_counter++ != consumer_counter_limit) + dev_dbg_ratelimited(&pdev->dev, "Consumer counter does not match limit in RDQ\n"); + + err = mlxsw_pci_rdq_skb_alloc(mlxsw_pci, elem_info); + if (err) { + dev_err_ratelimited(&pdev->dev, "Failed to alloc skb for RDQ\n"); + goto out; + } + + mlxsw_pci_wqe_frag_unmap(mlxsw_pci, wqe, 0, DMA_FROM_DEVICE); + + if (mlxsw_pci_cqe_lag_get(cqe_v, cqe)) { + rx_info.is_lag = true; + rx_info.u.lag_id = mlxsw_pci_cqe_lag_id_get(cqe_v, cqe); + rx_info.lag_port_index = + mlxsw_pci_cqe_lag_subport_get(cqe_v, cqe); + } else { + rx_info.is_lag = false; + rx_info.u.sys_port = mlxsw_pci_cqe_system_port_get(cqe); + } + + rx_info.trap_id = mlxsw_pci_cqe_trap_id_get(cqe); + + if (rx_info.trap_id == MLXSW_TRAP_ID_DISCARD_INGRESS_ACL || + rx_info.trap_id == MLXSW_TRAP_ID_DISCARD_EGRESS_ACL) { + u32 cookie_index = 0; + + if (mlxsw_pci->max_cqe_ver >= MLXSW_PCI_CQE_V2) + cookie_index = mlxsw_pci_cqe2_user_def_val_orig_pkt_len_get(cqe); + mlxsw_skb_cb(skb)->rx_md_info.cookie_index = cookie_index; + } else if (rx_info.trap_id >= MLXSW_TRAP_ID_MIRROR_SESSION0 && + rx_info.trap_id <= MLXSW_TRAP_ID_MIRROR_SESSION7 && + mlxsw_pci->max_cqe_ver >= MLXSW_PCI_CQE_V2) { + rx_info.mirror_reason = mlxsw_pci_cqe2_mirror_reason_get(cqe); + mlxsw_pci_cqe_rdq_md_init(skb, cqe); + } else if (rx_info.trap_id == MLXSW_TRAP_ID_PKT_SAMPLE && + mlxsw_pci->max_cqe_ver >= MLXSW_PCI_CQE_V2) { + mlxsw_pci_cqe_rdq_md_tx_port_init(skb, cqe); + } + + mlxsw_pci_skb_cb_ts_set(mlxsw_pci, skb, cqe_v, cqe); + + byte_count = mlxsw_pci_cqe_byte_count_get(cqe); + if (mlxsw_pci_cqe_crc_get(cqe_v, cqe)) + byte_count -= ETH_FCS_LEN; + skb_put(skb, byte_count); + mlxsw_core_skb_receive(mlxsw_pci->core, skb, &rx_info); + +out: + /* Everything is set up, ring doorbell to pass elem to HW */ + q->producer_counter++; + mlxsw_pci_queue_doorbell_producer_ring(mlxsw_pci, q); + return; +} + +static char *mlxsw_pci_cq_sw_cqe_get(struct mlxsw_pci_queue *q) +{ + struct mlxsw_pci_queue_elem_info *elem_info; + char *elem; + bool owner_bit; + + elem_info = mlxsw_pci_queue_elem_info_consumer_get(q); + elem = elem_info->elem; + owner_bit = mlxsw_pci_cqe_owner_get(q->u.cq.v, elem); + if (mlxsw_pci_elem_hw_owned(q, owner_bit)) + return NULL; + q->consumer_counter++; + rmb(); /* make sure we read owned bit before the rest of elem */ + return elem; +} + +static void mlxsw_pci_cq_tasklet(struct tasklet_struct *t) +{ + struct mlxsw_pci_queue *q = from_tasklet(q, t, tasklet); + struct mlxsw_pci *mlxsw_pci = q->pci; + char *cqe; + int items = 0; + int credits = q->count >> 1; + + while ((cqe = mlxsw_pci_cq_sw_cqe_get(q))) { + u16 wqe_counter = mlxsw_pci_cqe_wqe_counter_get(cqe); + u8 sendq = mlxsw_pci_cqe_sr_get(q->u.cq.v, cqe); + u8 dqn = mlxsw_pci_cqe_dqn_get(q->u.cq.v, cqe); + char ncqe[MLXSW_PCI_CQE_SIZE_MAX]; + + memcpy(ncqe, cqe, q->elem_size); + mlxsw_pci_queue_doorbell_consumer_ring(mlxsw_pci, q); + + if (sendq) { + struct mlxsw_pci_queue *sdq; + + sdq = mlxsw_pci_sdq_get(mlxsw_pci, dqn); + mlxsw_pci_cqe_sdq_handle(mlxsw_pci, sdq, + wqe_counter, q->u.cq.v, ncqe); + q->u.cq.comp_sdq_count++; + } else { + struct mlxsw_pci_queue *rdq; + + rdq = mlxsw_pci_rdq_get(mlxsw_pci, dqn); + mlxsw_pci_cqe_rdq_handle(mlxsw_pci, rdq, + wqe_counter, q->u.cq.v, ncqe); + q->u.cq.comp_rdq_count++; + } + if (++items == credits) + break; + } + if (items) + mlxsw_pci_queue_doorbell_arm_consumer_ring(mlxsw_pci, q); +} + +static u16 mlxsw_pci_cq_elem_count(const struct mlxsw_pci_queue *q) +{ + return q->u.cq.v == MLXSW_PCI_CQE_V2 ? MLXSW_PCI_CQE2_COUNT : + MLXSW_PCI_CQE01_COUNT; +} + +static u8 mlxsw_pci_cq_elem_size(const struct mlxsw_pci_queue *q) +{ + return q->u.cq.v == MLXSW_PCI_CQE_V2 ? MLXSW_PCI_CQE2_SIZE : + MLXSW_PCI_CQE01_SIZE; +} + +static int mlxsw_pci_eq_init(struct mlxsw_pci *mlxsw_pci, char *mbox, + struct mlxsw_pci_queue *q) +{ + int i; + int err; + + q->consumer_counter = 0; + + for (i = 0; i < q->count; i++) { + char *elem = mlxsw_pci_queue_elem_get(q, i); + + mlxsw_pci_eqe_owner_set(elem, 1); + } + + mlxsw_cmd_mbox_sw2hw_eq_int_msix_set(mbox, 1); /* MSI-X used */ + mlxsw_cmd_mbox_sw2hw_eq_st_set(mbox, 1); /* armed */ + mlxsw_cmd_mbox_sw2hw_eq_log_eq_size_set(mbox, ilog2(q->count)); + for (i = 0; i < MLXSW_PCI_AQ_PAGES; i++) { + dma_addr_t mapaddr = __mlxsw_pci_queue_page_get(q, i); + + mlxsw_cmd_mbox_sw2hw_eq_pa_set(mbox, i, mapaddr); + } + err = mlxsw_cmd_sw2hw_eq(mlxsw_pci->core, mbox, q->num); + if (err) + return err; + mlxsw_pci_queue_doorbell_consumer_ring(mlxsw_pci, q); + mlxsw_pci_queue_doorbell_arm_consumer_ring(mlxsw_pci, q); + return 0; +} + +static void mlxsw_pci_eq_fini(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q) +{ + mlxsw_cmd_hw2sw_eq(mlxsw_pci->core, q->num); +} + +static void mlxsw_pci_eq_cmd_event(struct mlxsw_pci *mlxsw_pci, char *eqe) +{ + mlxsw_pci->cmd.comp.status = mlxsw_pci_eqe_cmd_status_get(eqe); + mlxsw_pci->cmd.comp.out_param = + ((u64) mlxsw_pci_eqe_cmd_out_param_h_get(eqe)) << 32 | + mlxsw_pci_eqe_cmd_out_param_l_get(eqe); + mlxsw_pci->cmd.wait_done = true; + wake_up(&mlxsw_pci->cmd.wait); +} + +static char *mlxsw_pci_eq_sw_eqe_get(struct mlxsw_pci_queue *q) +{ + struct mlxsw_pci_queue_elem_info *elem_info; + char *elem; + bool owner_bit; + + elem_info = mlxsw_pci_queue_elem_info_consumer_get(q); + elem = elem_info->elem; + owner_bit = mlxsw_pci_eqe_owner_get(elem); + if (mlxsw_pci_elem_hw_owned(q, owner_bit)) + return NULL; + q->consumer_counter++; + rmb(); /* make sure we read owned bit before the rest of elem */ + return elem; +} + +static void mlxsw_pci_eq_tasklet(struct tasklet_struct *t) +{ + struct mlxsw_pci_queue *q = from_tasklet(q, t, tasklet); + struct mlxsw_pci *mlxsw_pci = q->pci; + u8 cq_count = mlxsw_pci_cq_count(mlxsw_pci); + unsigned long active_cqns[BITS_TO_LONGS(MLXSW_PCI_CQS_MAX)]; + char *eqe; + u8 cqn; + bool cq_handle = false; + int items = 0; + int credits = q->count >> 1; + + memset(&active_cqns, 0, sizeof(active_cqns)); + + while ((eqe = mlxsw_pci_eq_sw_eqe_get(q))) { + + /* Command interface completion events are always received on + * queue MLXSW_PCI_EQ_ASYNC_NUM (EQ0) and completion events + * are mapped to queue MLXSW_PCI_EQ_COMP_NUM (EQ1). + */ + switch (q->num) { + case MLXSW_PCI_EQ_ASYNC_NUM: + mlxsw_pci_eq_cmd_event(mlxsw_pci, eqe); + q->u.eq.ev_cmd_count++; + break; + case MLXSW_PCI_EQ_COMP_NUM: + cqn = mlxsw_pci_eqe_cqn_get(eqe); + set_bit(cqn, active_cqns); + cq_handle = true; + q->u.eq.ev_comp_count++; + break; + default: + q->u.eq.ev_other_count++; + } + if (++items == credits) + break; + } + if (items) { + mlxsw_pci_queue_doorbell_consumer_ring(mlxsw_pci, q); + mlxsw_pci_queue_doorbell_arm_consumer_ring(mlxsw_pci, q); + } + + if (!cq_handle) + return; + for_each_set_bit(cqn, active_cqns, cq_count) { + q = mlxsw_pci_cq_get(mlxsw_pci, cqn); + mlxsw_pci_queue_tasklet_schedule(q); + } +} + +struct mlxsw_pci_queue_ops { + const char *name; + enum mlxsw_pci_queue_type type; + void (*pre_init)(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q); + int (*init)(struct mlxsw_pci *mlxsw_pci, char *mbox, + struct mlxsw_pci_queue *q); + void (*fini)(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_queue *q); + void (*tasklet)(struct tasklet_struct *t); + u16 (*elem_count_f)(const struct mlxsw_pci_queue *q); + u8 (*elem_size_f)(const struct mlxsw_pci_queue *q); + u16 elem_count; + u8 elem_size; +}; + +static const struct mlxsw_pci_queue_ops mlxsw_pci_sdq_ops = { + .type = MLXSW_PCI_QUEUE_TYPE_SDQ, + .init = mlxsw_pci_sdq_init, + .fini = mlxsw_pci_sdq_fini, + .elem_count = MLXSW_PCI_WQE_COUNT, + .elem_size = MLXSW_PCI_WQE_SIZE, +}; + +static const struct mlxsw_pci_queue_ops mlxsw_pci_rdq_ops = { + .type = MLXSW_PCI_QUEUE_TYPE_RDQ, + .init = mlxsw_pci_rdq_init, + .fini = mlxsw_pci_rdq_fini, + .elem_count = MLXSW_PCI_WQE_COUNT, + .elem_size = MLXSW_PCI_WQE_SIZE +}; + +static const struct mlxsw_pci_queue_ops mlxsw_pci_cq_ops = { + .type = MLXSW_PCI_QUEUE_TYPE_CQ, + .pre_init = mlxsw_pci_cq_pre_init, + .init = mlxsw_pci_cq_init, + .fini = mlxsw_pci_cq_fini, + .tasklet = mlxsw_pci_cq_tasklet, + .elem_count_f = mlxsw_pci_cq_elem_count, + .elem_size_f = mlxsw_pci_cq_elem_size +}; + +static const struct mlxsw_pci_queue_ops mlxsw_pci_eq_ops = { + .type = MLXSW_PCI_QUEUE_TYPE_EQ, + .init = mlxsw_pci_eq_init, + .fini = mlxsw_pci_eq_fini, + .tasklet = mlxsw_pci_eq_tasklet, + .elem_count = MLXSW_PCI_EQE_COUNT, + .elem_size = MLXSW_PCI_EQE_SIZE +}; + +static int mlxsw_pci_queue_init(struct mlxsw_pci *mlxsw_pci, char *mbox, + const struct mlxsw_pci_queue_ops *q_ops, + struct mlxsw_pci_queue *q, u8 q_num) +{ + struct mlxsw_pci_mem_item *mem_item = &q->mem_item; + int i; + int err; + + q->num = q_num; + if (q_ops->pre_init) + q_ops->pre_init(mlxsw_pci, q); + + spin_lock_init(&q->lock); + q->count = q_ops->elem_count_f ? q_ops->elem_count_f(q) : + q_ops->elem_count; + q->elem_size = q_ops->elem_size_f ? q_ops->elem_size_f(q) : + q_ops->elem_size; + q->type = q_ops->type; + q->pci = mlxsw_pci; + + if (q_ops->tasklet) + tasklet_setup(&q->tasklet, q_ops->tasklet); + + mem_item->size = MLXSW_PCI_AQ_SIZE; + mem_item->buf = dma_alloc_coherent(&mlxsw_pci->pdev->dev, + mem_item->size, &mem_item->mapaddr, + GFP_KERNEL); + if (!mem_item->buf) + return -ENOMEM; + + q->elem_info = kcalloc(q->count, sizeof(*q->elem_info), GFP_KERNEL); + if (!q->elem_info) { + err = -ENOMEM; + goto err_elem_info_alloc; + } + + /* Initialize dma mapped elements info elem_info for + * future easy access. + */ + for (i = 0; i < q->count; i++) { + struct mlxsw_pci_queue_elem_info *elem_info; + + elem_info = mlxsw_pci_queue_elem_info_get(q, i); + elem_info->elem = + __mlxsw_pci_queue_elem_get(q, q->elem_size, i); + } + + mlxsw_cmd_mbox_zero(mbox); + err = q_ops->init(mlxsw_pci, mbox, q); + if (err) + goto err_q_ops_init; + return 0; + +err_q_ops_init: + kfree(q->elem_info); +err_elem_info_alloc: + dma_free_coherent(&mlxsw_pci->pdev->dev, mem_item->size, + mem_item->buf, mem_item->mapaddr); + return err; +} + +static void mlxsw_pci_queue_fini(struct mlxsw_pci *mlxsw_pci, + const struct mlxsw_pci_queue_ops *q_ops, + struct mlxsw_pci_queue *q) +{ + struct mlxsw_pci_mem_item *mem_item = &q->mem_item; + + q_ops->fini(mlxsw_pci, q); + kfree(q->elem_info); + dma_free_coherent(&mlxsw_pci->pdev->dev, mem_item->size, + mem_item->buf, mem_item->mapaddr); +} + +static int mlxsw_pci_queue_group_init(struct mlxsw_pci *mlxsw_pci, char *mbox, + const struct mlxsw_pci_queue_ops *q_ops, + u8 num_qs) +{ + struct mlxsw_pci_queue_type_group *queue_group; + int i; + int err; + + queue_group = mlxsw_pci_queue_type_group_get(mlxsw_pci, q_ops->type); + queue_group->q = kcalloc(num_qs, sizeof(*queue_group->q), GFP_KERNEL); + if (!queue_group->q) + return -ENOMEM; + + for (i = 0; i < num_qs; i++) { + err = mlxsw_pci_queue_init(mlxsw_pci, mbox, q_ops, + &queue_group->q[i], i); + if (err) + goto err_queue_init; + } + queue_group->count = num_qs; + + return 0; + +err_queue_init: + for (i--; i >= 0; i--) + mlxsw_pci_queue_fini(mlxsw_pci, q_ops, &queue_group->q[i]); + kfree(queue_group->q); + return err; +} + +static void mlxsw_pci_queue_group_fini(struct mlxsw_pci *mlxsw_pci, + const struct mlxsw_pci_queue_ops *q_ops) +{ + struct mlxsw_pci_queue_type_group *queue_group; + int i; + + queue_group = mlxsw_pci_queue_type_group_get(mlxsw_pci, q_ops->type); + for (i = 0; i < queue_group->count; i++) + mlxsw_pci_queue_fini(mlxsw_pci, q_ops, &queue_group->q[i]); + kfree(queue_group->q); +} + +static int mlxsw_pci_aqs_init(struct mlxsw_pci *mlxsw_pci, char *mbox) +{ + struct pci_dev *pdev = mlxsw_pci->pdev; + u8 num_sdqs; + u8 sdq_log2sz; + u8 num_rdqs; + u8 rdq_log2sz; + u8 num_cqs; + u8 cq_log2sz; + u8 cqv2_log2sz; + u8 num_eqs; + u8 eq_log2sz; + int err; + + mlxsw_cmd_mbox_zero(mbox); + err = mlxsw_cmd_query_aq_cap(mlxsw_pci->core, mbox); + if (err) + return err; + + num_sdqs = mlxsw_cmd_mbox_query_aq_cap_max_num_sdqs_get(mbox); + sdq_log2sz = mlxsw_cmd_mbox_query_aq_cap_log_max_sdq_sz_get(mbox); + num_rdqs = mlxsw_cmd_mbox_query_aq_cap_max_num_rdqs_get(mbox); + rdq_log2sz = mlxsw_cmd_mbox_query_aq_cap_log_max_rdq_sz_get(mbox); + num_cqs = mlxsw_cmd_mbox_query_aq_cap_max_num_cqs_get(mbox); + cq_log2sz = mlxsw_cmd_mbox_query_aq_cap_log_max_cq_sz_get(mbox); + cqv2_log2sz = mlxsw_cmd_mbox_query_aq_cap_log_max_cqv2_sz_get(mbox); + num_eqs = mlxsw_cmd_mbox_query_aq_cap_max_num_eqs_get(mbox); + eq_log2sz = mlxsw_cmd_mbox_query_aq_cap_log_max_eq_sz_get(mbox); + + if (num_sdqs + num_rdqs > num_cqs || + num_sdqs < MLXSW_PCI_SDQS_MIN || + num_cqs > MLXSW_PCI_CQS_MAX || num_eqs != MLXSW_PCI_EQS_COUNT) { + dev_err(&pdev->dev, "Unsupported number of queues\n"); + return -EINVAL; + } + + if ((1 << sdq_log2sz != MLXSW_PCI_WQE_COUNT) || + (1 << rdq_log2sz != MLXSW_PCI_WQE_COUNT) || + (1 << cq_log2sz != MLXSW_PCI_CQE01_COUNT) || + (mlxsw_pci->max_cqe_ver == MLXSW_PCI_CQE_V2 && + (1 << cqv2_log2sz != MLXSW_PCI_CQE2_COUNT)) || + (1 << eq_log2sz != MLXSW_PCI_EQE_COUNT)) { + dev_err(&pdev->dev, "Unsupported number of async queue descriptors\n"); + return -EINVAL; + } + + mlxsw_pci->num_sdq_cqs = num_sdqs; + + err = mlxsw_pci_queue_group_init(mlxsw_pci, mbox, &mlxsw_pci_eq_ops, + num_eqs); + if (err) { + dev_err(&pdev->dev, "Failed to initialize event queues\n"); + return err; + } + + err = mlxsw_pci_queue_group_init(mlxsw_pci, mbox, &mlxsw_pci_cq_ops, + num_cqs); + if (err) { + dev_err(&pdev->dev, "Failed to initialize completion queues\n"); + goto err_cqs_init; + } + + err = mlxsw_pci_queue_group_init(mlxsw_pci, mbox, &mlxsw_pci_sdq_ops, + num_sdqs); + if (err) { + dev_err(&pdev->dev, "Failed to initialize send descriptor queues\n"); + goto err_sdqs_init; + } + + err = mlxsw_pci_queue_group_init(mlxsw_pci, mbox, &mlxsw_pci_rdq_ops, + num_rdqs); + if (err) { + dev_err(&pdev->dev, "Failed to initialize receive descriptor queues\n"); + goto err_rdqs_init; + } + + /* We have to poll in command interface until queues are initialized */ + mlxsw_pci->cmd.nopoll = true; + return 0; + +err_rdqs_init: + mlxsw_pci_queue_group_fini(mlxsw_pci, &mlxsw_pci_sdq_ops); +err_sdqs_init: + mlxsw_pci_queue_group_fini(mlxsw_pci, &mlxsw_pci_cq_ops); +err_cqs_init: + mlxsw_pci_queue_group_fini(mlxsw_pci, &mlxsw_pci_eq_ops); + return err; +} + +static void mlxsw_pci_aqs_fini(struct mlxsw_pci *mlxsw_pci) +{ + mlxsw_pci->cmd.nopoll = false; + mlxsw_pci_queue_group_fini(mlxsw_pci, &mlxsw_pci_rdq_ops); + mlxsw_pci_queue_group_fini(mlxsw_pci, &mlxsw_pci_sdq_ops); + mlxsw_pci_queue_group_fini(mlxsw_pci, &mlxsw_pci_cq_ops); + mlxsw_pci_queue_group_fini(mlxsw_pci, &mlxsw_pci_eq_ops); +} + +static void +mlxsw_pci_config_profile_swid_config(struct mlxsw_pci *mlxsw_pci, + char *mbox, int index, + const struct mlxsw_swid_config *swid) +{ + u8 mask = 0; + + if (swid->used_type) { + mlxsw_cmd_mbox_config_profile_swid_config_type_set( + mbox, index, swid->type); + mask |= 1; + } + if (swid->used_properties) { + mlxsw_cmd_mbox_config_profile_swid_config_properties_set( + mbox, index, swid->properties); + mask |= 2; + } + mlxsw_cmd_mbox_config_profile_swid_config_mask_set(mbox, index, mask); +} + +static int +mlxsw_pci_profile_get_kvd_sizes(const struct mlxsw_pci *mlxsw_pci, + const struct mlxsw_config_profile *profile, + struct mlxsw_res *res) +{ + u64 single_size, double_size, linear_size; + int err; + + err = mlxsw_core_kvd_sizes_get(mlxsw_pci->core, profile, + &single_size, &double_size, + &linear_size); + if (err) + return err; + + MLXSW_RES_SET(res, KVD_SINGLE_SIZE, single_size); + MLXSW_RES_SET(res, KVD_DOUBLE_SIZE, double_size); + MLXSW_RES_SET(res, KVD_LINEAR_SIZE, linear_size); + + return 0; +} + +static int mlxsw_pci_config_profile(struct mlxsw_pci *mlxsw_pci, char *mbox, + const struct mlxsw_config_profile *profile, + struct mlxsw_res *res) +{ + int i; + int err; + + mlxsw_cmd_mbox_zero(mbox); + + if (profile->used_max_vepa_channels) { + mlxsw_cmd_mbox_config_profile_set_max_vepa_channels_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_max_vepa_channels_set( + mbox, profile->max_vepa_channels); + } + if (profile->used_max_lag) { + mlxsw_cmd_mbox_config_profile_set_max_lag_set(mbox, 1); + mlxsw_cmd_mbox_config_profile_max_lag_set(mbox, + profile->max_lag); + } + if (profile->used_max_mid) { + mlxsw_cmd_mbox_config_profile_set_max_mid_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_max_mid_set( + mbox, profile->max_mid); + } + if (profile->used_max_pgt) { + mlxsw_cmd_mbox_config_profile_set_max_pgt_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_max_pgt_set( + mbox, profile->max_pgt); + } + if (profile->used_max_system_port) { + mlxsw_cmd_mbox_config_profile_set_max_system_port_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_max_system_port_set( + mbox, profile->max_system_port); + } + if (profile->used_max_vlan_groups) { + mlxsw_cmd_mbox_config_profile_set_max_vlan_groups_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_max_vlan_groups_set( + mbox, profile->max_vlan_groups); + } + if (profile->used_max_regions) { + mlxsw_cmd_mbox_config_profile_set_max_regions_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_max_regions_set( + mbox, profile->max_regions); + } + if (profile->used_flood_tables) { + mlxsw_cmd_mbox_config_profile_set_flood_tables_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_max_flood_tables_set( + mbox, profile->max_flood_tables); + mlxsw_cmd_mbox_config_profile_max_vid_flood_tables_set( + mbox, profile->max_vid_flood_tables); + mlxsw_cmd_mbox_config_profile_max_fid_offset_flood_tables_set( + mbox, profile->max_fid_offset_flood_tables); + mlxsw_cmd_mbox_config_profile_fid_offset_flood_table_size_set( + mbox, profile->fid_offset_flood_table_size); + mlxsw_cmd_mbox_config_profile_max_fid_flood_tables_set( + mbox, profile->max_fid_flood_tables); + mlxsw_cmd_mbox_config_profile_fid_flood_table_size_set( + mbox, profile->fid_flood_table_size); + } + if (profile->used_flood_mode) { + mlxsw_cmd_mbox_config_profile_set_flood_mode_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_flood_mode_set( + mbox, profile->flood_mode); + } + if (profile->used_max_ib_mc) { + mlxsw_cmd_mbox_config_profile_set_max_ib_mc_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_max_ib_mc_set( + mbox, profile->max_ib_mc); + } + if (profile->used_max_pkey) { + mlxsw_cmd_mbox_config_profile_set_max_pkey_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_max_pkey_set( + mbox, profile->max_pkey); + } + if (profile->used_ar_sec) { + mlxsw_cmd_mbox_config_profile_set_ar_sec_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_ar_sec_set( + mbox, profile->ar_sec); + } + if (profile->used_adaptive_routing_group_cap) { + mlxsw_cmd_mbox_config_profile_set_adaptive_routing_group_cap_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_adaptive_routing_group_cap_set( + mbox, profile->adaptive_routing_group_cap); + } + if (profile->used_ubridge) { + mlxsw_cmd_mbox_config_profile_set_ubridge_set(mbox, 1); + mlxsw_cmd_mbox_config_profile_ubridge_set(mbox, + profile->ubridge); + } + if (profile->used_kvd_sizes && MLXSW_RES_VALID(res, KVD_SIZE)) { + err = mlxsw_pci_profile_get_kvd_sizes(mlxsw_pci, profile, res); + if (err) + return err; + + mlxsw_cmd_mbox_config_profile_set_kvd_linear_size_set(mbox, 1); + mlxsw_cmd_mbox_config_profile_kvd_linear_size_set(mbox, + MLXSW_RES_GET(res, KVD_LINEAR_SIZE)); + mlxsw_cmd_mbox_config_profile_set_kvd_hash_single_size_set(mbox, + 1); + mlxsw_cmd_mbox_config_profile_kvd_hash_single_size_set(mbox, + MLXSW_RES_GET(res, KVD_SINGLE_SIZE)); + mlxsw_cmd_mbox_config_profile_set_kvd_hash_double_size_set( + mbox, 1); + mlxsw_cmd_mbox_config_profile_kvd_hash_double_size_set(mbox, + MLXSW_RES_GET(res, KVD_DOUBLE_SIZE)); + } + + for (i = 0; i < MLXSW_CONFIG_PROFILE_SWID_COUNT; i++) + mlxsw_pci_config_profile_swid_config(mlxsw_pci, mbox, i, + &profile->swid_config[i]); + + if (mlxsw_pci->max_cqe_ver > MLXSW_PCI_CQE_V0) { + mlxsw_cmd_mbox_config_profile_set_cqe_version_set(mbox, 1); + mlxsw_cmd_mbox_config_profile_cqe_version_set(mbox, 1); + } + + if (profile->used_cqe_time_stamp_type) { + mlxsw_cmd_mbox_config_profile_set_cqe_time_stamp_type_set(mbox, + 1); + mlxsw_cmd_mbox_config_profile_cqe_time_stamp_type_set(mbox, + profile->cqe_time_stamp_type); + } + + return mlxsw_cmd_config_profile_set(mlxsw_pci->core, mbox); +} + +static int mlxsw_pci_boardinfo(struct mlxsw_pci *mlxsw_pci, char *mbox) +{ + struct mlxsw_bus_info *bus_info = &mlxsw_pci->bus_info; + int err; + + mlxsw_cmd_mbox_zero(mbox); + err = mlxsw_cmd_boardinfo(mlxsw_pci->core, mbox); + if (err) + return err; + mlxsw_cmd_mbox_boardinfo_vsd_memcpy_from(mbox, bus_info->vsd); + mlxsw_cmd_mbox_boardinfo_psid_memcpy_from(mbox, bus_info->psid); + return 0; +} + +static int mlxsw_pci_fw_area_init(struct mlxsw_pci *mlxsw_pci, char *mbox, + u16 num_pages) +{ + struct mlxsw_pci_mem_item *mem_item; + int nent = 0; + int i; + int err; + + mlxsw_pci->fw_area.items = kcalloc(num_pages, sizeof(*mem_item), + GFP_KERNEL); + if (!mlxsw_pci->fw_area.items) + return -ENOMEM; + mlxsw_pci->fw_area.count = num_pages; + + mlxsw_cmd_mbox_zero(mbox); + for (i = 0; i < num_pages; i++) { + mem_item = &mlxsw_pci->fw_area.items[i]; + + mem_item->size = MLXSW_PCI_PAGE_SIZE; + mem_item->buf = dma_alloc_coherent(&mlxsw_pci->pdev->dev, + mem_item->size, + &mem_item->mapaddr, GFP_KERNEL); + if (!mem_item->buf) { + err = -ENOMEM; + goto err_alloc; + } + mlxsw_cmd_mbox_map_fa_pa_set(mbox, nent, mem_item->mapaddr); + mlxsw_cmd_mbox_map_fa_log2size_set(mbox, nent, 0); /* 1 page */ + if (++nent == MLXSW_CMD_MAP_FA_VPM_ENTRIES_MAX) { + err = mlxsw_cmd_map_fa(mlxsw_pci->core, mbox, nent); + if (err) + goto err_cmd_map_fa; + nent = 0; + mlxsw_cmd_mbox_zero(mbox); + } + } + + if (nent) { + err = mlxsw_cmd_map_fa(mlxsw_pci->core, mbox, nent); + if (err) + goto err_cmd_map_fa; + } + + return 0; + +err_cmd_map_fa: +err_alloc: + for (i--; i >= 0; i--) { + mem_item = &mlxsw_pci->fw_area.items[i]; + + dma_free_coherent(&mlxsw_pci->pdev->dev, mem_item->size, + mem_item->buf, mem_item->mapaddr); + } + kfree(mlxsw_pci->fw_area.items); + return err; +} + +static void mlxsw_pci_fw_area_fini(struct mlxsw_pci *mlxsw_pci) +{ + struct mlxsw_pci_mem_item *mem_item; + int i; + + mlxsw_cmd_unmap_fa(mlxsw_pci->core); + + for (i = 0; i < mlxsw_pci->fw_area.count; i++) { + mem_item = &mlxsw_pci->fw_area.items[i]; + + dma_free_coherent(&mlxsw_pci->pdev->dev, mem_item->size, + mem_item->buf, mem_item->mapaddr); + } + kfree(mlxsw_pci->fw_area.items); +} + +static irqreturn_t mlxsw_pci_eq_irq_handler(int irq, void *dev_id) +{ + struct mlxsw_pci *mlxsw_pci = dev_id; + struct mlxsw_pci_queue *q; + int i; + + for (i = 0; i < MLXSW_PCI_EQS_COUNT; i++) { + q = mlxsw_pci_eq_get(mlxsw_pci, i); + mlxsw_pci_queue_tasklet_schedule(q); + } + return IRQ_HANDLED; +} + +static int mlxsw_pci_mbox_alloc(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_mem_item *mbox) +{ + struct pci_dev *pdev = mlxsw_pci->pdev; + int err = 0; + + mbox->size = MLXSW_CMD_MBOX_SIZE; + mbox->buf = dma_alloc_coherent(&pdev->dev, MLXSW_CMD_MBOX_SIZE, + &mbox->mapaddr, GFP_KERNEL); + if (!mbox->buf) { + dev_err(&pdev->dev, "Failed allocating memory for mailbox\n"); + err = -ENOMEM; + } + + return err; +} + +static void mlxsw_pci_mbox_free(struct mlxsw_pci *mlxsw_pci, + struct mlxsw_pci_mem_item *mbox) +{ + struct pci_dev *pdev = mlxsw_pci->pdev; + + dma_free_coherent(&pdev->dev, MLXSW_CMD_MBOX_SIZE, mbox->buf, + mbox->mapaddr); +} + +static int mlxsw_pci_sys_ready_wait(struct mlxsw_pci *mlxsw_pci, + const struct pci_device_id *id, + u32 *p_sys_status) +{ + unsigned long end; + u32 val; + + /* We must wait for the HW to become responsive. */ + msleep(MLXSW_PCI_SW_RESET_WAIT_MSECS); + + end = jiffies + msecs_to_jiffies(MLXSW_PCI_SW_RESET_TIMEOUT_MSECS); + do { + val = mlxsw_pci_read32(mlxsw_pci, FW_READY); + if ((val & MLXSW_PCI_FW_READY_MASK) == MLXSW_PCI_FW_READY_MAGIC) + return 0; + cond_resched(); + } while (time_before(jiffies, end)); + + *p_sys_status = val & MLXSW_PCI_FW_READY_MASK; + + return -EBUSY; +} + +static int mlxsw_pci_sw_reset(struct mlxsw_pci *mlxsw_pci, + const struct pci_device_id *id) +{ + struct pci_dev *pdev = mlxsw_pci->pdev; + char mrsr_pl[MLXSW_REG_MRSR_LEN]; + u32 sys_status; + int err; + + err = mlxsw_pci_sys_ready_wait(mlxsw_pci, id, &sys_status); + if (err) { + dev_err(&pdev->dev, "Failed to reach system ready status before reset. Status is 0x%x\n", + sys_status); + return err; + } + + mlxsw_reg_mrsr_pack(mrsr_pl); + err = mlxsw_reg_write(mlxsw_pci->core, MLXSW_REG(mrsr), mrsr_pl); + if (err) + return err; + + err = mlxsw_pci_sys_ready_wait(mlxsw_pci, id, &sys_status); + if (err) { + dev_err(&pdev->dev, "Failed to reach system ready status after reset. Status is 0x%x\n", + sys_status); + return err; + } + + return 0; +} + +static int mlxsw_pci_alloc_irq_vectors(struct mlxsw_pci *mlxsw_pci) +{ + int err; + + err = pci_alloc_irq_vectors(mlxsw_pci->pdev, 1, 1, PCI_IRQ_MSIX); + if (err < 0) + dev_err(&mlxsw_pci->pdev->dev, "MSI-X init failed\n"); + return err; +} + +static void mlxsw_pci_free_irq_vectors(struct mlxsw_pci *mlxsw_pci) +{ + pci_free_irq_vectors(mlxsw_pci->pdev); +} + +static int mlxsw_pci_init(void *bus_priv, struct mlxsw_core *mlxsw_core, + const struct mlxsw_config_profile *profile, + struct mlxsw_res *res) +{ + struct mlxsw_pci *mlxsw_pci = bus_priv; + struct pci_dev *pdev = mlxsw_pci->pdev; + char *mbox; + u16 num_pages; + int err; + + mlxsw_pci->core = mlxsw_core; + + mbox = mlxsw_cmd_mbox_alloc(); + if (!mbox) + return -ENOMEM; + + err = mlxsw_pci_sw_reset(mlxsw_pci, mlxsw_pci->id); + if (err) + goto err_sw_reset; + + err = mlxsw_pci_alloc_irq_vectors(mlxsw_pci); + if (err < 0) { + dev_err(&pdev->dev, "MSI-X init failed\n"); + goto err_alloc_irq; + } + + err = mlxsw_cmd_query_fw(mlxsw_core, mbox); + if (err) + goto err_query_fw; + + mlxsw_pci->bus_info.fw_rev.major = + mlxsw_cmd_mbox_query_fw_fw_rev_major_get(mbox); + mlxsw_pci->bus_info.fw_rev.minor = + mlxsw_cmd_mbox_query_fw_fw_rev_minor_get(mbox); + mlxsw_pci->bus_info.fw_rev.subminor = + mlxsw_cmd_mbox_query_fw_fw_rev_subminor_get(mbox); + + if (mlxsw_cmd_mbox_query_fw_cmd_interface_rev_get(mbox) != 1) { + dev_err(&pdev->dev, "Unsupported cmd interface revision ID queried from hw\n"); + err = -EINVAL; + goto err_iface_rev; + } + if (mlxsw_cmd_mbox_query_fw_doorbell_page_bar_get(mbox) != 0) { + dev_err(&pdev->dev, "Unsupported doorbell page bar queried from hw\n"); + err = -EINVAL; + goto err_doorbell_page_bar; + } + + mlxsw_pci->doorbell_offset = + mlxsw_cmd_mbox_query_fw_doorbell_page_offset_get(mbox); + + if (mlxsw_cmd_mbox_query_fw_fr_rn_clk_bar_get(mbox) != 0) { + dev_err(&pdev->dev, "Unsupported free running clock BAR queried from hw\n"); + err = -EINVAL; + goto err_fr_rn_clk_bar; + } + + mlxsw_pci->free_running_clock_offset = + mlxsw_cmd_mbox_query_fw_free_running_clock_offset_get(mbox); + + if (mlxsw_cmd_mbox_query_fw_utc_sec_bar_get(mbox) != 0) { + dev_err(&pdev->dev, "Unsupported UTC sec BAR queried from hw\n"); + err = -EINVAL; + goto err_utc_sec_bar; + } + + mlxsw_pci->utc_sec_offset = + mlxsw_cmd_mbox_query_fw_utc_sec_offset_get(mbox); + + if (mlxsw_cmd_mbox_query_fw_utc_nsec_bar_get(mbox) != 0) { + dev_err(&pdev->dev, "Unsupported UTC nsec BAR queried from hw\n"); + err = -EINVAL; + goto err_utc_nsec_bar; + } + + mlxsw_pci->utc_nsec_offset = + mlxsw_cmd_mbox_query_fw_utc_nsec_offset_get(mbox); + + num_pages = mlxsw_cmd_mbox_query_fw_fw_pages_get(mbox); + err = mlxsw_pci_fw_area_init(mlxsw_pci, mbox, num_pages); + if (err) + goto err_fw_area_init; + + err = mlxsw_pci_boardinfo(mlxsw_pci, mbox); + if (err) + goto err_boardinfo; + + err = mlxsw_core_resources_query(mlxsw_core, mbox, res); + if (err) + goto err_query_resources; + + if (MLXSW_CORE_RES_VALID(mlxsw_core, CQE_V2) && + MLXSW_CORE_RES_GET(mlxsw_core, CQE_V2)) + mlxsw_pci->max_cqe_ver = MLXSW_PCI_CQE_V2; + else if (MLXSW_CORE_RES_VALID(mlxsw_core, CQE_V1) && + MLXSW_CORE_RES_GET(mlxsw_core, CQE_V1)) + mlxsw_pci->max_cqe_ver = MLXSW_PCI_CQE_V1; + else if ((MLXSW_CORE_RES_VALID(mlxsw_core, CQE_V0) && + MLXSW_CORE_RES_GET(mlxsw_core, CQE_V0)) || + !MLXSW_CORE_RES_VALID(mlxsw_core, CQE_V0)) { + mlxsw_pci->max_cqe_ver = MLXSW_PCI_CQE_V0; + } else { + dev_err(&pdev->dev, "Invalid supported CQE version combination reported\n"); + goto err_cqe_v_check; + } + + err = mlxsw_pci_config_profile(mlxsw_pci, mbox, profile, res); + if (err) + goto err_config_profile; + + /* Some resources depend on unified bridge model, which is configured + * as part of config_profile. Query the resources again to get correct + * values. + */ + err = mlxsw_core_resources_query(mlxsw_core, mbox, res); + if (err) + goto err_requery_resources; + + err = mlxsw_pci_aqs_init(mlxsw_pci, mbox); + if (err) + goto err_aqs_init; + + err = request_irq(pci_irq_vector(pdev, 0), + mlxsw_pci_eq_irq_handler, 0, + mlxsw_pci->bus_info.device_kind, mlxsw_pci); + if (err) { + dev_err(&pdev->dev, "IRQ request failed\n"); + goto err_request_eq_irq; + } + + goto mbox_put; + +err_request_eq_irq: + mlxsw_pci_aqs_fini(mlxsw_pci); +err_aqs_init: +err_requery_resources: +err_config_profile: +err_cqe_v_check: +err_query_resources: +err_boardinfo: + mlxsw_pci_fw_area_fini(mlxsw_pci); +err_fw_area_init: +err_utc_nsec_bar: +err_utc_sec_bar: +err_fr_rn_clk_bar: +err_doorbell_page_bar: +err_iface_rev: +err_query_fw: + mlxsw_pci_free_irq_vectors(mlxsw_pci); +err_alloc_irq: +err_sw_reset: +mbox_put: + mlxsw_cmd_mbox_free(mbox); + return err; +} + +static void mlxsw_pci_fini(void *bus_priv) +{ + struct mlxsw_pci *mlxsw_pci = bus_priv; + + free_irq(pci_irq_vector(mlxsw_pci->pdev, 0), mlxsw_pci); + mlxsw_pci_aqs_fini(mlxsw_pci); + mlxsw_pci_fw_area_fini(mlxsw_pci); + mlxsw_pci_free_irq_vectors(mlxsw_pci); +} + +static struct mlxsw_pci_queue * +mlxsw_pci_sdq_pick(struct mlxsw_pci *mlxsw_pci, + const struct mlxsw_tx_info *tx_info) +{ + u8 ctl_sdq_count = mlxsw_pci_sdq_count(mlxsw_pci) - 1; + u8 sdqn; + + if (tx_info->is_emad) { + sdqn = MLXSW_PCI_SDQ_EMAD_INDEX; + } else { + BUILD_BUG_ON(MLXSW_PCI_SDQ_EMAD_INDEX != 0); + sdqn = 1 + (tx_info->local_port % ctl_sdq_count); + } + + return mlxsw_pci_sdq_get(mlxsw_pci, sdqn); +} + +static bool mlxsw_pci_skb_transmit_busy(void *bus_priv, + const struct mlxsw_tx_info *tx_info) +{ + struct mlxsw_pci *mlxsw_pci = bus_priv; + struct mlxsw_pci_queue *q = mlxsw_pci_sdq_pick(mlxsw_pci, tx_info); + + return !mlxsw_pci_queue_elem_info_producer_get(q); +} + +static int mlxsw_pci_skb_transmit(void *bus_priv, struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info) +{ + struct mlxsw_pci *mlxsw_pci = bus_priv; + struct mlxsw_pci_queue *q; + struct mlxsw_pci_queue_elem_info *elem_info; + char *wqe; + int i; + int err; + + if (skb_shinfo(skb)->nr_frags > MLXSW_PCI_WQE_SG_ENTRIES - 1) { + err = skb_linearize(skb); + if (err) + return err; + } + + q = mlxsw_pci_sdq_pick(mlxsw_pci, tx_info); + spin_lock_bh(&q->lock); + elem_info = mlxsw_pci_queue_elem_info_producer_get(q); + if (!elem_info) { + /* queue is full */ + err = -EAGAIN; + goto unlock; + } + mlxsw_skb_cb(skb)->tx_info = *tx_info; + elem_info->u.sdq.skb = skb; + + wqe = elem_info->elem; + mlxsw_pci_wqe_c_set(wqe, 1); /* always report completion */ + mlxsw_pci_wqe_lp_set(wqe, 0); + mlxsw_pci_wqe_type_set(wqe, MLXSW_PCI_WQE_TYPE_ETHERNET); + + err = mlxsw_pci_wqe_frag_map(mlxsw_pci, wqe, 0, skb->data, + skb_headlen(skb), DMA_TO_DEVICE); + if (err) + goto unlock; + + for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) { + const skb_frag_t *frag = &skb_shinfo(skb)->frags[i]; + + err = mlxsw_pci_wqe_frag_map(mlxsw_pci, wqe, i + 1, + skb_frag_address(frag), + skb_frag_size(frag), + DMA_TO_DEVICE); + if (err) + goto unmap_frags; + } + + if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)) + skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS; + + /* Set unused sq entries byte count to zero. */ + for (i++; i < MLXSW_PCI_WQE_SG_ENTRIES; i++) + mlxsw_pci_wqe_byte_count_set(wqe, i, 0); + + /* Everything is set up, ring producer doorbell to get HW going */ + q->producer_counter++; + mlxsw_pci_queue_doorbell_producer_ring(mlxsw_pci, q); + + goto unlock; + +unmap_frags: + for (; i >= 0; i--) + mlxsw_pci_wqe_frag_unmap(mlxsw_pci, wqe, i, DMA_TO_DEVICE); +unlock: + spin_unlock_bh(&q->lock); + return err; +} + +static int mlxsw_pci_cmd_exec(void *bus_priv, u16 opcode, u8 opcode_mod, + u32 in_mod, bool out_mbox_direct, + char *in_mbox, size_t in_mbox_size, + char *out_mbox, size_t out_mbox_size, + u8 *p_status) +{ + struct mlxsw_pci *mlxsw_pci = bus_priv; + dma_addr_t in_mapaddr = 0, out_mapaddr = 0; + bool evreq = mlxsw_pci->cmd.nopoll; + unsigned long timeout = msecs_to_jiffies(MLXSW_PCI_CIR_TIMEOUT_MSECS); + bool *p_wait_done = &mlxsw_pci->cmd.wait_done; + int err; + + *p_status = MLXSW_CMD_STATUS_OK; + + err = mutex_lock_interruptible(&mlxsw_pci->cmd.lock); + if (err) + return err; + + if (in_mbox) { + memcpy(mlxsw_pci->cmd.in_mbox.buf, in_mbox, in_mbox_size); + in_mapaddr = mlxsw_pci->cmd.in_mbox.mapaddr; + } + mlxsw_pci_write32(mlxsw_pci, CIR_IN_PARAM_HI, upper_32_bits(in_mapaddr)); + mlxsw_pci_write32(mlxsw_pci, CIR_IN_PARAM_LO, lower_32_bits(in_mapaddr)); + + if (out_mbox) + out_mapaddr = mlxsw_pci->cmd.out_mbox.mapaddr; + mlxsw_pci_write32(mlxsw_pci, CIR_OUT_PARAM_HI, upper_32_bits(out_mapaddr)); + mlxsw_pci_write32(mlxsw_pci, CIR_OUT_PARAM_LO, lower_32_bits(out_mapaddr)); + + mlxsw_pci_write32(mlxsw_pci, CIR_IN_MODIFIER, in_mod); + mlxsw_pci_write32(mlxsw_pci, CIR_TOKEN, 0); + + *p_wait_done = false; + + wmb(); /* all needs to be written before we write control register */ + mlxsw_pci_write32(mlxsw_pci, CIR_CTRL, + MLXSW_PCI_CIR_CTRL_GO_BIT | + (evreq ? MLXSW_PCI_CIR_CTRL_EVREQ_BIT : 0) | + (opcode_mod << MLXSW_PCI_CIR_CTRL_OPCODE_MOD_SHIFT) | + opcode); + + if (!evreq) { + unsigned long end; + + end = jiffies + timeout; + do { + u32 ctrl = mlxsw_pci_read32(mlxsw_pci, CIR_CTRL); + + if (!(ctrl & MLXSW_PCI_CIR_CTRL_GO_BIT)) { + *p_wait_done = true; + *p_status = ctrl >> MLXSW_PCI_CIR_CTRL_STATUS_SHIFT; + break; + } + cond_resched(); + } while (time_before(jiffies, end)); + } else { + wait_event_timeout(mlxsw_pci->cmd.wait, *p_wait_done, timeout); + *p_status = mlxsw_pci->cmd.comp.status; + } + + err = 0; + if (*p_wait_done) { + if (*p_status) + err = -EIO; + } else { + err = -ETIMEDOUT; + } + + if (!err && out_mbox && out_mbox_direct) { + /* Some commands don't use output param as address to mailbox + * but they store output directly into registers. In that case, + * copy registers into mbox buffer. + */ + __be32 tmp; + + if (!evreq) { + tmp = cpu_to_be32(mlxsw_pci_read32(mlxsw_pci, + CIR_OUT_PARAM_HI)); + memcpy(out_mbox, &tmp, sizeof(tmp)); + tmp = cpu_to_be32(mlxsw_pci_read32(mlxsw_pci, + CIR_OUT_PARAM_LO)); + memcpy(out_mbox + sizeof(tmp), &tmp, sizeof(tmp)); + } + } else if (!err && out_mbox) { + memcpy(out_mbox, mlxsw_pci->cmd.out_mbox.buf, out_mbox_size); + } + + mutex_unlock(&mlxsw_pci->cmd.lock); + + return err; +} + +static u32 mlxsw_pci_read_frc_h(void *bus_priv) +{ + struct mlxsw_pci *mlxsw_pci = bus_priv; + u64 frc_offset_h; + + frc_offset_h = mlxsw_pci->free_running_clock_offset; + return mlxsw_pci_read32_off(mlxsw_pci, frc_offset_h); +} + +static u32 mlxsw_pci_read_frc_l(void *bus_priv) +{ + struct mlxsw_pci *mlxsw_pci = bus_priv; + u64 frc_offset_l; + + frc_offset_l = mlxsw_pci->free_running_clock_offset + 4; + return mlxsw_pci_read32_off(mlxsw_pci, frc_offset_l); +} + +static u32 mlxsw_pci_read_utc_sec(void *bus_priv) +{ + struct mlxsw_pci *mlxsw_pci = bus_priv; + + return mlxsw_pci_read32_off(mlxsw_pci, mlxsw_pci->utc_sec_offset); +} + +static u32 mlxsw_pci_read_utc_nsec(void *bus_priv) +{ + struct mlxsw_pci *mlxsw_pci = bus_priv; + + return mlxsw_pci_read32_off(mlxsw_pci, mlxsw_pci->utc_nsec_offset); +} + +static const struct mlxsw_bus mlxsw_pci_bus = { + .kind = "pci", + .init = mlxsw_pci_init, + .fini = mlxsw_pci_fini, + .skb_transmit_busy = mlxsw_pci_skb_transmit_busy, + .skb_transmit = mlxsw_pci_skb_transmit, + .cmd_exec = mlxsw_pci_cmd_exec, + .read_frc_h = mlxsw_pci_read_frc_h, + .read_frc_l = mlxsw_pci_read_frc_l, + .read_utc_sec = mlxsw_pci_read_utc_sec, + .read_utc_nsec = mlxsw_pci_read_utc_nsec, + .features = MLXSW_BUS_F_TXRX | MLXSW_BUS_F_RESET, +}; + +static int mlxsw_pci_cmd_init(struct mlxsw_pci *mlxsw_pci) +{ + int err; + + mutex_init(&mlxsw_pci->cmd.lock); + init_waitqueue_head(&mlxsw_pci->cmd.wait); + + err = mlxsw_pci_mbox_alloc(mlxsw_pci, &mlxsw_pci->cmd.in_mbox); + if (err) + goto err_in_mbox_alloc; + + err = mlxsw_pci_mbox_alloc(mlxsw_pci, &mlxsw_pci->cmd.out_mbox); + if (err) + goto err_out_mbox_alloc; + + return 0; + +err_out_mbox_alloc: + mlxsw_pci_mbox_free(mlxsw_pci, &mlxsw_pci->cmd.in_mbox); +err_in_mbox_alloc: + mutex_destroy(&mlxsw_pci->cmd.lock); + return err; +} + +static void mlxsw_pci_cmd_fini(struct mlxsw_pci *mlxsw_pci) +{ + mlxsw_pci_mbox_free(mlxsw_pci, &mlxsw_pci->cmd.out_mbox); + mlxsw_pci_mbox_free(mlxsw_pci, &mlxsw_pci->cmd.in_mbox); + mutex_destroy(&mlxsw_pci->cmd.lock); +} + +static int mlxsw_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + const char *driver_name = dev_driver_string(&pdev->dev); + struct mlxsw_pci *mlxsw_pci; + int err; + + mlxsw_pci = kzalloc(sizeof(*mlxsw_pci), GFP_KERNEL); + if (!mlxsw_pci) + return -ENOMEM; + + err = pci_enable_device(pdev); + if (err) { + dev_err(&pdev->dev, "pci_enable_device failed\n"); + goto err_pci_enable_device; + } + + err = pci_request_regions(pdev, driver_name); + if (err) { + dev_err(&pdev->dev, "pci_request_regions failed\n"); + goto err_pci_request_regions; + } + + err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64)); + if (err) { + err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32)); + if (err) { + dev_err(&pdev->dev, "dma_set_mask failed\n"); + goto err_pci_set_dma_mask; + } + } + + if (pci_resource_len(pdev, 0) < MLXSW_PCI_BAR0_SIZE) { + dev_err(&pdev->dev, "invalid PCI region size\n"); + err = -EINVAL; + goto err_pci_resource_len_check; + } + + mlxsw_pci->hw_addr = ioremap(pci_resource_start(pdev, 0), + pci_resource_len(pdev, 0)); + if (!mlxsw_pci->hw_addr) { + dev_err(&pdev->dev, "ioremap failed\n"); + err = -EIO; + goto err_ioremap; + } + pci_set_master(pdev); + + mlxsw_pci->pdev = pdev; + pci_set_drvdata(pdev, mlxsw_pci); + + err = mlxsw_pci_cmd_init(mlxsw_pci); + if (err) + goto err_pci_cmd_init; + + mlxsw_pci->bus_info.device_kind = driver_name; + mlxsw_pci->bus_info.device_name = pci_name(mlxsw_pci->pdev); + mlxsw_pci->bus_info.dev = &pdev->dev; + mlxsw_pci->bus_info.read_clock_capable = true; + mlxsw_pci->id = id; + + err = mlxsw_core_bus_device_register(&mlxsw_pci->bus_info, + &mlxsw_pci_bus, mlxsw_pci, false, + NULL, NULL); + if (err) { + dev_err(&pdev->dev, "cannot register bus device\n"); + goto err_bus_device_register; + } + + return 0; + +err_bus_device_register: + mlxsw_pci_cmd_fini(mlxsw_pci); +err_pci_cmd_init: + iounmap(mlxsw_pci->hw_addr); +err_ioremap: +err_pci_resource_len_check: +err_pci_set_dma_mask: + pci_release_regions(pdev); +err_pci_request_regions: + pci_disable_device(pdev); +err_pci_enable_device: + kfree(mlxsw_pci); + return err; +} + +static void mlxsw_pci_remove(struct pci_dev *pdev) +{ + struct mlxsw_pci *mlxsw_pci = pci_get_drvdata(pdev); + + mlxsw_core_bus_device_unregister(mlxsw_pci->core, false); + mlxsw_pci_cmd_fini(mlxsw_pci); + iounmap(mlxsw_pci->hw_addr); + pci_release_regions(mlxsw_pci->pdev); + pci_disable_device(mlxsw_pci->pdev); + kfree(mlxsw_pci); +} + +int mlxsw_pci_driver_register(struct pci_driver *pci_driver) +{ + pci_driver->probe = mlxsw_pci_probe; + pci_driver->remove = mlxsw_pci_remove; + pci_driver->shutdown = mlxsw_pci_remove; + return pci_register_driver(pci_driver); +} +EXPORT_SYMBOL(mlxsw_pci_driver_register); + +void mlxsw_pci_driver_unregister(struct pci_driver *pci_driver) +{ + pci_unregister_driver(pci_driver); +} +EXPORT_SYMBOL(mlxsw_pci_driver_unregister); + +static int __init mlxsw_pci_module_init(void) +{ + return 0; +} + +static void __exit mlxsw_pci_module_exit(void) +{ +} + +module_init(mlxsw_pci_module_init); +module_exit(mlxsw_pci_module_exit); + +MODULE_LICENSE("Dual BSD/GPL"); +MODULE_AUTHOR("Jiri Pirko <jiri@mellanox.com>"); +MODULE_DESCRIPTION("Mellanox switch PCI interface driver"); diff --git a/drivers/net/ethernet/mellanox/mlxsw/pci.h b/drivers/net/ethernet/mellanox/mlxsw/pci.h new file mode 100644 index 000000000..cacc2f9fa --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/pci.h @@ -0,0 +1,34 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2016-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_PCI_H +#define _MLXSW_PCI_H + +#include <linux/pci.h> + +#define PCI_DEVICE_ID_MELLANOX_SPECTRUM 0xcb84 +#define PCI_DEVICE_ID_MELLANOX_SPECTRUM2 0xcf6c +#define PCI_DEVICE_ID_MELLANOX_SPECTRUM3 0xcf70 +#define PCI_DEVICE_ID_MELLANOX_SPECTRUM4 0xcf80 + +#if IS_ENABLED(CONFIG_MLXSW_PCI) + +int mlxsw_pci_driver_register(struct pci_driver *pci_driver); +void mlxsw_pci_driver_unregister(struct pci_driver *pci_driver); + +#else + +static inline int +mlxsw_pci_driver_register(struct pci_driver *pci_driver) +{ + return 0; +} + +static inline void +mlxsw_pci_driver_unregister(struct pci_driver *pci_driver) +{ +} + +#endif + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/pci_hw.h b/drivers/net/ethernet/mellanox/mlxsw/pci_hw.h new file mode 100644 index 000000000..7cdf0ce24 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/pci_hw.h @@ -0,0 +1,416 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_PCI_HW_H +#define _MLXSW_PCI_HW_H + +#include <linux/bitops.h> + +#include "item.h" + +#define MLXSW_PCI_BAR0_SIZE (1024 * 1024) /* 1MB */ +#define MLXSW_PCI_PAGE_SIZE 4096 + +#define MLXSW_PCI_CIR_BASE 0x71000 +#define MLXSW_PCI_CIR_IN_PARAM_HI MLXSW_PCI_CIR_BASE +#define MLXSW_PCI_CIR_IN_PARAM_LO (MLXSW_PCI_CIR_BASE + 0x04) +#define MLXSW_PCI_CIR_IN_MODIFIER (MLXSW_PCI_CIR_BASE + 0x08) +#define MLXSW_PCI_CIR_OUT_PARAM_HI (MLXSW_PCI_CIR_BASE + 0x0C) +#define MLXSW_PCI_CIR_OUT_PARAM_LO (MLXSW_PCI_CIR_BASE + 0x10) +#define MLXSW_PCI_CIR_TOKEN (MLXSW_PCI_CIR_BASE + 0x14) +#define MLXSW_PCI_CIR_CTRL (MLXSW_PCI_CIR_BASE + 0x18) +#define MLXSW_PCI_CIR_CTRL_GO_BIT BIT(23) +#define MLXSW_PCI_CIR_CTRL_EVREQ_BIT BIT(22) +#define MLXSW_PCI_CIR_CTRL_OPCODE_MOD_SHIFT 12 +#define MLXSW_PCI_CIR_CTRL_STATUS_SHIFT 24 +#define MLXSW_PCI_CIR_TIMEOUT_MSECS 1000 + +#define MLXSW_PCI_SW_RESET_TIMEOUT_MSECS 900000 +#define MLXSW_PCI_SW_RESET_WAIT_MSECS 400 +#define MLXSW_PCI_FW_READY 0xA1844 +#define MLXSW_PCI_FW_READY_MASK 0xFFFF +#define MLXSW_PCI_FW_READY_MAGIC 0x5E + +#define MLXSW_PCI_DOORBELL_SDQ_OFFSET 0x000 +#define MLXSW_PCI_DOORBELL_RDQ_OFFSET 0x200 +#define MLXSW_PCI_DOORBELL_CQ_OFFSET 0x400 +#define MLXSW_PCI_DOORBELL_EQ_OFFSET 0x600 +#define MLXSW_PCI_DOORBELL_ARM_CQ_OFFSET 0x800 +#define MLXSW_PCI_DOORBELL_ARM_EQ_OFFSET 0xA00 + +#define MLXSW_PCI_DOORBELL(offset, type_offset, num) \ + ((offset) + (type_offset) + (num) * 4) + +#define MLXSW_PCI_CQS_MAX 96 +#define MLXSW_PCI_EQS_COUNT 2 +#define MLXSW_PCI_EQ_ASYNC_NUM 0 +#define MLXSW_PCI_EQ_COMP_NUM 1 + +#define MLXSW_PCI_SDQS_MIN 2 /* EMAD and control traffic */ +#define MLXSW_PCI_SDQ_EMAD_INDEX 0 +#define MLXSW_PCI_SDQ_EMAD_TC 0 +#define MLXSW_PCI_SDQ_CTL_TC 3 + +#define MLXSW_PCI_AQ_PAGES 8 +#define MLXSW_PCI_AQ_SIZE (MLXSW_PCI_PAGE_SIZE * MLXSW_PCI_AQ_PAGES) +#define MLXSW_PCI_WQE_SIZE 32 /* 32 bytes per element */ +#define MLXSW_PCI_CQE01_SIZE 16 /* 16 bytes per element */ +#define MLXSW_PCI_CQE2_SIZE 32 /* 32 bytes per element */ +#define MLXSW_PCI_CQE_SIZE_MAX MLXSW_PCI_CQE2_SIZE +#define MLXSW_PCI_EQE_SIZE 16 /* 16 bytes per element */ +#define MLXSW_PCI_WQE_COUNT (MLXSW_PCI_AQ_SIZE / MLXSW_PCI_WQE_SIZE) +#define MLXSW_PCI_CQE01_COUNT (MLXSW_PCI_AQ_SIZE / MLXSW_PCI_CQE01_SIZE) +#define MLXSW_PCI_CQE2_COUNT (MLXSW_PCI_AQ_SIZE / MLXSW_PCI_CQE2_SIZE) +#define MLXSW_PCI_EQE_COUNT (MLXSW_PCI_AQ_SIZE / MLXSW_PCI_EQE_SIZE) +#define MLXSW_PCI_EQE_UPDATE_COUNT 0x80 + +#define MLXSW_PCI_WQE_SG_ENTRIES 3 +#define MLXSW_PCI_WQE_TYPE_ETHERNET 0xA + +/* pci_wqe_c + * If set it indicates that a completion should be reported upon + * execution of this descriptor. + */ +MLXSW_ITEM32(pci, wqe, c, 0x00, 31, 1); + +/* pci_wqe_lp + * Local Processing, set if packet should be processed by the local + * switch hardware: + * For Ethernet EMAD (Direct Route and non Direct Route) - + * must be set if packet destination is local device + * For InfiniBand CTL - must be set if packet destination is local device + * Otherwise it must be clear + * Local Process packets must not exceed the size of 2K (including payload + * and headers). + */ +MLXSW_ITEM32(pci, wqe, lp, 0x00, 30, 1); + +/* pci_wqe_type + * Packet type. + */ +MLXSW_ITEM32(pci, wqe, type, 0x00, 23, 4); + +/* pci_wqe_byte_count + * Size of i-th scatter/gather entry, 0 if entry is unused. + */ +MLXSW_ITEM16_INDEXED(pci, wqe, byte_count, 0x02, 0, 14, 0x02, 0x00, false); + +/* pci_wqe_address + * Physical address of i-th scatter/gather entry. + * Gather Entries must be 2Byte aligned. + */ +MLXSW_ITEM64_INDEXED(pci, wqe, address, 0x08, 0, 64, 0x8, 0x0, false); + +enum mlxsw_pci_cqe_v { + MLXSW_PCI_CQE_V0, + MLXSW_PCI_CQE_V1, + MLXSW_PCI_CQE_V2, +}; + +#define mlxsw_pci_cqe_item_helpers(name, v0, v1, v2) \ +static inline u32 mlxsw_pci_cqe_##name##_get(enum mlxsw_pci_cqe_v v, char *cqe) \ +{ \ + switch (v) { \ + default: \ + case MLXSW_PCI_CQE_V0: \ + return mlxsw_pci_cqe##v0##_##name##_get(cqe); \ + case MLXSW_PCI_CQE_V1: \ + return mlxsw_pci_cqe##v1##_##name##_get(cqe); \ + case MLXSW_PCI_CQE_V2: \ + return mlxsw_pci_cqe##v2##_##name##_get(cqe); \ + } \ +} \ +static inline void mlxsw_pci_cqe_##name##_set(enum mlxsw_pci_cqe_v v, \ + char *cqe, u32 val) \ +{ \ + switch (v) { \ + default: \ + case MLXSW_PCI_CQE_V0: \ + mlxsw_pci_cqe##v0##_##name##_set(cqe, val); \ + break; \ + case MLXSW_PCI_CQE_V1: \ + mlxsw_pci_cqe##v1##_##name##_set(cqe, val); \ + break; \ + case MLXSW_PCI_CQE_V2: \ + mlxsw_pci_cqe##v2##_##name##_set(cqe, val); \ + break; \ + } \ +} + +/* pci_cqe_lag + * Packet arrives from a port which is a LAG + */ +MLXSW_ITEM32(pci, cqe0, lag, 0x00, 23, 1); +MLXSW_ITEM32(pci, cqe12, lag, 0x00, 24, 1); +mlxsw_pci_cqe_item_helpers(lag, 0, 12, 12); + +/* pci_cqe_system_port/lag_id + * When lag=0: System port on which the packet was received + * When lag=1: + * bits [15:4] LAG ID on which the packet was received + * bits [3:0] sub_port on which the packet was received + */ +MLXSW_ITEM32(pci, cqe, system_port, 0x00, 0, 16); +MLXSW_ITEM32(pci, cqe0, lag_id, 0x00, 4, 12); +MLXSW_ITEM32(pci, cqe12, lag_id, 0x00, 0, 16); +mlxsw_pci_cqe_item_helpers(lag_id, 0, 12, 12); +MLXSW_ITEM32(pci, cqe0, lag_subport, 0x00, 0, 4); +MLXSW_ITEM32(pci, cqe12, lag_subport, 0x00, 16, 8); +mlxsw_pci_cqe_item_helpers(lag_subport, 0, 12, 12); + +/* pci_cqe_wqe_counter + * WQE count of the WQEs completed on the associated dqn + */ +MLXSW_ITEM32(pci, cqe, wqe_counter, 0x04, 16, 16); + +/* pci_cqe_byte_count + * Byte count of received packets including additional two + * Reserved Bytes that are append to the end of the frame. + * Reserved for Send CQE. + */ +MLXSW_ITEM32(pci, cqe, byte_count, 0x04, 0, 14); + +#define MLXSW_PCI_CQE2_MIRROR_CONG_INVALID 0xFFFF + +/* pci_cqe_mirror_cong_high + * Congestion level in units of 8KB of the egress traffic class of the original + * packet that does mirroring to the CPU. Value of 0xFFFF means that the + * congestion level is invalid. + */ +MLXSW_ITEM32(pci, cqe2, mirror_cong_high, 0x08, 16, 4); + +/* pci_cqe_trap_id + * Trap ID that captured the packet. + */ +MLXSW_ITEM32(pci, cqe, trap_id, 0x08, 0, 10); + +/* pci_cqe_crc + * Length include CRC. Indicates the length field includes + * the packet's CRC. + */ +MLXSW_ITEM32(pci, cqe0, crc, 0x0C, 8, 1); +MLXSW_ITEM32(pci, cqe12, crc, 0x0C, 9, 1); +mlxsw_pci_cqe_item_helpers(crc, 0, 12, 12); + +/* pci_cqe_e + * CQE with Error. + */ +MLXSW_ITEM32(pci, cqe0, e, 0x0C, 7, 1); +MLXSW_ITEM32(pci, cqe12, e, 0x00, 27, 1); +mlxsw_pci_cqe_item_helpers(e, 0, 12, 12); + +/* pci_cqe_sr + * 1 - Send Queue + * 0 - Receive Queue + */ +MLXSW_ITEM32(pci, cqe0, sr, 0x0C, 6, 1); +MLXSW_ITEM32(pci, cqe12, sr, 0x00, 26, 1); +mlxsw_pci_cqe_item_helpers(sr, 0, 12, 12); + +/* pci_cqe_dqn + * Descriptor Queue (DQ) Number. + */ +MLXSW_ITEM32(pci, cqe0, dqn, 0x0C, 1, 5); +MLXSW_ITEM32(pci, cqe12, dqn, 0x0C, 1, 6); +mlxsw_pci_cqe_item_helpers(dqn, 0, 12, 12); + +/* pci_cqe_time_stamp_low + * Time stamp of the CQE + * Format according to time_stamp_type: + * 0: uSec - 1.024uSec (default for devices which do not support + * time_stamp_type). Only bits 15:0 are valid + * 1: FRC - Free Running Clock - units of 1nSec + * 2: UTC - time_stamp[37:30] = Sec + * - time_stamp[29:0] = nSec + * 3: Mirror_UTC. UTC time stamp of the original packet that has + * MIRROR_SESSION traps + * - time_stamp[37:30] = Sec + * - time_stamp[29:0] = nSec + * Formats 0..2 are configured by + * CONFIG_PROFILE.cqe_time_stamp_type for PTP traps + * Format 3 is used for MIRROR_SESSION traps + * Note that Spectrum does not reveal FRC, UTC and Mirror_UTC + */ +MLXSW_ITEM32(pci, cqe2, time_stamp_low, 0x0C, 16, 16); + +#define MLXSW_PCI_CQE2_MIRROR_TCLASS_INVALID 0x1F + +/* pci_cqe_mirror_tclass + * The egress traffic class of the original packet that does mirroring to the + * CPU. Value of 0x1F means that the traffic class is invalid. + */ +MLXSW_ITEM32(pci, cqe2, mirror_tclass, 0x10, 27, 5); + +/* pci_cqe_tx_lag + * The Tx port of a packet that is mirrored / sampled to the CPU is a LAG. + */ +MLXSW_ITEM32(pci, cqe2, tx_lag, 0x10, 24, 1); + +/* pci_cqe_tx_lag_subport + * The port index within the LAG of a packet that is mirrored / sampled to the + * CPU. Reserved when tx_lag is 0. + */ +MLXSW_ITEM32(pci, cqe2, tx_lag_subport, 0x10, 16, 8); + +#define MLXSW_PCI_CQE2_TX_PORT_MULTI_PORT 0xFFFE +#define MLXSW_PCI_CQE2_TX_PORT_INVALID 0xFFFF + +/* pci_cqe_tx_lag_id + * The Tx LAG ID of the original packet that is mirrored / sampled to the CPU. + * Value of 0xFFFE means multi-port. Value fo 0xFFFF means that the Tx LAG ID + * is invalid. Reserved when tx_lag is 0. + */ +MLXSW_ITEM32(pci, cqe2, tx_lag_id, 0x10, 0, 16); + +/* pci_cqe_tx_system_port + * The Tx port of the original packet that is mirrored / sampled to the CPU. + * Value of 0xFFFE means multi-port. Value fo 0xFFFF means that the Tx port is + * invalid. Reserved when tx_lag is 1. + */ +MLXSW_ITEM32(pci, cqe2, tx_system_port, 0x10, 0, 16); + +/* pci_cqe_mirror_cong_low + * Congestion level in units of 8KB of the egress traffic class of the original + * packet that does mirroring to the CPU. Value of 0xFFFF means that the + * congestion level is invalid. + */ +MLXSW_ITEM32(pci, cqe2, mirror_cong_low, 0x14, 20, 12); + +#define MLXSW_PCI_CQE2_MIRROR_CONG_SHIFT 13 /* Units of 8KB. */ + +static inline u16 mlxsw_pci_cqe2_mirror_cong_get(const char *cqe) +{ + u16 cong_high = mlxsw_pci_cqe2_mirror_cong_high_get(cqe); + u16 cong_low = mlxsw_pci_cqe2_mirror_cong_low_get(cqe); + + return cong_high << 12 | cong_low; +} + +/* pci_cqe_user_def_val_orig_pkt_len + * When trap_id is an ACL: User defined value from policy engine action. + */ +MLXSW_ITEM32(pci, cqe2, user_def_val_orig_pkt_len, 0x14, 0, 20); + +/* pci_cqe_mirror_reason + * Mirror reason. + */ +MLXSW_ITEM32(pci, cqe2, mirror_reason, 0x18, 24, 8); + +enum mlxsw_pci_cqe_time_stamp_type { + MLXSW_PCI_CQE_TIME_STAMP_TYPE_USEC, + MLXSW_PCI_CQE_TIME_STAMP_TYPE_FRC, + MLXSW_PCI_CQE_TIME_STAMP_TYPE_UTC, + MLXSW_PCI_CQE_TIME_STAMP_TYPE_MIRROR_UTC, +}; + +/* pci_cqe_time_stamp_type + * Time stamp type: + * 0: uSec - 1.024uSec (default for devices which do not support + * time_stamp_type) + * 1: FRC - Free Running Clock - units of 1nSec + * 2: UTC + * 3: Mirror_UTC. UTC time stamp of the original packet that has + * MIRROR_SESSION traps + */ +MLXSW_ITEM32(pci, cqe2, time_stamp_type, 0x18, 22, 2); + +#define MLXSW_PCI_CQE2_MIRROR_LATENCY_INVALID 0xFFFFFF + +/* pci_cqe_time_stamp_high + * Time stamp of the CQE + * Format according to time_stamp_type: + * 0: uSec - 1.024uSec (default for devices which do not support + * time_stamp_type). Only bits 15:0 are valid + * 1: FRC - Free Running Clock - units of 1nSec + * 2: UTC - time_stamp[37:30] = Sec + * - time_stamp[29:0] = nSec + * 3: Mirror_UTC. UTC time stamp of the original packet that has + * MIRROR_SESSION traps + * - time_stamp[37:30] = Sec + * - time_stamp[29:0] = nSec + * Formats 0..2 are configured by + * CONFIG_PROFILE.cqe_time_stamp_type for PTP traps + * Format 3 is used for MIRROR_SESSION traps + * Note that Spectrum does not reveal FRC, UTC and Mirror_UTC + */ +MLXSW_ITEM32(pci, cqe2, time_stamp_high, 0x18, 0, 22); + +static inline u64 mlxsw_pci_cqe2_time_stamp_get(const char *cqe) +{ + u64 ts_high = mlxsw_pci_cqe2_time_stamp_high_get(cqe); + u64 ts_low = mlxsw_pci_cqe2_time_stamp_low_get(cqe); + + return ts_high << 16 | ts_low; +} + +static inline u8 mlxsw_pci_cqe2_time_stamp_sec_get(const char *cqe) +{ + u64 full_ts = mlxsw_pci_cqe2_time_stamp_get(cqe); + + return full_ts >> 30 & 0xFF; +} + +static inline u32 mlxsw_pci_cqe2_time_stamp_nsec_get(const char *cqe) +{ + u64 full_ts = mlxsw_pci_cqe2_time_stamp_get(cqe); + + return full_ts & 0x3FFFFFFF; +} + +/* pci_cqe_mirror_latency + * End-to-end latency of the original packet that does mirroring to the CPU. + * Value of 0xFFFFFF means that the latency is invalid. Units are according to + * MOGCR.mirror_latency_units. + */ +MLXSW_ITEM32(pci, cqe2, mirror_latency, 0x1C, 8, 24); + +/* pci_cqe_owner + * Ownership bit. + */ +MLXSW_ITEM32(pci, cqe01, owner, 0x0C, 0, 1); +MLXSW_ITEM32(pci, cqe2, owner, 0x1C, 0, 1); +mlxsw_pci_cqe_item_helpers(owner, 01, 01, 2); + +/* pci_eqe_event_type + * Event type. + */ +MLXSW_ITEM32(pci, eqe, event_type, 0x0C, 24, 8); +#define MLXSW_PCI_EQE_EVENT_TYPE_COMP 0x00 +#define MLXSW_PCI_EQE_EVENT_TYPE_CMD 0x0A + +/* pci_eqe_event_sub_type + * Event type. + */ +MLXSW_ITEM32(pci, eqe, event_sub_type, 0x0C, 16, 8); + +/* pci_eqe_cqn + * Completion Queue that triggered this EQE. + */ +MLXSW_ITEM32(pci, eqe, cqn, 0x0C, 8, 7); + +/* pci_eqe_owner + * Ownership bit. + */ +MLXSW_ITEM32(pci, eqe, owner, 0x0C, 0, 1); + +/* pci_eqe_cmd_token + * Command completion event - token + */ +MLXSW_ITEM32(pci, eqe, cmd_token, 0x00, 16, 16); + +/* pci_eqe_cmd_status + * Command completion event - status + */ +MLXSW_ITEM32(pci, eqe, cmd_status, 0x00, 0, 8); + +/* pci_eqe_cmd_out_param_h + * Command completion event - output parameter - higher part + */ +MLXSW_ITEM32(pci, eqe, cmd_out_param_h, 0x04, 0, 32); + +/* pci_eqe_cmd_out_param_l + * Command completion event - output parameter - lower part + */ +MLXSW_ITEM32(pci, eqe, cmd_out_param_l, 0x08, 0, 32); + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/port.h b/drivers/net/ethernet/mellanox/mlxsw/port.h new file mode 100644 index 000000000..ac4d4ea51 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/port.h @@ -0,0 +1,38 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_PORT_H +#define _MLXSW_PORT_H + +#include <linux/types.h> + +#define MLXSW_PORT_MAX_MTU 10000 + +#define MLXSW_PORT_DEFAULT_VID 1 + +#define MLXSW_PORT_SWID_DISABLED_PORT 255 +#define MLXSW_PORT_SWID_ALL_SWIDS 254 +#define MLXSW_PORT_SWID_TYPE_IB 1 +#define MLXSW_PORT_SWID_TYPE_ETH 2 + +#define MLXSW_PORT_MAX_IB_PHY_PORTS 36 +#define MLXSW_PORT_MAX_IB_PORTS (MLXSW_PORT_MAX_IB_PHY_PORTS + 1) + +#define MLXSW_PORT_CPU_PORT 0x0 + +#define MLXSW_PORT_DONT_CARE 0xFF + +enum mlxsw_port_admin_status { + MLXSW_PORT_ADMIN_STATUS_UP = 1, + MLXSW_PORT_ADMIN_STATUS_DOWN = 2, + MLXSW_PORT_ADMIN_STATUS_UP_ONCE = 3, + MLXSW_PORT_ADMIN_STATUS_DISABLED = 4, +}; + +enum mlxsw_reg_pude_oper_status { + MLXSW_PORT_OPER_STATUS_UP = 1, + MLXSW_PORT_OPER_STATUS_DOWN = 2, + MLXSW_PORT_OPER_STATUS_FAILURE = 4, /* Can be set to up again. */ +}; + +#endif /* _MLXSW_PORT_H */ diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h b/drivers/net/ethernet/mellanox/mlxsw/reg.h new file mode 100644 index 000000000..a34ff19c5 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/reg.h @@ -0,0 +1,12937 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_REG_H +#define _MLXSW_REG_H + +#include <linux/kernel.h> +#include <linux/string.h> +#include <linux/bitops.h> +#include <linux/if_vlan.h> + +#include "item.h" +#include "port.h" + +struct mlxsw_reg_info { + u16 id; + u16 len; /* In u8 */ + const char *name; +}; + +#define MLXSW_REG_DEFINE(_name, _id, _len) \ +static const struct mlxsw_reg_info mlxsw_reg_##_name = { \ + .id = _id, \ + .len = _len, \ + .name = #_name, \ +} + +#define MLXSW_REG(type) (&mlxsw_reg_##type) +#define MLXSW_REG_LEN(type) MLXSW_REG(type)->len +#define MLXSW_REG_ZERO(type, payload) memset(payload, 0, MLXSW_REG(type)->len) + +/* SGCR - Switch General Configuration Register + * -------------------------------------------- + * This register is used for configuration of the switch capabilities. + */ +#define MLXSW_REG_SGCR_ID 0x2000 +#define MLXSW_REG_SGCR_LEN 0x10 + +MLXSW_REG_DEFINE(sgcr, MLXSW_REG_SGCR_ID, MLXSW_REG_SGCR_LEN); + +/* reg_sgcr_llb + * Link Local Broadcast (Default=0) + * When set, all Link Local packets (224.0.0.X) will be treated as broadcast + * packets and ignore the IGMP snooping entries. + * Access: RW + */ +MLXSW_ITEM32(reg, sgcr, llb, 0x04, 0, 1); + +static inline void mlxsw_reg_sgcr_pack(char *payload, bool llb) +{ + MLXSW_REG_ZERO(sgcr, payload); + mlxsw_reg_sgcr_llb_set(payload, !!llb); +} + +/* SPAD - Switch Physical Address Register + * --------------------------------------- + * The SPAD register configures the switch physical MAC address. + */ +#define MLXSW_REG_SPAD_ID 0x2002 +#define MLXSW_REG_SPAD_LEN 0x10 + +MLXSW_REG_DEFINE(spad, MLXSW_REG_SPAD_ID, MLXSW_REG_SPAD_LEN); + +/* reg_spad_base_mac + * Base MAC address for the switch partitions. + * Per switch partition MAC address is equal to: + * base_mac + swid + * Access: RW + */ +MLXSW_ITEM_BUF(reg, spad, base_mac, 0x02, 6); + +/* SSPR - Switch System Port Record Register + * ----------------------------------------- + * Configures the system port to local port mapping. + */ +#define MLXSW_REG_SSPR_ID 0x2008 +#define MLXSW_REG_SSPR_LEN 0x8 + +MLXSW_REG_DEFINE(sspr, MLXSW_REG_SSPR_ID, MLXSW_REG_SSPR_LEN); + +/* reg_sspr_m + * Master - if set, then the record describes the master system port. + * This is needed in case a local port is mapped into several system ports + * (for multipathing). That number will be reported as the source system + * port when packets are forwarded to the CPU. Only one master port is allowed + * per local port. + * + * Note: Must be set for Spectrum. + * Access: RW + */ +MLXSW_ITEM32(reg, sspr, m, 0x00, 31, 1); + +/* reg_sspr_local_port + * Local port number. + * + * Access: RW + */ +MLXSW_ITEM32_LP(reg, sspr, 0x00, 16, 0x00, 12); + +/* reg_sspr_system_port + * Unique identifier within the stacking domain that represents all the ports + * that are available in the system (external ports). + * + * Currently, only single-ASIC configurations are supported, so we default to + * 1:1 mapping between system ports and local ports. + * Access: Index + */ +MLXSW_ITEM32(reg, sspr, system_port, 0x04, 0, 16); + +static inline void mlxsw_reg_sspr_pack(char *payload, u16 local_port) +{ + MLXSW_REG_ZERO(sspr, payload); + mlxsw_reg_sspr_m_set(payload, 1); + mlxsw_reg_sspr_local_port_set(payload, local_port); + mlxsw_reg_sspr_system_port_set(payload, local_port); +} + +/* SFDAT - Switch Filtering Database Aging Time + * -------------------------------------------- + * Controls the Switch aging time. Aging time is able to be set per Switch + * Partition. + */ +#define MLXSW_REG_SFDAT_ID 0x2009 +#define MLXSW_REG_SFDAT_LEN 0x8 + +MLXSW_REG_DEFINE(sfdat, MLXSW_REG_SFDAT_ID, MLXSW_REG_SFDAT_LEN); + +/* reg_sfdat_swid + * Switch partition ID. + * Access: Index + */ +MLXSW_ITEM32(reg, sfdat, swid, 0x00, 24, 8); + +/* reg_sfdat_age_time + * Aging time in seconds + * Min - 10 seconds + * Max - 1,000,000 seconds + * Default is 300 seconds. + * Access: RW + */ +MLXSW_ITEM32(reg, sfdat, age_time, 0x04, 0, 20); + +static inline void mlxsw_reg_sfdat_pack(char *payload, u32 age_time) +{ + MLXSW_REG_ZERO(sfdat, payload); + mlxsw_reg_sfdat_swid_set(payload, 0); + mlxsw_reg_sfdat_age_time_set(payload, age_time); +} + +/* SFD - Switch Filtering Database + * ------------------------------- + * The following register defines the access to the filtering database. + * The register supports querying, adding, removing and modifying the database. + * The access is optimized for bulk updates in which case more than one + * FDB record is present in the same command. + */ +#define MLXSW_REG_SFD_ID 0x200A +#define MLXSW_REG_SFD_BASE_LEN 0x10 /* base length, without records */ +#define MLXSW_REG_SFD_REC_LEN 0x10 /* record length */ +#define MLXSW_REG_SFD_REC_MAX_COUNT 64 +#define MLXSW_REG_SFD_LEN (MLXSW_REG_SFD_BASE_LEN + \ + MLXSW_REG_SFD_REC_LEN * MLXSW_REG_SFD_REC_MAX_COUNT) + +MLXSW_REG_DEFINE(sfd, MLXSW_REG_SFD_ID, MLXSW_REG_SFD_LEN); + +/* reg_sfd_swid + * Switch partition ID for queries. Reserved on Write. + * Access: Index + */ +MLXSW_ITEM32(reg, sfd, swid, 0x00, 24, 8); + +enum mlxsw_reg_sfd_op { + /* Dump entire FDB a (process according to record_locator) */ + MLXSW_REG_SFD_OP_QUERY_DUMP = 0, + /* Query records by {MAC, VID/FID} value */ + MLXSW_REG_SFD_OP_QUERY_QUERY = 1, + /* Query and clear activity. Query records by {MAC, VID/FID} value */ + MLXSW_REG_SFD_OP_QUERY_QUERY_AND_CLEAR_ACTIVITY = 2, + /* Test. Response indicates if each of the records could be + * added to the FDB. + */ + MLXSW_REG_SFD_OP_WRITE_TEST = 0, + /* Add/modify. Aged-out records cannot be added. This command removes + * the learning notification of the {MAC, VID/FID}. Response includes + * the entries that were added to the FDB. + */ + MLXSW_REG_SFD_OP_WRITE_EDIT = 1, + /* Remove record by {MAC, VID/FID}. This command also removes + * the learning notification and aged-out notifications + * of the {MAC, VID/FID}. The response provides current (pre-removal) + * entries as non-aged-out. + */ + MLXSW_REG_SFD_OP_WRITE_REMOVE = 2, + /* Remove learned notification by {MAC, VID/FID}. The response provides + * the removed learning notification. + */ + MLXSW_REG_SFD_OP_WRITE_REMOVE_NOTIFICATION = 2, +}; + +/* reg_sfd_op + * Operation. + * Access: OP + */ +MLXSW_ITEM32(reg, sfd, op, 0x04, 30, 2); + +/* reg_sfd_record_locator + * Used for querying the FDB. Use record_locator=0 to initiate the + * query. When a record is returned, a new record_locator is + * returned to be used in the subsequent query. + * Reserved for database update. + * Access: Index + */ +MLXSW_ITEM32(reg, sfd, record_locator, 0x04, 0, 30); + +/* reg_sfd_num_rec + * Request: Number of records to read/add/modify/remove + * Response: Number of records read/added/replaced/removed + * See above description for more details. + * Ranges 0..64 + * Access: RW + */ +MLXSW_ITEM32(reg, sfd, num_rec, 0x08, 0, 8); + +static inline void mlxsw_reg_sfd_pack(char *payload, enum mlxsw_reg_sfd_op op, + u32 record_locator) +{ + MLXSW_REG_ZERO(sfd, payload); + mlxsw_reg_sfd_op_set(payload, op); + mlxsw_reg_sfd_record_locator_set(payload, record_locator); +} + +/* reg_sfd_rec_swid + * Switch partition ID. + * Access: Index + */ +MLXSW_ITEM32_INDEXED(reg, sfd, rec_swid, MLXSW_REG_SFD_BASE_LEN, 24, 8, + MLXSW_REG_SFD_REC_LEN, 0x00, false); + +enum mlxsw_reg_sfd_rec_type { + MLXSW_REG_SFD_REC_TYPE_UNICAST = 0x0, + MLXSW_REG_SFD_REC_TYPE_UNICAST_LAG = 0x1, + MLXSW_REG_SFD_REC_TYPE_MULTICAST = 0x2, + MLXSW_REG_SFD_REC_TYPE_UNICAST_TUNNEL = 0xC, +}; + +/* reg_sfd_rec_type + * FDB record type. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sfd, rec_type, MLXSW_REG_SFD_BASE_LEN, 20, 4, + MLXSW_REG_SFD_REC_LEN, 0x00, false); + +enum mlxsw_reg_sfd_rec_policy { + /* Replacement disabled, aging disabled. */ + MLXSW_REG_SFD_REC_POLICY_STATIC_ENTRY = 0, + /* (mlag remote): Replacement enabled, aging disabled, + * learning notification enabled on this port. + */ + MLXSW_REG_SFD_REC_POLICY_DYNAMIC_ENTRY_MLAG = 1, + /* (ingress device): Replacement enabled, aging enabled. */ + MLXSW_REG_SFD_REC_POLICY_DYNAMIC_ENTRY_INGRESS = 3, +}; + +/* reg_sfd_rec_policy + * Policy. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sfd, rec_policy, MLXSW_REG_SFD_BASE_LEN, 18, 2, + MLXSW_REG_SFD_REC_LEN, 0x00, false); + +/* reg_sfd_rec_a + * Activity. Set for new static entries. Set for static entries if a frame SMAC + * lookup hits on the entry. + * To clear the a bit, use "query and clear activity" op. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sfd, rec_a, MLXSW_REG_SFD_BASE_LEN, 16, 1, + MLXSW_REG_SFD_REC_LEN, 0x00, false); + +/* reg_sfd_rec_mac + * MAC address. + * Access: Index + */ +MLXSW_ITEM_BUF_INDEXED(reg, sfd, rec_mac, MLXSW_REG_SFD_BASE_LEN, 6, + MLXSW_REG_SFD_REC_LEN, 0x02); + +enum mlxsw_reg_sfd_rec_action { + /* forward */ + MLXSW_REG_SFD_REC_ACTION_NOP = 0, + /* forward and trap, trap_id is FDB_TRAP */ + MLXSW_REG_SFD_REC_ACTION_MIRROR_TO_CPU = 1, + /* trap and do not forward, trap_id is FDB_TRAP */ + MLXSW_REG_SFD_REC_ACTION_TRAP = 2, + /* forward to IP router */ + MLXSW_REG_SFD_REC_ACTION_FORWARD_IP_ROUTER = 3, + MLXSW_REG_SFD_REC_ACTION_DISCARD_ERROR = 15, +}; + +/* reg_sfd_rec_action + * Action to apply on the packet. + * Note: Dynamic entries can only be configured with NOP action. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sfd, rec_action, MLXSW_REG_SFD_BASE_LEN, 28, 4, + MLXSW_REG_SFD_REC_LEN, 0x0C, false); + +/* reg_sfd_uc_sub_port + * VEPA channel on local port. + * Valid only if local port is a non-stacking port. Must be 0 if multichannel + * VEPA is not enabled. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_sub_port, MLXSW_REG_SFD_BASE_LEN, 16, 8, + MLXSW_REG_SFD_REC_LEN, 0x08, false); + +/* reg_sfd_uc_set_vid + * Set VID. + * 0 - Do not update VID. + * 1 - Set VID. + * For Spectrum-2 when set_vid=0 and smpe_valid=1, the smpe will modify the vid. + * Access: RW + * + * Note: Reserved when legacy bridge model is used. + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_set_vid, MLXSW_REG_SFD_BASE_LEN, 31, 1, + MLXSW_REG_SFD_REC_LEN, 0x08, false); + +/* reg_sfd_uc_fid_vid + * Filtering ID or VLAN ID + * For SwitchX and SwitchX-2: + * - Dynamic entries (policy 2,3) use FID + * - Static entries (policy 0) use VID + * - When independent learning is configured, VID=FID + * For Spectrum: use FID for both Dynamic and Static entries. + * VID should not be used. + * Access: Index + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_fid_vid, MLXSW_REG_SFD_BASE_LEN, 0, 16, + MLXSW_REG_SFD_REC_LEN, 0x08, false); + +/* reg_sfd_uc_vid + * New VID when set_vid=1. + * Access: RW + * + * Note: Reserved when legacy bridge model is used and when set_vid=0. + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_vid, MLXSW_REG_SFD_BASE_LEN, 16, 12, + MLXSW_REG_SFD_REC_LEN, 0x0C, false); + +/* reg_sfd_uc_system_port + * Unique port identifier for the final destination of the packet. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_system_port, MLXSW_REG_SFD_BASE_LEN, 0, 16, + MLXSW_REG_SFD_REC_LEN, 0x0C, false); + +static inline void mlxsw_reg_sfd_rec_pack(char *payload, int rec_index, + enum mlxsw_reg_sfd_rec_type rec_type, + const char *mac, + enum mlxsw_reg_sfd_rec_action action) +{ + u8 num_rec = mlxsw_reg_sfd_num_rec_get(payload); + + if (rec_index >= num_rec) + mlxsw_reg_sfd_num_rec_set(payload, rec_index + 1); + mlxsw_reg_sfd_rec_swid_set(payload, rec_index, 0); + mlxsw_reg_sfd_rec_type_set(payload, rec_index, rec_type); + mlxsw_reg_sfd_rec_mac_memcpy_to(payload, rec_index, mac); + mlxsw_reg_sfd_rec_action_set(payload, rec_index, action); +} + +static inline void mlxsw_reg_sfd_uc_pack(char *payload, int rec_index, + enum mlxsw_reg_sfd_rec_policy policy, + const char *mac, u16 fid_vid, u16 vid, + enum mlxsw_reg_sfd_rec_action action, + u16 local_port) +{ + mlxsw_reg_sfd_rec_pack(payload, rec_index, + MLXSW_REG_SFD_REC_TYPE_UNICAST, mac, action); + mlxsw_reg_sfd_rec_policy_set(payload, rec_index, policy); + mlxsw_reg_sfd_uc_sub_port_set(payload, rec_index, 0); + mlxsw_reg_sfd_uc_fid_vid_set(payload, rec_index, fid_vid); + mlxsw_reg_sfd_uc_set_vid_set(payload, rec_index, vid ? true : false); + mlxsw_reg_sfd_uc_vid_set(payload, rec_index, vid); + mlxsw_reg_sfd_uc_system_port_set(payload, rec_index, local_port); +} + +/* reg_sfd_uc_lag_sub_port + * LAG sub port. + * Must be 0 if multichannel VEPA is not enabled. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_lag_sub_port, MLXSW_REG_SFD_BASE_LEN, 16, 8, + MLXSW_REG_SFD_REC_LEN, 0x08, false); + +/* reg_sfd_uc_lag_set_vid + * Set VID. + * 0 - Do not update VID. + * 1 - Set VID. + * For Spectrum-2 when set_vid=0 and smpe_valid=1, the smpe will modify the vid. + * Access: RW + * + * Note: Reserved when legacy bridge model is used. + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_lag_set_vid, MLXSW_REG_SFD_BASE_LEN, 31, 1, + MLXSW_REG_SFD_REC_LEN, 0x08, false); + +/* reg_sfd_uc_lag_fid_vid + * Filtering ID or VLAN ID + * For SwitchX and SwitchX-2: + * - Dynamic entries (policy 2,3) use FID + * - Static entries (policy 0) use VID + * - When independent learning is configured, VID=FID + * For Spectrum: use FID for both Dynamic and Static entries. + * VID should not be used. + * Access: Index + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_lag_fid_vid, MLXSW_REG_SFD_BASE_LEN, 0, 16, + MLXSW_REG_SFD_REC_LEN, 0x08, false); + +/* reg_sfd_uc_lag_lag_vid + * New vlan ID. + * Access: RW + * + * Note: Reserved when legacy bridge model is used and set_vid=0. + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_lag_lag_vid, MLXSW_REG_SFD_BASE_LEN, 16, 12, + MLXSW_REG_SFD_REC_LEN, 0x0C, false); + +/* reg_sfd_uc_lag_lag_id + * LAG Identifier - pointer into the LAG descriptor table. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_lag_lag_id, MLXSW_REG_SFD_BASE_LEN, 0, 10, + MLXSW_REG_SFD_REC_LEN, 0x0C, false); + +static inline void +mlxsw_reg_sfd_uc_lag_pack(char *payload, int rec_index, + enum mlxsw_reg_sfd_rec_policy policy, + const char *mac, u16 fid_vid, + enum mlxsw_reg_sfd_rec_action action, u16 lag_vid, + u16 lag_id) +{ + mlxsw_reg_sfd_rec_pack(payload, rec_index, + MLXSW_REG_SFD_REC_TYPE_UNICAST_LAG, + mac, action); + mlxsw_reg_sfd_rec_policy_set(payload, rec_index, policy); + mlxsw_reg_sfd_uc_lag_sub_port_set(payload, rec_index, 0); + mlxsw_reg_sfd_uc_lag_fid_vid_set(payload, rec_index, fid_vid); + mlxsw_reg_sfd_uc_lag_set_vid_set(payload, rec_index, true); + mlxsw_reg_sfd_uc_lag_lag_vid_set(payload, rec_index, lag_vid); + mlxsw_reg_sfd_uc_lag_lag_id_set(payload, rec_index, lag_id); +} + +/* reg_sfd_mc_pgi + * + * Multicast port group index - index into the port group table. + * Value 0x1FFF indicates the pgi should point to the MID entry. + * For Spectrum this value must be set to 0x1FFF + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sfd, mc_pgi, MLXSW_REG_SFD_BASE_LEN, 16, 13, + MLXSW_REG_SFD_REC_LEN, 0x08, false); + +/* reg_sfd_mc_fid_vid + * + * Filtering ID or VLAN ID + * Access: Index + */ +MLXSW_ITEM32_INDEXED(reg, sfd, mc_fid_vid, MLXSW_REG_SFD_BASE_LEN, 0, 16, + MLXSW_REG_SFD_REC_LEN, 0x08, false); + +/* reg_sfd_mc_mid + * + * Multicast identifier - global identifier that represents the multicast + * group across all devices. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sfd, mc_mid, MLXSW_REG_SFD_BASE_LEN, 0, 16, + MLXSW_REG_SFD_REC_LEN, 0x0C, false); + +static inline void +mlxsw_reg_sfd_mc_pack(char *payload, int rec_index, + const char *mac, u16 fid_vid, + enum mlxsw_reg_sfd_rec_action action, u16 mid) +{ + mlxsw_reg_sfd_rec_pack(payload, rec_index, + MLXSW_REG_SFD_REC_TYPE_MULTICAST, mac, action); + mlxsw_reg_sfd_mc_pgi_set(payload, rec_index, 0x1FFF); + mlxsw_reg_sfd_mc_fid_vid_set(payload, rec_index, fid_vid); + mlxsw_reg_sfd_mc_mid_set(payload, rec_index, mid); +} + +/* reg_sfd_uc_tunnel_uip_msb + * When protocol is IPv4, the most significant byte of the underlay IPv4 + * destination IP. + * When protocol is IPv6, reserved. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_tunnel_uip_msb, MLXSW_REG_SFD_BASE_LEN, 24, + 8, MLXSW_REG_SFD_REC_LEN, 0x08, false); + +/* reg_sfd_uc_tunnel_fid + * Filtering ID. + * Access: Index + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_tunnel_fid, MLXSW_REG_SFD_BASE_LEN, 0, 16, + MLXSW_REG_SFD_REC_LEN, 0x08, false); + +enum mlxsw_reg_sfd_uc_tunnel_protocol { + MLXSW_REG_SFD_UC_TUNNEL_PROTOCOL_IPV4, + MLXSW_REG_SFD_UC_TUNNEL_PROTOCOL_IPV6, +}; + +/* reg_sfd_uc_tunnel_protocol + * IP protocol. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_tunnel_protocol, MLXSW_REG_SFD_BASE_LEN, 27, + 1, MLXSW_REG_SFD_REC_LEN, 0x0C, false); + +/* reg_sfd_uc_tunnel_uip_lsb + * When protocol is IPv4, the least significant bytes of the underlay + * IPv4 destination IP. + * When protocol is IPv6, pointer to the underlay IPv6 destination IP + * which is configured by RIPS. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sfd, uc_tunnel_uip_lsb, MLXSW_REG_SFD_BASE_LEN, 0, + 24, MLXSW_REG_SFD_REC_LEN, 0x0C, false); + +static inline void +mlxsw_reg_sfd_uc_tunnel_pack(char *payload, int rec_index, + enum mlxsw_reg_sfd_rec_policy policy, + const char *mac, u16 fid, + enum mlxsw_reg_sfd_rec_action action, + enum mlxsw_reg_sfd_uc_tunnel_protocol proto) +{ + mlxsw_reg_sfd_rec_pack(payload, rec_index, + MLXSW_REG_SFD_REC_TYPE_UNICAST_TUNNEL, mac, + action); + mlxsw_reg_sfd_rec_policy_set(payload, rec_index, policy); + mlxsw_reg_sfd_uc_tunnel_fid_set(payload, rec_index, fid); + mlxsw_reg_sfd_uc_tunnel_protocol_set(payload, rec_index, proto); +} + +static inline void +mlxsw_reg_sfd_uc_tunnel_pack4(char *payload, int rec_index, + enum mlxsw_reg_sfd_rec_policy policy, + const char *mac, u16 fid, + enum mlxsw_reg_sfd_rec_action action, u32 uip) +{ + mlxsw_reg_sfd_uc_tunnel_uip_msb_set(payload, rec_index, uip >> 24); + mlxsw_reg_sfd_uc_tunnel_uip_lsb_set(payload, rec_index, uip); + mlxsw_reg_sfd_uc_tunnel_pack(payload, rec_index, policy, mac, fid, + action, + MLXSW_REG_SFD_UC_TUNNEL_PROTOCOL_IPV4); +} + +static inline void +mlxsw_reg_sfd_uc_tunnel_pack6(char *payload, int rec_index, const char *mac, + u16 fid, enum mlxsw_reg_sfd_rec_action action, + u32 uip_ptr) +{ + mlxsw_reg_sfd_uc_tunnel_uip_lsb_set(payload, rec_index, uip_ptr); + /* Only static policy is supported for IPv6 unicast tunnel entry. */ + mlxsw_reg_sfd_uc_tunnel_pack(payload, rec_index, + MLXSW_REG_SFD_REC_POLICY_STATIC_ENTRY, + mac, fid, action, + MLXSW_REG_SFD_UC_TUNNEL_PROTOCOL_IPV6); +} + +enum mlxsw_reg_tunnel_port { + MLXSW_REG_TUNNEL_PORT_NVE, + MLXSW_REG_TUNNEL_PORT_VPLS, + MLXSW_REG_TUNNEL_PORT_FLEX_TUNNEL0, + MLXSW_REG_TUNNEL_PORT_FLEX_TUNNEL1, +}; + +/* SFN - Switch FDB Notification Register + * ------------------------------------------- + * The switch provides notifications on newly learned FDB entries and + * aged out entries. The notifications can be polled by software. + */ +#define MLXSW_REG_SFN_ID 0x200B +#define MLXSW_REG_SFN_BASE_LEN 0x10 /* base length, without records */ +#define MLXSW_REG_SFN_REC_LEN 0x10 /* record length */ +#define MLXSW_REG_SFN_REC_MAX_COUNT 64 +#define MLXSW_REG_SFN_LEN (MLXSW_REG_SFN_BASE_LEN + \ + MLXSW_REG_SFN_REC_LEN * MLXSW_REG_SFN_REC_MAX_COUNT) + +MLXSW_REG_DEFINE(sfn, MLXSW_REG_SFN_ID, MLXSW_REG_SFN_LEN); + +/* reg_sfn_swid + * Switch partition ID. + * Access: Index + */ +MLXSW_ITEM32(reg, sfn, swid, 0x00, 24, 8); + +/* reg_sfn_end + * Forces the current session to end. + * Access: OP + */ +MLXSW_ITEM32(reg, sfn, end, 0x04, 20, 1); + +/* reg_sfn_num_rec + * Request: Number of learned notifications and aged-out notification + * records requested. + * Response: Number of notification records returned (must be smaller + * than or equal to the value requested) + * Ranges 0..64 + * Access: OP + */ +MLXSW_ITEM32(reg, sfn, num_rec, 0x04, 0, 8); + +static inline void mlxsw_reg_sfn_pack(char *payload) +{ + MLXSW_REG_ZERO(sfn, payload); + mlxsw_reg_sfn_swid_set(payload, 0); + mlxsw_reg_sfn_end_set(payload, 0); + mlxsw_reg_sfn_num_rec_set(payload, MLXSW_REG_SFN_REC_MAX_COUNT); +} + +/* reg_sfn_rec_swid + * Switch partition ID. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sfn, rec_swid, MLXSW_REG_SFN_BASE_LEN, 24, 8, + MLXSW_REG_SFN_REC_LEN, 0x00, false); + +enum mlxsw_reg_sfn_rec_type { + /* MAC addresses learned on a regular port. */ + MLXSW_REG_SFN_REC_TYPE_LEARNED_MAC = 0x5, + /* MAC addresses learned on a LAG port. */ + MLXSW_REG_SFN_REC_TYPE_LEARNED_MAC_LAG = 0x6, + /* Aged-out MAC address on a regular port. */ + MLXSW_REG_SFN_REC_TYPE_AGED_OUT_MAC = 0x7, + /* Aged-out MAC address on a LAG port. */ + MLXSW_REG_SFN_REC_TYPE_AGED_OUT_MAC_LAG = 0x8, + /* Learned unicast tunnel record. */ + MLXSW_REG_SFN_REC_TYPE_LEARNED_UNICAST_TUNNEL = 0xD, + /* Aged-out unicast tunnel record. */ + MLXSW_REG_SFN_REC_TYPE_AGED_OUT_UNICAST_TUNNEL = 0xE, +}; + +/* reg_sfn_rec_type + * Notification record type. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sfn, rec_type, MLXSW_REG_SFN_BASE_LEN, 20, 4, + MLXSW_REG_SFN_REC_LEN, 0x00, false); + +/* reg_sfn_rec_mac + * MAC address. + * Access: RO + */ +MLXSW_ITEM_BUF_INDEXED(reg, sfn, rec_mac, MLXSW_REG_SFN_BASE_LEN, 6, + MLXSW_REG_SFN_REC_LEN, 0x02); + +/* reg_sfn_mac_sub_port + * VEPA channel on the local port. + * 0 if multichannel VEPA is not enabled. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sfn, mac_sub_port, MLXSW_REG_SFN_BASE_LEN, 16, 8, + MLXSW_REG_SFN_REC_LEN, 0x08, false); + +/* reg_sfn_mac_fid + * Filtering identifier. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sfn, mac_fid, MLXSW_REG_SFN_BASE_LEN, 0, 16, + MLXSW_REG_SFN_REC_LEN, 0x08, false); + +/* reg_sfn_mac_system_port + * Unique port identifier for the final destination of the packet. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sfn, mac_system_port, MLXSW_REG_SFN_BASE_LEN, 0, 16, + MLXSW_REG_SFN_REC_LEN, 0x0C, false); + +static inline void mlxsw_reg_sfn_mac_unpack(char *payload, int rec_index, + char *mac, u16 *p_vid, + u16 *p_local_port) +{ + mlxsw_reg_sfn_rec_mac_memcpy_from(payload, rec_index, mac); + *p_vid = mlxsw_reg_sfn_mac_fid_get(payload, rec_index); + *p_local_port = mlxsw_reg_sfn_mac_system_port_get(payload, rec_index); +} + +/* reg_sfn_mac_lag_lag_id + * LAG ID (pointer into the LAG descriptor table). + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sfn, mac_lag_lag_id, MLXSW_REG_SFN_BASE_LEN, 0, 10, + MLXSW_REG_SFN_REC_LEN, 0x0C, false); + +static inline void mlxsw_reg_sfn_mac_lag_unpack(char *payload, int rec_index, + char *mac, u16 *p_vid, + u16 *p_lag_id) +{ + mlxsw_reg_sfn_rec_mac_memcpy_from(payload, rec_index, mac); + *p_vid = mlxsw_reg_sfn_mac_fid_get(payload, rec_index); + *p_lag_id = mlxsw_reg_sfn_mac_lag_lag_id_get(payload, rec_index); +} + +/* reg_sfn_uc_tunnel_uip_msb + * When protocol is IPv4, the most significant byte of the underlay IPv4 + * address of the remote VTEP. + * When protocol is IPv6, reserved. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sfn, uc_tunnel_uip_msb, MLXSW_REG_SFN_BASE_LEN, 24, + 8, MLXSW_REG_SFN_REC_LEN, 0x08, false); + +enum mlxsw_reg_sfn_uc_tunnel_protocol { + MLXSW_REG_SFN_UC_TUNNEL_PROTOCOL_IPV4, + MLXSW_REG_SFN_UC_TUNNEL_PROTOCOL_IPV6, +}; + +/* reg_sfn_uc_tunnel_protocol + * IP protocol. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sfn, uc_tunnel_protocol, MLXSW_REG_SFN_BASE_LEN, 27, + 1, MLXSW_REG_SFN_REC_LEN, 0x0C, false); + +/* reg_sfn_uc_tunnel_uip_lsb + * When protocol is IPv4, the least significant bytes of the underlay + * IPv4 address of the remote VTEP. + * When protocol is IPv6, ipv6_id to be queried from TNIPSD. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sfn, uc_tunnel_uip_lsb, MLXSW_REG_SFN_BASE_LEN, 0, + 24, MLXSW_REG_SFN_REC_LEN, 0x0C, false); + +/* reg_sfn_uc_tunnel_port + * Tunnel port. + * Reserved on Spectrum. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sfn, tunnel_port, MLXSW_REG_SFN_BASE_LEN, 0, 4, + MLXSW_REG_SFN_REC_LEN, 0x10, false); + +static inline void +mlxsw_reg_sfn_uc_tunnel_unpack(char *payload, int rec_index, char *mac, + u16 *p_fid, u32 *p_uip, + enum mlxsw_reg_sfn_uc_tunnel_protocol *p_proto) +{ + u32 uip_msb, uip_lsb; + + mlxsw_reg_sfn_rec_mac_memcpy_from(payload, rec_index, mac); + *p_fid = mlxsw_reg_sfn_mac_fid_get(payload, rec_index); + uip_msb = mlxsw_reg_sfn_uc_tunnel_uip_msb_get(payload, rec_index); + uip_lsb = mlxsw_reg_sfn_uc_tunnel_uip_lsb_get(payload, rec_index); + *p_uip = uip_msb << 24 | uip_lsb; + *p_proto = mlxsw_reg_sfn_uc_tunnel_protocol_get(payload, rec_index); +} + +/* SPMS - Switch Port MSTP/RSTP State Register + * ------------------------------------------- + * Configures the spanning tree state of a physical port. + */ +#define MLXSW_REG_SPMS_ID 0x200D +#define MLXSW_REG_SPMS_LEN 0x404 + +MLXSW_REG_DEFINE(spms, MLXSW_REG_SPMS_ID, MLXSW_REG_SPMS_LEN); + +/* reg_spms_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, spms, 0x00, 16, 0x00, 12); + +enum mlxsw_reg_spms_state { + MLXSW_REG_SPMS_STATE_NO_CHANGE, + MLXSW_REG_SPMS_STATE_DISCARDING, + MLXSW_REG_SPMS_STATE_LEARNING, + MLXSW_REG_SPMS_STATE_FORWARDING, +}; + +/* reg_spms_state + * Spanning tree state of each VLAN ID (VID) of the local port. + * 0 - Do not change spanning tree state (used only when writing). + * 1 - Discarding. No learning or forwarding to/from this port (default). + * 2 - Learning. Port is learning, but not forwarding. + * 3 - Forwarding. Port is learning and forwarding. + * Access: RW + */ +MLXSW_ITEM_BIT_ARRAY(reg, spms, state, 0x04, 0x400, 2); + +static inline void mlxsw_reg_spms_pack(char *payload, u16 local_port) +{ + MLXSW_REG_ZERO(spms, payload); + mlxsw_reg_spms_local_port_set(payload, local_port); +} + +static inline void mlxsw_reg_spms_vid_pack(char *payload, u16 vid, + enum mlxsw_reg_spms_state state) +{ + mlxsw_reg_spms_state_set(payload, vid, state); +} + +/* SPVID - Switch Port VID + * ----------------------- + * The switch port VID configures the default VID for a port. + */ +#define MLXSW_REG_SPVID_ID 0x200E +#define MLXSW_REG_SPVID_LEN 0x08 + +MLXSW_REG_DEFINE(spvid, MLXSW_REG_SPVID_ID, MLXSW_REG_SPVID_LEN); + +/* reg_spvid_tport + * Port is tunnel port. + * Reserved when SwitchX/-2 or Spectrum-1. + * Access: Index + */ +MLXSW_ITEM32(reg, spvid, tport, 0x00, 24, 1); + +/* reg_spvid_local_port + * When tport = 0: Local port number. Not supported for CPU port. + * When tport = 1: Tunnel port. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, spvid, 0x00, 16, 0x00, 12); + +/* reg_spvid_sub_port + * Virtual port within the physical port. + * Should be set to 0 when virtual ports are not enabled on the port. + * Access: Index + */ +MLXSW_ITEM32(reg, spvid, sub_port, 0x00, 8, 8); + +/* reg_spvid_egr_et_set + * When VLAN is pushed at ingress (for untagged packets or for + * QinQ push mode) then the EtherType is decided at the egress port. + * Reserved when Spectrum-1. + * Access: RW + */ +MLXSW_ITEM32(reg, spvid, egr_et_set, 0x04, 24, 1); + +/* reg_spvid_et_vlan + * EtherType used for when VLAN is pushed at ingress (for untagged + * packets or for QinQ push mode). + * 0: ether_type0 - (default) + * 1: ether_type1 + * 2: ether_type2 - Reserved when Spectrum-1, supported by Spectrum-2 + * Ethertype IDs are configured by SVER. + * Reserved when egr_et_set = 1. + * Access: RW + */ +MLXSW_ITEM32(reg, spvid, et_vlan, 0x04, 16, 2); + +/* reg_spvid_pvid + * Port default VID + * Access: RW + */ +MLXSW_ITEM32(reg, spvid, pvid, 0x04, 0, 12); + +static inline void mlxsw_reg_spvid_pack(char *payload, u16 local_port, u16 pvid, + u8 et_vlan) +{ + MLXSW_REG_ZERO(spvid, payload); + mlxsw_reg_spvid_local_port_set(payload, local_port); + mlxsw_reg_spvid_pvid_set(payload, pvid); + mlxsw_reg_spvid_et_vlan_set(payload, et_vlan); +} + +/* SPVM - Switch Port VLAN Membership + * ---------------------------------- + * The Switch Port VLAN Membership register configures the VLAN membership + * of a port in a VLAN denoted by VID. VLAN membership is managed per + * virtual port. The register can be used to add and remove VID(s) from a port. + */ +#define MLXSW_REG_SPVM_ID 0x200F +#define MLXSW_REG_SPVM_BASE_LEN 0x04 /* base length, without records */ +#define MLXSW_REG_SPVM_REC_LEN 0x04 /* record length */ +#define MLXSW_REG_SPVM_REC_MAX_COUNT 255 +#define MLXSW_REG_SPVM_LEN (MLXSW_REG_SPVM_BASE_LEN + \ + MLXSW_REG_SPVM_REC_LEN * MLXSW_REG_SPVM_REC_MAX_COUNT) + +MLXSW_REG_DEFINE(spvm, MLXSW_REG_SPVM_ID, MLXSW_REG_SPVM_LEN); + +/* reg_spvm_pt + * Priority tagged. If this bit is set, packets forwarded to the port with + * untagged VLAN membership (u bit is set) will be tagged with priority tag + * (VID=0) + * Access: RW + */ +MLXSW_ITEM32(reg, spvm, pt, 0x00, 31, 1); + +/* reg_spvm_pte + * Priority Tagged Update Enable. On Write operations, if this bit is cleared, + * the pt bit will NOT be updated. To update the pt bit, pte must be set. + * Access: WO + */ +MLXSW_ITEM32(reg, spvm, pte, 0x00, 30, 1); + +/* reg_spvm_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, spvm, 0x00, 16, 0x00, 12); + +/* reg_spvm_sub_port + * Virtual port within the physical port. + * Should be set to 0 when virtual ports are not enabled on the port. + * Access: Index + */ +MLXSW_ITEM32(reg, spvm, sub_port, 0x00, 8, 8); + +/* reg_spvm_num_rec + * Number of records to update. Each record contains: i, e, u, vid. + * Access: OP + */ +MLXSW_ITEM32(reg, spvm, num_rec, 0x00, 0, 8); + +/* reg_spvm_rec_i + * Ingress membership in VLAN ID. + * Access: Index + */ +MLXSW_ITEM32_INDEXED(reg, spvm, rec_i, + MLXSW_REG_SPVM_BASE_LEN, 14, 1, + MLXSW_REG_SPVM_REC_LEN, 0, false); + +/* reg_spvm_rec_e + * Egress membership in VLAN ID. + * Access: Index + */ +MLXSW_ITEM32_INDEXED(reg, spvm, rec_e, + MLXSW_REG_SPVM_BASE_LEN, 13, 1, + MLXSW_REG_SPVM_REC_LEN, 0, false); + +/* reg_spvm_rec_u + * Untagged - port is an untagged member - egress transmission uses untagged + * frames on VID<n> + * Access: Index + */ +MLXSW_ITEM32_INDEXED(reg, spvm, rec_u, + MLXSW_REG_SPVM_BASE_LEN, 12, 1, + MLXSW_REG_SPVM_REC_LEN, 0, false); + +/* reg_spvm_rec_vid + * Egress membership in VLAN ID. + * Access: Index + */ +MLXSW_ITEM32_INDEXED(reg, spvm, rec_vid, + MLXSW_REG_SPVM_BASE_LEN, 0, 12, + MLXSW_REG_SPVM_REC_LEN, 0, false); + +static inline void mlxsw_reg_spvm_pack(char *payload, u16 local_port, + u16 vid_begin, u16 vid_end, + bool is_member, bool untagged) +{ + int size = vid_end - vid_begin + 1; + int i; + + MLXSW_REG_ZERO(spvm, payload); + mlxsw_reg_spvm_local_port_set(payload, local_port); + mlxsw_reg_spvm_num_rec_set(payload, size); + + for (i = 0; i < size; i++) { + mlxsw_reg_spvm_rec_i_set(payload, i, is_member); + mlxsw_reg_spvm_rec_e_set(payload, i, is_member); + mlxsw_reg_spvm_rec_u_set(payload, i, untagged); + mlxsw_reg_spvm_rec_vid_set(payload, i, vid_begin + i); + } +} + +/* SPAFT - Switch Port Acceptable Frame Types + * ------------------------------------------ + * The Switch Port Acceptable Frame Types register configures the frame + * admittance of the port. + */ +#define MLXSW_REG_SPAFT_ID 0x2010 +#define MLXSW_REG_SPAFT_LEN 0x08 + +MLXSW_REG_DEFINE(spaft, MLXSW_REG_SPAFT_ID, MLXSW_REG_SPAFT_LEN); + +/* reg_spaft_local_port + * Local port number. + * Access: Index + * + * Note: CPU port is not supported (all tag types are allowed). + */ +MLXSW_ITEM32_LP(reg, spaft, 0x00, 16, 0x00, 12); + +/* reg_spaft_sub_port + * Virtual port within the physical port. + * Should be set to 0 when virtual ports are not enabled on the port. + * Access: RW + */ +MLXSW_ITEM32(reg, spaft, sub_port, 0x00, 8, 8); + +/* reg_spaft_allow_untagged + * When set, untagged frames on the ingress are allowed (default). + * Access: RW + */ +MLXSW_ITEM32(reg, spaft, allow_untagged, 0x04, 31, 1); + +/* reg_spaft_allow_prio_tagged + * When set, priority tagged frames on the ingress are allowed (default). + * Access: RW + */ +MLXSW_ITEM32(reg, spaft, allow_prio_tagged, 0x04, 30, 1); + +/* reg_spaft_allow_tagged + * When set, tagged frames on the ingress are allowed (default). + * Access: RW + */ +MLXSW_ITEM32(reg, spaft, allow_tagged, 0x04, 29, 1); + +static inline void mlxsw_reg_spaft_pack(char *payload, u16 local_port, + bool allow_untagged) +{ + MLXSW_REG_ZERO(spaft, payload); + mlxsw_reg_spaft_local_port_set(payload, local_port); + mlxsw_reg_spaft_allow_untagged_set(payload, allow_untagged); + mlxsw_reg_spaft_allow_prio_tagged_set(payload, allow_untagged); + mlxsw_reg_spaft_allow_tagged_set(payload, true); +} + +/* SFGC - Switch Flooding Group Configuration + * ------------------------------------------ + * The following register controls the association of flooding tables and MIDs + * to packet types used for flooding. + */ +#define MLXSW_REG_SFGC_ID 0x2011 +#define MLXSW_REG_SFGC_LEN 0x14 + +MLXSW_REG_DEFINE(sfgc, MLXSW_REG_SFGC_ID, MLXSW_REG_SFGC_LEN); + +enum mlxsw_reg_sfgc_type { + MLXSW_REG_SFGC_TYPE_BROADCAST, + MLXSW_REG_SFGC_TYPE_UNKNOWN_UNICAST, + MLXSW_REG_SFGC_TYPE_UNREGISTERED_MULTICAST_IPV4, + MLXSW_REG_SFGC_TYPE_UNREGISTERED_MULTICAST_IPV6, + MLXSW_REG_SFGC_TYPE_RESERVED, + MLXSW_REG_SFGC_TYPE_UNREGISTERED_MULTICAST_NON_IP, + MLXSW_REG_SFGC_TYPE_IPV4_LINK_LOCAL, + MLXSW_REG_SFGC_TYPE_IPV6_ALL_HOST, + MLXSW_REG_SFGC_TYPE_MAX, +}; + +/* reg_sfgc_type + * The traffic type to reach the flooding table. + * Access: Index + */ +MLXSW_ITEM32(reg, sfgc, type, 0x00, 0, 4); + +/* bridge_type is used in SFGC and SFMR. */ +enum mlxsw_reg_bridge_type { + MLXSW_REG_BRIDGE_TYPE_0 = 0, /* Used for .1q FIDs. */ + MLXSW_REG_BRIDGE_TYPE_1 = 1, /* Used for .1d FIDs. */ +}; + +/* reg_sfgc_bridge_type + * Access: Index + * + * Note: SwitchX-2 only supports 802.1Q mode. + */ +MLXSW_ITEM32(reg, sfgc, bridge_type, 0x04, 24, 3); + +enum mlxsw_flood_table_type { + MLXSW_REG_SFGC_TABLE_TYPE_VID = 1, + MLXSW_REG_SFGC_TABLE_TYPE_SINGLE = 2, + MLXSW_REG_SFGC_TABLE_TYPE_ANY = 0, + MLXSW_REG_SFGC_TABLE_TYPE_FID_OFFSET = 3, + MLXSW_REG_SFGC_TABLE_TYPE_FID = 4, +}; + +/* reg_sfgc_table_type + * See mlxsw_flood_table_type + * Access: RW + * + * Note: FID offset and FID types are not supported in SwitchX-2. + */ +MLXSW_ITEM32(reg, sfgc, table_type, 0x04, 16, 3); + +/* reg_sfgc_flood_table + * Flooding table index to associate with the specific type on the specific + * switch partition. + * Access: RW + */ +MLXSW_ITEM32(reg, sfgc, flood_table, 0x04, 0, 6); + +/* reg_sfgc_counter_set_type + * Counter Set Type for flow counters. + * Access: RW + */ +MLXSW_ITEM32(reg, sfgc, counter_set_type, 0x0C, 24, 8); + +/* reg_sfgc_counter_index + * Counter Index for flow counters. + * Access: RW + */ +MLXSW_ITEM32(reg, sfgc, counter_index, 0x0C, 0, 24); + +/* reg_sfgc_mid_base + * MID Base. + * Access: RW + * + * Note: Reserved when legacy bridge model is used. + */ +MLXSW_ITEM32(reg, sfgc, mid_base, 0x10, 0, 16); + +static inline void +mlxsw_reg_sfgc_pack(char *payload, enum mlxsw_reg_sfgc_type type, + enum mlxsw_reg_bridge_type bridge_type, + enum mlxsw_flood_table_type table_type, + unsigned int flood_table, u16 mid_base) +{ + MLXSW_REG_ZERO(sfgc, payload); + mlxsw_reg_sfgc_type_set(payload, type); + mlxsw_reg_sfgc_bridge_type_set(payload, bridge_type); + mlxsw_reg_sfgc_table_type_set(payload, table_type); + mlxsw_reg_sfgc_flood_table_set(payload, flood_table); + mlxsw_reg_sfgc_mid_base_set(payload, mid_base); +} + +/* SFDF - Switch Filtering DB Flush + * -------------------------------- + * The switch filtering DB flush register is used to flush the FDB. + * Note that FDB notifications are flushed as well. + */ +#define MLXSW_REG_SFDF_ID 0x2013 +#define MLXSW_REG_SFDF_LEN 0x14 + +MLXSW_REG_DEFINE(sfdf, MLXSW_REG_SFDF_ID, MLXSW_REG_SFDF_LEN); + +/* reg_sfdf_swid + * Switch partition ID. + * Access: Index + */ +MLXSW_ITEM32(reg, sfdf, swid, 0x00, 24, 8); + +enum mlxsw_reg_sfdf_flush_type { + MLXSW_REG_SFDF_FLUSH_PER_SWID, + MLXSW_REG_SFDF_FLUSH_PER_FID, + MLXSW_REG_SFDF_FLUSH_PER_PORT, + MLXSW_REG_SFDF_FLUSH_PER_PORT_AND_FID, + MLXSW_REG_SFDF_FLUSH_PER_LAG, + MLXSW_REG_SFDF_FLUSH_PER_LAG_AND_FID, + MLXSW_REG_SFDF_FLUSH_PER_NVE, + MLXSW_REG_SFDF_FLUSH_PER_NVE_AND_FID, +}; + +/* reg_sfdf_flush_type + * Flush type. + * 0 - All SWID dynamic entries are flushed. + * 1 - All FID dynamic entries are flushed. + * 2 - All dynamic entries pointing to port are flushed. + * 3 - All FID dynamic entries pointing to port are flushed. + * 4 - All dynamic entries pointing to LAG are flushed. + * 5 - All FID dynamic entries pointing to LAG are flushed. + * 6 - All entries of type "Unicast Tunnel" or "Multicast Tunnel" are + * flushed. + * 7 - All entries of type "Unicast Tunnel" or "Multicast Tunnel" are + * flushed, per FID. + * Access: RW + */ +MLXSW_ITEM32(reg, sfdf, flush_type, 0x04, 28, 4); + +/* reg_sfdf_flush_static + * Static. + * 0 - Flush only dynamic entries. + * 1 - Flush both dynamic and static entries. + * Access: RW + */ +MLXSW_ITEM32(reg, sfdf, flush_static, 0x04, 24, 1); + +static inline void mlxsw_reg_sfdf_pack(char *payload, + enum mlxsw_reg_sfdf_flush_type type) +{ + MLXSW_REG_ZERO(sfdf, payload); + mlxsw_reg_sfdf_flush_type_set(payload, type); + mlxsw_reg_sfdf_flush_static_set(payload, true); +} + +/* reg_sfdf_fid + * FID to flush. + * Access: RW + */ +MLXSW_ITEM32(reg, sfdf, fid, 0x0C, 0, 16); + +/* reg_sfdf_system_port + * Port to flush. + * Access: RW + */ +MLXSW_ITEM32(reg, sfdf, system_port, 0x0C, 0, 16); + +/* reg_sfdf_port_fid_system_port + * Port to flush, pointed to by FID. + * Access: RW + */ +MLXSW_ITEM32(reg, sfdf, port_fid_system_port, 0x08, 0, 16); + +/* reg_sfdf_lag_id + * LAG ID to flush. + * Access: RW + */ +MLXSW_ITEM32(reg, sfdf, lag_id, 0x0C, 0, 10); + +/* reg_sfdf_lag_fid_lag_id + * LAG ID to flush, pointed to by FID. + * Access: RW + */ +MLXSW_ITEM32(reg, sfdf, lag_fid_lag_id, 0x08, 0, 10); + +/* SLDR - Switch LAG Descriptor Register + * ----------------------------------------- + * The switch LAG descriptor register is populated by LAG descriptors. + * Each LAG descriptor is indexed by lag_id. The LAG ID runs from 0 to + * max_lag-1. + */ +#define MLXSW_REG_SLDR_ID 0x2014 +#define MLXSW_REG_SLDR_LEN 0x0C /* counting in only one port in list */ + +MLXSW_REG_DEFINE(sldr, MLXSW_REG_SLDR_ID, MLXSW_REG_SLDR_LEN); + +enum mlxsw_reg_sldr_op { + /* Indicates a creation of a new LAG-ID, lag_id must be valid */ + MLXSW_REG_SLDR_OP_LAG_CREATE, + MLXSW_REG_SLDR_OP_LAG_DESTROY, + /* Ports that appear in the list have the Distributor enabled */ + MLXSW_REG_SLDR_OP_LAG_ADD_PORT_LIST, + /* Removes ports from the disributor list */ + MLXSW_REG_SLDR_OP_LAG_REMOVE_PORT_LIST, +}; + +/* reg_sldr_op + * Operation. + * Access: RW + */ +MLXSW_ITEM32(reg, sldr, op, 0x00, 29, 3); + +/* reg_sldr_lag_id + * LAG identifier. The lag_id is the index into the LAG descriptor table. + * Access: Index + */ +MLXSW_ITEM32(reg, sldr, lag_id, 0x00, 0, 10); + +static inline void mlxsw_reg_sldr_lag_create_pack(char *payload, u8 lag_id) +{ + MLXSW_REG_ZERO(sldr, payload); + mlxsw_reg_sldr_op_set(payload, MLXSW_REG_SLDR_OP_LAG_CREATE); + mlxsw_reg_sldr_lag_id_set(payload, lag_id); +} + +static inline void mlxsw_reg_sldr_lag_destroy_pack(char *payload, u8 lag_id) +{ + MLXSW_REG_ZERO(sldr, payload); + mlxsw_reg_sldr_op_set(payload, MLXSW_REG_SLDR_OP_LAG_DESTROY); + mlxsw_reg_sldr_lag_id_set(payload, lag_id); +} + +/* reg_sldr_num_ports + * The number of member ports of the LAG. + * Reserved for Create / Destroy operations + * For Add / Remove operations - indicates the number of ports in the list. + * Access: RW + */ +MLXSW_ITEM32(reg, sldr, num_ports, 0x04, 24, 8); + +/* reg_sldr_system_port + * System port. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, sldr, system_port, 0x08, 0, 16, 4, 0, false); + +static inline void mlxsw_reg_sldr_lag_add_port_pack(char *payload, u8 lag_id, + u16 local_port) +{ + MLXSW_REG_ZERO(sldr, payload); + mlxsw_reg_sldr_op_set(payload, MLXSW_REG_SLDR_OP_LAG_ADD_PORT_LIST); + mlxsw_reg_sldr_lag_id_set(payload, lag_id); + mlxsw_reg_sldr_num_ports_set(payload, 1); + mlxsw_reg_sldr_system_port_set(payload, 0, local_port); +} + +static inline void mlxsw_reg_sldr_lag_remove_port_pack(char *payload, u8 lag_id, + u16 local_port) +{ + MLXSW_REG_ZERO(sldr, payload); + mlxsw_reg_sldr_op_set(payload, MLXSW_REG_SLDR_OP_LAG_REMOVE_PORT_LIST); + mlxsw_reg_sldr_lag_id_set(payload, lag_id); + mlxsw_reg_sldr_num_ports_set(payload, 1); + mlxsw_reg_sldr_system_port_set(payload, 0, local_port); +} + +/* SLCR - Switch LAG Configuration 2 Register + * ------------------------------------------- + * The Switch LAG Configuration register is used for configuring the + * LAG properties of the switch. + */ +#define MLXSW_REG_SLCR_ID 0x2015 +#define MLXSW_REG_SLCR_LEN 0x10 + +MLXSW_REG_DEFINE(slcr, MLXSW_REG_SLCR_ID, MLXSW_REG_SLCR_LEN); + +enum mlxsw_reg_slcr_pp { + /* Global Configuration (for all ports) */ + MLXSW_REG_SLCR_PP_GLOBAL, + /* Per port configuration, based on local_port field */ + MLXSW_REG_SLCR_PP_PER_PORT, +}; + +/* reg_slcr_pp + * Per Port Configuration + * Note: Reading at Global mode results in reading port 1 configuration. + * Access: Index + */ +MLXSW_ITEM32(reg, slcr, pp, 0x00, 24, 1); + +/* reg_slcr_local_port + * Local port number + * Supported from CPU port + * Not supported from router port + * Reserved when pp = Global Configuration + * Access: Index + */ +MLXSW_ITEM32_LP(reg, slcr, 0x00, 16, 0x00, 12); + +enum mlxsw_reg_slcr_type { + MLXSW_REG_SLCR_TYPE_CRC, /* default */ + MLXSW_REG_SLCR_TYPE_XOR, + MLXSW_REG_SLCR_TYPE_RANDOM, +}; + +/* reg_slcr_type + * Hash type + * Access: RW + */ +MLXSW_ITEM32(reg, slcr, type, 0x00, 0, 4); + +/* Ingress port */ +#define MLXSW_REG_SLCR_LAG_HASH_IN_PORT BIT(0) +/* SMAC - for IPv4 and IPv6 packets */ +#define MLXSW_REG_SLCR_LAG_HASH_SMAC_IP BIT(1) +/* SMAC - for non-IP packets */ +#define MLXSW_REG_SLCR_LAG_HASH_SMAC_NONIP BIT(2) +#define MLXSW_REG_SLCR_LAG_HASH_SMAC \ + (MLXSW_REG_SLCR_LAG_HASH_SMAC_IP | \ + MLXSW_REG_SLCR_LAG_HASH_SMAC_NONIP) +/* DMAC - for IPv4 and IPv6 packets */ +#define MLXSW_REG_SLCR_LAG_HASH_DMAC_IP BIT(3) +/* DMAC - for non-IP packets */ +#define MLXSW_REG_SLCR_LAG_HASH_DMAC_NONIP BIT(4) +#define MLXSW_REG_SLCR_LAG_HASH_DMAC \ + (MLXSW_REG_SLCR_LAG_HASH_DMAC_IP | \ + MLXSW_REG_SLCR_LAG_HASH_DMAC_NONIP) +/* Ethertype - for IPv4 and IPv6 packets */ +#define MLXSW_REG_SLCR_LAG_HASH_ETHERTYPE_IP BIT(5) +/* Ethertype - for non-IP packets */ +#define MLXSW_REG_SLCR_LAG_HASH_ETHERTYPE_NONIP BIT(6) +#define MLXSW_REG_SLCR_LAG_HASH_ETHERTYPE \ + (MLXSW_REG_SLCR_LAG_HASH_ETHERTYPE_IP | \ + MLXSW_REG_SLCR_LAG_HASH_ETHERTYPE_NONIP) +/* VLAN ID - for IPv4 and IPv6 packets */ +#define MLXSW_REG_SLCR_LAG_HASH_VLANID_IP BIT(7) +/* VLAN ID - for non-IP packets */ +#define MLXSW_REG_SLCR_LAG_HASH_VLANID_NONIP BIT(8) +#define MLXSW_REG_SLCR_LAG_HASH_VLANID \ + (MLXSW_REG_SLCR_LAG_HASH_VLANID_IP | \ + MLXSW_REG_SLCR_LAG_HASH_VLANID_NONIP) +/* Source IP address (can be IPv4 or IPv6) */ +#define MLXSW_REG_SLCR_LAG_HASH_SIP BIT(9) +/* Destination IP address (can be IPv4 or IPv6) */ +#define MLXSW_REG_SLCR_LAG_HASH_DIP BIT(10) +/* TCP/UDP source port */ +#define MLXSW_REG_SLCR_LAG_HASH_SPORT BIT(11) +/* TCP/UDP destination port*/ +#define MLXSW_REG_SLCR_LAG_HASH_DPORT BIT(12) +/* IPv4 Protocol/IPv6 Next Header */ +#define MLXSW_REG_SLCR_LAG_HASH_IPPROTO BIT(13) +/* IPv6 Flow label */ +#define MLXSW_REG_SLCR_LAG_HASH_FLOWLABEL BIT(14) +/* SID - FCoE source ID */ +#define MLXSW_REG_SLCR_LAG_HASH_FCOE_SID BIT(15) +/* DID - FCoE destination ID */ +#define MLXSW_REG_SLCR_LAG_HASH_FCOE_DID BIT(16) +/* OXID - FCoE originator exchange ID */ +#define MLXSW_REG_SLCR_LAG_HASH_FCOE_OXID BIT(17) +/* Destination QP number - for RoCE packets */ +#define MLXSW_REG_SLCR_LAG_HASH_ROCE_DQP BIT(19) + +/* reg_slcr_lag_hash + * LAG hashing configuration. This is a bitmask, in which each set + * bit includes the corresponding item in the LAG hash calculation. + * The default lag_hash contains SMAC, DMAC, VLANID and + * Ethertype (for all packet types). + * Access: RW + */ +MLXSW_ITEM32(reg, slcr, lag_hash, 0x04, 0, 20); + +/* reg_slcr_seed + * LAG seed value. The seed is the same for all ports. + * Access: RW + */ +MLXSW_ITEM32(reg, slcr, seed, 0x08, 0, 32); + +static inline void mlxsw_reg_slcr_pack(char *payload, u16 lag_hash, u32 seed) +{ + MLXSW_REG_ZERO(slcr, payload); + mlxsw_reg_slcr_pp_set(payload, MLXSW_REG_SLCR_PP_GLOBAL); + mlxsw_reg_slcr_type_set(payload, MLXSW_REG_SLCR_TYPE_CRC); + mlxsw_reg_slcr_lag_hash_set(payload, lag_hash); + mlxsw_reg_slcr_seed_set(payload, seed); +} + +/* SLCOR - Switch LAG Collector Register + * ------------------------------------- + * The Switch LAG Collector register controls the Local Port membership + * in a LAG and enablement of the collector. + */ +#define MLXSW_REG_SLCOR_ID 0x2016 +#define MLXSW_REG_SLCOR_LEN 0x10 + +MLXSW_REG_DEFINE(slcor, MLXSW_REG_SLCOR_ID, MLXSW_REG_SLCOR_LEN); + +enum mlxsw_reg_slcor_col { + /* Port is added with collector disabled */ + MLXSW_REG_SLCOR_COL_LAG_ADD_PORT, + MLXSW_REG_SLCOR_COL_LAG_COLLECTOR_ENABLED, + MLXSW_REG_SLCOR_COL_LAG_COLLECTOR_DISABLED, + MLXSW_REG_SLCOR_COL_LAG_REMOVE_PORT, +}; + +/* reg_slcor_col + * Collector configuration + * Access: RW + */ +MLXSW_ITEM32(reg, slcor, col, 0x00, 30, 2); + +/* reg_slcor_local_port + * Local port number + * Not supported for CPU port + * Access: Index + */ +MLXSW_ITEM32_LP(reg, slcor, 0x00, 16, 0x00, 12); + +/* reg_slcor_lag_id + * LAG Identifier. Index into the LAG descriptor table. + * Access: Index + */ +MLXSW_ITEM32(reg, slcor, lag_id, 0x00, 0, 10); + +/* reg_slcor_port_index + * Port index in the LAG list. Only valid on Add Port to LAG col. + * Valid range is from 0 to cap_max_lag_members-1 + * Access: RW + */ +MLXSW_ITEM32(reg, slcor, port_index, 0x04, 0, 10); + +static inline void mlxsw_reg_slcor_pack(char *payload, + u16 local_port, u16 lag_id, + enum mlxsw_reg_slcor_col col) +{ + MLXSW_REG_ZERO(slcor, payload); + mlxsw_reg_slcor_col_set(payload, col); + mlxsw_reg_slcor_local_port_set(payload, local_port); + mlxsw_reg_slcor_lag_id_set(payload, lag_id); +} + +static inline void mlxsw_reg_slcor_port_add_pack(char *payload, + u16 local_port, u16 lag_id, + u8 port_index) +{ + mlxsw_reg_slcor_pack(payload, local_port, lag_id, + MLXSW_REG_SLCOR_COL_LAG_ADD_PORT); + mlxsw_reg_slcor_port_index_set(payload, port_index); +} + +static inline void mlxsw_reg_slcor_port_remove_pack(char *payload, + u16 local_port, u16 lag_id) +{ + mlxsw_reg_slcor_pack(payload, local_port, lag_id, + MLXSW_REG_SLCOR_COL_LAG_REMOVE_PORT); +} + +static inline void mlxsw_reg_slcor_col_enable_pack(char *payload, + u16 local_port, u16 lag_id) +{ + mlxsw_reg_slcor_pack(payload, local_port, lag_id, + MLXSW_REG_SLCOR_COL_LAG_COLLECTOR_ENABLED); +} + +static inline void mlxsw_reg_slcor_col_disable_pack(char *payload, + u16 local_port, u16 lag_id) +{ + mlxsw_reg_slcor_pack(payload, local_port, lag_id, + MLXSW_REG_SLCOR_COL_LAG_COLLECTOR_ENABLED); +} + +/* SPMLR - Switch Port MAC Learning Register + * ----------------------------------------- + * Controls the Switch MAC learning policy per port. + */ +#define MLXSW_REG_SPMLR_ID 0x2018 +#define MLXSW_REG_SPMLR_LEN 0x8 + +MLXSW_REG_DEFINE(spmlr, MLXSW_REG_SPMLR_ID, MLXSW_REG_SPMLR_LEN); + +/* reg_spmlr_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, spmlr, 0x00, 16, 0x00, 12); + +/* reg_spmlr_sub_port + * Virtual port within the physical port. + * Should be set to 0 when virtual ports are not enabled on the port. + * Access: Index + */ +MLXSW_ITEM32(reg, spmlr, sub_port, 0x00, 8, 8); + +enum mlxsw_reg_spmlr_learn_mode { + MLXSW_REG_SPMLR_LEARN_MODE_DISABLE = 0, + MLXSW_REG_SPMLR_LEARN_MODE_ENABLE = 2, + MLXSW_REG_SPMLR_LEARN_MODE_SEC = 3, +}; + +/* reg_spmlr_learn_mode + * Learning mode on the port. + * 0 - Learning disabled. + * 2 - Learning enabled. + * 3 - Security mode. + * + * In security mode the switch does not learn MACs on the port, but uses the + * SMAC to see if it exists on another ingress port. If so, the packet is + * classified as a bad packet and is discarded unless the software registers + * to receive port security error packets usign HPKT. + */ +MLXSW_ITEM32(reg, spmlr, learn_mode, 0x04, 30, 2); + +static inline void mlxsw_reg_spmlr_pack(char *payload, u16 local_port, + enum mlxsw_reg_spmlr_learn_mode mode) +{ + MLXSW_REG_ZERO(spmlr, payload); + mlxsw_reg_spmlr_local_port_set(payload, local_port); + mlxsw_reg_spmlr_sub_port_set(payload, 0); + mlxsw_reg_spmlr_learn_mode_set(payload, mode); +} + +/* SVFA - Switch VID to FID Allocation Register + * -------------------------------------------- + * Controls the VID to FID mapping and {Port, VID} to FID mapping for + * virtualized ports. + */ +#define MLXSW_REG_SVFA_ID 0x201C +#define MLXSW_REG_SVFA_LEN 0x18 + +MLXSW_REG_DEFINE(svfa, MLXSW_REG_SVFA_ID, MLXSW_REG_SVFA_LEN); + +/* reg_svfa_swid + * Switch partition ID. + * Access: Index + */ +MLXSW_ITEM32(reg, svfa, swid, 0x00, 24, 8); + +/* reg_svfa_local_port + * Local port number. + * Access: Index + * + * Note: Reserved for 802.1Q FIDs. + */ +MLXSW_ITEM32_LP(reg, svfa, 0x00, 16, 0x00, 12); + +enum mlxsw_reg_svfa_mt { + MLXSW_REG_SVFA_MT_VID_TO_FID, + MLXSW_REG_SVFA_MT_PORT_VID_TO_FID, + MLXSW_REG_SVFA_MT_VNI_TO_FID, +}; + +/* reg_svfa_mapping_table + * Mapping table: + * 0 - VID to FID + * 1 - {Port, VID} to FID + * Access: Index + * + * Note: Reserved for SwitchX-2. + */ +MLXSW_ITEM32(reg, svfa, mapping_table, 0x00, 8, 3); + +/* reg_svfa_v + * Valid. + * Valid if set. + * Access: RW + * + * Note: Reserved for SwitchX-2. + */ +MLXSW_ITEM32(reg, svfa, v, 0x00, 0, 1); + +/* reg_svfa_fid + * Filtering ID. + * Access: RW + */ +MLXSW_ITEM32(reg, svfa, fid, 0x04, 16, 16); + +/* reg_svfa_vid + * VLAN ID. + * Access: Index + */ +MLXSW_ITEM32(reg, svfa, vid, 0x04, 0, 12); + +/* reg_svfa_counter_set_type + * Counter set type for flow counters. + * Access: RW + * + * Note: Reserved for SwitchX-2. + */ +MLXSW_ITEM32(reg, svfa, counter_set_type, 0x08, 24, 8); + +/* reg_svfa_counter_index + * Counter index for flow counters. + * Access: RW + * + * Note: Reserved for SwitchX-2. + */ +MLXSW_ITEM32(reg, svfa, counter_index, 0x08, 0, 24); + +/* reg_svfa_vni + * Virtual Network Identifier. + * Access: Index + * + * Note: Reserved when mapping_table is not 2 (VNI mapping table). + */ +MLXSW_ITEM32(reg, svfa, vni, 0x10, 0, 24); + +/* reg_svfa_irif_v + * Ingress RIF valid. + * 0 - Ingress RIF is not valid, no ingress RIF assigned. + * 1 - Ingress RIF valid. + * Must not be set for a non enabled RIF. + * Access: RW + * + * Note: Reserved when legacy bridge model is used. + */ +MLXSW_ITEM32(reg, svfa, irif_v, 0x14, 24, 1); + +/* reg_svfa_irif + * Ingress RIF (Router Interface). + * Range is 0..cap_max_router_interfaces-1. + * Access: RW + * + * Note: Reserved when legacy bridge model is used and when irif_v=0. + */ +MLXSW_ITEM32(reg, svfa, irif, 0x14, 0, 16); + +static inline void __mlxsw_reg_svfa_pack(char *payload, + enum mlxsw_reg_svfa_mt mt, bool valid, + u16 fid, bool irif_v, u16 irif) +{ + MLXSW_REG_ZERO(svfa, payload); + mlxsw_reg_svfa_swid_set(payload, 0); + mlxsw_reg_svfa_mapping_table_set(payload, mt); + mlxsw_reg_svfa_v_set(payload, valid); + mlxsw_reg_svfa_fid_set(payload, fid); + mlxsw_reg_svfa_irif_v_set(payload, irif_v); + mlxsw_reg_svfa_irif_set(payload, irif_v ? irif : 0); +} + +static inline void mlxsw_reg_svfa_port_vid_pack(char *payload, u16 local_port, + bool valid, u16 fid, u16 vid, + bool irif_v, u16 irif) +{ + enum mlxsw_reg_svfa_mt mt = MLXSW_REG_SVFA_MT_PORT_VID_TO_FID; + + __mlxsw_reg_svfa_pack(payload, mt, valid, fid, irif_v, irif); + mlxsw_reg_svfa_local_port_set(payload, local_port); + mlxsw_reg_svfa_vid_set(payload, vid); +} + +static inline void mlxsw_reg_svfa_vid_pack(char *payload, bool valid, u16 fid, + u16 vid, bool irif_v, u16 irif) +{ + enum mlxsw_reg_svfa_mt mt = MLXSW_REG_SVFA_MT_VID_TO_FID; + + __mlxsw_reg_svfa_pack(payload, mt, valid, fid, irif_v, irif); + mlxsw_reg_svfa_vid_set(payload, vid); +} + +static inline void mlxsw_reg_svfa_vni_pack(char *payload, bool valid, u16 fid, + u32 vni, bool irif_v, u16 irif) +{ + enum mlxsw_reg_svfa_mt mt = MLXSW_REG_SVFA_MT_VNI_TO_FID; + + __mlxsw_reg_svfa_pack(payload, mt, valid, fid, irif_v, irif); + mlxsw_reg_svfa_vni_set(payload, vni); +} + +/* SPVTR - Switch Port VLAN Stacking Register + * ------------------------------------------ + * The Switch Port VLAN Stacking register configures the VLAN mode of the port + * to enable VLAN stacking. + */ +#define MLXSW_REG_SPVTR_ID 0x201D +#define MLXSW_REG_SPVTR_LEN 0x10 + +MLXSW_REG_DEFINE(spvtr, MLXSW_REG_SPVTR_ID, MLXSW_REG_SPVTR_LEN); + +/* reg_spvtr_tport + * Port is tunnel port. + * Access: Index + * + * Note: Reserved when SwitchX/-2 or Spectrum-1. + */ +MLXSW_ITEM32(reg, spvtr, tport, 0x00, 24, 1); + +/* reg_spvtr_local_port + * When tport = 0: local port number (Not supported from/to CPU). + * When tport = 1: tunnel port. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, spvtr, 0x00, 16, 0x00, 12); + +/* reg_spvtr_ippe + * Ingress Port Prio Mode Update Enable. + * When set, the Port Prio Mode is updated with the provided ipprio_mode field. + * Reserved on Get operations. + * Access: OP + */ +MLXSW_ITEM32(reg, spvtr, ippe, 0x04, 31, 1); + +/* reg_spvtr_ipve + * Ingress Port VID Mode Update Enable. + * When set, the Ingress Port VID Mode is updated with the provided ipvid_mode + * field. + * Reserved on Get operations. + * Access: OP + */ +MLXSW_ITEM32(reg, spvtr, ipve, 0x04, 30, 1); + +/* reg_spvtr_epve + * Egress Port VID Mode Update Enable. + * When set, the Egress Port VID Mode is updated with the provided epvid_mode + * field. + * Access: OP + */ +MLXSW_ITEM32(reg, spvtr, epve, 0x04, 29, 1); + +/* reg_spvtr_ipprio_mode + * Ingress Port Priority Mode. + * This controls the PCP and DEI of the new outer VLAN + * Note: for SwitchX/-2 the DEI is not affected. + * 0: use port default PCP and DEI (configured by QPDPC). + * 1: use C-VLAN PCP and DEI. + * Has no effect when ipvid_mode = 0. + * Reserved when tport = 1. + * Access: RW + */ +MLXSW_ITEM32(reg, spvtr, ipprio_mode, 0x04, 20, 4); + +enum mlxsw_reg_spvtr_ipvid_mode { + /* IEEE Compliant PVID (default) */ + MLXSW_REG_SPVTR_IPVID_MODE_IEEE_COMPLIANT_PVID, + /* Push VLAN (for VLAN stacking, except prio tagged packets) */ + MLXSW_REG_SPVTR_IPVID_MODE_PUSH_VLAN_FOR_UNTAGGED_PACKET, + /* Always push VLAN (also for prio tagged packets) */ + MLXSW_REG_SPVTR_IPVID_MODE_ALWAYS_PUSH_VLAN, +}; + +/* reg_spvtr_ipvid_mode + * Ingress Port VLAN-ID Mode. + * For Spectrum family, this affects the values of SPVM.i + * Access: RW + */ +MLXSW_ITEM32(reg, spvtr, ipvid_mode, 0x04, 16, 4); + +enum mlxsw_reg_spvtr_epvid_mode { + /* IEEE Compliant VLAN membership */ + MLXSW_REG_SPVTR_EPVID_MODE_IEEE_COMPLIANT_VLAN_MEMBERSHIP, + /* Pop VLAN (for VLAN stacking) */ + MLXSW_REG_SPVTR_EPVID_MODE_POP_VLAN, +}; + +/* reg_spvtr_epvid_mode + * Egress Port VLAN-ID Mode. + * For Spectrum family, this affects the values of SPVM.e,u,pt. + * Access: WO + */ +MLXSW_ITEM32(reg, spvtr, epvid_mode, 0x04, 0, 4); + +static inline void mlxsw_reg_spvtr_pack(char *payload, bool tport, + u16 local_port, + enum mlxsw_reg_spvtr_ipvid_mode ipvid_mode) +{ + MLXSW_REG_ZERO(spvtr, payload); + mlxsw_reg_spvtr_tport_set(payload, tport); + mlxsw_reg_spvtr_local_port_set(payload, local_port); + mlxsw_reg_spvtr_ipvid_mode_set(payload, ipvid_mode); + mlxsw_reg_spvtr_ipve_set(payload, true); +} + +/* SVPE - Switch Virtual-Port Enabling Register + * -------------------------------------------- + * Enables port virtualization. + */ +#define MLXSW_REG_SVPE_ID 0x201E +#define MLXSW_REG_SVPE_LEN 0x4 + +MLXSW_REG_DEFINE(svpe, MLXSW_REG_SVPE_ID, MLXSW_REG_SVPE_LEN); + +/* reg_svpe_local_port + * Local port number + * Access: Index + * + * Note: CPU port is not supported (uses VLAN mode only). + */ +MLXSW_ITEM32_LP(reg, svpe, 0x00, 16, 0x00, 12); + +/* reg_svpe_vp_en + * Virtual port enable. + * 0 - Disable, VLAN mode (VID to FID). + * 1 - Enable, Virtual port mode ({Port, VID} to FID). + * Access: RW + */ +MLXSW_ITEM32(reg, svpe, vp_en, 0x00, 8, 1); + +static inline void mlxsw_reg_svpe_pack(char *payload, u16 local_port, + bool enable) +{ + MLXSW_REG_ZERO(svpe, payload); + mlxsw_reg_svpe_local_port_set(payload, local_port); + mlxsw_reg_svpe_vp_en_set(payload, enable); +} + +/* SFMR - Switch FID Management Register + * ------------------------------------- + * Creates and configures FIDs. + */ +#define MLXSW_REG_SFMR_ID 0x201F +#define MLXSW_REG_SFMR_LEN 0x30 + +MLXSW_REG_DEFINE(sfmr, MLXSW_REG_SFMR_ID, MLXSW_REG_SFMR_LEN); + +enum mlxsw_reg_sfmr_op { + MLXSW_REG_SFMR_OP_CREATE_FID, + MLXSW_REG_SFMR_OP_DESTROY_FID, +}; + +/* reg_sfmr_op + * Operation. + * 0 - Create or edit FID. + * 1 - Destroy FID. + * Access: WO + */ +MLXSW_ITEM32(reg, sfmr, op, 0x00, 24, 4); + +/* reg_sfmr_fid + * Filtering ID. + * Access: Index + */ +MLXSW_ITEM32(reg, sfmr, fid, 0x00, 0, 16); + +/* reg_sfmr_flood_rsp + * Router sub-port flooding table. + * 0 - Regular flooding table. + * 1 - Router sub-port flooding table. For this FID the flooding is per + * router-sub-port local_port. Must not be set for a FID which is not a + * router-sub-port and must be set prior to enabling the relevant RIF. + * Access: RW + * + * Note: Reserved when legacy bridge model is used. + */ +MLXSW_ITEM32(reg, sfmr, flood_rsp, 0x08, 31, 1); + +/* reg_sfmr_flood_bridge_type + * Flood bridge type (see SFGC.bridge_type). + * 0 - type_0. + * 1 - type_1. + * Access: RW + * + * Note: Reserved when legacy bridge model is used and when flood_rsp=1. + */ +MLXSW_ITEM32(reg, sfmr, flood_bridge_type, 0x08, 28, 1); + +/* reg_sfmr_fid_offset + * FID offset. + * Used to point into the flooding table selected by SFGC register if + * the table is of type FID-Offset. Otherwise, this field is reserved. + * Access: RW + */ +MLXSW_ITEM32(reg, sfmr, fid_offset, 0x08, 0, 16); + +/* reg_sfmr_vtfp + * Valid Tunnel Flood Pointer. + * If not set, then nve_tunnel_flood_ptr is reserved and considered NULL. + * Access: RW + * + * Note: Reserved for 802.1Q FIDs. + */ +MLXSW_ITEM32(reg, sfmr, vtfp, 0x0C, 31, 1); + +/* reg_sfmr_nve_tunnel_flood_ptr + * Underlay Flooding and BC Pointer. + * Used as a pointer to the first entry of the group based link lists of + * flooding or BC entries (for NVE tunnels). + * Access: RW + */ +MLXSW_ITEM32(reg, sfmr, nve_tunnel_flood_ptr, 0x0C, 0, 24); + +/* reg_sfmr_vv + * VNI Valid. + * If not set, then vni is reserved. + * Access: RW + * + * Note: Reserved for 802.1Q FIDs. + */ +MLXSW_ITEM32(reg, sfmr, vv, 0x10, 31, 1); + +/* reg_sfmr_vni + * Virtual Network Identifier. + * When legacy bridge model is used, a given VNI can only be assigned to one + * FID. When unified bridge model is used, it configures only the FID->VNI, + * the VNI->FID is done by SVFA. + * Access: RW + */ +MLXSW_ITEM32(reg, sfmr, vni, 0x10, 0, 24); + +/* reg_sfmr_irif_v + * Ingress RIF valid. + * 0 - Ingress RIF is not valid, no ingress RIF assigned. + * 1 - Ingress RIF valid. + * Must not be set for a non valid RIF. + * Access: RW + * + * Note: Reserved when legacy bridge model is used. + */ +MLXSW_ITEM32(reg, sfmr, irif_v, 0x14, 24, 1); + +/* reg_sfmr_irif + * Ingress RIF (Router Interface). + * Range is 0..cap_max_router_interfaces-1. + * Access: RW + * + * Note: Reserved when legacy bridge model is used and when irif_v=0. + */ +MLXSW_ITEM32(reg, sfmr, irif, 0x14, 0, 16); + +/* reg_sfmr_smpe_valid + * SMPE is valid. + * Access: RW + * + * Note: Reserved when legacy bridge model is used, when flood_rsp=1 and on + * Spectrum-1. + */ +MLXSW_ITEM32(reg, sfmr, smpe_valid, 0x28, 20, 1); + +/* reg_sfmr_smpe + * Switch multicast port to egress VID. + * Range is 0..cap_max_rmpe-1 + * Access: RW + * + * Note: Reserved when legacy bridge model is used, when flood_rsp=1 and on + * Spectrum-1. + */ +MLXSW_ITEM32(reg, sfmr, smpe, 0x28, 0, 16); + +static inline void mlxsw_reg_sfmr_pack(char *payload, + enum mlxsw_reg_sfmr_op op, u16 fid, + u16 fid_offset, bool flood_rsp, + enum mlxsw_reg_bridge_type bridge_type, + bool smpe_valid, u16 smpe) +{ + MLXSW_REG_ZERO(sfmr, payload); + mlxsw_reg_sfmr_op_set(payload, op); + mlxsw_reg_sfmr_fid_set(payload, fid); + mlxsw_reg_sfmr_fid_offset_set(payload, fid_offset); + mlxsw_reg_sfmr_vtfp_set(payload, false); + mlxsw_reg_sfmr_vv_set(payload, false); + mlxsw_reg_sfmr_flood_rsp_set(payload, flood_rsp); + mlxsw_reg_sfmr_flood_bridge_type_set(payload, bridge_type); + mlxsw_reg_sfmr_smpe_valid_set(payload, smpe_valid); + mlxsw_reg_sfmr_smpe_set(payload, smpe); +} + +/* SPVMLR - Switch Port VLAN MAC Learning Register + * ----------------------------------------------- + * Controls the switch MAC learning policy per {Port, VID}. + */ +#define MLXSW_REG_SPVMLR_ID 0x2020 +#define MLXSW_REG_SPVMLR_BASE_LEN 0x04 /* base length, without records */ +#define MLXSW_REG_SPVMLR_REC_LEN 0x04 /* record length */ +#define MLXSW_REG_SPVMLR_REC_MAX_COUNT 255 +#define MLXSW_REG_SPVMLR_LEN (MLXSW_REG_SPVMLR_BASE_LEN + \ + MLXSW_REG_SPVMLR_REC_LEN * \ + MLXSW_REG_SPVMLR_REC_MAX_COUNT) + +MLXSW_REG_DEFINE(spvmlr, MLXSW_REG_SPVMLR_ID, MLXSW_REG_SPVMLR_LEN); + +/* reg_spvmlr_local_port + * Local ingress port. + * Access: Index + * + * Note: CPU port is not supported. + */ +MLXSW_ITEM32_LP(reg, spvmlr, 0x00, 16, 0x00, 12); + +/* reg_spvmlr_num_rec + * Number of records to update. + * Access: OP + */ +MLXSW_ITEM32(reg, spvmlr, num_rec, 0x00, 0, 8); + +/* reg_spvmlr_rec_learn_enable + * 0 - Disable learning for {Port, VID}. + * 1 - Enable learning for {Port, VID}. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, spvmlr, rec_learn_enable, MLXSW_REG_SPVMLR_BASE_LEN, + 31, 1, MLXSW_REG_SPVMLR_REC_LEN, 0x00, false); + +/* reg_spvmlr_rec_vid + * VLAN ID to be added/removed from port or for querying. + * Access: Index + */ +MLXSW_ITEM32_INDEXED(reg, spvmlr, rec_vid, MLXSW_REG_SPVMLR_BASE_LEN, 0, 12, + MLXSW_REG_SPVMLR_REC_LEN, 0x00, false); + +static inline void mlxsw_reg_spvmlr_pack(char *payload, u16 local_port, + u16 vid_begin, u16 vid_end, + bool learn_enable) +{ + int num_rec = vid_end - vid_begin + 1; + int i; + + WARN_ON(num_rec < 1 || num_rec > MLXSW_REG_SPVMLR_REC_MAX_COUNT); + + MLXSW_REG_ZERO(spvmlr, payload); + mlxsw_reg_spvmlr_local_port_set(payload, local_port); + mlxsw_reg_spvmlr_num_rec_set(payload, num_rec); + + for (i = 0; i < num_rec; i++) { + mlxsw_reg_spvmlr_rec_learn_enable_set(payload, i, learn_enable); + mlxsw_reg_spvmlr_rec_vid_set(payload, i, vid_begin + i); + } +} + +/* SPVC - Switch Port VLAN Classification Register + * ----------------------------------------------- + * Configures the port to identify packets as untagged / single tagged / + * double packets based on the packet EtherTypes. + * Ethertype IDs are configured by SVER. + */ +#define MLXSW_REG_SPVC_ID 0x2026 +#define MLXSW_REG_SPVC_LEN 0x0C + +MLXSW_REG_DEFINE(spvc, MLXSW_REG_SPVC_ID, MLXSW_REG_SPVC_LEN); + +/* reg_spvc_local_port + * Local port. + * Access: Index + * + * Note: applies both to Rx port and Tx port, so if a packet traverses + * through Rx port i and a Tx port j then port i and port j must have the + * same configuration. + */ +MLXSW_ITEM32_LP(reg, spvc, 0x00, 16, 0x00, 12); + +/* reg_spvc_inner_et2 + * Vlan Tag1 EtherType2 enable. + * Packet is initially classified as double VLAN Tag if in addition to + * being classified with a tag0 VLAN Tag its tag1 EtherType value is + * equal to ether_type2. + * 0: disable (default) + * 1: enable + * Access: RW + */ +MLXSW_ITEM32(reg, spvc, inner_et2, 0x08, 17, 1); + +/* reg_spvc_et2 + * Vlan Tag0 EtherType2 enable. + * Packet is initially classified as VLAN Tag if its tag0 EtherType is + * equal to ether_type2. + * 0: disable (default) + * 1: enable + * Access: RW + */ +MLXSW_ITEM32(reg, spvc, et2, 0x08, 16, 1); + +/* reg_spvc_inner_et1 + * Vlan Tag1 EtherType1 enable. + * Packet is initially classified as double VLAN Tag if in addition to + * being classified with a tag0 VLAN Tag its tag1 EtherType value is + * equal to ether_type1. + * 0: disable + * 1: enable (default) + * Access: RW + */ +MLXSW_ITEM32(reg, spvc, inner_et1, 0x08, 9, 1); + +/* reg_spvc_et1 + * Vlan Tag0 EtherType1 enable. + * Packet is initially classified as VLAN Tag if its tag0 EtherType is + * equal to ether_type1. + * 0: disable + * 1: enable (default) + * Access: RW + */ +MLXSW_ITEM32(reg, spvc, et1, 0x08, 8, 1); + +/* reg_inner_et0 + * Vlan Tag1 EtherType0 enable. + * Packet is initially classified as double VLAN Tag if in addition to + * being classified with a tag0 VLAN Tag its tag1 EtherType value is + * equal to ether_type0. + * 0: disable + * 1: enable (default) + * Access: RW + */ +MLXSW_ITEM32(reg, spvc, inner_et0, 0x08, 1, 1); + +/* reg_et0 + * Vlan Tag0 EtherType0 enable. + * Packet is initially classified as VLAN Tag if its tag0 EtherType is + * equal to ether_type0. + * 0: disable + * 1: enable (default) + * Access: RW + */ +MLXSW_ITEM32(reg, spvc, et0, 0x08, 0, 1); + +static inline void mlxsw_reg_spvc_pack(char *payload, u16 local_port, bool et1, + bool et0) +{ + MLXSW_REG_ZERO(spvc, payload); + mlxsw_reg_spvc_local_port_set(payload, local_port); + /* Enable inner_et1 and inner_et0 to enable identification of double + * tagged packets. + */ + mlxsw_reg_spvc_inner_et1_set(payload, 1); + mlxsw_reg_spvc_inner_et0_set(payload, 1); + mlxsw_reg_spvc_et1_set(payload, et1); + mlxsw_reg_spvc_et0_set(payload, et0); +} + +/* SPEVET - Switch Port Egress VLAN EtherType + * ------------------------------------------ + * The switch port egress VLAN EtherType configures which EtherType to push at + * egress for packets incoming through a local port for which 'SPVID.egr_et_set' + * is set. + */ +#define MLXSW_REG_SPEVET_ID 0x202A +#define MLXSW_REG_SPEVET_LEN 0x08 + +MLXSW_REG_DEFINE(spevet, MLXSW_REG_SPEVET_ID, MLXSW_REG_SPEVET_LEN); + +/* reg_spevet_local_port + * Egress Local port number. + * Not supported to CPU port. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, spevet, 0x00, 16, 0x00, 12); + +/* reg_spevet_et_vlan + * Egress EtherType VLAN to push when SPVID.egr_et_set field set for the packet: + * 0: ether_type0 - (default) + * 1: ether_type1 + * 2: ether_type2 + * Access: RW + */ +MLXSW_ITEM32(reg, spevet, et_vlan, 0x04, 16, 2); + +static inline void mlxsw_reg_spevet_pack(char *payload, u16 local_port, + u8 et_vlan) +{ + MLXSW_REG_ZERO(spevet, payload); + mlxsw_reg_spevet_local_port_set(payload, local_port); + mlxsw_reg_spevet_et_vlan_set(payload, et_vlan); +} + +/* SMPE - Switch Multicast Port to Egress VID + * ------------------------------------------ + * The switch multicast port to egress VID maps + * {egress_port, SMPE index} -> {VID}. + */ +#define MLXSW_REG_SMPE_ID 0x202B +#define MLXSW_REG_SMPE_LEN 0x0C + +MLXSW_REG_DEFINE(smpe, MLXSW_REG_SMPE_ID, MLXSW_REG_SMPE_LEN); + +/* reg_smpe_local_port + * Local port number. + * CPU port is not supported. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, smpe, 0x00, 16, 0x00, 12); + +/* reg_smpe_smpe_index + * Switch multicast port to egress VID. + * Range is 0..cap_max_rmpe-1. + * Access: Index + */ +MLXSW_ITEM32(reg, smpe, smpe_index, 0x04, 0, 16); + +/* reg_smpe_evid + * Egress VID. + * Access: RW + */ +MLXSW_ITEM32(reg, smpe, evid, 0x08, 0, 12); + +static inline void mlxsw_reg_smpe_pack(char *payload, u16 local_port, + u16 smpe_index, u16 evid) +{ + MLXSW_REG_ZERO(smpe, payload); + mlxsw_reg_smpe_local_port_set(payload, local_port); + mlxsw_reg_smpe_smpe_index_set(payload, smpe_index); + mlxsw_reg_smpe_evid_set(payload, evid); +} + +/* SMID-V2 - Switch Multicast ID Version 2 Register + * ------------------------------------------------ + * The MID record maps from a MID (Multicast ID), which is a unique identifier + * of the multicast group within the stacking domain, into a list of local + * ports into which the packet is replicated. + */ +#define MLXSW_REG_SMID2_ID 0x2034 +#define MLXSW_REG_SMID2_LEN 0x120 + +MLXSW_REG_DEFINE(smid2, MLXSW_REG_SMID2_ID, MLXSW_REG_SMID2_LEN); + +/* reg_smid2_swid + * Switch partition ID. + * Access: Index + */ +MLXSW_ITEM32(reg, smid2, swid, 0x00, 24, 8); + +/* reg_smid2_mid + * Multicast identifier - global identifier that represents the multicast group + * across all devices. + * Access: Index + */ +MLXSW_ITEM32(reg, smid2, mid, 0x00, 0, 16); + +/* reg_smid2_smpe_valid + * SMPE is valid. + * When not valid, the egress VID will not be modified by the SMPE table. + * Access: RW + * + * Note: Reserved when legacy bridge model is used and on Spectrum-2. + */ +MLXSW_ITEM32(reg, smid2, smpe_valid, 0x08, 20, 1); + +/* reg_smid2_smpe + * Switch multicast port to egress VID. + * Access: RW + * + * Note: Reserved when legacy bridge model is used and on Spectrum-2. + */ +MLXSW_ITEM32(reg, smid2, smpe, 0x08, 0, 16); + +/* reg_smid2_port + * Local port memebership (1 bit per port). + * Access: RW + */ +MLXSW_ITEM_BIT_ARRAY(reg, smid2, port, 0x20, 0x80, 1); + +/* reg_smid2_port_mask + * Local port mask (1 bit per port). + * Access: WO + */ +MLXSW_ITEM_BIT_ARRAY(reg, smid2, port_mask, 0xA0, 0x80, 1); + +static inline void mlxsw_reg_smid2_pack(char *payload, u16 mid, u16 port, + bool set, bool smpe_valid, u16 smpe) +{ + MLXSW_REG_ZERO(smid2, payload); + mlxsw_reg_smid2_swid_set(payload, 0); + mlxsw_reg_smid2_mid_set(payload, mid); + mlxsw_reg_smid2_port_set(payload, port, set); + mlxsw_reg_smid2_port_mask_set(payload, port, 1); + mlxsw_reg_smid2_smpe_valid_set(payload, smpe_valid); + mlxsw_reg_smid2_smpe_set(payload, smpe_valid ? smpe : 0); +} + +/* CWTP - Congetion WRED ECN TClass Profile + * ---------------------------------------- + * Configures the profiles for queues of egress port and traffic class + */ +#define MLXSW_REG_CWTP_ID 0x2802 +#define MLXSW_REG_CWTP_BASE_LEN 0x28 +#define MLXSW_REG_CWTP_PROFILE_DATA_REC_LEN 0x08 +#define MLXSW_REG_CWTP_LEN 0x40 + +MLXSW_REG_DEFINE(cwtp, MLXSW_REG_CWTP_ID, MLXSW_REG_CWTP_LEN); + +/* reg_cwtp_local_port + * Local port number + * Not supported for CPU port + * Access: Index + */ +MLXSW_ITEM32_LP(reg, cwtp, 0x00, 16, 0x00, 12); + +/* reg_cwtp_traffic_class + * Traffic Class to configure + * Access: Index + */ +MLXSW_ITEM32(reg, cwtp, traffic_class, 32, 0, 8); + +/* reg_cwtp_profile_min + * Minimum Average Queue Size of the profile in cells. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, cwtp, profile_min, MLXSW_REG_CWTP_BASE_LEN, + 0, 20, MLXSW_REG_CWTP_PROFILE_DATA_REC_LEN, 0, false); + +/* reg_cwtp_profile_percent + * Percentage of WRED and ECN marking for maximum Average Queue size + * Range is 0 to 100, units of integer percentage + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, cwtp, profile_percent, MLXSW_REG_CWTP_BASE_LEN, + 24, 7, MLXSW_REG_CWTP_PROFILE_DATA_REC_LEN, 4, false); + +/* reg_cwtp_profile_max + * Maximum Average Queue size of the profile in cells + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, cwtp, profile_max, MLXSW_REG_CWTP_BASE_LEN, + 0, 20, MLXSW_REG_CWTP_PROFILE_DATA_REC_LEN, 4, false); + +#define MLXSW_REG_CWTP_MIN_VALUE 64 +#define MLXSW_REG_CWTP_MAX_PROFILE 2 +#define MLXSW_REG_CWTP_DEFAULT_PROFILE 1 + +static inline void mlxsw_reg_cwtp_pack(char *payload, u16 local_port, + u8 traffic_class) +{ + int i; + + MLXSW_REG_ZERO(cwtp, payload); + mlxsw_reg_cwtp_local_port_set(payload, local_port); + mlxsw_reg_cwtp_traffic_class_set(payload, traffic_class); + + for (i = 0; i <= MLXSW_REG_CWTP_MAX_PROFILE; i++) { + mlxsw_reg_cwtp_profile_min_set(payload, i, + MLXSW_REG_CWTP_MIN_VALUE); + mlxsw_reg_cwtp_profile_max_set(payload, i, + MLXSW_REG_CWTP_MIN_VALUE); + } +} + +#define MLXSW_REG_CWTP_PROFILE_TO_INDEX(profile) (profile - 1) + +static inline void +mlxsw_reg_cwtp_profile_pack(char *payload, u8 profile, u32 min, u32 max, + u32 probability) +{ + u8 index = MLXSW_REG_CWTP_PROFILE_TO_INDEX(profile); + + mlxsw_reg_cwtp_profile_min_set(payload, index, min); + mlxsw_reg_cwtp_profile_max_set(payload, index, max); + mlxsw_reg_cwtp_profile_percent_set(payload, index, probability); +} + +/* CWTPM - Congestion WRED ECN TClass and Pool Mapping + * --------------------------------------------------- + * The CWTPM register maps each egress port and traffic class to profile num. + */ +#define MLXSW_REG_CWTPM_ID 0x2803 +#define MLXSW_REG_CWTPM_LEN 0x44 + +MLXSW_REG_DEFINE(cwtpm, MLXSW_REG_CWTPM_ID, MLXSW_REG_CWTPM_LEN); + +/* reg_cwtpm_local_port + * Local port number + * Not supported for CPU port + * Access: Index + */ +MLXSW_ITEM32_LP(reg, cwtpm, 0x00, 16, 0x00, 12); + +/* reg_cwtpm_traffic_class + * Traffic Class to configure + * Access: Index + */ +MLXSW_ITEM32(reg, cwtpm, traffic_class, 32, 0, 8); + +/* reg_cwtpm_ew + * Control enablement of WRED for traffic class: + * 0 - Disable + * 1 - Enable + * Access: RW + */ +MLXSW_ITEM32(reg, cwtpm, ew, 36, 1, 1); + +/* reg_cwtpm_ee + * Control enablement of ECN for traffic class: + * 0 - Disable + * 1 - Enable + * Access: RW + */ +MLXSW_ITEM32(reg, cwtpm, ee, 36, 0, 1); + +/* reg_cwtpm_tcp_g + * TCP Green Profile. + * Index of the profile within {port, traffic class} to use. + * 0 for disabling both WRED and ECN for this type of traffic. + * Access: RW + */ +MLXSW_ITEM32(reg, cwtpm, tcp_g, 52, 0, 2); + +/* reg_cwtpm_tcp_y + * TCP Yellow Profile. + * Index of the profile within {port, traffic class} to use. + * 0 for disabling both WRED and ECN for this type of traffic. + * Access: RW + */ +MLXSW_ITEM32(reg, cwtpm, tcp_y, 56, 16, 2); + +/* reg_cwtpm_tcp_r + * TCP Red Profile. + * Index of the profile within {port, traffic class} to use. + * 0 for disabling both WRED and ECN for this type of traffic. + * Access: RW + */ +MLXSW_ITEM32(reg, cwtpm, tcp_r, 56, 0, 2); + +/* reg_cwtpm_ntcp_g + * Non-TCP Green Profile. + * Index of the profile within {port, traffic class} to use. + * 0 for disabling both WRED and ECN for this type of traffic. + * Access: RW + */ +MLXSW_ITEM32(reg, cwtpm, ntcp_g, 60, 0, 2); + +/* reg_cwtpm_ntcp_y + * Non-TCP Yellow Profile. + * Index of the profile within {port, traffic class} to use. + * 0 for disabling both WRED and ECN for this type of traffic. + * Access: RW + */ +MLXSW_ITEM32(reg, cwtpm, ntcp_y, 64, 16, 2); + +/* reg_cwtpm_ntcp_r + * Non-TCP Red Profile. + * Index of the profile within {port, traffic class} to use. + * 0 for disabling both WRED and ECN for this type of traffic. + * Access: RW + */ +MLXSW_ITEM32(reg, cwtpm, ntcp_r, 64, 0, 2); + +#define MLXSW_REG_CWTPM_RESET_PROFILE 0 + +static inline void mlxsw_reg_cwtpm_pack(char *payload, u16 local_port, + u8 traffic_class, u8 profile, + bool wred, bool ecn) +{ + MLXSW_REG_ZERO(cwtpm, payload); + mlxsw_reg_cwtpm_local_port_set(payload, local_port); + mlxsw_reg_cwtpm_traffic_class_set(payload, traffic_class); + mlxsw_reg_cwtpm_ew_set(payload, wred); + mlxsw_reg_cwtpm_ee_set(payload, ecn); + mlxsw_reg_cwtpm_tcp_g_set(payload, profile); + mlxsw_reg_cwtpm_tcp_y_set(payload, profile); + mlxsw_reg_cwtpm_tcp_r_set(payload, profile); + mlxsw_reg_cwtpm_ntcp_g_set(payload, profile); + mlxsw_reg_cwtpm_ntcp_y_set(payload, profile); + mlxsw_reg_cwtpm_ntcp_r_set(payload, profile); +} + +/* PGCR - Policy-Engine General Configuration Register + * --------------------------------------------------- + * This register configures general Policy-Engine settings. + */ +#define MLXSW_REG_PGCR_ID 0x3001 +#define MLXSW_REG_PGCR_LEN 0x20 + +MLXSW_REG_DEFINE(pgcr, MLXSW_REG_PGCR_ID, MLXSW_REG_PGCR_LEN); + +/* reg_pgcr_default_action_pointer_base + * Default action pointer base. Each region has a default action pointer + * which is equal to default_action_pointer_base + region_id. + * Access: RW + */ +MLXSW_ITEM32(reg, pgcr, default_action_pointer_base, 0x1C, 0, 24); + +static inline void mlxsw_reg_pgcr_pack(char *payload, u32 pointer_base) +{ + MLXSW_REG_ZERO(pgcr, payload); + mlxsw_reg_pgcr_default_action_pointer_base_set(payload, pointer_base); +} + +/* PPBT - Policy-Engine Port Binding Table + * --------------------------------------- + * This register is used for configuration of the Port Binding Table. + */ +#define MLXSW_REG_PPBT_ID 0x3002 +#define MLXSW_REG_PPBT_LEN 0x14 + +MLXSW_REG_DEFINE(ppbt, MLXSW_REG_PPBT_ID, MLXSW_REG_PPBT_LEN); + +enum mlxsw_reg_pxbt_e { + MLXSW_REG_PXBT_E_IACL, + MLXSW_REG_PXBT_E_EACL, +}; + +/* reg_ppbt_e + * Access: Index + */ +MLXSW_ITEM32(reg, ppbt, e, 0x00, 31, 1); + +enum mlxsw_reg_pxbt_op { + MLXSW_REG_PXBT_OP_BIND, + MLXSW_REG_PXBT_OP_UNBIND, +}; + +/* reg_ppbt_op + * Access: RW + */ +MLXSW_ITEM32(reg, ppbt, op, 0x00, 28, 3); + +/* reg_ppbt_local_port + * Local port. Not including CPU port. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, ppbt, 0x00, 16, 0x00, 12); + +/* reg_ppbt_g + * group - When set, the binding is of an ACL group. When cleared, + * the binding is of an ACL. + * Must be set to 1 for Spectrum. + * Access: RW + */ +MLXSW_ITEM32(reg, ppbt, g, 0x10, 31, 1); + +/* reg_ppbt_acl_info + * ACL/ACL group identifier. If the g bit is set, this field should hold + * the acl_group_id, else it should hold the acl_id. + * Access: RW + */ +MLXSW_ITEM32(reg, ppbt, acl_info, 0x10, 0, 16); + +static inline void mlxsw_reg_ppbt_pack(char *payload, enum mlxsw_reg_pxbt_e e, + enum mlxsw_reg_pxbt_op op, + u16 local_port, u16 acl_info) +{ + MLXSW_REG_ZERO(ppbt, payload); + mlxsw_reg_ppbt_e_set(payload, e); + mlxsw_reg_ppbt_op_set(payload, op); + mlxsw_reg_ppbt_local_port_set(payload, local_port); + mlxsw_reg_ppbt_g_set(payload, true); + mlxsw_reg_ppbt_acl_info_set(payload, acl_info); +} + +/* PACL - Policy-Engine ACL Register + * --------------------------------- + * This register is used for configuration of the ACL. + */ +#define MLXSW_REG_PACL_ID 0x3004 +#define MLXSW_REG_PACL_LEN 0x70 + +MLXSW_REG_DEFINE(pacl, MLXSW_REG_PACL_ID, MLXSW_REG_PACL_LEN); + +/* reg_pacl_v + * Valid. Setting the v bit makes the ACL valid. It should not be cleared + * while the ACL is bounded to either a port, VLAN or ACL rule. + * Access: RW + */ +MLXSW_ITEM32(reg, pacl, v, 0x00, 24, 1); + +/* reg_pacl_acl_id + * An identifier representing the ACL (managed by software) + * Range 0 .. cap_max_acl_regions - 1 + * Access: Index + */ +MLXSW_ITEM32(reg, pacl, acl_id, 0x08, 0, 16); + +#define MLXSW_REG_PXXX_TCAM_REGION_INFO_LEN 16 + +/* reg_pacl_tcam_region_info + * Opaque object that represents a TCAM region. + * Obtained through PTAR register. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, pacl, tcam_region_info, 0x30, + MLXSW_REG_PXXX_TCAM_REGION_INFO_LEN); + +static inline void mlxsw_reg_pacl_pack(char *payload, u16 acl_id, + bool valid, const char *tcam_region_info) +{ + MLXSW_REG_ZERO(pacl, payload); + mlxsw_reg_pacl_acl_id_set(payload, acl_id); + mlxsw_reg_pacl_v_set(payload, valid); + mlxsw_reg_pacl_tcam_region_info_memcpy_to(payload, tcam_region_info); +} + +/* PAGT - Policy-Engine ACL Group Table + * ------------------------------------ + * This register is used for configuration of the ACL Group Table. + */ +#define MLXSW_REG_PAGT_ID 0x3005 +#define MLXSW_REG_PAGT_BASE_LEN 0x30 +#define MLXSW_REG_PAGT_ACL_LEN 4 +#define MLXSW_REG_PAGT_ACL_MAX_NUM 16 +#define MLXSW_REG_PAGT_LEN (MLXSW_REG_PAGT_BASE_LEN + \ + MLXSW_REG_PAGT_ACL_MAX_NUM * MLXSW_REG_PAGT_ACL_LEN) + +MLXSW_REG_DEFINE(pagt, MLXSW_REG_PAGT_ID, MLXSW_REG_PAGT_LEN); + +/* reg_pagt_size + * Number of ACLs in the group. + * Size 0 invalidates a group. + * Range 0 .. cap_max_acl_group_size (hard coded to 16 for now) + * Total number of ACLs in all groups must be lower or equal + * to cap_max_acl_tot_groups + * Note: a group which is binded must not be invalidated + * Access: Index + */ +MLXSW_ITEM32(reg, pagt, size, 0x00, 0, 8); + +/* reg_pagt_acl_group_id + * An identifier (numbered from 0..cap_max_acl_groups-1) representing + * the ACL Group identifier (managed by software). + * Access: Index + */ +MLXSW_ITEM32(reg, pagt, acl_group_id, 0x08, 0, 16); + +/* reg_pagt_multi + * Multi-ACL + * 0 - This ACL is the last ACL in the multi-ACL + * 1 - This ACL is part of a multi-ACL + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, pagt, multi, 0x30, 31, 1, 0x04, 0x00, false); + +/* reg_pagt_acl_id + * ACL identifier + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, pagt, acl_id, 0x30, 0, 16, 0x04, 0x00, false); + +static inline void mlxsw_reg_pagt_pack(char *payload, u16 acl_group_id) +{ + MLXSW_REG_ZERO(pagt, payload); + mlxsw_reg_pagt_acl_group_id_set(payload, acl_group_id); +} + +static inline void mlxsw_reg_pagt_acl_id_pack(char *payload, int index, + u16 acl_id, bool multi) +{ + u8 size = mlxsw_reg_pagt_size_get(payload); + + if (index >= size) + mlxsw_reg_pagt_size_set(payload, index + 1); + mlxsw_reg_pagt_multi_set(payload, index, multi); + mlxsw_reg_pagt_acl_id_set(payload, index, acl_id); +} + +/* PTAR - Policy-Engine TCAM Allocation Register + * --------------------------------------------- + * This register is used for allocation of regions in the TCAM. + * Note: Query method is not supported on this register. + */ +#define MLXSW_REG_PTAR_ID 0x3006 +#define MLXSW_REG_PTAR_BASE_LEN 0x20 +#define MLXSW_REG_PTAR_KEY_ID_LEN 1 +#define MLXSW_REG_PTAR_KEY_ID_MAX_NUM 16 +#define MLXSW_REG_PTAR_LEN (MLXSW_REG_PTAR_BASE_LEN + \ + MLXSW_REG_PTAR_KEY_ID_MAX_NUM * MLXSW_REG_PTAR_KEY_ID_LEN) + +MLXSW_REG_DEFINE(ptar, MLXSW_REG_PTAR_ID, MLXSW_REG_PTAR_LEN); + +enum mlxsw_reg_ptar_op { + /* allocate a TCAM region */ + MLXSW_REG_PTAR_OP_ALLOC, + /* resize a TCAM region */ + MLXSW_REG_PTAR_OP_RESIZE, + /* deallocate TCAM region */ + MLXSW_REG_PTAR_OP_FREE, + /* test allocation */ + MLXSW_REG_PTAR_OP_TEST, +}; + +/* reg_ptar_op + * Access: OP + */ +MLXSW_ITEM32(reg, ptar, op, 0x00, 28, 4); + +/* reg_ptar_action_set_type + * Type of action set to be used on this region. + * For Spectrum and Spectrum-2, this is always type 2 - "flexible" + * Access: WO + */ +MLXSW_ITEM32(reg, ptar, action_set_type, 0x00, 16, 8); + +enum mlxsw_reg_ptar_key_type { + MLXSW_REG_PTAR_KEY_TYPE_FLEX = 0x50, /* Spetrum */ + MLXSW_REG_PTAR_KEY_TYPE_FLEX2 = 0x51, /* Spectrum-2 */ +}; + +/* reg_ptar_key_type + * TCAM key type for the region. + * Access: WO + */ +MLXSW_ITEM32(reg, ptar, key_type, 0x00, 0, 8); + +/* reg_ptar_region_size + * TCAM region size. When allocating/resizing this is the requested size, + * the response is the actual size. Note that actual size may be + * larger than requested. + * Allowed range 1 .. cap_max_rules-1 + * Reserved during op deallocate. + * Access: WO + */ +MLXSW_ITEM32(reg, ptar, region_size, 0x04, 0, 16); + +/* reg_ptar_region_id + * Region identifier + * Range 0 .. cap_max_regions-1 + * Access: Index + */ +MLXSW_ITEM32(reg, ptar, region_id, 0x08, 0, 16); + +/* reg_ptar_tcam_region_info + * Opaque object that represents the TCAM region. + * Returned when allocating a region. + * Provided by software for ACL generation and region deallocation and resize. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, ptar, tcam_region_info, 0x10, + MLXSW_REG_PXXX_TCAM_REGION_INFO_LEN); + +/* reg_ptar_flexible_key_id + * Identifier of the Flexible Key. + * Only valid if key_type == "FLEX_KEY" + * The key size will be rounded up to one of the following values: + * 9B, 18B, 36B, 54B. + * This field is reserved for in resize operation. + * Access: WO + */ +MLXSW_ITEM8_INDEXED(reg, ptar, flexible_key_id, 0x20, 0, 8, + MLXSW_REG_PTAR_KEY_ID_LEN, 0x00, false); + +static inline void mlxsw_reg_ptar_pack(char *payload, enum mlxsw_reg_ptar_op op, + enum mlxsw_reg_ptar_key_type key_type, + u16 region_size, u16 region_id, + const char *tcam_region_info) +{ + MLXSW_REG_ZERO(ptar, payload); + mlxsw_reg_ptar_op_set(payload, op); + mlxsw_reg_ptar_action_set_type_set(payload, 2); /* "flexible" */ + mlxsw_reg_ptar_key_type_set(payload, key_type); + mlxsw_reg_ptar_region_size_set(payload, region_size); + mlxsw_reg_ptar_region_id_set(payload, region_id); + mlxsw_reg_ptar_tcam_region_info_memcpy_to(payload, tcam_region_info); +} + +static inline void mlxsw_reg_ptar_key_id_pack(char *payload, int index, + u16 key_id) +{ + mlxsw_reg_ptar_flexible_key_id_set(payload, index, key_id); +} + +static inline void mlxsw_reg_ptar_unpack(char *payload, char *tcam_region_info) +{ + mlxsw_reg_ptar_tcam_region_info_memcpy_from(payload, tcam_region_info); +} + +/* PPBS - Policy-Engine Policy Based Switching Register + * ---------------------------------------------------- + * This register retrieves and sets Policy Based Switching Table entries. + */ +#define MLXSW_REG_PPBS_ID 0x300C +#define MLXSW_REG_PPBS_LEN 0x14 + +MLXSW_REG_DEFINE(ppbs, MLXSW_REG_PPBS_ID, MLXSW_REG_PPBS_LEN); + +/* reg_ppbs_pbs_ptr + * Index into the PBS table. + * For Spectrum, the index points to the KVD Linear. + * Access: Index + */ +MLXSW_ITEM32(reg, ppbs, pbs_ptr, 0x08, 0, 24); + +/* reg_ppbs_system_port + * Unique port identifier for the final destination of the packet. + * Access: RW + */ +MLXSW_ITEM32(reg, ppbs, system_port, 0x10, 0, 16); + +static inline void mlxsw_reg_ppbs_pack(char *payload, u32 pbs_ptr, + u16 system_port) +{ + MLXSW_REG_ZERO(ppbs, payload); + mlxsw_reg_ppbs_pbs_ptr_set(payload, pbs_ptr); + mlxsw_reg_ppbs_system_port_set(payload, system_port); +} + +/* PRCR - Policy-Engine Rules Copy Register + * ---------------------------------------- + * This register is used for accessing rules within a TCAM region. + */ +#define MLXSW_REG_PRCR_ID 0x300D +#define MLXSW_REG_PRCR_LEN 0x40 + +MLXSW_REG_DEFINE(prcr, MLXSW_REG_PRCR_ID, MLXSW_REG_PRCR_LEN); + +enum mlxsw_reg_prcr_op { + /* Move rules. Moves the rules from "tcam_region_info" starting + * at offset "offset" to "dest_tcam_region_info" + * at offset "dest_offset." + */ + MLXSW_REG_PRCR_OP_MOVE, + /* Copy rules. Copies the rules from "tcam_region_info" starting + * at offset "offset" to "dest_tcam_region_info" + * at offset "dest_offset." + */ + MLXSW_REG_PRCR_OP_COPY, +}; + +/* reg_prcr_op + * Access: OP + */ +MLXSW_ITEM32(reg, prcr, op, 0x00, 28, 4); + +/* reg_prcr_offset + * Offset within the source region to copy/move from. + * Access: Index + */ +MLXSW_ITEM32(reg, prcr, offset, 0x00, 0, 16); + +/* reg_prcr_size + * The number of rules to copy/move. + * Access: WO + */ +MLXSW_ITEM32(reg, prcr, size, 0x04, 0, 16); + +/* reg_prcr_tcam_region_info + * Opaque object that represents the source TCAM region. + * Access: Index + */ +MLXSW_ITEM_BUF(reg, prcr, tcam_region_info, 0x10, + MLXSW_REG_PXXX_TCAM_REGION_INFO_LEN); + +/* reg_prcr_dest_offset + * Offset within the source region to copy/move to. + * Access: Index + */ +MLXSW_ITEM32(reg, prcr, dest_offset, 0x20, 0, 16); + +/* reg_prcr_dest_tcam_region_info + * Opaque object that represents the destination TCAM region. + * Access: Index + */ +MLXSW_ITEM_BUF(reg, prcr, dest_tcam_region_info, 0x30, + MLXSW_REG_PXXX_TCAM_REGION_INFO_LEN); + +static inline void mlxsw_reg_prcr_pack(char *payload, enum mlxsw_reg_prcr_op op, + const char *src_tcam_region_info, + u16 src_offset, + const char *dest_tcam_region_info, + u16 dest_offset, u16 size) +{ + MLXSW_REG_ZERO(prcr, payload); + mlxsw_reg_prcr_op_set(payload, op); + mlxsw_reg_prcr_offset_set(payload, src_offset); + mlxsw_reg_prcr_size_set(payload, size); + mlxsw_reg_prcr_tcam_region_info_memcpy_to(payload, + src_tcam_region_info); + mlxsw_reg_prcr_dest_offset_set(payload, dest_offset); + mlxsw_reg_prcr_dest_tcam_region_info_memcpy_to(payload, + dest_tcam_region_info); +} + +/* PEFA - Policy-Engine Extended Flexible Action Register + * ------------------------------------------------------ + * This register is used for accessing an extended flexible action entry + * in the central KVD Linear Database. + */ +#define MLXSW_REG_PEFA_ID 0x300F +#define MLXSW_REG_PEFA_LEN 0xB0 + +MLXSW_REG_DEFINE(pefa, MLXSW_REG_PEFA_ID, MLXSW_REG_PEFA_LEN); + +/* reg_pefa_index + * Index in the KVD Linear Centralized Database. + * Access: Index + */ +MLXSW_ITEM32(reg, pefa, index, 0x00, 0, 24); + +/* reg_pefa_a + * Index in the KVD Linear Centralized Database. + * Activity + * For a new entry: set if ca=0, clear if ca=1 + * Set if a packet lookup has hit on the specific entry + * Access: RO + */ +MLXSW_ITEM32(reg, pefa, a, 0x04, 29, 1); + +/* reg_pefa_ca + * Clear activity + * When write: activity is according to this field + * When read: after reading the activity is cleared according to ca + * Access: OP + */ +MLXSW_ITEM32(reg, pefa, ca, 0x04, 24, 1); + +#define MLXSW_REG_FLEX_ACTION_SET_LEN 0xA8 + +/* reg_pefa_flex_action_set + * Action-set to perform when rule is matched. + * Must be zero padded if action set is shorter. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, pefa, flex_action_set, 0x08, MLXSW_REG_FLEX_ACTION_SET_LEN); + +static inline void mlxsw_reg_pefa_pack(char *payload, u32 index, bool ca, + const char *flex_action_set) +{ + MLXSW_REG_ZERO(pefa, payload); + mlxsw_reg_pefa_index_set(payload, index); + mlxsw_reg_pefa_ca_set(payload, ca); + if (flex_action_set) + mlxsw_reg_pefa_flex_action_set_memcpy_to(payload, + flex_action_set); +} + +static inline void mlxsw_reg_pefa_unpack(char *payload, bool *p_a) +{ + *p_a = mlxsw_reg_pefa_a_get(payload); +} + +/* PEMRBT - Policy-Engine Multicast Router Binding Table Register + * -------------------------------------------------------------- + * This register is used for binding Multicast router to an ACL group + * that serves the MC router. + * This register is not supported by SwitchX/-2 and Spectrum. + */ +#define MLXSW_REG_PEMRBT_ID 0x3014 +#define MLXSW_REG_PEMRBT_LEN 0x14 + +MLXSW_REG_DEFINE(pemrbt, MLXSW_REG_PEMRBT_ID, MLXSW_REG_PEMRBT_LEN); + +enum mlxsw_reg_pemrbt_protocol { + MLXSW_REG_PEMRBT_PROTO_IPV4, + MLXSW_REG_PEMRBT_PROTO_IPV6, +}; + +/* reg_pemrbt_protocol + * Access: Index + */ +MLXSW_ITEM32(reg, pemrbt, protocol, 0x00, 0, 1); + +/* reg_pemrbt_group_id + * ACL group identifier. + * Range 0..cap_max_acl_groups-1 + * Access: RW + */ +MLXSW_ITEM32(reg, pemrbt, group_id, 0x10, 0, 16); + +static inline void +mlxsw_reg_pemrbt_pack(char *payload, enum mlxsw_reg_pemrbt_protocol protocol, + u16 group_id) +{ + MLXSW_REG_ZERO(pemrbt, payload); + mlxsw_reg_pemrbt_protocol_set(payload, protocol); + mlxsw_reg_pemrbt_group_id_set(payload, group_id); +} + +/* PTCE-V2 - Policy-Engine TCAM Entry Register Version 2 + * ----------------------------------------------------- + * This register is used for accessing rules within a TCAM region. + * It is a new version of PTCE in order to support wider key, + * mask and action within a TCAM region. This register is not supported + * by SwitchX and SwitchX-2. + */ +#define MLXSW_REG_PTCE2_ID 0x3017 +#define MLXSW_REG_PTCE2_LEN 0x1D8 + +MLXSW_REG_DEFINE(ptce2, MLXSW_REG_PTCE2_ID, MLXSW_REG_PTCE2_LEN); + +/* reg_ptce2_v + * Valid. + * Access: RW + */ +MLXSW_ITEM32(reg, ptce2, v, 0x00, 31, 1); + +/* reg_ptce2_a + * Activity. Set if a packet lookup has hit on the specific entry. + * To clear the "a" bit, use "clear activity" op or "clear on read" op. + * Access: RO + */ +MLXSW_ITEM32(reg, ptce2, a, 0x00, 30, 1); + +enum mlxsw_reg_ptce2_op { + /* Read operation. */ + MLXSW_REG_PTCE2_OP_QUERY_READ = 0, + /* clear on read operation. Used to read entry + * and clear Activity bit. + */ + MLXSW_REG_PTCE2_OP_QUERY_CLEAR_ON_READ = 1, + /* Write operation. Used to write a new entry to the table. + * All R/W fields are relevant for new entry. Activity bit is set + * for new entries - Note write with v = 0 will delete the entry. + */ + MLXSW_REG_PTCE2_OP_WRITE_WRITE = 0, + /* Update action. Only action set will be updated. */ + MLXSW_REG_PTCE2_OP_WRITE_UPDATE = 1, + /* Clear activity. A bit is cleared for the entry. */ + MLXSW_REG_PTCE2_OP_WRITE_CLEAR_ACTIVITY = 2, +}; + +/* reg_ptce2_op + * Access: OP + */ +MLXSW_ITEM32(reg, ptce2, op, 0x00, 20, 3); + +/* reg_ptce2_offset + * Access: Index + */ +MLXSW_ITEM32(reg, ptce2, offset, 0x00, 0, 16); + +/* reg_ptce2_priority + * Priority of the rule, higher values win. The range is 1..cap_kvd_size-1. + * Note: priority does not have to be unique per rule. + * Within a region, higher priority should have lower offset (no limitation + * between regions in a multi-region). + * Access: RW + */ +MLXSW_ITEM32(reg, ptce2, priority, 0x04, 0, 24); + +/* reg_ptce2_tcam_region_info + * Opaque object that represents the TCAM region. + * Access: Index + */ +MLXSW_ITEM_BUF(reg, ptce2, tcam_region_info, 0x10, + MLXSW_REG_PXXX_TCAM_REGION_INFO_LEN); + +#define MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN 96 + +/* reg_ptce2_flex_key_blocks + * ACL Key. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, ptce2, flex_key_blocks, 0x20, + MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN); + +/* reg_ptce2_mask + * mask- in the same size as key. A bit that is set directs the TCAM + * to compare the corresponding bit in key. A bit that is clear directs + * the TCAM to ignore the corresponding bit in key. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, ptce2, mask, 0x80, + MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN); + +/* reg_ptce2_flex_action_set + * ACL action set. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, ptce2, flex_action_set, 0xE0, + MLXSW_REG_FLEX_ACTION_SET_LEN); + +static inline void mlxsw_reg_ptce2_pack(char *payload, bool valid, + enum mlxsw_reg_ptce2_op op, + const char *tcam_region_info, + u16 offset, u32 priority) +{ + MLXSW_REG_ZERO(ptce2, payload); + mlxsw_reg_ptce2_v_set(payload, valid); + mlxsw_reg_ptce2_op_set(payload, op); + mlxsw_reg_ptce2_offset_set(payload, offset); + mlxsw_reg_ptce2_priority_set(payload, priority); + mlxsw_reg_ptce2_tcam_region_info_memcpy_to(payload, tcam_region_info); +} + +/* PERPT - Policy-Engine ERP Table Register + * ---------------------------------------- + * This register adds and removes eRPs from the eRP table. + */ +#define MLXSW_REG_PERPT_ID 0x3021 +#define MLXSW_REG_PERPT_LEN 0x80 + +MLXSW_REG_DEFINE(perpt, MLXSW_REG_PERPT_ID, MLXSW_REG_PERPT_LEN); + +/* reg_perpt_erpt_bank + * eRP table bank. + * Range 0 .. cap_max_erp_table_banks - 1 + * Access: Index + */ +MLXSW_ITEM32(reg, perpt, erpt_bank, 0x00, 16, 4); + +/* reg_perpt_erpt_index + * Index to eRP table within the eRP bank. + * Range is 0 .. cap_max_erp_table_bank_size - 1 + * Access: Index + */ +MLXSW_ITEM32(reg, perpt, erpt_index, 0x00, 0, 8); + +enum mlxsw_reg_perpt_key_size { + MLXSW_REG_PERPT_KEY_SIZE_2KB, + MLXSW_REG_PERPT_KEY_SIZE_4KB, + MLXSW_REG_PERPT_KEY_SIZE_8KB, + MLXSW_REG_PERPT_KEY_SIZE_12KB, +}; + +/* reg_perpt_key_size + * Access: OP + */ +MLXSW_ITEM32(reg, perpt, key_size, 0x04, 0, 4); + +/* reg_perpt_bf_bypass + * 0 - The eRP is used only if bloom filter state is set for the given + * rule. + * 1 - The eRP is used regardless of bloom filter state. + * The bypass is an OR condition of region_id or eRP. See PERCR.bf_bypass + * Access: RW + */ +MLXSW_ITEM32(reg, perpt, bf_bypass, 0x08, 8, 1); + +/* reg_perpt_erp_id + * eRP ID for use by the rules. + * Access: RW + */ +MLXSW_ITEM32(reg, perpt, erp_id, 0x08, 0, 4); + +/* reg_perpt_erpt_base_bank + * Base eRP table bank, points to head of erp_vector + * Range is 0 .. cap_max_erp_table_banks - 1 + * Access: OP + */ +MLXSW_ITEM32(reg, perpt, erpt_base_bank, 0x0C, 16, 4); + +/* reg_perpt_erpt_base_index + * Base index to eRP table within the eRP bank + * Range is 0 .. cap_max_erp_table_bank_size - 1 + * Access: OP + */ +MLXSW_ITEM32(reg, perpt, erpt_base_index, 0x0C, 0, 8); + +/* reg_perpt_erp_index_in_vector + * eRP index in the vector. + * Access: OP + */ +MLXSW_ITEM32(reg, perpt, erp_index_in_vector, 0x10, 0, 4); + +/* reg_perpt_erp_vector + * eRP vector. + * Access: OP + */ +MLXSW_ITEM_BIT_ARRAY(reg, perpt, erp_vector, 0x14, 4, 1); + +/* reg_perpt_mask + * Mask + * 0 - A-TCAM will ignore the bit in key + * 1 - A-TCAM will compare the bit in key + * Access: RW + */ +MLXSW_ITEM_BUF(reg, perpt, mask, 0x20, MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN); + +static inline void mlxsw_reg_perpt_erp_vector_pack(char *payload, + unsigned long *erp_vector, + unsigned long size) +{ + unsigned long bit; + + for_each_set_bit(bit, erp_vector, size) + mlxsw_reg_perpt_erp_vector_set(payload, bit, true); +} + +static inline void +mlxsw_reg_perpt_pack(char *payload, u8 erpt_bank, u8 erpt_index, + enum mlxsw_reg_perpt_key_size key_size, u8 erp_id, + u8 erpt_base_bank, u8 erpt_base_index, u8 erp_index, + char *mask) +{ + MLXSW_REG_ZERO(perpt, payload); + mlxsw_reg_perpt_erpt_bank_set(payload, erpt_bank); + mlxsw_reg_perpt_erpt_index_set(payload, erpt_index); + mlxsw_reg_perpt_key_size_set(payload, key_size); + mlxsw_reg_perpt_bf_bypass_set(payload, false); + mlxsw_reg_perpt_erp_id_set(payload, erp_id); + mlxsw_reg_perpt_erpt_base_bank_set(payload, erpt_base_bank); + mlxsw_reg_perpt_erpt_base_index_set(payload, erpt_base_index); + mlxsw_reg_perpt_erp_index_in_vector_set(payload, erp_index); + mlxsw_reg_perpt_mask_memcpy_to(payload, mask); +} + +/* PERAR - Policy-Engine Region Association Register + * ------------------------------------------------- + * This register associates a hw region for region_id's. Changing on the fly + * is supported by the device. + */ +#define MLXSW_REG_PERAR_ID 0x3026 +#define MLXSW_REG_PERAR_LEN 0x08 + +MLXSW_REG_DEFINE(perar, MLXSW_REG_PERAR_ID, MLXSW_REG_PERAR_LEN); + +/* reg_perar_region_id + * Region identifier + * Range 0 .. cap_max_regions-1 + * Access: Index + */ +MLXSW_ITEM32(reg, perar, region_id, 0x00, 0, 16); + +static inline unsigned int +mlxsw_reg_perar_hw_regions_needed(unsigned int block_num) +{ + return DIV_ROUND_UP(block_num, 4); +} + +/* reg_perar_hw_region + * HW Region + * Range 0 .. cap_max_regions-1 + * Default: hw_region = region_id + * For a 8 key block region, 2 consecutive regions are used + * For a 12 key block region, 3 consecutive regions are used + * Access: RW + */ +MLXSW_ITEM32(reg, perar, hw_region, 0x04, 0, 16); + +static inline void mlxsw_reg_perar_pack(char *payload, u16 region_id, + u16 hw_region) +{ + MLXSW_REG_ZERO(perar, payload); + mlxsw_reg_perar_region_id_set(payload, region_id); + mlxsw_reg_perar_hw_region_set(payload, hw_region); +} + +/* PTCE-V3 - Policy-Engine TCAM Entry Register Version 3 + * ----------------------------------------------------- + * This register is a new version of PTCE-V2 in order to support the + * A-TCAM. This register is not supported by SwitchX/-2 and Spectrum. + */ +#define MLXSW_REG_PTCE3_ID 0x3027 +#define MLXSW_REG_PTCE3_LEN 0xF0 + +MLXSW_REG_DEFINE(ptce3, MLXSW_REG_PTCE3_ID, MLXSW_REG_PTCE3_LEN); + +/* reg_ptce3_v + * Valid. + * Access: RW + */ +MLXSW_ITEM32(reg, ptce3, v, 0x00, 31, 1); + +enum mlxsw_reg_ptce3_op { + /* Write operation. Used to write a new entry to the table. + * All R/W fields are relevant for new entry. Activity bit is set + * for new entries. Write with v = 0 will delete the entry. Must + * not be used if an entry exists. + */ + MLXSW_REG_PTCE3_OP_WRITE_WRITE = 0, + /* Update operation */ + MLXSW_REG_PTCE3_OP_WRITE_UPDATE = 1, + /* Read operation */ + MLXSW_REG_PTCE3_OP_QUERY_READ = 0, +}; + +/* reg_ptce3_op + * Access: OP + */ +MLXSW_ITEM32(reg, ptce3, op, 0x00, 20, 3); + +/* reg_ptce3_priority + * Priority of the rule. Higher values win. + * For Spectrum-2 range is 1..cap_kvd_size - 1 + * Note: Priority does not have to be unique per rule. + * Access: RW + */ +MLXSW_ITEM32(reg, ptce3, priority, 0x04, 0, 24); + +/* reg_ptce3_tcam_region_info + * Opaque object that represents the TCAM region. + * Access: Index + */ +MLXSW_ITEM_BUF(reg, ptce3, tcam_region_info, 0x10, + MLXSW_REG_PXXX_TCAM_REGION_INFO_LEN); + +/* reg_ptce3_flex2_key_blocks + * ACL key. The key must be masked according to eRP (if exists) or + * according to master mask. + * Access: Index + */ +MLXSW_ITEM_BUF(reg, ptce3, flex2_key_blocks, 0x20, + MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN); + +/* reg_ptce3_erp_id + * eRP ID. + * Access: Index + */ +MLXSW_ITEM32(reg, ptce3, erp_id, 0x80, 0, 4); + +/* reg_ptce3_delta_start + * Start point of delta_value and delta_mask, in bits. Must not exceed + * num_key_blocks * 36 - 8. Reserved when delta_mask = 0. + * Access: Index + */ +MLXSW_ITEM32(reg, ptce3, delta_start, 0x84, 0, 10); + +/* reg_ptce3_delta_mask + * Delta mask. + * 0 - Ignore relevant bit in delta_value + * 1 - Compare relevant bit in delta_value + * Delta mask must not be set for reserved fields in the key blocks. + * Note: No delta when no eRPs. Thus, for regions with + * PERERP.erpt_pointer_valid = 0 the delta mask must be 0. + * Access: Index + */ +MLXSW_ITEM32(reg, ptce3, delta_mask, 0x88, 16, 8); + +/* reg_ptce3_delta_value + * Delta value. + * Bits which are masked by delta_mask must be 0. + * Access: Index + */ +MLXSW_ITEM32(reg, ptce3, delta_value, 0x88, 0, 8); + +/* reg_ptce3_prune_vector + * Pruning vector relative to the PERPT.erp_id. + * Used for reducing lookups. + * 0 - NEED: Do a lookup using the eRP. + * 1 - PRUNE: Do not perform a lookup using the eRP. + * Maybe be modified by PEAPBL and PEAPBM. + * Note: In Spectrum-2, a region of 8 key blocks must be set to either + * all 1's or all 0's. + * Access: RW + */ +MLXSW_ITEM_BIT_ARRAY(reg, ptce3, prune_vector, 0x90, 4, 1); + +/* reg_ptce3_prune_ctcam + * Pruning on C-TCAM. Used for reducing lookups. + * 0 - NEED: Do a lookup in the C-TCAM. + * 1 - PRUNE: Do not perform a lookup in the C-TCAM. + * Access: RW + */ +MLXSW_ITEM32(reg, ptce3, prune_ctcam, 0x94, 31, 1); + +/* reg_ptce3_large_exists + * Large entry key ID exists. + * Within the region: + * 0 - SINGLE: The large_entry_key_id is not currently in use. + * For rule insert: The MSB of the key (blocks 6..11) will be added. + * For rule delete: The MSB of the key will be removed. + * 1 - NON_SINGLE: The large_entry_key_id is currently in use. + * For rule insert: The MSB of the key (blocks 6..11) will not be added. + * For rule delete: The MSB of the key will not be removed. + * Access: WO + */ +MLXSW_ITEM32(reg, ptce3, large_exists, 0x98, 31, 1); + +/* reg_ptce3_large_entry_key_id + * Large entry key ID. + * A key for 12 key blocks rules. Reserved when region has less than 12 key + * blocks. Must be different for different keys which have the same common + * 6 key blocks (MSB, blocks 6..11) key within a region. + * Range is 0..cap_max_pe_large_key_id - 1 + * Access: RW + */ +MLXSW_ITEM32(reg, ptce3, large_entry_key_id, 0x98, 0, 24); + +/* reg_ptce3_action_pointer + * Pointer to action. + * Range is 0..cap_max_kvd_action_sets - 1 + * Access: RW + */ +MLXSW_ITEM32(reg, ptce3, action_pointer, 0xA0, 0, 24); + +static inline void mlxsw_reg_ptce3_pack(char *payload, bool valid, + enum mlxsw_reg_ptce3_op op, + u32 priority, + const char *tcam_region_info, + const char *key, u8 erp_id, + u16 delta_start, u8 delta_mask, + u8 delta_value, bool large_exists, + u32 lkey_id, u32 action_pointer) +{ + MLXSW_REG_ZERO(ptce3, payload); + mlxsw_reg_ptce3_v_set(payload, valid); + mlxsw_reg_ptce3_op_set(payload, op); + mlxsw_reg_ptce3_priority_set(payload, priority); + mlxsw_reg_ptce3_tcam_region_info_memcpy_to(payload, tcam_region_info); + mlxsw_reg_ptce3_flex2_key_blocks_memcpy_to(payload, key); + mlxsw_reg_ptce3_erp_id_set(payload, erp_id); + mlxsw_reg_ptce3_delta_start_set(payload, delta_start); + mlxsw_reg_ptce3_delta_mask_set(payload, delta_mask); + mlxsw_reg_ptce3_delta_value_set(payload, delta_value); + mlxsw_reg_ptce3_large_exists_set(payload, large_exists); + mlxsw_reg_ptce3_large_entry_key_id_set(payload, lkey_id); + mlxsw_reg_ptce3_action_pointer_set(payload, action_pointer); +} + +/* PERCR - Policy-Engine Region Configuration Register + * --------------------------------------------------- + * This register configures the region parameters. The region_id must be + * allocated. + */ +#define MLXSW_REG_PERCR_ID 0x302A +#define MLXSW_REG_PERCR_LEN 0x80 + +MLXSW_REG_DEFINE(percr, MLXSW_REG_PERCR_ID, MLXSW_REG_PERCR_LEN); + +/* reg_percr_region_id + * Region identifier. + * Range 0..cap_max_regions-1 + * Access: Index + */ +MLXSW_ITEM32(reg, percr, region_id, 0x00, 0, 16); + +/* reg_percr_atcam_ignore_prune + * Ignore prune_vector by other A-TCAM rules. Used e.g., for a new rule. + * Access: RW + */ +MLXSW_ITEM32(reg, percr, atcam_ignore_prune, 0x04, 25, 1); + +/* reg_percr_ctcam_ignore_prune + * Ignore prune_ctcam by other A-TCAM rules. Used e.g., for a new rule. + * Access: RW + */ +MLXSW_ITEM32(reg, percr, ctcam_ignore_prune, 0x04, 24, 1); + +/* reg_percr_bf_bypass + * Bloom filter bypass. + * 0 - Bloom filter is used (default) + * 1 - Bloom filter is bypassed. The bypass is an OR condition of + * region_id or eRP. See PERPT.bf_bypass + * Access: RW + */ +MLXSW_ITEM32(reg, percr, bf_bypass, 0x04, 16, 1); + +/* reg_percr_master_mask + * Master mask. Logical OR mask of all masks of all rules of a region + * (both A-TCAM and C-TCAM). When there are no eRPs + * (erpt_pointer_valid = 0), then this provides the mask. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, percr, master_mask, 0x20, 96); + +static inline void mlxsw_reg_percr_pack(char *payload, u16 region_id) +{ + MLXSW_REG_ZERO(percr, payload); + mlxsw_reg_percr_region_id_set(payload, region_id); + mlxsw_reg_percr_atcam_ignore_prune_set(payload, false); + mlxsw_reg_percr_ctcam_ignore_prune_set(payload, false); + mlxsw_reg_percr_bf_bypass_set(payload, false); +} + +/* PERERP - Policy-Engine Region eRP Register + * ------------------------------------------ + * This register configures the region eRP. The region_id must be + * allocated. + */ +#define MLXSW_REG_PERERP_ID 0x302B +#define MLXSW_REG_PERERP_LEN 0x1C + +MLXSW_REG_DEFINE(pererp, MLXSW_REG_PERERP_ID, MLXSW_REG_PERERP_LEN); + +/* reg_pererp_region_id + * Region identifier. + * Range 0..cap_max_regions-1 + * Access: Index + */ +MLXSW_ITEM32(reg, pererp, region_id, 0x00, 0, 16); + +/* reg_pererp_ctcam_le + * C-TCAM lookup enable. Reserved when erpt_pointer_valid = 0. + * Access: RW + */ +MLXSW_ITEM32(reg, pererp, ctcam_le, 0x04, 28, 1); + +/* reg_pererp_erpt_pointer_valid + * erpt_pointer is valid. + * Access: RW + */ +MLXSW_ITEM32(reg, pererp, erpt_pointer_valid, 0x10, 31, 1); + +/* reg_pererp_erpt_bank_pointer + * Pointer to eRP table bank. May be modified at any time. + * Range 0..cap_max_erp_table_banks-1 + * Reserved when erpt_pointer_valid = 0 + */ +MLXSW_ITEM32(reg, pererp, erpt_bank_pointer, 0x10, 16, 4); + +/* reg_pererp_erpt_pointer + * Pointer to eRP table within the eRP bank. Can be changed for an + * existing region. + * Range 0..cap_max_erp_table_size-1 + * Reserved when erpt_pointer_valid = 0 + * Access: RW + */ +MLXSW_ITEM32(reg, pererp, erpt_pointer, 0x10, 0, 8); + +/* reg_pererp_erpt_vector + * Vector of allowed eRP indexes starting from erpt_pointer within the + * erpt_bank_pointer. Next entries will be in next bank. + * Note that eRP index is used and not eRP ID. + * Reserved when erpt_pointer_valid = 0 + * Access: RW + */ +MLXSW_ITEM_BIT_ARRAY(reg, pererp, erpt_vector, 0x14, 4, 1); + +/* reg_pererp_master_rp_id + * Master RP ID. When there are no eRPs, then this provides the eRP ID + * for the lookup. Can be changed for an existing region. + * Reserved when erpt_pointer_valid = 1 + * Access: RW + */ +MLXSW_ITEM32(reg, pererp, master_rp_id, 0x18, 0, 4); + +static inline void mlxsw_reg_pererp_erp_vector_pack(char *payload, + unsigned long *erp_vector, + unsigned long size) +{ + unsigned long bit; + + for_each_set_bit(bit, erp_vector, size) + mlxsw_reg_pererp_erpt_vector_set(payload, bit, true); +} + +static inline void mlxsw_reg_pererp_pack(char *payload, u16 region_id, + bool ctcam_le, bool erpt_pointer_valid, + u8 erpt_bank_pointer, u8 erpt_pointer, + u8 master_rp_id) +{ + MLXSW_REG_ZERO(pererp, payload); + mlxsw_reg_pererp_region_id_set(payload, region_id); + mlxsw_reg_pererp_ctcam_le_set(payload, ctcam_le); + mlxsw_reg_pererp_erpt_pointer_valid_set(payload, erpt_pointer_valid); + mlxsw_reg_pererp_erpt_bank_pointer_set(payload, erpt_bank_pointer); + mlxsw_reg_pererp_erpt_pointer_set(payload, erpt_pointer); + mlxsw_reg_pererp_master_rp_id_set(payload, master_rp_id); +} + +/* PEABFE - Policy-Engine Algorithmic Bloom Filter Entries Register + * ---------------------------------------------------------------- + * This register configures the Bloom filter entries. + */ +#define MLXSW_REG_PEABFE_ID 0x3022 +#define MLXSW_REG_PEABFE_BASE_LEN 0x10 +#define MLXSW_REG_PEABFE_BF_REC_LEN 0x4 +#define MLXSW_REG_PEABFE_BF_REC_MAX_COUNT 256 +#define MLXSW_REG_PEABFE_LEN (MLXSW_REG_PEABFE_BASE_LEN + \ + MLXSW_REG_PEABFE_BF_REC_LEN * \ + MLXSW_REG_PEABFE_BF_REC_MAX_COUNT) + +MLXSW_REG_DEFINE(peabfe, MLXSW_REG_PEABFE_ID, MLXSW_REG_PEABFE_LEN); + +/* reg_peabfe_size + * Number of BF entries to be updated. + * Range 1..256 + * Access: Op + */ +MLXSW_ITEM32(reg, peabfe, size, 0x00, 0, 9); + +/* reg_peabfe_bf_entry_state + * Bloom filter state + * 0 - Clear + * 1 - Set + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, peabfe, bf_entry_state, + MLXSW_REG_PEABFE_BASE_LEN, 31, 1, + MLXSW_REG_PEABFE_BF_REC_LEN, 0x00, false); + +/* reg_peabfe_bf_entry_bank + * Bloom filter bank ID + * Range 0..cap_max_erp_table_banks-1 + * Access: Index + */ +MLXSW_ITEM32_INDEXED(reg, peabfe, bf_entry_bank, + MLXSW_REG_PEABFE_BASE_LEN, 24, 4, + MLXSW_REG_PEABFE_BF_REC_LEN, 0x00, false); + +/* reg_peabfe_bf_entry_index + * Bloom filter entry index + * Range 0..2^cap_max_bf_log-1 + * Access: Index + */ +MLXSW_ITEM32_INDEXED(reg, peabfe, bf_entry_index, + MLXSW_REG_PEABFE_BASE_LEN, 0, 24, + MLXSW_REG_PEABFE_BF_REC_LEN, 0x00, false); + +static inline void mlxsw_reg_peabfe_pack(char *payload) +{ + MLXSW_REG_ZERO(peabfe, payload); +} + +static inline void mlxsw_reg_peabfe_rec_pack(char *payload, int rec_index, + u8 state, u8 bank, u32 bf_index) +{ + u8 num_rec = mlxsw_reg_peabfe_size_get(payload); + + if (rec_index >= num_rec) + mlxsw_reg_peabfe_size_set(payload, rec_index + 1); + mlxsw_reg_peabfe_bf_entry_state_set(payload, rec_index, state); + mlxsw_reg_peabfe_bf_entry_bank_set(payload, rec_index, bank); + mlxsw_reg_peabfe_bf_entry_index_set(payload, rec_index, bf_index); +} + +/* IEDR - Infrastructure Entry Delete Register + * ---------------------------------------------------- + * This register is used for deleting entries from the entry tables. + * It is legitimate to attempt to delete a nonexisting entry (the device will + * respond as a good flow). + */ +#define MLXSW_REG_IEDR_ID 0x3804 +#define MLXSW_REG_IEDR_BASE_LEN 0x10 /* base length, without records */ +#define MLXSW_REG_IEDR_REC_LEN 0x8 /* record length */ +#define MLXSW_REG_IEDR_REC_MAX_COUNT 64 +#define MLXSW_REG_IEDR_LEN (MLXSW_REG_IEDR_BASE_LEN + \ + MLXSW_REG_IEDR_REC_LEN * \ + MLXSW_REG_IEDR_REC_MAX_COUNT) + +MLXSW_REG_DEFINE(iedr, MLXSW_REG_IEDR_ID, MLXSW_REG_IEDR_LEN); + +/* reg_iedr_num_rec + * Number of records. + * Access: OP + */ +MLXSW_ITEM32(reg, iedr, num_rec, 0x00, 0, 8); + +/* reg_iedr_rec_type + * Resource type. + * Access: OP + */ +MLXSW_ITEM32_INDEXED(reg, iedr, rec_type, MLXSW_REG_IEDR_BASE_LEN, 24, 8, + MLXSW_REG_IEDR_REC_LEN, 0x00, false); + +/* reg_iedr_rec_size + * Size of entries do be deleted. The unit is 1 entry, regardless of entry type. + * Access: OP + */ +MLXSW_ITEM32_INDEXED(reg, iedr, rec_size, MLXSW_REG_IEDR_BASE_LEN, 0, 13, + MLXSW_REG_IEDR_REC_LEN, 0x00, false); + +/* reg_iedr_rec_index_start + * Resource index start. + * Access: OP + */ +MLXSW_ITEM32_INDEXED(reg, iedr, rec_index_start, MLXSW_REG_IEDR_BASE_LEN, 0, 24, + MLXSW_REG_IEDR_REC_LEN, 0x04, false); + +static inline void mlxsw_reg_iedr_pack(char *payload) +{ + MLXSW_REG_ZERO(iedr, payload); +} + +static inline void mlxsw_reg_iedr_rec_pack(char *payload, int rec_index, + u8 rec_type, u16 rec_size, + u32 rec_index_start) +{ + u8 num_rec = mlxsw_reg_iedr_num_rec_get(payload); + + if (rec_index >= num_rec) + mlxsw_reg_iedr_num_rec_set(payload, rec_index + 1); + mlxsw_reg_iedr_rec_type_set(payload, rec_index, rec_type); + mlxsw_reg_iedr_rec_size_set(payload, rec_index, rec_size); + mlxsw_reg_iedr_rec_index_start_set(payload, rec_index, rec_index_start); +} + +/* QPTS - QoS Priority Trust State Register + * ---------------------------------------- + * This register controls the port policy to calculate the switch priority and + * packet color based on incoming packet fields. + */ +#define MLXSW_REG_QPTS_ID 0x4002 +#define MLXSW_REG_QPTS_LEN 0x8 + +MLXSW_REG_DEFINE(qpts, MLXSW_REG_QPTS_ID, MLXSW_REG_QPTS_LEN); + +/* reg_qpts_local_port + * Local port number. + * Access: Index + * + * Note: CPU port is supported. + */ +MLXSW_ITEM32_LP(reg, qpts, 0x00, 16, 0x00, 12); + +enum mlxsw_reg_qpts_trust_state { + MLXSW_REG_QPTS_TRUST_STATE_PCP = 1, + MLXSW_REG_QPTS_TRUST_STATE_DSCP = 2, /* For MPLS, trust EXP. */ +}; + +/* reg_qpts_trust_state + * Trust state for a given port. + * Access: RW + */ +MLXSW_ITEM32(reg, qpts, trust_state, 0x04, 0, 3); + +static inline void mlxsw_reg_qpts_pack(char *payload, u16 local_port, + enum mlxsw_reg_qpts_trust_state ts) +{ + MLXSW_REG_ZERO(qpts, payload); + + mlxsw_reg_qpts_local_port_set(payload, local_port); + mlxsw_reg_qpts_trust_state_set(payload, ts); +} + +/* QPCR - QoS Policer Configuration Register + * ----------------------------------------- + * The QPCR register is used to create policers - that limit + * the rate of bytes or packets via some trap group. + */ +#define MLXSW_REG_QPCR_ID 0x4004 +#define MLXSW_REG_QPCR_LEN 0x28 + +MLXSW_REG_DEFINE(qpcr, MLXSW_REG_QPCR_ID, MLXSW_REG_QPCR_LEN); + +enum mlxsw_reg_qpcr_g { + MLXSW_REG_QPCR_G_GLOBAL = 2, + MLXSW_REG_QPCR_G_STORM_CONTROL = 3, +}; + +/* reg_qpcr_g + * The policer type. + * Access: Index + */ +MLXSW_ITEM32(reg, qpcr, g, 0x00, 14, 2); + +/* reg_qpcr_pid + * Policer ID. + * Access: Index + */ +MLXSW_ITEM32(reg, qpcr, pid, 0x00, 0, 14); + +/* reg_qpcr_clear_counter + * Clear counters. + * Access: OP + */ +MLXSW_ITEM32(reg, qpcr, clear_counter, 0x04, 31, 1); + +/* reg_qpcr_color_aware + * Is the policer aware of colors. + * Must be 0 (unaware) for cpu port. + * Access: RW for unbounded policer. RO for bounded policer. + */ +MLXSW_ITEM32(reg, qpcr, color_aware, 0x04, 15, 1); + +/* reg_qpcr_bytes + * Is policer limit is for bytes per sec or packets per sec. + * 0 - packets + * 1 - bytes + * Access: RW for unbounded policer. RO for bounded policer. + */ +MLXSW_ITEM32(reg, qpcr, bytes, 0x04, 14, 1); + +enum mlxsw_reg_qpcr_ir_units { + MLXSW_REG_QPCR_IR_UNITS_M, + MLXSW_REG_QPCR_IR_UNITS_K, +}; + +/* reg_qpcr_ir_units + * Policer's units for cir and eir fields (for bytes limits only) + * 1 - 10^3 + * 0 - 10^6 + * Access: OP + */ +MLXSW_ITEM32(reg, qpcr, ir_units, 0x04, 12, 1); + +enum mlxsw_reg_qpcr_rate_type { + MLXSW_REG_QPCR_RATE_TYPE_SINGLE = 1, + MLXSW_REG_QPCR_RATE_TYPE_DOUBLE = 2, +}; + +/* reg_qpcr_rate_type + * Policer can have one limit (single rate) or 2 limits with specific operation + * for packets that exceed the lower rate but not the upper one. + * (For cpu port must be single rate) + * Access: RW for unbounded policer. RO for bounded policer. + */ +MLXSW_ITEM32(reg, qpcr, rate_type, 0x04, 8, 2); + +/* reg_qpc_cbs + * Policer's committed burst size. + * The policer is working with time slices of 50 nano sec. By default every + * slice is granted the proportionate share of the committed rate. If we want to + * allow a slice to exceed that share (while still keeping the rate per sec) we + * can allow burst. The burst size is between the default proportionate share + * (and no lower than 8) to 32Gb. (Even though giving a number higher than the + * committed rate will result in exceeding the rate). The burst size must be a + * log of 2 and will be determined by 2^cbs. + * Access: RW + */ +MLXSW_ITEM32(reg, qpcr, cbs, 0x08, 24, 6); + +/* reg_qpcr_cir + * Policer's committed rate. + * The rate used for sungle rate, the lower rate for double rate. + * For bytes limits, the rate will be this value * the unit from ir_units. + * (Resolution error is up to 1%). + * Access: RW + */ +MLXSW_ITEM32(reg, qpcr, cir, 0x0C, 0, 32); + +/* reg_qpcr_eir + * Policer's exceed rate. + * The higher rate for double rate, reserved for single rate. + * Lower rate for double rate policer. + * For bytes limits, the rate will be this value * the unit from ir_units. + * (Resolution error is up to 1%). + * Access: RW + */ +MLXSW_ITEM32(reg, qpcr, eir, 0x10, 0, 32); + +#define MLXSW_REG_QPCR_DOUBLE_RATE_ACTION 2 + +/* reg_qpcr_exceed_action. + * What to do with packets between the 2 limits for double rate. + * Access: RW for unbounded policer. RO for bounded policer. + */ +MLXSW_ITEM32(reg, qpcr, exceed_action, 0x14, 0, 4); + +enum mlxsw_reg_qpcr_action { + /* Discard */ + MLXSW_REG_QPCR_ACTION_DISCARD = 1, + /* Forward and set color to red. + * If the packet is intended to cpu port, it will be dropped. + */ + MLXSW_REG_QPCR_ACTION_FORWARD = 2, +}; + +/* reg_qpcr_violate_action + * What to do with packets that cross the cir limit (for single rate) or the eir + * limit (for double rate). + * Access: RW for unbounded policer. RO for bounded policer. + */ +MLXSW_ITEM32(reg, qpcr, violate_action, 0x18, 0, 4); + +/* reg_qpcr_violate_count + * Counts the number of times violate_action happened on this PID. + * Access: RW + */ +MLXSW_ITEM64(reg, qpcr, violate_count, 0x20, 0, 64); + +/* Packets */ +#define MLXSW_REG_QPCR_LOWEST_CIR 1 +#define MLXSW_REG_QPCR_HIGHEST_CIR (2 * 1000 * 1000 * 1000) /* 2Gpps */ +#define MLXSW_REG_QPCR_LOWEST_CBS 4 +#define MLXSW_REG_QPCR_HIGHEST_CBS 24 + +/* Bandwidth */ +#define MLXSW_REG_QPCR_LOWEST_CIR_BITS 1024 /* bps */ +#define MLXSW_REG_QPCR_HIGHEST_CIR_BITS 2000000000000ULL /* 2Tbps */ +#define MLXSW_REG_QPCR_LOWEST_CBS_BITS_SP1 4 +#define MLXSW_REG_QPCR_LOWEST_CBS_BITS_SP2 4 +#define MLXSW_REG_QPCR_HIGHEST_CBS_BITS_SP1 25 +#define MLXSW_REG_QPCR_HIGHEST_CBS_BITS_SP2 31 + +static inline void mlxsw_reg_qpcr_pack(char *payload, u16 pid, + enum mlxsw_reg_qpcr_ir_units ir_units, + bool bytes, u32 cir, u16 cbs) +{ + MLXSW_REG_ZERO(qpcr, payload); + mlxsw_reg_qpcr_pid_set(payload, pid); + mlxsw_reg_qpcr_g_set(payload, MLXSW_REG_QPCR_G_GLOBAL); + mlxsw_reg_qpcr_rate_type_set(payload, MLXSW_REG_QPCR_RATE_TYPE_SINGLE); + mlxsw_reg_qpcr_violate_action_set(payload, + MLXSW_REG_QPCR_ACTION_DISCARD); + mlxsw_reg_qpcr_cir_set(payload, cir); + mlxsw_reg_qpcr_ir_units_set(payload, ir_units); + mlxsw_reg_qpcr_bytes_set(payload, bytes); + mlxsw_reg_qpcr_cbs_set(payload, cbs); +} + +/* QTCT - QoS Switch Traffic Class Table + * ------------------------------------- + * Configures the mapping between the packet switch priority and the + * traffic class on the transmit port. + */ +#define MLXSW_REG_QTCT_ID 0x400A +#define MLXSW_REG_QTCT_LEN 0x08 + +MLXSW_REG_DEFINE(qtct, MLXSW_REG_QTCT_ID, MLXSW_REG_QTCT_LEN); + +/* reg_qtct_local_port + * Local port number. + * Access: Index + * + * Note: CPU port is not supported. + */ +MLXSW_ITEM32_LP(reg, qtct, 0x00, 16, 0x00, 12); + +/* reg_qtct_sub_port + * Virtual port within the physical port. + * Should be set to 0 when virtual ports are not enabled on the port. + * Access: Index + */ +MLXSW_ITEM32(reg, qtct, sub_port, 0x00, 8, 8); + +/* reg_qtct_switch_prio + * Switch priority. + * Access: Index + */ +MLXSW_ITEM32(reg, qtct, switch_prio, 0x00, 0, 4); + +/* reg_qtct_tclass + * Traffic class. + * Default values: + * switch_prio 0 : tclass 1 + * switch_prio 1 : tclass 0 + * switch_prio i : tclass i, for i > 1 + * Access: RW + */ +MLXSW_ITEM32(reg, qtct, tclass, 0x04, 0, 4); + +static inline void mlxsw_reg_qtct_pack(char *payload, u16 local_port, + u8 switch_prio, u8 tclass) +{ + MLXSW_REG_ZERO(qtct, payload); + mlxsw_reg_qtct_local_port_set(payload, local_port); + mlxsw_reg_qtct_switch_prio_set(payload, switch_prio); + mlxsw_reg_qtct_tclass_set(payload, tclass); +} + +/* QEEC - QoS ETS Element Configuration Register + * --------------------------------------------- + * Configures the ETS elements. + */ +#define MLXSW_REG_QEEC_ID 0x400D +#define MLXSW_REG_QEEC_LEN 0x20 + +MLXSW_REG_DEFINE(qeec, MLXSW_REG_QEEC_ID, MLXSW_REG_QEEC_LEN); + +/* reg_qeec_local_port + * Local port number. + * Access: Index + * + * Note: CPU port is supported. + */ +MLXSW_ITEM32_LP(reg, qeec, 0x00, 16, 0x00, 12); + +enum mlxsw_reg_qeec_hr { + MLXSW_REG_QEEC_HR_PORT, + MLXSW_REG_QEEC_HR_GROUP, + MLXSW_REG_QEEC_HR_SUBGROUP, + MLXSW_REG_QEEC_HR_TC, +}; + +/* reg_qeec_element_hierarchy + * 0 - Port + * 1 - Group + * 2 - Subgroup + * 3 - Traffic Class + * Access: Index + */ +MLXSW_ITEM32(reg, qeec, element_hierarchy, 0x04, 16, 4); + +/* reg_qeec_element_index + * The index of the element in the hierarchy. + * Access: Index + */ +MLXSW_ITEM32(reg, qeec, element_index, 0x04, 0, 8); + +/* reg_qeec_next_element_index + * The index of the next (lower) element in the hierarchy. + * Access: RW + * + * Note: Reserved for element_hierarchy 0. + */ +MLXSW_ITEM32(reg, qeec, next_element_index, 0x08, 0, 8); + +/* reg_qeec_mise + * Min shaper configuration enable. Enables configuration of the min + * shaper on this ETS element + * 0 - Disable + * 1 - Enable + * Access: RW + */ +MLXSW_ITEM32(reg, qeec, mise, 0x0C, 31, 1); + +/* reg_qeec_ptps + * PTP shaper + * 0: regular shaper mode + * 1: PTP oriented shaper + * Allowed only for hierarchy 0 + * Not supported for CPU port + * Note that ptps mode may affect the shaper rates of all hierarchies + * Supported only on Spectrum-1 + * Access: RW + */ +MLXSW_ITEM32(reg, qeec, ptps, 0x0C, 29, 1); + +enum { + MLXSW_REG_QEEC_BYTES_MODE, + MLXSW_REG_QEEC_PACKETS_MODE, +}; + +/* reg_qeec_pb + * Packets or bytes mode. + * 0 - Bytes mode + * 1 - Packets mode + * Access: RW + * + * Note: Used for max shaper configuration. For Spectrum, packets mode + * is supported only for traffic classes of CPU port. + */ +MLXSW_ITEM32(reg, qeec, pb, 0x0C, 28, 1); + +/* The smallest permitted min shaper rate. */ +#define MLXSW_REG_QEEC_MIS_MIN 200000 /* Kbps */ + +/* reg_qeec_min_shaper_rate + * Min shaper information rate. + * For CPU port, can only be configured for port hierarchy. + * When in bytes mode, value is specified in units of 1000bps. + * Access: RW + */ +MLXSW_ITEM32(reg, qeec, min_shaper_rate, 0x0C, 0, 28); + +/* reg_qeec_mase + * Max shaper configuration enable. Enables configuration of the max + * shaper on this ETS element. + * 0 - Disable + * 1 - Enable + * Access: RW + */ +MLXSW_ITEM32(reg, qeec, mase, 0x10, 31, 1); + +/* The largest max shaper value possible to disable the shaper. */ +#define MLXSW_REG_QEEC_MAS_DIS ((1u << 31) - 1) /* Kbps */ + +/* reg_qeec_max_shaper_rate + * Max shaper information rate. + * For CPU port, can only be configured for port hierarchy. + * When in bytes mode, value is specified in units of 1000bps. + * Access: RW + */ +MLXSW_ITEM32(reg, qeec, max_shaper_rate, 0x10, 0, 31); + +/* reg_qeec_de + * DWRR configuration enable. Enables configuration of the dwrr and + * dwrr_weight. + * 0 - Disable + * 1 - Enable + * Access: RW + */ +MLXSW_ITEM32(reg, qeec, de, 0x18, 31, 1); + +/* reg_qeec_dwrr + * Transmission selection algorithm to use on the link going down from + * the ETS element. + * 0 - Strict priority + * 1 - DWRR + * Access: RW + */ +MLXSW_ITEM32(reg, qeec, dwrr, 0x18, 15, 1); + +/* reg_qeec_dwrr_weight + * DWRR weight on the link going down from the ETS element. The + * percentage of bandwidth guaranteed to an ETS element within + * its hierarchy. The sum of all weights across all ETS elements + * within one hierarchy should be equal to 100. Reserved when + * transmission selection algorithm is strict priority. + * Access: RW + */ +MLXSW_ITEM32(reg, qeec, dwrr_weight, 0x18, 0, 8); + +/* reg_qeec_max_shaper_bs + * Max shaper burst size + * Burst size is 2^max_shaper_bs * 512 bits + * For Spectrum-1: Range is: 5..25 + * For Spectrum-2: Range is: 11..25 + * Reserved when ptps = 1 + * Access: RW + */ +MLXSW_ITEM32(reg, qeec, max_shaper_bs, 0x1C, 0, 6); + +#define MLXSW_REG_QEEC_HIGHEST_SHAPER_BS 25 +#define MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP1 5 +#define MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP2 11 +#define MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP3 11 +#define MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP4 11 + +static inline void mlxsw_reg_qeec_pack(char *payload, u16 local_port, + enum mlxsw_reg_qeec_hr hr, u8 index, + u8 next_index) +{ + MLXSW_REG_ZERO(qeec, payload); + mlxsw_reg_qeec_local_port_set(payload, local_port); + mlxsw_reg_qeec_element_hierarchy_set(payload, hr); + mlxsw_reg_qeec_element_index_set(payload, index); + mlxsw_reg_qeec_next_element_index_set(payload, next_index); +} + +static inline void mlxsw_reg_qeec_ptps_pack(char *payload, u16 local_port, + bool ptps) +{ + MLXSW_REG_ZERO(qeec, payload); + mlxsw_reg_qeec_local_port_set(payload, local_port); + mlxsw_reg_qeec_element_hierarchy_set(payload, MLXSW_REG_QEEC_HR_PORT); + mlxsw_reg_qeec_ptps_set(payload, ptps); +} + +/* QRWE - QoS ReWrite Enable + * ------------------------- + * This register configures the rewrite enable per receive port. + */ +#define MLXSW_REG_QRWE_ID 0x400F +#define MLXSW_REG_QRWE_LEN 0x08 + +MLXSW_REG_DEFINE(qrwe, MLXSW_REG_QRWE_ID, MLXSW_REG_QRWE_LEN); + +/* reg_qrwe_local_port + * Local port number. + * Access: Index + * + * Note: CPU port is supported. No support for router port. + */ +MLXSW_ITEM32_LP(reg, qrwe, 0x00, 16, 0x00, 12); + +/* reg_qrwe_dscp + * Whether to enable DSCP rewrite (default is 0, don't rewrite). + * Access: RW + */ +MLXSW_ITEM32(reg, qrwe, dscp, 0x04, 1, 1); + +/* reg_qrwe_pcp + * Whether to enable PCP and DEI rewrite (default is 0, don't rewrite). + * Access: RW + */ +MLXSW_ITEM32(reg, qrwe, pcp, 0x04, 0, 1); + +static inline void mlxsw_reg_qrwe_pack(char *payload, u16 local_port, + bool rewrite_pcp, bool rewrite_dscp) +{ + MLXSW_REG_ZERO(qrwe, payload); + mlxsw_reg_qrwe_local_port_set(payload, local_port); + mlxsw_reg_qrwe_pcp_set(payload, rewrite_pcp); + mlxsw_reg_qrwe_dscp_set(payload, rewrite_dscp); +} + +/* QPDSM - QoS Priority to DSCP Mapping + * ------------------------------------ + * QoS Priority to DSCP Mapping Register + */ +#define MLXSW_REG_QPDSM_ID 0x4011 +#define MLXSW_REG_QPDSM_BASE_LEN 0x04 /* base length, without records */ +#define MLXSW_REG_QPDSM_PRIO_ENTRY_REC_LEN 0x4 /* record length */ +#define MLXSW_REG_QPDSM_PRIO_ENTRY_REC_MAX_COUNT 16 +#define MLXSW_REG_QPDSM_LEN (MLXSW_REG_QPDSM_BASE_LEN + \ + MLXSW_REG_QPDSM_PRIO_ENTRY_REC_LEN * \ + MLXSW_REG_QPDSM_PRIO_ENTRY_REC_MAX_COUNT) + +MLXSW_REG_DEFINE(qpdsm, MLXSW_REG_QPDSM_ID, MLXSW_REG_QPDSM_LEN); + +/* reg_qpdsm_local_port + * Local Port. Supported for data packets from CPU port. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, qpdsm, 0x00, 16, 0x00, 12); + +/* reg_qpdsm_prio_entry_color0_e + * Enable update of the entry for color 0 and a given port. + * Access: WO + */ +MLXSW_ITEM32_INDEXED(reg, qpdsm, prio_entry_color0_e, + MLXSW_REG_QPDSM_BASE_LEN, 31, 1, + MLXSW_REG_QPDSM_PRIO_ENTRY_REC_LEN, 0x00, false); + +/* reg_qpdsm_prio_entry_color0_dscp + * DSCP field in the outer label of the packet for color 0 and a given port. + * Reserved when e=0. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, qpdsm, prio_entry_color0_dscp, + MLXSW_REG_QPDSM_BASE_LEN, 24, 6, + MLXSW_REG_QPDSM_PRIO_ENTRY_REC_LEN, 0x00, false); + +/* reg_qpdsm_prio_entry_color1_e + * Enable update of the entry for color 1 and a given port. + * Access: WO + */ +MLXSW_ITEM32_INDEXED(reg, qpdsm, prio_entry_color1_e, + MLXSW_REG_QPDSM_BASE_LEN, 23, 1, + MLXSW_REG_QPDSM_PRIO_ENTRY_REC_LEN, 0x00, false); + +/* reg_qpdsm_prio_entry_color1_dscp + * DSCP field in the outer label of the packet for color 1 and a given port. + * Reserved when e=0. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, qpdsm, prio_entry_color1_dscp, + MLXSW_REG_QPDSM_BASE_LEN, 16, 6, + MLXSW_REG_QPDSM_PRIO_ENTRY_REC_LEN, 0x00, false); + +/* reg_qpdsm_prio_entry_color2_e + * Enable update of the entry for color 2 and a given port. + * Access: WO + */ +MLXSW_ITEM32_INDEXED(reg, qpdsm, prio_entry_color2_e, + MLXSW_REG_QPDSM_BASE_LEN, 15, 1, + MLXSW_REG_QPDSM_PRIO_ENTRY_REC_LEN, 0x00, false); + +/* reg_qpdsm_prio_entry_color2_dscp + * DSCP field in the outer label of the packet for color 2 and a given port. + * Reserved when e=0. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, qpdsm, prio_entry_color2_dscp, + MLXSW_REG_QPDSM_BASE_LEN, 8, 6, + MLXSW_REG_QPDSM_PRIO_ENTRY_REC_LEN, 0x00, false); + +static inline void mlxsw_reg_qpdsm_pack(char *payload, u16 local_port) +{ + MLXSW_REG_ZERO(qpdsm, payload); + mlxsw_reg_qpdsm_local_port_set(payload, local_port); +} + +static inline void +mlxsw_reg_qpdsm_prio_pack(char *payload, unsigned short prio, u8 dscp) +{ + mlxsw_reg_qpdsm_prio_entry_color0_e_set(payload, prio, 1); + mlxsw_reg_qpdsm_prio_entry_color0_dscp_set(payload, prio, dscp); + mlxsw_reg_qpdsm_prio_entry_color1_e_set(payload, prio, 1); + mlxsw_reg_qpdsm_prio_entry_color1_dscp_set(payload, prio, dscp); + mlxsw_reg_qpdsm_prio_entry_color2_e_set(payload, prio, 1); + mlxsw_reg_qpdsm_prio_entry_color2_dscp_set(payload, prio, dscp); +} + +/* QPDP - QoS Port DSCP to Priority Mapping Register + * ------------------------------------------------- + * This register controls the port default Switch Priority and Color. The + * default Switch Priority and Color are used for frames where the trust state + * uses default values. All member ports of a LAG should be configured with the + * same default values. + */ +#define MLXSW_REG_QPDP_ID 0x4007 +#define MLXSW_REG_QPDP_LEN 0x8 + +MLXSW_REG_DEFINE(qpdp, MLXSW_REG_QPDP_ID, MLXSW_REG_QPDP_LEN); + +/* reg_qpdp_local_port + * Local Port. Supported for data packets from CPU port. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, qpdp, 0x00, 16, 0x00, 12); + +/* reg_qpdp_switch_prio + * Default port Switch Priority (default 0) + * Access: RW + */ +MLXSW_ITEM32(reg, qpdp, switch_prio, 0x04, 0, 4); + +static inline void mlxsw_reg_qpdp_pack(char *payload, u16 local_port, + u8 switch_prio) +{ + MLXSW_REG_ZERO(qpdp, payload); + mlxsw_reg_qpdp_local_port_set(payload, local_port); + mlxsw_reg_qpdp_switch_prio_set(payload, switch_prio); +} + +/* QPDPM - QoS Port DSCP to Priority Mapping Register + * -------------------------------------------------- + * This register controls the mapping from DSCP field to + * Switch Priority for IP packets. + */ +#define MLXSW_REG_QPDPM_ID 0x4013 +#define MLXSW_REG_QPDPM_BASE_LEN 0x4 /* base length, without records */ +#define MLXSW_REG_QPDPM_DSCP_ENTRY_REC_LEN 0x2 /* record length */ +#define MLXSW_REG_QPDPM_DSCP_ENTRY_REC_MAX_COUNT 64 +#define MLXSW_REG_QPDPM_LEN (MLXSW_REG_QPDPM_BASE_LEN + \ + MLXSW_REG_QPDPM_DSCP_ENTRY_REC_LEN * \ + MLXSW_REG_QPDPM_DSCP_ENTRY_REC_MAX_COUNT) + +MLXSW_REG_DEFINE(qpdpm, MLXSW_REG_QPDPM_ID, MLXSW_REG_QPDPM_LEN); + +/* reg_qpdpm_local_port + * Local Port. Supported for data packets from CPU port. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, qpdpm, 0x00, 16, 0x00, 12); + +/* reg_qpdpm_dscp_e + * Enable update of the specific entry. When cleared, the switch_prio and color + * fields are ignored and the previous switch_prio and color values are + * preserved. + * Access: WO + */ +MLXSW_ITEM16_INDEXED(reg, qpdpm, dscp_entry_e, MLXSW_REG_QPDPM_BASE_LEN, 15, 1, + MLXSW_REG_QPDPM_DSCP_ENTRY_REC_LEN, 0x00, false); + +/* reg_qpdpm_dscp_prio + * The new Switch Priority value for the relevant DSCP value. + * Access: RW + */ +MLXSW_ITEM16_INDEXED(reg, qpdpm, dscp_entry_prio, + MLXSW_REG_QPDPM_BASE_LEN, 0, 4, + MLXSW_REG_QPDPM_DSCP_ENTRY_REC_LEN, 0x00, false); + +static inline void mlxsw_reg_qpdpm_pack(char *payload, u16 local_port) +{ + MLXSW_REG_ZERO(qpdpm, payload); + mlxsw_reg_qpdpm_local_port_set(payload, local_port); +} + +static inline void +mlxsw_reg_qpdpm_dscp_pack(char *payload, unsigned short dscp, u8 prio) +{ + mlxsw_reg_qpdpm_dscp_entry_e_set(payload, dscp, 1); + mlxsw_reg_qpdpm_dscp_entry_prio_set(payload, dscp, prio); +} + +/* QTCTM - QoS Switch Traffic Class Table is Multicast-Aware Register + * ------------------------------------------------------------------ + * This register configures if the Switch Priority to Traffic Class mapping is + * based on Multicast packet indication. If so, then multicast packets will get + * a Traffic Class that is plus (cap_max_tclass_data/2) the value configured by + * QTCT. + * By default, Switch Priority to Traffic Class mapping is not based on + * Multicast packet indication. + */ +#define MLXSW_REG_QTCTM_ID 0x401A +#define MLXSW_REG_QTCTM_LEN 0x08 + +MLXSW_REG_DEFINE(qtctm, MLXSW_REG_QTCTM_ID, MLXSW_REG_QTCTM_LEN); + +/* reg_qtctm_local_port + * Local port number. + * No support for CPU port. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, qtctm, 0x00, 16, 0x00, 12); + +/* reg_qtctm_mc + * Multicast Mode + * Whether Switch Priority to Traffic Class mapping is based on Multicast packet + * indication (default is 0, not based on Multicast packet indication). + */ +MLXSW_ITEM32(reg, qtctm, mc, 0x04, 0, 1); + +static inline void +mlxsw_reg_qtctm_pack(char *payload, u16 local_port, bool mc) +{ + MLXSW_REG_ZERO(qtctm, payload); + mlxsw_reg_qtctm_local_port_set(payload, local_port); + mlxsw_reg_qtctm_mc_set(payload, mc); +} + +/* QPSC - QoS PTP Shaper Configuration Register + * -------------------------------------------- + * The QPSC allows advanced configuration of the shapers when QEEC.ptps=1. + * Supported only on Spectrum-1. + */ +#define MLXSW_REG_QPSC_ID 0x401B +#define MLXSW_REG_QPSC_LEN 0x28 + +MLXSW_REG_DEFINE(qpsc, MLXSW_REG_QPSC_ID, MLXSW_REG_QPSC_LEN); + +enum mlxsw_reg_qpsc_port_speed { + MLXSW_REG_QPSC_PORT_SPEED_100M, + MLXSW_REG_QPSC_PORT_SPEED_1G, + MLXSW_REG_QPSC_PORT_SPEED_10G, + MLXSW_REG_QPSC_PORT_SPEED_25G, +}; + +/* reg_qpsc_port_speed + * Port speed. + * Access: Index + */ +MLXSW_ITEM32(reg, qpsc, port_speed, 0x00, 0, 4); + +/* reg_qpsc_shaper_time_exp + * The base-time-interval for updating the shapers tokens (for all hierarchies). + * shaper_update_rate = 2 ^ shaper_time_exp * (1 + shaper_time_mantissa) * 32nSec + * shaper_rate = 64bit * shaper_inc / shaper_update_rate + * Access: RW + */ +MLXSW_ITEM32(reg, qpsc, shaper_time_exp, 0x04, 16, 4); + +/* reg_qpsc_shaper_time_mantissa + * The base-time-interval for updating the shapers tokens (for all hierarchies). + * shaper_update_rate = 2 ^ shaper_time_exp * (1 + shaper_time_mantissa) * 32nSec + * shaper_rate = 64bit * shaper_inc / shaper_update_rate + * Access: RW + */ +MLXSW_ITEM32(reg, qpsc, shaper_time_mantissa, 0x04, 0, 5); + +/* reg_qpsc_shaper_inc + * Number of tokens added to shaper on each update. + * Units of 8B. + * Access: RW + */ +MLXSW_ITEM32(reg, qpsc, shaper_inc, 0x08, 0, 5); + +/* reg_qpsc_shaper_bs + * Max shaper Burst size. + * Burst size is 2 ^ max_shaper_bs * 512 [bits] + * Range is: 5..25 (from 2KB..2GB) + * Access: RW + */ +MLXSW_ITEM32(reg, qpsc, shaper_bs, 0x0C, 0, 6); + +/* reg_qpsc_ptsc_we + * Write enable to port_to_shaper_credits. + * Access: WO + */ +MLXSW_ITEM32(reg, qpsc, ptsc_we, 0x10, 31, 1); + +/* reg_qpsc_port_to_shaper_credits + * For split ports: range 1..57 + * For non-split ports: range 1..112 + * Written only when ptsc_we is set. + * Access: RW + */ +MLXSW_ITEM32(reg, qpsc, port_to_shaper_credits, 0x10, 0, 8); + +/* reg_qpsc_ing_timestamp_inc + * Ingress timestamp increment. + * 2's complement. + * The timestamp of MTPPTR at ingress will be incremented by this value. Global + * value for all ports. + * Same units as used by MTPPTR. + * Access: RW + */ +MLXSW_ITEM32(reg, qpsc, ing_timestamp_inc, 0x20, 0, 32); + +/* reg_qpsc_egr_timestamp_inc + * Egress timestamp increment. + * 2's complement. + * The timestamp of MTPPTR at egress will be incremented by this value. Global + * value for all ports. + * Same units as used by MTPPTR. + * Access: RW + */ +MLXSW_ITEM32(reg, qpsc, egr_timestamp_inc, 0x24, 0, 32); + +static inline void +mlxsw_reg_qpsc_pack(char *payload, enum mlxsw_reg_qpsc_port_speed port_speed, + u8 shaper_time_exp, u8 shaper_time_mantissa, u8 shaper_inc, + u8 shaper_bs, u8 port_to_shaper_credits, + int ing_timestamp_inc, int egr_timestamp_inc) +{ + MLXSW_REG_ZERO(qpsc, payload); + mlxsw_reg_qpsc_port_speed_set(payload, port_speed); + mlxsw_reg_qpsc_shaper_time_exp_set(payload, shaper_time_exp); + mlxsw_reg_qpsc_shaper_time_mantissa_set(payload, shaper_time_mantissa); + mlxsw_reg_qpsc_shaper_inc_set(payload, shaper_inc); + mlxsw_reg_qpsc_shaper_bs_set(payload, shaper_bs); + mlxsw_reg_qpsc_ptsc_we_set(payload, true); + mlxsw_reg_qpsc_port_to_shaper_credits_set(payload, port_to_shaper_credits); + mlxsw_reg_qpsc_ing_timestamp_inc_set(payload, ing_timestamp_inc); + mlxsw_reg_qpsc_egr_timestamp_inc_set(payload, egr_timestamp_inc); +} + +/* PMLP - Ports Module to Local Port Register + * ------------------------------------------ + * Configures the assignment of modules to local ports. + */ +#define MLXSW_REG_PMLP_ID 0x5002 +#define MLXSW_REG_PMLP_LEN 0x40 + +MLXSW_REG_DEFINE(pmlp, MLXSW_REG_PMLP_ID, MLXSW_REG_PMLP_LEN); + +/* reg_pmlp_rxtx + * 0 - Tx value is used for both Tx and Rx. + * 1 - Rx value is taken from a separte field. + * Access: RW + */ +MLXSW_ITEM32(reg, pmlp, rxtx, 0x00, 31, 1); + +/* reg_pmlp_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, pmlp, 0x00, 16, 0x00, 12); + +/* reg_pmlp_width + * 0 - Unmap local port. + * 1 - Lane 0 is used. + * 2 - Lanes 0 and 1 are used. + * 4 - Lanes 0, 1, 2 and 3 are used. + * 8 - Lanes 0-7 are used. + * Access: RW + */ +MLXSW_ITEM32(reg, pmlp, width, 0x00, 0, 8); + +/* reg_pmlp_module + * Module number. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, pmlp, module, 0x04, 0, 8, 0x04, 0x00, false); + +/* reg_pmlp_slot_index + * Module number. + * Slot_index + * Slot_index = 0 represent the onboard (motherboard). + * In case of non-modular system only slot_index = 0 is available. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, pmlp, slot_index, 0x04, 8, 4, 0x04, 0x00, false); + +/* reg_pmlp_tx_lane + * Tx Lane. When rxtx field is cleared, this field is used for Rx as well. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, pmlp, tx_lane, 0x04, 16, 4, 0x04, 0x00, false); + +/* reg_pmlp_rx_lane + * Rx Lane. When rxtx field is cleared, this field is ignored and Rx lane is + * equal to Tx lane. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, pmlp, rx_lane, 0x04, 24, 4, 0x04, 0x00, false); + +static inline void mlxsw_reg_pmlp_pack(char *payload, u16 local_port) +{ + MLXSW_REG_ZERO(pmlp, payload); + mlxsw_reg_pmlp_local_port_set(payload, local_port); +} + +/* PMTU - Port MTU Register + * ------------------------ + * Configures and reports the port MTU. + */ +#define MLXSW_REG_PMTU_ID 0x5003 +#define MLXSW_REG_PMTU_LEN 0x10 + +MLXSW_REG_DEFINE(pmtu, MLXSW_REG_PMTU_ID, MLXSW_REG_PMTU_LEN); + +/* reg_pmtu_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, pmtu, 0x00, 16, 0x00, 12); + +/* reg_pmtu_max_mtu + * Maximum MTU. + * When port type (e.g. Ethernet) is configured, the relevant MTU is + * reported, otherwise the minimum between the max_mtu of the different + * types is reported. + * Access: RO + */ +MLXSW_ITEM32(reg, pmtu, max_mtu, 0x04, 16, 16); + +/* reg_pmtu_admin_mtu + * MTU value to set port to. Must be smaller or equal to max_mtu. + * Note: If port type is Infiniband, then port must be disabled, when its + * MTU is set. + * Access: RW + */ +MLXSW_ITEM32(reg, pmtu, admin_mtu, 0x08, 16, 16); + +/* reg_pmtu_oper_mtu + * The actual MTU configured on the port. Packets exceeding this size + * will be dropped. + * Note: In Ethernet and FC oper_mtu == admin_mtu, however, in Infiniband + * oper_mtu might be smaller than admin_mtu. + * Access: RO + */ +MLXSW_ITEM32(reg, pmtu, oper_mtu, 0x0C, 16, 16); + +static inline void mlxsw_reg_pmtu_pack(char *payload, u16 local_port, + u16 new_mtu) +{ + MLXSW_REG_ZERO(pmtu, payload); + mlxsw_reg_pmtu_local_port_set(payload, local_port); + mlxsw_reg_pmtu_max_mtu_set(payload, 0); + mlxsw_reg_pmtu_admin_mtu_set(payload, new_mtu); + mlxsw_reg_pmtu_oper_mtu_set(payload, 0); +} + +/* PTYS - Port Type and Speed Register + * ----------------------------------- + * Configures and reports the port speed type. + * + * Note: When set while the link is up, the changes will not take effect + * until the port transitions from down to up state. + */ +#define MLXSW_REG_PTYS_ID 0x5004 +#define MLXSW_REG_PTYS_LEN 0x40 + +MLXSW_REG_DEFINE(ptys, MLXSW_REG_PTYS_ID, MLXSW_REG_PTYS_LEN); + +/* an_disable_admin + * Auto negotiation disable administrative configuration + * 0 - Device doesn't support AN disable. + * 1 - Device supports AN disable. + * Access: RW + */ +MLXSW_ITEM32(reg, ptys, an_disable_admin, 0x00, 30, 1); + +/* reg_ptys_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, ptys, 0x00, 16, 0x00, 12); + +#define MLXSW_REG_PTYS_PROTO_MASK_IB BIT(0) +#define MLXSW_REG_PTYS_PROTO_MASK_ETH BIT(2) + +/* reg_ptys_proto_mask + * Protocol mask. Indicates which protocol is used. + * 0 - Infiniband. + * 1 - Fibre Channel. + * 2 - Ethernet. + * Access: Index + */ +MLXSW_ITEM32(reg, ptys, proto_mask, 0x00, 0, 3); + +enum { + MLXSW_REG_PTYS_AN_STATUS_NA, + MLXSW_REG_PTYS_AN_STATUS_OK, + MLXSW_REG_PTYS_AN_STATUS_FAIL, +}; + +/* reg_ptys_an_status + * Autonegotiation status. + * Access: RO + */ +MLXSW_ITEM32(reg, ptys, an_status, 0x04, 28, 4); + +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_SGMII_100M BIT(0) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_1000BASE_X_SGMII BIT(1) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_5GBASE_R BIT(3) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_XFI_XAUI_1_10G BIT(4) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_XLAUI_4_XLPPI_4_40G BIT(5) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_25GAUI_1_25GBASE_CR_KR BIT(6) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_50GAUI_2_LAUI_2_50GBASE_CR2_KR2 BIT(7) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_50GAUI_1_LAUI_1_50GBASE_CR_KR BIT(8) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_CAUI_4_100GBASE_CR4_KR4 BIT(9) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_100GAUI_2_100GBASE_CR2_KR2 BIT(10) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_200GAUI_4_200GBASE_CR4_KR4 BIT(12) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_400GAUI_8 BIT(15) + +/* reg_ptys_ext_eth_proto_cap + * Extended Ethernet port supported speeds and protocols. + * Access: RO + */ +MLXSW_ITEM32(reg, ptys, ext_eth_proto_cap, 0x08, 0, 32); + +#define MLXSW_REG_PTYS_ETH_SPEED_SGMII BIT(0) +#define MLXSW_REG_PTYS_ETH_SPEED_1000BASE_KX BIT(1) +#define MLXSW_REG_PTYS_ETH_SPEED_10GBASE_CX4 BIT(2) +#define MLXSW_REG_PTYS_ETH_SPEED_10GBASE_KX4 BIT(3) +#define MLXSW_REG_PTYS_ETH_SPEED_10GBASE_KR BIT(4) +#define MLXSW_REG_PTYS_ETH_SPEED_40GBASE_CR4 BIT(6) +#define MLXSW_REG_PTYS_ETH_SPEED_40GBASE_KR4 BIT(7) +#define MLXSW_REG_PTYS_ETH_SPEED_10GBASE_CR BIT(12) +#define MLXSW_REG_PTYS_ETH_SPEED_10GBASE_SR BIT(13) +#define MLXSW_REG_PTYS_ETH_SPEED_10GBASE_ER_LR BIT(14) +#define MLXSW_REG_PTYS_ETH_SPEED_40GBASE_SR4 BIT(15) +#define MLXSW_REG_PTYS_ETH_SPEED_40GBASE_LR4_ER4 BIT(16) +#define MLXSW_REG_PTYS_ETH_SPEED_50GBASE_SR2 BIT(18) +#define MLXSW_REG_PTYS_ETH_SPEED_50GBASE_KR4 BIT(19) +#define MLXSW_REG_PTYS_ETH_SPEED_100GBASE_CR4 BIT(20) +#define MLXSW_REG_PTYS_ETH_SPEED_100GBASE_SR4 BIT(21) +#define MLXSW_REG_PTYS_ETH_SPEED_100GBASE_KR4 BIT(22) +#define MLXSW_REG_PTYS_ETH_SPEED_100GBASE_LR4_ER4 BIT(23) +#define MLXSW_REG_PTYS_ETH_SPEED_100BASE_T BIT(24) +#define MLXSW_REG_PTYS_ETH_SPEED_1000BASE_T BIT(25) +#define MLXSW_REG_PTYS_ETH_SPEED_25GBASE_CR BIT(27) +#define MLXSW_REG_PTYS_ETH_SPEED_25GBASE_KR BIT(28) +#define MLXSW_REG_PTYS_ETH_SPEED_25GBASE_SR BIT(29) +#define MLXSW_REG_PTYS_ETH_SPEED_50GBASE_CR2 BIT(30) +#define MLXSW_REG_PTYS_ETH_SPEED_50GBASE_KR2 BIT(31) + +/* reg_ptys_eth_proto_cap + * Ethernet port supported speeds and protocols. + * Access: RO + */ +MLXSW_ITEM32(reg, ptys, eth_proto_cap, 0x0C, 0, 32); + +/* reg_ptys_ext_eth_proto_admin + * Extended speed and protocol to set port to. + * Access: RW + */ +MLXSW_ITEM32(reg, ptys, ext_eth_proto_admin, 0x14, 0, 32); + +/* reg_ptys_eth_proto_admin + * Speed and protocol to set port to. + * Access: RW + */ +MLXSW_ITEM32(reg, ptys, eth_proto_admin, 0x18, 0, 32); + +/* reg_ptys_ext_eth_proto_oper + * The extended current speed and protocol configured for the port. + * Access: RO + */ +MLXSW_ITEM32(reg, ptys, ext_eth_proto_oper, 0x20, 0, 32); + +/* reg_ptys_eth_proto_oper + * The current speed and protocol configured for the port. + * Access: RO + */ +MLXSW_ITEM32(reg, ptys, eth_proto_oper, 0x24, 0, 32); + +enum mlxsw_reg_ptys_connector_type { + MLXSW_REG_PTYS_CONNECTOR_TYPE_UNKNOWN_OR_NO_CONNECTOR, + MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_NONE, + MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_TP, + MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_AUI, + MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_BNC, + MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_MII, + MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_FIBRE, + MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_DA, + MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_OTHER, +}; + +/* reg_ptys_connector_type + * Connector type indication. + * Access: RO + */ +MLXSW_ITEM32(reg, ptys, connector_type, 0x2C, 0, 4); + +static inline void mlxsw_reg_ptys_eth_pack(char *payload, u16 local_port, + u32 proto_admin, bool autoneg) +{ + MLXSW_REG_ZERO(ptys, payload); + mlxsw_reg_ptys_local_port_set(payload, local_port); + mlxsw_reg_ptys_proto_mask_set(payload, MLXSW_REG_PTYS_PROTO_MASK_ETH); + mlxsw_reg_ptys_eth_proto_admin_set(payload, proto_admin); + mlxsw_reg_ptys_an_disable_admin_set(payload, !autoneg); +} + +static inline void mlxsw_reg_ptys_ext_eth_pack(char *payload, u16 local_port, + u32 proto_admin, bool autoneg) +{ + MLXSW_REG_ZERO(ptys, payload); + mlxsw_reg_ptys_local_port_set(payload, local_port); + mlxsw_reg_ptys_proto_mask_set(payload, MLXSW_REG_PTYS_PROTO_MASK_ETH); + mlxsw_reg_ptys_ext_eth_proto_admin_set(payload, proto_admin); + mlxsw_reg_ptys_an_disable_admin_set(payload, !autoneg); +} + +static inline void mlxsw_reg_ptys_eth_unpack(char *payload, + u32 *p_eth_proto_cap, + u32 *p_eth_proto_admin, + u32 *p_eth_proto_oper) +{ + if (p_eth_proto_cap) + *p_eth_proto_cap = + mlxsw_reg_ptys_eth_proto_cap_get(payload); + if (p_eth_proto_admin) + *p_eth_proto_admin = + mlxsw_reg_ptys_eth_proto_admin_get(payload); + if (p_eth_proto_oper) + *p_eth_proto_oper = + mlxsw_reg_ptys_eth_proto_oper_get(payload); +} + +static inline void mlxsw_reg_ptys_ext_eth_unpack(char *payload, + u32 *p_eth_proto_cap, + u32 *p_eth_proto_admin, + u32 *p_eth_proto_oper) +{ + if (p_eth_proto_cap) + *p_eth_proto_cap = + mlxsw_reg_ptys_ext_eth_proto_cap_get(payload); + if (p_eth_proto_admin) + *p_eth_proto_admin = + mlxsw_reg_ptys_ext_eth_proto_admin_get(payload); + if (p_eth_proto_oper) + *p_eth_proto_oper = + mlxsw_reg_ptys_ext_eth_proto_oper_get(payload); +} + +/* PPAD - Port Physical Address Register + * ------------------------------------- + * The PPAD register configures the per port physical MAC address. + */ +#define MLXSW_REG_PPAD_ID 0x5005 +#define MLXSW_REG_PPAD_LEN 0x10 + +MLXSW_REG_DEFINE(ppad, MLXSW_REG_PPAD_ID, MLXSW_REG_PPAD_LEN); + +/* reg_ppad_single_base_mac + * 0: base_mac, local port should be 0 and mac[7:0] is + * reserved. HW will set incremental + * 1: single_mac - mac of the local_port + * Access: RW + */ +MLXSW_ITEM32(reg, ppad, single_base_mac, 0x00, 28, 1); + +/* reg_ppad_local_port + * port number, if single_base_mac = 0 then local_port is reserved + * Access: RW + */ +MLXSW_ITEM32_LP(reg, ppad, 0x00, 16, 0x00, 24); + +/* reg_ppad_mac + * If single_base_mac = 0 - base MAC address, mac[7:0] is reserved. + * If single_base_mac = 1 - the per port MAC address + * Access: RW + */ +MLXSW_ITEM_BUF(reg, ppad, mac, 0x02, 6); + +static inline void mlxsw_reg_ppad_pack(char *payload, bool single_base_mac, + u16 local_port) +{ + MLXSW_REG_ZERO(ppad, payload); + mlxsw_reg_ppad_single_base_mac_set(payload, !!single_base_mac); + mlxsw_reg_ppad_local_port_set(payload, local_port); +} + +/* PAOS - Ports Administrative and Operational Status Register + * ----------------------------------------------------------- + * Configures and retrieves per port administrative and operational status. + */ +#define MLXSW_REG_PAOS_ID 0x5006 +#define MLXSW_REG_PAOS_LEN 0x10 + +MLXSW_REG_DEFINE(paos, MLXSW_REG_PAOS_ID, MLXSW_REG_PAOS_LEN); + +/* reg_paos_swid + * Switch partition ID with which to associate the port. + * Note: while external ports uses unique local port numbers (and thus swid is + * redundant), router ports use the same local port number where swid is the + * only indication for the relevant port. + * Access: Index + */ +MLXSW_ITEM32(reg, paos, swid, 0x00, 24, 8); + +/* reg_paos_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, paos, 0x00, 16, 0x00, 12); + +/* reg_paos_admin_status + * Port administrative state (the desired state of the port): + * 1 - Up. + * 2 - Down. + * 3 - Up once. This means that in case of link failure, the port won't go + * into polling mode, but will wait to be re-enabled by software. + * 4 - Disabled by system. Can only be set by hardware. + * Access: RW + */ +MLXSW_ITEM32(reg, paos, admin_status, 0x00, 8, 4); + +/* reg_paos_oper_status + * Port operational state (the current state): + * 1 - Up. + * 2 - Down. + * 3 - Down by port failure. This means that the device will not let the + * port up again until explicitly specified by software. + * Access: RO + */ +MLXSW_ITEM32(reg, paos, oper_status, 0x00, 0, 4); + +/* reg_paos_ase + * Admin state update enabled. + * Access: WO + */ +MLXSW_ITEM32(reg, paos, ase, 0x04, 31, 1); + +/* reg_paos_ee + * Event update enable. If this bit is set, event generation will be + * updated based on the e field. + * Access: WO + */ +MLXSW_ITEM32(reg, paos, ee, 0x04, 30, 1); + +/* reg_paos_e + * Event generation on operational state change: + * 0 - Do not generate event. + * 1 - Generate Event. + * 2 - Generate Single Event. + * Access: RW + */ +MLXSW_ITEM32(reg, paos, e, 0x04, 0, 2); + +static inline void mlxsw_reg_paos_pack(char *payload, u16 local_port, + enum mlxsw_port_admin_status status) +{ + MLXSW_REG_ZERO(paos, payload); + mlxsw_reg_paos_swid_set(payload, 0); + mlxsw_reg_paos_local_port_set(payload, local_port); + mlxsw_reg_paos_admin_status_set(payload, status); + mlxsw_reg_paos_oper_status_set(payload, 0); + mlxsw_reg_paos_ase_set(payload, 1); + mlxsw_reg_paos_ee_set(payload, 1); + mlxsw_reg_paos_e_set(payload, 1); +} + +/* PFCC - Ports Flow Control Configuration Register + * ------------------------------------------------ + * Configures and retrieves the per port flow control configuration. + */ +#define MLXSW_REG_PFCC_ID 0x5007 +#define MLXSW_REG_PFCC_LEN 0x20 + +MLXSW_REG_DEFINE(pfcc, MLXSW_REG_PFCC_ID, MLXSW_REG_PFCC_LEN); + +/* reg_pfcc_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, pfcc, 0x00, 16, 0x00, 12); + +/* reg_pfcc_pnat + * Port number access type. Determines the way local_port is interpreted: + * 0 - Local port number. + * 1 - IB / label port number. + * Access: Index + */ +MLXSW_ITEM32(reg, pfcc, pnat, 0x00, 14, 2); + +/* reg_pfcc_shl_cap + * Send to higher layers capabilities: + * 0 - No capability of sending Pause and PFC frames to higher layers. + * 1 - Device has capability of sending Pause and PFC frames to higher + * layers. + * Access: RO + */ +MLXSW_ITEM32(reg, pfcc, shl_cap, 0x00, 1, 1); + +/* reg_pfcc_shl_opr + * Send to higher layers operation: + * 0 - Pause and PFC frames are handled by the port (default). + * 1 - Pause and PFC frames are handled by the port and also sent to + * higher layers. Only valid if shl_cap = 1. + * Access: RW + */ +MLXSW_ITEM32(reg, pfcc, shl_opr, 0x00, 0, 1); + +/* reg_pfcc_ppan + * Pause policy auto negotiation. + * 0 - Disabled. Generate / ignore Pause frames based on pptx / pprtx. + * 1 - Enabled. When auto-negotiation is performed, set the Pause policy + * based on the auto-negotiation resolution. + * Access: RW + * + * Note: The auto-negotiation advertisement is set according to pptx and + * pprtx. When PFC is set on Tx / Rx, ppan must be set to 0. + */ +MLXSW_ITEM32(reg, pfcc, ppan, 0x04, 28, 4); + +/* reg_pfcc_prio_mask_tx + * Bit per priority indicating if Tx flow control policy should be + * updated based on bit pfctx. + * Access: WO + */ +MLXSW_ITEM32(reg, pfcc, prio_mask_tx, 0x04, 16, 8); + +/* reg_pfcc_prio_mask_rx + * Bit per priority indicating if Rx flow control policy should be + * updated based on bit pfcrx. + * Access: WO + */ +MLXSW_ITEM32(reg, pfcc, prio_mask_rx, 0x04, 0, 8); + +/* reg_pfcc_pptx + * Admin Pause policy on Tx. + * 0 - Never generate Pause frames (default). + * 1 - Generate Pause frames according to Rx buffer threshold. + * Access: RW + */ +MLXSW_ITEM32(reg, pfcc, pptx, 0x08, 31, 1); + +/* reg_pfcc_aptx + * Active (operational) Pause policy on Tx. + * 0 - Never generate Pause frames. + * 1 - Generate Pause frames according to Rx buffer threshold. + * Access: RO + */ +MLXSW_ITEM32(reg, pfcc, aptx, 0x08, 30, 1); + +/* reg_pfcc_pfctx + * Priority based flow control policy on Tx[7:0]. Per-priority bit mask: + * 0 - Never generate priority Pause frames on the specified priority + * (default). + * 1 - Generate priority Pause frames according to Rx buffer threshold on + * the specified priority. + * Access: RW + * + * Note: pfctx and pptx must be mutually exclusive. + */ +MLXSW_ITEM32(reg, pfcc, pfctx, 0x08, 16, 8); + +/* reg_pfcc_pprx + * Admin Pause policy on Rx. + * 0 - Ignore received Pause frames (default). + * 1 - Respect received Pause frames. + * Access: RW + */ +MLXSW_ITEM32(reg, pfcc, pprx, 0x0C, 31, 1); + +/* reg_pfcc_aprx + * Active (operational) Pause policy on Rx. + * 0 - Ignore received Pause frames. + * 1 - Respect received Pause frames. + * Access: RO + */ +MLXSW_ITEM32(reg, pfcc, aprx, 0x0C, 30, 1); + +/* reg_pfcc_pfcrx + * Priority based flow control policy on Rx[7:0]. Per-priority bit mask: + * 0 - Ignore incoming priority Pause frames on the specified priority + * (default). + * 1 - Respect incoming priority Pause frames on the specified priority. + * Access: RW + */ +MLXSW_ITEM32(reg, pfcc, pfcrx, 0x0C, 16, 8); + +#define MLXSW_REG_PFCC_ALL_PRIO 0xFF + +static inline void mlxsw_reg_pfcc_prio_pack(char *payload, u8 pfc_en) +{ + mlxsw_reg_pfcc_prio_mask_tx_set(payload, MLXSW_REG_PFCC_ALL_PRIO); + mlxsw_reg_pfcc_prio_mask_rx_set(payload, MLXSW_REG_PFCC_ALL_PRIO); + mlxsw_reg_pfcc_pfctx_set(payload, pfc_en); + mlxsw_reg_pfcc_pfcrx_set(payload, pfc_en); +} + +static inline void mlxsw_reg_pfcc_pack(char *payload, u16 local_port) +{ + MLXSW_REG_ZERO(pfcc, payload); + mlxsw_reg_pfcc_local_port_set(payload, local_port); +} + +/* PPCNT - Ports Performance Counters Register + * ------------------------------------------- + * The PPCNT register retrieves per port performance counters. + */ +#define MLXSW_REG_PPCNT_ID 0x5008 +#define MLXSW_REG_PPCNT_LEN 0x100 +#define MLXSW_REG_PPCNT_COUNTERS_OFFSET 0x08 + +MLXSW_REG_DEFINE(ppcnt, MLXSW_REG_PPCNT_ID, MLXSW_REG_PPCNT_LEN); + +/* reg_ppcnt_swid + * For HCA: must be always 0. + * Switch partition ID to associate port with. + * Switch partitions are numbered from 0 to 7 inclusively. + * Switch partition 254 indicates stacking ports. + * Switch partition 255 indicates all switch partitions. + * Only valid on Set() operation with local_port=255. + * Access: Index + */ +MLXSW_ITEM32(reg, ppcnt, swid, 0x00, 24, 8); + +/* reg_ppcnt_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, ppcnt, 0x00, 16, 0x00, 12); + +/* reg_ppcnt_pnat + * Port number access type: + * 0 - Local port number + * 1 - IB port number + * Access: Index + */ +MLXSW_ITEM32(reg, ppcnt, pnat, 0x00, 14, 2); + +enum mlxsw_reg_ppcnt_grp { + MLXSW_REG_PPCNT_IEEE_8023_CNT = 0x0, + MLXSW_REG_PPCNT_RFC_2863_CNT = 0x1, + MLXSW_REG_PPCNT_RFC_2819_CNT = 0x2, + MLXSW_REG_PPCNT_RFC_3635_CNT = 0x3, + MLXSW_REG_PPCNT_EXT_CNT = 0x5, + MLXSW_REG_PPCNT_DISCARD_CNT = 0x6, + MLXSW_REG_PPCNT_PRIO_CNT = 0x10, + MLXSW_REG_PPCNT_TC_CNT = 0x11, + MLXSW_REG_PPCNT_TC_CONG_CNT = 0x13, +}; + +/* reg_ppcnt_grp + * Performance counter group. + * Group 63 indicates all groups. Only valid on Set() operation with + * clr bit set. + * 0x0: IEEE 802.3 Counters + * 0x1: RFC 2863 Counters + * 0x2: RFC 2819 Counters + * 0x3: RFC 3635 Counters + * 0x5: Ethernet Extended Counters + * 0x6: Ethernet Discard Counters + * 0x8: Link Level Retransmission Counters + * 0x10: Per Priority Counters + * 0x11: Per Traffic Class Counters + * 0x12: Physical Layer Counters + * 0x13: Per Traffic Class Congestion Counters + * Access: Index + */ +MLXSW_ITEM32(reg, ppcnt, grp, 0x00, 0, 6); + +/* reg_ppcnt_clr + * Clear counters. Setting the clr bit will reset the counter value + * for all counters in the counter group. This bit can be set + * for both Set() and Get() operation. + * Access: OP + */ +MLXSW_ITEM32(reg, ppcnt, clr, 0x04, 31, 1); + +/* reg_ppcnt_lp_gl + * Local port global variable. + * 0: local_port 255 = all ports of the device. + * 1: local_port indicates local port number for all ports. + * Access: OP + */ +MLXSW_ITEM32(reg, ppcnt, lp_gl, 0x04, 30, 1); + +/* reg_ppcnt_prio_tc + * Priority for counter set that support per priority, valid values: 0-7. + * Traffic class for counter set that support per traffic class, + * valid values: 0- cap_max_tclass-1 . + * For HCA: cap_max_tclass is always 8. + * Otherwise must be 0. + * Access: Index + */ +MLXSW_ITEM32(reg, ppcnt, prio_tc, 0x04, 0, 5); + +/* Ethernet IEEE 802.3 Counter Group */ + +/* reg_ppcnt_a_frames_transmitted_ok + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_frames_transmitted_ok, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x00, 0, 64); + +/* reg_ppcnt_a_frames_received_ok + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_frames_received_ok, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x08, 0, 64); + +/* reg_ppcnt_a_frame_check_sequence_errors + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_frame_check_sequence_errors, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x10, 0, 64); + +/* reg_ppcnt_a_alignment_errors + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_alignment_errors, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x18, 0, 64); + +/* reg_ppcnt_a_octets_transmitted_ok + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_octets_transmitted_ok, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x20, 0, 64); + +/* reg_ppcnt_a_octets_received_ok + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_octets_received_ok, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x28, 0, 64); + +/* reg_ppcnt_a_multicast_frames_xmitted_ok + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_multicast_frames_xmitted_ok, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x30, 0, 64); + +/* reg_ppcnt_a_broadcast_frames_xmitted_ok + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_broadcast_frames_xmitted_ok, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x38, 0, 64); + +/* reg_ppcnt_a_multicast_frames_received_ok + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_multicast_frames_received_ok, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x40, 0, 64); + +/* reg_ppcnt_a_broadcast_frames_received_ok + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_broadcast_frames_received_ok, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x48, 0, 64); + +/* reg_ppcnt_a_in_range_length_errors + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_in_range_length_errors, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x50, 0, 64); + +/* reg_ppcnt_a_out_of_range_length_field + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_out_of_range_length_field, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x58, 0, 64); + +/* reg_ppcnt_a_frame_too_long_errors + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_frame_too_long_errors, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x60, 0, 64); + +/* reg_ppcnt_a_symbol_error_during_carrier + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_symbol_error_during_carrier, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x68, 0, 64); + +/* reg_ppcnt_a_mac_control_frames_transmitted + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_mac_control_frames_transmitted, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x70, 0, 64); + +/* reg_ppcnt_a_mac_control_frames_received + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_mac_control_frames_received, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x78, 0, 64); + +/* reg_ppcnt_a_unsupported_opcodes_received + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_unsupported_opcodes_received, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x80, 0, 64); + +/* reg_ppcnt_a_pause_mac_ctrl_frames_received + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_pause_mac_ctrl_frames_received, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x88, 0, 64); + +/* reg_ppcnt_a_pause_mac_ctrl_frames_transmitted + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, a_pause_mac_ctrl_frames_transmitted, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x90, 0, 64); + +/* Ethernet RFC 2863 Counter Group */ + +/* reg_ppcnt_if_in_discards + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, if_in_discards, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x10, 0, 64); + +/* reg_ppcnt_if_out_discards + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, if_out_discards, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x38, 0, 64); + +/* reg_ppcnt_if_out_errors + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, if_out_errors, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x40, 0, 64); + +/* Ethernet RFC 2819 Counter Group */ + +/* reg_ppcnt_ether_stats_undersize_pkts + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_undersize_pkts, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x30, 0, 64); + +/* reg_ppcnt_ether_stats_oversize_pkts + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_oversize_pkts, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x38, 0, 64); + +/* reg_ppcnt_ether_stats_fragments + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_fragments, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x40, 0, 64); + +/* reg_ppcnt_ether_stats_pkts64octets + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_pkts64octets, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x58, 0, 64); + +/* reg_ppcnt_ether_stats_pkts65to127octets + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_pkts65to127octets, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x60, 0, 64); + +/* reg_ppcnt_ether_stats_pkts128to255octets + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_pkts128to255octets, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x68, 0, 64); + +/* reg_ppcnt_ether_stats_pkts256to511octets + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_pkts256to511octets, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x70, 0, 64); + +/* reg_ppcnt_ether_stats_pkts512to1023octets + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_pkts512to1023octets, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x78, 0, 64); + +/* reg_ppcnt_ether_stats_pkts1024to1518octets + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_pkts1024to1518octets, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x80, 0, 64); + +/* reg_ppcnt_ether_stats_pkts1519to2047octets + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_pkts1519to2047octets, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x88, 0, 64); + +/* reg_ppcnt_ether_stats_pkts2048to4095octets + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_pkts2048to4095octets, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x90, 0, 64); + +/* reg_ppcnt_ether_stats_pkts4096to8191octets + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_pkts4096to8191octets, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x98, 0, 64); + +/* reg_ppcnt_ether_stats_pkts8192to10239octets + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ether_stats_pkts8192to10239octets, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0xA0, 0, 64); + +/* Ethernet RFC 3635 Counter Group */ + +/* reg_ppcnt_dot3stats_fcs_errors + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, dot3stats_fcs_errors, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x08, 0, 64); + +/* reg_ppcnt_dot3stats_symbol_errors + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, dot3stats_symbol_errors, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x60, 0, 64); + +/* reg_ppcnt_dot3control_in_unknown_opcodes + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, dot3control_in_unknown_opcodes, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x68, 0, 64); + +/* reg_ppcnt_dot3in_pause_frames + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, dot3in_pause_frames, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x70, 0, 64); + +/* Ethernet Extended Counter Group Counters */ + +/* reg_ppcnt_ecn_marked + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ecn_marked, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x08, 0, 64); + +/* Ethernet Discard Counter Group Counters */ + +/* reg_ppcnt_ingress_general + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ingress_general, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x00, 0, 64); + +/* reg_ppcnt_ingress_policy_engine + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ingress_policy_engine, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x08, 0, 64); + +/* reg_ppcnt_ingress_vlan_membership + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ingress_vlan_membership, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x10, 0, 64); + +/* reg_ppcnt_ingress_tag_frame_type + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ingress_tag_frame_type, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x18, 0, 64); + +/* reg_ppcnt_egress_vlan_membership + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, egress_vlan_membership, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x20, 0, 64); + +/* reg_ppcnt_loopback_filter + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, loopback_filter, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x28, 0, 64); + +/* reg_ppcnt_egress_general + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, egress_general, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x30, 0, 64); + +/* reg_ppcnt_egress_hoq + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, egress_hoq, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x40, 0, 64); + +/* reg_ppcnt_egress_policy_engine + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, egress_policy_engine, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x50, 0, 64); + +/* reg_ppcnt_ingress_tx_link_down + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ingress_tx_link_down, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x58, 0, 64); + +/* reg_ppcnt_egress_stp_filter + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, egress_stp_filter, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x60, 0, 64); + +/* reg_ppcnt_egress_sll + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, egress_sll, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x70, 0, 64); + +/* Ethernet Per Priority Group Counters */ + +/* reg_ppcnt_rx_octets + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, rx_octets, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x00, 0, 64); + +/* reg_ppcnt_rx_frames + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, rx_frames, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x20, 0, 64); + +/* reg_ppcnt_tx_octets + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, tx_octets, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x28, 0, 64); + +/* reg_ppcnt_tx_frames + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, tx_frames, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x48, 0, 64); + +/* reg_ppcnt_rx_pause + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, rx_pause, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x50, 0, 64); + +/* reg_ppcnt_rx_pause_duration + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, rx_pause_duration, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x58, 0, 64); + +/* reg_ppcnt_tx_pause + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, tx_pause, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x60, 0, 64); + +/* reg_ppcnt_tx_pause_duration + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, tx_pause_duration, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x68, 0, 64); + +/* reg_ppcnt_rx_pause_transition + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, tx_pause_transition, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x70, 0, 64); + +/* Ethernet Per Traffic Class Counters */ + +/* reg_ppcnt_tc_transmit_queue + * Contains the transmit queue depth in cells of traffic class + * selected by prio_tc and the port selected by local_port. + * The field cannot be cleared. + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, tc_transmit_queue, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x00, 0, 64); + +/* reg_ppcnt_tc_no_buffer_discard_uc + * The number of unicast packets dropped due to lack of shared + * buffer resources. + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, tc_no_buffer_discard_uc, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x08, 0, 64); + +/* Ethernet Per Traffic Class Congestion Group Counters */ + +/* reg_ppcnt_wred_discard + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, wred_discard, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x00, 0, 64); + +/* reg_ppcnt_ecn_marked_tc + * Access: RO + */ +MLXSW_ITEM64(reg, ppcnt, ecn_marked_tc, + MLXSW_REG_PPCNT_COUNTERS_OFFSET + 0x08, 0, 64); + +static inline void mlxsw_reg_ppcnt_pack(char *payload, u16 local_port, + enum mlxsw_reg_ppcnt_grp grp, + u8 prio_tc) +{ + MLXSW_REG_ZERO(ppcnt, payload); + mlxsw_reg_ppcnt_swid_set(payload, 0); + mlxsw_reg_ppcnt_local_port_set(payload, local_port); + mlxsw_reg_ppcnt_pnat_set(payload, 0); + mlxsw_reg_ppcnt_grp_set(payload, grp); + mlxsw_reg_ppcnt_clr_set(payload, 0); + mlxsw_reg_ppcnt_lp_gl_set(payload, 1); + mlxsw_reg_ppcnt_prio_tc_set(payload, prio_tc); +} + +/* PPTB - Port Prio To Buffer Register + * ----------------------------------- + * Configures the switch priority to buffer table. + */ +#define MLXSW_REG_PPTB_ID 0x500B +#define MLXSW_REG_PPTB_LEN 0x10 + +MLXSW_REG_DEFINE(pptb, MLXSW_REG_PPTB_ID, MLXSW_REG_PPTB_LEN); + +enum { + MLXSW_REG_PPTB_MM_UM, + MLXSW_REG_PPTB_MM_UNICAST, + MLXSW_REG_PPTB_MM_MULTICAST, +}; + +/* reg_pptb_mm + * Mapping mode. + * 0 - Map both unicast and multicast packets to the same buffer. + * 1 - Map only unicast packets. + * 2 - Map only multicast packets. + * Access: Index + * + * Note: SwitchX-2 only supports the first option. + */ +MLXSW_ITEM32(reg, pptb, mm, 0x00, 28, 2); + +/* reg_pptb_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, pptb, 0x00, 16, 0x00, 12); + +/* reg_pptb_um + * Enables the update of the untagged_buf field. + * Access: RW + */ +MLXSW_ITEM32(reg, pptb, um, 0x00, 8, 1); + +/* reg_pptb_pm + * Enables the update of the prio_to_buff field. + * Bit <i> is a flag for updating the mapping for switch priority <i>. + * Access: RW + */ +MLXSW_ITEM32(reg, pptb, pm, 0x00, 0, 8); + +/* reg_pptb_prio_to_buff + * Mapping of switch priority <i> to one of the allocated receive port + * buffers. + * Access: RW + */ +MLXSW_ITEM_BIT_ARRAY(reg, pptb, prio_to_buff, 0x04, 0x04, 4); + +/* reg_pptb_pm_msb + * Enables the update of the prio_to_buff field. + * Bit <i> is a flag for updating the mapping for switch priority <i+8>. + * Access: RW + */ +MLXSW_ITEM32(reg, pptb, pm_msb, 0x08, 24, 8); + +/* reg_pptb_untagged_buff + * Mapping of untagged frames to one of the allocated receive port buffers. + * Access: RW + * + * Note: In SwitchX-2 this field must be mapped to buffer 8. Reserved for + * Spectrum, as it maps untagged packets based on the default switch priority. + */ +MLXSW_ITEM32(reg, pptb, untagged_buff, 0x08, 0, 4); + +/* reg_pptb_prio_to_buff_msb + * Mapping of switch priority <i+8> to one of the allocated receive port + * buffers. + * Access: RW + */ +MLXSW_ITEM_BIT_ARRAY(reg, pptb, prio_to_buff_msb, 0x0C, 0x04, 4); + +#define MLXSW_REG_PPTB_ALL_PRIO 0xFF + +static inline void mlxsw_reg_pptb_pack(char *payload, u16 local_port) +{ + MLXSW_REG_ZERO(pptb, payload); + mlxsw_reg_pptb_mm_set(payload, MLXSW_REG_PPTB_MM_UM); + mlxsw_reg_pptb_local_port_set(payload, local_port); + mlxsw_reg_pptb_pm_set(payload, MLXSW_REG_PPTB_ALL_PRIO); + mlxsw_reg_pptb_pm_msb_set(payload, MLXSW_REG_PPTB_ALL_PRIO); +} + +static inline void mlxsw_reg_pptb_prio_to_buff_pack(char *payload, u8 prio, + u8 buff) +{ + mlxsw_reg_pptb_prio_to_buff_set(payload, prio, buff); + mlxsw_reg_pptb_prio_to_buff_msb_set(payload, prio, buff); +} + +/* PBMC - Port Buffer Management Control Register + * ---------------------------------------------- + * The PBMC register configures and retrieves the port packet buffer + * allocation for different Prios, and the Pause threshold management. + */ +#define MLXSW_REG_PBMC_ID 0x500C +#define MLXSW_REG_PBMC_LEN 0x6C + +MLXSW_REG_DEFINE(pbmc, MLXSW_REG_PBMC_ID, MLXSW_REG_PBMC_LEN); + +/* reg_pbmc_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, pbmc, 0x00, 16, 0x00, 12); + +/* reg_pbmc_xoff_timer_value + * When device generates a pause frame, it uses this value as the pause + * timer (time for the peer port to pause in quota-512 bit time). + * Access: RW + */ +MLXSW_ITEM32(reg, pbmc, xoff_timer_value, 0x04, 16, 16); + +/* reg_pbmc_xoff_refresh + * The time before a new pause frame should be sent to refresh the pause RW + * state. Using the same units as xoff_timer_value above (in quota-512 bit + * time). + * Access: RW + */ +MLXSW_ITEM32(reg, pbmc, xoff_refresh, 0x04, 0, 16); + +#define MLXSW_REG_PBMC_PORT_SHARED_BUF_IDX 11 + +/* reg_pbmc_buf_lossy + * The field indicates if the buffer is lossy. + * 0 - Lossless + * 1 - Lossy + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, pbmc, buf_lossy, 0x0C, 25, 1, 0x08, 0x00, false); + +/* reg_pbmc_buf_epsb + * Eligible for Port Shared buffer. + * If epsb is set, packets assigned to buffer are allowed to insert the port + * shared buffer. + * When buf_lossy is MLXSW_REG_PBMC_LOSSY_LOSSY this field is reserved. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, pbmc, buf_epsb, 0x0C, 24, 1, 0x08, 0x00, false); + +/* reg_pbmc_buf_size + * The part of the packet buffer array is allocated for the specific buffer. + * Units are represented in cells. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, pbmc, buf_size, 0x0C, 0, 16, 0x08, 0x00, false); + +/* reg_pbmc_buf_xoff_threshold + * Once the amount of data in the buffer goes above this value, device + * starts sending PFC frames for all priorities associated with the + * buffer. Units are represented in cells. Reserved in case of lossy + * buffer. + * Access: RW + * + * Note: In Spectrum, reserved for buffer[9]. + */ +MLXSW_ITEM32_INDEXED(reg, pbmc, buf_xoff_threshold, 0x0C, 16, 16, + 0x08, 0x04, false); + +/* reg_pbmc_buf_xon_threshold + * When the amount of data in the buffer goes below this value, device + * stops sending PFC frames for the priorities associated with the + * buffer. Units are represented in cells. Reserved in case of lossy + * buffer. + * Access: RW + * + * Note: In Spectrum, reserved for buffer[9]. + */ +MLXSW_ITEM32_INDEXED(reg, pbmc, buf_xon_threshold, 0x0C, 0, 16, + 0x08, 0x04, false); + +static inline void mlxsw_reg_pbmc_pack(char *payload, u16 local_port, + u16 xoff_timer_value, u16 xoff_refresh) +{ + MLXSW_REG_ZERO(pbmc, payload); + mlxsw_reg_pbmc_local_port_set(payload, local_port); + mlxsw_reg_pbmc_xoff_timer_value_set(payload, xoff_timer_value); + mlxsw_reg_pbmc_xoff_refresh_set(payload, xoff_refresh); +} + +static inline void mlxsw_reg_pbmc_lossy_buffer_pack(char *payload, + int buf_index, + u16 size) +{ + mlxsw_reg_pbmc_buf_lossy_set(payload, buf_index, 1); + mlxsw_reg_pbmc_buf_epsb_set(payload, buf_index, 0); + mlxsw_reg_pbmc_buf_size_set(payload, buf_index, size); +} + +static inline void mlxsw_reg_pbmc_lossless_buffer_pack(char *payload, + int buf_index, u16 size, + u16 threshold) +{ + mlxsw_reg_pbmc_buf_lossy_set(payload, buf_index, 0); + mlxsw_reg_pbmc_buf_epsb_set(payload, buf_index, 0); + mlxsw_reg_pbmc_buf_size_set(payload, buf_index, size); + mlxsw_reg_pbmc_buf_xoff_threshold_set(payload, buf_index, threshold); + mlxsw_reg_pbmc_buf_xon_threshold_set(payload, buf_index, threshold); +} + +/* PSPA - Port Switch Partition Allocation + * --------------------------------------- + * Controls the association of a port with a switch partition and enables + * configuring ports as stacking ports. + */ +#define MLXSW_REG_PSPA_ID 0x500D +#define MLXSW_REG_PSPA_LEN 0x8 + +MLXSW_REG_DEFINE(pspa, MLXSW_REG_PSPA_ID, MLXSW_REG_PSPA_LEN); + +/* reg_pspa_swid + * Switch partition ID. + * Access: RW + */ +MLXSW_ITEM32(reg, pspa, swid, 0x00, 24, 8); + +/* reg_pspa_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, pspa, 0x00, 16, 0x00, 0); + +/* reg_pspa_sub_port + * Virtual port within the local port. Set to 0 when virtual ports are + * disabled on the local port. + * Access: Index + */ +MLXSW_ITEM32(reg, pspa, sub_port, 0x00, 8, 8); + +static inline void mlxsw_reg_pspa_pack(char *payload, u8 swid, u16 local_port) +{ + MLXSW_REG_ZERO(pspa, payload); + mlxsw_reg_pspa_swid_set(payload, swid); + mlxsw_reg_pspa_local_port_set(payload, local_port); + mlxsw_reg_pspa_sub_port_set(payload, 0); +} + +/* PMAOS - Ports Module Administrative and Operational Status + * ---------------------------------------------------------- + * This register configures and retrieves the per module status. + */ +#define MLXSW_REG_PMAOS_ID 0x5012 +#define MLXSW_REG_PMAOS_LEN 0x10 + +MLXSW_REG_DEFINE(pmaos, MLXSW_REG_PMAOS_ID, MLXSW_REG_PMAOS_LEN); + +/* reg_pmaos_rst + * Module reset toggle. + * Note: Setting reset while module is plugged-in will result in transition to + * "initializing" operational state. + * Access: OP + */ +MLXSW_ITEM32(reg, pmaos, rst, 0x00, 31, 1); + +/* reg_pmaos_slot_index + * Slot index. + * Access: Index + */ +MLXSW_ITEM32(reg, pmaos, slot_index, 0x00, 24, 4); + +/* reg_pmaos_module + * Module number. + * Access: Index + */ +MLXSW_ITEM32(reg, pmaos, module, 0x00, 16, 8); + +enum mlxsw_reg_pmaos_admin_status { + MLXSW_REG_PMAOS_ADMIN_STATUS_ENABLED = 1, + MLXSW_REG_PMAOS_ADMIN_STATUS_DISABLED = 2, + /* If the module is active and then unplugged, or experienced an error + * event, the operational status should go to "disabled" and can only + * be enabled upon explicit enable command. + */ + MLXSW_REG_PMAOS_ADMIN_STATUS_ENABLED_ONCE = 3, +}; + +/* reg_pmaos_admin_status + * Module administrative state (the desired state of the module). + * Note: To disable a module, all ports associated with the port must be + * administatively down first. + * Access: RW + */ +MLXSW_ITEM32(reg, pmaos, admin_status, 0x00, 8, 4); + +/* reg_pmaos_ase + * Admin state update enable. + * If this bit is set, admin state will be updated based on admin_state field. + * Only relevant on Set() operations. + * Access: WO + */ +MLXSW_ITEM32(reg, pmaos, ase, 0x04, 31, 1); + +/* reg_pmaos_ee + * Event update enable. + * If this bit is set, event generation will be updated based on the e field. + * Only relevant on Set operations. + * Access: WO + */ +MLXSW_ITEM32(reg, pmaos, ee, 0x04, 30, 1); + +enum mlxsw_reg_pmaos_e { + MLXSW_REG_PMAOS_E_DO_NOT_GENERATE_EVENT, + MLXSW_REG_PMAOS_E_GENERATE_EVENT, + MLXSW_REG_PMAOS_E_GENERATE_SINGLE_EVENT, +}; + +/* reg_pmaos_e + * Event Generation on operational state change. + * Access: RW + */ +MLXSW_ITEM32(reg, pmaos, e, 0x04, 0, 2); + +static inline void mlxsw_reg_pmaos_pack(char *payload, u8 slot_index, u8 module) +{ + MLXSW_REG_ZERO(pmaos, payload); + mlxsw_reg_pmaos_slot_index_set(payload, slot_index); + mlxsw_reg_pmaos_module_set(payload, module); +} + +/* PPLR - Port Physical Loopback Register + * -------------------------------------- + * This register allows configuration of the port's loopback mode. + */ +#define MLXSW_REG_PPLR_ID 0x5018 +#define MLXSW_REG_PPLR_LEN 0x8 + +MLXSW_REG_DEFINE(pplr, MLXSW_REG_PPLR_ID, MLXSW_REG_PPLR_LEN); + +/* reg_pplr_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, pplr, 0x00, 16, 0x00, 12); + +/* Phy local loopback. When set the port's egress traffic is looped back + * to the receiver and the port transmitter is disabled. + */ +#define MLXSW_REG_PPLR_LB_TYPE_BIT_PHY_LOCAL BIT(1) + +/* reg_pplr_lb_en + * Loopback enable. + * Access: RW + */ +MLXSW_ITEM32(reg, pplr, lb_en, 0x04, 0, 8); + +static inline void mlxsw_reg_pplr_pack(char *payload, u16 local_port, + bool phy_local) +{ + MLXSW_REG_ZERO(pplr, payload); + mlxsw_reg_pplr_local_port_set(payload, local_port); + mlxsw_reg_pplr_lb_en_set(payload, + phy_local ? + MLXSW_REG_PPLR_LB_TYPE_BIT_PHY_LOCAL : 0); +} + +/* PMTDB - Port Module To local DataBase Register + * ---------------------------------------------- + * The PMTDB register allows to query the possible module<->local port + * mapping than can be used in PMLP. It does not represent the actual/current + * mapping of the local to module. Actual mapping is only defined by PMLP. + */ +#define MLXSW_REG_PMTDB_ID 0x501A +#define MLXSW_REG_PMTDB_LEN 0x40 + +MLXSW_REG_DEFINE(pmtdb, MLXSW_REG_PMTDB_ID, MLXSW_REG_PMTDB_LEN); + +/* reg_pmtdb_slot_index + * Slot index (0: Main board). + * Access: Index + */ +MLXSW_ITEM32(reg, pmtdb, slot_index, 0x00, 24, 4); + +/* reg_pmtdb_module + * Module number. + * Access: Index + */ +MLXSW_ITEM32(reg, pmtdb, module, 0x00, 16, 8); + +/* reg_pmtdb_ports_width + * Port's width + * Access: Index + */ +MLXSW_ITEM32(reg, pmtdb, ports_width, 0x00, 12, 4); + +/* reg_pmtdb_num_ports + * Number of ports in a single module (split/breakout) + * Access: Index + */ +MLXSW_ITEM32(reg, pmtdb, num_ports, 0x00, 8, 4); + +enum mlxsw_reg_pmtdb_status { + MLXSW_REG_PMTDB_STATUS_SUCCESS, +}; + +/* reg_pmtdb_status + * Status + * Access: RO + */ +MLXSW_ITEM32(reg, pmtdb, status, 0x00, 0, 4); + +/* reg_pmtdb_port_num + * The local_port value which can be assigned to the module. + * In case of more than one port, port<x> represent the /<x> port of + * the module. + * Access: RO + */ +MLXSW_ITEM16_INDEXED(reg, pmtdb, port_num, 0x04, 0, 10, 0x02, 0x00, false); + +static inline void mlxsw_reg_pmtdb_pack(char *payload, u8 slot_index, u8 module, + u8 ports_width, u8 num_ports) +{ + MLXSW_REG_ZERO(pmtdb, payload); + mlxsw_reg_pmtdb_slot_index_set(payload, slot_index); + mlxsw_reg_pmtdb_module_set(payload, module); + mlxsw_reg_pmtdb_ports_width_set(payload, ports_width); + mlxsw_reg_pmtdb_num_ports_set(payload, num_ports); +} + +/* PMECR - Ports Mapping Event Configuration Register + * -------------------------------------------------- + * The PMECR register is used to enable/disable event triggering + * in case of local port mapping change. + */ +#define MLXSW_REG_PMECR_ID 0x501B +#define MLXSW_REG_PMECR_LEN 0x20 + +MLXSW_REG_DEFINE(pmecr, MLXSW_REG_PMECR_ID, MLXSW_REG_PMECR_LEN); + +/* reg_pmecr_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, pmecr, 0x00, 16, 0x00, 12); + +/* reg_pmecr_ee + * Event update enable. If this bit is set, event generation will be updated + * based on the e field. Only relevant on Set operations. + * Access: WO + */ +MLXSW_ITEM32(reg, pmecr, ee, 0x04, 30, 1); + +/* reg_pmecr_eswi + * Software ignore enable bit. If this bit is set, the value of swi is used. + * If this bit is clear, the value of swi is ignored. + * Only relevant on Set operations. + * Access: WO + */ +MLXSW_ITEM32(reg, pmecr, eswi, 0x04, 24, 1); + +/* reg_pmecr_swi + * Software ignore. If this bit is set, the device shouldn't generate events + * in case of PMLP SET operation but only upon self local port mapping change + * (if applicable according to e configuration). This is supplementary + * configuration on top of e value. + * Access: RW + */ +MLXSW_ITEM32(reg, pmecr, swi, 0x04, 8, 1); + +enum mlxsw_reg_pmecr_e { + MLXSW_REG_PMECR_E_DO_NOT_GENERATE_EVENT, + MLXSW_REG_PMECR_E_GENERATE_EVENT, + MLXSW_REG_PMECR_E_GENERATE_SINGLE_EVENT, +}; + +/* reg_pmecr_e + * Event generation on local port mapping change. + * Access: RW + */ +MLXSW_ITEM32(reg, pmecr, e, 0x04, 0, 2); + +static inline void mlxsw_reg_pmecr_pack(char *payload, u16 local_port, + enum mlxsw_reg_pmecr_e e) +{ + MLXSW_REG_ZERO(pmecr, payload); + mlxsw_reg_pmecr_local_port_set(payload, local_port); + mlxsw_reg_pmecr_e_set(payload, e); + mlxsw_reg_pmecr_ee_set(payload, true); + mlxsw_reg_pmecr_swi_set(payload, true); + mlxsw_reg_pmecr_eswi_set(payload, true); +} + +/* PMPE - Port Module Plug/Unplug Event Register + * --------------------------------------------- + * This register reports any operational status change of a module. + * A change in the module’s state will generate an event only if the change + * happens after arming the event mechanism. Any changes to the module state + * while the event mechanism is not armed will not be reported. Software can + * query the PMPE register for module status. + */ +#define MLXSW_REG_PMPE_ID 0x5024 +#define MLXSW_REG_PMPE_LEN 0x10 + +MLXSW_REG_DEFINE(pmpe, MLXSW_REG_PMPE_ID, MLXSW_REG_PMPE_LEN); + +/* reg_pmpe_slot_index + * Slot index. + * Access: Index + */ +MLXSW_ITEM32(reg, pmpe, slot_index, 0x00, 24, 4); + +/* reg_pmpe_module + * Module number. + * Access: Index + */ +MLXSW_ITEM32(reg, pmpe, module, 0x00, 16, 8); + +enum mlxsw_reg_pmpe_module_status { + MLXSW_REG_PMPE_MODULE_STATUS_PLUGGED_ENABLED = 1, + MLXSW_REG_PMPE_MODULE_STATUS_UNPLUGGED, + MLXSW_REG_PMPE_MODULE_STATUS_PLUGGED_ERROR, + MLXSW_REG_PMPE_MODULE_STATUS_PLUGGED_DISABLED, +}; + +/* reg_pmpe_module_status + * Module status. + * Access: RO + */ +MLXSW_ITEM32(reg, pmpe, module_status, 0x00, 0, 4); + +/* reg_pmpe_error_type + * Module error details. + * Access: RO + */ +MLXSW_ITEM32(reg, pmpe, error_type, 0x04, 8, 4); + +/* PDDR - Port Diagnostics Database Register + * ----------------------------------------- + * The PDDR enables to read the Phy debug database + */ +#define MLXSW_REG_PDDR_ID 0x5031 +#define MLXSW_REG_PDDR_LEN 0x100 + +MLXSW_REG_DEFINE(pddr, MLXSW_REG_PDDR_ID, MLXSW_REG_PDDR_LEN); + +/* reg_pddr_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, pddr, 0x00, 16, 0x00, 12); + +enum mlxsw_reg_pddr_page_select { + MLXSW_REG_PDDR_PAGE_SELECT_TROUBLESHOOTING_INFO = 1, +}; + +/* reg_pddr_page_select + * Page select index. + * Access: Index + */ +MLXSW_ITEM32(reg, pddr, page_select, 0x04, 0, 8); + +enum mlxsw_reg_pddr_trblsh_group_opcode { + /* Monitor opcodes */ + MLXSW_REG_PDDR_TRBLSH_GROUP_OPCODE_MONITOR, +}; + +/* reg_pddr_group_opcode + * Group selector. + * Access: Index + */ +MLXSW_ITEM32(reg, pddr, trblsh_group_opcode, 0x08, 0, 16); + +/* reg_pddr_status_opcode + * Group selector. + * Access: RO + */ +MLXSW_ITEM32(reg, pddr, trblsh_status_opcode, 0x0C, 0, 16); + +static inline void mlxsw_reg_pddr_pack(char *payload, u16 local_port, + u8 page_select) +{ + MLXSW_REG_ZERO(pddr, payload); + mlxsw_reg_pddr_local_port_set(payload, local_port); + mlxsw_reg_pddr_page_select_set(payload, page_select); +} + +/* PMMP - Port Module Memory Map Properties Register + * ------------------------------------------------- + * The PMMP register allows to override the module memory map advertisement. + * The register can only be set when the module is disabled by PMAOS register. + */ +#define MLXSW_REG_PMMP_ID 0x5044 +#define MLXSW_REG_PMMP_LEN 0x2C + +MLXSW_REG_DEFINE(pmmp, MLXSW_REG_PMMP_ID, MLXSW_REG_PMMP_LEN); + +/* reg_pmmp_module + * Module number. + * Access: Index + */ +MLXSW_ITEM32(reg, pmmp, module, 0x00, 16, 8); + +/* reg_pmmp_slot_index + * Slot index. + * Access: Index + */ +MLXSW_ITEM32(reg, pmmp, slot_index, 0x00, 24, 4); + +/* reg_pmmp_sticky + * When set, will keep eeprom_override values after plug-out event. + * Access: OP + */ +MLXSW_ITEM32(reg, pmmp, sticky, 0x00, 0, 1); + +/* reg_pmmp_eeprom_override_mask + * Write mask bit (negative polarity). + * 0 - Allow write + * 1 - Ignore write + * On write, indicates which of the bits from eeprom_override field are + * updated. + * Access: WO + */ +MLXSW_ITEM32(reg, pmmp, eeprom_override_mask, 0x04, 16, 16); + +enum { + /* Set module to low power mode */ + MLXSW_REG_PMMP_EEPROM_OVERRIDE_LOW_POWER_MASK = BIT(8), +}; + +/* reg_pmmp_eeprom_override + * Override / ignore EEPROM advertisement properties bitmask + * Access: RW + */ +MLXSW_ITEM32(reg, pmmp, eeprom_override, 0x04, 0, 16); + +static inline void mlxsw_reg_pmmp_pack(char *payload, u8 slot_index, u8 module) +{ + MLXSW_REG_ZERO(pmmp, payload); + mlxsw_reg_pmmp_slot_index_set(payload, slot_index); + mlxsw_reg_pmmp_module_set(payload, module); +} + +/* PLLP - Port Local port to Label Port mapping Register + * ----------------------------------------------------- + * The PLLP register returns the mapping from Local Port into Label Port. + */ +#define MLXSW_REG_PLLP_ID 0x504A +#define MLXSW_REG_PLLP_LEN 0x10 + +MLXSW_REG_DEFINE(pllp, MLXSW_REG_PLLP_ID, MLXSW_REG_PLLP_LEN); + +/* reg_pllp_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, pllp, 0x00, 16, 0x00, 12); + +/* reg_pllp_label_port + * Front panel label of the port. + * Access: RO + */ +MLXSW_ITEM32(reg, pllp, label_port, 0x00, 0, 8); + +/* reg_pllp_split_num + * Label split mapping for local_port. + * Access: RO + */ +MLXSW_ITEM32(reg, pllp, split_num, 0x04, 0, 4); + +/* reg_pllp_slot_index + * Slot index (0: Main board). + * Access: RO + */ +MLXSW_ITEM32(reg, pllp, slot_index, 0x08, 0, 4); + +static inline void mlxsw_reg_pllp_pack(char *payload, u16 local_port) +{ + MLXSW_REG_ZERO(pllp, payload); + mlxsw_reg_pllp_local_port_set(payload, local_port); +} + +static inline void mlxsw_reg_pllp_unpack(char *payload, u8 *label_port, + u8 *split_num, u8 *slot_index) +{ + *label_port = mlxsw_reg_pllp_label_port_get(payload); + *split_num = mlxsw_reg_pllp_split_num_get(payload); + *slot_index = mlxsw_reg_pllp_slot_index_get(payload); +} + +/* PMTM - Port Module Type Mapping Register + * ---------------------------------------- + * The PMTM register allows query or configuration of module types. + * The register can only be set when the module is disabled by PMAOS register + */ +#define MLXSW_REG_PMTM_ID 0x5067 +#define MLXSW_REG_PMTM_LEN 0x10 + +MLXSW_REG_DEFINE(pmtm, MLXSW_REG_PMTM_ID, MLXSW_REG_PMTM_LEN); + +/* reg_pmtm_slot_index + * Slot index. + * Access: Index + */ +MLXSW_ITEM32(reg, pmtm, slot_index, 0x00, 24, 4); + +/* reg_pmtm_module + * Module number. + * Access: Index + */ +MLXSW_ITEM32(reg, pmtm, module, 0x00, 16, 8); + +enum mlxsw_reg_pmtm_module_type { + MLXSW_REG_PMTM_MODULE_TYPE_BACKPLANE_4_LANES = 0, + MLXSW_REG_PMTM_MODULE_TYPE_QSFP = 1, + MLXSW_REG_PMTM_MODULE_TYPE_SFP = 2, + MLXSW_REG_PMTM_MODULE_TYPE_BACKPLANE_SINGLE_LANE = 4, + MLXSW_REG_PMTM_MODULE_TYPE_BACKPLANE_2_LANES = 8, + MLXSW_REG_PMTM_MODULE_TYPE_CHIP2CHIP4X = 10, + MLXSW_REG_PMTM_MODULE_TYPE_CHIP2CHIP2X = 11, + MLXSW_REG_PMTM_MODULE_TYPE_CHIP2CHIP1X = 12, + MLXSW_REG_PMTM_MODULE_TYPE_QSFP_DD = 14, + MLXSW_REG_PMTM_MODULE_TYPE_OSFP = 15, + MLXSW_REG_PMTM_MODULE_TYPE_SFP_DD = 16, + MLXSW_REG_PMTM_MODULE_TYPE_DSFP = 17, + MLXSW_REG_PMTM_MODULE_TYPE_CHIP2CHIP8X = 18, + MLXSW_REG_PMTM_MODULE_TYPE_TWISTED_PAIR = 19, +}; + +/* reg_pmtm_module_type + * Module type. + * Access: RW + */ +MLXSW_ITEM32(reg, pmtm, module_type, 0x04, 0, 5); + +static inline void mlxsw_reg_pmtm_pack(char *payload, u8 slot_index, u8 module) +{ + MLXSW_REG_ZERO(pmtm, payload); + mlxsw_reg_pmtm_slot_index_set(payload, slot_index); + mlxsw_reg_pmtm_module_set(payload, module); +} + +/* HTGT - Host Trap Group Table + * ---------------------------- + * Configures the properties for forwarding to CPU. + */ +#define MLXSW_REG_HTGT_ID 0x7002 +#define MLXSW_REG_HTGT_LEN 0x20 + +MLXSW_REG_DEFINE(htgt, MLXSW_REG_HTGT_ID, MLXSW_REG_HTGT_LEN); + +/* reg_htgt_swid + * Switch partition ID. + * Access: Index + */ +MLXSW_ITEM32(reg, htgt, swid, 0x00, 24, 8); + +#define MLXSW_REG_HTGT_PATH_TYPE_LOCAL 0x0 /* For locally attached CPU */ + +/* reg_htgt_type + * CPU path type. + * Access: RW + */ +MLXSW_ITEM32(reg, htgt, type, 0x00, 8, 4); + +enum mlxsw_reg_htgt_trap_group { + MLXSW_REG_HTGT_TRAP_GROUP_EMAD, + MLXSW_REG_HTGT_TRAP_GROUP_CORE_EVENT, + MLXSW_REG_HTGT_TRAP_GROUP_SP_STP, + MLXSW_REG_HTGT_TRAP_GROUP_SP_LACP, + MLXSW_REG_HTGT_TRAP_GROUP_SP_LLDP, + MLXSW_REG_HTGT_TRAP_GROUP_SP_MC_SNOOPING, + MLXSW_REG_HTGT_TRAP_GROUP_SP_BGP, + MLXSW_REG_HTGT_TRAP_GROUP_SP_OSPF, + MLXSW_REG_HTGT_TRAP_GROUP_SP_PIM, + MLXSW_REG_HTGT_TRAP_GROUP_SP_MULTICAST, + MLXSW_REG_HTGT_TRAP_GROUP_SP_NEIGH_DISCOVERY, + MLXSW_REG_HTGT_TRAP_GROUP_SP_ROUTER_EXP, + MLXSW_REG_HTGT_TRAP_GROUP_SP_EXTERNAL_ROUTE, + MLXSW_REG_HTGT_TRAP_GROUP_SP_IP2ME, + MLXSW_REG_HTGT_TRAP_GROUP_SP_DHCP, + MLXSW_REG_HTGT_TRAP_GROUP_SP_EVENT, + MLXSW_REG_HTGT_TRAP_GROUP_SP_IPV6, + MLXSW_REG_HTGT_TRAP_GROUP_SP_LBERROR, + MLXSW_REG_HTGT_TRAP_GROUP_SP_PTP0, + MLXSW_REG_HTGT_TRAP_GROUP_SP_PTP1, + MLXSW_REG_HTGT_TRAP_GROUP_SP_VRRP, + MLXSW_REG_HTGT_TRAP_GROUP_SP_PKT_SAMPLE, + MLXSW_REG_HTGT_TRAP_GROUP_SP_FLOW_LOGGING, + MLXSW_REG_HTGT_TRAP_GROUP_SP_FID_MISS, + MLXSW_REG_HTGT_TRAP_GROUP_SP_BFD, + MLXSW_REG_HTGT_TRAP_GROUP_SP_DUMMY, + MLXSW_REG_HTGT_TRAP_GROUP_SP_L2_DISCARDS, + MLXSW_REG_HTGT_TRAP_GROUP_SP_L3_DISCARDS, + MLXSW_REG_HTGT_TRAP_GROUP_SP_L3_EXCEPTIONS, + MLXSW_REG_HTGT_TRAP_GROUP_SP_TUNNEL_DISCARDS, + MLXSW_REG_HTGT_TRAP_GROUP_SP_ACL_DISCARDS, + MLXSW_REG_HTGT_TRAP_GROUP_SP_BUFFER_DISCARDS, + + __MLXSW_REG_HTGT_TRAP_GROUP_MAX, + MLXSW_REG_HTGT_TRAP_GROUP_MAX = __MLXSW_REG_HTGT_TRAP_GROUP_MAX - 1 +}; + +/* reg_htgt_trap_group + * Trap group number. User defined number specifying which trap groups + * should be forwarded to the CPU. The mapping between trap IDs and trap + * groups is configured using HPKT register. + * Access: Index + */ +MLXSW_ITEM32(reg, htgt, trap_group, 0x00, 0, 8); + +enum { + MLXSW_REG_HTGT_POLICER_DISABLE, + MLXSW_REG_HTGT_POLICER_ENABLE, +}; + +/* reg_htgt_pide + * Enable policer ID specified using 'pid' field. + * Access: RW + */ +MLXSW_ITEM32(reg, htgt, pide, 0x04, 15, 1); + +#define MLXSW_REG_HTGT_INVALID_POLICER 0xff + +/* reg_htgt_pid + * Policer ID for the trap group. + * Access: RW + */ +MLXSW_ITEM32(reg, htgt, pid, 0x04, 0, 8); + +#define MLXSW_REG_HTGT_TRAP_TO_CPU 0x0 + +/* reg_htgt_mirror_action + * Mirror action to use. + * 0 - Trap to CPU. + * 1 - Trap to CPU and mirror to a mirroring agent. + * 2 - Mirror to a mirroring agent and do not trap to CPU. + * Access: RW + * + * Note: Mirroring to a mirroring agent is only supported in Spectrum. + */ +MLXSW_ITEM32(reg, htgt, mirror_action, 0x08, 8, 2); + +/* reg_htgt_mirroring_agent + * Mirroring agent. + * Access: RW + */ +MLXSW_ITEM32(reg, htgt, mirroring_agent, 0x08, 0, 3); + +#define MLXSW_REG_HTGT_DEFAULT_PRIORITY 0 + +/* reg_htgt_priority + * Trap group priority. + * In case a packet matches multiple classification rules, the packet will + * only be trapped once, based on the trap ID associated with the group (via + * register HPKT) with the highest priority. + * Supported values are 0-7, with 7 represnting the highest priority. + * Access: RW + * + * Note: In SwitchX-2 this field is ignored and the priority value is replaced + * by the 'trap_group' field. + */ +MLXSW_ITEM32(reg, htgt, priority, 0x0C, 0, 4); + +#define MLXSW_REG_HTGT_DEFAULT_TC 7 + +/* reg_htgt_local_path_cpu_tclass + * CPU ingress traffic class for the trap group. + * Access: RW + */ +MLXSW_ITEM32(reg, htgt, local_path_cpu_tclass, 0x10, 16, 6); + +enum mlxsw_reg_htgt_local_path_rdq { + MLXSW_REG_HTGT_LOCAL_PATH_RDQ_SX2_CTRL = 0x13, + MLXSW_REG_HTGT_LOCAL_PATH_RDQ_SX2_RX = 0x14, + MLXSW_REG_HTGT_LOCAL_PATH_RDQ_SX2_EMAD = 0x15, + MLXSW_REG_HTGT_LOCAL_PATH_RDQ_SIB_EMAD = 0x15, +}; +/* reg_htgt_local_path_rdq + * Receive descriptor queue (RDQ) to use for the trap group. + * Access: RW + */ +MLXSW_ITEM32(reg, htgt, local_path_rdq, 0x10, 0, 6); + +static inline void mlxsw_reg_htgt_pack(char *payload, u8 group, u8 policer_id, + u8 priority, u8 tc) +{ + MLXSW_REG_ZERO(htgt, payload); + + if (policer_id == MLXSW_REG_HTGT_INVALID_POLICER) { + mlxsw_reg_htgt_pide_set(payload, + MLXSW_REG_HTGT_POLICER_DISABLE); + } else { + mlxsw_reg_htgt_pide_set(payload, + MLXSW_REG_HTGT_POLICER_ENABLE); + mlxsw_reg_htgt_pid_set(payload, policer_id); + } + + mlxsw_reg_htgt_type_set(payload, MLXSW_REG_HTGT_PATH_TYPE_LOCAL); + mlxsw_reg_htgt_trap_group_set(payload, group); + mlxsw_reg_htgt_mirror_action_set(payload, MLXSW_REG_HTGT_TRAP_TO_CPU); + mlxsw_reg_htgt_mirroring_agent_set(payload, 0); + mlxsw_reg_htgt_priority_set(payload, priority); + mlxsw_reg_htgt_local_path_cpu_tclass_set(payload, tc); + mlxsw_reg_htgt_local_path_rdq_set(payload, group); +} + +/* HPKT - Host Packet Trap + * ----------------------- + * Configures trap IDs inside trap groups. + */ +#define MLXSW_REG_HPKT_ID 0x7003 +#define MLXSW_REG_HPKT_LEN 0x10 + +MLXSW_REG_DEFINE(hpkt, MLXSW_REG_HPKT_ID, MLXSW_REG_HPKT_LEN); + +enum { + MLXSW_REG_HPKT_ACK_NOT_REQUIRED, + MLXSW_REG_HPKT_ACK_REQUIRED, +}; + +/* reg_hpkt_ack + * Require acknowledgements from the host for events. + * If set, then the device will wait for the event it sent to be acknowledged + * by the host. This option is only relevant for event trap IDs. + * Access: RW + * + * Note: Currently not supported by firmware. + */ +MLXSW_ITEM32(reg, hpkt, ack, 0x00, 24, 1); + +enum mlxsw_reg_hpkt_action { + MLXSW_REG_HPKT_ACTION_FORWARD, + MLXSW_REG_HPKT_ACTION_TRAP_TO_CPU, + MLXSW_REG_HPKT_ACTION_MIRROR_TO_CPU, + MLXSW_REG_HPKT_ACTION_DISCARD, + MLXSW_REG_HPKT_ACTION_SOFT_DISCARD, + MLXSW_REG_HPKT_ACTION_TRAP_AND_SOFT_DISCARD, + MLXSW_REG_HPKT_ACTION_TRAP_EXCEPTION_TO_CPU, + MLXSW_REG_HPKT_ACTION_SET_FW_DEFAULT = 15, +}; + +/* reg_hpkt_action + * Action to perform on packet when trapped. + * 0 - No action. Forward to CPU based on switching rules. + * 1 - Trap to CPU (CPU receives sole copy). + * 2 - Mirror to CPU (CPU receives a replica of the packet). + * 3 - Discard. + * 4 - Soft discard (allow other traps to act on the packet). + * 5 - Trap and soft discard (allow other traps to overwrite this trap). + * 6 - Trap to CPU (CPU receives sole copy) and count it as error. + * 15 - Restore the firmware's default action. + * Access: RW + * + * Note: Must be set to 0 (forward) for event trap IDs, as they are already + * addressed to the CPU. + */ +MLXSW_ITEM32(reg, hpkt, action, 0x00, 20, 3); + +/* reg_hpkt_trap_group + * Trap group to associate the trap with. + * Access: RW + */ +MLXSW_ITEM32(reg, hpkt, trap_group, 0x00, 12, 6); + +/* reg_hpkt_trap_id + * Trap ID. + * Access: Index + * + * Note: A trap ID can only be associated with a single trap group. The device + * will associate the trap ID with the last trap group configured. + */ +MLXSW_ITEM32(reg, hpkt, trap_id, 0x00, 0, 10); + +enum { + MLXSW_REG_HPKT_CTRL_PACKET_DEFAULT, + MLXSW_REG_HPKT_CTRL_PACKET_NO_BUFFER, + MLXSW_REG_HPKT_CTRL_PACKET_USE_BUFFER, +}; + +/* reg_hpkt_ctrl + * Configure dedicated buffer resources for control packets. + * Ignored by SwitchX-2. + * 0 - Keep factory defaults. + * 1 - Do not use control buffer for this trap ID. + * 2 - Use control buffer for this trap ID. + * Access: RW + */ +MLXSW_ITEM32(reg, hpkt, ctrl, 0x04, 16, 2); + +static inline void mlxsw_reg_hpkt_pack(char *payload, u8 action, u16 trap_id, + enum mlxsw_reg_htgt_trap_group trap_group, + bool is_ctrl) +{ + MLXSW_REG_ZERO(hpkt, payload); + mlxsw_reg_hpkt_ack_set(payload, MLXSW_REG_HPKT_ACK_NOT_REQUIRED); + mlxsw_reg_hpkt_action_set(payload, action); + mlxsw_reg_hpkt_trap_group_set(payload, trap_group); + mlxsw_reg_hpkt_trap_id_set(payload, trap_id); + mlxsw_reg_hpkt_ctrl_set(payload, is_ctrl ? + MLXSW_REG_HPKT_CTRL_PACKET_USE_BUFFER : + MLXSW_REG_HPKT_CTRL_PACKET_NO_BUFFER); +} + +/* RGCR - Router General Configuration Register + * -------------------------------------------- + * The register is used for setting up the router configuration. + */ +#define MLXSW_REG_RGCR_ID 0x8001 +#define MLXSW_REG_RGCR_LEN 0x28 + +MLXSW_REG_DEFINE(rgcr, MLXSW_REG_RGCR_ID, MLXSW_REG_RGCR_LEN); + +/* reg_rgcr_ipv4_en + * IPv4 router enable. + * Access: RW + */ +MLXSW_ITEM32(reg, rgcr, ipv4_en, 0x00, 31, 1); + +/* reg_rgcr_ipv6_en + * IPv6 router enable. + * Access: RW + */ +MLXSW_ITEM32(reg, rgcr, ipv6_en, 0x00, 30, 1); + +/* reg_rgcr_max_router_interfaces + * Defines the maximum number of active router interfaces for all virtual + * routers. + * Access: RW + */ +MLXSW_ITEM32(reg, rgcr, max_router_interfaces, 0x10, 0, 16); + +/* reg_rgcr_usp + * Update switch priority and packet color. + * 0 - Preserve the value of Switch Priority and packet color. + * 1 - Recalculate the value of Switch Priority and packet color. + * Access: RW + * + * Note: Not supported by SwitchX and SwitchX-2. + */ +MLXSW_ITEM32(reg, rgcr, usp, 0x18, 20, 1); + +/* reg_rgcr_pcp_rw + * Indicates how to handle the pcp_rewrite_en value: + * 0 - Preserve the value of pcp_rewrite_en. + * 2 - Disable PCP rewrite. + * 3 - Enable PCP rewrite. + * Access: RW + * + * Note: Not supported by SwitchX and SwitchX-2. + */ +MLXSW_ITEM32(reg, rgcr, pcp_rw, 0x18, 16, 2); + +/* reg_rgcr_activity_dis + * Activity disable: + * 0 - Activity will be set when an entry is hit (default). + * 1 - Activity will not be set when an entry is hit. + * + * Bit 0 - Disable activity bit in Router Algorithmic LPM Unicast Entry + * (RALUE). + * Bit 1 - Disable activity bit in Router Algorithmic LPM Unicast Host + * Entry (RAUHT). + * Bits 2:7 are reserved. + * Access: RW + * + * Note: Not supported by SwitchX, SwitchX-2 and Switch-IB. + */ +MLXSW_ITEM32(reg, rgcr, activity_dis, 0x20, 0, 8); + +static inline void mlxsw_reg_rgcr_pack(char *payload, bool ipv4_en, + bool ipv6_en) +{ + MLXSW_REG_ZERO(rgcr, payload); + mlxsw_reg_rgcr_ipv4_en_set(payload, ipv4_en); + mlxsw_reg_rgcr_ipv6_en_set(payload, ipv6_en); +} + +/* RITR - Router Interface Table Register + * -------------------------------------- + * The register is used to configure the router interface table. + */ +#define MLXSW_REG_RITR_ID 0x8002 +#define MLXSW_REG_RITR_LEN 0x40 + +MLXSW_REG_DEFINE(ritr, MLXSW_REG_RITR_ID, MLXSW_REG_RITR_LEN); + +/* reg_ritr_enable + * Enables routing on the router interface. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, enable, 0x00, 31, 1); + +/* reg_ritr_ipv4 + * IPv4 routing enable. Enables routing of IPv4 traffic on the router + * interface. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, ipv4, 0x00, 29, 1); + +/* reg_ritr_ipv6 + * IPv6 routing enable. Enables routing of IPv6 traffic on the router + * interface. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, ipv6, 0x00, 28, 1); + +/* reg_ritr_ipv4_mc + * IPv4 multicast routing enable. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, ipv4_mc, 0x00, 27, 1); + +/* reg_ritr_ipv6_mc + * IPv6 multicast routing enable. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, ipv6_mc, 0x00, 26, 1); + +enum mlxsw_reg_ritr_if_type { + /* VLAN interface. */ + MLXSW_REG_RITR_VLAN_IF, + /* FID interface. */ + MLXSW_REG_RITR_FID_IF, + /* Sub-port interface. */ + MLXSW_REG_RITR_SP_IF, + /* Loopback Interface. */ + MLXSW_REG_RITR_LOOPBACK_IF, +}; + +/* reg_ritr_type + * Router interface type as per enum mlxsw_reg_ritr_if_type. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, type, 0x00, 23, 3); + +enum { + MLXSW_REG_RITR_RIF_CREATE, + MLXSW_REG_RITR_RIF_DEL, +}; + +/* reg_ritr_op + * Opcode: + * 0 - Create or edit RIF. + * 1 - Delete RIF. + * Reserved for SwitchX-2. For Spectrum, editing of interface properties + * is not supported. An interface must be deleted and re-created in order + * to update properties. + * Access: WO + */ +MLXSW_ITEM32(reg, ritr, op, 0x00, 20, 2); + +/* reg_ritr_rif + * Router interface index. A pointer to the Router Interface Table. + * Access: Index + */ +MLXSW_ITEM32(reg, ritr, rif, 0x00, 0, 16); + +/* reg_ritr_ipv4_fe + * IPv4 Forwarding Enable. + * Enables routing of IPv4 traffic on the router interface. When disabled, + * forwarding is blocked but local traffic (traps and IP2ME) will be enabled. + * Not supported in SwitchX-2. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, ipv4_fe, 0x04, 29, 1); + +/* reg_ritr_ipv6_fe + * IPv6 Forwarding Enable. + * Enables routing of IPv6 traffic on the router interface. When disabled, + * forwarding is blocked but local traffic (traps and IP2ME) will be enabled. + * Not supported in SwitchX-2. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, ipv6_fe, 0x04, 28, 1); + +/* reg_ritr_ipv4_mc_fe + * IPv4 Multicast Forwarding Enable. + * When disabled, forwarding is blocked but local traffic (traps and IP to me) + * will be enabled. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, ipv4_mc_fe, 0x04, 27, 1); + +/* reg_ritr_ipv6_mc_fe + * IPv6 Multicast Forwarding Enable. + * When disabled, forwarding is blocked but local traffic (traps and IP to me) + * will be enabled. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, ipv6_mc_fe, 0x04, 26, 1); + +/* reg_ritr_lb_en + * Loop-back filter enable for unicast packets. + * If the flag is set then loop-back filter for unicast packets is + * implemented on the RIF. Multicast packets are always subject to + * loop-back filtering. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, lb_en, 0x04, 24, 1); + +/* reg_ritr_virtual_router + * Virtual router ID associated with the router interface. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, virtual_router, 0x04, 0, 16); + +/* reg_ritr_mtu + * Router interface MTU. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, mtu, 0x34, 0, 16); + +/* reg_ritr_if_swid + * Switch partition ID. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, if_swid, 0x08, 24, 8); + +/* reg_ritr_if_mac_profile_id + * MAC msb profile ID. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, if_mac_profile_id, 0x10, 16, 4); + +/* reg_ritr_if_mac + * Router interface MAC address. + * In Spectrum, all MAC addresses must have the same 38 MSBits. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, ritr, if_mac, 0x12, 6); + +/* reg_ritr_if_vrrp_id_ipv6 + * VRRP ID for IPv6 + * Note: Reserved for RIF types other than VLAN, FID and Sub-port. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, if_vrrp_id_ipv6, 0x1C, 8, 8); + +/* reg_ritr_if_vrrp_id_ipv4 + * VRRP ID for IPv4 + * Note: Reserved for RIF types other than VLAN, FID and Sub-port. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, if_vrrp_id_ipv4, 0x1C, 0, 8); + +/* VLAN Interface */ + +/* reg_ritr_vlan_if_vlan_id + * VLAN ID. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, vlan_if_vlan_id, 0x08, 0, 12); + +/* reg_ritr_vlan_if_efid + * Egress FID. + * Used to connect the RIF to a bridge. + * Access: RW + * + * Note: Reserved when legacy bridge model is used and on Spectrum-1. + */ +MLXSW_ITEM32(reg, ritr, vlan_if_efid, 0x0C, 0, 16); + +/* FID Interface */ + +/* reg_ritr_fid_if_fid + * Filtering ID. Used to connect a bridge to the router. + * When legacy bridge model is used, only FIDs from the vFID range are + * supported. When unified bridge model is used, this is the egress FID for + * router to bridge. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, fid_if_fid, 0x08, 0, 16); + +/* Sub-port Interface */ + +/* reg_ritr_sp_if_lag + * LAG indication. When this bit is set the system_port field holds the + * LAG identifier. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, sp_if_lag, 0x08, 24, 1); + +/* reg_ritr_sp_system_port + * Port unique indentifier. When lag bit is set, this field holds the + * lag_id in bits 0:9. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, sp_if_system_port, 0x08, 0, 16); + +/* reg_ritr_sp_if_efid + * Egress filtering ID. + * Used to connect the eRIF to a bridge if eRIF-ACL has modified the DMAC or + * the VID. + * Access: RW + * + * Note: Reserved when legacy bridge model is used. + */ +MLXSW_ITEM32(reg, ritr, sp_if_efid, 0x0C, 0, 16); + +/* reg_ritr_sp_if_vid + * VLAN ID. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, sp_if_vid, 0x18, 0, 12); + +/* Loopback Interface */ + +enum mlxsw_reg_ritr_loopback_protocol { + /* IPinIP IPv4 underlay Unicast */ + MLXSW_REG_RITR_LOOPBACK_PROTOCOL_IPIP_IPV4, + /* IPinIP IPv6 underlay Unicast */ + MLXSW_REG_RITR_LOOPBACK_PROTOCOL_IPIP_IPV6, + /* IPinIP generic - used for Spectrum-2 underlay RIF */ + MLXSW_REG_RITR_LOOPBACK_GENERIC, +}; + +/* reg_ritr_loopback_protocol + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, loopback_protocol, 0x08, 28, 4); + +enum mlxsw_reg_ritr_loopback_ipip_type { + /* Tunnel is IPinIP. */ + MLXSW_REG_RITR_LOOPBACK_IPIP_TYPE_IP_IN_IP, + /* Tunnel is GRE, no key. */ + MLXSW_REG_RITR_LOOPBACK_IPIP_TYPE_IP_IN_GRE_IN_IP, + /* Tunnel is GRE, with a key. */ + MLXSW_REG_RITR_LOOPBACK_IPIP_TYPE_IP_IN_GRE_KEY_IN_IP, +}; + +/* reg_ritr_loopback_ipip_type + * Encapsulation type. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, loopback_ipip_type, 0x10, 24, 4); + +enum mlxsw_reg_ritr_loopback_ipip_options { + /* The key is defined by gre_key. */ + MLXSW_REG_RITR_LOOPBACK_IPIP_OPTIONS_GRE_KEY_PRESET, +}; + +/* reg_ritr_loopback_ipip_options + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, loopback_ipip_options, 0x10, 20, 4); + +/* reg_ritr_loopback_ipip_uvr + * Underlay Virtual Router ID. + * Range is 0..cap_max_virtual_routers-1. + * Reserved for Spectrum-2. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, loopback_ipip_uvr, 0x10, 0, 16); + +/* reg_ritr_loopback_ipip_underlay_rif + * Underlay ingress router interface. + * Reserved for Spectrum. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, loopback_ipip_underlay_rif, 0x14, 0, 16); + +/* reg_ritr_loopback_ipip_usip* + * Encapsulation Underlay source IP. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, ritr, loopback_ipip_usip6, 0x18, 16); +MLXSW_ITEM32(reg, ritr, loopback_ipip_usip4, 0x24, 0, 32); + +/* reg_ritr_loopback_ipip_gre_key + * GRE Key. + * Reserved when ipip_type is not IP_IN_GRE_KEY_IN_IP. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, loopback_ipip_gre_key, 0x28, 0, 32); + +/* Shared between ingress/egress */ +enum mlxsw_reg_ritr_counter_set_type { + /* No Count. */ + MLXSW_REG_RITR_COUNTER_SET_TYPE_NO_COUNT = 0x0, + /* Basic. Used for router interfaces, counting the following: + * - Error and Discard counters. + * - Unicast, Multicast and Broadcast counters. Sharing the + * same set of counters for the different type of traffic + * (IPv4, IPv6 and mpls). + */ + MLXSW_REG_RITR_COUNTER_SET_TYPE_BASIC = 0x9, +}; + +/* reg_ritr_ingress_counter_index + * Counter Index for flow counter. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, ingress_counter_index, 0x38, 0, 24); + +/* reg_ritr_ingress_counter_set_type + * Igress Counter Set Type for router interface counter. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, ingress_counter_set_type, 0x38, 24, 8); + +/* reg_ritr_egress_counter_index + * Counter Index for flow counter. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, egress_counter_index, 0x3C, 0, 24); + +/* reg_ritr_egress_counter_set_type + * Egress Counter Set Type for router interface counter. + * Access: RW + */ +MLXSW_ITEM32(reg, ritr, egress_counter_set_type, 0x3C, 24, 8); + +static inline void mlxsw_reg_ritr_counter_pack(char *payload, u32 index, + bool enable, bool egress) +{ + enum mlxsw_reg_ritr_counter_set_type set_type; + + if (enable) + set_type = MLXSW_REG_RITR_COUNTER_SET_TYPE_BASIC; + else + set_type = MLXSW_REG_RITR_COUNTER_SET_TYPE_NO_COUNT; + + if (egress) { + mlxsw_reg_ritr_egress_counter_set_type_set(payload, set_type); + mlxsw_reg_ritr_egress_counter_index_set(payload, index); + } else { + mlxsw_reg_ritr_ingress_counter_set_type_set(payload, set_type); + mlxsw_reg_ritr_ingress_counter_index_set(payload, index); + } +} + +static inline void mlxsw_reg_ritr_rif_pack(char *payload, u16 rif) +{ + MLXSW_REG_ZERO(ritr, payload); + mlxsw_reg_ritr_rif_set(payload, rif); +} + +static inline void mlxsw_reg_ritr_sp_if_pack(char *payload, bool lag, + u16 system_port, u16 efid, u16 vid) +{ + mlxsw_reg_ritr_sp_if_lag_set(payload, lag); + mlxsw_reg_ritr_sp_if_system_port_set(payload, system_port); + mlxsw_reg_ritr_sp_if_efid_set(payload, efid); + mlxsw_reg_ritr_sp_if_vid_set(payload, vid); +} + +static inline void mlxsw_reg_ritr_pack(char *payload, bool enable, + enum mlxsw_reg_ritr_if_type type, + u16 rif, u16 vr_id, u16 mtu) +{ + bool op = enable ? MLXSW_REG_RITR_RIF_CREATE : MLXSW_REG_RITR_RIF_DEL; + + MLXSW_REG_ZERO(ritr, payload); + mlxsw_reg_ritr_enable_set(payload, enable); + mlxsw_reg_ritr_ipv4_set(payload, 1); + mlxsw_reg_ritr_ipv6_set(payload, 1); + mlxsw_reg_ritr_ipv4_mc_set(payload, 1); + mlxsw_reg_ritr_ipv6_mc_set(payload, 1); + mlxsw_reg_ritr_type_set(payload, type); + mlxsw_reg_ritr_op_set(payload, op); + mlxsw_reg_ritr_rif_set(payload, rif); + mlxsw_reg_ritr_ipv4_fe_set(payload, 1); + mlxsw_reg_ritr_ipv6_fe_set(payload, 1); + mlxsw_reg_ritr_ipv4_mc_fe_set(payload, 1); + mlxsw_reg_ritr_ipv6_mc_fe_set(payload, 1); + mlxsw_reg_ritr_lb_en_set(payload, 1); + mlxsw_reg_ritr_virtual_router_set(payload, vr_id); + mlxsw_reg_ritr_mtu_set(payload, mtu); +} + +static inline void mlxsw_reg_ritr_mac_pack(char *payload, const char *mac) +{ + mlxsw_reg_ritr_if_mac_memcpy_to(payload, mac); +} + +static inline void +mlxsw_reg_ritr_vlan_if_pack(char *payload, bool enable, u16 rif, u16 vr_id, + u16 mtu, const char *mac, u8 mac_profile_id, + u16 vlan_id, u16 efid) +{ + enum mlxsw_reg_ritr_if_type type = MLXSW_REG_RITR_VLAN_IF; + + mlxsw_reg_ritr_pack(payload, enable, type, rif, vr_id, mtu); + mlxsw_reg_ritr_if_mac_memcpy_to(payload, mac); + mlxsw_reg_ritr_if_mac_profile_id_set(payload, mac_profile_id); + mlxsw_reg_ritr_vlan_if_vlan_id_set(payload, vlan_id); + mlxsw_reg_ritr_vlan_if_efid_set(payload, efid); +} + +static inline void +mlxsw_reg_ritr_loopback_ipip_common_pack(char *payload, + enum mlxsw_reg_ritr_loopback_ipip_type ipip_type, + enum mlxsw_reg_ritr_loopback_ipip_options options, + u16 uvr_id, u16 underlay_rif, u32 gre_key) +{ + mlxsw_reg_ritr_loopback_ipip_type_set(payload, ipip_type); + mlxsw_reg_ritr_loopback_ipip_options_set(payload, options); + mlxsw_reg_ritr_loopback_ipip_uvr_set(payload, uvr_id); + mlxsw_reg_ritr_loopback_ipip_underlay_rif_set(payload, underlay_rif); + mlxsw_reg_ritr_loopback_ipip_gre_key_set(payload, gre_key); +} + +static inline void +mlxsw_reg_ritr_loopback_ipip4_pack(char *payload, + enum mlxsw_reg_ritr_loopback_ipip_type ipip_type, + enum mlxsw_reg_ritr_loopback_ipip_options options, + u16 uvr_id, u16 underlay_rif, u32 usip, u32 gre_key) +{ + mlxsw_reg_ritr_loopback_protocol_set(payload, + MLXSW_REG_RITR_LOOPBACK_PROTOCOL_IPIP_IPV4); + mlxsw_reg_ritr_loopback_ipip_common_pack(payload, ipip_type, options, + uvr_id, underlay_rif, gre_key); + mlxsw_reg_ritr_loopback_ipip_usip4_set(payload, usip); +} + +static inline void +mlxsw_reg_ritr_loopback_ipip6_pack(char *payload, + enum mlxsw_reg_ritr_loopback_ipip_type ipip_type, + enum mlxsw_reg_ritr_loopback_ipip_options options, + u16 uvr_id, u16 underlay_rif, + const struct in6_addr *usip, u32 gre_key) +{ + enum mlxsw_reg_ritr_loopback_protocol protocol = + MLXSW_REG_RITR_LOOPBACK_PROTOCOL_IPIP_IPV6; + + mlxsw_reg_ritr_loopback_protocol_set(payload, protocol); + mlxsw_reg_ritr_loopback_ipip_common_pack(payload, ipip_type, options, + uvr_id, underlay_rif, gre_key); + mlxsw_reg_ritr_loopback_ipip_usip6_memcpy_to(payload, + (const char *)usip); +} + +/* RTAR - Router TCAM Allocation Register + * -------------------------------------- + * This register is used for allocation of regions in the TCAM table. + */ +#define MLXSW_REG_RTAR_ID 0x8004 +#define MLXSW_REG_RTAR_LEN 0x20 + +MLXSW_REG_DEFINE(rtar, MLXSW_REG_RTAR_ID, MLXSW_REG_RTAR_LEN); + +enum mlxsw_reg_rtar_op { + MLXSW_REG_RTAR_OP_ALLOCATE, + MLXSW_REG_RTAR_OP_RESIZE, + MLXSW_REG_RTAR_OP_DEALLOCATE, +}; + +/* reg_rtar_op + * Access: WO + */ +MLXSW_ITEM32(reg, rtar, op, 0x00, 28, 4); + +enum mlxsw_reg_rtar_key_type { + MLXSW_REG_RTAR_KEY_TYPE_IPV4_MULTICAST = 1, + MLXSW_REG_RTAR_KEY_TYPE_IPV6_MULTICAST = 3 +}; + +/* reg_rtar_key_type + * TCAM key type for the region. + * Access: WO + */ +MLXSW_ITEM32(reg, rtar, key_type, 0x00, 0, 8); + +/* reg_rtar_region_size + * TCAM region size. When allocating/resizing this is the requested + * size, the response is the actual size. + * Note: Actual size may be larger than requested. + * Reserved for op = Deallocate + * Access: WO + */ +MLXSW_ITEM32(reg, rtar, region_size, 0x04, 0, 16); + +static inline void mlxsw_reg_rtar_pack(char *payload, + enum mlxsw_reg_rtar_op op, + enum mlxsw_reg_rtar_key_type key_type, + u16 region_size) +{ + MLXSW_REG_ZERO(rtar, payload); + mlxsw_reg_rtar_op_set(payload, op); + mlxsw_reg_rtar_key_type_set(payload, key_type); + mlxsw_reg_rtar_region_size_set(payload, region_size); +} + +/* RATR - Router Adjacency Table Register + * -------------------------------------- + * The RATR register is used to configure the Router Adjacency (next-hop) + * Table. + */ +#define MLXSW_REG_RATR_ID 0x8008 +#define MLXSW_REG_RATR_LEN 0x2C + +MLXSW_REG_DEFINE(ratr, MLXSW_REG_RATR_ID, MLXSW_REG_RATR_LEN); + +enum mlxsw_reg_ratr_op { + /* Read */ + MLXSW_REG_RATR_OP_QUERY_READ = 0, + /* Read and clear activity */ + MLXSW_REG_RATR_OP_QUERY_READ_CLEAR = 2, + /* Write Adjacency entry */ + MLXSW_REG_RATR_OP_WRITE_WRITE_ENTRY = 1, + /* Write Adjacency entry only if the activity is cleared. + * The write may not succeed if the activity is set. There is not + * direct feedback if the write has succeeded or not, however + * the get will reveal the actual entry (SW can compare the get + * response to the set command). + */ + MLXSW_REG_RATR_OP_WRITE_WRITE_ENTRY_ON_ACTIVITY = 3, +}; + +/* reg_ratr_op + * Note that Write operation may also be used for updating + * counter_set_type and counter_index. In this case all other + * fields must not be updated. + * Access: OP + */ +MLXSW_ITEM32(reg, ratr, op, 0x00, 28, 4); + +/* reg_ratr_v + * Valid bit. Indicates if the adjacency entry is valid. + * Note: the device may need some time before reusing an invalidated + * entry. During this time the entry can not be reused. It is + * recommended to use another entry before reusing an invalidated + * entry (e.g. software can put it at the end of the list for + * reusing). Trying to access an invalidated entry not yet cleared + * by the device results with failure indicating "Try Again" status. + * When valid is '0' then egress_router_interface,trap_action, + * adjacency_parameters and counters are reserved + * Access: RW + */ +MLXSW_ITEM32(reg, ratr, v, 0x00, 24, 1); + +/* reg_ratr_a + * Activity. Set for new entries. Set if a packet lookup has hit on + * the specific entry. To clear the a bit, use "clear activity". + * Access: RO + */ +MLXSW_ITEM32(reg, ratr, a, 0x00, 16, 1); + +enum mlxsw_reg_ratr_type { + /* Ethernet */ + MLXSW_REG_RATR_TYPE_ETHERNET, + /* IPoIB Unicast without GRH. + * Reserved for Spectrum. + */ + MLXSW_REG_RATR_TYPE_IPOIB_UC, + /* IPoIB Unicast with GRH. Supported only in table 0 (Ethernet unicast + * adjacency). + * Reserved for Spectrum. + */ + MLXSW_REG_RATR_TYPE_IPOIB_UC_W_GRH, + /* IPoIB Multicast. + * Reserved for Spectrum. + */ + MLXSW_REG_RATR_TYPE_IPOIB_MC, + /* MPLS. + * Reserved for SwitchX/-2. + */ + MLXSW_REG_RATR_TYPE_MPLS, + /* IPinIP Encap. + * Reserved for SwitchX/-2. + */ + MLXSW_REG_RATR_TYPE_IPIP, +}; + +/* reg_ratr_type + * Adjacency entry type. + * Access: RW + */ +MLXSW_ITEM32(reg, ratr, type, 0x04, 28, 4); + +/* reg_ratr_adjacency_index_low + * Bits 15:0 of index into the adjacency table. + * For SwitchX and SwitchX-2, the adjacency table is linear and + * used for adjacency entries only. + * For Spectrum, the index is to the KVD linear. + * Access: Index + */ +MLXSW_ITEM32(reg, ratr, adjacency_index_low, 0x04, 0, 16); + +/* reg_ratr_egress_router_interface + * Range is 0 .. cap_max_router_interfaces - 1 + * Access: RW + */ +MLXSW_ITEM32(reg, ratr, egress_router_interface, 0x08, 0, 16); + +enum mlxsw_reg_ratr_trap_action { + MLXSW_REG_RATR_TRAP_ACTION_NOP, + MLXSW_REG_RATR_TRAP_ACTION_TRAP, + MLXSW_REG_RATR_TRAP_ACTION_MIRROR_TO_CPU, + MLXSW_REG_RATR_TRAP_ACTION_MIRROR, + MLXSW_REG_RATR_TRAP_ACTION_DISCARD_ERRORS, +}; + +/* reg_ratr_trap_action + * see mlxsw_reg_ratr_trap_action + * Access: RW + */ +MLXSW_ITEM32(reg, ratr, trap_action, 0x0C, 28, 4); + +/* reg_ratr_adjacency_index_high + * Bits 23:16 of the adjacency_index. + * Access: Index + */ +MLXSW_ITEM32(reg, ratr, adjacency_index_high, 0x0C, 16, 8); + +enum mlxsw_reg_ratr_trap_id { + MLXSW_REG_RATR_TRAP_ID_RTR_EGRESS0, + MLXSW_REG_RATR_TRAP_ID_RTR_EGRESS1, +}; + +/* reg_ratr_trap_id + * Trap ID to be reported to CPU. + * Trap-ID is RTR_EGRESS0 or RTR_EGRESS1. + * For trap_action of NOP, MIRROR and DISCARD_ERROR + * Access: RW + */ +MLXSW_ITEM32(reg, ratr, trap_id, 0x0C, 0, 8); + +/* reg_ratr_eth_destination_mac + * MAC address of the destination next-hop. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, ratr, eth_destination_mac, 0x12, 6); + +enum mlxsw_reg_ratr_ipip_type { + /* IPv4, address set by mlxsw_reg_ratr_ipip_ipv4_udip. */ + MLXSW_REG_RATR_IPIP_TYPE_IPV4, + /* IPv6, address set by mlxsw_reg_ratr_ipip_ipv6_ptr. */ + MLXSW_REG_RATR_IPIP_TYPE_IPV6, +}; + +/* reg_ratr_ipip_type + * Underlay destination ip type. + * Note: the type field must match the protocol of the router interface. + * Access: RW + */ +MLXSW_ITEM32(reg, ratr, ipip_type, 0x10, 16, 4); + +/* reg_ratr_ipip_ipv4_udip + * Underlay ipv4 dip. + * Reserved when ipip_type is IPv6. + * Access: RW + */ +MLXSW_ITEM32(reg, ratr, ipip_ipv4_udip, 0x18, 0, 32); + +/* reg_ratr_ipip_ipv6_ptr + * Pointer to IPv6 underlay destination ip address. + * For Spectrum: Pointer to KVD linear space. + * Access: RW + */ +MLXSW_ITEM32(reg, ratr, ipip_ipv6_ptr, 0x1C, 0, 24); + +enum mlxsw_reg_flow_counter_set_type { + /* No count */ + MLXSW_REG_FLOW_COUNTER_SET_TYPE_NO_COUNT = 0x00, + /* Count packets and bytes */ + MLXSW_REG_FLOW_COUNTER_SET_TYPE_PACKETS_BYTES = 0x03, + /* Count only packets */ + MLXSW_REG_FLOW_COUNTER_SET_TYPE_PACKETS = 0x05, +}; + +/* reg_ratr_counter_set_type + * Counter set type for flow counters + * Access: RW + */ +MLXSW_ITEM32(reg, ratr, counter_set_type, 0x28, 24, 8); + +/* reg_ratr_counter_index + * Counter index for flow counters + * Access: RW + */ +MLXSW_ITEM32(reg, ratr, counter_index, 0x28, 0, 24); + +static inline void +mlxsw_reg_ratr_pack(char *payload, + enum mlxsw_reg_ratr_op op, bool valid, + enum mlxsw_reg_ratr_type type, + u32 adjacency_index, u16 egress_rif) +{ + MLXSW_REG_ZERO(ratr, payload); + mlxsw_reg_ratr_op_set(payload, op); + mlxsw_reg_ratr_v_set(payload, valid); + mlxsw_reg_ratr_type_set(payload, type); + mlxsw_reg_ratr_adjacency_index_low_set(payload, adjacency_index); + mlxsw_reg_ratr_adjacency_index_high_set(payload, adjacency_index >> 16); + mlxsw_reg_ratr_egress_router_interface_set(payload, egress_rif); +} + +static inline void mlxsw_reg_ratr_eth_entry_pack(char *payload, + const char *dest_mac) +{ + mlxsw_reg_ratr_eth_destination_mac_memcpy_to(payload, dest_mac); +} + +static inline void mlxsw_reg_ratr_ipip4_entry_pack(char *payload, u32 ipv4_udip) +{ + mlxsw_reg_ratr_ipip_type_set(payload, MLXSW_REG_RATR_IPIP_TYPE_IPV4); + mlxsw_reg_ratr_ipip_ipv4_udip_set(payload, ipv4_udip); +} + +static inline void mlxsw_reg_ratr_ipip6_entry_pack(char *payload, u32 ipv6_ptr) +{ + mlxsw_reg_ratr_ipip_type_set(payload, MLXSW_REG_RATR_IPIP_TYPE_IPV6); + mlxsw_reg_ratr_ipip_ipv6_ptr_set(payload, ipv6_ptr); +} + +static inline void mlxsw_reg_ratr_counter_pack(char *payload, u64 counter_index, + bool counter_enable) +{ + enum mlxsw_reg_flow_counter_set_type set_type; + + if (counter_enable) + set_type = MLXSW_REG_FLOW_COUNTER_SET_TYPE_PACKETS_BYTES; + else + set_type = MLXSW_REG_FLOW_COUNTER_SET_TYPE_NO_COUNT; + + mlxsw_reg_ratr_counter_index_set(payload, counter_index); + mlxsw_reg_ratr_counter_set_type_set(payload, set_type); +} + +/* RDPM - Router DSCP to Priority Mapping + * -------------------------------------- + * Controls the mapping from DSCP field to switch priority on routed packets + */ +#define MLXSW_REG_RDPM_ID 0x8009 +#define MLXSW_REG_RDPM_BASE_LEN 0x00 +#define MLXSW_REG_RDPM_DSCP_ENTRY_REC_LEN 0x01 +#define MLXSW_REG_RDPM_DSCP_ENTRY_REC_MAX_COUNT 64 +#define MLXSW_REG_RDPM_LEN 0x40 +#define MLXSW_REG_RDPM_LAST_ENTRY (MLXSW_REG_RDPM_BASE_LEN + \ + MLXSW_REG_RDPM_LEN - \ + MLXSW_REG_RDPM_DSCP_ENTRY_REC_LEN) + +MLXSW_REG_DEFINE(rdpm, MLXSW_REG_RDPM_ID, MLXSW_REG_RDPM_LEN); + +/* reg_dscp_entry_e + * Enable update of the specific entry + * Access: Index + */ +MLXSW_ITEM8_INDEXED(reg, rdpm, dscp_entry_e, MLXSW_REG_RDPM_LAST_ENTRY, 7, 1, + -MLXSW_REG_RDPM_DSCP_ENTRY_REC_LEN, 0x00, false); + +/* reg_dscp_entry_prio + * Switch Priority + * Access: RW + */ +MLXSW_ITEM8_INDEXED(reg, rdpm, dscp_entry_prio, MLXSW_REG_RDPM_LAST_ENTRY, 0, 4, + -MLXSW_REG_RDPM_DSCP_ENTRY_REC_LEN, 0x00, false); + +static inline void mlxsw_reg_rdpm_pack(char *payload, unsigned short index, + u8 prio) +{ + mlxsw_reg_rdpm_dscp_entry_e_set(payload, index, 1); + mlxsw_reg_rdpm_dscp_entry_prio_set(payload, index, prio); +} + +/* RICNT - Router Interface Counter Register + * ----------------------------------------- + * The RICNT register retrieves per port performance counters + */ +#define MLXSW_REG_RICNT_ID 0x800B +#define MLXSW_REG_RICNT_LEN 0x100 + +MLXSW_REG_DEFINE(ricnt, MLXSW_REG_RICNT_ID, MLXSW_REG_RICNT_LEN); + +/* reg_ricnt_counter_index + * Counter index + * Access: RW + */ +MLXSW_ITEM32(reg, ricnt, counter_index, 0x04, 0, 24); + +enum mlxsw_reg_ricnt_counter_set_type { + /* No Count. */ + MLXSW_REG_RICNT_COUNTER_SET_TYPE_NO_COUNT = 0x00, + /* Basic. Used for router interfaces, counting the following: + * - Error and Discard counters. + * - Unicast, Multicast and Broadcast counters. Sharing the + * same set of counters for the different type of traffic + * (IPv4, IPv6 and mpls). + */ + MLXSW_REG_RICNT_COUNTER_SET_TYPE_BASIC = 0x09, +}; + +/* reg_ricnt_counter_set_type + * Counter Set Type for router interface counter + * Access: RW + */ +MLXSW_ITEM32(reg, ricnt, counter_set_type, 0x04, 24, 8); + +enum mlxsw_reg_ricnt_opcode { + /* Nop. Supported only for read access*/ + MLXSW_REG_RICNT_OPCODE_NOP = 0x00, + /* Clear. Setting the clr bit will reset the counter value for + * all counters of the specified Router Interface. + */ + MLXSW_REG_RICNT_OPCODE_CLEAR = 0x08, +}; + +/* reg_ricnt_opcode + * Opcode + * Access: RW + */ +MLXSW_ITEM32(reg, ricnt, op, 0x00, 28, 4); + +/* reg_ricnt_good_unicast_packets + * good unicast packets. + * Access: RW + */ +MLXSW_ITEM64(reg, ricnt, good_unicast_packets, 0x08, 0, 64); + +/* reg_ricnt_good_multicast_packets + * good multicast packets. + * Access: RW + */ +MLXSW_ITEM64(reg, ricnt, good_multicast_packets, 0x10, 0, 64); + +/* reg_ricnt_good_broadcast_packets + * good broadcast packets + * Access: RW + */ +MLXSW_ITEM64(reg, ricnt, good_broadcast_packets, 0x18, 0, 64); + +/* reg_ricnt_good_unicast_bytes + * A count of L3 data and padding octets not including L2 headers + * for good unicast frames. + * Access: RW + */ +MLXSW_ITEM64(reg, ricnt, good_unicast_bytes, 0x20, 0, 64); + +/* reg_ricnt_good_multicast_bytes + * A count of L3 data and padding octets not including L2 headers + * for good multicast frames. + * Access: RW + */ +MLXSW_ITEM64(reg, ricnt, good_multicast_bytes, 0x28, 0, 64); + +/* reg_ritr_good_broadcast_bytes + * A count of L3 data and padding octets not including L2 headers + * for good broadcast frames. + * Access: RW + */ +MLXSW_ITEM64(reg, ricnt, good_broadcast_bytes, 0x30, 0, 64); + +/* reg_ricnt_error_packets + * A count of errored frames that do not pass the router checks. + * Access: RW + */ +MLXSW_ITEM64(reg, ricnt, error_packets, 0x38, 0, 64); + +/* reg_ricnt_discrad_packets + * A count of non-errored frames that do not pass the router checks. + * Access: RW + */ +MLXSW_ITEM64(reg, ricnt, discard_packets, 0x40, 0, 64); + +/* reg_ricnt_error_bytes + * A count of L3 data and padding octets not including L2 headers + * for errored frames. + * Access: RW + */ +MLXSW_ITEM64(reg, ricnt, error_bytes, 0x48, 0, 64); + +/* reg_ricnt_discard_bytes + * A count of L3 data and padding octets not including L2 headers + * for non-errored frames that do not pass the router checks. + * Access: RW + */ +MLXSW_ITEM64(reg, ricnt, discard_bytes, 0x50, 0, 64); + +static inline void mlxsw_reg_ricnt_pack(char *payload, u32 index, + enum mlxsw_reg_ricnt_opcode op) +{ + MLXSW_REG_ZERO(ricnt, payload); + mlxsw_reg_ricnt_op_set(payload, op); + mlxsw_reg_ricnt_counter_index_set(payload, index); + mlxsw_reg_ricnt_counter_set_type_set(payload, + MLXSW_REG_RICNT_COUNTER_SET_TYPE_BASIC); +} + +/* RRCR - Router Rules Copy Register Layout + * ---------------------------------------- + * This register is used for moving and copying route entry rules. + */ +#define MLXSW_REG_RRCR_ID 0x800F +#define MLXSW_REG_RRCR_LEN 0x24 + +MLXSW_REG_DEFINE(rrcr, MLXSW_REG_RRCR_ID, MLXSW_REG_RRCR_LEN); + +enum mlxsw_reg_rrcr_op { + /* Move rules */ + MLXSW_REG_RRCR_OP_MOVE, + /* Copy rules */ + MLXSW_REG_RRCR_OP_COPY, +}; + +/* reg_rrcr_op + * Access: WO + */ +MLXSW_ITEM32(reg, rrcr, op, 0x00, 28, 4); + +/* reg_rrcr_offset + * Offset within the region from which to copy/move. + * Access: Index + */ +MLXSW_ITEM32(reg, rrcr, offset, 0x00, 0, 16); + +/* reg_rrcr_size + * The number of rules to copy/move. + * Access: WO + */ +MLXSW_ITEM32(reg, rrcr, size, 0x04, 0, 16); + +/* reg_rrcr_table_id + * Identifier of the table on which to perform the operation. Encoding is the + * same as in RTAR.key_type + * Access: Index + */ +MLXSW_ITEM32(reg, rrcr, table_id, 0x10, 0, 4); + +/* reg_rrcr_dest_offset + * Offset within the region to which to copy/move + * Access: Index + */ +MLXSW_ITEM32(reg, rrcr, dest_offset, 0x20, 0, 16); + +static inline void mlxsw_reg_rrcr_pack(char *payload, enum mlxsw_reg_rrcr_op op, + u16 offset, u16 size, + enum mlxsw_reg_rtar_key_type table_id, + u16 dest_offset) +{ + MLXSW_REG_ZERO(rrcr, payload); + mlxsw_reg_rrcr_op_set(payload, op); + mlxsw_reg_rrcr_offset_set(payload, offset); + mlxsw_reg_rrcr_size_set(payload, size); + mlxsw_reg_rrcr_table_id_set(payload, table_id); + mlxsw_reg_rrcr_dest_offset_set(payload, dest_offset); +} + +/* RALTA - Router Algorithmic LPM Tree Allocation Register + * ------------------------------------------------------- + * RALTA is used to allocate the LPM trees of the SHSPM method. + */ +#define MLXSW_REG_RALTA_ID 0x8010 +#define MLXSW_REG_RALTA_LEN 0x04 + +MLXSW_REG_DEFINE(ralta, MLXSW_REG_RALTA_ID, MLXSW_REG_RALTA_LEN); + +/* reg_ralta_op + * opcode (valid for Write, must be 0 on Read) + * 0 - allocate a tree + * 1 - deallocate a tree + * Access: OP + */ +MLXSW_ITEM32(reg, ralta, op, 0x00, 28, 2); + +enum mlxsw_reg_ralxx_protocol { + MLXSW_REG_RALXX_PROTOCOL_IPV4, + MLXSW_REG_RALXX_PROTOCOL_IPV6, +}; + +/* reg_ralta_protocol + * Protocol. + * Deallocation opcode: Reserved. + * Access: RW + */ +MLXSW_ITEM32(reg, ralta, protocol, 0x00, 24, 4); + +/* reg_ralta_tree_id + * An identifier (numbered from 1..cap_shspm_max_trees-1) representing + * the tree identifier (managed by software). + * Note that tree_id 0 is allocated for a default-route tree. + * Access: Index + */ +MLXSW_ITEM32(reg, ralta, tree_id, 0x00, 0, 8); + +static inline void mlxsw_reg_ralta_pack(char *payload, bool alloc, + enum mlxsw_reg_ralxx_protocol protocol, + u8 tree_id) +{ + MLXSW_REG_ZERO(ralta, payload); + mlxsw_reg_ralta_op_set(payload, !alloc); + mlxsw_reg_ralta_protocol_set(payload, protocol); + mlxsw_reg_ralta_tree_id_set(payload, tree_id); +} + +/* RALST - Router Algorithmic LPM Structure Tree Register + * ------------------------------------------------------ + * RALST is used to set and query the structure of an LPM tree. + * The structure of the tree must be sorted as a sorted binary tree, while + * each node is a bin that is tagged as the length of the prefixes the lookup + * will refer to. Therefore, bin X refers to a set of entries with prefixes + * of X bits to match with the destination address. The bin 0 indicates + * the default action, when there is no match of any prefix. + */ +#define MLXSW_REG_RALST_ID 0x8011 +#define MLXSW_REG_RALST_LEN 0x104 + +MLXSW_REG_DEFINE(ralst, MLXSW_REG_RALST_ID, MLXSW_REG_RALST_LEN); + +/* reg_ralst_root_bin + * The bin number of the root bin. + * 0<root_bin=<(length of IP address) + * For a default-route tree configure 0xff + * Access: RW + */ +MLXSW_ITEM32(reg, ralst, root_bin, 0x00, 16, 8); + +/* reg_ralst_tree_id + * Tree identifier numbered from 1..(cap_shspm_max_trees-1). + * Access: Index + */ +MLXSW_ITEM32(reg, ralst, tree_id, 0x00, 0, 8); + +#define MLXSW_REG_RALST_BIN_NO_CHILD 0xff +#define MLXSW_REG_RALST_BIN_OFFSET 0x04 +#define MLXSW_REG_RALST_BIN_COUNT 128 + +/* reg_ralst_left_child_bin + * Holding the children of the bin according to the stored tree's structure. + * For trees composed of less than 4 blocks, the bins in excess are reserved. + * Note that tree_id 0 is allocated for a default-route tree, bins are 0xff + * Access: RW + */ +MLXSW_ITEM16_INDEXED(reg, ralst, left_child_bin, 0x04, 8, 8, 0x02, 0x00, false); + +/* reg_ralst_right_child_bin + * Holding the children of the bin according to the stored tree's structure. + * For trees composed of less than 4 blocks, the bins in excess are reserved. + * Note that tree_id 0 is allocated for a default-route tree, bins are 0xff + * Access: RW + */ +MLXSW_ITEM16_INDEXED(reg, ralst, right_child_bin, 0x04, 0, 8, 0x02, 0x00, + false); + +static inline void mlxsw_reg_ralst_pack(char *payload, u8 root_bin, u8 tree_id) +{ + MLXSW_REG_ZERO(ralst, payload); + + /* Initialize all bins to have no left or right child */ + memset(payload + MLXSW_REG_RALST_BIN_OFFSET, + MLXSW_REG_RALST_BIN_NO_CHILD, MLXSW_REG_RALST_BIN_COUNT * 2); + + mlxsw_reg_ralst_root_bin_set(payload, root_bin); + mlxsw_reg_ralst_tree_id_set(payload, tree_id); +} + +static inline void mlxsw_reg_ralst_bin_pack(char *payload, u8 bin_number, + u8 left_child_bin, + u8 right_child_bin) +{ + int bin_index = bin_number - 1; + + mlxsw_reg_ralst_left_child_bin_set(payload, bin_index, left_child_bin); + mlxsw_reg_ralst_right_child_bin_set(payload, bin_index, + right_child_bin); +} + +/* RALTB - Router Algorithmic LPM Tree Binding Register + * ---------------------------------------------------- + * RALTB is used to bind virtual router and protocol to an allocated LPM tree. + */ +#define MLXSW_REG_RALTB_ID 0x8012 +#define MLXSW_REG_RALTB_LEN 0x04 + +MLXSW_REG_DEFINE(raltb, MLXSW_REG_RALTB_ID, MLXSW_REG_RALTB_LEN); + +/* reg_raltb_virtual_router + * Virtual Router ID + * Range is 0..cap_max_virtual_routers-1 + * Access: Index + */ +MLXSW_ITEM32(reg, raltb, virtual_router, 0x00, 16, 16); + +/* reg_raltb_protocol + * Protocol. + * Access: Index + */ +MLXSW_ITEM32(reg, raltb, protocol, 0x00, 12, 4); + +/* reg_raltb_tree_id + * Tree to be used for the {virtual_router, protocol} + * Tree identifier numbered from 1..(cap_shspm_max_trees-1). + * By default, all Unicast IPv4 and IPv6 are bound to tree_id 0. + * Access: RW + */ +MLXSW_ITEM32(reg, raltb, tree_id, 0x00, 0, 8); + +static inline void mlxsw_reg_raltb_pack(char *payload, u16 virtual_router, + enum mlxsw_reg_ralxx_protocol protocol, + u8 tree_id) +{ + MLXSW_REG_ZERO(raltb, payload); + mlxsw_reg_raltb_virtual_router_set(payload, virtual_router); + mlxsw_reg_raltb_protocol_set(payload, protocol); + mlxsw_reg_raltb_tree_id_set(payload, tree_id); +} + +/* RALUE - Router Algorithmic LPM Unicast Entry Register + * ----------------------------------------------------- + * RALUE is used to configure and query LPM entries that serve + * the Unicast protocols. + */ +#define MLXSW_REG_RALUE_ID 0x8013 +#define MLXSW_REG_RALUE_LEN 0x38 + +MLXSW_REG_DEFINE(ralue, MLXSW_REG_RALUE_ID, MLXSW_REG_RALUE_LEN); + +/* reg_ralue_protocol + * Protocol. + * Access: Index + */ +MLXSW_ITEM32(reg, ralue, protocol, 0x00, 24, 4); + +enum mlxsw_reg_ralue_op { + /* Read operation. If entry doesn't exist, the operation fails. */ + MLXSW_REG_RALUE_OP_QUERY_READ = 0, + /* Clear on read operation. Used to read entry and + * clear Activity bit. + */ + MLXSW_REG_RALUE_OP_QUERY_CLEAR = 1, + /* Write operation. Used to write a new entry to the table. All RW + * fields are written for new entry. Activity bit is set + * for new entries. + */ + MLXSW_REG_RALUE_OP_WRITE_WRITE = 0, + /* Update operation. Used to update an existing route entry and + * only update the RW fields that are detailed in the field + * op_u_mask. If entry doesn't exist, the operation fails. + */ + MLXSW_REG_RALUE_OP_WRITE_UPDATE = 1, + /* Clear activity. The Activity bit (the field a) is cleared + * for the entry. + */ + MLXSW_REG_RALUE_OP_WRITE_CLEAR = 2, + /* Delete operation. Used to delete an existing entry. If entry + * doesn't exist, the operation fails. + */ + MLXSW_REG_RALUE_OP_WRITE_DELETE = 3, +}; + +/* reg_ralue_op + * Operation. + * Access: OP + */ +MLXSW_ITEM32(reg, ralue, op, 0x00, 20, 3); + +/* reg_ralue_a + * Activity. Set for new entries. Set if a packet lookup has hit on the + * specific entry, only if the entry is a route. To clear the a bit, use + * "clear activity" op. + * Enabled by activity_dis in RGCR + * Access: RO + */ +MLXSW_ITEM32(reg, ralue, a, 0x00, 16, 1); + +/* reg_ralue_virtual_router + * Virtual Router ID + * Range is 0..cap_max_virtual_routers-1 + * Access: Index + */ +MLXSW_ITEM32(reg, ralue, virtual_router, 0x04, 16, 16); + +#define MLXSW_REG_RALUE_OP_U_MASK_ENTRY_TYPE BIT(0) +#define MLXSW_REG_RALUE_OP_U_MASK_BMP_LEN BIT(1) +#define MLXSW_REG_RALUE_OP_U_MASK_ACTION BIT(2) + +/* reg_ralue_op_u_mask + * opcode update mask. + * On read operation, this field is reserved. + * This field is valid for update opcode, otherwise - reserved. + * This field is a bitmask of the fields that should be updated. + * Access: WO + */ +MLXSW_ITEM32(reg, ralue, op_u_mask, 0x04, 8, 3); + +/* reg_ralue_prefix_len + * Number of bits in the prefix of the LPM route. + * Note that for IPv6 prefixes, if prefix_len>64 the entry consumes + * two entries in the physical HW table. + * Access: Index + */ +MLXSW_ITEM32(reg, ralue, prefix_len, 0x08, 0, 8); + +/* reg_ralue_dip* + * The prefix of the route or of the marker that the object of the LPM + * is compared with. The most significant bits of the dip are the prefix. + * The least significant bits must be '0' if the prefix_len is smaller + * than 128 for IPv6 or smaller than 32 for IPv4. + * IPv4 address uses bits dip[31:0] and bits dip[127:32] are reserved. + * Access: Index + */ +MLXSW_ITEM32(reg, ralue, dip4, 0x18, 0, 32); +MLXSW_ITEM_BUF(reg, ralue, dip6, 0x0C, 16); + +enum mlxsw_reg_ralue_entry_type { + MLXSW_REG_RALUE_ENTRY_TYPE_MARKER_ENTRY = 1, + MLXSW_REG_RALUE_ENTRY_TYPE_ROUTE_ENTRY = 2, + MLXSW_REG_RALUE_ENTRY_TYPE_MARKER_AND_ROUTE_ENTRY = 3, +}; + +/* reg_ralue_entry_type + * Entry type. + * Note - for Marker entries, the action_type and action fields are reserved. + * Access: RW + */ +MLXSW_ITEM32(reg, ralue, entry_type, 0x1C, 30, 2); + +/* reg_ralue_bmp_len + * The best match prefix length in the case that there is no match for + * longer prefixes. + * If (entry_type != MARKER_ENTRY), bmp_len must be equal to prefix_len + * Note for any update operation with entry_type modification this + * field must be set. + * Access: RW + */ +MLXSW_ITEM32(reg, ralue, bmp_len, 0x1C, 16, 8); + +enum mlxsw_reg_ralue_action_type { + MLXSW_REG_RALUE_ACTION_TYPE_REMOTE, + MLXSW_REG_RALUE_ACTION_TYPE_LOCAL, + MLXSW_REG_RALUE_ACTION_TYPE_IP2ME, +}; + +/* reg_ralue_action_type + * Action Type + * Indicates how the IP address is connected. + * It can be connected to a local subnet through local_erif or can be + * on a remote subnet connected through a next-hop router, + * or transmitted to the CPU. + * Reserved when entry_type = MARKER_ENTRY + * Access: RW + */ +MLXSW_ITEM32(reg, ralue, action_type, 0x1C, 0, 2); + +enum mlxsw_reg_ralue_trap_action { + MLXSW_REG_RALUE_TRAP_ACTION_NOP, + MLXSW_REG_RALUE_TRAP_ACTION_TRAP, + MLXSW_REG_RALUE_TRAP_ACTION_MIRROR_TO_CPU, + MLXSW_REG_RALUE_TRAP_ACTION_MIRROR, + MLXSW_REG_RALUE_TRAP_ACTION_DISCARD_ERROR, +}; + +/* reg_ralue_trap_action + * Trap action. + * For IP2ME action, only NOP and MIRROR are possible. + * Access: RW + */ +MLXSW_ITEM32(reg, ralue, trap_action, 0x20, 28, 4); + +/* reg_ralue_trap_id + * Trap ID to be reported to CPU. + * Trap ID is RTR_INGRESS0 or RTR_INGRESS1. + * For trap_action of NOP, MIRROR and DISCARD_ERROR, trap_id is reserved. + * Access: RW + */ +MLXSW_ITEM32(reg, ralue, trap_id, 0x20, 0, 9); + +/* reg_ralue_adjacency_index + * Points to the first entry of the group-based ECMP. + * Only relevant in case of REMOTE action. + * Access: RW + */ +MLXSW_ITEM32(reg, ralue, adjacency_index, 0x24, 0, 24); + +/* reg_ralue_ecmp_size + * Amount of sequential entries starting + * from the adjacency_index (the number of ECMPs). + * The valid range is 1-64, 512, 1024, 2048 and 4096. + * Reserved when trap_action is TRAP or DISCARD_ERROR. + * Only relevant in case of REMOTE action. + * Access: RW + */ +MLXSW_ITEM32(reg, ralue, ecmp_size, 0x28, 0, 13); + +/* reg_ralue_local_erif + * Egress Router Interface. + * Only relevant in case of LOCAL action. + * Access: RW + */ +MLXSW_ITEM32(reg, ralue, local_erif, 0x24, 0, 16); + +/* reg_ralue_ip2me_v + * Valid bit for the tunnel_ptr field. + * If valid = 0 then trap to CPU as IP2ME trap ID. + * If valid = 1 and the packet format allows NVE or IPinIP tunnel + * decapsulation then tunnel decapsulation is done. + * If valid = 1 and packet format does not allow NVE or IPinIP tunnel + * decapsulation then trap as IP2ME trap ID. + * Only relevant in case of IP2ME action. + * Access: RW + */ +MLXSW_ITEM32(reg, ralue, ip2me_v, 0x24, 31, 1); + +/* reg_ralue_ip2me_tunnel_ptr + * Tunnel Pointer for NVE or IPinIP tunnel decapsulation. + * For Spectrum, pointer to KVD Linear. + * Only relevant in case of IP2ME action. + * Access: RW + */ +MLXSW_ITEM32(reg, ralue, ip2me_tunnel_ptr, 0x24, 0, 24); + +static inline void mlxsw_reg_ralue_pack(char *payload, + enum mlxsw_reg_ralxx_protocol protocol, + enum mlxsw_reg_ralue_op op, + u16 virtual_router, u8 prefix_len) +{ + MLXSW_REG_ZERO(ralue, payload); + mlxsw_reg_ralue_protocol_set(payload, protocol); + mlxsw_reg_ralue_op_set(payload, op); + mlxsw_reg_ralue_virtual_router_set(payload, virtual_router); + mlxsw_reg_ralue_prefix_len_set(payload, prefix_len); + mlxsw_reg_ralue_entry_type_set(payload, + MLXSW_REG_RALUE_ENTRY_TYPE_ROUTE_ENTRY); + mlxsw_reg_ralue_bmp_len_set(payload, prefix_len); +} + +static inline void mlxsw_reg_ralue_pack4(char *payload, + enum mlxsw_reg_ralxx_protocol protocol, + enum mlxsw_reg_ralue_op op, + u16 virtual_router, u8 prefix_len, + u32 dip) +{ + mlxsw_reg_ralue_pack(payload, protocol, op, virtual_router, prefix_len); + mlxsw_reg_ralue_dip4_set(payload, dip); +} + +static inline void mlxsw_reg_ralue_pack6(char *payload, + enum mlxsw_reg_ralxx_protocol protocol, + enum mlxsw_reg_ralue_op op, + u16 virtual_router, u8 prefix_len, + const void *dip) +{ + mlxsw_reg_ralue_pack(payload, protocol, op, virtual_router, prefix_len); + mlxsw_reg_ralue_dip6_memcpy_to(payload, dip); +} + +static inline void +mlxsw_reg_ralue_act_remote_pack(char *payload, + enum mlxsw_reg_ralue_trap_action trap_action, + u16 trap_id, u32 adjacency_index, u16 ecmp_size) +{ + mlxsw_reg_ralue_action_type_set(payload, + MLXSW_REG_RALUE_ACTION_TYPE_REMOTE); + mlxsw_reg_ralue_trap_action_set(payload, trap_action); + mlxsw_reg_ralue_trap_id_set(payload, trap_id); + mlxsw_reg_ralue_adjacency_index_set(payload, adjacency_index); + mlxsw_reg_ralue_ecmp_size_set(payload, ecmp_size); +} + +static inline void +mlxsw_reg_ralue_act_local_pack(char *payload, + enum mlxsw_reg_ralue_trap_action trap_action, + u16 trap_id, u16 local_erif) +{ + mlxsw_reg_ralue_action_type_set(payload, + MLXSW_REG_RALUE_ACTION_TYPE_LOCAL); + mlxsw_reg_ralue_trap_action_set(payload, trap_action); + mlxsw_reg_ralue_trap_id_set(payload, trap_id); + mlxsw_reg_ralue_local_erif_set(payload, local_erif); +} + +static inline void +mlxsw_reg_ralue_act_ip2me_pack(char *payload) +{ + mlxsw_reg_ralue_action_type_set(payload, + MLXSW_REG_RALUE_ACTION_TYPE_IP2ME); +} + +static inline void +mlxsw_reg_ralue_act_ip2me_tun_pack(char *payload, u32 tunnel_ptr) +{ + mlxsw_reg_ralue_action_type_set(payload, + MLXSW_REG_RALUE_ACTION_TYPE_IP2ME); + mlxsw_reg_ralue_ip2me_v_set(payload, 1); + mlxsw_reg_ralue_ip2me_tunnel_ptr_set(payload, tunnel_ptr); +} + +/* RAUHT - Router Algorithmic LPM Unicast Host Table Register + * ---------------------------------------------------------- + * The RAUHT register is used to configure and query the Unicast Host table in + * devices that implement the Algorithmic LPM. + */ +#define MLXSW_REG_RAUHT_ID 0x8014 +#define MLXSW_REG_RAUHT_LEN 0x74 + +MLXSW_REG_DEFINE(rauht, MLXSW_REG_RAUHT_ID, MLXSW_REG_RAUHT_LEN); + +enum mlxsw_reg_rauht_type { + MLXSW_REG_RAUHT_TYPE_IPV4, + MLXSW_REG_RAUHT_TYPE_IPV6, +}; + +/* reg_rauht_type + * Access: Index + */ +MLXSW_ITEM32(reg, rauht, type, 0x00, 24, 2); + +enum mlxsw_reg_rauht_op { + MLXSW_REG_RAUHT_OP_QUERY_READ = 0, + /* Read operation */ + MLXSW_REG_RAUHT_OP_QUERY_CLEAR_ON_READ = 1, + /* Clear on read operation. Used to read entry and clear + * activity bit. + */ + MLXSW_REG_RAUHT_OP_WRITE_ADD = 0, + /* Add. Used to write a new entry to the table. All R/W fields are + * relevant for new entry. Activity bit is set for new entries. + */ + MLXSW_REG_RAUHT_OP_WRITE_UPDATE = 1, + /* Update action. Used to update an existing route entry and + * only update the following fields: + * trap_action, trap_id, mac, counter_set_type, counter_index + */ + MLXSW_REG_RAUHT_OP_WRITE_CLEAR_ACTIVITY = 2, + /* Clear activity. A bit is cleared for the entry. */ + MLXSW_REG_RAUHT_OP_WRITE_DELETE = 3, + /* Delete entry */ + MLXSW_REG_RAUHT_OP_WRITE_DELETE_ALL = 4, + /* Delete all host entries on a RIF. In this command, dip + * field is reserved. + */ +}; + +/* reg_rauht_op + * Access: OP + */ +MLXSW_ITEM32(reg, rauht, op, 0x00, 20, 3); + +/* reg_rauht_a + * Activity. Set for new entries. Set if a packet lookup has hit on + * the specific entry. + * To clear the a bit, use "clear activity" op. + * Enabled by activity_dis in RGCR + * Access: RO + */ +MLXSW_ITEM32(reg, rauht, a, 0x00, 16, 1); + +/* reg_rauht_rif + * Router Interface + * Access: Index + */ +MLXSW_ITEM32(reg, rauht, rif, 0x00, 0, 16); + +/* reg_rauht_dip* + * Destination address. + * Access: Index + */ +MLXSW_ITEM32(reg, rauht, dip4, 0x1C, 0x0, 32); +MLXSW_ITEM_BUF(reg, rauht, dip6, 0x10, 16); + +enum mlxsw_reg_rauht_trap_action { + MLXSW_REG_RAUHT_TRAP_ACTION_NOP, + MLXSW_REG_RAUHT_TRAP_ACTION_TRAP, + MLXSW_REG_RAUHT_TRAP_ACTION_MIRROR_TO_CPU, + MLXSW_REG_RAUHT_TRAP_ACTION_MIRROR, + MLXSW_REG_RAUHT_TRAP_ACTION_DISCARD_ERRORS, +}; + +/* reg_rauht_trap_action + * Access: RW + */ +MLXSW_ITEM32(reg, rauht, trap_action, 0x60, 28, 4); + +enum mlxsw_reg_rauht_trap_id { + MLXSW_REG_RAUHT_TRAP_ID_RTR_EGRESS0, + MLXSW_REG_RAUHT_TRAP_ID_RTR_EGRESS1, +}; + +/* reg_rauht_trap_id + * Trap ID to be reported to CPU. + * Trap-ID is RTR_EGRESS0 or RTR_EGRESS1. + * For trap_action of NOP, MIRROR and DISCARD_ERROR, + * trap_id is reserved. + * Access: RW + */ +MLXSW_ITEM32(reg, rauht, trap_id, 0x60, 0, 9); + +/* reg_rauht_counter_set_type + * Counter set type for flow counters + * Access: RW + */ +MLXSW_ITEM32(reg, rauht, counter_set_type, 0x68, 24, 8); + +/* reg_rauht_counter_index + * Counter index for flow counters + * Access: RW + */ +MLXSW_ITEM32(reg, rauht, counter_index, 0x68, 0, 24); + +/* reg_rauht_mac + * MAC address. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, rauht, mac, 0x6E, 6); + +static inline void mlxsw_reg_rauht_pack(char *payload, + enum mlxsw_reg_rauht_op op, u16 rif, + const char *mac) +{ + MLXSW_REG_ZERO(rauht, payload); + mlxsw_reg_rauht_op_set(payload, op); + mlxsw_reg_rauht_rif_set(payload, rif); + mlxsw_reg_rauht_mac_memcpy_to(payload, mac); +} + +static inline void mlxsw_reg_rauht_pack4(char *payload, + enum mlxsw_reg_rauht_op op, u16 rif, + const char *mac, u32 dip) +{ + mlxsw_reg_rauht_pack(payload, op, rif, mac); + mlxsw_reg_rauht_dip4_set(payload, dip); +} + +static inline void mlxsw_reg_rauht_pack6(char *payload, + enum mlxsw_reg_rauht_op op, u16 rif, + const char *mac, const char *dip) +{ + mlxsw_reg_rauht_pack(payload, op, rif, mac); + mlxsw_reg_rauht_type_set(payload, MLXSW_REG_RAUHT_TYPE_IPV6); + mlxsw_reg_rauht_dip6_memcpy_to(payload, dip); +} + +static inline void mlxsw_reg_rauht_pack_counter(char *payload, + u64 counter_index) +{ + mlxsw_reg_rauht_counter_index_set(payload, counter_index); + mlxsw_reg_rauht_counter_set_type_set(payload, + MLXSW_REG_FLOW_COUNTER_SET_TYPE_PACKETS_BYTES); +} + +/* RALEU - Router Algorithmic LPM ECMP Update Register + * --------------------------------------------------- + * The register enables updating the ECMP section in the action for multiple + * LPM Unicast entries in a single operation. The update is executed to + * all entries of a {virtual router, protocol} tuple using the same ECMP group. + */ +#define MLXSW_REG_RALEU_ID 0x8015 +#define MLXSW_REG_RALEU_LEN 0x28 + +MLXSW_REG_DEFINE(raleu, MLXSW_REG_RALEU_ID, MLXSW_REG_RALEU_LEN); + +/* reg_raleu_protocol + * Protocol. + * Access: Index + */ +MLXSW_ITEM32(reg, raleu, protocol, 0x00, 24, 4); + +/* reg_raleu_virtual_router + * Virtual Router ID + * Range is 0..cap_max_virtual_routers-1 + * Access: Index + */ +MLXSW_ITEM32(reg, raleu, virtual_router, 0x00, 0, 16); + +/* reg_raleu_adjacency_index + * Adjacency Index used for matching on the existing entries. + * Access: Index + */ +MLXSW_ITEM32(reg, raleu, adjacency_index, 0x10, 0, 24); + +/* reg_raleu_ecmp_size + * ECMP Size used for matching on the existing entries. + * Access: Index + */ +MLXSW_ITEM32(reg, raleu, ecmp_size, 0x14, 0, 13); + +/* reg_raleu_new_adjacency_index + * New Adjacency Index. + * Access: WO + */ +MLXSW_ITEM32(reg, raleu, new_adjacency_index, 0x20, 0, 24); + +/* reg_raleu_new_ecmp_size + * New ECMP Size. + * Access: WO + */ +MLXSW_ITEM32(reg, raleu, new_ecmp_size, 0x24, 0, 13); + +static inline void mlxsw_reg_raleu_pack(char *payload, + enum mlxsw_reg_ralxx_protocol protocol, + u16 virtual_router, + u32 adjacency_index, u16 ecmp_size, + u32 new_adjacency_index, + u16 new_ecmp_size) +{ + MLXSW_REG_ZERO(raleu, payload); + mlxsw_reg_raleu_protocol_set(payload, protocol); + mlxsw_reg_raleu_virtual_router_set(payload, virtual_router); + mlxsw_reg_raleu_adjacency_index_set(payload, adjacency_index); + mlxsw_reg_raleu_ecmp_size_set(payload, ecmp_size); + mlxsw_reg_raleu_new_adjacency_index_set(payload, new_adjacency_index); + mlxsw_reg_raleu_new_ecmp_size_set(payload, new_ecmp_size); +} + +/* RAUHTD - Router Algorithmic LPM Unicast Host Table Dump Register + * ---------------------------------------------------------------- + * The RAUHTD register allows dumping entries from the Router Unicast Host + * Table. For a given session an entry is dumped no more than one time. The + * first RAUHTD access after reset is a new session. A session ends when the + * num_rec response is smaller than num_rec request or for IPv4 when the + * num_entries is smaller than 4. The clear activity affect the current session + * or the last session if a new session has not started. + */ +#define MLXSW_REG_RAUHTD_ID 0x8018 +#define MLXSW_REG_RAUHTD_BASE_LEN 0x20 +#define MLXSW_REG_RAUHTD_REC_LEN 0x20 +#define MLXSW_REG_RAUHTD_REC_MAX_NUM 32 +#define MLXSW_REG_RAUHTD_LEN (MLXSW_REG_RAUHTD_BASE_LEN + \ + MLXSW_REG_RAUHTD_REC_MAX_NUM * MLXSW_REG_RAUHTD_REC_LEN) +#define MLXSW_REG_RAUHTD_IPV4_ENT_PER_REC 4 + +MLXSW_REG_DEFINE(rauhtd, MLXSW_REG_RAUHTD_ID, MLXSW_REG_RAUHTD_LEN); + +#define MLXSW_REG_RAUHTD_FILTER_A BIT(0) +#define MLXSW_REG_RAUHTD_FILTER_RIF BIT(3) + +/* reg_rauhtd_filter_fields + * if a bit is '0' then the relevant field is ignored and dump is done + * regardless of the field value + * Bit0 - filter by activity: entry_a + * Bit3 - filter by entry rip: entry_rif + * Access: Index + */ +MLXSW_ITEM32(reg, rauhtd, filter_fields, 0x00, 0, 8); + +enum mlxsw_reg_rauhtd_op { + MLXSW_REG_RAUHTD_OP_DUMP, + MLXSW_REG_RAUHTD_OP_DUMP_AND_CLEAR, +}; + +/* reg_rauhtd_op + * Access: OP + */ +MLXSW_ITEM32(reg, rauhtd, op, 0x04, 24, 2); + +/* reg_rauhtd_num_rec + * At request: number of records requested + * At response: number of records dumped + * For IPv4, each record has 4 entries at request and up to 4 entries + * at response + * Range is 0..MLXSW_REG_RAUHTD_REC_MAX_NUM + * Access: Index + */ +MLXSW_ITEM32(reg, rauhtd, num_rec, 0x04, 0, 8); + +/* reg_rauhtd_entry_a + * Dump only if activity has value of entry_a + * Reserved if filter_fields bit0 is '0' + * Access: Index + */ +MLXSW_ITEM32(reg, rauhtd, entry_a, 0x08, 16, 1); + +enum mlxsw_reg_rauhtd_type { + MLXSW_REG_RAUHTD_TYPE_IPV4, + MLXSW_REG_RAUHTD_TYPE_IPV6, +}; + +/* reg_rauhtd_type + * Dump only if record type is: + * 0 - IPv4 + * 1 - IPv6 + * Access: Index + */ +MLXSW_ITEM32(reg, rauhtd, type, 0x08, 0, 4); + +/* reg_rauhtd_entry_rif + * Dump only if RIF has value of entry_rif + * Reserved if filter_fields bit3 is '0' + * Access: Index + */ +MLXSW_ITEM32(reg, rauhtd, entry_rif, 0x0C, 0, 16); + +static inline void mlxsw_reg_rauhtd_pack(char *payload, + enum mlxsw_reg_rauhtd_type type) +{ + MLXSW_REG_ZERO(rauhtd, payload); + mlxsw_reg_rauhtd_filter_fields_set(payload, MLXSW_REG_RAUHTD_FILTER_A); + mlxsw_reg_rauhtd_op_set(payload, MLXSW_REG_RAUHTD_OP_DUMP_AND_CLEAR); + mlxsw_reg_rauhtd_num_rec_set(payload, MLXSW_REG_RAUHTD_REC_MAX_NUM); + mlxsw_reg_rauhtd_entry_a_set(payload, 1); + mlxsw_reg_rauhtd_type_set(payload, type); +} + +/* reg_rauhtd_ipv4_rec_num_entries + * Number of valid entries in this record: + * 0 - 1 valid entry + * 1 - 2 valid entries + * 2 - 3 valid entries + * 3 - 4 valid entries + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, rauhtd, ipv4_rec_num_entries, + MLXSW_REG_RAUHTD_BASE_LEN, 28, 2, + MLXSW_REG_RAUHTD_REC_LEN, 0x00, false); + +/* reg_rauhtd_rec_type + * Record type. + * 0 - IPv4 + * 1 - IPv6 + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, rauhtd, rec_type, MLXSW_REG_RAUHTD_BASE_LEN, 24, 2, + MLXSW_REG_RAUHTD_REC_LEN, 0x00, false); + +#define MLXSW_REG_RAUHTD_IPV4_ENT_LEN 0x8 + +/* reg_rauhtd_ipv4_ent_a + * Activity. Set for new entries. Set if a packet lookup has hit on the + * specific entry. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, rauhtd, ipv4_ent_a, MLXSW_REG_RAUHTD_BASE_LEN, 16, 1, + MLXSW_REG_RAUHTD_IPV4_ENT_LEN, 0x00, false); + +/* reg_rauhtd_ipv4_ent_rif + * Router interface. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, rauhtd, ipv4_ent_rif, MLXSW_REG_RAUHTD_BASE_LEN, 0, + 16, MLXSW_REG_RAUHTD_IPV4_ENT_LEN, 0x00, false); + +/* reg_rauhtd_ipv4_ent_dip + * Destination IPv4 address. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, rauhtd, ipv4_ent_dip, MLXSW_REG_RAUHTD_BASE_LEN, 0, + 32, MLXSW_REG_RAUHTD_IPV4_ENT_LEN, 0x04, false); + +#define MLXSW_REG_RAUHTD_IPV6_ENT_LEN 0x20 + +/* reg_rauhtd_ipv6_ent_a + * Activity. Set for new entries. Set if a packet lookup has hit on the + * specific entry. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, rauhtd, ipv6_ent_a, MLXSW_REG_RAUHTD_BASE_LEN, 16, 1, + MLXSW_REG_RAUHTD_IPV6_ENT_LEN, 0x00, false); + +/* reg_rauhtd_ipv6_ent_rif + * Router interface. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, rauhtd, ipv6_ent_rif, MLXSW_REG_RAUHTD_BASE_LEN, 0, + 16, MLXSW_REG_RAUHTD_IPV6_ENT_LEN, 0x00, false); + +/* reg_rauhtd_ipv6_ent_dip + * Destination IPv6 address. + * Access: RO + */ +MLXSW_ITEM_BUF_INDEXED(reg, rauhtd, ipv6_ent_dip, MLXSW_REG_RAUHTD_BASE_LEN, + 16, MLXSW_REG_RAUHTD_IPV6_ENT_LEN, 0x10); + +static inline void mlxsw_reg_rauhtd_ent_ipv4_unpack(char *payload, + int ent_index, u16 *p_rif, + u32 *p_dip) +{ + *p_rif = mlxsw_reg_rauhtd_ipv4_ent_rif_get(payload, ent_index); + *p_dip = mlxsw_reg_rauhtd_ipv4_ent_dip_get(payload, ent_index); +} + +static inline void mlxsw_reg_rauhtd_ent_ipv6_unpack(char *payload, + int rec_index, u16 *p_rif, + char *p_dip) +{ + *p_rif = mlxsw_reg_rauhtd_ipv6_ent_rif_get(payload, rec_index); + mlxsw_reg_rauhtd_ipv6_ent_dip_memcpy_from(payload, rec_index, p_dip); +} + +/* RTDP - Routing Tunnel Decap Properties Register + * ----------------------------------------------- + * The RTDP register is used for configuring the tunnel decap properties of NVE + * and IPinIP. + */ +#define MLXSW_REG_RTDP_ID 0x8020 +#define MLXSW_REG_RTDP_LEN 0x44 + +MLXSW_REG_DEFINE(rtdp, MLXSW_REG_RTDP_ID, MLXSW_REG_RTDP_LEN); + +enum mlxsw_reg_rtdp_type { + MLXSW_REG_RTDP_TYPE_NVE, + MLXSW_REG_RTDP_TYPE_IPIP, +}; + +/* reg_rtdp_type + * Type of the RTDP entry as per enum mlxsw_reg_rtdp_type. + * Access: RW + */ +MLXSW_ITEM32(reg, rtdp, type, 0x00, 28, 4); + +/* reg_rtdp_tunnel_index + * Index to the Decap entry. + * For Spectrum, Index to KVD Linear. + * Access: Index + */ +MLXSW_ITEM32(reg, rtdp, tunnel_index, 0x00, 0, 24); + +/* reg_rtdp_egress_router_interface + * Underlay egress router interface. + * Valid range is from 0 to cap_max_router_interfaces - 1 + * Access: RW + */ +MLXSW_ITEM32(reg, rtdp, egress_router_interface, 0x40, 0, 16); + +/* IPinIP */ + +/* reg_rtdp_ipip_irif + * Ingress Router Interface for the overlay router + * Access: RW + */ +MLXSW_ITEM32(reg, rtdp, ipip_irif, 0x04, 16, 16); + +enum mlxsw_reg_rtdp_ipip_sip_check { + /* No sip checks. */ + MLXSW_REG_RTDP_IPIP_SIP_CHECK_NO, + /* Filter packet if underlay is not IPv4 or if underlay SIP does not + * equal ipv4_usip. + */ + MLXSW_REG_RTDP_IPIP_SIP_CHECK_FILTER_IPV4, + /* Filter packet if underlay is not IPv6 or if underlay SIP does not + * equal ipv6_usip. + */ + MLXSW_REG_RTDP_IPIP_SIP_CHECK_FILTER_IPV6 = 3, +}; + +/* reg_rtdp_ipip_sip_check + * SIP check to perform. If decapsulation failed due to these configurations + * then trap_id is IPIP_DECAP_ERROR. + * Access: RW + */ +MLXSW_ITEM32(reg, rtdp, ipip_sip_check, 0x04, 0, 3); + +/* If set, allow decapsulation of IPinIP (without GRE). */ +#define MLXSW_REG_RTDP_IPIP_TYPE_CHECK_ALLOW_IPIP BIT(0) +/* If set, allow decapsulation of IPinGREinIP without a key. */ +#define MLXSW_REG_RTDP_IPIP_TYPE_CHECK_ALLOW_GRE BIT(1) +/* If set, allow decapsulation of IPinGREinIP with a key. */ +#define MLXSW_REG_RTDP_IPIP_TYPE_CHECK_ALLOW_GRE_KEY BIT(2) + +/* reg_rtdp_ipip_type_check + * Flags as per MLXSW_REG_RTDP_IPIP_TYPE_CHECK_*. If decapsulation failed due to + * these configurations then trap_id is IPIP_DECAP_ERROR. + * Access: RW + */ +MLXSW_ITEM32(reg, rtdp, ipip_type_check, 0x08, 24, 3); + +/* reg_rtdp_ipip_gre_key_check + * Whether GRE key should be checked. When check is enabled: + * - A packet received as IPinIP (without GRE) will always pass. + * - A packet received as IPinGREinIP without a key will not pass the check. + * - A packet received as IPinGREinIP with a key will pass the check only if the + * key in the packet is equal to expected_gre_key. + * If decapsulation failed due to GRE key then trap_id is IPIP_DECAP_ERROR. + * Access: RW + */ +MLXSW_ITEM32(reg, rtdp, ipip_gre_key_check, 0x08, 23, 1); + +/* reg_rtdp_ipip_ipv4_usip + * Underlay IPv4 address for ipv4 source address check. + * Reserved when sip_check is not '1'. + * Access: RW + */ +MLXSW_ITEM32(reg, rtdp, ipip_ipv4_usip, 0x0C, 0, 32); + +/* reg_rtdp_ipip_ipv6_usip_ptr + * This field is valid when sip_check is "sipv6 check explicitly". This is a + * pointer to the IPv6 DIP which is configured by RIPS. For Spectrum, the index + * is to the KVD linear. + * Reserved when sip_check is not MLXSW_REG_RTDP_IPIP_SIP_CHECK_FILTER_IPV6. + * Access: RW + */ +MLXSW_ITEM32(reg, rtdp, ipip_ipv6_usip_ptr, 0x10, 0, 24); + +/* reg_rtdp_ipip_expected_gre_key + * GRE key for checking. + * Reserved when gre_key_check is '0'. + * Access: RW + */ +MLXSW_ITEM32(reg, rtdp, ipip_expected_gre_key, 0x14, 0, 32); + +static inline void mlxsw_reg_rtdp_pack(char *payload, + enum mlxsw_reg_rtdp_type type, + u32 tunnel_index) +{ + MLXSW_REG_ZERO(rtdp, payload); + mlxsw_reg_rtdp_type_set(payload, type); + mlxsw_reg_rtdp_tunnel_index_set(payload, tunnel_index); +} + +static inline void +mlxsw_reg_rtdp_ipip_pack(char *payload, u16 irif, + enum mlxsw_reg_rtdp_ipip_sip_check sip_check, + unsigned int type_check, bool gre_key_check, + u32 expected_gre_key) +{ + mlxsw_reg_rtdp_ipip_irif_set(payload, irif); + mlxsw_reg_rtdp_ipip_sip_check_set(payload, sip_check); + mlxsw_reg_rtdp_ipip_type_check_set(payload, type_check); + mlxsw_reg_rtdp_ipip_gre_key_check_set(payload, gre_key_check); + mlxsw_reg_rtdp_ipip_expected_gre_key_set(payload, expected_gre_key); +} + +static inline void +mlxsw_reg_rtdp_ipip4_pack(char *payload, u16 irif, + enum mlxsw_reg_rtdp_ipip_sip_check sip_check, + unsigned int type_check, bool gre_key_check, + u32 ipv4_usip, u32 expected_gre_key) +{ + mlxsw_reg_rtdp_ipip_pack(payload, irif, sip_check, type_check, + gre_key_check, expected_gre_key); + mlxsw_reg_rtdp_ipip_ipv4_usip_set(payload, ipv4_usip); +} + +static inline void +mlxsw_reg_rtdp_ipip6_pack(char *payload, u16 irif, + enum mlxsw_reg_rtdp_ipip_sip_check sip_check, + unsigned int type_check, bool gre_key_check, + u32 ipv6_usip_ptr, u32 expected_gre_key) +{ + mlxsw_reg_rtdp_ipip_pack(payload, irif, sip_check, type_check, + gre_key_check, expected_gre_key); + mlxsw_reg_rtdp_ipip_ipv6_usip_ptr_set(payload, ipv6_usip_ptr); +} + +/* RIPS - Router IP version Six Register + * ------------------------------------- + * The RIPS register is used to store IPv6 addresses for use by the NVE and + * IPinIP + */ +#define MLXSW_REG_RIPS_ID 0x8021 +#define MLXSW_REG_RIPS_LEN 0x14 + +MLXSW_REG_DEFINE(rips, MLXSW_REG_RIPS_ID, MLXSW_REG_RIPS_LEN); + +/* reg_rips_index + * Index to IPv6 address. + * For Spectrum, the index is to the KVD linear. + * Access: Index + */ +MLXSW_ITEM32(reg, rips, index, 0x00, 0, 24); + +/* reg_rips_ipv6 + * IPv6 address + * Access: RW + */ +MLXSW_ITEM_BUF(reg, rips, ipv6, 0x04, 16); + +static inline void mlxsw_reg_rips_pack(char *payload, u32 index, + const struct in6_addr *ipv6) +{ + MLXSW_REG_ZERO(rips, payload); + mlxsw_reg_rips_index_set(payload, index); + mlxsw_reg_rips_ipv6_memcpy_to(payload, (const char *)ipv6); +} + +/* RATRAD - Router Adjacency Table Activity Dump Register + * ------------------------------------------------------ + * The RATRAD register is used to dump and optionally clear activity bits of + * router adjacency table entries. + */ +#define MLXSW_REG_RATRAD_ID 0x8022 +#define MLXSW_REG_RATRAD_LEN 0x210 + +MLXSW_REG_DEFINE(ratrad, MLXSW_REG_RATRAD_ID, MLXSW_REG_RATRAD_LEN); + +enum { + /* Read activity */ + MLXSW_REG_RATRAD_OP_READ_ACTIVITY, + /* Read and clear activity */ + MLXSW_REG_RATRAD_OP_READ_CLEAR_ACTIVITY, +}; + +/* reg_ratrad_op + * Access: Operation + */ +MLXSW_ITEM32(reg, ratrad, op, 0x00, 30, 2); + +/* reg_ratrad_ecmp_size + * ecmp_size is the amount of sequential entries from adjacency_index. Valid + * ranges: + * Spectrum-1: 32-64, 512, 1024, 2048, 4096 + * Spectrum-2/3: 32-128, 256, 512, 1024, 2048, 4096 + * Access: Index + */ +MLXSW_ITEM32(reg, ratrad, ecmp_size, 0x00, 0, 13); + +/* reg_ratrad_adjacency_index + * Index into the adjacency table. + * Access: Index + */ +MLXSW_ITEM32(reg, ratrad, adjacency_index, 0x04, 0, 24); + +/* reg_ratrad_activity_vector + * Activity bit per adjacency index. + * Bits higher than ecmp_size are reserved. + * Access: RO + */ +MLXSW_ITEM_BIT_ARRAY(reg, ratrad, activity_vector, 0x10, 0x200, 1); + +static inline void mlxsw_reg_ratrad_pack(char *payload, u32 adjacency_index, + u16 ecmp_size) +{ + MLXSW_REG_ZERO(ratrad, payload); + mlxsw_reg_ratrad_op_set(payload, + MLXSW_REG_RATRAD_OP_READ_CLEAR_ACTIVITY); + mlxsw_reg_ratrad_ecmp_size_set(payload, ecmp_size); + mlxsw_reg_ratrad_adjacency_index_set(payload, adjacency_index); +} + +/* RIGR-V2 - Router Interface Group Register Version 2 + * --------------------------------------------------- + * The RIGR_V2 register is used to add, remove and query egress interface list + * of a multicast forwarding entry. + */ +#define MLXSW_REG_RIGR2_ID 0x8023 +#define MLXSW_REG_RIGR2_LEN 0xB0 + +#define MLXSW_REG_RIGR2_MAX_ERIFS 32 + +MLXSW_REG_DEFINE(rigr2, MLXSW_REG_RIGR2_ID, MLXSW_REG_RIGR2_LEN); + +/* reg_rigr2_rigr_index + * KVD Linear index. + * Access: Index + */ +MLXSW_ITEM32(reg, rigr2, rigr_index, 0x04, 0, 24); + +/* reg_rigr2_vnext + * Next RIGR Index is valid. + * Access: RW + */ +MLXSW_ITEM32(reg, rigr2, vnext, 0x08, 31, 1); + +/* reg_rigr2_next_rigr_index + * Next RIGR Index. The index is to the KVD linear. + * Reserved when vnxet = '0'. + * Access: RW + */ +MLXSW_ITEM32(reg, rigr2, next_rigr_index, 0x08, 0, 24); + +/* reg_rigr2_vrmid + * RMID Index is valid. + * Access: RW + */ +MLXSW_ITEM32(reg, rigr2, vrmid, 0x20, 31, 1); + +/* reg_rigr2_rmid_index + * RMID Index. + * Range 0 .. max_mid - 1 + * Reserved when vrmid = '0'. + * The index is to the Port Group Table (PGT) + * Access: RW + */ +MLXSW_ITEM32(reg, rigr2, rmid_index, 0x20, 0, 16); + +/* reg_rigr2_erif_entry_v + * Egress Router Interface is valid. + * Note that low-entries must be set if high-entries are set. For + * example: if erif_entry[2].v is set then erif_entry[1].v and + * erif_entry[0].v must be set. + * Index can be from 0 to cap_mc_erif_list_entries-1 + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, rigr2, erif_entry_v, 0x24, 31, 1, 4, 0, false); + +/* reg_rigr2_erif_entry_erif + * Egress Router Interface. + * Valid range is from 0 to cap_max_router_interfaces - 1 + * Index can be from 0 to MLXSW_REG_RIGR2_MAX_ERIFS - 1 + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, rigr2, erif_entry_erif, 0x24, 0, 16, 4, 0, false); + +static inline void mlxsw_reg_rigr2_pack(char *payload, u32 rigr_index, + bool vnext, u32 next_rigr_index) +{ + MLXSW_REG_ZERO(rigr2, payload); + mlxsw_reg_rigr2_rigr_index_set(payload, rigr_index); + mlxsw_reg_rigr2_vnext_set(payload, vnext); + mlxsw_reg_rigr2_next_rigr_index_set(payload, next_rigr_index); + mlxsw_reg_rigr2_vrmid_set(payload, 0); + mlxsw_reg_rigr2_rmid_index_set(payload, 0); +} + +static inline void mlxsw_reg_rigr2_erif_entry_pack(char *payload, int index, + bool v, u16 erif) +{ + mlxsw_reg_rigr2_erif_entry_v_set(payload, index, v); + mlxsw_reg_rigr2_erif_entry_erif_set(payload, index, erif); +} + +/* RECR-V2 - Router ECMP Configuration Version 2 Register + * ------------------------------------------------------ + */ +#define MLXSW_REG_RECR2_ID 0x8025 +#define MLXSW_REG_RECR2_LEN 0x38 + +MLXSW_REG_DEFINE(recr2, MLXSW_REG_RECR2_ID, MLXSW_REG_RECR2_LEN); + +/* reg_recr2_pp + * Per-port configuration + * Access: Index + */ +MLXSW_ITEM32(reg, recr2, pp, 0x00, 24, 1); + +/* reg_recr2_sh + * Symmetric hash + * Access: RW + */ +MLXSW_ITEM32(reg, recr2, sh, 0x00, 8, 1); + +/* reg_recr2_seed + * Seed + * Access: RW + */ +MLXSW_ITEM32(reg, recr2, seed, 0x08, 0, 32); + +enum { + /* Enable IPv4 fields if packet is not TCP and not UDP */ + MLXSW_REG_RECR2_IPV4_EN_NOT_TCP_NOT_UDP = 3, + /* Enable IPv4 fields if packet is TCP or UDP */ + MLXSW_REG_RECR2_IPV4_EN_TCP_UDP = 4, + /* Enable IPv6 fields if packet is not TCP and not UDP */ + MLXSW_REG_RECR2_IPV6_EN_NOT_TCP_NOT_UDP = 5, + /* Enable IPv6 fields if packet is TCP or UDP */ + MLXSW_REG_RECR2_IPV6_EN_TCP_UDP = 6, + /* Enable TCP/UDP header fields if packet is IPv4 */ + MLXSW_REG_RECR2_TCP_UDP_EN_IPV4 = 7, + /* Enable TCP/UDP header fields if packet is IPv6 */ + MLXSW_REG_RECR2_TCP_UDP_EN_IPV6 = 8, + + __MLXSW_REG_RECR2_HEADER_CNT, +}; + +/* reg_recr2_outer_header_enables + * Bit mask where each bit enables a specific layer to be included in + * the hash calculation. + * Access: RW + */ +MLXSW_ITEM_BIT_ARRAY(reg, recr2, outer_header_enables, 0x10, 0x04, 1); + +enum { + /* IPv4 Source IP */ + MLXSW_REG_RECR2_IPV4_SIP0 = 9, + MLXSW_REG_RECR2_IPV4_SIP3 = 12, + /* IPv4 Destination IP */ + MLXSW_REG_RECR2_IPV4_DIP0 = 13, + MLXSW_REG_RECR2_IPV4_DIP3 = 16, + /* IP Protocol */ + MLXSW_REG_RECR2_IPV4_PROTOCOL = 17, + /* IPv6 Source IP */ + MLXSW_REG_RECR2_IPV6_SIP0_7 = 21, + MLXSW_REG_RECR2_IPV6_SIP8 = 29, + MLXSW_REG_RECR2_IPV6_SIP15 = 36, + /* IPv6 Destination IP */ + MLXSW_REG_RECR2_IPV6_DIP0_7 = 37, + MLXSW_REG_RECR2_IPV6_DIP8 = 45, + MLXSW_REG_RECR2_IPV6_DIP15 = 52, + /* IPv6 Next Header */ + MLXSW_REG_RECR2_IPV6_NEXT_HEADER = 53, + /* IPv6 Flow Label */ + MLXSW_REG_RECR2_IPV6_FLOW_LABEL = 57, + /* TCP/UDP Source Port */ + MLXSW_REG_RECR2_TCP_UDP_SPORT = 74, + /* TCP/UDP Destination Port */ + MLXSW_REG_RECR2_TCP_UDP_DPORT = 75, + + __MLXSW_REG_RECR2_FIELD_CNT, +}; + +/* reg_recr2_outer_header_fields_enable + * Packet fields to enable for ECMP hash subject to outer_header_enable. + * Access: RW + */ +MLXSW_ITEM_BIT_ARRAY(reg, recr2, outer_header_fields_enable, 0x14, 0x14, 1); + +/* reg_recr2_inner_header_enables + * Bit mask where each bit enables a specific inner layer to be included in the + * hash calculation. Same values as reg_recr2_outer_header_enables. + * Access: RW + */ +MLXSW_ITEM_BIT_ARRAY(reg, recr2, inner_header_enables, 0x2C, 0x04, 1); + +enum { + /* Inner IPv4 Source IP */ + MLXSW_REG_RECR2_INNER_IPV4_SIP0 = 3, + MLXSW_REG_RECR2_INNER_IPV4_SIP3 = 6, + /* Inner IPv4 Destination IP */ + MLXSW_REG_RECR2_INNER_IPV4_DIP0 = 7, + MLXSW_REG_RECR2_INNER_IPV4_DIP3 = 10, + /* Inner IP Protocol */ + MLXSW_REG_RECR2_INNER_IPV4_PROTOCOL = 11, + /* Inner IPv6 Source IP */ + MLXSW_REG_RECR2_INNER_IPV6_SIP0_7 = 12, + MLXSW_REG_RECR2_INNER_IPV6_SIP8 = 20, + MLXSW_REG_RECR2_INNER_IPV6_SIP15 = 27, + /* Inner IPv6 Destination IP */ + MLXSW_REG_RECR2_INNER_IPV6_DIP0_7 = 28, + MLXSW_REG_RECR2_INNER_IPV6_DIP8 = 36, + MLXSW_REG_RECR2_INNER_IPV6_DIP15 = 43, + /* Inner IPv6 Next Header */ + MLXSW_REG_RECR2_INNER_IPV6_NEXT_HEADER = 44, + /* Inner IPv6 Flow Label */ + MLXSW_REG_RECR2_INNER_IPV6_FLOW_LABEL = 45, + /* Inner TCP/UDP Source Port */ + MLXSW_REG_RECR2_INNER_TCP_UDP_SPORT = 46, + /* Inner TCP/UDP Destination Port */ + MLXSW_REG_RECR2_INNER_TCP_UDP_DPORT = 47, + + __MLXSW_REG_RECR2_INNER_FIELD_CNT, +}; + +/* reg_recr2_inner_header_fields_enable + * Inner packet fields to enable for ECMP hash subject to inner_header_enables. + * Access: RW + */ +MLXSW_ITEM_BIT_ARRAY(reg, recr2, inner_header_fields_enable, 0x30, 0x08, 1); + +static inline void mlxsw_reg_recr2_pack(char *payload, u32 seed) +{ + MLXSW_REG_ZERO(recr2, payload); + mlxsw_reg_recr2_pp_set(payload, false); + mlxsw_reg_recr2_sh_set(payload, true); + mlxsw_reg_recr2_seed_set(payload, seed); +} + +/* RMFT-V2 - Router Multicast Forwarding Table Version 2 Register + * -------------------------------------------------------------- + * The RMFT_V2 register is used to configure and query the multicast table. + */ +#define MLXSW_REG_RMFT2_ID 0x8027 +#define MLXSW_REG_RMFT2_LEN 0x174 + +MLXSW_REG_DEFINE(rmft2, MLXSW_REG_RMFT2_ID, MLXSW_REG_RMFT2_LEN); + +/* reg_rmft2_v + * Valid + * Access: RW + */ +MLXSW_ITEM32(reg, rmft2, v, 0x00, 31, 1); + +enum mlxsw_reg_rmft2_type { + MLXSW_REG_RMFT2_TYPE_IPV4, + MLXSW_REG_RMFT2_TYPE_IPV6 +}; + +/* reg_rmft2_type + * Access: Index + */ +MLXSW_ITEM32(reg, rmft2, type, 0x00, 28, 2); + +enum mlxsw_sp_reg_rmft2_op { + /* For Write: + * Write operation. Used to write a new entry to the table. All RW + * fields are relevant for new entry. Activity bit is set for new + * entries - Note write with v (Valid) 0 will delete the entry. + * For Query: + * Read operation + */ + MLXSW_REG_RMFT2_OP_READ_WRITE, +}; + +/* reg_rmft2_op + * Operation. + * Access: OP + */ +MLXSW_ITEM32(reg, rmft2, op, 0x00, 20, 2); + +/* reg_rmft2_a + * Activity. Set for new entries. Set if a packet lookup has hit on the specific + * entry. + * Access: RO + */ +MLXSW_ITEM32(reg, rmft2, a, 0x00, 16, 1); + +/* reg_rmft2_offset + * Offset within the multicast forwarding table to write to. + * Access: Index + */ +MLXSW_ITEM32(reg, rmft2, offset, 0x00, 0, 16); + +/* reg_rmft2_virtual_router + * Virtual Router ID. Range from 0..cap_max_virtual_routers-1 + * Access: RW + */ +MLXSW_ITEM32(reg, rmft2, virtual_router, 0x04, 0, 16); + +enum mlxsw_reg_rmft2_irif_mask { + MLXSW_REG_RMFT2_IRIF_MASK_IGNORE, + MLXSW_REG_RMFT2_IRIF_MASK_COMPARE +}; + +/* reg_rmft2_irif_mask + * Ingress RIF mask. + * Access: RW + */ +MLXSW_ITEM32(reg, rmft2, irif_mask, 0x08, 24, 1); + +/* reg_rmft2_irif + * Ingress RIF index. + * Access: RW + */ +MLXSW_ITEM32(reg, rmft2, irif, 0x08, 0, 16); + +/* reg_rmft2_dip{4,6} + * Destination IPv4/6 address + * Access: RW + */ +MLXSW_ITEM_BUF(reg, rmft2, dip6, 0x10, 16); +MLXSW_ITEM32(reg, rmft2, dip4, 0x1C, 0, 32); + +/* reg_rmft2_dip{4,6}_mask + * A bit that is set directs the TCAM to compare the corresponding bit in key. A + * bit that is clear directs the TCAM to ignore the corresponding bit in key. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, rmft2, dip6_mask, 0x20, 16); +MLXSW_ITEM32(reg, rmft2, dip4_mask, 0x2C, 0, 32); + +/* reg_rmft2_sip{4,6} + * Source IPv4/6 address + * Access: RW + */ +MLXSW_ITEM_BUF(reg, rmft2, sip6, 0x30, 16); +MLXSW_ITEM32(reg, rmft2, sip4, 0x3C, 0, 32); + +/* reg_rmft2_sip{4,6}_mask + * A bit that is set directs the TCAM to compare the corresponding bit in key. A + * bit that is clear directs the TCAM to ignore the corresponding bit in key. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, rmft2, sip6_mask, 0x40, 16); +MLXSW_ITEM32(reg, rmft2, sip4_mask, 0x4C, 0, 32); + +/* reg_rmft2_flexible_action_set + * ACL action set. The only supported action types in this field and in any + * action-set pointed from here are as follows: + * 00h: ACTION_NULL + * 01h: ACTION_MAC_TTL, only TTL configuration is supported. + * 03h: ACTION_TRAP + * 06h: ACTION_QOS + * 08h: ACTION_POLICING_MONITORING + * 10h: ACTION_ROUTER_MC + * Access: RW + */ +MLXSW_ITEM_BUF(reg, rmft2, flexible_action_set, 0x80, + MLXSW_REG_FLEX_ACTION_SET_LEN); + +static inline void +mlxsw_reg_rmft2_common_pack(char *payload, bool v, u16 offset, + u16 virtual_router, + enum mlxsw_reg_rmft2_irif_mask irif_mask, u16 irif, + const char *flex_action_set) +{ + MLXSW_REG_ZERO(rmft2, payload); + mlxsw_reg_rmft2_v_set(payload, v); + mlxsw_reg_rmft2_op_set(payload, MLXSW_REG_RMFT2_OP_READ_WRITE); + mlxsw_reg_rmft2_offset_set(payload, offset); + mlxsw_reg_rmft2_virtual_router_set(payload, virtual_router); + mlxsw_reg_rmft2_irif_mask_set(payload, irif_mask); + mlxsw_reg_rmft2_irif_set(payload, irif); + if (flex_action_set) + mlxsw_reg_rmft2_flexible_action_set_memcpy_to(payload, + flex_action_set); +} + +static inline void +mlxsw_reg_rmft2_ipv4_pack(char *payload, bool v, u16 offset, u16 virtual_router, + enum mlxsw_reg_rmft2_irif_mask irif_mask, u16 irif, + u32 dip4, u32 dip4_mask, u32 sip4, u32 sip4_mask, + const char *flexible_action_set) +{ + mlxsw_reg_rmft2_common_pack(payload, v, offset, virtual_router, + irif_mask, irif, flexible_action_set); + mlxsw_reg_rmft2_type_set(payload, MLXSW_REG_RMFT2_TYPE_IPV4); + mlxsw_reg_rmft2_dip4_set(payload, dip4); + mlxsw_reg_rmft2_dip4_mask_set(payload, dip4_mask); + mlxsw_reg_rmft2_sip4_set(payload, sip4); + mlxsw_reg_rmft2_sip4_mask_set(payload, sip4_mask); +} + +static inline void +mlxsw_reg_rmft2_ipv6_pack(char *payload, bool v, u16 offset, u16 virtual_router, + enum mlxsw_reg_rmft2_irif_mask irif_mask, u16 irif, + struct in6_addr dip6, struct in6_addr dip6_mask, + struct in6_addr sip6, struct in6_addr sip6_mask, + const char *flexible_action_set) +{ + mlxsw_reg_rmft2_common_pack(payload, v, offset, virtual_router, + irif_mask, irif, flexible_action_set); + mlxsw_reg_rmft2_type_set(payload, MLXSW_REG_RMFT2_TYPE_IPV6); + mlxsw_reg_rmft2_dip6_memcpy_to(payload, (void *)&dip6); + mlxsw_reg_rmft2_dip6_mask_memcpy_to(payload, (void *)&dip6_mask); + mlxsw_reg_rmft2_sip6_memcpy_to(payload, (void *)&sip6); + mlxsw_reg_rmft2_sip6_mask_memcpy_to(payload, (void *)&sip6_mask); +} + +/* REIV - Router Egress Interface to VID Register + * ---------------------------------------------- + * The REIV register maps {eRIF, egress_port} -> VID. + * This mapping is done at the egress, after the ACLs. + * This mapping always takes effect after router, regardless of cast + * (for unicast/multicast/port-base multicast), regardless of eRIF type and + * regardless of bridge decisions (e.g. SFD for unicast or SMPE). + * Reserved when the RIF is a loopback RIF. + * + * Note: Reserved when legacy bridge model is used. + */ +#define MLXSW_REG_REIV_ID 0x8034 +#define MLXSW_REG_REIV_BASE_LEN 0x20 /* base length, without records */ +#define MLXSW_REG_REIV_REC_LEN 0x04 /* record length */ +#define MLXSW_REG_REIV_REC_MAX_COUNT 256 /* firmware limitation */ +#define MLXSW_REG_REIV_LEN (MLXSW_REG_REIV_BASE_LEN + \ + MLXSW_REG_REIV_REC_LEN * \ + MLXSW_REG_REIV_REC_MAX_COUNT) + +MLXSW_REG_DEFINE(reiv, MLXSW_REG_REIV_ID, MLXSW_REG_REIV_LEN); + +/* reg_reiv_port_page + * Port page - elport_record[0] is 256*port_page. + * Access: Index + */ +MLXSW_ITEM32(reg, reiv, port_page, 0x00, 0, 4); + +/* reg_reiv_erif + * Egress RIF. + * Range is 0..cap_max_router_interfaces-1. + * Access: Index + */ +MLXSW_ITEM32(reg, reiv, erif, 0x04, 0, 16); + +/* reg_reiv_rec_update + * Update enable (when write): + * 0 - Do not update the entry. + * 1 - Update the entry. + * Access: OP + */ +MLXSW_ITEM32_INDEXED(reg, reiv, rec_update, MLXSW_REG_REIV_BASE_LEN, 31, 1, + MLXSW_REG_REIV_REC_LEN, 0x00, false); + +/* reg_reiv_rec_evid + * Egress VID. + * Range is 0..4095. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, reiv, rec_evid, MLXSW_REG_REIV_BASE_LEN, 0, 12, + MLXSW_REG_REIV_REC_LEN, 0x00, false); + +static inline void mlxsw_reg_reiv_pack(char *payload, u8 port_page, u16 erif) +{ + MLXSW_REG_ZERO(reiv, payload); + mlxsw_reg_reiv_port_page_set(payload, port_page); + mlxsw_reg_reiv_erif_set(payload, erif); +} + +/* MFCR - Management Fan Control Register + * -------------------------------------- + * This register controls the settings of the Fan Speed PWM mechanism. + */ +#define MLXSW_REG_MFCR_ID 0x9001 +#define MLXSW_REG_MFCR_LEN 0x08 + +MLXSW_REG_DEFINE(mfcr, MLXSW_REG_MFCR_ID, MLXSW_REG_MFCR_LEN); + +enum mlxsw_reg_mfcr_pwm_frequency { + MLXSW_REG_MFCR_PWM_FEQ_11HZ = 0x00, + MLXSW_REG_MFCR_PWM_FEQ_14_7HZ = 0x01, + MLXSW_REG_MFCR_PWM_FEQ_22_1HZ = 0x02, + MLXSW_REG_MFCR_PWM_FEQ_1_4KHZ = 0x40, + MLXSW_REG_MFCR_PWM_FEQ_5KHZ = 0x41, + MLXSW_REG_MFCR_PWM_FEQ_20KHZ = 0x42, + MLXSW_REG_MFCR_PWM_FEQ_22_5KHZ = 0x43, + MLXSW_REG_MFCR_PWM_FEQ_25KHZ = 0x44, +}; + +/* reg_mfcr_pwm_frequency + * Controls the frequency of the PWM signal. + * Access: RW + */ +MLXSW_ITEM32(reg, mfcr, pwm_frequency, 0x00, 0, 7); + +#define MLXSW_MFCR_TACHOS_MAX 10 + +/* reg_mfcr_tacho_active + * Indicates which of the tachometer is active (bit per tachometer). + * Access: RO + */ +MLXSW_ITEM32(reg, mfcr, tacho_active, 0x04, 16, MLXSW_MFCR_TACHOS_MAX); + +#define MLXSW_MFCR_PWMS_MAX 5 + +/* reg_mfcr_pwm_active + * Indicates which of the PWM control is active (bit per PWM). + * Access: RO + */ +MLXSW_ITEM32(reg, mfcr, pwm_active, 0x04, 0, MLXSW_MFCR_PWMS_MAX); + +static inline void +mlxsw_reg_mfcr_pack(char *payload, + enum mlxsw_reg_mfcr_pwm_frequency pwm_frequency) +{ + MLXSW_REG_ZERO(mfcr, payload); + mlxsw_reg_mfcr_pwm_frequency_set(payload, pwm_frequency); +} + +static inline void +mlxsw_reg_mfcr_unpack(char *payload, + enum mlxsw_reg_mfcr_pwm_frequency *p_pwm_frequency, + u16 *p_tacho_active, u8 *p_pwm_active) +{ + *p_pwm_frequency = mlxsw_reg_mfcr_pwm_frequency_get(payload); + *p_tacho_active = mlxsw_reg_mfcr_tacho_active_get(payload); + *p_pwm_active = mlxsw_reg_mfcr_pwm_active_get(payload); +} + +/* MFSC - Management Fan Speed Control Register + * -------------------------------------------- + * This register controls the settings of the Fan Speed PWM mechanism. + */ +#define MLXSW_REG_MFSC_ID 0x9002 +#define MLXSW_REG_MFSC_LEN 0x08 + +MLXSW_REG_DEFINE(mfsc, MLXSW_REG_MFSC_ID, MLXSW_REG_MFSC_LEN); + +/* reg_mfsc_pwm + * Fan pwm to control / monitor. + * Access: Index + */ +MLXSW_ITEM32(reg, mfsc, pwm, 0x00, 24, 3); + +/* reg_mfsc_pwm_duty_cycle + * Controls the duty cycle of the PWM. Value range from 0..255 to + * represent duty cycle of 0%...100%. + * Access: RW + */ +MLXSW_ITEM32(reg, mfsc, pwm_duty_cycle, 0x04, 0, 8); + +static inline void mlxsw_reg_mfsc_pack(char *payload, u8 pwm, + u8 pwm_duty_cycle) +{ + MLXSW_REG_ZERO(mfsc, payload); + mlxsw_reg_mfsc_pwm_set(payload, pwm); + mlxsw_reg_mfsc_pwm_duty_cycle_set(payload, pwm_duty_cycle); +} + +/* MFSM - Management Fan Speed Measurement + * --------------------------------------- + * This register controls the settings of the Tacho measurements and + * enables reading the Tachometer measurements. + */ +#define MLXSW_REG_MFSM_ID 0x9003 +#define MLXSW_REG_MFSM_LEN 0x08 + +MLXSW_REG_DEFINE(mfsm, MLXSW_REG_MFSM_ID, MLXSW_REG_MFSM_LEN); + +/* reg_mfsm_tacho + * Fan tachometer index. + * Access: Index + */ +MLXSW_ITEM32(reg, mfsm, tacho, 0x00, 24, 4); + +/* reg_mfsm_rpm + * Fan speed (round per minute). + * Access: RO + */ +MLXSW_ITEM32(reg, mfsm, rpm, 0x04, 0, 16); + +static inline void mlxsw_reg_mfsm_pack(char *payload, u8 tacho) +{ + MLXSW_REG_ZERO(mfsm, payload); + mlxsw_reg_mfsm_tacho_set(payload, tacho); +} + +/* MFSL - Management Fan Speed Limit Register + * ------------------------------------------ + * The Fan Speed Limit register is used to configure the fan speed + * event / interrupt notification mechanism. Fan speed threshold are + * defined for both under-speed and over-speed. + */ +#define MLXSW_REG_MFSL_ID 0x9004 +#define MLXSW_REG_MFSL_LEN 0x0C + +MLXSW_REG_DEFINE(mfsl, MLXSW_REG_MFSL_ID, MLXSW_REG_MFSL_LEN); + +/* reg_mfsl_tacho + * Fan tachometer index. + * Access: Index + */ +MLXSW_ITEM32(reg, mfsl, tacho, 0x00, 24, 4); + +/* reg_mfsl_tach_min + * Tachometer minimum value (minimum RPM). + * Access: RW + */ +MLXSW_ITEM32(reg, mfsl, tach_min, 0x04, 0, 16); + +/* reg_mfsl_tach_max + * Tachometer maximum value (maximum RPM). + * Access: RW + */ +MLXSW_ITEM32(reg, mfsl, tach_max, 0x08, 0, 16); + +static inline void mlxsw_reg_mfsl_pack(char *payload, u8 tacho, + u16 tach_min, u16 tach_max) +{ + MLXSW_REG_ZERO(mfsl, payload); + mlxsw_reg_mfsl_tacho_set(payload, tacho); + mlxsw_reg_mfsl_tach_min_set(payload, tach_min); + mlxsw_reg_mfsl_tach_max_set(payload, tach_max); +} + +static inline void mlxsw_reg_mfsl_unpack(char *payload, u8 tacho, + u16 *p_tach_min, u16 *p_tach_max) +{ + if (p_tach_min) + *p_tach_min = mlxsw_reg_mfsl_tach_min_get(payload); + + if (p_tach_max) + *p_tach_max = mlxsw_reg_mfsl_tach_max_get(payload); +} + +/* FORE - Fan Out of Range Event Register + * -------------------------------------- + * This register reports the status of the controlled fans compared to the + * range defined by the MFSL register. + */ +#define MLXSW_REG_FORE_ID 0x9007 +#define MLXSW_REG_FORE_LEN 0x0C + +MLXSW_REG_DEFINE(fore, MLXSW_REG_FORE_ID, MLXSW_REG_FORE_LEN); + +/* fan_under_limit + * Fan speed is below the low limit defined in MFSL register. Each bit relates + * to a single tachometer and indicates the specific tachometer reading is + * below the threshold. + * Access: RO + */ +MLXSW_ITEM32(reg, fore, fan_under_limit, 0x00, 16, 10); + +static inline void mlxsw_reg_fore_unpack(char *payload, u8 tacho, + bool *fault) +{ + u16 limit; + + if (fault) { + limit = mlxsw_reg_fore_fan_under_limit_get(payload); + *fault = limit & BIT(tacho); + } +} + +/* MTCAP - Management Temperature Capabilities + * ------------------------------------------- + * This register exposes the capabilities of the device and + * system temperature sensing. + */ +#define MLXSW_REG_MTCAP_ID 0x9009 +#define MLXSW_REG_MTCAP_LEN 0x08 + +MLXSW_REG_DEFINE(mtcap, MLXSW_REG_MTCAP_ID, MLXSW_REG_MTCAP_LEN); + +/* reg_mtcap_sensor_count + * Number of sensors supported by the device. + * This includes the QSFP module sensors (if exists in the QSFP module). + * Access: RO + */ +MLXSW_ITEM32(reg, mtcap, sensor_count, 0x00, 0, 7); + +/* MTMP - Management Temperature + * ----------------------------- + * This register controls the settings of the temperature measurements + * and enables reading the temperature measurements. Note that temperature + * is in 0.125 degrees Celsius. + */ +#define MLXSW_REG_MTMP_ID 0x900A +#define MLXSW_REG_MTMP_LEN 0x20 + +MLXSW_REG_DEFINE(mtmp, MLXSW_REG_MTMP_ID, MLXSW_REG_MTMP_LEN); + +/* reg_mtmp_slot_index + * Slot index (0: Main board). + * Access: Index + */ +MLXSW_ITEM32(reg, mtmp, slot_index, 0x00, 16, 4); + +#define MLXSW_REG_MTMP_MODULE_INDEX_MIN 64 +#define MLXSW_REG_MTMP_GBOX_INDEX_MIN 256 +/* reg_mtmp_sensor_index + * Sensors index to access. + * 64-127 of sensor_index are mapped to the SFP+/QSFP modules sequentially + * (module 0 is mapped to sensor_index 64). + * Access: Index + */ +MLXSW_ITEM32(reg, mtmp, sensor_index, 0x00, 0, 12); + +/* Convert to milli degrees Celsius */ +#define MLXSW_REG_MTMP_TEMP_TO_MC(val) ({ typeof(val) v_ = (val); \ + ((v_) >= 0) ? ((v_) * 125) : \ + ((s16)((GENMASK(15, 0) + (v_) + 1) \ + * 125)); }) + +/* reg_mtmp_max_operational_temperature + * The highest temperature in the nominal operational range. Reading is in + * 0.125 Celsius degrees units. + * In case of module this is SFF critical temperature threshold. + * Access: RO + */ +MLXSW_ITEM32(reg, mtmp, max_operational_temperature, 0x04, 16, 16); + +/* reg_mtmp_temperature + * Temperature reading from the sensor. Reading is in 0.125 Celsius + * degrees units. + * Access: RO + */ +MLXSW_ITEM32(reg, mtmp, temperature, 0x04, 0, 16); + +/* reg_mtmp_mte + * Max Temperature Enable - enables measuring the max temperature on a sensor. + * Access: RW + */ +MLXSW_ITEM32(reg, mtmp, mte, 0x08, 31, 1); + +/* reg_mtmp_mtr + * Max Temperature Reset - clears the value of the max temperature register. + * Access: WO + */ +MLXSW_ITEM32(reg, mtmp, mtr, 0x08, 30, 1); + +/* reg_mtmp_max_temperature + * The highest measured temperature from the sensor. + * When the bit mte is cleared, the field max_temperature is reserved. + * Access: RO + */ +MLXSW_ITEM32(reg, mtmp, max_temperature, 0x08, 0, 16); + +/* reg_mtmp_tee + * Temperature Event Enable. + * 0 - Do not generate event + * 1 - Generate event + * 2 - Generate single event + * Access: RW + */ + +enum mlxsw_reg_mtmp_tee { + MLXSW_REG_MTMP_TEE_NO_EVENT, + MLXSW_REG_MTMP_TEE_GENERATE_EVENT, + MLXSW_REG_MTMP_TEE_GENERATE_SINGLE_EVENT, +}; + +MLXSW_ITEM32(reg, mtmp, tee, 0x0C, 30, 2); + +#define MLXSW_REG_MTMP_THRESH_HI 0x348 /* 105 Celsius */ + +/* reg_mtmp_temperature_threshold_hi + * High threshold for Temperature Warning Event. In 0.125 Celsius. + * Access: RW + */ +MLXSW_ITEM32(reg, mtmp, temperature_threshold_hi, 0x0C, 0, 16); + +#define MLXSW_REG_MTMP_HYSTERESIS_TEMP 0x28 /* 5 Celsius */ +/* reg_mtmp_temperature_threshold_lo + * Low threshold for Temperature Warning Event. In 0.125 Celsius. + * Access: RW + */ +MLXSW_ITEM32(reg, mtmp, temperature_threshold_lo, 0x10, 0, 16); + +#define MLXSW_REG_MTMP_SENSOR_NAME_SIZE 8 + +/* reg_mtmp_sensor_name + * Sensor Name + * Access: RO + */ +MLXSW_ITEM_BUF(reg, mtmp, sensor_name, 0x18, MLXSW_REG_MTMP_SENSOR_NAME_SIZE); + +static inline void mlxsw_reg_mtmp_pack(char *payload, u8 slot_index, + u16 sensor_index, bool max_temp_enable, + bool max_temp_reset) +{ + MLXSW_REG_ZERO(mtmp, payload); + mlxsw_reg_mtmp_slot_index_set(payload, slot_index); + mlxsw_reg_mtmp_sensor_index_set(payload, sensor_index); + mlxsw_reg_mtmp_mte_set(payload, max_temp_enable); + mlxsw_reg_mtmp_mtr_set(payload, max_temp_reset); + mlxsw_reg_mtmp_temperature_threshold_hi_set(payload, + MLXSW_REG_MTMP_THRESH_HI); +} + +static inline void mlxsw_reg_mtmp_unpack(char *payload, int *p_temp, + int *p_max_temp, int *p_temp_hi, + int *p_max_oper_temp, + char *sensor_name) +{ + s16 temp; + + if (p_temp) { + temp = mlxsw_reg_mtmp_temperature_get(payload); + *p_temp = MLXSW_REG_MTMP_TEMP_TO_MC(temp); + } + if (p_max_temp) { + temp = mlxsw_reg_mtmp_max_temperature_get(payload); + *p_max_temp = MLXSW_REG_MTMP_TEMP_TO_MC(temp); + } + if (p_temp_hi) { + temp = mlxsw_reg_mtmp_temperature_threshold_hi_get(payload); + *p_temp_hi = MLXSW_REG_MTMP_TEMP_TO_MC(temp); + } + if (p_max_oper_temp) { + temp = mlxsw_reg_mtmp_max_operational_temperature_get(payload); + *p_max_oper_temp = MLXSW_REG_MTMP_TEMP_TO_MC(temp); + } + if (sensor_name) + mlxsw_reg_mtmp_sensor_name_memcpy_from(payload, sensor_name); +} + +/* MTWE - Management Temperature Warning Event + * ------------------------------------------- + * This register is used for over temperature warning. + */ +#define MLXSW_REG_MTWE_ID 0x900B +#define MLXSW_REG_MTWE_LEN 0x10 + +MLXSW_REG_DEFINE(mtwe, MLXSW_REG_MTWE_ID, MLXSW_REG_MTWE_LEN); + +/* reg_mtwe_sensor_warning + * Bit vector indicating which of the sensor reading is above threshold. + * Address 00h bit31 is sensor_warning[127]. + * Address 0Ch bit0 is sensor_warning[0]. + * Access: RO + */ +MLXSW_ITEM_BIT_ARRAY(reg, mtwe, sensor_warning, 0x0, 0x10, 1); + +/* MTBR - Management Temperature Bulk Register + * ------------------------------------------- + * This register is used for bulk temperature reading. + */ +#define MLXSW_REG_MTBR_ID 0x900F +#define MLXSW_REG_MTBR_BASE_LEN 0x10 /* base length, without records */ +#define MLXSW_REG_MTBR_REC_LEN 0x04 /* record length */ +#define MLXSW_REG_MTBR_REC_MAX_COUNT 47 /* firmware limitation */ +#define MLXSW_REG_MTBR_LEN (MLXSW_REG_MTBR_BASE_LEN + \ + MLXSW_REG_MTBR_REC_LEN * \ + MLXSW_REG_MTBR_REC_MAX_COUNT) + +MLXSW_REG_DEFINE(mtbr, MLXSW_REG_MTBR_ID, MLXSW_REG_MTBR_LEN); + +/* reg_mtbr_slot_index + * Slot index (0: Main board). + * Access: Index + */ +MLXSW_ITEM32(reg, mtbr, slot_index, 0x00, 16, 4); + +/* reg_mtbr_base_sensor_index + * Base sensors index to access (0 - ASIC sensor, 1-63 - ambient sensors, + * 64-127 are mapped to the SFP+/QSFP modules sequentially). + * Access: Index + */ +MLXSW_ITEM32(reg, mtbr, base_sensor_index, 0x00, 0, 12); + +/* reg_mtbr_num_rec + * Request: Number of records to read + * Response: Number of records read + * See above description for more details. + * Range 1..255 + * Access: RW + */ +MLXSW_ITEM32(reg, mtbr, num_rec, 0x04, 0, 8); + +/* reg_mtbr_rec_max_temp + * The highest measured temperature from the sensor. + * When the bit mte is cleared, the field max_temperature is reserved. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, mtbr, rec_max_temp, MLXSW_REG_MTBR_BASE_LEN, 16, + 16, MLXSW_REG_MTBR_REC_LEN, 0x00, false); + +/* reg_mtbr_rec_temp + * Temperature reading from the sensor. Reading is in 0..125 Celsius + * degrees units. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, mtbr, rec_temp, MLXSW_REG_MTBR_BASE_LEN, 0, 16, + MLXSW_REG_MTBR_REC_LEN, 0x00, false); + +static inline void mlxsw_reg_mtbr_pack(char *payload, u8 slot_index, + u16 base_sensor_index, u8 num_rec) +{ + MLXSW_REG_ZERO(mtbr, payload); + mlxsw_reg_mtbr_slot_index_set(payload, slot_index); + mlxsw_reg_mtbr_base_sensor_index_set(payload, base_sensor_index); + mlxsw_reg_mtbr_num_rec_set(payload, num_rec); +} + +/* Error codes from temperatute reading */ +enum mlxsw_reg_mtbr_temp_status { + MLXSW_REG_MTBR_NO_CONN = 0x8000, + MLXSW_REG_MTBR_NO_TEMP_SENS = 0x8001, + MLXSW_REG_MTBR_INDEX_NA = 0x8002, + MLXSW_REG_MTBR_BAD_SENS_INFO = 0x8003, +}; + +/* Base index for reading modules temperature */ +#define MLXSW_REG_MTBR_BASE_MODULE_INDEX 64 + +static inline void mlxsw_reg_mtbr_temp_unpack(char *payload, int rec_ind, + u16 *p_temp, u16 *p_max_temp) +{ + if (p_temp) + *p_temp = mlxsw_reg_mtbr_rec_temp_get(payload, rec_ind); + if (p_max_temp) + *p_max_temp = mlxsw_reg_mtbr_rec_max_temp_get(payload, rec_ind); +} + +/* MCIA - Management Cable Info Access + * ----------------------------------- + * MCIA register is used to access the SFP+ and QSFP connector's EPROM. + */ + +#define MLXSW_REG_MCIA_ID 0x9014 +#define MLXSW_REG_MCIA_LEN 0x40 + +MLXSW_REG_DEFINE(mcia, MLXSW_REG_MCIA_ID, MLXSW_REG_MCIA_LEN); + +/* reg_mcia_l + * Lock bit. Setting this bit will lock the access to the specific + * cable. Used for updating a full page in a cable EPROM. Any access + * other then subsequence writes will fail while the port is locked. + * Access: RW + */ +MLXSW_ITEM32(reg, mcia, l, 0x00, 31, 1); + +/* reg_mcia_module + * Module number. + * Access: Index + */ +MLXSW_ITEM32(reg, mcia, module, 0x00, 16, 8); + +/* reg_mcia_slot_index + * Slot index (0: Main board) + * Access: Index + */ +MLXSW_ITEM32(reg, mcia, slot, 0x00, 12, 4); + +enum { + MLXSW_REG_MCIA_STATUS_GOOD = 0, + /* No response from module's EEPROM. */ + MLXSW_REG_MCIA_STATUS_NO_EEPROM_MODULE = 1, + /* Module type not supported by the device. */ + MLXSW_REG_MCIA_STATUS_MODULE_NOT_SUPPORTED = 2, + /* No module present indication. */ + MLXSW_REG_MCIA_STATUS_MODULE_NOT_CONNECTED = 3, + /* Error occurred while trying to access module's EEPROM using I2C. */ + MLXSW_REG_MCIA_STATUS_I2C_ERROR = 9, + /* Module is disabled. */ + MLXSW_REG_MCIA_STATUS_MODULE_DISABLED = 16, +}; + +/* reg_mcia_status + * Module status. + * Access: RO + */ +MLXSW_ITEM32(reg, mcia, status, 0x00, 0, 8); + +/* reg_mcia_i2c_device_address + * I2C device address. + * Access: RW + */ +MLXSW_ITEM32(reg, mcia, i2c_device_address, 0x04, 24, 8); + +/* reg_mcia_page_number + * Page number. + * Access: RW + */ +MLXSW_ITEM32(reg, mcia, page_number, 0x04, 16, 8); + +/* reg_mcia_device_address + * Device address. + * Access: RW + */ +MLXSW_ITEM32(reg, mcia, device_address, 0x04, 0, 16); + +/* reg_mcia_bank_number + * Bank number. + * Access: Index + */ +MLXSW_ITEM32(reg, mcia, bank_number, 0x08, 16, 8); + +/* reg_mcia_size + * Number of bytes to read/write (up to 48 bytes). + * Access: RW + */ +MLXSW_ITEM32(reg, mcia, size, 0x08, 0, 16); + +#define MLXSW_REG_MCIA_EEPROM_PAGE_LENGTH 256 +#define MLXSW_REG_MCIA_EEPROM_UP_PAGE_LENGTH 128 +#define MLXSW_REG_MCIA_EEPROM_SIZE 48 +#define MLXSW_REG_MCIA_I2C_ADDR_LOW 0x50 +#define MLXSW_REG_MCIA_I2C_ADDR_HIGH 0x51 +#define MLXSW_REG_MCIA_PAGE0_LO_OFF 0xa0 +#define MLXSW_REG_MCIA_TH_ITEM_SIZE 2 +#define MLXSW_REG_MCIA_TH_PAGE_NUM 3 +#define MLXSW_REG_MCIA_TH_PAGE_CMIS_NUM 2 +#define MLXSW_REG_MCIA_PAGE0_LO 0 +#define MLXSW_REG_MCIA_TH_PAGE_OFF 0x80 +#define MLXSW_REG_MCIA_EEPROM_CMIS_FLAT_MEMORY BIT(7) + +enum mlxsw_reg_mcia_eeprom_module_info_rev_id { + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_REV_ID_UNSPC = 0x00, + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_REV_ID_8436 = 0x01, + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_REV_ID_8636 = 0x03, +}; + +enum mlxsw_reg_mcia_eeprom_module_info_id { + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_SFP = 0x03, + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP = 0x0C, + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP_PLUS = 0x0D, + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP28 = 0x11, + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_QSFP_DD = 0x18, + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID_OSFP = 0x19, +}; + +enum mlxsw_reg_mcia_eeprom_module_info { + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_ID, + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_REV_ID, + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_TYPE_ID, + MLXSW_REG_MCIA_EEPROM_MODULE_INFO_SIZE, +}; + +/* reg_mcia_eeprom + * Bytes to read/write. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, mcia, eeprom, 0x10, MLXSW_REG_MCIA_EEPROM_SIZE); + +/* This is used to access the optional upper pages (1-3) in the QSFP+ + * memory map. Page 1 is available on offset 256 through 383, page 2 - + * on offset 384 through 511, page 3 - on offset 512 through 639. + */ +#define MLXSW_REG_MCIA_PAGE_GET(off) (((off) - \ + MLXSW_REG_MCIA_EEPROM_PAGE_LENGTH) / \ + MLXSW_REG_MCIA_EEPROM_UP_PAGE_LENGTH + 1) + +static inline void mlxsw_reg_mcia_pack(char *payload, u8 slot_index, u8 module, + u8 lock, u8 page_number, + u16 device_addr, u8 size, + u8 i2c_device_addr) +{ + MLXSW_REG_ZERO(mcia, payload); + mlxsw_reg_mcia_slot_set(payload, slot_index); + mlxsw_reg_mcia_module_set(payload, module); + mlxsw_reg_mcia_l_set(payload, lock); + mlxsw_reg_mcia_page_number_set(payload, page_number); + mlxsw_reg_mcia_device_address_set(payload, device_addr); + mlxsw_reg_mcia_size_set(payload, size); + mlxsw_reg_mcia_i2c_device_address_set(payload, i2c_device_addr); +} + +/* MPAT - Monitoring Port Analyzer Table + * ------------------------------------- + * MPAT Register is used to query and configure the Switch PortAnalyzer Table. + * For an enabled analyzer, all fields except e (enable) cannot be modified. + */ +#define MLXSW_REG_MPAT_ID 0x901A +#define MLXSW_REG_MPAT_LEN 0x78 + +MLXSW_REG_DEFINE(mpat, MLXSW_REG_MPAT_ID, MLXSW_REG_MPAT_LEN); + +/* reg_mpat_pa_id + * Port Analyzer ID. + * Access: Index + */ +MLXSW_ITEM32(reg, mpat, pa_id, 0x00, 28, 4); + +/* reg_mpat_session_id + * Mirror Session ID. + * Used for MIRROR_SESSION<i> trap. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, session_id, 0x00, 24, 4); + +/* reg_mpat_system_port + * A unique port identifier for the final destination of the packet. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, system_port, 0x00, 0, 16); + +/* reg_mpat_e + * Enable. Indicating the Port Analyzer is enabled. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, e, 0x04, 31, 1); + +/* reg_mpat_qos + * Quality Of Service Mode. + * 0: CONFIGURED - QoS parameters (Switch Priority, and encapsulation + * PCP, DEI, DSCP or VL) are configured. + * 1: MAINTAIN - QoS parameters (Switch Priority, Color) are the + * same as in the original packet that has triggered the mirroring. For + * SPAN also the pcp,dei are maintained. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, qos, 0x04, 26, 1); + +/* reg_mpat_be + * Best effort mode. Indicates mirroring traffic should not cause packet + * drop or back pressure, but will discard the mirrored packets. Mirrored + * packets will be forwarded on a best effort manner. + * 0: Do not discard mirrored packets + * 1: Discard mirrored packets if causing congestion + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, be, 0x04, 25, 1); + +enum mlxsw_reg_mpat_span_type { + /* Local SPAN Ethernet. + * The original packet is not encapsulated. + */ + MLXSW_REG_MPAT_SPAN_TYPE_LOCAL_ETH = 0x0, + + /* Remote SPAN Ethernet VLAN. + * The packet is forwarded to the monitoring port on the monitoring + * VLAN. + */ + MLXSW_REG_MPAT_SPAN_TYPE_REMOTE_ETH = 0x1, + + /* Encapsulated Remote SPAN Ethernet L3 GRE. + * The packet is encapsulated with GRE header. + */ + MLXSW_REG_MPAT_SPAN_TYPE_REMOTE_ETH_L3 = 0x3, +}; + +/* reg_mpat_span_type + * SPAN type. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, span_type, 0x04, 0, 4); + +/* reg_mpat_pide + * Policer enable. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, pide, 0x0C, 15, 1); + +/* reg_mpat_pid + * Policer ID. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, pid, 0x0C, 0, 14); + +/* Remote SPAN - Ethernet VLAN + * - - - - - - - - - - - - - - + */ + +/* reg_mpat_eth_rspan_vid + * Encapsulation header VLAN ID. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, eth_rspan_vid, 0x18, 0, 12); + +/* Encapsulated Remote SPAN - Ethernet L2 + * - - - - - - - - - - - - - - - - - - - + */ + +enum mlxsw_reg_mpat_eth_rspan_version { + MLXSW_REG_MPAT_ETH_RSPAN_VERSION_NO_HEADER = 15, +}; + +/* reg_mpat_eth_rspan_version + * RSPAN mirror header version. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, eth_rspan_version, 0x10, 18, 4); + +/* reg_mpat_eth_rspan_mac + * Destination MAC address. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, mpat, eth_rspan_mac, 0x12, 6); + +/* reg_mpat_eth_rspan_tp + * Tag Packet. Indicates whether the mirroring header should be VLAN tagged. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, eth_rspan_tp, 0x18, 16, 1); + +/* Encapsulated Remote SPAN - Ethernet L3 + * - - - - - - - - - - - - - - - - - - - + */ + +enum mlxsw_reg_mpat_eth_rspan_protocol { + MLXSW_REG_MPAT_ETH_RSPAN_PROTOCOL_IPV4, + MLXSW_REG_MPAT_ETH_RSPAN_PROTOCOL_IPV6, +}; + +/* reg_mpat_eth_rspan_protocol + * SPAN encapsulation protocol. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, eth_rspan_protocol, 0x18, 24, 4); + +/* reg_mpat_eth_rspan_ttl + * Encapsulation header Time-to-Live/HopLimit. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, eth_rspan_ttl, 0x1C, 4, 8); + +/* reg_mpat_eth_rspan_smac + * Source MAC address + * Access: RW + */ +MLXSW_ITEM_BUF(reg, mpat, eth_rspan_smac, 0x22, 6); + +/* reg_mpat_eth_rspan_dip* + * Destination IP address. The IP version is configured by protocol. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, eth_rspan_dip4, 0x4C, 0, 32); +MLXSW_ITEM_BUF(reg, mpat, eth_rspan_dip6, 0x40, 16); + +/* reg_mpat_eth_rspan_sip* + * Source IP address. The IP version is configured by protocol. + * Access: RW + */ +MLXSW_ITEM32(reg, mpat, eth_rspan_sip4, 0x5C, 0, 32); +MLXSW_ITEM_BUF(reg, mpat, eth_rspan_sip6, 0x50, 16); + +static inline void mlxsw_reg_mpat_pack(char *payload, u8 pa_id, + u16 system_port, bool e, + enum mlxsw_reg_mpat_span_type span_type) +{ + MLXSW_REG_ZERO(mpat, payload); + mlxsw_reg_mpat_pa_id_set(payload, pa_id); + mlxsw_reg_mpat_system_port_set(payload, system_port); + mlxsw_reg_mpat_e_set(payload, e); + mlxsw_reg_mpat_qos_set(payload, 1); + mlxsw_reg_mpat_be_set(payload, 1); + mlxsw_reg_mpat_span_type_set(payload, span_type); +} + +static inline void mlxsw_reg_mpat_eth_rspan_pack(char *payload, u16 vid) +{ + mlxsw_reg_mpat_eth_rspan_vid_set(payload, vid); +} + +static inline void +mlxsw_reg_mpat_eth_rspan_l2_pack(char *payload, + enum mlxsw_reg_mpat_eth_rspan_version version, + const char *mac, + bool tp) +{ + mlxsw_reg_mpat_eth_rspan_version_set(payload, version); + mlxsw_reg_mpat_eth_rspan_mac_memcpy_to(payload, mac); + mlxsw_reg_mpat_eth_rspan_tp_set(payload, tp); +} + +static inline void +mlxsw_reg_mpat_eth_rspan_l3_ipv4_pack(char *payload, u8 ttl, + const char *smac, + u32 sip, u32 dip) +{ + mlxsw_reg_mpat_eth_rspan_ttl_set(payload, ttl); + mlxsw_reg_mpat_eth_rspan_smac_memcpy_to(payload, smac); + mlxsw_reg_mpat_eth_rspan_protocol_set(payload, + MLXSW_REG_MPAT_ETH_RSPAN_PROTOCOL_IPV4); + mlxsw_reg_mpat_eth_rspan_sip4_set(payload, sip); + mlxsw_reg_mpat_eth_rspan_dip4_set(payload, dip); +} + +static inline void +mlxsw_reg_mpat_eth_rspan_l3_ipv6_pack(char *payload, u8 ttl, + const char *smac, + struct in6_addr sip, struct in6_addr dip) +{ + mlxsw_reg_mpat_eth_rspan_ttl_set(payload, ttl); + mlxsw_reg_mpat_eth_rspan_smac_memcpy_to(payload, smac); + mlxsw_reg_mpat_eth_rspan_protocol_set(payload, + MLXSW_REG_MPAT_ETH_RSPAN_PROTOCOL_IPV6); + mlxsw_reg_mpat_eth_rspan_sip6_memcpy_to(payload, (void *)&sip); + mlxsw_reg_mpat_eth_rspan_dip6_memcpy_to(payload, (void *)&dip); +} + +/* MPAR - Monitoring Port Analyzer Register + * ---------------------------------------- + * MPAR register is used to query and configure the port analyzer port mirroring + * properties. + */ +#define MLXSW_REG_MPAR_ID 0x901B +#define MLXSW_REG_MPAR_LEN 0x0C + +MLXSW_REG_DEFINE(mpar, MLXSW_REG_MPAR_ID, MLXSW_REG_MPAR_LEN); + +/* reg_mpar_local_port + * The local port to mirror the packets from. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, mpar, 0x00, 16, 0x00, 4); + +enum mlxsw_reg_mpar_i_e { + MLXSW_REG_MPAR_TYPE_EGRESS, + MLXSW_REG_MPAR_TYPE_INGRESS, +}; + +/* reg_mpar_i_e + * Ingress/Egress + * Access: Index + */ +MLXSW_ITEM32(reg, mpar, i_e, 0x00, 0, 4); + +/* reg_mpar_enable + * Enable mirroring + * By default, port mirroring is disabled for all ports. + * Access: RW + */ +MLXSW_ITEM32(reg, mpar, enable, 0x04, 31, 1); + +/* reg_mpar_pa_id + * Port Analyzer ID. + * Access: RW + */ +MLXSW_ITEM32(reg, mpar, pa_id, 0x04, 0, 4); + +#define MLXSW_REG_MPAR_RATE_MAX 3500000000UL + +/* reg_mpar_probability_rate + * Sampling rate. + * Valid values are: 1 to 3.5*10^9 + * Value of 1 means "sample all". Default is 1. + * Reserved when Spectrum-1. + * Access: RW + */ +MLXSW_ITEM32(reg, mpar, probability_rate, 0x08, 0, 32); + +static inline void mlxsw_reg_mpar_pack(char *payload, u16 local_port, + enum mlxsw_reg_mpar_i_e i_e, + bool enable, u8 pa_id, + u32 probability_rate) +{ + MLXSW_REG_ZERO(mpar, payload); + mlxsw_reg_mpar_local_port_set(payload, local_port); + mlxsw_reg_mpar_enable_set(payload, enable); + mlxsw_reg_mpar_i_e_set(payload, i_e); + mlxsw_reg_mpar_pa_id_set(payload, pa_id); + mlxsw_reg_mpar_probability_rate_set(payload, probability_rate); +} + +/* MGIR - Management General Information Register + * ---------------------------------------------- + * MGIR register allows software to query the hardware and firmware general + * information. + */ +#define MLXSW_REG_MGIR_ID 0x9020 +#define MLXSW_REG_MGIR_LEN 0x9C + +MLXSW_REG_DEFINE(mgir, MLXSW_REG_MGIR_ID, MLXSW_REG_MGIR_LEN); + +/* reg_mgir_hw_info_device_hw_revision + * Access: RO + */ +MLXSW_ITEM32(reg, mgir, hw_info_device_hw_revision, 0x0, 16, 16); + +#define MLXSW_REG_MGIR_FW_INFO_PSID_SIZE 16 + +/* reg_mgir_fw_info_psid + * PSID (ASCII string). + * Access: RO + */ +MLXSW_ITEM_BUF(reg, mgir, fw_info_psid, 0x30, MLXSW_REG_MGIR_FW_INFO_PSID_SIZE); + +/* reg_mgir_fw_info_extended_major + * Access: RO + */ +MLXSW_ITEM32(reg, mgir, fw_info_extended_major, 0x44, 0, 32); + +/* reg_mgir_fw_info_extended_minor + * Access: RO + */ +MLXSW_ITEM32(reg, mgir, fw_info_extended_minor, 0x48, 0, 32); + +/* reg_mgir_fw_info_extended_sub_minor + * Access: RO + */ +MLXSW_ITEM32(reg, mgir, fw_info_extended_sub_minor, 0x4C, 0, 32); + +static inline void mlxsw_reg_mgir_pack(char *payload) +{ + MLXSW_REG_ZERO(mgir, payload); +} + +static inline void +mlxsw_reg_mgir_unpack(char *payload, u32 *hw_rev, char *fw_info_psid, + u32 *fw_major, u32 *fw_minor, u32 *fw_sub_minor) +{ + *hw_rev = mlxsw_reg_mgir_hw_info_device_hw_revision_get(payload); + mlxsw_reg_mgir_fw_info_psid_memcpy_from(payload, fw_info_psid); + *fw_major = mlxsw_reg_mgir_fw_info_extended_major_get(payload); + *fw_minor = mlxsw_reg_mgir_fw_info_extended_minor_get(payload); + *fw_sub_minor = mlxsw_reg_mgir_fw_info_extended_sub_minor_get(payload); +} + +/* MRSR - Management Reset and Shutdown Register + * --------------------------------------------- + * MRSR register is used to reset or shutdown the switch or + * the entire system (when applicable). + */ +#define MLXSW_REG_MRSR_ID 0x9023 +#define MLXSW_REG_MRSR_LEN 0x08 + +MLXSW_REG_DEFINE(mrsr, MLXSW_REG_MRSR_ID, MLXSW_REG_MRSR_LEN); + +/* reg_mrsr_command + * Reset/shutdown command + * 0 - do nothing + * 1 - software reset + * Access: WO + */ +MLXSW_ITEM32(reg, mrsr, command, 0x00, 0, 4); + +static inline void mlxsw_reg_mrsr_pack(char *payload) +{ + MLXSW_REG_ZERO(mrsr, payload); + mlxsw_reg_mrsr_command_set(payload, 1); +} + +/* MLCR - Management LED Control Register + * -------------------------------------- + * Controls the system LEDs. + */ +#define MLXSW_REG_MLCR_ID 0x902B +#define MLXSW_REG_MLCR_LEN 0x0C + +MLXSW_REG_DEFINE(mlcr, MLXSW_REG_MLCR_ID, MLXSW_REG_MLCR_LEN); + +/* reg_mlcr_local_port + * Local port number. + * Access: RW + */ +MLXSW_ITEM32_LP(reg, mlcr, 0x00, 16, 0x00, 24); + +#define MLXSW_REG_MLCR_DURATION_MAX 0xFFFF + +/* reg_mlcr_beacon_duration + * Duration of the beacon to be active, in seconds. + * 0x0 - Will turn off the beacon. + * 0xFFFF - Will turn on the beacon until explicitly turned off. + * Access: RW + */ +MLXSW_ITEM32(reg, mlcr, beacon_duration, 0x04, 0, 16); + +/* reg_mlcr_beacon_remain + * Remaining duration of the beacon, in seconds. + * 0xFFFF indicates an infinite amount of time. + * Access: RO + */ +MLXSW_ITEM32(reg, mlcr, beacon_remain, 0x08, 0, 16); + +static inline void mlxsw_reg_mlcr_pack(char *payload, u16 local_port, + bool active) +{ + MLXSW_REG_ZERO(mlcr, payload); + mlxsw_reg_mlcr_local_port_set(payload, local_port); + mlxsw_reg_mlcr_beacon_duration_set(payload, active ? + MLXSW_REG_MLCR_DURATION_MAX : 0); +} + +/* MCION - Management Cable IO and Notifications Register + * ------------------------------------------------------ + * The MCION register is used to query transceiver modules' IO pins and other + * notifications. + */ +#define MLXSW_REG_MCION_ID 0x9052 +#define MLXSW_REG_MCION_LEN 0x18 + +MLXSW_REG_DEFINE(mcion, MLXSW_REG_MCION_ID, MLXSW_REG_MCION_LEN); + +/* reg_mcion_module + * Module number. + * Access: Index + */ +MLXSW_ITEM32(reg, mcion, module, 0x00, 16, 8); + +/* reg_mcion_slot_index + * Slot index. + * Access: Index + */ +MLXSW_ITEM32(reg, mcion, slot_index, 0x00, 12, 4); + +enum { + MLXSW_REG_MCION_MODULE_STATUS_BITS_PRESENT_MASK = BIT(0), + MLXSW_REG_MCION_MODULE_STATUS_BITS_LOW_POWER_MASK = BIT(8), +}; + +/* reg_mcion_module_status_bits + * Module IO status as defined by SFF. + * Access: RO + */ +MLXSW_ITEM32(reg, mcion, module_status_bits, 0x04, 0, 16); + +static inline void mlxsw_reg_mcion_pack(char *payload, u8 slot_index, u8 module) +{ + MLXSW_REG_ZERO(mcion, payload); + mlxsw_reg_mcion_slot_index_set(payload, slot_index); + mlxsw_reg_mcion_module_set(payload, module); +} + +/* MTPPS - Management Pulse Per Second Register + * -------------------------------------------- + * This register provides the device PPS capabilities, configure the PPS in and + * out modules and holds the PPS in time stamp. + */ +#define MLXSW_REG_MTPPS_ID 0x9053 +#define MLXSW_REG_MTPPS_LEN 0x3C + +MLXSW_REG_DEFINE(mtpps, MLXSW_REG_MTPPS_ID, MLXSW_REG_MTPPS_LEN); + +/* reg_mtpps_enable + * Enables the PPS functionality the specific pin. + * A boolean variable. + * Access: RW + */ +MLXSW_ITEM32(reg, mtpps, enable, 0x20, 31, 1); + +enum mlxsw_reg_mtpps_pin_mode { + MLXSW_REG_MTPPS_PIN_MODE_VIRTUAL_PIN = 0x2, +}; + +/* reg_mtpps_pin_mode + * Pin mode to be used. The mode must comply with the supported modes of the + * requested pin. + * Access: RW + */ +MLXSW_ITEM32(reg, mtpps, pin_mode, 0x20, 8, 4); + +#define MLXSW_REG_MTPPS_PIN_SP_VIRTUAL_PIN 7 + +/* reg_mtpps_pin + * Pin to be configured or queried out of the supported pins. + * Access: Index + */ +MLXSW_ITEM32(reg, mtpps, pin, 0x20, 0, 8); + +/* reg_mtpps_time_stamp + * When pin_mode = pps_in, the latched device time when it was triggered from + * the external GPIO pin. + * When pin_mode = pps_out or virtual_pin or pps_out_and_virtual_pin, the target + * time to generate next output signal. + * Time is in units of device clock. + * Access: RW + */ +MLXSW_ITEM64(reg, mtpps, time_stamp, 0x28, 0, 64); + +static inline void +mlxsw_reg_mtpps_vpin_pack(char *payload, u64 time_stamp) +{ + MLXSW_REG_ZERO(mtpps, payload); + mlxsw_reg_mtpps_pin_set(payload, MLXSW_REG_MTPPS_PIN_SP_VIRTUAL_PIN); + mlxsw_reg_mtpps_pin_mode_set(payload, + MLXSW_REG_MTPPS_PIN_MODE_VIRTUAL_PIN); + mlxsw_reg_mtpps_enable_set(payload, true); + mlxsw_reg_mtpps_time_stamp_set(payload, time_stamp); +} + +/* MTUTC - Management UTC Register + * ------------------------------- + * Configures the HW UTC counter. + */ +#define MLXSW_REG_MTUTC_ID 0x9055 +#define MLXSW_REG_MTUTC_LEN 0x1C + +MLXSW_REG_DEFINE(mtutc, MLXSW_REG_MTUTC_ID, MLXSW_REG_MTUTC_LEN); + +enum mlxsw_reg_mtutc_operation { + MLXSW_REG_MTUTC_OPERATION_SET_TIME_AT_NEXT_SEC = 0, + MLXSW_REG_MTUTC_OPERATION_SET_TIME_IMMEDIATE = 1, + MLXSW_REG_MTUTC_OPERATION_ADJUST_TIME = 2, + MLXSW_REG_MTUTC_OPERATION_ADJUST_FREQ = 3, +}; + +/* reg_mtutc_operation + * Operation. + * Access: OP + */ +MLXSW_ITEM32(reg, mtutc, operation, 0x00, 0, 4); + +/* reg_mtutc_freq_adjustment + * Frequency adjustment: Every PPS the HW frequency will be + * adjusted by this value. Units of HW clock, where HW counts + * 10^9 HW clocks for 1 HW second. Range is from -50,000,000 to +50,000,000. + * In Spectrum-2, the field is reversed, positive values mean to decrease the + * frequency. + * Access: RW + */ +MLXSW_ITEM32(reg, mtutc, freq_adjustment, 0x04, 0, 32); + +#define MLXSW_REG_MTUTC_MAX_FREQ_ADJ (50 * 1000 * 1000) + +/* reg_mtutc_utc_sec + * UTC seconds. + * Access: WO + */ +MLXSW_ITEM32(reg, mtutc, utc_sec, 0x10, 0, 32); + +/* reg_mtutc_utc_nsec + * UTC nSecs. + * Range 0..(10^9-1) + * Updated when operation is SET_TIME_IMMEDIATE. + * Reserved on Spectrum-1. + * Access: WO + */ +MLXSW_ITEM32(reg, mtutc, utc_nsec, 0x14, 0, 30); + +/* reg_mtutc_time_adjustment + * Time adjustment. + * Units of nSec. + * Range is from -32768 to +32767. + * Updated when operation is ADJUST_TIME. + * Reserved on Spectrum-1. + * Access: WO + */ +MLXSW_ITEM32(reg, mtutc, time_adjustment, 0x18, 0, 32); + +static inline void +mlxsw_reg_mtutc_pack(char *payload, enum mlxsw_reg_mtutc_operation oper, + u32 freq_adj, u32 utc_sec, u32 utc_nsec, u32 time_adj) +{ + MLXSW_REG_ZERO(mtutc, payload); + mlxsw_reg_mtutc_operation_set(payload, oper); + mlxsw_reg_mtutc_freq_adjustment_set(payload, freq_adj); + mlxsw_reg_mtutc_utc_sec_set(payload, utc_sec); + mlxsw_reg_mtutc_utc_nsec_set(payload, utc_nsec); + mlxsw_reg_mtutc_time_adjustment_set(payload, time_adj); +} + +/* MCQI - Management Component Query Information + * --------------------------------------------- + * This register allows querying information about firmware components. + */ +#define MLXSW_REG_MCQI_ID 0x9061 +#define MLXSW_REG_MCQI_BASE_LEN 0x18 +#define MLXSW_REG_MCQI_CAP_LEN 0x14 +#define MLXSW_REG_MCQI_LEN (MLXSW_REG_MCQI_BASE_LEN + MLXSW_REG_MCQI_CAP_LEN) + +MLXSW_REG_DEFINE(mcqi, MLXSW_REG_MCQI_ID, MLXSW_REG_MCQI_LEN); + +/* reg_mcqi_component_index + * Index of the accessed component. + * Access: Index + */ +MLXSW_ITEM32(reg, mcqi, component_index, 0x00, 0, 16); + +enum mlxfw_reg_mcqi_info_type { + MLXSW_REG_MCQI_INFO_TYPE_CAPABILITIES, +}; + +/* reg_mcqi_info_type + * Component properties set. + * Access: RW + */ +MLXSW_ITEM32(reg, mcqi, info_type, 0x08, 0, 5); + +/* reg_mcqi_offset + * The requested/returned data offset from the section start, given in bytes. + * Must be DWORD aligned. + * Access: RW + */ +MLXSW_ITEM32(reg, mcqi, offset, 0x10, 0, 32); + +/* reg_mcqi_data_size + * The requested/returned data size, given in bytes. If data_size is not DWORD + * aligned, the last bytes are zero padded. + * Access: RW + */ +MLXSW_ITEM32(reg, mcqi, data_size, 0x14, 0, 16); + +/* reg_mcqi_cap_max_component_size + * Maximum size for this component, given in bytes. + * Access: RO + */ +MLXSW_ITEM32(reg, mcqi, cap_max_component_size, 0x20, 0, 32); + +/* reg_mcqi_cap_log_mcda_word_size + * Log 2 of the access word size in bytes. Read and write access must be aligned + * to the word size. Write access must be done for an integer number of words. + * Access: RO + */ +MLXSW_ITEM32(reg, mcqi, cap_log_mcda_word_size, 0x24, 28, 4); + +/* reg_mcqi_cap_mcda_max_write_size + * Maximal write size for MCDA register + * Access: RO + */ +MLXSW_ITEM32(reg, mcqi, cap_mcda_max_write_size, 0x24, 0, 16); + +static inline void mlxsw_reg_mcqi_pack(char *payload, u16 component_index) +{ + MLXSW_REG_ZERO(mcqi, payload); + mlxsw_reg_mcqi_component_index_set(payload, component_index); + mlxsw_reg_mcqi_info_type_set(payload, + MLXSW_REG_MCQI_INFO_TYPE_CAPABILITIES); + mlxsw_reg_mcqi_offset_set(payload, 0); + mlxsw_reg_mcqi_data_size_set(payload, MLXSW_REG_MCQI_CAP_LEN); +} + +static inline void mlxsw_reg_mcqi_unpack(char *payload, + u32 *p_cap_max_component_size, + u8 *p_cap_log_mcda_word_size, + u16 *p_cap_mcda_max_write_size) +{ + *p_cap_max_component_size = + mlxsw_reg_mcqi_cap_max_component_size_get(payload); + *p_cap_log_mcda_word_size = + mlxsw_reg_mcqi_cap_log_mcda_word_size_get(payload); + *p_cap_mcda_max_write_size = + mlxsw_reg_mcqi_cap_mcda_max_write_size_get(payload); +} + +/* MCC - Management Component Control + * ---------------------------------- + * Controls the firmware component and updates the FSM. + */ +#define MLXSW_REG_MCC_ID 0x9062 +#define MLXSW_REG_MCC_LEN 0x1C + +MLXSW_REG_DEFINE(mcc, MLXSW_REG_MCC_ID, MLXSW_REG_MCC_LEN); + +enum mlxsw_reg_mcc_instruction { + MLXSW_REG_MCC_INSTRUCTION_LOCK_UPDATE_HANDLE = 0x01, + MLXSW_REG_MCC_INSTRUCTION_RELEASE_UPDATE_HANDLE = 0x02, + MLXSW_REG_MCC_INSTRUCTION_UPDATE_COMPONENT = 0x03, + MLXSW_REG_MCC_INSTRUCTION_VERIFY_COMPONENT = 0x04, + MLXSW_REG_MCC_INSTRUCTION_ACTIVATE = 0x06, + MLXSW_REG_MCC_INSTRUCTION_CANCEL = 0x08, +}; + +/* reg_mcc_instruction + * Command to be executed by the FSM. + * Applicable for write operation only. + * Access: RW + */ +MLXSW_ITEM32(reg, mcc, instruction, 0x00, 0, 8); + +/* reg_mcc_component_index + * Index of the accessed component. Applicable only for commands that + * refer to components. Otherwise, this field is reserved. + * Access: Index + */ +MLXSW_ITEM32(reg, mcc, component_index, 0x04, 0, 16); + +/* reg_mcc_update_handle + * Token representing the current flow executed by the FSM. + * Access: WO + */ +MLXSW_ITEM32(reg, mcc, update_handle, 0x08, 0, 24); + +/* reg_mcc_error_code + * Indicates the successful completion of the instruction, or the reason it + * failed + * Access: RO + */ +MLXSW_ITEM32(reg, mcc, error_code, 0x0C, 8, 8); + +/* reg_mcc_control_state + * Current FSM state + * Access: RO + */ +MLXSW_ITEM32(reg, mcc, control_state, 0x0C, 0, 4); + +/* reg_mcc_component_size + * Component size in bytes. Valid for UPDATE_COMPONENT instruction. Specifying + * the size may shorten the update time. Value 0x0 means that size is + * unspecified. + * Access: WO + */ +MLXSW_ITEM32(reg, mcc, component_size, 0x10, 0, 32); + +static inline void mlxsw_reg_mcc_pack(char *payload, + enum mlxsw_reg_mcc_instruction instr, + u16 component_index, u32 update_handle, + u32 component_size) +{ + MLXSW_REG_ZERO(mcc, payload); + mlxsw_reg_mcc_instruction_set(payload, instr); + mlxsw_reg_mcc_component_index_set(payload, component_index); + mlxsw_reg_mcc_update_handle_set(payload, update_handle); + mlxsw_reg_mcc_component_size_set(payload, component_size); +} + +static inline void mlxsw_reg_mcc_unpack(char *payload, u32 *p_update_handle, + u8 *p_error_code, u8 *p_control_state) +{ + if (p_update_handle) + *p_update_handle = mlxsw_reg_mcc_update_handle_get(payload); + if (p_error_code) + *p_error_code = mlxsw_reg_mcc_error_code_get(payload); + if (p_control_state) + *p_control_state = mlxsw_reg_mcc_control_state_get(payload); +} + +/* MCDA - Management Component Data Access + * --------------------------------------- + * This register allows reading and writing a firmware component. + */ +#define MLXSW_REG_MCDA_ID 0x9063 +#define MLXSW_REG_MCDA_BASE_LEN 0x10 +#define MLXSW_REG_MCDA_MAX_DATA_LEN 0x80 +#define MLXSW_REG_MCDA_LEN \ + (MLXSW_REG_MCDA_BASE_LEN + MLXSW_REG_MCDA_MAX_DATA_LEN) + +MLXSW_REG_DEFINE(mcda, MLXSW_REG_MCDA_ID, MLXSW_REG_MCDA_LEN); + +/* reg_mcda_update_handle + * Token representing the current flow executed by the FSM. + * Access: RW + */ +MLXSW_ITEM32(reg, mcda, update_handle, 0x00, 0, 24); + +/* reg_mcda_offset + * Offset of accessed address relative to component start. Accesses must be in + * accordance to log_mcda_word_size in MCQI reg. + * Access: RW + */ +MLXSW_ITEM32(reg, mcda, offset, 0x04, 0, 32); + +/* reg_mcda_size + * Size of the data accessed, given in bytes. + * Access: RW + */ +MLXSW_ITEM32(reg, mcda, size, 0x08, 0, 16); + +/* reg_mcda_data + * Data block accessed. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, mcda, data, 0x10, 0, 32, 4, 0, false); + +static inline void mlxsw_reg_mcda_pack(char *payload, u32 update_handle, + u32 offset, u16 size, u8 *data) +{ + int i; + + MLXSW_REG_ZERO(mcda, payload); + mlxsw_reg_mcda_update_handle_set(payload, update_handle); + mlxsw_reg_mcda_offset_set(payload, offset); + mlxsw_reg_mcda_size_set(payload, size); + + for (i = 0; i < size / 4; i++) + mlxsw_reg_mcda_data_set(payload, i, *(u32 *) &data[i * 4]); +} + +/* MPSC - Monitoring Packet Sampling Configuration Register + * -------------------------------------------------------- + * MPSC Register is used to configure the Packet Sampling mechanism. + */ +#define MLXSW_REG_MPSC_ID 0x9080 +#define MLXSW_REG_MPSC_LEN 0x1C + +MLXSW_REG_DEFINE(mpsc, MLXSW_REG_MPSC_ID, MLXSW_REG_MPSC_LEN); + +/* reg_mpsc_local_port + * Local port number + * Not supported for CPU port + * Access: Index + */ +MLXSW_ITEM32_LP(reg, mpsc, 0x00, 16, 0x00, 12); + +/* reg_mpsc_e + * Enable sampling on port local_port + * Access: RW + */ +MLXSW_ITEM32(reg, mpsc, e, 0x04, 30, 1); + +#define MLXSW_REG_MPSC_RATE_MAX 3500000000UL + +/* reg_mpsc_rate + * Sampling rate = 1 out of rate packets (with randomization around + * the point). Valid values are: 1 to MLXSW_REG_MPSC_RATE_MAX + * Access: RW + */ +MLXSW_ITEM32(reg, mpsc, rate, 0x08, 0, 32); + +static inline void mlxsw_reg_mpsc_pack(char *payload, u16 local_port, bool e, + u32 rate) +{ + MLXSW_REG_ZERO(mpsc, payload); + mlxsw_reg_mpsc_local_port_set(payload, local_port); + mlxsw_reg_mpsc_e_set(payload, e); + mlxsw_reg_mpsc_rate_set(payload, rate); +} + +/* MGPC - Monitoring General Purpose Counter Set Register + * The MGPC register retrieves and sets the General Purpose Counter Set. + */ +#define MLXSW_REG_MGPC_ID 0x9081 +#define MLXSW_REG_MGPC_LEN 0x18 + +MLXSW_REG_DEFINE(mgpc, MLXSW_REG_MGPC_ID, MLXSW_REG_MGPC_LEN); + +/* reg_mgpc_counter_set_type + * Counter set type. + * Access: OP + */ +MLXSW_ITEM32(reg, mgpc, counter_set_type, 0x00, 24, 8); + +/* reg_mgpc_counter_index + * Counter index. + * Access: Index + */ +MLXSW_ITEM32(reg, mgpc, counter_index, 0x00, 0, 24); + +enum mlxsw_reg_mgpc_opcode { + /* Nop */ + MLXSW_REG_MGPC_OPCODE_NOP = 0x00, + /* Clear counters */ + MLXSW_REG_MGPC_OPCODE_CLEAR = 0x08, +}; + +/* reg_mgpc_opcode + * Opcode. + * Access: OP + */ +MLXSW_ITEM32(reg, mgpc, opcode, 0x04, 28, 4); + +/* reg_mgpc_byte_counter + * Byte counter value. + * Access: RW + */ +MLXSW_ITEM64(reg, mgpc, byte_counter, 0x08, 0, 64); + +/* reg_mgpc_packet_counter + * Packet counter value. + * Access: RW + */ +MLXSW_ITEM64(reg, mgpc, packet_counter, 0x10, 0, 64); + +static inline void mlxsw_reg_mgpc_pack(char *payload, u32 counter_index, + enum mlxsw_reg_mgpc_opcode opcode, + enum mlxsw_reg_flow_counter_set_type set_type) +{ + MLXSW_REG_ZERO(mgpc, payload); + mlxsw_reg_mgpc_counter_index_set(payload, counter_index); + mlxsw_reg_mgpc_counter_set_type_set(payload, set_type); + mlxsw_reg_mgpc_opcode_set(payload, opcode); +} + +/* MPRS - Monitoring Parsing State Register + * ---------------------------------------- + * The MPRS register is used for setting up the parsing for hash, + * policy-engine and routing. + */ +#define MLXSW_REG_MPRS_ID 0x9083 +#define MLXSW_REG_MPRS_LEN 0x14 + +MLXSW_REG_DEFINE(mprs, MLXSW_REG_MPRS_ID, MLXSW_REG_MPRS_LEN); + +/* reg_mprs_parsing_depth + * Minimum parsing depth. + * Need to enlarge parsing depth according to L3, MPLS, tunnels, ACL + * rules, traps, hash, etc. Default is 96 bytes. Reserved when SwitchX-2. + * Access: RW + */ +MLXSW_ITEM32(reg, mprs, parsing_depth, 0x00, 0, 16); + +/* reg_mprs_parsing_en + * Parsing enable. + * Bit 0 - Enable parsing of NVE of types VxLAN, VxLAN-GPE, GENEVE and + * NVGRE. Default is enabled. Reserved when SwitchX-2. + * Access: RW + */ +MLXSW_ITEM32(reg, mprs, parsing_en, 0x04, 0, 16); + +/* reg_mprs_vxlan_udp_dport + * VxLAN UDP destination port. + * Used for identifying VxLAN packets and for dport field in + * encapsulation. Default is 4789. + * Access: RW + */ +MLXSW_ITEM32(reg, mprs, vxlan_udp_dport, 0x10, 0, 16); + +static inline void mlxsw_reg_mprs_pack(char *payload, u16 parsing_depth, + u16 vxlan_udp_dport) +{ + MLXSW_REG_ZERO(mprs, payload); + mlxsw_reg_mprs_parsing_depth_set(payload, parsing_depth); + mlxsw_reg_mprs_parsing_en_set(payload, true); + mlxsw_reg_mprs_vxlan_udp_dport_set(payload, vxlan_udp_dport); +} + +/* MOGCR - Monitoring Global Configuration Register + * ------------------------------------------------ + */ +#define MLXSW_REG_MOGCR_ID 0x9086 +#define MLXSW_REG_MOGCR_LEN 0x20 + +MLXSW_REG_DEFINE(mogcr, MLXSW_REG_MOGCR_ID, MLXSW_REG_MOGCR_LEN); + +/* reg_mogcr_ptp_iftc + * PTP Ingress FIFO Trap Clear + * The PTP_ING_FIFO trap provides MTPPTR with clr according + * to this value. Default 0. + * Reserved when IB switches and when SwitchX/-2, Spectrum-2 + * Access: RW + */ +MLXSW_ITEM32(reg, mogcr, ptp_iftc, 0x00, 1, 1); + +/* reg_mogcr_ptp_eftc + * PTP Egress FIFO Trap Clear + * The PTP_EGR_FIFO trap provides MTPPTR with clr according + * to this value. Default 0. + * Reserved when IB switches and when SwitchX/-2, Spectrum-2 + * Access: RW + */ +MLXSW_ITEM32(reg, mogcr, ptp_eftc, 0x00, 0, 1); + +/* reg_mogcr_mirroring_pid_base + * Base policer id for mirroring policers. + * Must have an even value (e.g. 1000, not 1001). + * Reserved when SwitchX/-2, Switch-IB/2, Spectrum-1 and Quantum. + * Access: RW + */ +MLXSW_ITEM32(reg, mogcr, mirroring_pid_base, 0x0C, 0, 14); + +/* MPAGR - Monitoring Port Analyzer Global Register + * ------------------------------------------------ + * This register is used for global port analyzer configurations. + * Note: This register is not supported by current FW versions for Spectrum-1. + */ +#define MLXSW_REG_MPAGR_ID 0x9089 +#define MLXSW_REG_MPAGR_LEN 0x0C + +MLXSW_REG_DEFINE(mpagr, MLXSW_REG_MPAGR_ID, MLXSW_REG_MPAGR_LEN); + +enum mlxsw_reg_mpagr_trigger { + MLXSW_REG_MPAGR_TRIGGER_EGRESS, + MLXSW_REG_MPAGR_TRIGGER_INGRESS, + MLXSW_REG_MPAGR_TRIGGER_INGRESS_WRED, + MLXSW_REG_MPAGR_TRIGGER_INGRESS_SHARED_BUFFER, + MLXSW_REG_MPAGR_TRIGGER_INGRESS_ING_CONG, + MLXSW_REG_MPAGR_TRIGGER_INGRESS_EGR_CONG, + MLXSW_REG_MPAGR_TRIGGER_EGRESS_ECN, + MLXSW_REG_MPAGR_TRIGGER_EGRESS_HIGH_LATENCY, +}; + +/* reg_mpagr_trigger + * Mirror trigger. + * Access: Index + */ +MLXSW_ITEM32(reg, mpagr, trigger, 0x00, 0, 4); + +/* reg_mpagr_pa_id + * Port analyzer ID. + * Access: RW + */ +MLXSW_ITEM32(reg, mpagr, pa_id, 0x04, 0, 4); + +#define MLXSW_REG_MPAGR_RATE_MAX 3500000000UL + +/* reg_mpagr_probability_rate + * Sampling rate. + * Valid values are: 1 to 3.5*10^9 + * Value of 1 means "sample all". Default is 1. + * Access: RW + */ +MLXSW_ITEM32(reg, mpagr, probability_rate, 0x08, 0, 32); + +static inline void mlxsw_reg_mpagr_pack(char *payload, + enum mlxsw_reg_mpagr_trigger trigger, + u8 pa_id, u32 probability_rate) +{ + MLXSW_REG_ZERO(mpagr, payload); + mlxsw_reg_mpagr_trigger_set(payload, trigger); + mlxsw_reg_mpagr_pa_id_set(payload, pa_id); + mlxsw_reg_mpagr_probability_rate_set(payload, probability_rate); +} + +/* MOMTE - Monitoring Mirror Trigger Enable Register + * ------------------------------------------------- + * This register is used to configure the mirror enable for different mirror + * reasons. + */ +#define MLXSW_REG_MOMTE_ID 0x908D +#define MLXSW_REG_MOMTE_LEN 0x10 + +MLXSW_REG_DEFINE(momte, MLXSW_REG_MOMTE_ID, MLXSW_REG_MOMTE_LEN); + +/* reg_momte_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, momte, 0x00, 16, 0x00, 12); + +enum mlxsw_reg_momte_type { + MLXSW_REG_MOMTE_TYPE_WRED = 0x20, + MLXSW_REG_MOMTE_TYPE_SHARED_BUFFER_TCLASS = 0x31, + MLXSW_REG_MOMTE_TYPE_SHARED_BUFFER_TCLASS_DESCRIPTORS = 0x32, + MLXSW_REG_MOMTE_TYPE_SHARED_BUFFER_EGRESS_PORT = 0x33, + MLXSW_REG_MOMTE_TYPE_ING_CONG = 0x40, + MLXSW_REG_MOMTE_TYPE_EGR_CONG = 0x50, + MLXSW_REG_MOMTE_TYPE_ECN = 0x60, + MLXSW_REG_MOMTE_TYPE_HIGH_LATENCY = 0x70, +}; + +/* reg_momte_type + * Type of mirroring. + * Access: Index + */ +MLXSW_ITEM32(reg, momte, type, 0x04, 0, 8); + +/* reg_momte_tclass_en + * TClass/PG mirror enable. Each bit represents corresponding tclass. + * 0: disable (default) + * 1: enable + * Access: RW + */ +MLXSW_ITEM_BIT_ARRAY(reg, momte, tclass_en, 0x08, 0x08, 1); + +static inline void mlxsw_reg_momte_pack(char *payload, u16 local_port, + enum mlxsw_reg_momte_type type) +{ + MLXSW_REG_ZERO(momte, payload); + mlxsw_reg_momte_local_port_set(payload, local_port); + mlxsw_reg_momte_type_set(payload, type); +} + +/* MTPPPC - Time Precision Packet Port Configuration + * ------------------------------------------------- + * This register serves for configuration of which PTP messages should be + * timestamped. This is a global configuration, despite the register name. + * + * Reserved when Spectrum-2. + */ +#define MLXSW_REG_MTPPPC_ID 0x9090 +#define MLXSW_REG_MTPPPC_LEN 0x28 + +MLXSW_REG_DEFINE(mtpppc, MLXSW_REG_MTPPPC_ID, MLXSW_REG_MTPPPC_LEN); + +/* reg_mtpppc_ing_timestamp_message_type + * Bitwise vector of PTP message types to timestamp at ingress. + * MessageType field as defined by IEEE 1588 + * Each bit corresponds to a value (e.g. Bit0: Sync, Bit1: Delay_Req) + * Default all 0 + * Access: RW + */ +MLXSW_ITEM32(reg, mtpppc, ing_timestamp_message_type, 0x08, 0, 16); + +/* reg_mtpppc_egr_timestamp_message_type + * Bitwise vector of PTP message types to timestamp at egress. + * MessageType field as defined by IEEE 1588 + * Each bit corresponds to a value (e.g. Bit0: Sync, Bit1: Delay_Req) + * Default all 0 + * Access: RW + */ +MLXSW_ITEM32(reg, mtpppc, egr_timestamp_message_type, 0x0C, 0, 16); + +static inline void mlxsw_reg_mtpppc_pack(char *payload, u16 ing, u16 egr) +{ + MLXSW_REG_ZERO(mtpppc, payload); + mlxsw_reg_mtpppc_ing_timestamp_message_type_set(payload, ing); + mlxsw_reg_mtpppc_egr_timestamp_message_type_set(payload, egr); +} + +/* MTPPTR - Time Precision Packet Timestamping Reading + * --------------------------------------------------- + * The MTPPTR is used for reading the per port PTP timestamp FIFO. + * There is a trap for packets which are latched to the timestamp FIFO, thus the + * SW knows which FIFO to read. Note that packets enter the FIFO before been + * trapped. The sequence number is used to synchronize the timestamp FIFO + * entries and the trapped packets. + * Reserved when Spectrum-2. + */ + +#define MLXSW_REG_MTPPTR_ID 0x9091 +#define MLXSW_REG_MTPPTR_BASE_LEN 0x10 /* base length, without records */ +#define MLXSW_REG_MTPPTR_REC_LEN 0x10 /* record length */ +#define MLXSW_REG_MTPPTR_REC_MAX_COUNT 4 +#define MLXSW_REG_MTPPTR_LEN (MLXSW_REG_MTPPTR_BASE_LEN + \ + MLXSW_REG_MTPPTR_REC_LEN * MLXSW_REG_MTPPTR_REC_MAX_COUNT) + +MLXSW_REG_DEFINE(mtpptr, MLXSW_REG_MTPPTR_ID, MLXSW_REG_MTPPTR_LEN); + +/* reg_mtpptr_local_port + * Not supported for CPU port. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, mtpptr, 0x00, 16, 0x00, 12); + +enum mlxsw_reg_mtpptr_dir { + MLXSW_REG_MTPPTR_DIR_INGRESS, + MLXSW_REG_MTPPTR_DIR_EGRESS, +}; + +/* reg_mtpptr_dir + * Direction. + * Access: Index + */ +MLXSW_ITEM32(reg, mtpptr, dir, 0x00, 0, 1); + +/* reg_mtpptr_clr + * Clear the records. + * Access: OP + */ +MLXSW_ITEM32(reg, mtpptr, clr, 0x04, 31, 1); + +/* reg_mtpptr_num_rec + * Number of valid records in the response + * Range 0.. cap_ptp_timestamp_fifo + * Access: RO + */ +MLXSW_ITEM32(reg, mtpptr, num_rec, 0x08, 0, 4); + +/* reg_mtpptr_rec_message_type + * MessageType field as defined by IEEE 1588 Each bit corresponds to a value + * (e.g. Bit0: Sync, Bit1: Delay_Req) + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, mtpptr, rec_message_type, + MLXSW_REG_MTPPTR_BASE_LEN, 8, 4, + MLXSW_REG_MTPPTR_REC_LEN, 0, false); + +/* reg_mtpptr_rec_domain_number + * DomainNumber field as defined by IEEE 1588 + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, mtpptr, rec_domain_number, + MLXSW_REG_MTPPTR_BASE_LEN, 0, 8, + MLXSW_REG_MTPPTR_REC_LEN, 0, false); + +/* reg_mtpptr_rec_sequence_id + * SequenceId field as defined by IEEE 1588 + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, mtpptr, rec_sequence_id, + MLXSW_REG_MTPPTR_BASE_LEN, 0, 16, + MLXSW_REG_MTPPTR_REC_LEN, 0x4, false); + +/* reg_mtpptr_rec_timestamp_high + * Timestamp of when the PTP packet has passed through the port Units of PLL + * clock time. + * For Spectrum-1 the PLL clock is 156.25Mhz and PLL clock time is 6.4nSec. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, mtpptr, rec_timestamp_high, + MLXSW_REG_MTPPTR_BASE_LEN, 0, 32, + MLXSW_REG_MTPPTR_REC_LEN, 0x8, false); + +/* reg_mtpptr_rec_timestamp_low + * See rec_timestamp_high. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, mtpptr, rec_timestamp_low, + MLXSW_REG_MTPPTR_BASE_LEN, 0, 32, + MLXSW_REG_MTPPTR_REC_LEN, 0xC, false); + +static inline void mlxsw_reg_mtpptr_unpack(const char *payload, + unsigned int rec, + u8 *p_message_type, + u8 *p_domain_number, + u16 *p_sequence_id, + u64 *p_timestamp) +{ + u32 timestamp_high, timestamp_low; + + *p_message_type = mlxsw_reg_mtpptr_rec_message_type_get(payload, rec); + *p_domain_number = mlxsw_reg_mtpptr_rec_domain_number_get(payload, rec); + *p_sequence_id = mlxsw_reg_mtpptr_rec_sequence_id_get(payload, rec); + timestamp_high = mlxsw_reg_mtpptr_rec_timestamp_high_get(payload, rec); + timestamp_low = mlxsw_reg_mtpptr_rec_timestamp_low_get(payload, rec); + *p_timestamp = (u64)timestamp_high << 32 | timestamp_low; +} + +/* MTPTPT - Monitoring Precision Time Protocol Trap Register + * --------------------------------------------------------- + * This register is used for configuring under which trap to deliver PTP + * packets depending on type of the packet. + */ +#define MLXSW_REG_MTPTPT_ID 0x9092 +#define MLXSW_REG_MTPTPT_LEN 0x08 + +MLXSW_REG_DEFINE(mtptpt, MLXSW_REG_MTPTPT_ID, MLXSW_REG_MTPTPT_LEN); + +enum mlxsw_reg_mtptpt_trap_id { + MLXSW_REG_MTPTPT_TRAP_ID_PTP0, + MLXSW_REG_MTPTPT_TRAP_ID_PTP1, +}; + +/* reg_mtptpt_trap_id + * Trap id. + * Access: Index + */ +MLXSW_ITEM32(reg, mtptpt, trap_id, 0x00, 0, 4); + +/* reg_mtptpt_message_type + * Bitwise vector of PTP message types to trap. This is a necessary but + * non-sufficient condition since need to enable also per port. See MTPPPC. + * Message types are defined by IEEE 1588 Each bit corresponds to a value (e.g. + * Bit0: Sync, Bit1: Delay_Req) + */ +MLXSW_ITEM32(reg, mtptpt, message_type, 0x04, 0, 16); + +static inline void mlxsw_reg_mtptpt_pack(char *payload, + enum mlxsw_reg_mtptpt_trap_id trap_id, + u16 message_type) +{ + MLXSW_REG_ZERO(mtptpt, payload); + mlxsw_reg_mtptpt_trap_id_set(payload, trap_id); + mlxsw_reg_mtptpt_message_type_set(payload, message_type); +} + +/* MTPCPC - Monitoring Time Precision Correction Port Configuration Register + * ------------------------------------------------------------------------- + */ +#define MLXSW_REG_MTPCPC_ID 0x9093 +#define MLXSW_REG_MTPCPC_LEN 0x2C + +MLXSW_REG_DEFINE(mtpcpc, MLXSW_REG_MTPCPC_ID, MLXSW_REG_MTPCPC_LEN); + +/* reg_mtpcpc_pport + * Per port: + * 0: config is global. When reading - the local_port is 1. + * 1: config is per port. + * Access: Index + */ +MLXSW_ITEM32(reg, mtpcpc, pport, 0x00, 31, 1); + +/* reg_mtpcpc_local_port + * Local port number. + * Supported to/from CPU port. + * Reserved when pport = 0. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, mtpcpc, 0x00, 16, 0x00, 12); + +/* reg_mtpcpc_ptp_trap_en + * Enable PTP traps. + * The trap_id is configured by MTPTPT. + * Access: RW + */ +MLXSW_ITEM32(reg, mtpcpc, ptp_trap_en, 0x04, 0, 1); + +/* reg_mtpcpc_ing_correction_message_type + * Bitwise vector of PTP message types to update correction-field at ingress. + * MessageType field as defined by IEEE 1588 Each bit corresponds to a value + * (e.g. Bit0: Sync, Bit1: Delay_Req). Supported also from CPU port. + * Default all 0 + * Access: RW + */ +MLXSW_ITEM32(reg, mtpcpc, ing_correction_message_type, 0x10, 0, 16); + +/* reg_mtpcpc_egr_correction_message_type + * Bitwise vector of PTP message types to update correction-field at egress. + * MessageType field as defined by IEEE 1588 Each bit corresponds to a value + * (e.g. Bit0: Sync, Bit1: Delay_Req). Supported also from CPU port. + * Default all 0 + * Access: RW + */ +MLXSW_ITEM32(reg, mtpcpc, egr_correction_message_type, 0x14, 0, 16); + +static inline void mlxsw_reg_mtpcpc_pack(char *payload, bool pport, + u16 local_port, bool ptp_trap_en, + u16 ing, u16 egr) +{ + MLXSW_REG_ZERO(mtpcpc, payload); + mlxsw_reg_mtpcpc_pport_set(payload, pport); + mlxsw_reg_mtpcpc_local_port_set(payload, pport ? local_port : 0); + mlxsw_reg_mtpcpc_ptp_trap_en_set(payload, ptp_trap_en); + mlxsw_reg_mtpcpc_ing_correction_message_type_set(payload, ing); + mlxsw_reg_mtpcpc_egr_correction_message_type_set(payload, egr); +} + +/* MFGD - Monitoring FW General Debug Register + * ------------------------------------------- + */ +#define MLXSW_REG_MFGD_ID 0x90F0 +#define MLXSW_REG_MFGD_LEN 0x0C + +MLXSW_REG_DEFINE(mfgd, MLXSW_REG_MFGD_ID, MLXSW_REG_MFGD_LEN); + +/* reg_mfgd_fw_fatal_event_mode + * 0 - don't check FW fatal (default) + * 1 - check FW fatal - enable MFDE trap + * Access: RW + */ +MLXSW_ITEM32(reg, mfgd, fatal_event_mode, 0x00, 9, 2); + +/* reg_mfgd_trigger_test + * Access: WO + */ +MLXSW_ITEM32(reg, mfgd, trigger_test, 0x00, 11, 1); + +/* MGPIR - Management General Peripheral Information Register + * ---------------------------------------------------------- + * MGPIR register allows software to query the hardware and + * firmware general information of peripheral entities. + */ +#define MLXSW_REG_MGPIR_ID 0x9100 +#define MLXSW_REG_MGPIR_LEN 0xA0 + +MLXSW_REG_DEFINE(mgpir, MLXSW_REG_MGPIR_ID, MLXSW_REG_MGPIR_LEN); + +enum mlxsw_reg_mgpir_device_type { + MLXSW_REG_MGPIR_DEVICE_TYPE_NONE, + MLXSW_REG_MGPIR_DEVICE_TYPE_GEARBOX_DIE, +}; + +/* mgpir_slot_index + * Slot index (0: Main board). + * Access: Index + */ +MLXSW_ITEM32(reg, mgpir, slot_index, 0x00, 28, 4); + +/* mgpir_device_type + * Access: RO + */ +MLXSW_ITEM32(reg, mgpir, device_type, 0x00, 24, 4); + +/* mgpir_devices_per_flash + * Number of devices of device_type per flash (can be shared by few devices). + * Access: RO + */ +MLXSW_ITEM32(reg, mgpir, devices_per_flash, 0x00, 16, 8); + +/* mgpir_num_of_devices + * Number of devices of device_type. + * Access: RO + */ +MLXSW_ITEM32(reg, mgpir, num_of_devices, 0x00, 0, 8); + +/* max_modules_per_slot + * Maximum number of modules that can be connected per slot. + * Access: RO + */ +MLXSW_ITEM32(reg, mgpir, max_modules_per_slot, 0x04, 16, 8); + +/* mgpir_num_of_slots + * Number of slots in the system. + * Access: RO + */ +MLXSW_ITEM32(reg, mgpir, num_of_slots, 0x04, 8, 8); + +/* mgpir_num_of_modules + * Number of modules. + * Access: RO + */ +MLXSW_ITEM32(reg, mgpir, num_of_modules, 0x04, 0, 8); + +static inline void mlxsw_reg_mgpir_pack(char *payload, u8 slot_index) +{ + MLXSW_REG_ZERO(mgpir, payload); + mlxsw_reg_mgpir_slot_index_set(payload, slot_index); +} + +static inline void +mlxsw_reg_mgpir_unpack(char *payload, u8 *num_of_devices, + enum mlxsw_reg_mgpir_device_type *device_type, + u8 *devices_per_flash, u8 *num_of_modules, + u8 *num_of_slots) +{ + if (num_of_devices) + *num_of_devices = mlxsw_reg_mgpir_num_of_devices_get(payload); + if (device_type) + *device_type = mlxsw_reg_mgpir_device_type_get(payload); + if (devices_per_flash) + *devices_per_flash = + mlxsw_reg_mgpir_devices_per_flash_get(payload); + if (num_of_modules) + *num_of_modules = mlxsw_reg_mgpir_num_of_modules_get(payload); + if (num_of_slots) + *num_of_slots = mlxsw_reg_mgpir_num_of_slots_get(payload); +} + +/* MBCT - Management Binary Code Transfer Register + * ----------------------------------------------- + * This register allows to transfer binary codes from the host to + * the management FW by transferring it by chunks of maximum 1KB. + */ +#define MLXSW_REG_MBCT_ID 0x9120 +#define MLXSW_REG_MBCT_LEN 0x420 + +MLXSW_REG_DEFINE(mbct, MLXSW_REG_MBCT_ID, MLXSW_REG_MBCT_LEN); + +/* reg_mbct_slot_index + * Slot index. 0 is reserved. + * Access: Index + */ +MLXSW_ITEM32(reg, mbct, slot_index, 0x00, 0, 4); + +/* reg_mbct_data_size + * Actual data field size in bytes for the current data transfer. + * Access: WO + */ +MLXSW_ITEM32(reg, mbct, data_size, 0x04, 0, 11); + +enum mlxsw_reg_mbct_op { + MLXSW_REG_MBCT_OP_ERASE_INI_IMAGE = 1, + MLXSW_REG_MBCT_OP_DATA_TRANSFER, /* Download */ + MLXSW_REG_MBCT_OP_ACTIVATE, + MLXSW_REG_MBCT_OP_CLEAR_ERRORS = 6, + MLXSW_REG_MBCT_OP_QUERY_STATUS, +}; + +/* reg_mbct_op + * Access: WO + */ +MLXSW_ITEM32(reg, mbct, op, 0x08, 28, 4); + +/* reg_mbct_last + * Indicates that the current data field is the last chunk of the INI. + * Access: WO + */ +MLXSW_ITEM32(reg, mbct, last, 0x08, 26, 1); + +/* reg_mbct_oee + * Opcode Event Enable. When set a BCTOE event will be sent once the opcode + * was executed and the fsm_state has changed. + * Access: WO + */ +MLXSW_ITEM32(reg, mbct, oee, 0x08, 25, 1); + +enum mlxsw_reg_mbct_status { + /* Partial data transfer completed successfully and ready for next + * data transfer. + */ + MLXSW_REG_MBCT_STATUS_PART_DATA = 2, + MLXSW_REG_MBCT_STATUS_LAST_DATA, + MLXSW_REG_MBCT_STATUS_ERASE_COMPLETE, + /* Error - trying to erase INI while it being used. */ + MLXSW_REG_MBCT_STATUS_ERROR_INI_IN_USE, + /* Last data transfer completed, applying magic pattern. */ + MLXSW_REG_MBCT_STATUS_ERASE_FAILED = 7, + MLXSW_REG_MBCT_STATUS_INI_ERROR, + MLXSW_REG_MBCT_STATUS_ACTIVATION_FAILED, + MLXSW_REG_MBCT_STATUS_ILLEGAL_OPERATION = 11, +}; + +/* reg_mbct_status + * Status. + * Access: RO + */ +MLXSW_ITEM32(reg, mbct, status, 0x0C, 24, 5); + +enum mlxsw_reg_mbct_fsm_state { + MLXSW_REG_MBCT_FSM_STATE_INI_IN_USE = 5, + MLXSW_REG_MBCT_FSM_STATE_ERROR, +}; + +/* reg_mbct_fsm_state + * FSM state. + * Access: RO + */ +MLXSW_ITEM32(reg, mbct, fsm_state, 0x0C, 16, 4); + +#define MLXSW_REG_MBCT_DATA_LEN 1024 + +/* reg_mbct_data + * Up to 1KB of data. + * Access: WO + */ +MLXSW_ITEM_BUF(reg, mbct, data, 0x20, MLXSW_REG_MBCT_DATA_LEN); + +static inline void mlxsw_reg_mbct_pack(char *payload, u8 slot_index, + enum mlxsw_reg_mbct_op op, bool oee) +{ + MLXSW_REG_ZERO(mbct, payload); + mlxsw_reg_mbct_slot_index_set(payload, slot_index); + mlxsw_reg_mbct_op_set(payload, op); + mlxsw_reg_mbct_oee_set(payload, oee); +} + +static inline void mlxsw_reg_mbct_dt_pack(char *payload, + u16 data_size, bool last, + const char *data) +{ + if (WARN_ON(data_size > MLXSW_REG_MBCT_DATA_LEN)) + return; + mlxsw_reg_mbct_data_size_set(payload, data_size); + mlxsw_reg_mbct_last_set(payload, last); + mlxsw_reg_mbct_data_memcpy_to(payload, data); +} + +static inline void +mlxsw_reg_mbct_unpack(const char *payload, u8 *p_slot_index, + enum mlxsw_reg_mbct_status *p_status, + enum mlxsw_reg_mbct_fsm_state *p_fsm_state) +{ + if (p_slot_index) + *p_slot_index = mlxsw_reg_mbct_slot_index_get(payload); + *p_status = mlxsw_reg_mbct_status_get(payload); + if (p_fsm_state) + *p_fsm_state = mlxsw_reg_mbct_fsm_state_get(payload); +} + +/* MDDT - Management DownStream Device Tunneling Register + * ------------------------------------------------------ + * This register allows to deliver query and request messages (PRM registers, + * commands) to a DownStream device. + */ +#define MLXSW_REG_MDDT_ID 0x9160 +#define MLXSW_REG_MDDT_LEN 0x110 + +MLXSW_REG_DEFINE(mddt, MLXSW_REG_MDDT_ID, MLXSW_REG_MDDT_LEN); + +/* reg_mddt_slot_index + * Slot index. + * Access: Index + */ +MLXSW_ITEM32(reg, mddt, slot_index, 0x00, 8, 4); + +/* reg_mddt_device_index + * Device index. + * Access: Index + */ +MLXSW_ITEM32(reg, mddt, device_index, 0x00, 0, 8); + +/* reg_mddt_read_size + * Read size in D-Words. + * Access: OP + */ +MLXSW_ITEM32(reg, mddt, read_size, 0x04, 24, 8); + +/* reg_mddt_write_size + * Write size in D-Words. + * Access: OP + */ +MLXSW_ITEM32(reg, mddt, write_size, 0x04, 16, 8); + +enum mlxsw_reg_mddt_status { + MLXSW_REG_MDDT_STATUS_OK, +}; + +/* reg_mddt_status + * Return code of the Downstream Device to the register that was sent. + * Access: RO + */ +MLXSW_ITEM32(reg, mddt, status, 0x0C, 24, 8); + +enum mlxsw_reg_mddt_method { + MLXSW_REG_MDDT_METHOD_QUERY, + MLXSW_REG_MDDT_METHOD_WRITE, +}; + +/* reg_mddt_method + * Access: OP + */ +MLXSW_ITEM32(reg, mddt, method, 0x0C, 22, 2); + +/* reg_mddt_register_id + * Access: Index + */ +MLXSW_ITEM32(reg, mddt, register_id, 0x0C, 0, 16); + +#define MLXSW_REG_MDDT_PAYLOAD_OFFSET 0x0C +#define MLXSW_REG_MDDT_PRM_REGISTER_HEADER_LEN 4 + +static inline char *mlxsw_reg_mddt_inner_payload(char *payload) +{ + return payload + MLXSW_REG_MDDT_PAYLOAD_OFFSET + + MLXSW_REG_MDDT_PRM_REGISTER_HEADER_LEN; +} + +static inline void mlxsw_reg_mddt_pack(char *payload, u8 slot_index, + u8 device_index, + enum mlxsw_reg_mddt_method method, + const struct mlxsw_reg_info *reg, + char **inner_payload) +{ + int len = reg->len + MLXSW_REG_MDDT_PRM_REGISTER_HEADER_LEN; + + if (WARN_ON(len + MLXSW_REG_MDDT_PAYLOAD_OFFSET > MLXSW_REG_MDDT_LEN)) + len = MLXSW_REG_MDDT_LEN - MLXSW_REG_MDDT_PAYLOAD_OFFSET; + + MLXSW_REG_ZERO(mddt, payload); + mlxsw_reg_mddt_slot_index_set(payload, slot_index); + mlxsw_reg_mddt_device_index_set(payload, device_index); + mlxsw_reg_mddt_method_set(payload, method); + mlxsw_reg_mddt_register_id_set(payload, reg->id); + mlxsw_reg_mddt_read_size_set(payload, len / 4); + mlxsw_reg_mddt_write_size_set(payload, len / 4); + *inner_payload = mlxsw_reg_mddt_inner_payload(payload); +} + +/* MDDQ - Management DownStream Device Query Register + * -------------------------------------------------- + * This register allows to query the DownStream device properties. The desired + * information is chosen upon the query_type field and is delivered by 32B + * of data blocks. + */ +#define MLXSW_REG_MDDQ_ID 0x9161 +#define MLXSW_REG_MDDQ_LEN 0x30 + +MLXSW_REG_DEFINE(mddq, MLXSW_REG_MDDQ_ID, MLXSW_REG_MDDQ_LEN); + +/* reg_mddq_sie + * Slot info event enable. + * When set to '1', each change in the slot_info.provisioned / sr_valid / + * active / ready will generate a DSDSC event. + * Access: RW + */ +MLXSW_ITEM32(reg, mddq, sie, 0x00, 31, 1); + +enum mlxsw_reg_mddq_query_type { + MLXSW_REG_MDDQ_QUERY_TYPE_SLOT_INFO = 1, + MLXSW_REG_MDDQ_QUERY_TYPE_DEVICE_INFO, /* If there are no devices + * on the slot, data_valid + * will be '0'. + */ + MLXSW_REG_MDDQ_QUERY_TYPE_SLOT_NAME, +}; + +/* reg_mddq_query_type + * Access: Index + */ +MLXSW_ITEM32(reg, mddq, query_type, 0x00, 16, 8); + +/* reg_mddq_slot_index + * Slot index. 0 is reserved. + * Access: Index + */ +MLXSW_ITEM32(reg, mddq, slot_index, 0x00, 0, 4); + +/* reg_mddq_response_msg_seq + * Response message sequential number. For a specific request, the response + * message sequential number is the following one. In addition, the last + * message should be 0. + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, response_msg_seq, 0x04, 16, 8); + +/* reg_mddq_request_msg_seq + * Request message sequential number. + * The first message number should be 0. + * Access: Index + */ +MLXSW_ITEM32(reg, mddq, request_msg_seq, 0x04, 0, 8); + +/* reg_mddq_data_valid + * If set, the data in the data field is valid and contain the information + * for the queried index. + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, data_valid, 0x08, 31, 1); + +/* reg_mddq_slot_info_provisioned + * If set, the INI file is applied and the card is provisioned. + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, slot_info_provisioned, 0x10, 31, 1); + +/* reg_mddq_slot_info_sr_valid + * If set, Shift Register is valid (after being provisioned) and data + * can be sent from the switch ASIC to the line-card CPLD over Shift-Register. + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, slot_info_sr_valid, 0x10, 30, 1); + +enum mlxsw_reg_mddq_slot_info_ready { + MLXSW_REG_MDDQ_SLOT_INFO_READY_NOT_READY, + MLXSW_REG_MDDQ_SLOT_INFO_READY_READY, + MLXSW_REG_MDDQ_SLOT_INFO_READY_ERROR, +}; + +/* reg_mddq_slot_info_lc_ready + * If set, the LC is powered on, matching the INI version and a new FW + * version can be burnt (if necessary). + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, slot_info_lc_ready, 0x10, 28, 2); + +/* reg_mddq_slot_info_active + * If set, the FW has completed the MDDC.device_enable command. + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, slot_info_active, 0x10, 27, 1); + +/* reg_mddq_slot_info_hw_revision + * Major user-configured version number of the current INI file. + * Valid only when active or ready are '1'. + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, slot_info_hw_revision, 0x14, 16, 16); + +/* reg_mddq_slot_info_ini_file_version + * User-configured version number of the current INI file. + * Valid only when active or lc_ready are '1'. + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, slot_info_ini_file_version, 0x14, 0, 16); + +/* reg_mddq_slot_info_card_type + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, slot_info_card_type, 0x18, 0, 8); + +static inline void +__mlxsw_reg_mddq_pack(char *payload, u8 slot_index, + enum mlxsw_reg_mddq_query_type query_type) +{ + MLXSW_REG_ZERO(mddq, payload); + mlxsw_reg_mddq_slot_index_set(payload, slot_index); + mlxsw_reg_mddq_query_type_set(payload, query_type); +} + +static inline void +mlxsw_reg_mddq_slot_info_pack(char *payload, u8 slot_index, bool sie) +{ + __mlxsw_reg_mddq_pack(payload, slot_index, + MLXSW_REG_MDDQ_QUERY_TYPE_SLOT_INFO); + mlxsw_reg_mddq_sie_set(payload, sie); +} + +static inline void +mlxsw_reg_mddq_slot_info_unpack(const char *payload, u8 *p_slot_index, + bool *p_provisioned, bool *p_sr_valid, + enum mlxsw_reg_mddq_slot_info_ready *p_lc_ready, + bool *p_active, u16 *p_hw_revision, + u16 *p_ini_file_version, + u8 *p_card_type) +{ + *p_slot_index = mlxsw_reg_mddq_slot_index_get(payload); + *p_provisioned = mlxsw_reg_mddq_slot_info_provisioned_get(payload); + *p_sr_valid = mlxsw_reg_mddq_slot_info_sr_valid_get(payload); + *p_lc_ready = mlxsw_reg_mddq_slot_info_lc_ready_get(payload); + *p_active = mlxsw_reg_mddq_slot_info_active_get(payload); + *p_hw_revision = mlxsw_reg_mddq_slot_info_hw_revision_get(payload); + *p_ini_file_version = mlxsw_reg_mddq_slot_info_ini_file_version_get(payload); + *p_card_type = mlxsw_reg_mddq_slot_info_card_type_get(payload); +} + +/* reg_mddq_device_info_flash_owner + * If set, the device is the flash owner. Otherwise, a shared flash + * is used by this device (another device is the flash owner). + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, device_info_flash_owner, 0x10, 30, 1); + +/* reg_mddq_device_info_device_index + * Device index. The first device should number 0. + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, device_info_device_index, 0x10, 0, 8); + +/* reg_mddq_device_info_fw_major + * Major FW version number. + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, device_info_fw_major, 0x14, 16, 16); + +/* reg_mddq_device_info_fw_minor + * Minor FW version number. + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, device_info_fw_minor, 0x18, 16, 16); + +/* reg_mddq_device_info_fw_sub_minor + * Sub-minor FW version number. + * Access: RO + */ +MLXSW_ITEM32(reg, mddq, device_info_fw_sub_minor, 0x18, 0, 16); + +static inline void +mlxsw_reg_mddq_device_info_pack(char *payload, u8 slot_index, + u8 request_msg_seq) +{ + __mlxsw_reg_mddq_pack(payload, slot_index, + MLXSW_REG_MDDQ_QUERY_TYPE_DEVICE_INFO); + mlxsw_reg_mddq_request_msg_seq_set(payload, request_msg_seq); +} + +static inline void +mlxsw_reg_mddq_device_info_unpack(const char *payload, u8 *p_response_msg_seq, + bool *p_data_valid, bool *p_flash_owner, + u8 *p_device_index, u16 *p_fw_major, + u16 *p_fw_minor, u16 *p_fw_sub_minor) +{ + *p_response_msg_seq = mlxsw_reg_mddq_response_msg_seq_get(payload); + *p_data_valid = mlxsw_reg_mddq_data_valid_get(payload); + *p_flash_owner = mlxsw_reg_mddq_device_info_flash_owner_get(payload); + *p_device_index = mlxsw_reg_mddq_device_info_device_index_get(payload); + *p_fw_major = mlxsw_reg_mddq_device_info_fw_major_get(payload); + *p_fw_minor = mlxsw_reg_mddq_device_info_fw_minor_get(payload); + *p_fw_sub_minor = mlxsw_reg_mddq_device_info_fw_sub_minor_get(payload); +} + +#define MLXSW_REG_MDDQ_SLOT_ASCII_NAME_LEN 20 + +/* reg_mddq_slot_ascii_name + * Slot's ASCII name. + * Access: RO + */ +MLXSW_ITEM_BUF(reg, mddq, slot_ascii_name, 0x10, + MLXSW_REG_MDDQ_SLOT_ASCII_NAME_LEN); + +static inline void +mlxsw_reg_mddq_slot_name_pack(char *payload, u8 slot_index) +{ + __mlxsw_reg_mddq_pack(payload, slot_index, + MLXSW_REG_MDDQ_QUERY_TYPE_SLOT_NAME); +} + +static inline void +mlxsw_reg_mddq_slot_name_unpack(const char *payload, char *slot_ascii_name) +{ + mlxsw_reg_mddq_slot_ascii_name_memcpy_from(payload, slot_ascii_name); +} + +/* MDDC - Management DownStream Device Control Register + * ---------------------------------------------------- + * This register allows to control downstream devices and line cards. + */ +#define MLXSW_REG_MDDC_ID 0x9163 +#define MLXSW_REG_MDDC_LEN 0x30 + +MLXSW_REG_DEFINE(mddc, MLXSW_REG_MDDC_ID, MLXSW_REG_MDDC_LEN); + +/* reg_mddc_slot_index + * Slot index. 0 is reserved. + * Access: Index + */ +MLXSW_ITEM32(reg, mddc, slot_index, 0x00, 0, 4); + +/* reg_mddc_rst + * Reset request. + * Access: OP + */ +MLXSW_ITEM32(reg, mddc, rst, 0x04, 29, 1); + +/* reg_mddc_device_enable + * When set, FW is the manager and allowed to program the downstream device. + * Access: RW + */ +MLXSW_ITEM32(reg, mddc, device_enable, 0x04, 28, 1); + +static inline void mlxsw_reg_mddc_pack(char *payload, u8 slot_index, bool rst, + bool device_enable) +{ + MLXSW_REG_ZERO(mddc, payload); + mlxsw_reg_mddc_slot_index_set(payload, slot_index); + mlxsw_reg_mddc_rst_set(payload, rst); + mlxsw_reg_mddc_device_enable_set(payload, device_enable); +} + +/* MFDE - Monitoring FW Debug Register + * ----------------------------------- + */ +#define MLXSW_REG_MFDE_ID 0x9200 +#define MLXSW_REG_MFDE_LEN 0x30 + +MLXSW_REG_DEFINE(mfde, MLXSW_REG_MFDE_ID, MLXSW_REG_MFDE_LEN); + +/* reg_mfde_irisc_id + * Which irisc triggered the event + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, irisc_id, 0x00, 24, 8); + +enum mlxsw_reg_mfde_severity { + /* Unrecoverable switch behavior */ + MLXSW_REG_MFDE_SEVERITY_FATL = 2, + /* Unexpected state with possible systemic failure */ + MLXSW_REG_MFDE_SEVERITY_NRML = 3, + /* Unexpected state without systemic failure */ + MLXSW_REG_MFDE_SEVERITY_INTR = 5, +}; + +/* reg_mfde_severity + * The severity of the event. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, severity, 0x00, 16, 8); + +enum mlxsw_reg_mfde_event_id { + /* CRspace timeout */ + MLXSW_REG_MFDE_EVENT_ID_CRSPACE_TO = 1, + /* KVD insertion machine stopped */ + MLXSW_REG_MFDE_EVENT_ID_KVD_IM_STOP, + /* Triggered by MFGD.trigger_test */ + MLXSW_REG_MFDE_EVENT_ID_TEST, + /* Triggered when firmware hits an assert */ + MLXSW_REG_MFDE_EVENT_ID_FW_ASSERT, + /* Fatal error interrupt from hardware */ + MLXSW_REG_MFDE_EVENT_ID_FATAL_CAUSE, +}; + +/* reg_mfde_event_id + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, event_id, 0x00, 0, 16); + +enum mlxsw_reg_mfde_method { + MLXSW_REG_MFDE_METHOD_QUERY, + MLXSW_REG_MFDE_METHOD_WRITE, +}; + +/* reg_mfde_method + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, method, 0x04, 29, 1); + +/* reg_mfde_long_process + * Indicates if the command is in long_process mode. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, long_process, 0x04, 28, 1); + +enum mlxsw_reg_mfde_command_type { + MLXSW_REG_MFDE_COMMAND_TYPE_MAD, + MLXSW_REG_MFDE_COMMAND_TYPE_EMAD, + MLXSW_REG_MFDE_COMMAND_TYPE_CMDIF, +}; + +/* reg_mfde_command_type + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, command_type, 0x04, 24, 2); + +/* reg_mfde_reg_attr_id + * EMAD - register id, MAD - attibute id + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, reg_attr_id, 0x04, 0, 16); + +/* reg_mfde_crspace_to_log_address + * crspace address accessed, which resulted in timeout. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, crspace_to_log_address, 0x10, 0, 32); + +/* reg_mfde_crspace_to_oe + * 0 - New event + * 1 - Old event, occurred before MFGD activation. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, crspace_to_oe, 0x14, 24, 1); + +/* reg_mfde_crspace_to_log_id + * Which irisc triggered the timeout. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, crspace_to_log_id, 0x14, 0, 4); + +/* reg_mfde_crspace_to_log_ip + * IP (instruction pointer) that triggered the timeout. + * Access: RO + */ +MLXSW_ITEM64(reg, mfde, crspace_to_log_ip, 0x18, 0, 64); + +/* reg_mfde_kvd_im_stop_oe + * 0 - New event + * 1 - Old event, occurred before MFGD activation. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, kvd_im_stop_oe, 0x10, 24, 1); + +/* reg_mfde_kvd_im_stop_pipes_mask + * Bit per kvh pipe. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, kvd_im_stop_pipes_mask, 0x10, 0, 16); + +/* reg_mfde_fw_assert_var0-4 + * Variables passed to assert. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, fw_assert_var0, 0x10, 0, 32); +MLXSW_ITEM32(reg, mfde, fw_assert_var1, 0x14, 0, 32); +MLXSW_ITEM32(reg, mfde, fw_assert_var2, 0x18, 0, 32); +MLXSW_ITEM32(reg, mfde, fw_assert_var3, 0x1C, 0, 32); +MLXSW_ITEM32(reg, mfde, fw_assert_var4, 0x20, 0, 32); + +/* reg_mfde_fw_assert_existptr + * The instruction pointer when assert was triggered. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, fw_assert_existptr, 0x24, 0, 32); + +/* reg_mfde_fw_assert_callra + * The return address after triggering assert. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, fw_assert_callra, 0x28, 0, 32); + +/* reg_mfde_fw_assert_oe + * 0 - New event + * 1 - Old event, occurred before MFGD activation. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, fw_assert_oe, 0x2C, 24, 1); + +/* reg_mfde_fw_assert_tile_v + * 0: The assert was from main + * 1: The assert was from a tile + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, fw_assert_tile_v, 0x2C, 23, 1); + +/* reg_mfde_fw_assert_tile_index + * When tile_v=1, the tile_index that caused the assert. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, fw_assert_tile_index, 0x2C, 16, 6); + +/* reg_mfde_fw_assert_ext_synd + * A generated one-to-one identifier which is specific per-assert. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, fw_assert_ext_synd, 0x2C, 0, 16); + +/* reg_mfde_fatal_cause_id + * HW interrupt cause id. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, fatal_cause_id, 0x10, 0, 18); + +/* reg_mfde_fatal_cause_tile_v + * 0: The assert was from main + * 1: The assert was from a tile + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, fatal_cause_tile_v, 0x14, 23, 1); + +/* reg_mfde_fatal_cause_tile_index + * When tile_v=1, the tile_index that caused the assert. + * Access: RO + */ +MLXSW_ITEM32(reg, mfde, fatal_cause_tile_index, 0x14, 16, 6); + +/* TNGCR - Tunneling NVE General Configuration Register + * ---------------------------------------------------- + * The TNGCR register is used for setting up the NVE Tunneling configuration. + */ +#define MLXSW_REG_TNGCR_ID 0xA001 +#define MLXSW_REG_TNGCR_LEN 0x44 + +MLXSW_REG_DEFINE(tngcr, MLXSW_REG_TNGCR_ID, MLXSW_REG_TNGCR_LEN); + +enum mlxsw_reg_tngcr_type { + MLXSW_REG_TNGCR_TYPE_VXLAN, + MLXSW_REG_TNGCR_TYPE_VXLAN_GPE, + MLXSW_REG_TNGCR_TYPE_GENEVE, + MLXSW_REG_TNGCR_TYPE_NVGRE, +}; + +/* reg_tngcr_type + * Tunnel type for encapsulation and decapsulation. The types are mutually + * exclusive. + * Note: For Spectrum the NVE parsing must be enabled in MPRS. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, type, 0x00, 0, 4); + +/* reg_tngcr_nve_valid + * The VTEP is valid. Allows adding FDB entries for tunnel encapsulation. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, nve_valid, 0x04, 31, 1); + +/* reg_tngcr_nve_ttl_uc + * The TTL for NVE tunnel encapsulation underlay unicast packets. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, nve_ttl_uc, 0x04, 0, 8); + +/* reg_tngcr_nve_ttl_mc + * The TTL for NVE tunnel encapsulation underlay multicast packets. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, nve_ttl_mc, 0x08, 0, 8); + +enum { + /* Do not copy flow label. Calculate flow label using nve_flh. */ + MLXSW_REG_TNGCR_FL_NO_COPY, + /* Copy flow label from inner packet if packet is IPv6 and + * encapsulation is by IPv6. Otherwise, calculate flow label using + * nve_flh. + */ + MLXSW_REG_TNGCR_FL_COPY, +}; + +/* reg_tngcr_nve_flc + * For NVE tunnel encapsulation: Flow label copy from inner packet. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, nve_flc, 0x0C, 25, 1); + +enum { + /* Flow label is static. In Spectrum this means '0'. Spectrum-2 + * uses {nve_fl_prefix, nve_fl_suffix}. + */ + MLXSW_REG_TNGCR_FL_NO_HASH, + /* 8 LSBs of the flow label are calculated from ECMP hash of the + * inner packet. 12 MSBs are configured by nve_fl_prefix. + */ + MLXSW_REG_TNGCR_FL_HASH, +}; + +/* reg_tngcr_nve_flh + * NVE flow label hash. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, nve_flh, 0x0C, 24, 1); + +/* reg_tngcr_nve_fl_prefix + * NVE flow label prefix. Constant 12 MSBs of the flow label. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, nve_fl_prefix, 0x0C, 8, 12); + +/* reg_tngcr_nve_fl_suffix + * NVE flow label suffix. Constant 8 LSBs of the flow label. + * Reserved when nve_flh=1 and for Spectrum. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, nve_fl_suffix, 0x0C, 0, 8); + +enum { + /* Source UDP port is fixed (default '0') */ + MLXSW_REG_TNGCR_UDP_SPORT_NO_HASH, + /* Source UDP port is calculated based on hash */ + MLXSW_REG_TNGCR_UDP_SPORT_HASH, +}; + +/* reg_tngcr_nve_udp_sport_type + * NVE UDP source port type. + * Spectrum uses LAG hash (SLCRv2). Spectrum-2 uses ECMP hash (RECRv2). + * When the source UDP port is calculated based on hash, then the 8 LSBs + * are calculated from hash the 8 MSBs are configured by + * nve_udp_sport_prefix. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, nve_udp_sport_type, 0x10, 24, 1); + +/* reg_tngcr_nve_udp_sport_prefix + * NVE UDP source port prefix. Constant 8 MSBs of the UDP source port. + * Reserved when NVE type is NVGRE. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, nve_udp_sport_prefix, 0x10, 8, 8); + +/* reg_tngcr_nve_group_size_mc + * The amount of sequential linked lists of MC entries. The first linked + * list is configured by SFD.underlay_mc_ptr. + * Valid values: 1, 2, 4, 8, 16, 32, 64 + * The linked list are configured by TNUMT. + * The hash is set by LAG hash. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, nve_group_size_mc, 0x18, 0, 8); + +/* reg_tngcr_nve_group_size_flood + * The amount of sequential linked lists of flooding entries. The first + * linked list is configured by SFMR.nve_tunnel_flood_ptr + * Valid values: 1, 2, 4, 8, 16, 32, 64 + * The linked list are configured by TNUMT. + * The hash is set by LAG hash. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, nve_group_size_flood, 0x1C, 0, 8); + +/* reg_tngcr_learn_enable + * During decapsulation, whether to learn from NVE port. + * Reserved when Spectrum-2. See TNPC. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, learn_enable, 0x20, 31, 1); + +/* reg_tngcr_underlay_virtual_router + * Underlay virtual router. + * Reserved when Spectrum-2. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, underlay_virtual_router, 0x20, 0, 16); + +/* reg_tngcr_underlay_rif + * Underlay ingress router interface. RIF type should be loopback generic. + * Reserved when Spectrum. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, underlay_rif, 0x24, 0, 16); + +/* reg_tngcr_usipv4 + * Underlay source IPv4 address of the NVE. + * Access: RW + */ +MLXSW_ITEM32(reg, tngcr, usipv4, 0x28, 0, 32); + +/* reg_tngcr_usipv6 + * Underlay source IPv6 address of the NVE. For Spectrum, must not be + * modified under traffic of NVE tunneling encapsulation. + * Access: RW + */ +MLXSW_ITEM_BUF(reg, tngcr, usipv6, 0x30, 16); + +static inline void mlxsw_reg_tngcr_pack(char *payload, + enum mlxsw_reg_tngcr_type type, + bool valid, u8 ttl) +{ + MLXSW_REG_ZERO(tngcr, payload); + mlxsw_reg_tngcr_type_set(payload, type); + mlxsw_reg_tngcr_nve_valid_set(payload, valid); + mlxsw_reg_tngcr_nve_ttl_uc_set(payload, ttl); + mlxsw_reg_tngcr_nve_ttl_mc_set(payload, ttl); + mlxsw_reg_tngcr_nve_flc_set(payload, MLXSW_REG_TNGCR_FL_NO_COPY); + mlxsw_reg_tngcr_nve_flh_set(payload, 0); + mlxsw_reg_tngcr_nve_udp_sport_type_set(payload, + MLXSW_REG_TNGCR_UDP_SPORT_HASH); + mlxsw_reg_tngcr_nve_udp_sport_prefix_set(payload, 0); + mlxsw_reg_tngcr_nve_group_size_mc_set(payload, 1); + mlxsw_reg_tngcr_nve_group_size_flood_set(payload, 1); +} + +/* TNUMT - Tunneling NVE Underlay Multicast Table Register + * ------------------------------------------------------- + * The TNUMT register is for building the underlay MC table. It is used + * for MC, flooding and BC traffic into the NVE tunnel. + */ +#define MLXSW_REG_TNUMT_ID 0xA003 +#define MLXSW_REG_TNUMT_LEN 0x20 + +MLXSW_REG_DEFINE(tnumt, MLXSW_REG_TNUMT_ID, MLXSW_REG_TNUMT_LEN); + +enum mlxsw_reg_tnumt_record_type { + MLXSW_REG_TNUMT_RECORD_TYPE_IPV4, + MLXSW_REG_TNUMT_RECORD_TYPE_IPV6, + MLXSW_REG_TNUMT_RECORD_TYPE_LABEL, +}; + +/* reg_tnumt_record_type + * Record type. + * Access: RW + */ +MLXSW_ITEM32(reg, tnumt, record_type, 0x00, 28, 4); + +/* reg_tnumt_tunnel_port + * Tunnel port. + * Access: RW + */ +MLXSW_ITEM32(reg, tnumt, tunnel_port, 0x00, 24, 4); + +/* reg_tnumt_underlay_mc_ptr + * Index to the underlay multicast table. + * For Spectrum the index is to the KVD linear. + * Access: Index + */ +MLXSW_ITEM32(reg, tnumt, underlay_mc_ptr, 0x00, 0, 24); + +/* reg_tnumt_vnext + * The next_underlay_mc_ptr is valid. + * Access: RW + */ +MLXSW_ITEM32(reg, tnumt, vnext, 0x04, 31, 1); + +/* reg_tnumt_next_underlay_mc_ptr + * The next index to the underlay multicast table. + * Access: RW + */ +MLXSW_ITEM32(reg, tnumt, next_underlay_mc_ptr, 0x04, 0, 24); + +/* reg_tnumt_record_size + * Number of IP addresses in the record. + * Range is 1..cap_max_nve_mc_entries_ipv{4,6} + * Access: RW + */ +MLXSW_ITEM32(reg, tnumt, record_size, 0x08, 0, 3); + +/* reg_tnumt_udip + * The underlay IPv4 addresses. udip[i] is reserved if i >= size + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, tnumt, udip, 0x0C, 0, 32, 0x04, 0x00, false); + +/* reg_tnumt_udip_ptr + * The pointer to the underlay IPv6 addresses. udip_ptr[i] is reserved if + * i >= size. The IPv6 addresses are configured by RIPS. + * Access: RW + */ +MLXSW_ITEM32_INDEXED(reg, tnumt, udip_ptr, 0x0C, 0, 24, 0x04, 0x00, false); + +static inline void mlxsw_reg_tnumt_pack(char *payload, + enum mlxsw_reg_tnumt_record_type type, + enum mlxsw_reg_tunnel_port tport, + u32 underlay_mc_ptr, bool vnext, + u32 next_underlay_mc_ptr, + u8 record_size) +{ + MLXSW_REG_ZERO(tnumt, payload); + mlxsw_reg_tnumt_record_type_set(payload, type); + mlxsw_reg_tnumt_tunnel_port_set(payload, tport); + mlxsw_reg_tnumt_underlay_mc_ptr_set(payload, underlay_mc_ptr); + mlxsw_reg_tnumt_vnext_set(payload, vnext); + mlxsw_reg_tnumt_next_underlay_mc_ptr_set(payload, next_underlay_mc_ptr); + mlxsw_reg_tnumt_record_size_set(payload, record_size); +} + +/* TNQCR - Tunneling NVE QoS Configuration Register + * ------------------------------------------------ + * The TNQCR register configures how QoS is set in encapsulation into the + * underlay network. + */ +#define MLXSW_REG_TNQCR_ID 0xA010 +#define MLXSW_REG_TNQCR_LEN 0x0C + +MLXSW_REG_DEFINE(tnqcr, MLXSW_REG_TNQCR_ID, MLXSW_REG_TNQCR_LEN); + +/* reg_tnqcr_enc_set_dscp + * For encapsulation: How to set DSCP field: + * 0 - Copy the DSCP from the overlay (inner) IP header to the underlay + * (outer) IP header. If there is no IP header, use TNQDR.dscp + * 1 - Set the DSCP field as TNQDR.dscp + * Access: RW + */ +MLXSW_ITEM32(reg, tnqcr, enc_set_dscp, 0x04, 28, 1); + +static inline void mlxsw_reg_tnqcr_pack(char *payload) +{ + MLXSW_REG_ZERO(tnqcr, payload); + mlxsw_reg_tnqcr_enc_set_dscp_set(payload, 0); +} + +/* TNQDR - Tunneling NVE QoS Default Register + * ------------------------------------------ + * The TNQDR register configures the default QoS settings for NVE + * encapsulation. + */ +#define MLXSW_REG_TNQDR_ID 0xA011 +#define MLXSW_REG_TNQDR_LEN 0x08 + +MLXSW_REG_DEFINE(tnqdr, MLXSW_REG_TNQDR_ID, MLXSW_REG_TNQDR_LEN); + +/* reg_tnqdr_local_port + * Local port number (receive port). CPU port is supported. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, tnqdr, 0x00, 16, 0x00, 12); + +/* reg_tnqdr_dscp + * For encapsulation, the default DSCP. + * Access: RW + */ +MLXSW_ITEM32(reg, tnqdr, dscp, 0x04, 0, 6); + +static inline void mlxsw_reg_tnqdr_pack(char *payload, u16 local_port) +{ + MLXSW_REG_ZERO(tnqdr, payload); + mlxsw_reg_tnqdr_local_port_set(payload, local_port); + mlxsw_reg_tnqdr_dscp_set(payload, 0); +} + +/* TNEEM - Tunneling NVE Encapsulation ECN Mapping Register + * -------------------------------------------------------- + * The TNEEM register maps ECN of the IP header at the ingress to the + * encapsulation to the ECN of the underlay network. + */ +#define MLXSW_REG_TNEEM_ID 0xA012 +#define MLXSW_REG_TNEEM_LEN 0x0C + +MLXSW_REG_DEFINE(tneem, MLXSW_REG_TNEEM_ID, MLXSW_REG_TNEEM_LEN); + +/* reg_tneem_overlay_ecn + * ECN of the IP header in the overlay network. + * Access: Index + */ +MLXSW_ITEM32(reg, tneem, overlay_ecn, 0x04, 24, 2); + +/* reg_tneem_underlay_ecn + * ECN of the IP header in the underlay network. + * Access: RW + */ +MLXSW_ITEM32(reg, tneem, underlay_ecn, 0x04, 16, 2); + +static inline void mlxsw_reg_tneem_pack(char *payload, u8 overlay_ecn, + u8 underlay_ecn) +{ + MLXSW_REG_ZERO(tneem, payload); + mlxsw_reg_tneem_overlay_ecn_set(payload, overlay_ecn); + mlxsw_reg_tneem_underlay_ecn_set(payload, underlay_ecn); +} + +/* TNDEM - Tunneling NVE Decapsulation ECN Mapping Register + * -------------------------------------------------------- + * The TNDEM register configures the actions that are done in the + * decapsulation. + */ +#define MLXSW_REG_TNDEM_ID 0xA013 +#define MLXSW_REG_TNDEM_LEN 0x0C + +MLXSW_REG_DEFINE(tndem, MLXSW_REG_TNDEM_ID, MLXSW_REG_TNDEM_LEN); + +/* reg_tndem_underlay_ecn + * ECN field of the IP header in the underlay network. + * Access: Index + */ +MLXSW_ITEM32(reg, tndem, underlay_ecn, 0x04, 24, 2); + +/* reg_tndem_overlay_ecn + * ECN field of the IP header in the overlay network. + * Access: Index + */ +MLXSW_ITEM32(reg, tndem, overlay_ecn, 0x04, 16, 2); + +/* reg_tndem_eip_ecn + * Egress IP ECN. ECN field of the IP header of the packet which goes out + * from the decapsulation. + * Access: RW + */ +MLXSW_ITEM32(reg, tndem, eip_ecn, 0x04, 8, 2); + +/* reg_tndem_trap_en + * Trap enable: + * 0 - No trap due to decap ECN + * 1 - Trap enable with trap_id + * Access: RW + */ +MLXSW_ITEM32(reg, tndem, trap_en, 0x08, 28, 4); + +/* reg_tndem_trap_id + * Trap ID. Either DECAP_ECN0 or DECAP_ECN1. + * Reserved when trap_en is '0'. + * Access: RW + */ +MLXSW_ITEM32(reg, tndem, trap_id, 0x08, 0, 9); + +static inline void mlxsw_reg_tndem_pack(char *payload, u8 underlay_ecn, + u8 overlay_ecn, u8 ecn, bool trap_en, + u16 trap_id) +{ + MLXSW_REG_ZERO(tndem, payload); + mlxsw_reg_tndem_underlay_ecn_set(payload, underlay_ecn); + mlxsw_reg_tndem_overlay_ecn_set(payload, overlay_ecn); + mlxsw_reg_tndem_eip_ecn_set(payload, ecn); + mlxsw_reg_tndem_trap_en_set(payload, trap_en); + mlxsw_reg_tndem_trap_id_set(payload, trap_id); +} + +/* TNPC - Tunnel Port Configuration Register + * ----------------------------------------- + * The TNPC register is used for tunnel port configuration. + * Reserved when Spectrum. + */ +#define MLXSW_REG_TNPC_ID 0xA020 +#define MLXSW_REG_TNPC_LEN 0x18 + +MLXSW_REG_DEFINE(tnpc, MLXSW_REG_TNPC_ID, MLXSW_REG_TNPC_LEN); + +/* reg_tnpc_tunnel_port + * Tunnel port. + * Access: Index + */ +MLXSW_ITEM32(reg, tnpc, tunnel_port, 0x00, 0, 4); + +/* reg_tnpc_learn_enable_v6 + * During IPv6 underlay decapsulation, whether to learn from tunnel port. + * Access: RW + */ +MLXSW_ITEM32(reg, tnpc, learn_enable_v6, 0x04, 1, 1); + +/* reg_tnpc_learn_enable_v4 + * During IPv4 underlay decapsulation, whether to learn from tunnel port. + * Access: RW + */ +MLXSW_ITEM32(reg, tnpc, learn_enable_v4, 0x04, 0, 1); + +static inline void mlxsw_reg_tnpc_pack(char *payload, + enum mlxsw_reg_tunnel_port tport, + bool learn_enable) +{ + MLXSW_REG_ZERO(tnpc, payload); + mlxsw_reg_tnpc_tunnel_port_set(payload, tport); + mlxsw_reg_tnpc_learn_enable_v4_set(payload, learn_enable); + mlxsw_reg_tnpc_learn_enable_v6_set(payload, learn_enable); +} + +/* TIGCR - Tunneling IPinIP General Configuration Register + * ------------------------------------------------------- + * The TIGCR register is used for setting up the IPinIP Tunnel configuration. + */ +#define MLXSW_REG_TIGCR_ID 0xA801 +#define MLXSW_REG_TIGCR_LEN 0x10 + +MLXSW_REG_DEFINE(tigcr, MLXSW_REG_TIGCR_ID, MLXSW_REG_TIGCR_LEN); + +/* reg_tigcr_ipip_ttlc + * For IPinIP Tunnel encapsulation: whether to copy the ttl from the packet + * header. + * Access: RW + */ +MLXSW_ITEM32(reg, tigcr, ttlc, 0x04, 8, 1); + +/* reg_tigcr_ipip_ttl_uc + * The TTL for IPinIP Tunnel encapsulation of unicast packets if + * reg_tigcr_ipip_ttlc is unset. + * Access: RW + */ +MLXSW_ITEM32(reg, tigcr, ttl_uc, 0x04, 0, 8); + +static inline void mlxsw_reg_tigcr_pack(char *payload, bool ttlc, u8 ttl_uc) +{ + MLXSW_REG_ZERO(tigcr, payload); + mlxsw_reg_tigcr_ttlc_set(payload, ttlc); + mlxsw_reg_tigcr_ttl_uc_set(payload, ttl_uc); +} + +/* TIEEM - Tunneling IPinIP Encapsulation ECN Mapping Register + * ----------------------------------------------------------- + * The TIEEM register maps ECN of the IP header at the ingress to the + * encapsulation to the ECN of the underlay network. + */ +#define MLXSW_REG_TIEEM_ID 0xA812 +#define MLXSW_REG_TIEEM_LEN 0x0C + +MLXSW_REG_DEFINE(tieem, MLXSW_REG_TIEEM_ID, MLXSW_REG_TIEEM_LEN); + +/* reg_tieem_overlay_ecn + * ECN of the IP header in the overlay network. + * Access: Index + */ +MLXSW_ITEM32(reg, tieem, overlay_ecn, 0x04, 24, 2); + +/* reg_tineem_underlay_ecn + * ECN of the IP header in the underlay network. + * Access: RW + */ +MLXSW_ITEM32(reg, tieem, underlay_ecn, 0x04, 16, 2); + +static inline void mlxsw_reg_tieem_pack(char *payload, u8 overlay_ecn, + u8 underlay_ecn) +{ + MLXSW_REG_ZERO(tieem, payload); + mlxsw_reg_tieem_overlay_ecn_set(payload, overlay_ecn); + mlxsw_reg_tieem_underlay_ecn_set(payload, underlay_ecn); +} + +/* TIDEM - Tunneling IPinIP Decapsulation ECN Mapping Register + * ----------------------------------------------------------- + * The TIDEM register configures the actions that are done in the + * decapsulation. + */ +#define MLXSW_REG_TIDEM_ID 0xA813 +#define MLXSW_REG_TIDEM_LEN 0x0C + +MLXSW_REG_DEFINE(tidem, MLXSW_REG_TIDEM_ID, MLXSW_REG_TIDEM_LEN); + +/* reg_tidem_underlay_ecn + * ECN field of the IP header in the underlay network. + * Access: Index + */ +MLXSW_ITEM32(reg, tidem, underlay_ecn, 0x04, 24, 2); + +/* reg_tidem_overlay_ecn + * ECN field of the IP header in the overlay network. + * Access: Index + */ +MLXSW_ITEM32(reg, tidem, overlay_ecn, 0x04, 16, 2); + +/* reg_tidem_eip_ecn + * Egress IP ECN. ECN field of the IP header of the packet which goes out + * from the decapsulation. + * Access: RW + */ +MLXSW_ITEM32(reg, tidem, eip_ecn, 0x04, 8, 2); + +/* reg_tidem_trap_en + * Trap enable: + * 0 - No trap due to decap ECN + * 1 - Trap enable with trap_id + * Access: RW + */ +MLXSW_ITEM32(reg, tidem, trap_en, 0x08, 28, 4); + +/* reg_tidem_trap_id + * Trap ID. Either DECAP_ECN0 or DECAP_ECN1. + * Reserved when trap_en is '0'. + * Access: RW + */ +MLXSW_ITEM32(reg, tidem, trap_id, 0x08, 0, 9); + +static inline void mlxsw_reg_tidem_pack(char *payload, u8 underlay_ecn, + u8 overlay_ecn, u8 eip_ecn, + bool trap_en, u16 trap_id) +{ + MLXSW_REG_ZERO(tidem, payload); + mlxsw_reg_tidem_underlay_ecn_set(payload, underlay_ecn); + mlxsw_reg_tidem_overlay_ecn_set(payload, overlay_ecn); + mlxsw_reg_tidem_eip_ecn_set(payload, eip_ecn); + mlxsw_reg_tidem_trap_en_set(payload, trap_en); + mlxsw_reg_tidem_trap_id_set(payload, trap_id); +} + +/* SBPR - Shared Buffer Pools Register + * ----------------------------------- + * The SBPR configures and retrieves the shared buffer pools and configuration. + */ +#define MLXSW_REG_SBPR_ID 0xB001 +#define MLXSW_REG_SBPR_LEN 0x14 + +MLXSW_REG_DEFINE(sbpr, MLXSW_REG_SBPR_ID, MLXSW_REG_SBPR_LEN); + +/* reg_sbpr_desc + * When set, configures descriptor buffer. + * Access: Index + */ +MLXSW_ITEM32(reg, sbpr, desc, 0x00, 31, 1); + +/* shared direstion enum for SBPR, SBCM, SBPM */ +enum mlxsw_reg_sbxx_dir { + MLXSW_REG_SBXX_DIR_INGRESS, + MLXSW_REG_SBXX_DIR_EGRESS, +}; + +/* reg_sbpr_dir + * Direction. + * Access: Index + */ +MLXSW_ITEM32(reg, sbpr, dir, 0x00, 24, 2); + +/* reg_sbpr_pool + * Pool index. + * Access: Index + */ +MLXSW_ITEM32(reg, sbpr, pool, 0x00, 0, 4); + +/* reg_sbpr_infi_size + * Size is infinite. + * Access: RW + */ +MLXSW_ITEM32(reg, sbpr, infi_size, 0x04, 31, 1); + +/* reg_sbpr_size + * Pool size in buffer cells. + * Reserved when infi_size = 1. + * Access: RW + */ +MLXSW_ITEM32(reg, sbpr, size, 0x04, 0, 24); + +enum mlxsw_reg_sbpr_mode { + MLXSW_REG_SBPR_MODE_STATIC, + MLXSW_REG_SBPR_MODE_DYNAMIC, +}; + +/* reg_sbpr_mode + * Pool quota calculation mode. + * Access: RW + */ +MLXSW_ITEM32(reg, sbpr, mode, 0x08, 0, 4); + +static inline void mlxsw_reg_sbpr_pack(char *payload, u8 pool, + enum mlxsw_reg_sbxx_dir dir, + enum mlxsw_reg_sbpr_mode mode, u32 size, + bool infi_size) +{ + MLXSW_REG_ZERO(sbpr, payload); + mlxsw_reg_sbpr_pool_set(payload, pool); + mlxsw_reg_sbpr_dir_set(payload, dir); + mlxsw_reg_sbpr_mode_set(payload, mode); + mlxsw_reg_sbpr_size_set(payload, size); + mlxsw_reg_sbpr_infi_size_set(payload, infi_size); +} + +/* SBCM - Shared Buffer Class Management Register + * ---------------------------------------------- + * The SBCM register configures and retrieves the shared buffer allocation + * and configuration according to Port-PG, including the binding to pool + * and definition of the associated quota. + */ +#define MLXSW_REG_SBCM_ID 0xB002 +#define MLXSW_REG_SBCM_LEN 0x28 + +MLXSW_REG_DEFINE(sbcm, MLXSW_REG_SBCM_ID, MLXSW_REG_SBCM_LEN); + +/* reg_sbcm_local_port + * Local port number. + * For Ingress: excludes CPU port and Router port + * For Egress: excludes IP Router + * Access: Index + */ +MLXSW_ITEM32_LP(reg, sbcm, 0x00, 16, 0x00, 4); + +/* reg_sbcm_pg_buff + * PG buffer - Port PG (dir=ingress) / traffic class (dir=egress) + * For PG buffer: range is 0..cap_max_pg_buffers - 1 + * For traffic class: range is 0..cap_max_tclass - 1 + * Note that when traffic class is in MC aware mode then the traffic + * classes which are MC aware cannot be configured. + * Access: Index + */ +MLXSW_ITEM32(reg, sbcm, pg_buff, 0x00, 8, 6); + +/* reg_sbcm_dir + * Direction. + * Access: Index + */ +MLXSW_ITEM32(reg, sbcm, dir, 0x00, 0, 2); + +/* reg_sbcm_min_buff + * Minimum buffer size for the limiter, in cells. + * Access: RW + */ +MLXSW_ITEM32(reg, sbcm, min_buff, 0x18, 0, 24); + +/* shared max_buff limits for dynamic threshold for SBCM, SBPM */ +#define MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN 1 +#define MLXSW_REG_SBXX_DYN_MAX_BUFF_MAX 14 + +/* reg_sbcm_infi_max + * Max buffer is infinite. + * Access: RW + */ +MLXSW_ITEM32(reg, sbcm, infi_max, 0x1C, 31, 1); + +/* reg_sbcm_max_buff + * When the pool associated to the port-pg/tclass is configured to + * static, Maximum buffer size for the limiter configured in cells. + * When the pool associated to the port-pg/tclass is configured to + * dynamic, the max_buff holds the "alpha" parameter, supporting + * the following values: + * 0: 0 + * i: (1/128)*2^(i-1), for i=1..14 + * 0xFF: Infinity + * Reserved when infi_max = 1. + * Access: RW + */ +MLXSW_ITEM32(reg, sbcm, max_buff, 0x1C, 0, 24); + +/* reg_sbcm_pool + * Association of the port-priority to a pool. + * Access: RW + */ +MLXSW_ITEM32(reg, sbcm, pool, 0x24, 0, 4); + +static inline void mlxsw_reg_sbcm_pack(char *payload, u16 local_port, u8 pg_buff, + enum mlxsw_reg_sbxx_dir dir, + u32 min_buff, u32 max_buff, + bool infi_max, u8 pool) +{ + MLXSW_REG_ZERO(sbcm, payload); + mlxsw_reg_sbcm_local_port_set(payload, local_port); + mlxsw_reg_sbcm_pg_buff_set(payload, pg_buff); + mlxsw_reg_sbcm_dir_set(payload, dir); + mlxsw_reg_sbcm_min_buff_set(payload, min_buff); + mlxsw_reg_sbcm_max_buff_set(payload, max_buff); + mlxsw_reg_sbcm_infi_max_set(payload, infi_max); + mlxsw_reg_sbcm_pool_set(payload, pool); +} + +/* SBPM - Shared Buffer Port Management Register + * --------------------------------------------- + * The SBPM register configures and retrieves the shared buffer allocation + * and configuration according to Port-Pool, including the definition + * of the associated quota. + */ +#define MLXSW_REG_SBPM_ID 0xB003 +#define MLXSW_REG_SBPM_LEN 0x28 + +MLXSW_REG_DEFINE(sbpm, MLXSW_REG_SBPM_ID, MLXSW_REG_SBPM_LEN); + +/* reg_sbpm_local_port + * Local port number. + * For Ingress: excludes CPU port and Router port + * For Egress: excludes IP Router + * Access: Index + */ +MLXSW_ITEM32_LP(reg, sbpm, 0x00, 16, 0x00, 12); + +/* reg_sbpm_pool + * The pool associated to quota counting on the local_port. + * Access: Index + */ +MLXSW_ITEM32(reg, sbpm, pool, 0x00, 8, 4); + +/* reg_sbpm_dir + * Direction. + * Access: Index + */ +MLXSW_ITEM32(reg, sbpm, dir, 0x00, 0, 2); + +/* reg_sbpm_buff_occupancy + * Current buffer occupancy in cells. + * Access: RO + */ +MLXSW_ITEM32(reg, sbpm, buff_occupancy, 0x10, 0, 24); + +/* reg_sbpm_clr + * Clear Max Buffer Occupancy + * When this bit is set, max_buff_occupancy field is cleared (and a + * new max value is tracked from the time the clear was performed). + * Access: OP + */ +MLXSW_ITEM32(reg, sbpm, clr, 0x14, 31, 1); + +/* reg_sbpm_max_buff_occupancy + * Maximum value of buffer occupancy in cells monitored. Cleared by + * writing to the clr field. + * Access: RO + */ +MLXSW_ITEM32(reg, sbpm, max_buff_occupancy, 0x14, 0, 24); + +/* reg_sbpm_min_buff + * Minimum buffer size for the limiter, in cells. + * Access: RW + */ +MLXSW_ITEM32(reg, sbpm, min_buff, 0x18, 0, 24); + +/* reg_sbpm_max_buff + * When the pool associated to the port-pg/tclass is configured to + * static, Maximum buffer size for the limiter configured in cells. + * When the pool associated to the port-pg/tclass is configured to + * dynamic, the max_buff holds the "alpha" parameter, supporting + * the following values: + * 0: 0 + * i: (1/128)*2^(i-1), for i=1..14 + * 0xFF: Infinity + * Access: RW + */ +MLXSW_ITEM32(reg, sbpm, max_buff, 0x1C, 0, 24); + +static inline void mlxsw_reg_sbpm_pack(char *payload, u16 local_port, u8 pool, + enum mlxsw_reg_sbxx_dir dir, bool clr, + u32 min_buff, u32 max_buff) +{ + MLXSW_REG_ZERO(sbpm, payload); + mlxsw_reg_sbpm_local_port_set(payload, local_port); + mlxsw_reg_sbpm_pool_set(payload, pool); + mlxsw_reg_sbpm_dir_set(payload, dir); + mlxsw_reg_sbpm_clr_set(payload, clr); + mlxsw_reg_sbpm_min_buff_set(payload, min_buff); + mlxsw_reg_sbpm_max_buff_set(payload, max_buff); +} + +static inline void mlxsw_reg_sbpm_unpack(char *payload, u32 *p_buff_occupancy, + u32 *p_max_buff_occupancy) +{ + *p_buff_occupancy = mlxsw_reg_sbpm_buff_occupancy_get(payload); + *p_max_buff_occupancy = mlxsw_reg_sbpm_max_buff_occupancy_get(payload); +} + +/* SBMM - Shared Buffer Multicast Management Register + * -------------------------------------------------- + * The SBMM register configures and retrieves the shared buffer allocation + * and configuration for MC packets according to Switch-Priority, including + * the binding to pool and definition of the associated quota. + */ +#define MLXSW_REG_SBMM_ID 0xB004 +#define MLXSW_REG_SBMM_LEN 0x28 + +MLXSW_REG_DEFINE(sbmm, MLXSW_REG_SBMM_ID, MLXSW_REG_SBMM_LEN); + +/* reg_sbmm_prio + * Switch Priority. + * Access: Index + */ +MLXSW_ITEM32(reg, sbmm, prio, 0x00, 8, 4); + +/* reg_sbmm_min_buff + * Minimum buffer size for the limiter, in cells. + * Access: RW + */ +MLXSW_ITEM32(reg, sbmm, min_buff, 0x18, 0, 24); + +/* reg_sbmm_max_buff + * When the pool associated to the port-pg/tclass is configured to + * static, Maximum buffer size for the limiter configured in cells. + * When the pool associated to the port-pg/tclass is configured to + * dynamic, the max_buff holds the "alpha" parameter, supporting + * the following values: + * 0: 0 + * i: (1/128)*2^(i-1), for i=1..14 + * 0xFF: Infinity + * Access: RW + */ +MLXSW_ITEM32(reg, sbmm, max_buff, 0x1C, 0, 24); + +/* reg_sbmm_pool + * Association of the port-priority to a pool. + * Access: RW + */ +MLXSW_ITEM32(reg, sbmm, pool, 0x24, 0, 4); + +static inline void mlxsw_reg_sbmm_pack(char *payload, u8 prio, u32 min_buff, + u32 max_buff, u8 pool) +{ + MLXSW_REG_ZERO(sbmm, payload); + mlxsw_reg_sbmm_prio_set(payload, prio); + mlxsw_reg_sbmm_min_buff_set(payload, min_buff); + mlxsw_reg_sbmm_max_buff_set(payload, max_buff); + mlxsw_reg_sbmm_pool_set(payload, pool); +} + +/* SBSR - Shared Buffer Status Register + * ------------------------------------ + * The SBSR register retrieves the shared buffer occupancy according to + * Port-Pool. Note that this register enables reading a large amount of data. + * It is the user's responsibility to limit the amount of data to ensure the + * response can match the maximum transfer unit. In case the response exceeds + * the maximum transport unit, it will be truncated with no special notice. + */ +#define MLXSW_REG_SBSR_ID 0xB005 +#define MLXSW_REG_SBSR_BASE_LEN 0x5C /* base length, without records */ +#define MLXSW_REG_SBSR_REC_LEN 0x8 /* record length */ +#define MLXSW_REG_SBSR_REC_MAX_COUNT 120 +#define MLXSW_REG_SBSR_LEN (MLXSW_REG_SBSR_BASE_LEN + \ + MLXSW_REG_SBSR_REC_LEN * \ + MLXSW_REG_SBSR_REC_MAX_COUNT) + +MLXSW_REG_DEFINE(sbsr, MLXSW_REG_SBSR_ID, MLXSW_REG_SBSR_LEN); + +/* reg_sbsr_clr + * Clear Max Buffer Occupancy. When this bit is set, the max_buff_occupancy + * field is cleared (and a new max value is tracked from the time the clear + * was performed). + * Access: OP + */ +MLXSW_ITEM32(reg, sbsr, clr, 0x00, 31, 1); + +#define MLXSW_REG_SBSR_NUM_PORTS_IN_PAGE 256 + +/* reg_sbsr_port_page + * Determines the range of the ports specified in the 'ingress_port_mask' + * and 'egress_port_mask' bit masks. + * {ingress,egress}_port_mask[x] is (256 * port_page) + x + * Access: Index + */ +MLXSW_ITEM32(reg, sbsr, port_page, 0x04, 0, 4); + +/* reg_sbsr_ingress_port_mask + * Bit vector for all ingress network ports. + * Indicates which of the ports (for which the relevant bit is set) + * are affected by the set operation. Configuration of any other port + * does not change. + * Access: Index + */ +MLXSW_ITEM_BIT_ARRAY(reg, sbsr, ingress_port_mask, 0x10, 0x20, 1); + +/* reg_sbsr_pg_buff_mask + * Bit vector for all switch priority groups. + * Indicates which of the priorities (for which the relevant bit is set) + * are affected by the set operation. Configuration of any other priority + * does not change. + * Range is 0..cap_max_pg_buffers - 1 + * Access: Index + */ +MLXSW_ITEM_BIT_ARRAY(reg, sbsr, pg_buff_mask, 0x30, 0x4, 1); + +/* reg_sbsr_egress_port_mask + * Bit vector for all egress network ports. + * Indicates which of the ports (for which the relevant bit is set) + * are affected by the set operation. Configuration of any other port + * does not change. + * Access: Index + */ +MLXSW_ITEM_BIT_ARRAY(reg, sbsr, egress_port_mask, 0x34, 0x20, 1); + +/* reg_sbsr_tclass_mask + * Bit vector for all traffic classes. + * Indicates which of the traffic classes (for which the relevant bit is + * set) are affected by the set operation. Configuration of any other + * traffic class does not change. + * Range is 0..cap_max_tclass - 1 + * Access: Index + */ +MLXSW_ITEM_BIT_ARRAY(reg, sbsr, tclass_mask, 0x54, 0x8, 1); + +static inline void mlxsw_reg_sbsr_pack(char *payload, bool clr) +{ + MLXSW_REG_ZERO(sbsr, payload); + mlxsw_reg_sbsr_clr_set(payload, clr); +} + +/* reg_sbsr_rec_buff_occupancy + * Current buffer occupancy in cells. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sbsr, rec_buff_occupancy, MLXSW_REG_SBSR_BASE_LEN, + 0, 24, MLXSW_REG_SBSR_REC_LEN, 0x00, false); + +/* reg_sbsr_rec_max_buff_occupancy + * Maximum value of buffer occupancy in cells monitored. Cleared by + * writing to the clr field. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, sbsr, rec_max_buff_occupancy, MLXSW_REG_SBSR_BASE_LEN, + 0, 24, MLXSW_REG_SBSR_REC_LEN, 0x04, false); + +static inline void mlxsw_reg_sbsr_rec_unpack(char *payload, int rec_index, + u32 *p_buff_occupancy, + u32 *p_max_buff_occupancy) +{ + *p_buff_occupancy = + mlxsw_reg_sbsr_rec_buff_occupancy_get(payload, rec_index); + *p_max_buff_occupancy = + mlxsw_reg_sbsr_rec_max_buff_occupancy_get(payload, rec_index); +} + +/* SBIB - Shared Buffer Internal Buffer Register + * --------------------------------------------- + * The SBIB register configures per port buffers for internal use. The internal + * buffers consume memory on the port buffers (note that the port buffers are + * used also by PBMC). + * + * For Spectrum this is used for egress mirroring. + */ +#define MLXSW_REG_SBIB_ID 0xB006 +#define MLXSW_REG_SBIB_LEN 0x10 + +MLXSW_REG_DEFINE(sbib, MLXSW_REG_SBIB_ID, MLXSW_REG_SBIB_LEN); + +/* reg_sbib_local_port + * Local port number + * Not supported for CPU port and router port + * Access: Index + */ +MLXSW_ITEM32_LP(reg, sbib, 0x00, 16, 0x00, 12); + +/* reg_sbib_buff_size + * Units represented in cells + * Allowed range is 0 to (cap_max_headroom_size - 1) + * Default is 0 + * Access: RW + */ +MLXSW_ITEM32(reg, sbib, buff_size, 0x08, 0, 24); + +static inline void mlxsw_reg_sbib_pack(char *payload, u16 local_port, + u32 buff_size) +{ + MLXSW_REG_ZERO(sbib, payload); + mlxsw_reg_sbib_local_port_set(payload, local_port); + mlxsw_reg_sbib_buff_size_set(payload, buff_size); +} + +static const struct mlxsw_reg_info *mlxsw_reg_infos[] = { + MLXSW_REG(sgcr), + MLXSW_REG(spad), + MLXSW_REG(sspr), + MLXSW_REG(sfdat), + MLXSW_REG(sfd), + MLXSW_REG(sfn), + MLXSW_REG(spms), + MLXSW_REG(spvid), + MLXSW_REG(spvm), + MLXSW_REG(spaft), + MLXSW_REG(sfgc), + MLXSW_REG(sfdf), + MLXSW_REG(sldr), + MLXSW_REG(slcr), + MLXSW_REG(slcor), + MLXSW_REG(spmlr), + MLXSW_REG(svfa), + MLXSW_REG(spvtr), + MLXSW_REG(svpe), + MLXSW_REG(sfmr), + MLXSW_REG(spvmlr), + MLXSW_REG(spvc), + MLXSW_REG(spevet), + MLXSW_REG(smpe), + MLXSW_REG(smid2), + MLXSW_REG(cwtp), + MLXSW_REG(cwtpm), + MLXSW_REG(pgcr), + MLXSW_REG(ppbt), + MLXSW_REG(pacl), + MLXSW_REG(pagt), + MLXSW_REG(ptar), + MLXSW_REG(ppbs), + MLXSW_REG(prcr), + MLXSW_REG(pefa), + MLXSW_REG(pemrbt), + MLXSW_REG(ptce2), + MLXSW_REG(perpt), + MLXSW_REG(peabfe), + MLXSW_REG(perar), + MLXSW_REG(ptce3), + MLXSW_REG(percr), + MLXSW_REG(pererp), + MLXSW_REG(iedr), + MLXSW_REG(qpts), + MLXSW_REG(qpcr), + MLXSW_REG(qtct), + MLXSW_REG(qeec), + MLXSW_REG(qrwe), + MLXSW_REG(qpdsm), + MLXSW_REG(qpdp), + MLXSW_REG(qpdpm), + MLXSW_REG(qtctm), + MLXSW_REG(qpsc), + MLXSW_REG(pmlp), + MLXSW_REG(pmtu), + MLXSW_REG(ptys), + MLXSW_REG(ppad), + MLXSW_REG(paos), + MLXSW_REG(pfcc), + MLXSW_REG(ppcnt), + MLXSW_REG(pptb), + MLXSW_REG(pbmc), + MLXSW_REG(pspa), + MLXSW_REG(pmaos), + MLXSW_REG(pplr), + MLXSW_REG(pmtdb), + MLXSW_REG(pmecr), + MLXSW_REG(pmpe), + MLXSW_REG(pddr), + MLXSW_REG(pmmp), + MLXSW_REG(pllp), + MLXSW_REG(pmtm), + MLXSW_REG(htgt), + MLXSW_REG(hpkt), + MLXSW_REG(rgcr), + MLXSW_REG(ritr), + MLXSW_REG(rtar), + MLXSW_REG(ratr), + MLXSW_REG(rtdp), + MLXSW_REG(rips), + MLXSW_REG(ratrad), + MLXSW_REG(rdpm), + MLXSW_REG(ricnt), + MLXSW_REG(rrcr), + MLXSW_REG(ralta), + MLXSW_REG(ralst), + MLXSW_REG(raltb), + MLXSW_REG(ralue), + MLXSW_REG(rauht), + MLXSW_REG(raleu), + MLXSW_REG(rauhtd), + MLXSW_REG(rigr2), + MLXSW_REG(recr2), + MLXSW_REG(rmft2), + MLXSW_REG(reiv), + MLXSW_REG(mfcr), + MLXSW_REG(mfsc), + MLXSW_REG(mfsm), + MLXSW_REG(mfsl), + MLXSW_REG(fore), + MLXSW_REG(mtcap), + MLXSW_REG(mtmp), + MLXSW_REG(mtwe), + MLXSW_REG(mtbr), + MLXSW_REG(mcia), + MLXSW_REG(mpat), + MLXSW_REG(mpar), + MLXSW_REG(mgir), + MLXSW_REG(mrsr), + MLXSW_REG(mlcr), + MLXSW_REG(mcion), + MLXSW_REG(mtpps), + MLXSW_REG(mtutc), + MLXSW_REG(mpsc), + MLXSW_REG(mcqi), + MLXSW_REG(mcc), + MLXSW_REG(mcda), + MLXSW_REG(mgpc), + MLXSW_REG(mprs), + MLXSW_REG(mogcr), + MLXSW_REG(mpagr), + MLXSW_REG(momte), + MLXSW_REG(mtpppc), + MLXSW_REG(mtpptr), + MLXSW_REG(mtptpt), + MLXSW_REG(mtpcpc), + MLXSW_REG(mfgd), + MLXSW_REG(mgpir), + MLXSW_REG(mbct), + MLXSW_REG(mddt), + MLXSW_REG(mddq), + MLXSW_REG(mddc), + MLXSW_REG(mfde), + MLXSW_REG(tngcr), + MLXSW_REG(tnumt), + MLXSW_REG(tnqcr), + MLXSW_REG(tnqdr), + MLXSW_REG(tneem), + MLXSW_REG(tndem), + MLXSW_REG(tnpc), + MLXSW_REG(tigcr), + MLXSW_REG(tieem), + MLXSW_REG(tidem), + MLXSW_REG(sbpr), + MLXSW_REG(sbcm), + MLXSW_REG(sbpm), + MLXSW_REG(sbmm), + MLXSW_REG(sbsr), + MLXSW_REG(sbib), +}; + +static inline const char *mlxsw_reg_id_str(u16 reg_id) +{ + const struct mlxsw_reg_info *reg_info; + int i; + + for (i = 0; i < ARRAY_SIZE(mlxsw_reg_infos); i++) { + reg_info = mlxsw_reg_infos[i]; + if (reg_info->id == reg_id) + return reg_info->name; + } + return "*UNKNOWN*"; +} + +/* PUDE - Port Up / Down Event + * --------------------------- + * Reports the operational state change of a port. + */ +#define MLXSW_REG_PUDE_LEN 0x10 + +/* reg_pude_swid + * Switch partition ID with which to associate the port. + * Access: Index + */ +MLXSW_ITEM32(reg, pude, swid, 0x00, 24, 8); + +/* reg_pude_local_port + * Local port number. + * Access: Index + */ +MLXSW_ITEM32_LP(reg, pude, 0x00, 16, 0x00, 12); + +/* reg_pude_admin_status + * Port administrative state (the desired state). + * 1 - Up. + * 2 - Down. + * 3 - Up once. This means that in case of link failure, the port won't go + * into polling mode, but will wait to be re-enabled by software. + * 4 - Disabled by system. Can only be set by hardware. + * Access: RO + */ +MLXSW_ITEM32(reg, pude, admin_status, 0x00, 8, 4); + +/* reg_pude_oper_status + * Port operatioanl state. + * 1 - Up. + * 2 - Down. + * 3 - Down by port failure. This means that the device will not let the + * port up again until explicitly specified by software. + * Access: RO + */ +MLXSW_ITEM32(reg, pude, oper_status, 0x00, 0, 4); + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/resources.h b/drivers/net/ethernet/mellanox/mlxsw/resources.h new file mode 100644 index 000000000..19ae0d1c7 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/resources.h @@ -0,0 +1,168 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2016-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_RESOURCES_H +#define _MLXSW_RESOURCES_H + +#include <linux/kernel.h> +#include <linux/types.h> + +enum mlxsw_res_id { + MLXSW_RES_ID_KVD_SIZE, + MLXSW_RES_ID_KVD_SINGLE_MIN_SIZE, + MLXSW_RES_ID_KVD_DOUBLE_MIN_SIZE, + MLXSW_RES_ID_PGT_SIZE, + MLXSW_RES_ID_MAX_KVD_LINEAR_RANGE, + MLXSW_RES_ID_MAX_KVD_ACTION_SETS, + MLXSW_RES_ID_MAX_TRAP_GROUPS, + MLXSW_RES_ID_CQE_V0, + MLXSW_RES_ID_CQE_V1, + MLXSW_RES_ID_CQE_V2, + MLXSW_RES_ID_COUNTER_POOL_SIZE, + MLXSW_RES_ID_COUNTER_BANK_SIZE, + MLXSW_RES_ID_MAX_SPAN, + MLXSW_RES_ID_COUNTER_SIZE_PACKETS_BYTES, + MLXSW_RES_ID_COUNTER_SIZE_ROUTER_BASIC, + MLXSW_RES_ID_MAX_SYSTEM_PORT, + MLXSW_RES_ID_FID, + MLXSW_RES_ID_MAX_LAG, + MLXSW_RES_ID_MAX_LAG_MEMBERS, + MLXSW_RES_ID_GUARANTEED_SHARED_BUFFER, + MLXSW_RES_ID_CELL_SIZE, + MLXSW_RES_ID_MAX_HEADROOM_SIZE, + MLXSW_RES_ID_ACL_MAX_TCAM_REGIONS, + MLXSW_RES_ID_ACL_MAX_TCAM_RULES, + MLXSW_RES_ID_ACL_MAX_REGIONS, + MLXSW_RES_ID_ACL_MAX_GROUPS, + MLXSW_RES_ID_ACL_MAX_GROUP_SIZE, + MLXSW_RES_ID_ACL_MAX_DEFAULT_ACTIONS, + MLXSW_RES_ID_ACL_FLEX_KEYS, + MLXSW_RES_ID_ACL_MAX_ACTION_PER_RULE, + MLXSW_RES_ID_ACL_ACTIONS_PER_SET, + MLXSW_RES_ID_ACL_MAX_ERPT_BANKS, + MLXSW_RES_ID_ACL_MAX_ERPT_BANK_SIZE, + MLXSW_RES_ID_ACL_MAX_LARGE_KEY_ID, + MLXSW_RES_ID_ACL_ERPT_ENTRIES_2KB, + MLXSW_RES_ID_ACL_ERPT_ENTRIES_4KB, + MLXSW_RES_ID_ACL_ERPT_ENTRIES_8KB, + MLXSW_RES_ID_ACL_ERPT_ENTRIES_12KB, + MLXSW_RES_ID_ACL_MAX_BF_LOG, + MLXSW_RES_ID_MAX_GLOBAL_POLICERS, + MLXSW_RES_ID_MAX_CPU_POLICERS, + MLXSW_RES_ID_MAX_VRS, + MLXSW_RES_ID_MAX_RIFS, + MLXSW_RES_ID_MC_ERIF_LIST_ENTRIES, + MLXSW_RES_ID_MAX_RIF_MAC_PROFILES, + MLXSW_RES_ID_MAX_LPM_TREES, + MLXSW_RES_ID_MAX_NVE_MC_ENTRIES_IPV4, + MLXSW_RES_ID_MAX_NVE_MC_ENTRIES_IPV6, + + /* Internal resources. + * Determined by the SW, not queried from the HW. + */ + MLXSW_RES_ID_KVD_SINGLE_SIZE, + MLXSW_RES_ID_KVD_DOUBLE_SIZE, + MLXSW_RES_ID_KVD_LINEAR_SIZE, + + __MLXSW_RES_ID_MAX, +}; + +static u16 mlxsw_res_ids[] = { + [MLXSW_RES_ID_KVD_SIZE] = 0x1001, + [MLXSW_RES_ID_KVD_SINGLE_MIN_SIZE] = 0x1002, + [MLXSW_RES_ID_KVD_DOUBLE_MIN_SIZE] = 0x1003, + [MLXSW_RES_ID_PGT_SIZE] = 0x1004, + [MLXSW_RES_ID_MAX_KVD_LINEAR_RANGE] = 0x1005, + [MLXSW_RES_ID_MAX_KVD_ACTION_SETS] = 0x1007, + [MLXSW_RES_ID_MAX_TRAP_GROUPS] = 0x2201, + [MLXSW_RES_ID_CQE_V0] = 0x2210, + [MLXSW_RES_ID_CQE_V1] = 0x2211, + [MLXSW_RES_ID_CQE_V2] = 0x2212, + [MLXSW_RES_ID_COUNTER_POOL_SIZE] = 0x2410, + [MLXSW_RES_ID_COUNTER_BANK_SIZE] = 0x2411, + [MLXSW_RES_ID_MAX_SPAN] = 0x2420, + [MLXSW_RES_ID_COUNTER_SIZE_PACKETS_BYTES] = 0x2443, + [MLXSW_RES_ID_COUNTER_SIZE_ROUTER_BASIC] = 0x2449, + [MLXSW_RES_ID_MAX_SYSTEM_PORT] = 0x2502, + [MLXSW_RES_ID_FID] = 0x2512, + [MLXSW_RES_ID_MAX_LAG] = 0x2520, + [MLXSW_RES_ID_MAX_LAG_MEMBERS] = 0x2521, + [MLXSW_RES_ID_GUARANTEED_SHARED_BUFFER] = 0x2805, /* Bytes */ + [MLXSW_RES_ID_CELL_SIZE] = 0x2803, /* Bytes */ + [MLXSW_RES_ID_MAX_HEADROOM_SIZE] = 0x2811, /* Bytes */ + [MLXSW_RES_ID_ACL_MAX_TCAM_REGIONS] = 0x2901, + [MLXSW_RES_ID_ACL_MAX_TCAM_RULES] = 0x2902, + [MLXSW_RES_ID_ACL_MAX_REGIONS] = 0x2903, + [MLXSW_RES_ID_ACL_MAX_GROUPS] = 0x2904, + [MLXSW_RES_ID_ACL_MAX_GROUP_SIZE] = 0x2905, + [MLXSW_RES_ID_ACL_MAX_DEFAULT_ACTIONS] = 0x2908, + [MLXSW_RES_ID_ACL_FLEX_KEYS] = 0x2910, + [MLXSW_RES_ID_ACL_MAX_ACTION_PER_RULE] = 0x2911, + [MLXSW_RES_ID_ACL_ACTIONS_PER_SET] = 0x2912, + [MLXSW_RES_ID_ACL_MAX_ERPT_BANKS] = 0x2940, + [MLXSW_RES_ID_ACL_MAX_ERPT_BANK_SIZE] = 0x2941, + [MLXSW_RES_ID_ACL_MAX_LARGE_KEY_ID] = 0x2942, + [MLXSW_RES_ID_ACL_ERPT_ENTRIES_2KB] = 0x2950, + [MLXSW_RES_ID_ACL_ERPT_ENTRIES_4KB] = 0x2951, + [MLXSW_RES_ID_ACL_ERPT_ENTRIES_8KB] = 0x2952, + [MLXSW_RES_ID_ACL_ERPT_ENTRIES_12KB] = 0x2953, + [MLXSW_RES_ID_ACL_MAX_BF_LOG] = 0x2960, + [MLXSW_RES_ID_MAX_GLOBAL_POLICERS] = 0x2A10, + [MLXSW_RES_ID_MAX_CPU_POLICERS] = 0x2A13, + [MLXSW_RES_ID_MAX_VRS] = 0x2C01, + [MLXSW_RES_ID_MAX_RIFS] = 0x2C02, + [MLXSW_RES_ID_MC_ERIF_LIST_ENTRIES] = 0x2C10, + [MLXSW_RES_ID_MAX_RIF_MAC_PROFILES] = 0x2C14, + [MLXSW_RES_ID_MAX_LPM_TREES] = 0x2C30, + [MLXSW_RES_ID_MAX_NVE_MC_ENTRIES_IPV4] = 0x2E02, + [MLXSW_RES_ID_MAX_NVE_MC_ENTRIES_IPV6] = 0x2E03, +}; + +struct mlxsw_res { + bool valid[__MLXSW_RES_ID_MAX]; + u64 values[__MLXSW_RES_ID_MAX]; +}; + +static inline bool mlxsw_res_valid(struct mlxsw_res *res, + enum mlxsw_res_id res_id) +{ + return res->valid[res_id]; +} + +#define MLXSW_RES_VALID(res, short_res_id) \ + mlxsw_res_valid(res, MLXSW_RES_ID_##short_res_id) + +static inline u64 mlxsw_res_get(struct mlxsw_res *res, + enum mlxsw_res_id res_id) +{ + if (WARN_ON(!res->valid[res_id])) + return 0; + return res->values[res_id]; +} + +#define MLXSW_RES_GET(res, short_res_id) \ + mlxsw_res_get(res, MLXSW_RES_ID_##short_res_id) + +static inline void mlxsw_res_set(struct mlxsw_res *res, + enum mlxsw_res_id res_id, u64 value) +{ + res->valid[res_id] = true; + res->values[res_id] = value; +} + +#define MLXSW_RES_SET(res, short_res_id, value) \ + mlxsw_res_set(res, MLXSW_RES_ID_##short_res_id, value) + +static inline void mlxsw_res_parse(struct mlxsw_res *res, u16 id, u64 value) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(mlxsw_res_ids); i++) { + if (mlxsw_res_ids[i] == id) { + mlxsw_res_set(res, i, value); + return; + } + } +} + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c new file mode 100644 index 000000000..67ecdb9e7 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c @@ -0,0 +1,5331 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/types.h> +#include <linux/pci.h> +#include <linux/netdevice.h> +#include <linux/etherdevice.h> +#include <linux/ethtool.h> +#include <linux/slab.h> +#include <linux/device.h> +#include <linux/skbuff.h> +#include <linux/if_vlan.h> +#include <linux/if_bridge.h> +#include <linux/workqueue.h> +#include <linux/jiffies.h> +#include <linux/bitops.h> +#include <linux/list.h> +#include <linux/notifier.h> +#include <linux/dcbnl.h> +#include <linux/inetdevice.h> +#include <linux/netlink.h> +#include <linux/jhash.h> +#include <linux/log2.h> +#include <linux/refcount.h> +#include <linux/rhashtable.h> +#include <net/switchdev.h> +#include <net/pkt_cls.h> +#include <net/netevent.h> +#include <net/addrconf.h> +#include <linux/ptp_classify.h> + +#include "spectrum.h" +#include "pci.h" +#include "core.h" +#include "core_env.h" +#include "reg.h" +#include "port.h" +#include "trap.h" +#include "txheader.h" +#include "spectrum_cnt.h" +#include "spectrum_dpipe.h" +#include "spectrum_acl_flex_actions.h" +#include "spectrum_span.h" +#include "spectrum_ptp.h" +#include "spectrum_trap.h" + +#define MLXSW_SP_FWREV_MINOR 2010 +#define MLXSW_SP_FWREV_SUBMINOR 1006 + +#define MLXSW_SP1_FWREV_MAJOR 13 +#define MLXSW_SP1_FWREV_CAN_RESET_MINOR 1702 + +static const struct mlxsw_fw_rev mlxsw_sp1_fw_rev = { + .major = MLXSW_SP1_FWREV_MAJOR, + .minor = MLXSW_SP_FWREV_MINOR, + .subminor = MLXSW_SP_FWREV_SUBMINOR, + .can_reset_minor = MLXSW_SP1_FWREV_CAN_RESET_MINOR, +}; + +#define MLXSW_SP1_FW_FILENAME \ + "mellanox/mlxsw_spectrum-" __stringify(MLXSW_SP1_FWREV_MAJOR) \ + "." __stringify(MLXSW_SP_FWREV_MINOR) \ + "." __stringify(MLXSW_SP_FWREV_SUBMINOR) ".mfa2" + +#define MLXSW_SP2_FWREV_MAJOR 29 + +static const struct mlxsw_fw_rev mlxsw_sp2_fw_rev = { + .major = MLXSW_SP2_FWREV_MAJOR, + .minor = MLXSW_SP_FWREV_MINOR, + .subminor = MLXSW_SP_FWREV_SUBMINOR, +}; + +#define MLXSW_SP2_FW_FILENAME \ + "mellanox/mlxsw_spectrum2-" __stringify(MLXSW_SP2_FWREV_MAJOR) \ + "." __stringify(MLXSW_SP_FWREV_MINOR) \ + "." __stringify(MLXSW_SP_FWREV_SUBMINOR) ".mfa2" + +#define MLXSW_SP3_FWREV_MAJOR 30 + +static const struct mlxsw_fw_rev mlxsw_sp3_fw_rev = { + .major = MLXSW_SP3_FWREV_MAJOR, + .minor = MLXSW_SP_FWREV_MINOR, + .subminor = MLXSW_SP_FWREV_SUBMINOR, +}; + +#define MLXSW_SP3_FW_FILENAME \ + "mellanox/mlxsw_spectrum3-" __stringify(MLXSW_SP3_FWREV_MAJOR) \ + "." __stringify(MLXSW_SP_FWREV_MINOR) \ + "." __stringify(MLXSW_SP_FWREV_SUBMINOR) ".mfa2" + +#define MLXSW_SP_LINECARDS_INI_BUNDLE_FILENAME \ + "mellanox/lc_ini_bundle_" \ + __stringify(MLXSW_SP_FWREV_MINOR) "_" \ + __stringify(MLXSW_SP_FWREV_SUBMINOR) ".bin" + +static const char mlxsw_sp1_driver_name[] = "mlxsw_spectrum"; +static const char mlxsw_sp2_driver_name[] = "mlxsw_spectrum2"; +static const char mlxsw_sp3_driver_name[] = "mlxsw_spectrum3"; +static const char mlxsw_sp4_driver_name[] = "mlxsw_spectrum4"; + +static const unsigned char mlxsw_sp1_mac_mask[ETH_ALEN] = { + 0xff, 0xff, 0xff, 0xff, 0xfc, 0x00 +}; +static const unsigned char mlxsw_sp2_mac_mask[ETH_ALEN] = { + 0xff, 0xff, 0xff, 0xff, 0xf0, 0x00 +}; + +/* tx_hdr_version + * Tx header version. + * Must be set to 1. + */ +MLXSW_ITEM32(tx, hdr, version, 0x00, 28, 4); + +/* tx_hdr_ctl + * Packet control type. + * 0 - Ethernet control (e.g. EMADs, LACP) + * 1 - Ethernet data + */ +MLXSW_ITEM32(tx, hdr, ctl, 0x00, 26, 2); + +/* tx_hdr_proto + * Packet protocol type. Must be set to 1 (Ethernet). + */ +MLXSW_ITEM32(tx, hdr, proto, 0x00, 21, 3); + +/* tx_hdr_rx_is_router + * Packet is sent from the router. Valid for data packets only. + */ +MLXSW_ITEM32(tx, hdr, rx_is_router, 0x00, 19, 1); + +/* tx_hdr_fid_valid + * Indicates if the 'fid' field is valid and should be used for + * forwarding lookup. Valid for data packets only. + */ +MLXSW_ITEM32(tx, hdr, fid_valid, 0x00, 16, 1); + +/* tx_hdr_swid + * Switch partition ID. Must be set to 0. + */ +MLXSW_ITEM32(tx, hdr, swid, 0x00, 12, 3); + +/* tx_hdr_control_tclass + * Indicates if the packet should use the control TClass and not one + * of the data TClasses. + */ +MLXSW_ITEM32(tx, hdr, control_tclass, 0x00, 6, 1); + +/* tx_hdr_etclass + * Egress TClass to be used on the egress device on the egress port. + */ +MLXSW_ITEM32(tx, hdr, etclass, 0x00, 0, 4); + +/* tx_hdr_port_mid + * Destination local port for unicast packets. + * Destination multicast ID for multicast packets. + * + * Control packets are directed to a specific egress port, while data + * packets are transmitted through the CPU port (0) into the switch partition, + * where forwarding rules are applied. + */ +MLXSW_ITEM32(tx, hdr, port_mid, 0x04, 16, 16); + +/* tx_hdr_fid + * Forwarding ID used for L2 forwarding lookup. Valid only if 'fid_valid' is + * set, otherwise calculated based on the packet's VID using VID to FID mapping. + * Valid for data packets only. + */ +MLXSW_ITEM32(tx, hdr, fid, 0x08, 16, 16); + +/* tx_hdr_type + * 0 - Data packets + * 6 - Control packets + */ +MLXSW_ITEM32(tx, hdr, type, 0x0C, 0, 4); + +int mlxsw_sp_flow_counter_get(struct mlxsw_sp *mlxsw_sp, + unsigned int counter_index, u64 *packets, + u64 *bytes) +{ + char mgpc_pl[MLXSW_REG_MGPC_LEN]; + int err; + + mlxsw_reg_mgpc_pack(mgpc_pl, counter_index, MLXSW_REG_MGPC_OPCODE_NOP, + MLXSW_REG_FLOW_COUNTER_SET_TYPE_PACKETS_BYTES); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(mgpc), mgpc_pl); + if (err) + return err; + if (packets) + *packets = mlxsw_reg_mgpc_packet_counter_get(mgpc_pl); + if (bytes) + *bytes = mlxsw_reg_mgpc_byte_counter_get(mgpc_pl); + return 0; +} + +static int mlxsw_sp_flow_counter_clear(struct mlxsw_sp *mlxsw_sp, + unsigned int counter_index) +{ + char mgpc_pl[MLXSW_REG_MGPC_LEN]; + + mlxsw_reg_mgpc_pack(mgpc_pl, counter_index, MLXSW_REG_MGPC_OPCODE_CLEAR, + MLXSW_REG_FLOW_COUNTER_SET_TYPE_PACKETS_BYTES); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mgpc), mgpc_pl); +} + +int mlxsw_sp_flow_counter_alloc(struct mlxsw_sp *mlxsw_sp, + unsigned int *p_counter_index) +{ + int err; + + err = mlxsw_sp_counter_alloc(mlxsw_sp, MLXSW_SP_COUNTER_SUB_POOL_FLOW, + p_counter_index); + if (err) + return err; + err = mlxsw_sp_flow_counter_clear(mlxsw_sp, *p_counter_index); + if (err) + goto err_counter_clear; + return 0; + +err_counter_clear: + mlxsw_sp_counter_free(mlxsw_sp, MLXSW_SP_COUNTER_SUB_POOL_FLOW, + *p_counter_index); + return err; +} + +void mlxsw_sp_flow_counter_free(struct mlxsw_sp *mlxsw_sp, + unsigned int counter_index) +{ + mlxsw_sp_counter_free(mlxsw_sp, MLXSW_SP_COUNTER_SUB_POOL_FLOW, + counter_index); +} + +void mlxsw_sp_txhdr_construct(struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info) +{ + char *txhdr = skb_push(skb, MLXSW_TXHDR_LEN); + + memset(txhdr, 0, MLXSW_TXHDR_LEN); + + mlxsw_tx_hdr_version_set(txhdr, MLXSW_TXHDR_VERSION_1); + mlxsw_tx_hdr_ctl_set(txhdr, MLXSW_TXHDR_ETH_CTL); + mlxsw_tx_hdr_proto_set(txhdr, MLXSW_TXHDR_PROTO_ETH); + mlxsw_tx_hdr_swid_set(txhdr, 0); + mlxsw_tx_hdr_control_tclass_set(txhdr, 1); + mlxsw_tx_hdr_port_mid_set(txhdr, tx_info->local_port); + mlxsw_tx_hdr_type_set(txhdr, MLXSW_TXHDR_TYPE_CONTROL); +} + +int +mlxsw_sp_txhdr_ptp_data_construct(struct mlxsw_core *mlxsw_core, + struct mlxsw_sp_port *mlxsw_sp_port, + struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info) +{ + char *txhdr; + u16 max_fid; + int err; + + if (skb_cow_head(skb, MLXSW_TXHDR_LEN)) { + err = -ENOMEM; + goto err_skb_cow_head; + } + + if (!MLXSW_CORE_RES_VALID(mlxsw_core, FID)) { + err = -EIO; + goto err_res_valid; + } + max_fid = MLXSW_CORE_RES_GET(mlxsw_core, FID); + + txhdr = skb_push(skb, MLXSW_TXHDR_LEN); + memset(txhdr, 0, MLXSW_TXHDR_LEN); + + mlxsw_tx_hdr_version_set(txhdr, MLXSW_TXHDR_VERSION_1); + mlxsw_tx_hdr_proto_set(txhdr, MLXSW_TXHDR_PROTO_ETH); + mlxsw_tx_hdr_rx_is_router_set(txhdr, true); + mlxsw_tx_hdr_fid_valid_set(txhdr, true); + mlxsw_tx_hdr_fid_set(txhdr, max_fid + tx_info->local_port - 1); + mlxsw_tx_hdr_type_set(txhdr, MLXSW_TXHDR_TYPE_DATA); + return 0; + +err_res_valid: +err_skb_cow_head: + this_cpu_inc(mlxsw_sp_port->pcpu_stats->tx_dropped); + dev_kfree_skb_any(skb); + return err; +} + +static bool mlxsw_sp_skb_requires_ts(struct sk_buff *skb) +{ + unsigned int type; + + if (!(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)) + return false; + + type = ptp_classify_raw(skb); + return !!ptp_parse_header(skb, type); +} + +static int mlxsw_sp_txhdr_handle(struct mlxsw_core *mlxsw_core, + struct mlxsw_sp_port *mlxsw_sp_port, + struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + + /* In Spectrum-2 and Spectrum-3, PTP events that require a time stamp + * need special handling and cannot be transmitted as regular control + * packets. + */ + if (unlikely(mlxsw_sp_skb_requires_ts(skb))) + return mlxsw_sp->ptp_ops->txhdr_construct(mlxsw_core, + mlxsw_sp_port, skb, + tx_info); + + if (skb_cow_head(skb, MLXSW_TXHDR_LEN)) { + this_cpu_inc(mlxsw_sp_port->pcpu_stats->tx_dropped); + dev_kfree_skb_any(skb); + return -ENOMEM; + } + + mlxsw_sp_txhdr_construct(skb, tx_info); + return 0; +} + +enum mlxsw_reg_spms_state mlxsw_sp_stp_spms_state(u8 state) +{ + switch (state) { + case BR_STATE_FORWARDING: + return MLXSW_REG_SPMS_STATE_FORWARDING; + case BR_STATE_LEARNING: + return MLXSW_REG_SPMS_STATE_LEARNING; + case BR_STATE_LISTENING: + case BR_STATE_DISABLED: + case BR_STATE_BLOCKING: + return MLXSW_REG_SPMS_STATE_DISCARDING; + default: + BUG(); + } +} + +int mlxsw_sp_port_vid_stp_set(struct mlxsw_sp_port *mlxsw_sp_port, u16 vid, + u8 state) +{ + enum mlxsw_reg_spms_state spms_state = mlxsw_sp_stp_spms_state(state); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char *spms_pl; + int err; + + spms_pl = kmalloc(MLXSW_REG_SPMS_LEN, GFP_KERNEL); + if (!spms_pl) + return -ENOMEM; + mlxsw_reg_spms_pack(spms_pl, mlxsw_sp_port->local_port); + mlxsw_reg_spms_vid_pack(spms_pl, vid, spms_state); + + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(spms), spms_pl); + kfree(spms_pl); + return err; +} + +static int mlxsw_sp_base_mac_get(struct mlxsw_sp *mlxsw_sp) +{ + char spad_pl[MLXSW_REG_SPAD_LEN] = {0}; + int err; + + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(spad), spad_pl); + if (err) + return err; + mlxsw_reg_spad_base_mac_memcpy_from(spad_pl, mlxsw_sp->base_mac); + return 0; +} + +int mlxsw_sp_port_admin_status_set(struct mlxsw_sp_port *mlxsw_sp_port, + bool is_up) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char paos_pl[MLXSW_REG_PAOS_LEN]; + + mlxsw_reg_paos_pack(paos_pl, mlxsw_sp_port->local_port, + is_up ? MLXSW_PORT_ADMIN_STATUS_UP : + MLXSW_PORT_ADMIN_STATUS_DOWN); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(paos), paos_pl); +} + +static int mlxsw_sp_port_dev_addr_set(struct mlxsw_sp_port *mlxsw_sp_port, + const unsigned char *addr) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char ppad_pl[MLXSW_REG_PPAD_LEN]; + + mlxsw_reg_ppad_pack(ppad_pl, true, mlxsw_sp_port->local_port); + mlxsw_reg_ppad_mac_memcpy_to(ppad_pl, addr); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ppad), ppad_pl); +} + +static int mlxsw_sp_port_dev_addr_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + + eth_hw_addr_gen(mlxsw_sp_port->dev, mlxsw_sp->base_mac, + mlxsw_sp_port->local_port); + return mlxsw_sp_port_dev_addr_set(mlxsw_sp_port, + mlxsw_sp_port->dev->dev_addr); +} + +static int mlxsw_sp_port_max_mtu_get(struct mlxsw_sp_port *mlxsw_sp_port, int *p_max_mtu) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char pmtu_pl[MLXSW_REG_PMTU_LEN]; + int err; + + mlxsw_reg_pmtu_pack(pmtu_pl, mlxsw_sp_port->local_port, 0); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(pmtu), pmtu_pl); + if (err) + return err; + + *p_max_mtu = mlxsw_reg_pmtu_max_mtu_get(pmtu_pl); + return 0; +} + +static int mlxsw_sp_port_mtu_set(struct mlxsw_sp_port *mlxsw_sp_port, u16 mtu) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char pmtu_pl[MLXSW_REG_PMTU_LEN]; + + mtu += MLXSW_TXHDR_LEN + ETH_HLEN; + if (mtu > mlxsw_sp_port->max_mtu) + return -EINVAL; + + mlxsw_reg_pmtu_pack(pmtu_pl, mlxsw_sp_port->local_port, mtu); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pmtu), pmtu_pl); +} + +static int mlxsw_sp_port_swid_set(struct mlxsw_sp *mlxsw_sp, + u16 local_port, u8 swid) +{ + char pspa_pl[MLXSW_REG_PSPA_LEN]; + + mlxsw_reg_pspa_pack(pspa_pl, swid, local_port); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pspa), pspa_pl); +} + +int mlxsw_sp_port_vp_mode_set(struct mlxsw_sp_port *mlxsw_sp_port, bool enable) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char svpe_pl[MLXSW_REG_SVPE_LEN]; + + mlxsw_reg_svpe_pack(svpe_pl, mlxsw_sp_port->local_port, enable); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(svpe), svpe_pl); +} + +int mlxsw_sp_port_vid_learning_set(struct mlxsw_sp_port *mlxsw_sp_port, u16 vid, + bool learn_enable) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char *spvmlr_pl; + int err; + + spvmlr_pl = kmalloc(MLXSW_REG_SPVMLR_LEN, GFP_KERNEL); + if (!spvmlr_pl) + return -ENOMEM; + mlxsw_reg_spvmlr_pack(spvmlr_pl, mlxsw_sp_port->local_port, vid, vid, + learn_enable); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(spvmlr), spvmlr_pl); + kfree(spvmlr_pl); + return err; +} + +int mlxsw_sp_ethtype_to_sver_type(u16 ethtype, u8 *p_sver_type) +{ + switch (ethtype) { + case ETH_P_8021Q: + *p_sver_type = 0; + break; + case ETH_P_8021AD: + *p_sver_type = 1; + break; + default: + return -EINVAL; + } + + return 0; +} + +int mlxsw_sp_port_egress_ethtype_set(struct mlxsw_sp_port *mlxsw_sp_port, + u16 ethtype) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char spevet_pl[MLXSW_REG_SPEVET_LEN]; + u8 sver_type; + int err; + + err = mlxsw_sp_ethtype_to_sver_type(ethtype, &sver_type); + if (err) + return err; + + mlxsw_reg_spevet_pack(spevet_pl, mlxsw_sp_port->local_port, sver_type); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(spevet), spevet_pl); +} + +static int __mlxsw_sp_port_pvid_set(struct mlxsw_sp_port *mlxsw_sp_port, + u16 vid, u16 ethtype) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char spvid_pl[MLXSW_REG_SPVID_LEN]; + u8 sver_type; + int err; + + err = mlxsw_sp_ethtype_to_sver_type(ethtype, &sver_type); + if (err) + return err; + + mlxsw_reg_spvid_pack(spvid_pl, mlxsw_sp_port->local_port, vid, + sver_type); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(spvid), spvid_pl); +} + +static int mlxsw_sp_port_allow_untagged_set(struct mlxsw_sp_port *mlxsw_sp_port, + bool allow) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char spaft_pl[MLXSW_REG_SPAFT_LEN]; + + mlxsw_reg_spaft_pack(spaft_pl, mlxsw_sp_port->local_port, allow); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(spaft), spaft_pl); +} + +int mlxsw_sp_port_pvid_set(struct mlxsw_sp_port *mlxsw_sp_port, u16 vid, + u16 ethtype) +{ + int err; + + if (!vid) { + err = mlxsw_sp_port_allow_untagged_set(mlxsw_sp_port, false); + if (err) + return err; + } else { + err = __mlxsw_sp_port_pvid_set(mlxsw_sp_port, vid, ethtype); + if (err) + return err; + err = mlxsw_sp_port_allow_untagged_set(mlxsw_sp_port, true); + if (err) + goto err_port_allow_untagged_set; + } + + mlxsw_sp_port->pvid = vid; + return 0; + +err_port_allow_untagged_set: + __mlxsw_sp_port_pvid_set(mlxsw_sp_port, mlxsw_sp_port->pvid, ethtype); + return err; +} + +static int +mlxsw_sp_port_system_port_mapping_set(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char sspr_pl[MLXSW_REG_SSPR_LEN]; + + mlxsw_reg_sspr_pack(sspr_pl, mlxsw_sp_port->local_port); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sspr), sspr_pl); +} + +static int +mlxsw_sp_port_module_info_parse(struct mlxsw_sp *mlxsw_sp, + u16 local_port, char *pmlp_pl, + struct mlxsw_sp_port_mapping *port_mapping) +{ + bool separate_rxtx; + u8 first_lane; + u8 slot_index; + u8 module; + u8 width; + int i; + + module = mlxsw_reg_pmlp_module_get(pmlp_pl, 0); + slot_index = mlxsw_reg_pmlp_slot_index_get(pmlp_pl, 0); + width = mlxsw_reg_pmlp_width_get(pmlp_pl); + separate_rxtx = mlxsw_reg_pmlp_rxtx_get(pmlp_pl); + first_lane = mlxsw_reg_pmlp_tx_lane_get(pmlp_pl, 0); + + if (width && !is_power_of_2(width)) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Unsupported module config: width value is not power of 2\n", + local_port); + return -EINVAL; + } + + for (i = 0; i < width; i++) { + if (mlxsw_reg_pmlp_module_get(pmlp_pl, i) != module) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Unsupported module config: contains multiple modules\n", + local_port); + return -EINVAL; + } + if (mlxsw_reg_pmlp_slot_index_get(pmlp_pl, i) != slot_index) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Unsupported module config: contains multiple slot indexes\n", + local_port); + return -EINVAL; + } + if (separate_rxtx && + mlxsw_reg_pmlp_tx_lane_get(pmlp_pl, i) != + mlxsw_reg_pmlp_rx_lane_get(pmlp_pl, i)) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Unsupported module config: TX and RX lane numbers are different\n", + local_port); + return -EINVAL; + } + if (mlxsw_reg_pmlp_tx_lane_get(pmlp_pl, i) != i + first_lane) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Unsupported module config: TX and RX lane numbers are not sequential\n", + local_port); + return -EINVAL; + } + } + + port_mapping->module = module; + port_mapping->slot_index = slot_index; + port_mapping->width = width; + port_mapping->module_width = width; + port_mapping->lane = mlxsw_reg_pmlp_tx_lane_get(pmlp_pl, 0); + return 0; +} + +static int +mlxsw_sp_port_module_info_get(struct mlxsw_sp *mlxsw_sp, u16 local_port, + struct mlxsw_sp_port_mapping *port_mapping) +{ + char pmlp_pl[MLXSW_REG_PMLP_LEN]; + int err; + + mlxsw_reg_pmlp_pack(pmlp_pl, local_port); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(pmlp), pmlp_pl); + if (err) + return err; + return mlxsw_sp_port_module_info_parse(mlxsw_sp, local_port, + pmlp_pl, port_mapping); +} + +static int +mlxsw_sp_port_module_map(struct mlxsw_sp *mlxsw_sp, u16 local_port, + const struct mlxsw_sp_port_mapping *port_mapping) +{ + char pmlp_pl[MLXSW_REG_PMLP_LEN]; + int i, err; + + mlxsw_env_module_port_map(mlxsw_sp->core, port_mapping->slot_index, + port_mapping->module); + + mlxsw_reg_pmlp_pack(pmlp_pl, local_port); + mlxsw_reg_pmlp_width_set(pmlp_pl, port_mapping->width); + for (i = 0; i < port_mapping->width; i++) { + mlxsw_reg_pmlp_slot_index_set(pmlp_pl, i, + port_mapping->slot_index); + mlxsw_reg_pmlp_module_set(pmlp_pl, i, port_mapping->module); + mlxsw_reg_pmlp_tx_lane_set(pmlp_pl, i, port_mapping->lane + i); /* Rx & Tx */ + } + + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pmlp), pmlp_pl); + if (err) + goto err_pmlp_write; + return 0; + +err_pmlp_write: + mlxsw_env_module_port_unmap(mlxsw_sp->core, port_mapping->slot_index, + port_mapping->module); + return err; +} + +static void mlxsw_sp_port_module_unmap(struct mlxsw_sp *mlxsw_sp, u16 local_port, + u8 slot_index, u8 module) +{ + char pmlp_pl[MLXSW_REG_PMLP_LEN]; + + mlxsw_reg_pmlp_pack(pmlp_pl, local_port); + mlxsw_reg_pmlp_width_set(pmlp_pl, 0); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pmlp), pmlp_pl); + mlxsw_env_module_port_unmap(mlxsw_sp->core, slot_index, module); +} + +static int mlxsw_sp_port_open(struct net_device *dev) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + int err; + + err = mlxsw_env_module_port_up(mlxsw_sp->core, + mlxsw_sp_port->mapping.slot_index, + mlxsw_sp_port->mapping.module); + if (err) + return err; + err = mlxsw_sp_port_admin_status_set(mlxsw_sp_port, true); + if (err) + goto err_port_admin_status_set; + netif_start_queue(dev); + return 0; + +err_port_admin_status_set: + mlxsw_env_module_port_down(mlxsw_sp->core, + mlxsw_sp_port->mapping.slot_index, + mlxsw_sp_port->mapping.module); + return err; +} + +static int mlxsw_sp_port_stop(struct net_device *dev) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + + netif_stop_queue(dev); + mlxsw_sp_port_admin_status_set(mlxsw_sp_port, false); + mlxsw_env_module_port_down(mlxsw_sp->core, + mlxsw_sp_port->mapping.slot_index, + mlxsw_sp_port->mapping.module); + return 0; +} + +static netdev_tx_t mlxsw_sp_port_xmit(struct sk_buff *skb, + struct net_device *dev) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_port_pcpu_stats *pcpu_stats; + const struct mlxsw_tx_info tx_info = { + .local_port = mlxsw_sp_port->local_port, + .is_emad = false, + }; + u64 len; + int err; + + memset(skb->cb, 0, sizeof(struct mlxsw_skb_cb)); + + if (mlxsw_core_skb_transmit_busy(mlxsw_sp->core, &tx_info)) + return NETDEV_TX_BUSY; + + if (eth_skb_pad(skb)) { + this_cpu_inc(mlxsw_sp_port->pcpu_stats->tx_dropped); + return NETDEV_TX_OK; + } + + err = mlxsw_sp_txhdr_handle(mlxsw_sp->core, mlxsw_sp_port, skb, + &tx_info); + if (err) + return NETDEV_TX_OK; + + /* TX header is consumed by HW on the way so we shouldn't count its + * bytes as being sent. + */ + len = skb->len - MLXSW_TXHDR_LEN; + + /* Due to a race we might fail here because of a full queue. In that + * unlikely case we simply drop the packet. + */ + err = mlxsw_core_skb_transmit(mlxsw_sp->core, skb, &tx_info); + + if (!err) { + pcpu_stats = this_cpu_ptr(mlxsw_sp_port->pcpu_stats); + u64_stats_update_begin(&pcpu_stats->syncp); + pcpu_stats->tx_packets++; + pcpu_stats->tx_bytes += len; + u64_stats_update_end(&pcpu_stats->syncp); + } else { + this_cpu_inc(mlxsw_sp_port->pcpu_stats->tx_dropped); + dev_kfree_skb_any(skb); + } + return NETDEV_TX_OK; +} + +static void mlxsw_sp_set_rx_mode(struct net_device *dev) +{ +} + +static int mlxsw_sp_port_set_mac_address(struct net_device *dev, void *p) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct sockaddr *addr = p; + int err; + + if (!is_valid_ether_addr(addr->sa_data)) + return -EADDRNOTAVAIL; + + err = mlxsw_sp_port_dev_addr_set(mlxsw_sp_port, addr->sa_data); + if (err) + return err; + eth_hw_addr_set(dev, addr->sa_data); + return 0; +} + +static int mlxsw_sp_port_change_mtu(struct net_device *dev, int mtu) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp_hdroom orig_hdroom; + struct mlxsw_sp_hdroom hdroom; + int err; + + orig_hdroom = *mlxsw_sp_port->hdroom; + + hdroom = orig_hdroom; + hdroom.mtu = mtu; + mlxsw_sp_hdroom_bufs_reset_sizes(mlxsw_sp_port, &hdroom); + + err = mlxsw_sp_hdroom_configure(mlxsw_sp_port, &hdroom); + if (err) { + netdev_err(dev, "Failed to configure port's headroom\n"); + return err; + } + + err = mlxsw_sp_port_mtu_set(mlxsw_sp_port, mtu); + if (err) + goto err_port_mtu_set; + dev->mtu = mtu; + return 0; + +err_port_mtu_set: + mlxsw_sp_hdroom_configure(mlxsw_sp_port, &orig_hdroom); + return err; +} + +static int +mlxsw_sp_port_get_sw_stats64(const struct net_device *dev, + struct rtnl_link_stats64 *stats) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp_port_pcpu_stats *p; + u64 rx_packets, rx_bytes, tx_packets, tx_bytes; + u32 tx_dropped = 0; + unsigned int start; + int i; + + for_each_possible_cpu(i) { + p = per_cpu_ptr(mlxsw_sp_port->pcpu_stats, i); + do { + start = u64_stats_fetch_begin_irq(&p->syncp); + rx_packets = p->rx_packets; + rx_bytes = p->rx_bytes; + tx_packets = p->tx_packets; + tx_bytes = p->tx_bytes; + } while (u64_stats_fetch_retry_irq(&p->syncp, start)); + + stats->rx_packets += rx_packets; + stats->rx_bytes += rx_bytes; + stats->tx_packets += tx_packets; + stats->tx_bytes += tx_bytes; + /* tx_dropped is u32, updated without syncp protection. */ + tx_dropped += p->tx_dropped; + } + stats->tx_dropped = tx_dropped; + return 0; +} + +static bool mlxsw_sp_port_has_offload_stats(const struct net_device *dev, int attr_id) +{ + switch (attr_id) { + case IFLA_OFFLOAD_XSTATS_CPU_HIT: + return true; + } + + return false; +} + +static int mlxsw_sp_port_get_offload_stats(int attr_id, const struct net_device *dev, + void *sp) +{ + switch (attr_id) { + case IFLA_OFFLOAD_XSTATS_CPU_HIT: + return mlxsw_sp_port_get_sw_stats64(dev, sp); + } + + return -EINVAL; +} + +int mlxsw_sp_port_get_stats_raw(struct net_device *dev, int grp, + int prio, char *ppcnt_pl) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + + mlxsw_reg_ppcnt_pack(ppcnt_pl, mlxsw_sp_port->local_port, grp, prio); + return mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ppcnt), ppcnt_pl); +} + +static int mlxsw_sp_port_get_hw_stats(struct net_device *dev, + struct rtnl_link_stats64 *stats) +{ + char ppcnt_pl[MLXSW_REG_PPCNT_LEN]; + int err; + + err = mlxsw_sp_port_get_stats_raw(dev, MLXSW_REG_PPCNT_IEEE_8023_CNT, + 0, ppcnt_pl); + if (err) + goto out; + + stats->tx_packets = + mlxsw_reg_ppcnt_a_frames_transmitted_ok_get(ppcnt_pl); + stats->rx_packets = + mlxsw_reg_ppcnt_a_frames_received_ok_get(ppcnt_pl); + stats->tx_bytes = + mlxsw_reg_ppcnt_a_octets_transmitted_ok_get(ppcnt_pl); + stats->rx_bytes = + mlxsw_reg_ppcnt_a_octets_received_ok_get(ppcnt_pl); + stats->multicast = + mlxsw_reg_ppcnt_a_multicast_frames_received_ok_get(ppcnt_pl); + + stats->rx_crc_errors = + mlxsw_reg_ppcnt_a_frame_check_sequence_errors_get(ppcnt_pl); + stats->rx_frame_errors = + mlxsw_reg_ppcnt_a_alignment_errors_get(ppcnt_pl); + + stats->rx_length_errors = ( + mlxsw_reg_ppcnt_a_in_range_length_errors_get(ppcnt_pl) + + mlxsw_reg_ppcnt_a_out_of_range_length_field_get(ppcnt_pl) + + mlxsw_reg_ppcnt_a_frame_too_long_errors_get(ppcnt_pl)); + + stats->rx_errors = (stats->rx_crc_errors + + stats->rx_frame_errors + stats->rx_length_errors); + +out: + return err; +} + +static void +mlxsw_sp_port_get_hw_xstats(struct net_device *dev, + struct mlxsw_sp_port_xstats *xstats) +{ + char ppcnt_pl[MLXSW_REG_PPCNT_LEN]; + int err, i; + + err = mlxsw_sp_port_get_stats_raw(dev, MLXSW_REG_PPCNT_EXT_CNT, 0, + ppcnt_pl); + if (!err) + xstats->ecn = mlxsw_reg_ppcnt_ecn_marked_get(ppcnt_pl); + + for (i = 0; i < TC_MAX_QUEUE; i++) { + err = mlxsw_sp_port_get_stats_raw(dev, + MLXSW_REG_PPCNT_TC_CONG_CNT, + i, ppcnt_pl); + if (err) + goto tc_cnt; + + xstats->wred_drop[i] = + mlxsw_reg_ppcnt_wred_discard_get(ppcnt_pl); + xstats->tc_ecn[i] = mlxsw_reg_ppcnt_ecn_marked_tc_get(ppcnt_pl); + +tc_cnt: + err = mlxsw_sp_port_get_stats_raw(dev, MLXSW_REG_PPCNT_TC_CNT, + i, ppcnt_pl); + if (err) + continue; + + xstats->backlog[i] = + mlxsw_reg_ppcnt_tc_transmit_queue_get(ppcnt_pl); + xstats->tail_drop[i] = + mlxsw_reg_ppcnt_tc_no_buffer_discard_uc_get(ppcnt_pl); + } + + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + err = mlxsw_sp_port_get_stats_raw(dev, MLXSW_REG_PPCNT_PRIO_CNT, + i, ppcnt_pl); + if (err) + continue; + + xstats->tx_packets[i] = mlxsw_reg_ppcnt_tx_frames_get(ppcnt_pl); + xstats->tx_bytes[i] = mlxsw_reg_ppcnt_tx_octets_get(ppcnt_pl); + } +} + +static void update_stats_cache(struct work_struct *work) +{ + struct mlxsw_sp_port *mlxsw_sp_port = + container_of(work, struct mlxsw_sp_port, + periodic_hw_stats.update_dw.work); + + if (!netif_carrier_ok(mlxsw_sp_port->dev)) + /* Note: mlxsw_sp_port_down_wipe_counters() clears the cache as + * necessary when port goes down. + */ + goto out; + + mlxsw_sp_port_get_hw_stats(mlxsw_sp_port->dev, + &mlxsw_sp_port->periodic_hw_stats.stats); + mlxsw_sp_port_get_hw_xstats(mlxsw_sp_port->dev, + &mlxsw_sp_port->periodic_hw_stats.xstats); + +out: + mlxsw_core_schedule_dw(&mlxsw_sp_port->periodic_hw_stats.update_dw, + MLXSW_HW_STATS_UPDATE_TIME); +} + +/* Return the stats from a cache that is updated periodically, + * as this function might get called in an atomic context. + */ +static void +mlxsw_sp_port_get_stats64(struct net_device *dev, + struct rtnl_link_stats64 *stats) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + + memcpy(stats, &mlxsw_sp_port->periodic_hw_stats.stats, sizeof(*stats)); +} + +static int __mlxsw_sp_port_vlan_set(struct mlxsw_sp_port *mlxsw_sp_port, + u16 vid_begin, u16 vid_end, + bool is_member, bool untagged) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char *spvm_pl; + int err; + + spvm_pl = kmalloc(MLXSW_REG_SPVM_LEN, GFP_KERNEL); + if (!spvm_pl) + return -ENOMEM; + + mlxsw_reg_spvm_pack(spvm_pl, mlxsw_sp_port->local_port, vid_begin, + vid_end, is_member, untagged); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(spvm), spvm_pl); + kfree(spvm_pl); + return err; +} + +int mlxsw_sp_port_vlan_set(struct mlxsw_sp_port *mlxsw_sp_port, u16 vid_begin, + u16 vid_end, bool is_member, bool untagged) +{ + u16 vid, vid_e; + int err; + + for (vid = vid_begin; vid <= vid_end; + vid += MLXSW_REG_SPVM_REC_MAX_COUNT) { + vid_e = min((u16) (vid + MLXSW_REG_SPVM_REC_MAX_COUNT - 1), + vid_end); + + err = __mlxsw_sp_port_vlan_set(mlxsw_sp_port, vid, vid_e, + is_member, untagged); + if (err) + return err; + } + + return 0; +} + +static void mlxsw_sp_port_vlan_flush(struct mlxsw_sp_port *mlxsw_sp_port, + bool flush_default) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan, *tmp; + + list_for_each_entry_safe(mlxsw_sp_port_vlan, tmp, + &mlxsw_sp_port->vlans_list, list) { + if (!flush_default && + mlxsw_sp_port_vlan->vid == MLXSW_SP_DEFAULT_VID) + continue; + mlxsw_sp_port_vlan_destroy(mlxsw_sp_port_vlan); + } +} + +static void +mlxsw_sp_port_vlan_cleanup(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan) +{ + if (mlxsw_sp_port_vlan->bridge_port) + mlxsw_sp_port_vlan_bridge_leave(mlxsw_sp_port_vlan); + else if (mlxsw_sp_port_vlan->fid) + mlxsw_sp_port_vlan_router_leave(mlxsw_sp_port_vlan); +} + +struct mlxsw_sp_port_vlan * +mlxsw_sp_port_vlan_create(struct mlxsw_sp_port *mlxsw_sp_port, u16 vid) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + bool untagged = vid == MLXSW_SP_DEFAULT_VID; + int err; + + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_find_by_vid(mlxsw_sp_port, vid); + if (mlxsw_sp_port_vlan) + return ERR_PTR(-EEXIST); + + err = mlxsw_sp_port_vlan_set(mlxsw_sp_port, vid, vid, true, untagged); + if (err) + return ERR_PTR(err); + + mlxsw_sp_port_vlan = kzalloc(sizeof(*mlxsw_sp_port_vlan), GFP_KERNEL); + if (!mlxsw_sp_port_vlan) { + err = -ENOMEM; + goto err_port_vlan_alloc; + } + + mlxsw_sp_port_vlan->mlxsw_sp_port = mlxsw_sp_port; + mlxsw_sp_port_vlan->vid = vid; + list_add(&mlxsw_sp_port_vlan->list, &mlxsw_sp_port->vlans_list); + + return mlxsw_sp_port_vlan; + +err_port_vlan_alloc: + mlxsw_sp_port_vlan_set(mlxsw_sp_port, vid, vid, false, false); + return ERR_PTR(err); +} + +void mlxsw_sp_port_vlan_destroy(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan) +{ + struct mlxsw_sp_port *mlxsw_sp_port = mlxsw_sp_port_vlan->mlxsw_sp_port; + u16 vid = mlxsw_sp_port_vlan->vid; + + mlxsw_sp_port_vlan_cleanup(mlxsw_sp_port_vlan); + list_del(&mlxsw_sp_port_vlan->list); + kfree(mlxsw_sp_port_vlan); + mlxsw_sp_port_vlan_set(mlxsw_sp_port, vid, vid, false, false); +} + +static int mlxsw_sp_port_add_vid(struct net_device *dev, + __be16 __always_unused proto, u16 vid) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + + /* VLAN 0 is added to HW filter when device goes up, but it is + * reserved in our case, so simply return. + */ + if (!vid) + return 0; + + return PTR_ERR_OR_ZERO(mlxsw_sp_port_vlan_create(mlxsw_sp_port, vid)); +} + +static int mlxsw_sp_port_kill_vid(struct net_device *dev, + __be16 __always_unused proto, u16 vid) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + + /* VLAN 0 is removed from HW filter when device goes down, but + * it is reserved in our case, so simply return. + */ + if (!vid) + return 0; + + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_find_by_vid(mlxsw_sp_port, vid); + if (!mlxsw_sp_port_vlan) + return 0; + mlxsw_sp_port_vlan_destroy(mlxsw_sp_port_vlan); + + return 0; +} + +static int mlxsw_sp_setup_tc_block(struct mlxsw_sp_port *mlxsw_sp_port, + struct flow_block_offload *f) +{ + switch (f->binder_type) { + case FLOW_BLOCK_BINDER_TYPE_CLSACT_INGRESS: + return mlxsw_sp_setup_tc_block_clsact(mlxsw_sp_port, f, true); + case FLOW_BLOCK_BINDER_TYPE_CLSACT_EGRESS: + return mlxsw_sp_setup_tc_block_clsact(mlxsw_sp_port, f, false); + case FLOW_BLOCK_BINDER_TYPE_RED_EARLY_DROP: + return mlxsw_sp_setup_tc_block_qevent_early_drop(mlxsw_sp_port, f); + case FLOW_BLOCK_BINDER_TYPE_RED_MARK: + return mlxsw_sp_setup_tc_block_qevent_mark(mlxsw_sp_port, f); + default: + return -EOPNOTSUPP; + } +} + +static int mlxsw_sp_setup_tc(struct net_device *dev, enum tc_setup_type type, + void *type_data) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + + switch (type) { + case TC_SETUP_BLOCK: + return mlxsw_sp_setup_tc_block(mlxsw_sp_port, type_data); + case TC_SETUP_QDISC_RED: + return mlxsw_sp_setup_tc_red(mlxsw_sp_port, type_data); + case TC_SETUP_QDISC_PRIO: + return mlxsw_sp_setup_tc_prio(mlxsw_sp_port, type_data); + case TC_SETUP_QDISC_ETS: + return mlxsw_sp_setup_tc_ets(mlxsw_sp_port, type_data); + case TC_SETUP_QDISC_TBF: + return mlxsw_sp_setup_tc_tbf(mlxsw_sp_port, type_data); + case TC_SETUP_QDISC_FIFO: + return mlxsw_sp_setup_tc_fifo(mlxsw_sp_port, type_data); + default: + return -EOPNOTSUPP; + } +} + +static int mlxsw_sp_feature_hw_tc(struct net_device *dev, bool enable) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + + if (!enable) { + if (mlxsw_sp_flow_block_rule_count(mlxsw_sp_port->ing_flow_block) || + mlxsw_sp_flow_block_rule_count(mlxsw_sp_port->eg_flow_block)) { + netdev_err(dev, "Active offloaded tc filters, can't turn hw_tc_offload off\n"); + return -EINVAL; + } + mlxsw_sp_flow_block_disable_inc(mlxsw_sp_port->ing_flow_block); + mlxsw_sp_flow_block_disable_inc(mlxsw_sp_port->eg_flow_block); + } else { + mlxsw_sp_flow_block_disable_dec(mlxsw_sp_port->ing_flow_block); + mlxsw_sp_flow_block_disable_dec(mlxsw_sp_port->eg_flow_block); + } + return 0; +} + +static int mlxsw_sp_feature_loopback(struct net_device *dev, bool enable) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + char pplr_pl[MLXSW_REG_PPLR_LEN]; + int err; + + if (netif_running(dev)) + mlxsw_sp_port_admin_status_set(mlxsw_sp_port, false); + + mlxsw_reg_pplr_pack(pplr_pl, mlxsw_sp_port->local_port, enable); + err = mlxsw_reg_write(mlxsw_sp_port->mlxsw_sp->core, MLXSW_REG(pplr), + pplr_pl); + + if (netif_running(dev)) + mlxsw_sp_port_admin_status_set(mlxsw_sp_port, true); + + return err; +} + +typedef int (*mlxsw_sp_feature_handler)(struct net_device *dev, bool enable); + +static int mlxsw_sp_handle_feature(struct net_device *dev, + netdev_features_t wanted_features, + netdev_features_t feature, + mlxsw_sp_feature_handler feature_handler) +{ + netdev_features_t changes = wanted_features ^ dev->features; + bool enable = !!(wanted_features & feature); + int err; + + if (!(changes & feature)) + return 0; + + err = feature_handler(dev, enable); + if (err) { + netdev_err(dev, "%s feature %pNF failed, err %d\n", + enable ? "Enable" : "Disable", &feature, err); + return err; + } + + if (enable) + dev->features |= feature; + else + dev->features &= ~feature; + + return 0; +} +static int mlxsw_sp_set_features(struct net_device *dev, + netdev_features_t features) +{ + netdev_features_t oper_features = dev->features; + int err = 0; + + err |= mlxsw_sp_handle_feature(dev, features, NETIF_F_HW_TC, + mlxsw_sp_feature_hw_tc); + err |= mlxsw_sp_handle_feature(dev, features, NETIF_F_LOOPBACK, + mlxsw_sp_feature_loopback); + + if (err) { + dev->features = oper_features; + return -EINVAL; + } + + return 0; +} + +static struct devlink_port * +mlxsw_sp_port_get_devlink_port(struct net_device *dev) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + + return mlxsw_core_port_devlink_port_get(mlxsw_sp->core, + mlxsw_sp_port->local_port); +} + +static int mlxsw_sp_port_hwtstamp_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct ifreq *ifr) +{ + struct hwtstamp_config config; + int err; + + if (copy_from_user(&config, ifr->ifr_data, sizeof(config))) + return -EFAULT; + + err = mlxsw_sp_port->mlxsw_sp->ptp_ops->hwtstamp_set(mlxsw_sp_port, + &config); + if (err) + return err; + + if (copy_to_user(ifr->ifr_data, &config, sizeof(config))) + return -EFAULT; + + return 0; +} + +static int mlxsw_sp_port_hwtstamp_get(struct mlxsw_sp_port *mlxsw_sp_port, + struct ifreq *ifr) +{ + struct hwtstamp_config config; + int err; + + err = mlxsw_sp_port->mlxsw_sp->ptp_ops->hwtstamp_get(mlxsw_sp_port, + &config); + if (err) + return err; + + if (copy_to_user(ifr->ifr_data, &config, sizeof(config))) + return -EFAULT; + + return 0; +} + +static inline void mlxsw_sp_port_ptp_clear(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct hwtstamp_config config = {0}; + + mlxsw_sp_port->mlxsw_sp->ptp_ops->hwtstamp_set(mlxsw_sp_port, &config); +} + +static int +mlxsw_sp_port_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + + switch (cmd) { + case SIOCSHWTSTAMP: + return mlxsw_sp_port_hwtstamp_set(mlxsw_sp_port, ifr); + case SIOCGHWTSTAMP: + return mlxsw_sp_port_hwtstamp_get(mlxsw_sp_port, ifr); + default: + return -EOPNOTSUPP; + } +} + +static const struct net_device_ops mlxsw_sp_port_netdev_ops = { + .ndo_open = mlxsw_sp_port_open, + .ndo_stop = mlxsw_sp_port_stop, + .ndo_start_xmit = mlxsw_sp_port_xmit, + .ndo_setup_tc = mlxsw_sp_setup_tc, + .ndo_set_rx_mode = mlxsw_sp_set_rx_mode, + .ndo_set_mac_address = mlxsw_sp_port_set_mac_address, + .ndo_change_mtu = mlxsw_sp_port_change_mtu, + .ndo_get_stats64 = mlxsw_sp_port_get_stats64, + .ndo_has_offload_stats = mlxsw_sp_port_has_offload_stats, + .ndo_get_offload_stats = mlxsw_sp_port_get_offload_stats, + .ndo_vlan_rx_add_vid = mlxsw_sp_port_add_vid, + .ndo_vlan_rx_kill_vid = mlxsw_sp_port_kill_vid, + .ndo_set_features = mlxsw_sp_set_features, + .ndo_get_devlink_port = mlxsw_sp_port_get_devlink_port, + .ndo_eth_ioctl = mlxsw_sp_port_ioctl, +}; + +static int +mlxsw_sp_port_speed_by_width_set(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u32 eth_proto_cap, eth_proto_admin, eth_proto_oper; + const struct mlxsw_sp_port_type_speed_ops *ops; + char ptys_pl[MLXSW_REG_PTYS_LEN]; + u32 eth_proto_cap_masked; + int err; + + ops = mlxsw_sp->port_type_speed_ops; + + /* Set advertised speeds to speeds supported by both the driver + * and the device. + */ + ops->reg_ptys_eth_pack(mlxsw_sp, ptys_pl, mlxsw_sp_port->local_port, + 0, false); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ptys), ptys_pl); + if (err) + return err; + + ops->reg_ptys_eth_unpack(mlxsw_sp, ptys_pl, ð_proto_cap, + ð_proto_admin, ð_proto_oper); + eth_proto_cap_masked = ops->ptys_proto_cap_masked_get(eth_proto_cap); + ops->reg_ptys_eth_pack(mlxsw_sp, ptys_pl, mlxsw_sp_port->local_port, + eth_proto_cap_masked, + mlxsw_sp_port->link.autoneg); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptys), ptys_pl); +} + +int mlxsw_sp_port_speed_get(struct mlxsw_sp_port *mlxsw_sp_port, u32 *speed) +{ + const struct mlxsw_sp_port_type_speed_ops *port_type_speed_ops; + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char ptys_pl[MLXSW_REG_PTYS_LEN]; + u32 eth_proto_oper; + int err; + + port_type_speed_ops = mlxsw_sp->port_type_speed_ops; + port_type_speed_ops->reg_ptys_eth_pack(mlxsw_sp, ptys_pl, + mlxsw_sp_port->local_port, 0, + false); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ptys), ptys_pl); + if (err) + return err; + port_type_speed_ops->reg_ptys_eth_unpack(mlxsw_sp, ptys_pl, NULL, NULL, + ð_proto_oper); + *speed = port_type_speed_ops->from_ptys_speed(mlxsw_sp, eth_proto_oper); + return 0; +} + +int mlxsw_sp_port_ets_set(struct mlxsw_sp_port *mlxsw_sp_port, + enum mlxsw_reg_qeec_hr hr, u8 index, u8 next_index, + bool dwrr, u8 dwrr_weight) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char qeec_pl[MLXSW_REG_QEEC_LEN]; + + mlxsw_reg_qeec_pack(qeec_pl, mlxsw_sp_port->local_port, hr, index, + next_index); + mlxsw_reg_qeec_de_set(qeec_pl, true); + mlxsw_reg_qeec_dwrr_set(qeec_pl, dwrr); + mlxsw_reg_qeec_dwrr_weight_set(qeec_pl, dwrr_weight); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qeec), qeec_pl); +} + +int mlxsw_sp_port_ets_maxrate_set(struct mlxsw_sp_port *mlxsw_sp_port, + enum mlxsw_reg_qeec_hr hr, u8 index, + u8 next_index, u32 maxrate, u8 burst_size) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char qeec_pl[MLXSW_REG_QEEC_LEN]; + + mlxsw_reg_qeec_pack(qeec_pl, mlxsw_sp_port->local_port, hr, index, + next_index); + mlxsw_reg_qeec_mase_set(qeec_pl, true); + mlxsw_reg_qeec_max_shaper_rate_set(qeec_pl, maxrate); + mlxsw_reg_qeec_max_shaper_bs_set(qeec_pl, burst_size); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qeec), qeec_pl); +} + +static int mlxsw_sp_port_min_bw_set(struct mlxsw_sp_port *mlxsw_sp_port, + enum mlxsw_reg_qeec_hr hr, u8 index, + u8 next_index, u32 minrate) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char qeec_pl[MLXSW_REG_QEEC_LEN]; + + mlxsw_reg_qeec_pack(qeec_pl, mlxsw_sp_port->local_port, hr, index, + next_index); + mlxsw_reg_qeec_mise_set(qeec_pl, true); + mlxsw_reg_qeec_min_shaper_rate_set(qeec_pl, minrate); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qeec), qeec_pl); +} + +int mlxsw_sp_port_prio_tc_set(struct mlxsw_sp_port *mlxsw_sp_port, + u8 switch_prio, u8 tclass) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char qtct_pl[MLXSW_REG_QTCT_LEN]; + + mlxsw_reg_qtct_pack(qtct_pl, mlxsw_sp_port->local_port, switch_prio, + tclass); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qtct), qtct_pl); +} + +static int mlxsw_sp_port_ets_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + int err, i; + + /* Setup the elements hierarcy, so that each TC is linked to + * one subgroup, which are all member in the same group. + */ + err = mlxsw_sp_port_ets_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_GROUP, 0, 0, false, 0); + if (err) + return err; + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + err = mlxsw_sp_port_ets_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, i, + 0, false, 0); + if (err) + return err; + } + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + err = mlxsw_sp_port_ets_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_TC, i, i, + false, 0); + if (err) + return err; + + err = mlxsw_sp_port_ets_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_TC, + i + 8, i, + true, 100); + if (err) + return err; + } + + /* Make sure the max shaper is disabled in all hierarchies that support + * it. Note that this disables ptps (PTP shaper), but that is intended + * for the initial configuration. + */ + err = mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_PORT, 0, 0, + MLXSW_REG_QEEC_MAS_DIS, 0); + if (err) + return err; + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + err = mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, + i, 0, + MLXSW_REG_QEEC_MAS_DIS, 0); + if (err) + return err; + } + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + err = mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_TC, + i, i, + MLXSW_REG_QEEC_MAS_DIS, 0); + if (err) + return err; + + err = mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_TC, + i + 8, i, + MLXSW_REG_QEEC_MAS_DIS, 0); + if (err) + return err; + } + + /* Configure the min shaper for multicast TCs. */ + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + err = mlxsw_sp_port_min_bw_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_TC, + i + 8, i, + MLXSW_REG_QEEC_MIS_MIN); + if (err) + return err; + } + + /* Map all priorities to traffic class 0. */ + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + err = mlxsw_sp_port_prio_tc_set(mlxsw_sp_port, i, 0); + if (err) + return err; + } + + return 0; +} + +static int mlxsw_sp_port_tc_mc_mode_set(struct mlxsw_sp_port *mlxsw_sp_port, + bool enable) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char qtctm_pl[MLXSW_REG_QTCTM_LEN]; + + mlxsw_reg_qtctm_pack(qtctm_pl, mlxsw_sp_port->local_port, enable); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qtctm), qtctm_pl); +} + +static int mlxsw_sp_port_overheat_init_val_set(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u8 slot_index = mlxsw_sp_port->mapping.slot_index; + u8 module = mlxsw_sp_port->mapping.module; + u64 overheat_counter; + int err; + + err = mlxsw_env_module_overheat_counter_get(mlxsw_sp->core, slot_index, + module, &overheat_counter); + if (err) + return err; + + mlxsw_sp_port->module_overheat_initial_val = overheat_counter; + return 0; +} + +int +mlxsw_sp_port_vlan_classification_set(struct mlxsw_sp_port *mlxsw_sp_port, + bool is_8021ad_tagged, + bool is_8021q_tagged) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char spvc_pl[MLXSW_REG_SPVC_LEN]; + + mlxsw_reg_spvc_pack(spvc_pl, mlxsw_sp_port->local_port, + is_8021ad_tagged, is_8021q_tagged); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(spvc), spvc_pl); +} + +static int mlxsw_sp_port_label_info_get(struct mlxsw_sp *mlxsw_sp, + u16 local_port, u8 *port_number, + u8 *split_port_subnumber, + u8 *slot_index) +{ + char pllp_pl[MLXSW_REG_PLLP_LEN]; + int err; + + mlxsw_reg_pllp_pack(pllp_pl, local_port); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(pllp), pllp_pl); + if (err) + return err; + mlxsw_reg_pllp_unpack(pllp_pl, port_number, + split_port_subnumber, slot_index); + return 0; +} + +static int mlxsw_sp_port_create(struct mlxsw_sp *mlxsw_sp, u16 local_port, + bool split, + struct mlxsw_sp_port_mapping *port_mapping) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + struct mlxsw_sp_port *mlxsw_sp_port; + u32 lanes = port_mapping->width; + u8 split_port_subnumber; + struct net_device *dev; + u8 port_number; + u8 slot_index; + bool splittable; + int err; + + err = mlxsw_sp_port_module_map(mlxsw_sp, local_port, port_mapping); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to map module\n", + local_port); + return err; + } + + err = mlxsw_sp_port_swid_set(mlxsw_sp, local_port, 0); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to set SWID\n", + local_port); + goto err_port_swid_set; + } + + err = mlxsw_sp_port_label_info_get(mlxsw_sp, local_port, &port_number, + &split_port_subnumber, &slot_index); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to get port label information\n", + local_port); + goto err_port_label_info_get; + } + + splittable = lanes > 1 && !split; + err = mlxsw_core_port_init(mlxsw_sp->core, local_port, slot_index, + port_number, split, split_port_subnumber, + splittable, lanes, mlxsw_sp->base_mac, + sizeof(mlxsw_sp->base_mac)); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to init core port\n", + local_port); + goto err_core_port_init; + } + + dev = alloc_etherdev(sizeof(struct mlxsw_sp_port)); + if (!dev) { + err = -ENOMEM; + goto err_alloc_etherdev; + } + SET_NETDEV_DEV(dev, mlxsw_sp->bus_info->dev); + dev_net_set(dev, mlxsw_sp_net(mlxsw_sp)); + mlxsw_sp_port = netdev_priv(dev); + mlxsw_sp_port->dev = dev; + mlxsw_sp_port->mlxsw_sp = mlxsw_sp; + mlxsw_sp_port->local_port = local_port; + mlxsw_sp_port->pvid = MLXSW_SP_DEFAULT_VID; + mlxsw_sp_port->split = split; + mlxsw_sp_port->mapping = *port_mapping; + mlxsw_sp_port->link.autoneg = 1; + INIT_LIST_HEAD(&mlxsw_sp_port->vlans_list); + + mlxsw_sp_port->pcpu_stats = + netdev_alloc_pcpu_stats(struct mlxsw_sp_port_pcpu_stats); + if (!mlxsw_sp_port->pcpu_stats) { + err = -ENOMEM; + goto err_alloc_stats; + } + + INIT_DELAYED_WORK(&mlxsw_sp_port->periodic_hw_stats.update_dw, + &update_stats_cache); + + dev->netdev_ops = &mlxsw_sp_port_netdev_ops; + dev->ethtool_ops = &mlxsw_sp_port_ethtool_ops; + + err = mlxsw_sp_port_dev_addr_init(mlxsw_sp_port); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Unable to init port mac address\n", + mlxsw_sp_port->local_port); + goto err_dev_addr_init; + } + + netif_carrier_off(dev); + + dev->features |= NETIF_F_NETNS_LOCAL | NETIF_F_LLTX | NETIF_F_SG | + NETIF_F_HW_VLAN_CTAG_FILTER | NETIF_F_HW_TC; + dev->hw_features |= NETIF_F_HW_TC | NETIF_F_LOOPBACK; + + dev->min_mtu = 0; + dev->max_mtu = ETH_MAX_MTU; + + /* Each packet needs to have a Tx header (metadata) on top all other + * headers. + */ + dev->needed_headroom = MLXSW_TXHDR_LEN; + + err = mlxsw_sp_port_system_port_mapping_set(mlxsw_sp_port); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to set system port mapping\n", + mlxsw_sp_port->local_port); + goto err_port_system_port_mapping_set; + } + + err = mlxsw_sp_port_speed_by_width_set(mlxsw_sp_port); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to enable speeds\n", + mlxsw_sp_port->local_port); + goto err_port_speed_by_width_set; + } + + err = mlxsw_sp->port_type_speed_ops->ptys_max_speed(mlxsw_sp_port, + &mlxsw_sp_port->max_speed); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to get maximum speed\n", + mlxsw_sp_port->local_port); + goto err_max_speed_get; + } + + err = mlxsw_sp_port_max_mtu_get(mlxsw_sp_port, &mlxsw_sp_port->max_mtu); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to get maximum MTU\n", + mlxsw_sp_port->local_port); + goto err_port_max_mtu_get; + } + + err = mlxsw_sp_port_mtu_set(mlxsw_sp_port, ETH_DATA_LEN); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to set MTU\n", + mlxsw_sp_port->local_port); + goto err_port_mtu_set; + } + + err = mlxsw_sp_port_admin_status_set(mlxsw_sp_port, false); + if (err) + goto err_port_admin_status_set; + + err = mlxsw_sp_port_buffers_init(mlxsw_sp_port); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to initialize buffers\n", + mlxsw_sp_port->local_port); + goto err_port_buffers_init; + } + + err = mlxsw_sp_port_ets_init(mlxsw_sp_port); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to initialize ETS\n", + mlxsw_sp_port->local_port); + goto err_port_ets_init; + } + + err = mlxsw_sp_port_tc_mc_mode_set(mlxsw_sp_port, true); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to initialize TC MC mode\n", + mlxsw_sp_port->local_port); + goto err_port_tc_mc_mode; + } + + /* ETS and buffers must be initialized before DCB. */ + err = mlxsw_sp_port_dcb_init(mlxsw_sp_port); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to initialize DCB\n", + mlxsw_sp_port->local_port); + goto err_port_dcb_init; + } + + err = mlxsw_sp_port_fids_init(mlxsw_sp_port); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to initialize FIDs\n", + mlxsw_sp_port->local_port); + goto err_port_fids_init; + } + + err = mlxsw_sp_tc_qdisc_init(mlxsw_sp_port); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to initialize TC qdiscs\n", + mlxsw_sp_port->local_port); + goto err_port_qdiscs_init; + } + + err = mlxsw_sp_port_vlan_set(mlxsw_sp_port, 0, VLAN_N_VID - 1, false, + false); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to clear VLAN filter\n", + mlxsw_sp_port->local_port); + goto err_port_vlan_clear; + } + + err = mlxsw_sp_port_nve_init(mlxsw_sp_port); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to initialize NVE\n", + mlxsw_sp_port->local_port); + goto err_port_nve_init; + } + + err = mlxsw_sp_port_pvid_set(mlxsw_sp_port, MLXSW_SP_DEFAULT_VID, + ETH_P_8021Q); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to set PVID\n", + mlxsw_sp_port->local_port); + goto err_port_pvid_set; + } + + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_create(mlxsw_sp_port, + MLXSW_SP_DEFAULT_VID); + if (IS_ERR(mlxsw_sp_port_vlan)) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to create VID 1\n", + mlxsw_sp_port->local_port); + err = PTR_ERR(mlxsw_sp_port_vlan); + goto err_port_vlan_create; + } + mlxsw_sp_port->default_vlan = mlxsw_sp_port_vlan; + + /* Set SPVC.et0=true and SPVC.et1=false to make the local port to treat + * only packets with 802.1q header as tagged packets. + */ + err = mlxsw_sp_port_vlan_classification_set(mlxsw_sp_port, false, true); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to set default VLAN classification\n", + local_port); + goto err_port_vlan_classification_set; + } + + INIT_DELAYED_WORK(&mlxsw_sp_port->ptp.shaper_dw, + mlxsw_sp->ptp_ops->shaper_work); + + mlxsw_sp->ports[local_port] = mlxsw_sp_port; + + err = mlxsw_sp_port_overheat_init_val_set(mlxsw_sp_port); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to set overheat initial value\n", + mlxsw_sp_port->local_port); + goto err_port_overheat_init_val_set; + } + + err = register_netdev(dev); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to register netdev\n", + mlxsw_sp_port->local_port); + goto err_register_netdev; + } + + mlxsw_core_port_eth_set(mlxsw_sp->core, mlxsw_sp_port->local_port, + mlxsw_sp_port, dev); + mlxsw_core_schedule_dw(&mlxsw_sp_port->periodic_hw_stats.update_dw, 0); + return 0; + +err_register_netdev: +err_port_overheat_init_val_set: + mlxsw_sp_port_vlan_classification_set(mlxsw_sp_port, true, true); +err_port_vlan_classification_set: + mlxsw_sp->ports[local_port] = NULL; + mlxsw_sp_port_vlan_destroy(mlxsw_sp_port_vlan); +err_port_vlan_create: +err_port_pvid_set: + mlxsw_sp_port_nve_fini(mlxsw_sp_port); +err_port_nve_init: +err_port_vlan_clear: + mlxsw_sp_tc_qdisc_fini(mlxsw_sp_port); +err_port_qdiscs_init: + mlxsw_sp_port_fids_fini(mlxsw_sp_port); +err_port_fids_init: + mlxsw_sp_port_dcb_fini(mlxsw_sp_port); +err_port_dcb_init: + mlxsw_sp_port_tc_mc_mode_set(mlxsw_sp_port, false); +err_port_tc_mc_mode: +err_port_ets_init: + mlxsw_sp_port_buffers_fini(mlxsw_sp_port); +err_port_buffers_init: +err_port_admin_status_set: +err_port_mtu_set: +err_port_max_mtu_get: +err_max_speed_get: +err_port_speed_by_width_set: +err_port_system_port_mapping_set: +err_dev_addr_init: + free_percpu(mlxsw_sp_port->pcpu_stats); +err_alloc_stats: + free_netdev(dev); +err_alloc_etherdev: + mlxsw_core_port_fini(mlxsw_sp->core, local_port); +err_core_port_init: +err_port_label_info_get: + mlxsw_sp_port_swid_set(mlxsw_sp, local_port, + MLXSW_PORT_SWID_DISABLED_PORT); +err_port_swid_set: + mlxsw_sp_port_module_unmap(mlxsw_sp, local_port, + port_mapping->slot_index, + port_mapping->module); + return err; +} + +static void mlxsw_sp_port_remove(struct mlxsw_sp *mlxsw_sp, u16 local_port) +{ + struct mlxsw_sp_port *mlxsw_sp_port = mlxsw_sp->ports[local_port]; + u8 slot_index = mlxsw_sp_port->mapping.slot_index; + u8 module = mlxsw_sp_port->mapping.module; + + cancel_delayed_work_sync(&mlxsw_sp_port->periodic_hw_stats.update_dw); + cancel_delayed_work_sync(&mlxsw_sp_port->ptp.shaper_dw); + mlxsw_core_port_clear(mlxsw_sp->core, local_port, mlxsw_sp); + unregister_netdev(mlxsw_sp_port->dev); /* This calls ndo_stop */ + mlxsw_sp_port_ptp_clear(mlxsw_sp_port); + mlxsw_sp_port_vlan_classification_set(mlxsw_sp_port, true, true); + mlxsw_sp->ports[local_port] = NULL; + mlxsw_sp_port_vlan_flush(mlxsw_sp_port, true); + mlxsw_sp_port_nve_fini(mlxsw_sp_port); + mlxsw_sp_tc_qdisc_fini(mlxsw_sp_port); + mlxsw_sp_port_fids_fini(mlxsw_sp_port); + mlxsw_sp_port_dcb_fini(mlxsw_sp_port); + mlxsw_sp_port_tc_mc_mode_set(mlxsw_sp_port, false); + mlxsw_sp_port_buffers_fini(mlxsw_sp_port); + free_percpu(mlxsw_sp_port->pcpu_stats); + WARN_ON_ONCE(!list_empty(&mlxsw_sp_port->vlans_list)); + free_netdev(mlxsw_sp_port->dev); + mlxsw_core_port_fini(mlxsw_sp->core, local_port); + mlxsw_sp_port_swid_set(mlxsw_sp, local_port, + MLXSW_PORT_SWID_DISABLED_PORT); + mlxsw_sp_port_module_unmap(mlxsw_sp, local_port, slot_index, module); +} + +static int mlxsw_sp_cpu_port_create(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_port *mlxsw_sp_port; + int err; + + mlxsw_sp_port = kzalloc(sizeof(*mlxsw_sp_port), GFP_KERNEL); + if (!mlxsw_sp_port) + return -ENOMEM; + + mlxsw_sp_port->mlxsw_sp = mlxsw_sp; + mlxsw_sp_port->local_port = MLXSW_PORT_CPU_PORT; + + err = mlxsw_core_cpu_port_init(mlxsw_sp->core, + mlxsw_sp_port, + mlxsw_sp->base_mac, + sizeof(mlxsw_sp->base_mac)); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize core CPU port\n"); + goto err_core_cpu_port_init; + } + + mlxsw_sp->ports[MLXSW_PORT_CPU_PORT] = mlxsw_sp_port; + return 0; + +err_core_cpu_port_init: + kfree(mlxsw_sp_port); + return err; +} + +static void mlxsw_sp_cpu_port_remove(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_port *mlxsw_sp_port = + mlxsw_sp->ports[MLXSW_PORT_CPU_PORT]; + + mlxsw_core_cpu_port_fini(mlxsw_sp->core); + mlxsw_sp->ports[MLXSW_PORT_CPU_PORT] = NULL; + kfree(mlxsw_sp_port); +} + +static bool mlxsw_sp_local_port_valid(u16 local_port) +{ + return local_port != MLXSW_PORT_CPU_PORT; +} + +static bool mlxsw_sp_port_created(struct mlxsw_sp *mlxsw_sp, u16 local_port) +{ + if (!mlxsw_sp_local_port_valid(local_port)) + return false; + return mlxsw_sp->ports[local_port] != NULL; +} + +static int mlxsw_sp_port_mapping_event_set(struct mlxsw_sp *mlxsw_sp, + u16 local_port, bool enable) +{ + char pmecr_pl[MLXSW_REG_PMECR_LEN]; + + mlxsw_reg_pmecr_pack(pmecr_pl, local_port, + enable ? MLXSW_REG_PMECR_E_GENERATE_EVENT : + MLXSW_REG_PMECR_E_DO_NOT_GENERATE_EVENT); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pmecr), pmecr_pl); +} + +struct mlxsw_sp_port_mapping_event { + struct list_head list; + char pmlp_pl[MLXSW_REG_PMLP_LEN]; +}; + +static void mlxsw_sp_port_mapping_events_work(struct work_struct *work) +{ + struct mlxsw_sp_port_mapping_event *event, *next_event; + struct mlxsw_sp_port_mapping_events *events; + struct mlxsw_sp_port_mapping port_mapping; + struct mlxsw_sp *mlxsw_sp; + struct devlink *devlink; + LIST_HEAD(event_queue); + u16 local_port; + int err; + + events = container_of(work, struct mlxsw_sp_port_mapping_events, work); + mlxsw_sp = container_of(events, struct mlxsw_sp, port_mapping_events); + devlink = priv_to_devlink(mlxsw_sp->core); + + spin_lock_bh(&events->queue_lock); + list_splice_init(&events->queue, &event_queue); + spin_unlock_bh(&events->queue_lock); + + list_for_each_entry_safe(event, next_event, &event_queue, list) { + local_port = mlxsw_reg_pmlp_local_port_get(event->pmlp_pl); + err = mlxsw_sp_port_module_info_parse(mlxsw_sp, local_port, + event->pmlp_pl, &port_mapping); + if (err) + goto out; + + if (WARN_ON_ONCE(!port_mapping.width)) + goto out; + + devl_lock(devlink); + + if (!mlxsw_sp_port_created(mlxsw_sp, local_port)) + mlxsw_sp_port_create(mlxsw_sp, local_port, + false, &port_mapping); + else + WARN_ON_ONCE(1); + + devl_unlock(devlink); + + mlxsw_sp->port_mapping[local_port] = port_mapping; + +out: + kfree(event); + } +} + +static void +mlxsw_sp_port_mapping_listener_func(const struct mlxsw_reg_info *reg, + char *pmlp_pl, void *priv) +{ + struct mlxsw_sp_port_mapping_events *events; + struct mlxsw_sp_port_mapping_event *event; + struct mlxsw_sp *mlxsw_sp = priv; + u16 local_port; + + local_port = mlxsw_reg_pmlp_local_port_get(pmlp_pl); + if (WARN_ON_ONCE(!mlxsw_sp_local_port_is_valid(mlxsw_sp, local_port))) + return; + + events = &mlxsw_sp->port_mapping_events; + event = kmalloc(sizeof(*event), GFP_ATOMIC); + if (!event) + return; + memcpy(event->pmlp_pl, pmlp_pl, sizeof(event->pmlp_pl)); + spin_lock(&events->queue_lock); + list_add_tail(&event->list, &events->queue); + spin_unlock(&events->queue_lock); + mlxsw_core_schedule_work(&events->work); +} + +static void +__mlxsw_sp_port_mapping_events_cancel(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_port_mapping_event *event, *next_event; + struct mlxsw_sp_port_mapping_events *events; + + events = &mlxsw_sp->port_mapping_events; + + /* Caller needs to make sure that no new event is going to appear. */ + cancel_work_sync(&events->work); + list_for_each_entry_safe(event, next_event, &events->queue, list) { + list_del(&event->list); + kfree(event); + } +} + +static void mlxsw_sp_ports_remove(struct mlxsw_sp *mlxsw_sp) +{ + unsigned int max_ports = mlxsw_core_max_ports(mlxsw_sp->core); + int i; + + for (i = 1; i < max_ports; i++) + mlxsw_sp_port_mapping_event_set(mlxsw_sp, i, false); + /* Make sure all scheduled events are processed */ + __mlxsw_sp_port_mapping_events_cancel(mlxsw_sp); + + for (i = 1; i < max_ports; i++) + if (mlxsw_sp_port_created(mlxsw_sp, i)) + mlxsw_sp_port_remove(mlxsw_sp, i); + mlxsw_sp_cpu_port_remove(mlxsw_sp); + kfree(mlxsw_sp->ports); + mlxsw_sp->ports = NULL; +} + +static void +mlxsw_sp_ports_remove_selected(struct mlxsw_core *mlxsw_core, + bool (*selector)(void *priv, u16 local_port), + void *priv) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + unsigned int max_ports = mlxsw_core_max_ports(mlxsw_core); + int i; + + for (i = 1; i < max_ports; i++) + if (mlxsw_sp_port_created(mlxsw_sp, i) && selector(priv, i)) + mlxsw_sp_port_remove(mlxsw_sp, i); +} + +static int mlxsw_sp_ports_create(struct mlxsw_sp *mlxsw_sp) +{ + unsigned int max_ports = mlxsw_core_max_ports(mlxsw_sp->core); + struct mlxsw_sp_port_mapping_events *events; + struct mlxsw_sp_port_mapping *port_mapping; + size_t alloc_size; + int i; + int err; + + alloc_size = sizeof(struct mlxsw_sp_port *) * max_ports; + mlxsw_sp->ports = kzalloc(alloc_size, GFP_KERNEL); + if (!mlxsw_sp->ports) + return -ENOMEM; + + events = &mlxsw_sp->port_mapping_events; + INIT_LIST_HEAD(&events->queue); + spin_lock_init(&events->queue_lock); + INIT_WORK(&events->work, mlxsw_sp_port_mapping_events_work); + + for (i = 1; i < max_ports; i++) { + err = mlxsw_sp_port_mapping_event_set(mlxsw_sp, i, true); + if (err) + goto err_event_enable; + } + + err = mlxsw_sp_cpu_port_create(mlxsw_sp); + if (err) + goto err_cpu_port_create; + + for (i = 1; i < max_ports; i++) { + port_mapping = &mlxsw_sp->port_mapping[i]; + if (!port_mapping->width) + continue; + err = mlxsw_sp_port_create(mlxsw_sp, i, false, port_mapping); + if (err) + goto err_port_create; + } + return 0; + +err_port_create: + for (i--; i >= 1; i--) + if (mlxsw_sp_port_created(mlxsw_sp, i)) + mlxsw_sp_port_remove(mlxsw_sp, i); + i = max_ports; + mlxsw_sp_cpu_port_remove(mlxsw_sp); +err_cpu_port_create: +err_event_enable: + for (i--; i >= 1; i--) + mlxsw_sp_port_mapping_event_set(mlxsw_sp, i, false); + /* Make sure all scheduled events are processed */ + __mlxsw_sp_port_mapping_events_cancel(mlxsw_sp); + kfree(mlxsw_sp->ports); + mlxsw_sp->ports = NULL; + return err; +} + +static int mlxsw_sp_port_module_info_init(struct mlxsw_sp *mlxsw_sp) +{ + unsigned int max_ports = mlxsw_core_max_ports(mlxsw_sp->core); + struct mlxsw_sp_port_mapping *port_mapping; + int i; + int err; + + mlxsw_sp->port_mapping = kcalloc(max_ports, + sizeof(struct mlxsw_sp_port_mapping), + GFP_KERNEL); + if (!mlxsw_sp->port_mapping) + return -ENOMEM; + + for (i = 1; i < max_ports; i++) { + port_mapping = &mlxsw_sp->port_mapping[i]; + err = mlxsw_sp_port_module_info_get(mlxsw_sp, i, port_mapping); + if (err) + goto err_port_module_info_get; + } + return 0; + +err_port_module_info_get: + kfree(mlxsw_sp->port_mapping); + return err; +} + +static void mlxsw_sp_port_module_info_fini(struct mlxsw_sp *mlxsw_sp) +{ + kfree(mlxsw_sp->port_mapping); +} + +static int +mlxsw_sp_port_split_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_port_mapping *port_mapping, + unsigned int count, const char *pmtdb_pl) +{ + struct mlxsw_sp_port_mapping split_port_mapping; + int err, i; + + split_port_mapping = *port_mapping; + split_port_mapping.width /= count; + for (i = 0; i < count; i++) { + u16 s_local_port = mlxsw_reg_pmtdb_port_num_get(pmtdb_pl, i); + + if (!mlxsw_sp_local_port_valid(s_local_port)) + continue; + + err = mlxsw_sp_port_create(mlxsw_sp, s_local_port, + true, &split_port_mapping); + if (err) + goto err_port_create; + split_port_mapping.lane += split_port_mapping.width; + } + + return 0; + +err_port_create: + for (i--; i >= 0; i--) { + u16 s_local_port = mlxsw_reg_pmtdb_port_num_get(pmtdb_pl, i); + + if (mlxsw_sp_port_created(mlxsw_sp, s_local_port)) + mlxsw_sp_port_remove(mlxsw_sp, s_local_port); + } + return err; +} + +static void mlxsw_sp_port_unsplit_create(struct mlxsw_sp *mlxsw_sp, + unsigned int count, + const char *pmtdb_pl) +{ + struct mlxsw_sp_port_mapping *port_mapping; + int i; + + /* Go over original unsplit ports in the gap and recreate them. */ + for (i = 0; i < count; i++) { + u16 local_port = mlxsw_reg_pmtdb_port_num_get(pmtdb_pl, i); + + port_mapping = &mlxsw_sp->port_mapping[local_port]; + if (!port_mapping->width || !mlxsw_sp_local_port_valid(local_port)) + continue; + mlxsw_sp_port_create(mlxsw_sp, local_port, + false, port_mapping); + } +} + +static struct mlxsw_sp_port * +mlxsw_sp_port_get_by_local_port(struct mlxsw_sp *mlxsw_sp, u16 local_port) +{ + if (mlxsw_sp->ports && mlxsw_sp->ports[local_port]) + return mlxsw_sp->ports[local_port]; + return NULL; +} + +static int mlxsw_sp_port_split(struct mlxsw_core *mlxsw_core, u16 local_port, + unsigned int count, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + struct mlxsw_sp_port_mapping port_mapping; + struct mlxsw_sp_port *mlxsw_sp_port; + enum mlxsw_reg_pmtdb_status status; + char pmtdb_pl[MLXSW_REG_PMTDB_LEN]; + int i; + int err; + + mlxsw_sp_port = mlxsw_sp_port_get_by_local_port(mlxsw_sp, local_port); + if (!mlxsw_sp_port) { + dev_err(mlxsw_sp->bus_info->dev, "Port number \"%d\" does not exist\n", + local_port); + NL_SET_ERR_MSG_MOD(extack, "Port number does not exist"); + return -EINVAL; + } + + if (mlxsw_sp_port->split) { + NL_SET_ERR_MSG_MOD(extack, "Port is already split"); + return -EINVAL; + } + + mlxsw_reg_pmtdb_pack(pmtdb_pl, mlxsw_sp_port->mapping.slot_index, + mlxsw_sp_port->mapping.module, + mlxsw_sp_port->mapping.module_width / count, + count); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(pmtdb), pmtdb_pl); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to query split info"); + return err; + } + + status = mlxsw_reg_pmtdb_status_get(pmtdb_pl); + if (status != MLXSW_REG_PMTDB_STATUS_SUCCESS) { + NL_SET_ERR_MSG_MOD(extack, "Unsupported split configuration"); + return -EINVAL; + } + + port_mapping = mlxsw_sp_port->mapping; + + for (i = 0; i < count; i++) { + u16 s_local_port = mlxsw_reg_pmtdb_port_num_get(pmtdb_pl, i); + + if (mlxsw_sp_port_created(mlxsw_sp, s_local_port)) + mlxsw_sp_port_remove(mlxsw_sp, s_local_port); + } + + err = mlxsw_sp_port_split_create(mlxsw_sp, &port_mapping, + count, pmtdb_pl); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to create split ports\n"); + goto err_port_split_create; + } + + return 0; + +err_port_split_create: + mlxsw_sp_port_unsplit_create(mlxsw_sp, count, pmtdb_pl); + + return err; +} + +static int mlxsw_sp_port_unsplit(struct mlxsw_core *mlxsw_core, u16 local_port, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + struct mlxsw_sp_port *mlxsw_sp_port; + char pmtdb_pl[MLXSW_REG_PMTDB_LEN]; + unsigned int count; + int i; + int err; + + mlxsw_sp_port = mlxsw_sp_port_get_by_local_port(mlxsw_sp, local_port); + if (!mlxsw_sp_port) { + dev_err(mlxsw_sp->bus_info->dev, "Port number \"%d\" does not exist\n", + local_port); + NL_SET_ERR_MSG_MOD(extack, "Port number does not exist"); + return -EINVAL; + } + + if (!mlxsw_sp_port->split) { + NL_SET_ERR_MSG_MOD(extack, "Port was not split"); + return -EINVAL; + } + + count = mlxsw_sp_port->mapping.module_width / + mlxsw_sp_port->mapping.width; + + mlxsw_reg_pmtdb_pack(pmtdb_pl, mlxsw_sp_port->mapping.slot_index, + mlxsw_sp_port->mapping.module, + mlxsw_sp_port->mapping.module_width / count, + count); + err = mlxsw_reg_query(mlxsw_core, MLXSW_REG(pmtdb), pmtdb_pl); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to query split info"); + return err; + } + + for (i = 0; i < count; i++) { + u16 s_local_port = mlxsw_reg_pmtdb_port_num_get(pmtdb_pl, i); + + if (mlxsw_sp_port_created(mlxsw_sp, s_local_port)) + mlxsw_sp_port_remove(mlxsw_sp, s_local_port); + } + + mlxsw_sp_port_unsplit_create(mlxsw_sp, count, pmtdb_pl); + + return 0; +} + +static void +mlxsw_sp_port_down_wipe_counters(struct mlxsw_sp_port *mlxsw_sp_port) +{ + int i; + + for (i = 0; i < TC_MAX_QUEUE; i++) + mlxsw_sp_port->periodic_hw_stats.xstats.backlog[i] = 0; +} + +static void mlxsw_sp_pude_event_func(const struct mlxsw_reg_info *reg, + char *pude_pl, void *priv) +{ + struct mlxsw_sp *mlxsw_sp = priv; + struct mlxsw_sp_port *mlxsw_sp_port; + enum mlxsw_reg_pude_oper_status status; + u16 local_port; + + local_port = mlxsw_reg_pude_local_port_get(pude_pl); + + if (WARN_ON_ONCE(!mlxsw_sp_local_port_is_valid(mlxsw_sp, local_port))) + return; + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + if (!mlxsw_sp_port) + return; + + status = mlxsw_reg_pude_oper_status_get(pude_pl); + if (status == MLXSW_PORT_OPER_STATUS_UP) { + netdev_info(mlxsw_sp_port->dev, "link up\n"); + netif_carrier_on(mlxsw_sp_port->dev); + mlxsw_core_schedule_dw(&mlxsw_sp_port->ptp.shaper_dw, 0); + } else { + netdev_info(mlxsw_sp_port->dev, "link down\n"); + netif_carrier_off(mlxsw_sp_port->dev); + mlxsw_sp_port_down_wipe_counters(mlxsw_sp_port); + } +} + +static void mlxsw_sp1_ptp_fifo_event_func(struct mlxsw_sp *mlxsw_sp, + char *mtpptr_pl, bool ingress) +{ + u16 local_port; + u8 num_rec; + int i; + + local_port = mlxsw_reg_mtpptr_local_port_get(mtpptr_pl); + num_rec = mlxsw_reg_mtpptr_num_rec_get(mtpptr_pl); + for (i = 0; i < num_rec; i++) { + u8 domain_number; + u8 message_type; + u16 sequence_id; + u64 timestamp; + + mlxsw_reg_mtpptr_unpack(mtpptr_pl, i, &message_type, + &domain_number, &sequence_id, + ×tamp); + mlxsw_sp1_ptp_got_timestamp(mlxsw_sp, ingress, local_port, + message_type, domain_number, + sequence_id, timestamp); + } +} + +static void mlxsw_sp1_ptp_ing_fifo_event_func(const struct mlxsw_reg_info *reg, + char *mtpptr_pl, void *priv) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + mlxsw_sp1_ptp_fifo_event_func(mlxsw_sp, mtpptr_pl, true); +} + +static void mlxsw_sp1_ptp_egr_fifo_event_func(const struct mlxsw_reg_info *reg, + char *mtpptr_pl, void *priv) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + mlxsw_sp1_ptp_fifo_event_func(mlxsw_sp, mtpptr_pl, false); +} + +void mlxsw_sp_rx_listener_no_mark_func(struct sk_buff *skb, + u16 local_port, void *priv) +{ + struct mlxsw_sp *mlxsw_sp = priv; + struct mlxsw_sp_port *mlxsw_sp_port = mlxsw_sp->ports[local_port]; + struct mlxsw_sp_port_pcpu_stats *pcpu_stats; + + if (unlikely(!mlxsw_sp_port)) { + dev_warn_ratelimited(mlxsw_sp->bus_info->dev, "Port %d: skb received for non-existent port\n", + local_port); + return; + } + + skb->dev = mlxsw_sp_port->dev; + + pcpu_stats = this_cpu_ptr(mlxsw_sp_port->pcpu_stats); + u64_stats_update_begin(&pcpu_stats->syncp); + pcpu_stats->rx_packets++; + pcpu_stats->rx_bytes += skb->len; + u64_stats_update_end(&pcpu_stats->syncp); + + skb->protocol = eth_type_trans(skb, skb->dev); + netif_receive_skb(skb); +} + +static void mlxsw_sp_rx_listener_mark_func(struct sk_buff *skb, u16 local_port, + void *priv) +{ + skb->offload_fwd_mark = 1; + return mlxsw_sp_rx_listener_no_mark_func(skb, local_port, priv); +} + +static void mlxsw_sp_rx_listener_l3_mark_func(struct sk_buff *skb, + u16 local_port, void *priv) +{ + skb->offload_l3_fwd_mark = 1; + skb->offload_fwd_mark = 1; + return mlxsw_sp_rx_listener_no_mark_func(skb, local_port, priv); +} + +void mlxsw_sp_ptp_receive(struct mlxsw_sp *mlxsw_sp, struct sk_buff *skb, + u16 local_port) +{ + mlxsw_sp->ptp_ops->receive(mlxsw_sp, skb, local_port); +} + +#define MLXSW_SP_RXL_NO_MARK(_trap_id, _action, _trap_group, _is_ctrl) \ + MLXSW_RXL(mlxsw_sp_rx_listener_no_mark_func, _trap_id, _action, \ + _is_ctrl, SP_##_trap_group, DISCARD) + +#define MLXSW_SP_RXL_MARK(_trap_id, _action, _trap_group, _is_ctrl) \ + MLXSW_RXL(mlxsw_sp_rx_listener_mark_func, _trap_id, _action, \ + _is_ctrl, SP_##_trap_group, DISCARD) + +#define MLXSW_SP_RXL_L3_MARK(_trap_id, _action, _trap_group, _is_ctrl) \ + MLXSW_RXL(mlxsw_sp_rx_listener_l3_mark_func, _trap_id, _action, \ + _is_ctrl, SP_##_trap_group, DISCARD) + +#define MLXSW_SP_EVENTL(_func, _trap_id) \ + MLXSW_EVENTL(_func, _trap_id, SP_EVENT) + +static const struct mlxsw_listener mlxsw_sp_listener[] = { + /* Events */ + MLXSW_SP_EVENTL(mlxsw_sp_pude_event_func, PUDE), + /* L2 traps */ + MLXSW_SP_RXL_NO_MARK(FID_MISS, TRAP_TO_CPU, FID_MISS, false), + /* L3 traps */ + MLXSW_SP_RXL_MARK(IPV6_UNSPECIFIED_ADDRESS, TRAP_TO_CPU, ROUTER_EXP, + false), + MLXSW_SP_RXL_MARK(IPV6_LINK_LOCAL_SRC, TRAP_TO_CPU, ROUTER_EXP, false), + MLXSW_SP_RXL_MARK(IPV6_MC_LINK_LOCAL_DEST, TRAP_TO_CPU, ROUTER_EXP, + false), + MLXSW_SP_RXL_NO_MARK(DISCARD_ING_ROUTER_SIP_CLASS_E, FORWARD, + ROUTER_EXP, false), + MLXSW_SP_RXL_NO_MARK(DISCARD_ING_ROUTER_MC_DMAC, FORWARD, + ROUTER_EXP, false), + MLXSW_SP_RXL_NO_MARK(DISCARD_ING_ROUTER_SIP_DIP, FORWARD, + ROUTER_EXP, false), + MLXSW_SP_RXL_NO_MARK(DISCARD_ING_ROUTER_DIP_LINK_LOCAL, FORWARD, + ROUTER_EXP, false), + /* Multicast Router Traps */ + MLXSW_SP_RXL_MARK(ACL1, TRAP_TO_CPU, MULTICAST, false), + MLXSW_SP_RXL_L3_MARK(ACL2, TRAP_TO_CPU, MULTICAST, false), + /* NVE traps */ + MLXSW_SP_RXL_MARK(NVE_ENCAP_ARP, TRAP_TO_CPU, NEIGH_DISCOVERY, false), +}; + +static const struct mlxsw_listener mlxsw_sp1_listener[] = { + /* Events */ + MLXSW_EVENTL(mlxsw_sp1_ptp_egr_fifo_event_func, PTP_EGR_FIFO, SP_PTP0), + MLXSW_EVENTL(mlxsw_sp1_ptp_ing_fifo_event_func, PTP_ING_FIFO, SP_PTP0), +}; + +static const struct mlxsw_listener mlxsw_sp2_listener[] = { + /* Events */ + MLXSW_SP_EVENTL(mlxsw_sp_port_mapping_listener_func, PMLPE), +}; + +static int mlxsw_sp_cpu_policers_set(struct mlxsw_core *mlxsw_core) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + char qpcr_pl[MLXSW_REG_QPCR_LEN]; + enum mlxsw_reg_qpcr_ir_units ir_units; + int max_cpu_policers; + bool is_bytes; + u8 burst_size; + u32 rate; + int i, err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_core, MAX_CPU_POLICERS)) + return -EIO; + + max_cpu_policers = MLXSW_CORE_RES_GET(mlxsw_core, MAX_CPU_POLICERS); + + ir_units = MLXSW_REG_QPCR_IR_UNITS_M; + for (i = 0; i < max_cpu_policers; i++) { + is_bytes = false; + switch (i) { + case MLXSW_REG_HTGT_TRAP_GROUP_SP_ROUTER_EXP: + case MLXSW_REG_HTGT_TRAP_GROUP_SP_MULTICAST: + case MLXSW_REG_HTGT_TRAP_GROUP_SP_FID_MISS: + rate = 1024; + burst_size = 7; + break; + default: + continue; + } + + __set_bit(i, mlxsw_sp->trap->policers_usage); + mlxsw_reg_qpcr_pack(qpcr_pl, i, ir_units, is_bytes, rate, + burst_size); + err = mlxsw_reg_write(mlxsw_core, MLXSW_REG(qpcr), qpcr_pl); + if (err) + return err; + } + + return 0; +} + +static int mlxsw_sp_trap_groups_set(struct mlxsw_core *mlxsw_core) +{ + char htgt_pl[MLXSW_REG_HTGT_LEN]; + enum mlxsw_reg_htgt_trap_group i; + int max_cpu_policers; + int max_trap_groups; + u8 priority, tc; + u16 policer_id; + int err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_core, MAX_TRAP_GROUPS)) + return -EIO; + + max_trap_groups = MLXSW_CORE_RES_GET(mlxsw_core, MAX_TRAP_GROUPS); + max_cpu_policers = MLXSW_CORE_RES_GET(mlxsw_core, MAX_CPU_POLICERS); + + for (i = 0; i < max_trap_groups; i++) { + policer_id = i; + switch (i) { + case MLXSW_REG_HTGT_TRAP_GROUP_SP_ROUTER_EXP: + case MLXSW_REG_HTGT_TRAP_GROUP_SP_MULTICAST: + case MLXSW_REG_HTGT_TRAP_GROUP_SP_FID_MISS: + priority = 1; + tc = 1; + break; + case MLXSW_REG_HTGT_TRAP_GROUP_SP_EVENT: + priority = MLXSW_REG_HTGT_DEFAULT_PRIORITY; + tc = MLXSW_REG_HTGT_DEFAULT_TC; + policer_id = MLXSW_REG_HTGT_INVALID_POLICER; + break; + default: + continue; + } + + if (max_cpu_policers <= policer_id && + policer_id != MLXSW_REG_HTGT_INVALID_POLICER) + return -EIO; + + mlxsw_reg_htgt_pack(htgt_pl, i, policer_id, priority, tc); + err = mlxsw_reg_write(mlxsw_core, MLXSW_REG(htgt), htgt_pl); + if (err) + return err; + } + + return 0; +} + +static int mlxsw_sp_traps_init(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_trap *trap; + u64 max_policers; + int err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, MAX_CPU_POLICERS)) + return -EIO; + max_policers = MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_CPU_POLICERS); + trap = kzalloc(struct_size(trap, policers_usage, + BITS_TO_LONGS(max_policers)), GFP_KERNEL); + if (!trap) + return -ENOMEM; + trap->max_policers = max_policers; + mlxsw_sp->trap = trap; + + err = mlxsw_sp_cpu_policers_set(mlxsw_sp->core); + if (err) + goto err_cpu_policers_set; + + err = mlxsw_sp_trap_groups_set(mlxsw_sp->core); + if (err) + goto err_trap_groups_set; + + err = mlxsw_core_traps_register(mlxsw_sp->core, mlxsw_sp_listener, + ARRAY_SIZE(mlxsw_sp_listener), + mlxsw_sp); + if (err) + goto err_traps_register; + + err = mlxsw_core_traps_register(mlxsw_sp->core, mlxsw_sp->listeners, + mlxsw_sp->listeners_count, mlxsw_sp); + if (err) + goto err_extra_traps_init; + + return 0; + +err_extra_traps_init: + mlxsw_core_traps_unregister(mlxsw_sp->core, mlxsw_sp_listener, + ARRAY_SIZE(mlxsw_sp_listener), + mlxsw_sp); +err_traps_register: +err_trap_groups_set: +err_cpu_policers_set: + kfree(trap); + return err; +} + +static void mlxsw_sp_traps_fini(struct mlxsw_sp *mlxsw_sp) +{ + mlxsw_core_traps_unregister(mlxsw_sp->core, mlxsw_sp->listeners, + mlxsw_sp->listeners_count, + mlxsw_sp); + mlxsw_core_traps_unregister(mlxsw_sp->core, mlxsw_sp_listener, + ARRAY_SIZE(mlxsw_sp_listener), mlxsw_sp); + kfree(mlxsw_sp->trap); +} + +#define MLXSW_SP_LAG_SEED_INIT 0xcafecafe + +static int mlxsw_sp_lag_init(struct mlxsw_sp *mlxsw_sp) +{ + char slcr_pl[MLXSW_REG_SLCR_LEN]; + u16 max_lag; + u32 seed; + int err; + + seed = jhash(mlxsw_sp->base_mac, sizeof(mlxsw_sp->base_mac), + MLXSW_SP_LAG_SEED_INIT); + mlxsw_reg_slcr_pack(slcr_pl, MLXSW_REG_SLCR_LAG_HASH_SMAC | + MLXSW_REG_SLCR_LAG_HASH_DMAC | + MLXSW_REG_SLCR_LAG_HASH_ETHERTYPE | + MLXSW_REG_SLCR_LAG_HASH_VLANID | + MLXSW_REG_SLCR_LAG_HASH_SIP | + MLXSW_REG_SLCR_LAG_HASH_DIP | + MLXSW_REG_SLCR_LAG_HASH_SPORT | + MLXSW_REG_SLCR_LAG_HASH_DPORT | + MLXSW_REG_SLCR_LAG_HASH_IPPROTO, seed); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(slcr), slcr_pl); + if (err) + return err; + + err = mlxsw_core_max_lag(mlxsw_sp->core, &max_lag); + if (err) + return err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, MAX_LAG_MEMBERS)) + return -EIO; + + mlxsw_sp->lags = kcalloc(max_lag, sizeof(struct mlxsw_sp_upper), + GFP_KERNEL); + if (!mlxsw_sp->lags) + return -ENOMEM; + + return 0; +} + +static void mlxsw_sp_lag_fini(struct mlxsw_sp *mlxsw_sp) +{ + kfree(mlxsw_sp->lags); +} + +static const struct mlxsw_sp_ptp_ops mlxsw_sp1_ptp_ops = { + .clock_init = mlxsw_sp1_ptp_clock_init, + .clock_fini = mlxsw_sp1_ptp_clock_fini, + .init = mlxsw_sp1_ptp_init, + .fini = mlxsw_sp1_ptp_fini, + .receive = mlxsw_sp1_ptp_receive, + .transmitted = mlxsw_sp1_ptp_transmitted, + .hwtstamp_get = mlxsw_sp1_ptp_hwtstamp_get, + .hwtstamp_set = mlxsw_sp1_ptp_hwtstamp_set, + .shaper_work = mlxsw_sp1_ptp_shaper_work, + .get_ts_info = mlxsw_sp1_ptp_get_ts_info, + .get_stats_count = mlxsw_sp1_get_stats_count, + .get_stats_strings = mlxsw_sp1_get_stats_strings, + .get_stats = mlxsw_sp1_get_stats, + .txhdr_construct = mlxsw_sp_ptp_txhdr_construct, +}; + +static const struct mlxsw_sp_ptp_ops mlxsw_sp2_ptp_ops = { + .clock_init = mlxsw_sp2_ptp_clock_init, + .clock_fini = mlxsw_sp2_ptp_clock_fini, + .init = mlxsw_sp2_ptp_init, + .fini = mlxsw_sp2_ptp_fini, + .receive = mlxsw_sp2_ptp_receive, + .transmitted = mlxsw_sp2_ptp_transmitted, + .hwtstamp_get = mlxsw_sp2_ptp_hwtstamp_get, + .hwtstamp_set = mlxsw_sp2_ptp_hwtstamp_set, + .shaper_work = mlxsw_sp2_ptp_shaper_work, + .get_ts_info = mlxsw_sp2_ptp_get_ts_info, + .get_stats_count = mlxsw_sp2_get_stats_count, + .get_stats_strings = mlxsw_sp2_get_stats_strings, + .get_stats = mlxsw_sp2_get_stats, + .txhdr_construct = mlxsw_sp2_ptp_txhdr_construct, +}; + +static const struct mlxsw_sp_ptp_ops mlxsw_sp4_ptp_ops = { + .clock_init = mlxsw_sp2_ptp_clock_init, + .clock_fini = mlxsw_sp2_ptp_clock_fini, + .init = mlxsw_sp2_ptp_init, + .fini = mlxsw_sp2_ptp_fini, + .receive = mlxsw_sp2_ptp_receive, + .transmitted = mlxsw_sp2_ptp_transmitted, + .hwtstamp_get = mlxsw_sp2_ptp_hwtstamp_get, + .hwtstamp_set = mlxsw_sp2_ptp_hwtstamp_set, + .shaper_work = mlxsw_sp2_ptp_shaper_work, + .get_ts_info = mlxsw_sp2_ptp_get_ts_info, + .get_stats_count = mlxsw_sp2_get_stats_count, + .get_stats_strings = mlxsw_sp2_get_stats_strings, + .get_stats = mlxsw_sp2_get_stats, + .txhdr_construct = mlxsw_sp_ptp_txhdr_construct, +}; + +struct mlxsw_sp_sample_trigger_node { + struct mlxsw_sp_sample_trigger trigger; + struct mlxsw_sp_sample_params params; + struct rhash_head ht_node; + struct rcu_head rcu; + refcount_t refcount; +}; + +static const struct rhashtable_params mlxsw_sp_sample_trigger_ht_params = { + .key_offset = offsetof(struct mlxsw_sp_sample_trigger_node, trigger), + .head_offset = offsetof(struct mlxsw_sp_sample_trigger_node, ht_node), + .key_len = sizeof(struct mlxsw_sp_sample_trigger), + .automatic_shrinking = true, +}; + +static void +mlxsw_sp_sample_trigger_key_init(struct mlxsw_sp_sample_trigger *key, + const struct mlxsw_sp_sample_trigger *trigger) +{ + memset(key, 0, sizeof(*key)); + key->type = trigger->type; + key->local_port = trigger->local_port; +} + +/* RCU read lock must be held */ +struct mlxsw_sp_sample_params * +mlxsw_sp_sample_trigger_params_lookup(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_sample_trigger *trigger) +{ + struct mlxsw_sp_sample_trigger_node *trigger_node; + struct mlxsw_sp_sample_trigger key; + + mlxsw_sp_sample_trigger_key_init(&key, trigger); + trigger_node = rhashtable_lookup(&mlxsw_sp->sample_trigger_ht, &key, + mlxsw_sp_sample_trigger_ht_params); + if (!trigger_node) + return NULL; + + return &trigger_node->params; +} + +static int +mlxsw_sp_sample_trigger_node_init(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_sample_trigger *trigger, + const struct mlxsw_sp_sample_params *params) +{ + struct mlxsw_sp_sample_trigger_node *trigger_node; + int err; + + trigger_node = kzalloc(sizeof(*trigger_node), GFP_KERNEL); + if (!trigger_node) + return -ENOMEM; + + trigger_node->trigger = *trigger; + trigger_node->params = *params; + refcount_set(&trigger_node->refcount, 1); + + err = rhashtable_insert_fast(&mlxsw_sp->sample_trigger_ht, + &trigger_node->ht_node, + mlxsw_sp_sample_trigger_ht_params); + if (err) + goto err_rhashtable_insert; + + return 0; + +err_rhashtable_insert: + kfree(trigger_node); + return err; +} + +static void +mlxsw_sp_sample_trigger_node_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_sample_trigger_node *trigger_node) +{ + rhashtable_remove_fast(&mlxsw_sp->sample_trigger_ht, + &trigger_node->ht_node, + mlxsw_sp_sample_trigger_ht_params); + kfree_rcu(trigger_node, rcu); +} + +int +mlxsw_sp_sample_trigger_params_set(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_sample_trigger *trigger, + const struct mlxsw_sp_sample_params *params, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_sample_trigger_node *trigger_node; + struct mlxsw_sp_sample_trigger key; + + ASSERT_RTNL(); + + mlxsw_sp_sample_trigger_key_init(&key, trigger); + + trigger_node = rhashtable_lookup_fast(&mlxsw_sp->sample_trigger_ht, + &key, + mlxsw_sp_sample_trigger_ht_params); + if (!trigger_node) + return mlxsw_sp_sample_trigger_node_init(mlxsw_sp, &key, + params); + + if (trigger_node->trigger.local_port) { + NL_SET_ERR_MSG_MOD(extack, "Sampling already enabled on port"); + return -EINVAL; + } + + if (trigger_node->params.psample_group != params->psample_group || + trigger_node->params.truncate != params->truncate || + trigger_node->params.rate != params->rate || + trigger_node->params.trunc_size != params->trunc_size) { + NL_SET_ERR_MSG_MOD(extack, "Sampling parameters do not match for an existing sampling trigger"); + return -EINVAL; + } + + refcount_inc(&trigger_node->refcount); + + return 0; +} + +void +mlxsw_sp_sample_trigger_params_unset(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_sample_trigger *trigger) +{ + struct mlxsw_sp_sample_trigger_node *trigger_node; + struct mlxsw_sp_sample_trigger key; + + ASSERT_RTNL(); + + mlxsw_sp_sample_trigger_key_init(&key, trigger); + + trigger_node = rhashtable_lookup_fast(&mlxsw_sp->sample_trigger_ht, + &key, + mlxsw_sp_sample_trigger_ht_params); + if (!trigger_node) + return; + + if (!refcount_dec_and_test(&trigger_node->refcount)) + return; + + mlxsw_sp_sample_trigger_node_fini(mlxsw_sp, trigger_node); +} + +static int mlxsw_sp_netdevice_event(struct notifier_block *unused, + unsigned long event, void *ptr); + +#define MLXSW_SP_DEFAULT_PARSING_DEPTH 96 +#define MLXSW_SP_INCREASED_PARSING_DEPTH 128 +#define MLXSW_SP_DEFAULT_VXLAN_UDP_DPORT 4789 + +static void mlxsw_sp_parsing_init(struct mlxsw_sp *mlxsw_sp) +{ + refcount_set(&mlxsw_sp->parsing.parsing_depth_ref, 0); + mlxsw_sp->parsing.parsing_depth = MLXSW_SP_DEFAULT_PARSING_DEPTH; + mlxsw_sp->parsing.vxlan_udp_dport = MLXSW_SP_DEFAULT_VXLAN_UDP_DPORT; + mutex_init(&mlxsw_sp->parsing.lock); +} + +static void mlxsw_sp_parsing_fini(struct mlxsw_sp *mlxsw_sp) +{ + mutex_destroy(&mlxsw_sp->parsing.lock); + WARN_ON_ONCE(refcount_read(&mlxsw_sp->parsing.parsing_depth_ref)); +} + +struct mlxsw_sp_ipv6_addr_node { + struct in6_addr key; + struct rhash_head ht_node; + u32 kvdl_index; + refcount_t refcount; +}; + +static const struct rhashtable_params mlxsw_sp_ipv6_addr_ht_params = { + .key_offset = offsetof(struct mlxsw_sp_ipv6_addr_node, key), + .head_offset = offsetof(struct mlxsw_sp_ipv6_addr_node, ht_node), + .key_len = sizeof(struct in6_addr), + .automatic_shrinking = true, +}; + +static int +mlxsw_sp_ipv6_addr_init(struct mlxsw_sp *mlxsw_sp, const struct in6_addr *addr6, + u32 *p_kvdl_index) +{ + struct mlxsw_sp_ipv6_addr_node *node; + char rips_pl[MLXSW_REG_RIPS_LEN]; + int err; + + err = mlxsw_sp_kvdl_alloc(mlxsw_sp, + MLXSW_SP_KVDL_ENTRY_TYPE_IPV6_ADDRESS, 1, + p_kvdl_index); + if (err) + return err; + + mlxsw_reg_rips_pack(rips_pl, *p_kvdl_index, addr6); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rips), rips_pl); + if (err) + goto err_rips_write; + + node = kzalloc(sizeof(*node), GFP_KERNEL); + if (!node) { + err = -ENOMEM; + goto err_node_alloc; + } + + node->key = *addr6; + node->kvdl_index = *p_kvdl_index; + refcount_set(&node->refcount, 1); + + err = rhashtable_insert_fast(&mlxsw_sp->ipv6_addr_ht, + &node->ht_node, + mlxsw_sp_ipv6_addr_ht_params); + if (err) + goto err_rhashtable_insert; + + return 0; + +err_rhashtable_insert: + kfree(node); +err_node_alloc: +err_rips_write: + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_IPV6_ADDRESS, 1, + *p_kvdl_index); + return err; +} + +static void mlxsw_sp_ipv6_addr_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipv6_addr_node *node) +{ + u32 kvdl_index = node->kvdl_index; + + rhashtable_remove_fast(&mlxsw_sp->ipv6_addr_ht, &node->ht_node, + mlxsw_sp_ipv6_addr_ht_params); + kfree(node); + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_IPV6_ADDRESS, 1, + kvdl_index); +} + +int mlxsw_sp_ipv6_addr_kvdl_index_get(struct mlxsw_sp *mlxsw_sp, + const struct in6_addr *addr6, + u32 *p_kvdl_index) +{ + struct mlxsw_sp_ipv6_addr_node *node; + int err = 0; + + mutex_lock(&mlxsw_sp->ipv6_addr_ht_lock); + node = rhashtable_lookup_fast(&mlxsw_sp->ipv6_addr_ht, addr6, + mlxsw_sp_ipv6_addr_ht_params); + if (node) { + refcount_inc(&node->refcount); + *p_kvdl_index = node->kvdl_index; + goto out_unlock; + } + + err = mlxsw_sp_ipv6_addr_init(mlxsw_sp, addr6, p_kvdl_index); + +out_unlock: + mutex_unlock(&mlxsw_sp->ipv6_addr_ht_lock); + return err; +} + +void +mlxsw_sp_ipv6_addr_put(struct mlxsw_sp *mlxsw_sp, const struct in6_addr *addr6) +{ + struct mlxsw_sp_ipv6_addr_node *node; + + mutex_lock(&mlxsw_sp->ipv6_addr_ht_lock); + node = rhashtable_lookup_fast(&mlxsw_sp->ipv6_addr_ht, addr6, + mlxsw_sp_ipv6_addr_ht_params); + if (WARN_ON(!node)) + goto out_unlock; + + if (!refcount_dec_and_test(&node->refcount)) + goto out_unlock; + + mlxsw_sp_ipv6_addr_fini(mlxsw_sp, node); + +out_unlock: + mutex_unlock(&mlxsw_sp->ipv6_addr_ht_lock); +} + +static int mlxsw_sp_ipv6_addr_ht_init(struct mlxsw_sp *mlxsw_sp) +{ + int err; + + err = rhashtable_init(&mlxsw_sp->ipv6_addr_ht, + &mlxsw_sp_ipv6_addr_ht_params); + if (err) + return err; + + mutex_init(&mlxsw_sp->ipv6_addr_ht_lock); + return 0; +} + +static void mlxsw_sp_ipv6_addr_ht_fini(struct mlxsw_sp *mlxsw_sp) +{ + mutex_destroy(&mlxsw_sp->ipv6_addr_ht_lock); + rhashtable_destroy(&mlxsw_sp->ipv6_addr_ht); +} + +static int mlxsw_sp_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + int err; + + mlxsw_sp->core = mlxsw_core; + mlxsw_sp->bus_info = mlxsw_bus_info; + + mlxsw_sp_parsing_init(mlxsw_sp); + mlxsw_core_emad_string_tlv_enable(mlxsw_core); + + err = mlxsw_sp_base_mac_get(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to get base mac\n"); + return err; + } + + err = mlxsw_sp_kvdl_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize KVDL\n"); + return err; + } + + err = mlxsw_sp_pgt_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize PGT\n"); + goto err_pgt_init; + } + + err = mlxsw_sp_fids_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize FIDs\n"); + goto err_fids_init; + } + + err = mlxsw_sp_policers_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize policers\n"); + goto err_policers_init; + } + + err = mlxsw_sp_traps_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to set traps\n"); + goto err_traps_init; + } + + err = mlxsw_sp_devlink_traps_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize devlink traps\n"); + goto err_devlink_traps_init; + } + + err = mlxsw_sp_buffers_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize buffers\n"); + goto err_buffers_init; + } + + err = mlxsw_sp_lag_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize LAG\n"); + goto err_lag_init; + } + + /* Initialize SPAN before router and switchdev, so that those components + * can call mlxsw_sp_span_respin(). + */ + err = mlxsw_sp_span_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to init span system\n"); + goto err_span_init; + } + + err = mlxsw_sp_switchdev_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize switchdev\n"); + goto err_switchdev_init; + } + + err = mlxsw_sp_counter_pool_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to init counter pool\n"); + goto err_counter_pool_init; + } + + err = mlxsw_sp_afa_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize ACL actions\n"); + goto err_afa_init; + } + + err = mlxsw_sp_ipv6_addr_ht_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize hash table for IPv6 addresses\n"); + goto err_ipv6_addr_ht_init; + } + + err = mlxsw_sp_nve_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize NVE\n"); + goto err_nve_init; + } + + err = mlxsw_sp_acl_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize ACL\n"); + goto err_acl_init; + } + + err = mlxsw_sp_router_init(mlxsw_sp, extack); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize router\n"); + goto err_router_init; + } + + if (mlxsw_sp->bus_info->read_clock_capable) { + /* NULL is a valid return value from clock_init */ + mlxsw_sp->clock = + mlxsw_sp->ptp_ops->clock_init(mlxsw_sp, + mlxsw_sp->bus_info->dev); + if (IS_ERR(mlxsw_sp->clock)) { + err = PTR_ERR(mlxsw_sp->clock); + dev_err(mlxsw_sp->bus_info->dev, "Failed to init ptp clock\n"); + goto err_ptp_clock_init; + } + } + + if (mlxsw_sp->clock) { + /* NULL is a valid return value from ptp_ops->init */ + mlxsw_sp->ptp_state = mlxsw_sp->ptp_ops->init(mlxsw_sp); + if (IS_ERR(mlxsw_sp->ptp_state)) { + err = PTR_ERR(mlxsw_sp->ptp_state); + dev_err(mlxsw_sp->bus_info->dev, "Failed to initialize PTP\n"); + goto err_ptp_init; + } + } + + /* Initialize netdevice notifier after SPAN is initialized, so that the + * event handler can call SPAN respin. + */ + mlxsw_sp->netdevice_nb.notifier_call = mlxsw_sp_netdevice_event; + err = register_netdevice_notifier_net(mlxsw_sp_net(mlxsw_sp), + &mlxsw_sp->netdevice_nb); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to register netdev notifier\n"); + goto err_netdev_notifier; + } + + err = mlxsw_sp_dpipe_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to init pipeline debug\n"); + goto err_dpipe_init; + } + + err = mlxsw_sp_port_module_info_init(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to init port module info\n"); + goto err_port_module_info_init; + } + + err = rhashtable_init(&mlxsw_sp->sample_trigger_ht, + &mlxsw_sp_sample_trigger_ht_params); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to init sampling trigger hashtable\n"); + goto err_sample_trigger_init; + } + + err = mlxsw_sp_ports_create(mlxsw_sp); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to create ports\n"); + goto err_ports_create; + } + + return 0; + +err_ports_create: + rhashtable_destroy(&mlxsw_sp->sample_trigger_ht); +err_sample_trigger_init: + mlxsw_sp_port_module_info_fini(mlxsw_sp); +err_port_module_info_init: + mlxsw_sp_dpipe_fini(mlxsw_sp); +err_dpipe_init: + unregister_netdevice_notifier_net(mlxsw_sp_net(mlxsw_sp), + &mlxsw_sp->netdevice_nb); +err_netdev_notifier: + if (mlxsw_sp->clock) + mlxsw_sp->ptp_ops->fini(mlxsw_sp->ptp_state); +err_ptp_init: + if (mlxsw_sp->clock) + mlxsw_sp->ptp_ops->clock_fini(mlxsw_sp->clock); +err_ptp_clock_init: + mlxsw_sp_router_fini(mlxsw_sp); +err_router_init: + mlxsw_sp_acl_fini(mlxsw_sp); +err_acl_init: + mlxsw_sp_nve_fini(mlxsw_sp); +err_nve_init: + mlxsw_sp_ipv6_addr_ht_fini(mlxsw_sp); +err_ipv6_addr_ht_init: + mlxsw_sp_afa_fini(mlxsw_sp); +err_afa_init: + mlxsw_sp_counter_pool_fini(mlxsw_sp); +err_counter_pool_init: + mlxsw_sp_switchdev_fini(mlxsw_sp); +err_switchdev_init: + mlxsw_sp_span_fini(mlxsw_sp); +err_span_init: + mlxsw_sp_lag_fini(mlxsw_sp); +err_lag_init: + mlxsw_sp_buffers_fini(mlxsw_sp); +err_buffers_init: + mlxsw_sp_devlink_traps_fini(mlxsw_sp); +err_devlink_traps_init: + mlxsw_sp_traps_fini(mlxsw_sp); +err_traps_init: + mlxsw_sp_policers_fini(mlxsw_sp); +err_policers_init: + mlxsw_sp_fids_fini(mlxsw_sp); +err_fids_init: + mlxsw_sp_pgt_fini(mlxsw_sp); +err_pgt_init: + mlxsw_sp_kvdl_fini(mlxsw_sp); + mlxsw_sp_parsing_fini(mlxsw_sp); + return err; +} + +static int mlxsw_sp1_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + + mlxsw_sp->switchdev_ops = &mlxsw_sp1_switchdev_ops; + mlxsw_sp->kvdl_ops = &mlxsw_sp1_kvdl_ops; + mlxsw_sp->afa_ops = &mlxsw_sp1_act_afa_ops; + mlxsw_sp->afk_ops = &mlxsw_sp1_afk_ops; + mlxsw_sp->mr_tcam_ops = &mlxsw_sp1_mr_tcam_ops; + mlxsw_sp->acl_rulei_ops = &mlxsw_sp1_acl_rulei_ops; + mlxsw_sp->acl_tcam_ops = &mlxsw_sp1_acl_tcam_ops; + mlxsw_sp->nve_ops_arr = mlxsw_sp1_nve_ops_arr; + mlxsw_sp->mac_mask = mlxsw_sp1_mac_mask; + mlxsw_sp->sb_vals = &mlxsw_sp1_sb_vals; + mlxsw_sp->sb_ops = &mlxsw_sp1_sb_ops; + mlxsw_sp->port_type_speed_ops = &mlxsw_sp1_port_type_speed_ops; + mlxsw_sp->ptp_ops = &mlxsw_sp1_ptp_ops; + mlxsw_sp->span_ops = &mlxsw_sp1_span_ops; + mlxsw_sp->policer_core_ops = &mlxsw_sp1_policer_core_ops; + mlxsw_sp->trap_ops = &mlxsw_sp1_trap_ops; + mlxsw_sp->mall_ops = &mlxsw_sp1_mall_ops; + mlxsw_sp->router_ops = &mlxsw_sp1_router_ops; + mlxsw_sp->listeners = mlxsw_sp1_listener; + mlxsw_sp->listeners_count = ARRAY_SIZE(mlxsw_sp1_listener); + mlxsw_sp->fid_family_arr = mlxsw_sp1_fid_family_arr; + mlxsw_sp->lowest_shaper_bs = MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP1; + mlxsw_sp->pgt_smpe_index_valid = true; + + return mlxsw_sp_init(mlxsw_core, mlxsw_bus_info, extack); +} + +static int mlxsw_sp2_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + + mlxsw_sp->switchdev_ops = &mlxsw_sp2_switchdev_ops; + mlxsw_sp->kvdl_ops = &mlxsw_sp2_kvdl_ops; + mlxsw_sp->afa_ops = &mlxsw_sp2_act_afa_ops; + mlxsw_sp->afk_ops = &mlxsw_sp2_afk_ops; + mlxsw_sp->mr_tcam_ops = &mlxsw_sp2_mr_tcam_ops; + mlxsw_sp->acl_rulei_ops = &mlxsw_sp2_acl_rulei_ops; + mlxsw_sp->acl_tcam_ops = &mlxsw_sp2_acl_tcam_ops; + mlxsw_sp->acl_bf_ops = &mlxsw_sp2_acl_bf_ops; + mlxsw_sp->nve_ops_arr = mlxsw_sp2_nve_ops_arr; + mlxsw_sp->mac_mask = mlxsw_sp2_mac_mask; + mlxsw_sp->sb_vals = &mlxsw_sp2_sb_vals; + mlxsw_sp->sb_ops = &mlxsw_sp2_sb_ops; + mlxsw_sp->port_type_speed_ops = &mlxsw_sp2_port_type_speed_ops; + mlxsw_sp->ptp_ops = &mlxsw_sp2_ptp_ops; + mlxsw_sp->span_ops = &mlxsw_sp2_span_ops; + mlxsw_sp->policer_core_ops = &mlxsw_sp2_policer_core_ops; + mlxsw_sp->trap_ops = &mlxsw_sp2_trap_ops; + mlxsw_sp->mall_ops = &mlxsw_sp2_mall_ops; + mlxsw_sp->router_ops = &mlxsw_sp2_router_ops; + mlxsw_sp->listeners = mlxsw_sp2_listener; + mlxsw_sp->listeners_count = ARRAY_SIZE(mlxsw_sp2_listener); + mlxsw_sp->fid_family_arr = mlxsw_sp2_fid_family_arr; + mlxsw_sp->lowest_shaper_bs = MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP2; + mlxsw_sp->pgt_smpe_index_valid = false; + + return mlxsw_sp_init(mlxsw_core, mlxsw_bus_info, extack); +} + +static int mlxsw_sp3_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + + mlxsw_sp->switchdev_ops = &mlxsw_sp2_switchdev_ops; + mlxsw_sp->kvdl_ops = &mlxsw_sp2_kvdl_ops; + mlxsw_sp->afa_ops = &mlxsw_sp2_act_afa_ops; + mlxsw_sp->afk_ops = &mlxsw_sp2_afk_ops; + mlxsw_sp->mr_tcam_ops = &mlxsw_sp2_mr_tcam_ops; + mlxsw_sp->acl_rulei_ops = &mlxsw_sp2_acl_rulei_ops; + mlxsw_sp->acl_tcam_ops = &mlxsw_sp2_acl_tcam_ops; + mlxsw_sp->acl_bf_ops = &mlxsw_sp2_acl_bf_ops; + mlxsw_sp->nve_ops_arr = mlxsw_sp2_nve_ops_arr; + mlxsw_sp->mac_mask = mlxsw_sp2_mac_mask; + mlxsw_sp->sb_vals = &mlxsw_sp2_sb_vals; + mlxsw_sp->sb_ops = &mlxsw_sp3_sb_ops; + mlxsw_sp->port_type_speed_ops = &mlxsw_sp2_port_type_speed_ops; + mlxsw_sp->ptp_ops = &mlxsw_sp2_ptp_ops; + mlxsw_sp->span_ops = &mlxsw_sp3_span_ops; + mlxsw_sp->policer_core_ops = &mlxsw_sp2_policer_core_ops; + mlxsw_sp->trap_ops = &mlxsw_sp2_trap_ops; + mlxsw_sp->mall_ops = &mlxsw_sp2_mall_ops; + mlxsw_sp->router_ops = &mlxsw_sp2_router_ops; + mlxsw_sp->listeners = mlxsw_sp2_listener; + mlxsw_sp->listeners_count = ARRAY_SIZE(mlxsw_sp2_listener); + mlxsw_sp->fid_family_arr = mlxsw_sp2_fid_family_arr; + mlxsw_sp->lowest_shaper_bs = MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP3; + mlxsw_sp->pgt_smpe_index_valid = false; + + return mlxsw_sp_init(mlxsw_core, mlxsw_bus_info, extack); +} + +static int mlxsw_sp4_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + + mlxsw_sp->switchdev_ops = &mlxsw_sp2_switchdev_ops; + mlxsw_sp->kvdl_ops = &mlxsw_sp2_kvdl_ops; + mlxsw_sp->afa_ops = &mlxsw_sp2_act_afa_ops; + mlxsw_sp->afk_ops = &mlxsw_sp4_afk_ops; + mlxsw_sp->mr_tcam_ops = &mlxsw_sp2_mr_tcam_ops; + mlxsw_sp->acl_rulei_ops = &mlxsw_sp2_acl_rulei_ops; + mlxsw_sp->acl_tcam_ops = &mlxsw_sp2_acl_tcam_ops; + mlxsw_sp->acl_bf_ops = &mlxsw_sp4_acl_bf_ops; + mlxsw_sp->nve_ops_arr = mlxsw_sp2_nve_ops_arr; + mlxsw_sp->mac_mask = mlxsw_sp2_mac_mask; + mlxsw_sp->sb_vals = &mlxsw_sp2_sb_vals; + mlxsw_sp->sb_ops = &mlxsw_sp3_sb_ops; + mlxsw_sp->port_type_speed_ops = &mlxsw_sp2_port_type_speed_ops; + mlxsw_sp->ptp_ops = &mlxsw_sp4_ptp_ops; + mlxsw_sp->span_ops = &mlxsw_sp3_span_ops; + mlxsw_sp->policer_core_ops = &mlxsw_sp2_policer_core_ops; + mlxsw_sp->trap_ops = &mlxsw_sp2_trap_ops; + mlxsw_sp->mall_ops = &mlxsw_sp2_mall_ops; + mlxsw_sp->router_ops = &mlxsw_sp2_router_ops; + mlxsw_sp->listeners = mlxsw_sp2_listener; + mlxsw_sp->listeners_count = ARRAY_SIZE(mlxsw_sp2_listener); + mlxsw_sp->fid_family_arr = mlxsw_sp2_fid_family_arr; + mlxsw_sp->lowest_shaper_bs = MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP4; + mlxsw_sp->pgt_smpe_index_valid = false; + + return mlxsw_sp_init(mlxsw_core, mlxsw_bus_info, extack); +} + +static void mlxsw_sp_fini(struct mlxsw_core *mlxsw_core) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + + mlxsw_sp_ports_remove(mlxsw_sp); + rhashtable_destroy(&mlxsw_sp->sample_trigger_ht); + mlxsw_sp_port_module_info_fini(mlxsw_sp); + mlxsw_sp_dpipe_fini(mlxsw_sp); + unregister_netdevice_notifier_net(mlxsw_sp_net(mlxsw_sp), + &mlxsw_sp->netdevice_nb); + if (mlxsw_sp->clock) { + mlxsw_sp->ptp_ops->fini(mlxsw_sp->ptp_state); + mlxsw_sp->ptp_ops->clock_fini(mlxsw_sp->clock); + } + mlxsw_sp_router_fini(mlxsw_sp); + mlxsw_sp_acl_fini(mlxsw_sp); + mlxsw_sp_nve_fini(mlxsw_sp); + mlxsw_sp_ipv6_addr_ht_fini(mlxsw_sp); + mlxsw_sp_afa_fini(mlxsw_sp); + mlxsw_sp_counter_pool_fini(mlxsw_sp); + mlxsw_sp_switchdev_fini(mlxsw_sp); + mlxsw_sp_span_fini(mlxsw_sp); + mlxsw_sp_lag_fini(mlxsw_sp); + mlxsw_sp_buffers_fini(mlxsw_sp); + mlxsw_sp_devlink_traps_fini(mlxsw_sp); + mlxsw_sp_traps_fini(mlxsw_sp); + mlxsw_sp_policers_fini(mlxsw_sp); + mlxsw_sp_fids_fini(mlxsw_sp); + mlxsw_sp_pgt_fini(mlxsw_sp); + mlxsw_sp_kvdl_fini(mlxsw_sp); + mlxsw_sp_parsing_fini(mlxsw_sp); +} + +static const struct mlxsw_config_profile mlxsw_sp1_config_profile = { + .used_flood_mode = 1, + .flood_mode = MLXSW_CMD_MBOX_CONFIG_PROFILE_FLOOD_MODE_CONTROLLED, + .used_max_ib_mc = 1, + .max_ib_mc = 0, + .used_max_pkey = 1, + .max_pkey = 0, + .used_ubridge = 1, + .ubridge = 1, + .used_kvd_sizes = 1, + .kvd_hash_single_parts = 59, + .kvd_hash_double_parts = 41, + .kvd_linear_size = MLXSW_SP_KVD_LINEAR_SIZE, + .swid_config = { + { + .used_type = 1, + .type = MLXSW_PORT_SWID_TYPE_ETH, + } + }, +}; + +static const struct mlxsw_config_profile mlxsw_sp2_config_profile = { + .used_flood_mode = 1, + .flood_mode = MLXSW_CMD_MBOX_CONFIG_PROFILE_FLOOD_MODE_CONTROLLED, + .used_max_ib_mc = 1, + .max_ib_mc = 0, + .used_max_pkey = 1, + .max_pkey = 0, + .used_ubridge = 1, + .ubridge = 1, + .swid_config = { + { + .used_type = 1, + .type = MLXSW_PORT_SWID_TYPE_ETH, + } + }, + .used_cqe_time_stamp_type = 1, + .cqe_time_stamp_type = MLXSW_CMD_MBOX_CONFIG_PROFILE_CQE_TIME_STAMP_TYPE_UTC, +}; + +/* Reduce number of LAGs from full capacity (256) to the maximum supported LAGs + * in Spectrum-2/3, to avoid regression in number of free entries in the PGT + * table. + */ +#define MLXSW_SP4_CONFIG_PROFILE_MAX_LAG 128 + +static const struct mlxsw_config_profile mlxsw_sp4_config_profile = { + .used_max_lag = 1, + .max_lag = MLXSW_SP4_CONFIG_PROFILE_MAX_LAG, + .used_flood_mode = 1, + .flood_mode = MLXSW_CMD_MBOX_CONFIG_PROFILE_FLOOD_MODE_CONTROLLED, + .used_max_ib_mc = 1, + .max_ib_mc = 0, + .used_max_pkey = 1, + .max_pkey = 0, + .used_ubridge = 1, + .ubridge = 1, + .swid_config = { + { + .used_type = 1, + .type = MLXSW_PORT_SWID_TYPE_ETH, + } + }, + .used_cqe_time_stamp_type = 1, + .cqe_time_stamp_type = MLXSW_CMD_MBOX_CONFIG_PROFILE_CQE_TIME_STAMP_TYPE_UTC, +}; + +static void +mlxsw_sp_resource_size_params_prepare(struct mlxsw_core *mlxsw_core, + struct devlink_resource_size_params *kvd_size_params, + struct devlink_resource_size_params *linear_size_params, + struct devlink_resource_size_params *hash_double_size_params, + struct devlink_resource_size_params *hash_single_size_params) +{ + u32 single_size_min = MLXSW_CORE_RES_GET(mlxsw_core, + KVD_SINGLE_MIN_SIZE); + u32 double_size_min = MLXSW_CORE_RES_GET(mlxsw_core, + KVD_DOUBLE_MIN_SIZE); + u32 kvd_size = MLXSW_CORE_RES_GET(mlxsw_core, KVD_SIZE); + u32 linear_size_min = 0; + + devlink_resource_size_params_init(kvd_size_params, kvd_size, kvd_size, + MLXSW_SP_KVD_GRANULARITY, + DEVLINK_RESOURCE_UNIT_ENTRY); + devlink_resource_size_params_init(linear_size_params, linear_size_min, + kvd_size - single_size_min - + double_size_min, + MLXSW_SP_KVD_GRANULARITY, + DEVLINK_RESOURCE_UNIT_ENTRY); + devlink_resource_size_params_init(hash_double_size_params, + double_size_min, + kvd_size - single_size_min - + linear_size_min, + MLXSW_SP_KVD_GRANULARITY, + DEVLINK_RESOURCE_UNIT_ENTRY); + devlink_resource_size_params_init(hash_single_size_params, + single_size_min, + kvd_size - double_size_min - + linear_size_min, + MLXSW_SP_KVD_GRANULARITY, + DEVLINK_RESOURCE_UNIT_ENTRY); +} + +static int mlxsw_sp1_resources_kvd_register(struct mlxsw_core *mlxsw_core) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + struct devlink_resource_size_params hash_single_size_params; + struct devlink_resource_size_params hash_double_size_params; + struct devlink_resource_size_params linear_size_params; + struct devlink_resource_size_params kvd_size_params; + u32 kvd_size, single_size, double_size, linear_size; + const struct mlxsw_config_profile *profile; + int err; + + profile = &mlxsw_sp1_config_profile; + if (!MLXSW_CORE_RES_VALID(mlxsw_core, KVD_SIZE)) + return -EIO; + + mlxsw_sp_resource_size_params_prepare(mlxsw_core, &kvd_size_params, + &linear_size_params, + &hash_double_size_params, + &hash_single_size_params); + + kvd_size = MLXSW_CORE_RES_GET(mlxsw_core, KVD_SIZE); + err = devl_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_KVD, + kvd_size, MLXSW_SP_RESOURCE_KVD, + DEVLINK_RESOURCE_ID_PARENT_TOP, + &kvd_size_params); + if (err) + return err; + + linear_size = profile->kvd_linear_size; + err = devl_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_KVD_LINEAR, + linear_size, + MLXSW_SP_RESOURCE_KVD_LINEAR, + MLXSW_SP_RESOURCE_KVD, + &linear_size_params); + if (err) + return err; + + err = mlxsw_sp1_kvdl_resources_register(mlxsw_core); + if (err) + return err; + + double_size = kvd_size - linear_size; + double_size *= profile->kvd_hash_double_parts; + double_size /= profile->kvd_hash_double_parts + + profile->kvd_hash_single_parts; + double_size = rounddown(double_size, MLXSW_SP_KVD_GRANULARITY); + err = devl_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_KVD_HASH_DOUBLE, + double_size, + MLXSW_SP_RESOURCE_KVD_HASH_DOUBLE, + MLXSW_SP_RESOURCE_KVD, + &hash_double_size_params); + if (err) + return err; + + single_size = kvd_size - double_size - linear_size; + err = devl_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_KVD_HASH_SINGLE, + single_size, + MLXSW_SP_RESOURCE_KVD_HASH_SINGLE, + MLXSW_SP_RESOURCE_KVD, + &hash_single_size_params); + if (err) + return err; + + return 0; +} + +static int mlxsw_sp2_resources_kvd_register(struct mlxsw_core *mlxsw_core) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + struct devlink_resource_size_params kvd_size_params; + u32 kvd_size; + + if (!MLXSW_CORE_RES_VALID(mlxsw_core, KVD_SIZE)) + return -EIO; + + kvd_size = MLXSW_CORE_RES_GET(mlxsw_core, KVD_SIZE); + devlink_resource_size_params_init(&kvd_size_params, kvd_size, kvd_size, + MLXSW_SP_KVD_GRANULARITY, + DEVLINK_RESOURCE_UNIT_ENTRY); + + return devl_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_KVD, + kvd_size, MLXSW_SP_RESOURCE_KVD, + DEVLINK_RESOURCE_ID_PARENT_TOP, + &kvd_size_params); +} + +static int mlxsw_sp_resources_span_register(struct mlxsw_core *mlxsw_core) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + struct devlink_resource_size_params span_size_params; + u32 max_span; + + if (!MLXSW_CORE_RES_VALID(mlxsw_core, MAX_SPAN)) + return -EIO; + + max_span = MLXSW_CORE_RES_GET(mlxsw_core, MAX_SPAN); + devlink_resource_size_params_init(&span_size_params, max_span, max_span, + 1, DEVLINK_RESOURCE_UNIT_ENTRY); + + return devl_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_SPAN, + max_span, MLXSW_SP_RESOURCE_SPAN, + DEVLINK_RESOURCE_ID_PARENT_TOP, + &span_size_params); +} + +static int +mlxsw_sp_resources_rif_mac_profile_register(struct mlxsw_core *mlxsw_core) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + struct devlink_resource_size_params size_params; + u8 max_rif_mac_profiles; + + if (!MLXSW_CORE_RES_VALID(mlxsw_core, MAX_RIF_MAC_PROFILES)) + max_rif_mac_profiles = 1; + else + max_rif_mac_profiles = MLXSW_CORE_RES_GET(mlxsw_core, + MAX_RIF_MAC_PROFILES); + devlink_resource_size_params_init(&size_params, max_rif_mac_profiles, + max_rif_mac_profiles, 1, + DEVLINK_RESOURCE_UNIT_ENTRY); + + return devl_resource_register(devlink, + "rif_mac_profiles", + max_rif_mac_profiles, + MLXSW_SP_RESOURCE_RIF_MAC_PROFILES, + DEVLINK_RESOURCE_ID_PARENT_TOP, + &size_params); +} + +static int mlxsw_sp_resources_rifs_register(struct mlxsw_core *mlxsw_core) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + struct devlink_resource_size_params size_params; + u64 max_rifs; + + if (!MLXSW_CORE_RES_VALID(mlxsw_core, MAX_RIFS)) + return -EIO; + + max_rifs = MLXSW_CORE_RES_GET(mlxsw_core, MAX_RIFS); + devlink_resource_size_params_init(&size_params, max_rifs, max_rifs, + 1, DEVLINK_RESOURCE_UNIT_ENTRY); + + return devl_resource_register(devlink, "rifs", max_rifs, + MLXSW_SP_RESOURCE_RIFS, + DEVLINK_RESOURCE_ID_PARENT_TOP, + &size_params); +} + +static int mlxsw_sp1_resources_register(struct mlxsw_core *mlxsw_core) +{ + int err; + + err = mlxsw_sp1_resources_kvd_register(mlxsw_core); + if (err) + return err; + + err = mlxsw_sp_resources_span_register(mlxsw_core); + if (err) + goto err_resources_span_register; + + err = mlxsw_sp_counter_resources_register(mlxsw_core); + if (err) + goto err_resources_counter_register; + + err = mlxsw_sp_policer_resources_register(mlxsw_core); + if (err) + goto err_policer_resources_register; + + err = mlxsw_sp_resources_rif_mac_profile_register(mlxsw_core); + if (err) + goto err_resources_rif_mac_profile_register; + + err = mlxsw_sp_resources_rifs_register(mlxsw_core); + if (err) + goto err_resources_rifs_register; + + return 0; + +err_resources_rifs_register: +err_resources_rif_mac_profile_register: +err_policer_resources_register: +err_resources_counter_register: +err_resources_span_register: + devl_resources_unregister(priv_to_devlink(mlxsw_core)); + return err; +} + +static int mlxsw_sp2_resources_register(struct mlxsw_core *mlxsw_core) +{ + int err; + + err = mlxsw_sp2_resources_kvd_register(mlxsw_core); + if (err) + return err; + + err = mlxsw_sp_resources_span_register(mlxsw_core); + if (err) + goto err_resources_span_register; + + err = mlxsw_sp_counter_resources_register(mlxsw_core); + if (err) + goto err_resources_counter_register; + + err = mlxsw_sp_policer_resources_register(mlxsw_core); + if (err) + goto err_policer_resources_register; + + err = mlxsw_sp_resources_rif_mac_profile_register(mlxsw_core); + if (err) + goto err_resources_rif_mac_profile_register; + + err = mlxsw_sp_resources_rifs_register(mlxsw_core); + if (err) + goto err_resources_rifs_register; + + return 0; + +err_resources_rifs_register: +err_resources_rif_mac_profile_register: +err_policer_resources_register: +err_resources_counter_register: +err_resources_span_register: + devl_resources_unregister(priv_to_devlink(mlxsw_core)); + return err; +} + +static int mlxsw_sp_kvd_sizes_get(struct mlxsw_core *mlxsw_core, + const struct mlxsw_config_profile *profile, + u64 *p_single_size, u64 *p_double_size, + u64 *p_linear_size) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + u32 double_size; + int err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_core, KVD_SINGLE_MIN_SIZE) || + !MLXSW_CORE_RES_VALID(mlxsw_core, KVD_DOUBLE_MIN_SIZE)) + return -EIO; + + /* The hash part is what left of the kvd without the + * linear part. It is split to the single size and + * double size by the parts ratio from the profile. + * Both sizes must be a multiplications of the + * granularity from the profile. In case the user + * provided the sizes they are obtained via devlink. + */ + err = devl_resource_size_get(devlink, + MLXSW_SP_RESOURCE_KVD_LINEAR, + p_linear_size); + if (err) + *p_linear_size = profile->kvd_linear_size; + + err = devl_resource_size_get(devlink, + MLXSW_SP_RESOURCE_KVD_HASH_DOUBLE, + p_double_size); + if (err) { + double_size = MLXSW_CORE_RES_GET(mlxsw_core, KVD_SIZE) - + *p_linear_size; + double_size *= profile->kvd_hash_double_parts; + double_size /= profile->kvd_hash_double_parts + + profile->kvd_hash_single_parts; + *p_double_size = rounddown(double_size, + MLXSW_SP_KVD_GRANULARITY); + } + + err = devl_resource_size_get(devlink, + MLXSW_SP_RESOURCE_KVD_HASH_SINGLE, + p_single_size); + if (err) + *p_single_size = MLXSW_CORE_RES_GET(mlxsw_core, KVD_SIZE) - + *p_double_size - *p_linear_size; + + /* Check results are legal. */ + if (*p_single_size < MLXSW_CORE_RES_GET(mlxsw_core, KVD_SINGLE_MIN_SIZE) || + *p_double_size < MLXSW_CORE_RES_GET(mlxsw_core, KVD_DOUBLE_MIN_SIZE) || + MLXSW_CORE_RES_GET(mlxsw_core, KVD_SIZE) < *p_linear_size) + return -EIO; + + return 0; +} + +static int +mlxsw_sp_params_acl_region_rehash_intrvl_get(struct devlink *devlink, u32 id, + struct devlink_param_gset_ctx *ctx) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + + ctx->val.vu32 = mlxsw_sp_acl_region_rehash_intrvl_get(mlxsw_sp); + return 0; +} + +static int +mlxsw_sp_params_acl_region_rehash_intrvl_set(struct devlink *devlink, u32 id, + struct devlink_param_gset_ctx *ctx) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + + return mlxsw_sp_acl_region_rehash_intrvl_set(mlxsw_sp, ctx->val.vu32); +} + +static const struct devlink_param mlxsw_sp2_devlink_params[] = { + DEVLINK_PARAM_DRIVER(MLXSW_DEVLINK_PARAM_ID_ACL_REGION_REHASH_INTERVAL, + "acl_region_rehash_interval", + DEVLINK_PARAM_TYPE_U32, + BIT(DEVLINK_PARAM_CMODE_RUNTIME), + mlxsw_sp_params_acl_region_rehash_intrvl_get, + mlxsw_sp_params_acl_region_rehash_intrvl_set, + NULL), +}; + +static int mlxsw_sp2_params_register(struct mlxsw_core *mlxsw_core) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + union devlink_param_value value; + int err; + + err = devlink_params_register(devlink, mlxsw_sp2_devlink_params, + ARRAY_SIZE(mlxsw_sp2_devlink_params)); + if (err) + return err; + + value.vu32 = 0; + devlink_param_driverinit_value_set(devlink, + MLXSW_DEVLINK_PARAM_ID_ACL_REGION_REHASH_INTERVAL, + value); + return 0; +} + +static void mlxsw_sp2_params_unregister(struct mlxsw_core *mlxsw_core) +{ + devlink_params_unregister(priv_to_devlink(mlxsw_core), + mlxsw_sp2_devlink_params, + ARRAY_SIZE(mlxsw_sp2_devlink_params)); +} + +static void mlxsw_sp_ptp_transmitted(struct mlxsw_core *mlxsw_core, + struct sk_buff *skb, u16 local_port) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + + skb_pull(skb, MLXSW_TXHDR_LEN); + mlxsw_sp->ptp_ops->transmitted(mlxsw_sp, skb, local_port); +} + +static struct mlxsw_driver mlxsw_sp1_driver = { + .kind = mlxsw_sp1_driver_name, + .priv_size = sizeof(struct mlxsw_sp), + .fw_req_rev = &mlxsw_sp1_fw_rev, + .fw_filename = MLXSW_SP1_FW_FILENAME, + .init = mlxsw_sp1_init, + .fini = mlxsw_sp_fini, + .port_split = mlxsw_sp_port_split, + .port_unsplit = mlxsw_sp_port_unsplit, + .sb_pool_get = mlxsw_sp_sb_pool_get, + .sb_pool_set = mlxsw_sp_sb_pool_set, + .sb_port_pool_get = mlxsw_sp_sb_port_pool_get, + .sb_port_pool_set = mlxsw_sp_sb_port_pool_set, + .sb_tc_pool_bind_get = mlxsw_sp_sb_tc_pool_bind_get, + .sb_tc_pool_bind_set = mlxsw_sp_sb_tc_pool_bind_set, + .sb_occ_snapshot = mlxsw_sp_sb_occ_snapshot, + .sb_occ_max_clear = mlxsw_sp_sb_occ_max_clear, + .sb_occ_port_pool_get = mlxsw_sp_sb_occ_port_pool_get, + .sb_occ_tc_port_bind_get = mlxsw_sp_sb_occ_tc_port_bind_get, + .trap_init = mlxsw_sp_trap_init, + .trap_fini = mlxsw_sp_trap_fini, + .trap_action_set = mlxsw_sp_trap_action_set, + .trap_group_init = mlxsw_sp_trap_group_init, + .trap_group_set = mlxsw_sp_trap_group_set, + .trap_policer_init = mlxsw_sp_trap_policer_init, + .trap_policer_fini = mlxsw_sp_trap_policer_fini, + .trap_policer_set = mlxsw_sp_trap_policer_set, + .trap_policer_counter_get = mlxsw_sp_trap_policer_counter_get, + .txhdr_construct = mlxsw_sp_txhdr_construct, + .resources_register = mlxsw_sp1_resources_register, + .kvd_sizes_get = mlxsw_sp_kvd_sizes_get, + .ptp_transmitted = mlxsw_sp_ptp_transmitted, + .txhdr_len = MLXSW_TXHDR_LEN, + .profile = &mlxsw_sp1_config_profile, + .sdq_supports_cqe_v2 = false, +}; + +static struct mlxsw_driver mlxsw_sp2_driver = { + .kind = mlxsw_sp2_driver_name, + .priv_size = sizeof(struct mlxsw_sp), + .fw_req_rev = &mlxsw_sp2_fw_rev, + .fw_filename = MLXSW_SP2_FW_FILENAME, + .init = mlxsw_sp2_init, + .fini = mlxsw_sp_fini, + .port_split = mlxsw_sp_port_split, + .port_unsplit = mlxsw_sp_port_unsplit, + .ports_remove_selected = mlxsw_sp_ports_remove_selected, + .sb_pool_get = mlxsw_sp_sb_pool_get, + .sb_pool_set = mlxsw_sp_sb_pool_set, + .sb_port_pool_get = mlxsw_sp_sb_port_pool_get, + .sb_port_pool_set = mlxsw_sp_sb_port_pool_set, + .sb_tc_pool_bind_get = mlxsw_sp_sb_tc_pool_bind_get, + .sb_tc_pool_bind_set = mlxsw_sp_sb_tc_pool_bind_set, + .sb_occ_snapshot = mlxsw_sp_sb_occ_snapshot, + .sb_occ_max_clear = mlxsw_sp_sb_occ_max_clear, + .sb_occ_port_pool_get = mlxsw_sp_sb_occ_port_pool_get, + .sb_occ_tc_port_bind_get = mlxsw_sp_sb_occ_tc_port_bind_get, + .trap_init = mlxsw_sp_trap_init, + .trap_fini = mlxsw_sp_trap_fini, + .trap_action_set = mlxsw_sp_trap_action_set, + .trap_group_init = mlxsw_sp_trap_group_init, + .trap_group_set = mlxsw_sp_trap_group_set, + .trap_policer_init = mlxsw_sp_trap_policer_init, + .trap_policer_fini = mlxsw_sp_trap_policer_fini, + .trap_policer_set = mlxsw_sp_trap_policer_set, + .trap_policer_counter_get = mlxsw_sp_trap_policer_counter_get, + .txhdr_construct = mlxsw_sp_txhdr_construct, + .resources_register = mlxsw_sp2_resources_register, + .params_register = mlxsw_sp2_params_register, + .params_unregister = mlxsw_sp2_params_unregister, + .ptp_transmitted = mlxsw_sp_ptp_transmitted, + .txhdr_len = MLXSW_TXHDR_LEN, + .profile = &mlxsw_sp2_config_profile, + .sdq_supports_cqe_v2 = true, +}; + +static struct mlxsw_driver mlxsw_sp3_driver = { + .kind = mlxsw_sp3_driver_name, + .priv_size = sizeof(struct mlxsw_sp), + .fw_req_rev = &mlxsw_sp3_fw_rev, + .fw_filename = MLXSW_SP3_FW_FILENAME, + .init = mlxsw_sp3_init, + .fini = mlxsw_sp_fini, + .port_split = mlxsw_sp_port_split, + .port_unsplit = mlxsw_sp_port_unsplit, + .ports_remove_selected = mlxsw_sp_ports_remove_selected, + .sb_pool_get = mlxsw_sp_sb_pool_get, + .sb_pool_set = mlxsw_sp_sb_pool_set, + .sb_port_pool_get = mlxsw_sp_sb_port_pool_get, + .sb_port_pool_set = mlxsw_sp_sb_port_pool_set, + .sb_tc_pool_bind_get = mlxsw_sp_sb_tc_pool_bind_get, + .sb_tc_pool_bind_set = mlxsw_sp_sb_tc_pool_bind_set, + .sb_occ_snapshot = mlxsw_sp_sb_occ_snapshot, + .sb_occ_max_clear = mlxsw_sp_sb_occ_max_clear, + .sb_occ_port_pool_get = mlxsw_sp_sb_occ_port_pool_get, + .sb_occ_tc_port_bind_get = mlxsw_sp_sb_occ_tc_port_bind_get, + .trap_init = mlxsw_sp_trap_init, + .trap_fini = mlxsw_sp_trap_fini, + .trap_action_set = mlxsw_sp_trap_action_set, + .trap_group_init = mlxsw_sp_trap_group_init, + .trap_group_set = mlxsw_sp_trap_group_set, + .trap_policer_init = mlxsw_sp_trap_policer_init, + .trap_policer_fini = mlxsw_sp_trap_policer_fini, + .trap_policer_set = mlxsw_sp_trap_policer_set, + .trap_policer_counter_get = mlxsw_sp_trap_policer_counter_get, + .txhdr_construct = mlxsw_sp_txhdr_construct, + .resources_register = mlxsw_sp2_resources_register, + .params_register = mlxsw_sp2_params_register, + .params_unregister = mlxsw_sp2_params_unregister, + .ptp_transmitted = mlxsw_sp_ptp_transmitted, + .txhdr_len = MLXSW_TXHDR_LEN, + .profile = &mlxsw_sp2_config_profile, + .sdq_supports_cqe_v2 = true, +}; + +static struct mlxsw_driver mlxsw_sp4_driver = { + .kind = mlxsw_sp4_driver_name, + .priv_size = sizeof(struct mlxsw_sp), + .init = mlxsw_sp4_init, + .fini = mlxsw_sp_fini, + .port_split = mlxsw_sp_port_split, + .port_unsplit = mlxsw_sp_port_unsplit, + .ports_remove_selected = mlxsw_sp_ports_remove_selected, + .sb_pool_get = mlxsw_sp_sb_pool_get, + .sb_pool_set = mlxsw_sp_sb_pool_set, + .sb_port_pool_get = mlxsw_sp_sb_port_pool_get, + .sb_port_pool_set = mlxsw_sp_sb_port_pool_set, + .sb_tc_pool_bind_get = mlxsw_sp_sb_tc_pool_bind_get, + .sb_tc_pool_bind_set = mlxsw_sp_sb_tc_pool_bind_set, + .sb_occ_snapshot = mlxsw_sp_sb_occ_snapshot, + .sb_occ_max_clear = mlxsw_sp_sb_occ_max_clear, + .sb_occ_port_pool_get = mlxsw_sp_sb_occ_port_pool_get, + .sb_occ_tc_port_bind_get = mlxsw_sp_sb_occ_tc_port_bind_get, + .trap_init = mlxsw_sp_trap_init, + .trap_fini = mlxsw_sp_trap_fini, + .trap_action_set = mlxsw_sp_trap_action_set, + .trap_group_init = mlxsw_sp_trap_group_init, + .trap_group_set = mlxsw_sp_trap_group_set, + .trap_policer_init = mlxsw_sp_trap_policer_init, + .trap_policer_fini = mlxsw_sp_trap_policer_fini, + .trap_policer_set = mlxsw_sp_trap_policer_set, + .trap_policer_counter_get = mlxsw_sp_trap_policer_counter_get, + .txhdr_construct = mlxsw_sp_txhdr_construct, + .resources_register = mlxsw_sp2_resources_register, + .params_register = mlxsw_sp2_params_register, + .params_unregister = mlxsw_sp2_params_unregister, + .ptp_transmitted = mlxsw_sp_ptp_transmitted, + .txhdr_len = MLXSW_TXHDR_LEN, + .profile = &mlxsw_sp4_config_profile, + .sdq_supports_cqe_v2 = true, +}; + +bool mlxsw_sp_port_dev_check(const struct net_device *dev) +{ + return dev->netdev_ops == &mlxsw_sp_port_netdev_ops; +} + +static int mlxsw_sp_lower_dev_walk(struct net_device *lower_dev, + struct netdev_nested_priv *priv) +{ + int ret = 0; + + if (mlxsw_sp_port_dev_check(lower_dev)) { + priv->data = (void *)netdev_priv(lower_dev); + ret = 1; + } + + return ret; +} + +struct mlxsw_sp_port *mlxsw_sp_port_dev_lower_find(struct net_device *dev) +{ + struct netdev_nested_priv priv = { + .data = NULL, + }; + + if (mlxsw_sp_port_dev_check(dev)) + return netdev_priv(dev); + + netdev_walk_all_lower_dev(dev, mlxsw_sp_lower_dev_walk, &priv); + + return (struct mlxsw_sp_port *)priv.data; +} + +struct mlxsw_sp *mlxsw_sp_lower_get(struct net_device *dev) +{ + struct mlxsw_sp_port *mlxsw_sp_port; + + mlxsw_sp_port = mlxsw_sp_port_dev_lower_find(dev); + return mlxsw_sp_port ? mlxsw_sp_port->mlxsw_sp : NULL; +} + +struct mlxsw_sp_port *mlxsw_sp_port_dev_lower_find_rcu(struct net_device *dev) +{ + struct netdev_nested_priv priv = { + .data = NULL, + }; + + if (mlxsw_sp_port_dev_check(dev)) + return netdev_priv(dev); + + netdev_walk_all_lower_dev_rcu(dev, mlxsw_sp_lower_dev_walk, + &priv); + + return (struct mlxsw_sp_port *)priv.data; +} + +struct mlxsw_sp_port *mlxsw_sp_port_lower_dev_hold(struct net_device *dev) +{ + struct mlxsw_sp_port *mlxsw_sp_port; + + rcu_read_lock(); + mlxsw_sp_port = mlxsw_sp_port_dev_lower_find_rcu(dev); + if (mlxsw_sp_port) + dev_hold(mlxsw_sp_port->dev); + rcu_read_unlock(); + return mlxsw_sp_port; +} + +void mlxsw_sp_port_dev_put(struct mlxsw_sp_port *mlxsw_sp_port) +{ + dev_put(mlxsw_sp_port->dev); +} + +int mlxsw_sp_parsing_depth_inc(struct mlxsw_sp *mlxsw_sp) +{ + char mprs_pl[MLXSW_REG_MPRS_LEN]; + int err = 0; + + mutex_lock(&mlxsw_sp->parsing.lock); + + if (refcount_inc_not_zero(&mlxsw_sp->parsing.parsing_depth_ref)) + goto out_unlock; + + mlxsw_reg_mprs_pack(mprs_pl, MLXSW_SP_INCREASED_PARSING_DEPTH, + mlxsw_sp->parsing.vxlan_udp_dport); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mprs), mprs_pl); + if (err) + goto out_unlock; + + mlxsw_sp->parsing.parsing_depth = MLXSW_SP_INCREASED_PARSING_DEPTH; + refcount_set(&mlxsw_sp->parsing.parsing_depth_ref, 1); + +out_unlock: + mutex_unlock(&mlxsw_sp->parsing.lock); + return err; +} + +void mlxsw_sp_parsing_depth_dec(struct mlxsw_sp *mlxsw_sp) +{ + char mprs_pl[MLXSW_REG_MPRS_LEN]; + + mutex_lock(&mlxsw_sp->parsing.lock); + + if (!refcount_dec_and_test(&mlxsw_sp->parsing.parsing_depth_ref)) + goto out_unlock; + + mlxsw_reg_mprs_pack(mprs_pl, MLXSW_SP_DEFAULT_PARSING_DEPTH, + mlxsw_sp->parsing.vxlan_udp_dport); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mprs), mprs_pl); + mlxsw_sp->parsing.parsing_depth = MLXSW_SP_DEFAULT_PARSING_DEPTH; + +out_unlock: + mutex_unlock(&mlxsw_sp->parsing.lock); +} + +int mlxsw_sp_parsing_vxlan_udp_dport_set(struct mlxsw_sp *mlxsw_sp, + __be16 udp_dport) +{ + char mprs_pl[MLXSW_REG_MPRS_LEN]; + int err; + + mutex_lock(&mlxsw_sp->parsing.lock); + + mlxsw_reg_mprs_pack(mprs_pl, mlxsw_sp->parsing.parsing_depth, + be16_to_cpu(udp_dport)); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mprs), mprs_pl); + if (err) + goto out_unlock; + + mlxsw_sp->parsing.vxlan_udp_dport = be16_to_cpu(udp_dport); + +out_unlock: + mutex_unlock(&mlxsw_sp->parsing.lock); + return err; +} + +static void +mlxsw_sp_port_lag_uppers_cleanup(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *lag_dev) +{ + struct net_device *br_dev = netdev_master_upper_dev_get(lag_dev); + struct net_device *upper_dev; + struct list_head *iter; + + if (netif_is_bridge_port(lag_dev)) + mlxsw_sp_port_bridge_leave(mlxsw_sp_port, lag_dev, br_dev); + + netdev_for_each_upper_dev_rcu(lag_dev, upper_dev, iter) { + if (!netif_is_bridge_port(upper_dev)) + continue; + br_dev = netdev_master_upper_dev_get(upper_dev); + mlxsw_sp_port_bridge_leave(mlxsw_sp_port, upper_dev, br_dev); + } +} + +static int mlxsw_sp_lag_create(struct mlxsw_sp *mlxsw_sp, u16 lag_id) +{ + char sldr_pl[MLXSW_REG_SLDR_LEN]; + + mlxsw_reg_sldr_lag_create_pack(sldr_pl, lag_id); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sldr), sldr_pl); +} + +static int mlxsw_sp_lag_destroy(struct mlxsw_sp *mlxsw_sp, u16 lag_id) +{ + char sldr_pl[MLXSW_REG_SLDR_LEN]; + + mlxsw_reg_sldr_lag_destroy_pack(sldr_pl, lag_id); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sldr), sldr_pl); +} + +static int mlxsw_sp_lag_col_port_add(struct mlxsw_sp_port *mlxsw_sp_port, + u16 lag_id, u8 port_index) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char slcor_pl[MLXSW_REG_SLCOR_LEN]; + + mlxsw_reg_slcor_port_add_pack(slcor_pl, mlxsw_sp_port->local_port, + lag_id, port_index); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(slcor), slcor_pl); +} + +static int mlxsw_sp_lag_col_port_remove(struct mlxsw_sp_port *mlxsw_sp_port, + u16 lag_id) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char slcor_pl[MLXSW_REG_SLCOR_LEN]; + + mlxsw_reg_slcor_port_remove_pack(slcor_pl, mlxsw_sp_port->local_port, + lag_id); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(slcor), slcor_pl); +} + +static int mlxsw_sp_lag_col_port_enable(struct mlxsw_sp_port *mlxsw_sp_port, + u16 lag_id) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char slcor_pl[MLXSW_REG_SLCOR_LEN]; + + mlxsw_reg_slcor_col_enable_pack(slcor_pl, mlxsw_sp_port->local_port, + lag_id); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(slcor), slcor_pl); +} + +static int mlxsw_sp_lag_col_port_disable(struct mlxsw_sp_port *mlxsw_sp_port, + u16 lag_id) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char slcor_pl[MLXSW_REG_SLCOR_LEN]; + + mlxsw_reg_slcor_col_disable_pack(slcor_pl, mlxsw_sp_port->local_port, + lag_id); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(slcor), slcor_pl); +} + +static int mlxsw_sp_lag_index_get(struct mlxsw_sp *mlxsw_sp, + struct net_device *lag_dev, + u16 *p_lag_id) +{ + struct mlxsw_sp_upper *lag; + int free_lag_id = -1; + u16 max_lag; + int err, i; + + err = mlxsw_core_max_lag(mlxsw_sp->core, &max_lag); + if (err) + return err; + + for (i = 0; i < max_lag; i++) { + lag = mlxsw_sp_lag_get(mlxsw_sp, i); + if (lag->ref_count) { + if (lag->dev == lag_dev) { + *p_lag_id = i; + return 0; + } + } else if (free_lag_id < 0) { + free_lag_id = i; + } + } + if (free_lag_id < 0) + return -EBUSY; + *p_lag_id = free_lag_id; + return 0; +} + +static bool +mlxsw_sp_master_lag_check(struct mlxsw_sp *mlxsw_sp, + struct net_device *lag_dev, + struct netdev_lag_upper_info *lag_upper_info, + struct netlink_ext_ack *extack) +{ + u16 lag_id; + + if (mlxsw_sp_lag_index_get(mlxsw_sp, lag_dev, &lag_id) != 0) { + NL_SET_ERR_MSG_MOD(extack, "Exceeded number of supported LAG devices"); + return false; + } + if (lag_upper_info->tx_type != NETDEV_LAG_TX_TYPE_HASH) { + NL_SET_ERR_MSG_MOD(extack, "LAG device using unsupported Tx type"); + return false; + } + return true; +} + +static int mlxsw_sp_port_lag_index_get(struct mlxsw_sp *mlxsw_sp, + u16 lag_id, u8 *p_port_index) +{ + u64 max_lag_members; + int i; + + max_lag_members = MLXSW_CORE_RES_GET(mlxsw_sp->core, + MAX_LAG_MEMBERS); + for (i = 0; i < max_lag_members; i++) { + if (!mlxsw_sp_port_lagged_get(mlxsw_sp, lag_id, i)) { + *p_port_index = i; + return 0; + } + } + return -EBUSY; +} + +static int mlxsw_sp_port_lag_join(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *lag_dev, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_upper *lag; + u16 lag_id; + u8 port_index; + int err; + + err = mlxsw_sp_lag_index_get(mlxsw_sp, lag_dev, &lag_id); + if (err) + return err; + lag = mlxsw_sp_lag_get(mlxsw_sp, lag_id); + if (!lag->ref_count) { + err = mlxsw_sp_lag_create(mlxsw_sp, lag_id); + if (err) + return err; + lag->dev = lag_dev; + } + + err = mlxsw_sp_port_lag_index_get(mlxsw_sp, lag_id, &port_index); + if (err) + return err; + err = mlxsw_sp_lag_col_port_add(mlxsw_sp_port, lag_id, port_index); + if (err) + goto err_col_port_add; + + mlxsw_core_lag_mapping_set(mlxsw_sp->core, lag_id, port_index, + mlxsw_sp_port->local_port); + mlxsw_sp_port->lag_id = lag_id; + mlxsw_sp_port->lagged = 1; + lag->ref_count++; + + /* Port is no longer usable as a router interface */ + if (mlxsw_sp_port->default_vlan->fid) + mlxsw_sp_port_vlan_router_leave(mlxsw_sp_port->default_vlan); + + /* Join a router interface configured on the LAG, if exists */ + err = mlxsw_sp_port_vlan_router_join(mlxsw_sp_port->default_vlan, + lag_dev, extack); + if (err) + goto err_router_join; + + return 0; + +err_router_join: + lag->ref_count--; + mlxsw_sp_port->lagged = 0; + mlxsw_core_lag_mapping_clear(mlxsw_sp->core, lag_id, + mlxsw_sp_port->local_port); + mlxsw_sp_lag_col_port_remove(mlxsw_sp_port, lag_id); +err_col_port_add: + if (!lag->ref_count) + mlxsw_sp_lag_destroy(mlxsw_sp, lag_id); + return err; +} + +static void mlxsw_sp_port_lag_leave(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *lag_dev) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 lag_id = mlxsw_sp_port->lag_id; + struct mlxsw_sp_upper *lag; + + if (!mlxsw_sp_port->lagged) + return; + lag = mlxsw_sp_lag_get(mlxsw_sp, lag_id); + WARN_ON(lag->ref_count == 0); + + mlxsw_sp_lag_col_port_remove(mlxsw_sp_port, lag_id); + + /* Any VLANs configured on the port are no longer valid */ + mlxsw_sp_port_vlan_flush(mlxsw_sp_port, false); + mlxsw_sp_port_vlan_cleanup(mlxsw_sp_port->default_vlan); + /* Make the LAG and its directly linked uppers leave bridges they + * are memeber in + */ + mlxsw_sp_port_lag_uppers_cleanup(mlxsw_sp_port, lag_dev); + + if (lag->ref_count == 1) + mlxsw_sp_lag_destroy(mlxsw_sp, lag_id); + + mlxsw_core_lag_mapping_clear(mlxsw_sp->core, lag_id, + mlxsw_sp_port->local_port); + mlxsw_sp_port->lagged = 0; + lag->ref_count--; + + /* Make sure untagged frames are allowed to ingress */ + mlxsw_sp_port_pvid_set(mlxsw_sp_port, MLXSW_SP_DEFAULT_VID, + ETH_P_8021Q); +} + +static int mlxsw_sp_lag_dist_port_add(struct mlxsw_sp_port *mlxsw_sp_port, + u16 lag_id) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char sldr_pl[MLXSW_REG_SLDR_LEN]; + + mlxsw_reg_sldr_lag_add_port_pack(sldr_pl, lag_id, + mlxsw_sp_port->local_port); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sldr), sldr_pl); +} + +static int mlxsw_sp_lag_dist_port_remove(struct mlxsw_sp_port *mlxsw_sp_port, + u16 lag_id) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char sldr_pl[MLXSW_REG_SLDR_LEN]; + + mlxsw_reg_sldr_lag_remove_port_pack(sldr_pl, lag_id, + mlxsw_sp_port->local_port); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sldr), sldr_pl); +} + +static int +mlxsw_sp_port_lag_col_dist_enable(struct mlxsw_sp_port *mlxsw_sp_port) +{ + int err; + + err = mlxsw_sp_lag_col_port_enable(mlxsw_sp_port, + mlxsw_sp_port->lag_id); + if (err) + return err; + + err = mlxsw_sp_lag_dist_port_add(mlxsw_sp_port, mlxsw_sp_port->lag_id); + if (err) + goto err_dist_port_add; + + return 0; + +err_dist_port_add: + mlxsw_sp_lag_col_port_disable(mlxsw_sp_port, mlxsw_sp_port->lag_id); + return err; +} + +static int +mlxsw_sp_port_lag_col_dist_disable(struct mlxsw_sp_port *mlxsw_sp_port) +{ + int err; + + err = mlxsw_sp_lag_dist_port_remove(mlxsw_sp_port, + mlxsw_sp_port->lag_id); + if (err) + return err; + + err = mlxsw_sp_lag_col_port_disable(mlxsw_sp_port, + mlxsw_sp_port->lag_id); + if (err) + goto err_col_port_disable; + + return 0; + +err_col_port_disable: + mlxsw_sp_lag_dist_port_add(mlxsw_sp_port, mlxsw_sp_port->lag_id); + return err; +} + +static int mlxsw_sp_port_lag_changed(struct mlxsw_sp_port *mlxsw_sp_port, + struct netdev_lag_lower_state_info *info) +{ + if (info->tx_enabled) + return mlxsw_sp_port_lag_col_dist_enable(mlxsw_sp_port); + else + return mlxsw_sp_port_lag_col_dist_disable(mlxsw_sp_port); +} + +static int mlxsw_sp_port_stp_set(struct mlxsw_sp_port *mlxsw_sp_port, + bool enable) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + enum mlxsw_reg_spms_state spms_state; + char *spms_pl; + u16 vid; + int err; + + spms_state = enable ? MLXSW_REG_SPMS_STATE_FORWARDING : + MLXSW_REG_SPMS_STATE_DISCARDING; + + spms_pl = kmalloc(MLXSW_REG_SPMS_LEN, GFP_KERNEL); + if (!spms_pl) + return -ENOMEM; + mlxsw_reg_spms_pack(spms_pl, mlxsw_sp_port->local_port); + + for (vid = 0; vid < VLAN_N_VID; vid++) + mlxsw_reg_spms_vid_pack(spms_pl, vid, spms_state); + + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(spms), spms_pl); + kfree(spms_pl); + return err; +} + +static int mlxsw_sp_port_ovs_join(struct mlxsw_sp_port *mlxsw_sp_port) +{ + u16 vid = 1; + int err; + + err = mlxsw_sp_port_vp_mode_set(mlxsw_sp_port, true); + if (err) + return err; + err = mlxsw_sp_port_stp_set(mlxsw_sp_port, true); + if (err) + goto err_port_stp_set; + err = mlxsw_sp_port_vlan_set(mlxsw_sp_port, 1, VLAN_N_VID - 2, + true, false); + if (err) + goto err_port_vlan_set; + + for (; vid <= VLAN_N_VID - 1; vid++) { + err = mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, + vid, false); + if (err) + goto err_vid_learning_set; + } + + return 0; + +err_vid_learning_set: + for (vid--; vid >= 1; vid--) + mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, vid, true); +err_port_vlan_set: + mlxsw_sp_port_stp_set(mlxsw_sp_port, false); +err_port_stp_set: + mlxsw_sp_port_vp_mode_set(mlxsw_sp_port, false); + return err; +} + +static void mlxsw_sp_port_ovs_leave(struct mlxsw_sp_port *mlxsw_sp_port) +{ + u16 vid; + + for (vid = VLAN_N_VID - 1; vid >= 1; vid--) + mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, + vid, true); + + mlxsw_sp_port_vlan_set(mlxsw_sp_port, 1, VLAN_N_VID - 2, + false, false); + mlxsw_sp_port_stp_set(mlxsw_sp_port, false); + mlxsw_sp_port_vp_mode_set(mlxsw_sp_port, false); +} + +static bool mlxsw_sp_bridge_has_multiple_vxlans(struct net_device *br_dev) +{ + unsigned int num_vxlans = 0; + struct net_device *dev; + struct list_head *iter; + + netdev_for_each_lower_dev(br_dev, dev, iter) { + if (netif_is_vxlan(dev)) + num_vxlans++; + } + + return num_vxlans > 1; +} + +static bool mlxsw_sp_bridge_vxlan_vlan_is_valid(struct net_device *br_dev) +{ + DECLARE_BITMAP(vlans, VLAN_N_VID) = {0}; + struct net_device *dev; + struct list_head *iter; + + netdev_for_each_lower_dev(br_dev, dev, iter) { + u16 pvid; + int err; + + if (!netif_is_vxlan(dev)) + continue; + + err = mlxsw_sp_vxlan_mapped_vid(dev, &pvid); + if (err || !pvid) + continue; + + if (test_and_set_bit(pvid, vlans)) + return false; + } + + return true; +} + +static bool mlxsw_sp_bridge_vxlan_is_valid(struct net_device *br_dev, + struct netlink_ext_ack *extack) +{ + if (br_multicast_enabled(br_dev)) { + NL_SET_ERR_MSG_MOD(extack, "Multicast can not be enabled on a bridge with a VxLAN device"); + return false; + } + + if (!br_vlan_enabled(br_dev) && + mlxsw_sp_bridge_has_multiple_vxlans(br_dev)) { + NL_SET_ERR_MSG_MOD(extack, "Multiple VxLAN devices are not supported in a VLAN-unaware bridge"); + return false; + } + + if (br_vlan_enabled(br_dev) && + !mlxsw_sp_bridge_vxlan_vlan_is_valid(br_dev)) { + NL_SET_ERR_MSG_MOD(extack, "Multiple VxLAN devices cannot have the same VLAN as PVID and egress untagged"); + return false; + } + + return true; +} + +static int mlxsw_sp_netdevice_port_upper_event(struct net_device *lower_dev, + struct net_device *dev, + unsigned long event, void *ptr) +{ + struct netdev_notifier_changeupper_info *info; + struct mlxsw_sp_port *mlxsw_sp_port; + struct netlink_ext_ack *extack; + struct net_device *upper_dev; + struct mlxsw_sp *mlxsw_sp; + int err = 0; + u16 proto; + + mlxsw_sp_port = netdev_priv(dev); + mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + info = ptr; + extack = netdev_notifier_info_to_extack(&info->info); + + switch (event) { + case NETDEV_PRECHANGEUPPER: + upper_dev = info->upper_dev; + if (!is_vlan_dev(upper_dev) && + !netif_is_lag_master(upper_dev) && + !netif_is_bridge_master(upper_dev) && + !netif_is_ovs_master(upper_dev) && + !netif_is_macvlan(upper_dev) && + !netif_is_l3_master(upper_dev)) { + NL_SET_ERR_MSG_MOD(extack, "Unknown upper device type"); + return -EINVAL; + } + if (!info->linking) + break; + if (netif_is_bridge_master(upper_dev) && + !mlxsw_sp_bridge_device_is_offloaded(mlxsw_sp, upper_dev) && + mlxsw_sp_bridge_has_vxlan(upper_dev) && + !mlxsw_sp_bridge_vxlan_is_valid(upper_dev, extack)) + return -EOPNOTSUPP; + if (netdev_has_any_upper_dev(upper_dev) && + (!netif_is_bridge_master(upper_dev) || + !mlxsw_sp_bridge_device_is_offloaded(mlxsw_sp, + upper_dev))) { + NL_SET_ERR_MSG_MOD(extack, "Enslaving a port to a device that already has an upper device is not supported"); + return -EINVAL; + } + if (netif_is_lag_master(upper_dev) && + !mlxsw_sp_master_lag_check(mlxsw_sp, upper_dev, + info->upper_info, extack)) + return -EINVAL; + if (netif_is_lag_master(upper_dev) && vlan_uses_dev(dev)) { + NL_SET_ERR_MSG_MOD(extack, "Master device is a LAG master and this device has a VLAN"); + return -EINVAL; + } + if (netif_is_lag_port(dev) && is_vlan_dev(upper_dev) && + !netif_is_lag_master(vlan_dev_real_dev(upper_dev))) { + NL_SET_ERR_MSG_MOD(extack, "Can not put a VLAN on a LAG port"); + return -EINVAL; + } + if (netif_is_macvlan(upper_dev) && + !mlxsw_sp_rif_exists(mlxsw_sp, lower_dev)) { + NL_SET_ERR_MSG_MOD(extack, "macvlan is only supported on top of router interfaces"); + return -EOPNOTSUPP; + } + if (netif_is_ovs_master(upper_dev) && vlan_uses_dev(dev)) { + NL_SET_ERR_MSG_MOD(extack, "Master device is an OVS master and this device has a VLAN"); + return -EINVAL; + } + if (netif_is_ovs_port(dev) && is_vlan_dev(upper_dev)) { + NL_SET_ERR_MSG_MOD(extack, "Can not put a VLAN on an OVS port"); + return -EINVAL; + } + if (netif_is_bridge_master(upper_dev)) { + br_vlan_get_proto(upper_dev, &proto); + if (br_vlan_enabled(upper_dev) && + proto != ETH_P_8021Q && proto != ETH_P_8021AD) { + NL_SET_ERR_MSG_MOD(extack, "Enslaving a port to a bridge with unknown VLAN protocol is not supported"); + return -EOPNOTSUPP; + } + if (vlan_uses_dev(lower_dev) && + br_vlan_enabled(upper_dev) && + proto == ETH_P_8021AD) { + NL_SET_ERR_MSG_MOD(extack, "Enslaving a port that already has a VLAN upper to an 802.1ad bridge is not supported"); + return -EOPNOTSUPP; + } + } + if (netif_is_bridge_port(lower_dev) && is_vlan_dev(upper_dev)) { + struct net_device *br_dev = netdev_master_upper_dev_get(lower_dev); + + if (br_vlan_enabled(br_dev)) { + br_vlan_get_proto(br_dev, &proto); + if (proto == ETH_P_8021AD) { + NL_SET_ERR_MSG_MOD(extack, "VLAN uppers are not supported on a port enslaved to an 802.1ad bridge"); + return -EOPNOTSUPP; + } + } + } + if (is_vlan_dev(upper_dev) && + ntohs(vlan_dev_vlan_proto(upper_dev)) != ETH_P_8021Q) { + NL_SET_ERR_MSG_MOD(extack, "VLAN uppers are only supported with 802.1q VLAN protocol"); + return -EOPNOTSUPP; + } + break; + case NETDEV_CHANGEUPPER: + upper_dev = info->upper_dev; + if (netif_is_bridge_master(upper_dev)) { + if (info->linking) + err = mlxsw_sp_port_bridge_join(mlxsw_sp_port, + lower_dev, + upper_dev, + extack); + else + mlxsw_sp_port_bridge_leave(mlxsw_sp_port, + lower_dev, + upper_dev); + } else if (netif_is_lag_master(upper_dev)) { + if (info->linking) { + err = mlxsw_sp_port_lag_join(mlxsw_sp_port, + upper_dev, extack); + } else { + mlxsw_sp_port_lag_col_dist_disable(mlxsw_sp_port); + mlxsw_sp_port_lag_leave(mlxsw_sp_port, + upper_dev); + } + } else if (netif_is_ovs_master(upper_dev)) { + if (info->linking) + err = mlxsw_sp_port_ovs_join(mlxsw_sp_port); + else + mlxsw_sp_port_ovs_leave(mlxsw_sp_port); + } else if (netif_is_macvlan(upper_dev)) { + if (!info->linking) + mlxsw_sp_rif_macvlan_del(mlxsw_sp, upper_dev); + } else if (is_vlan_dev(upper_dev)) { + struct net_device *br_dev; + + if (!netif_is_bridge_port(upper_dev)) + break; + if (info->linking) + break; + br_dev = netdev_master_upper_dev_get(upper_dev); + mlxsw_sp_port_bridge_leave(mlxsw_sp_port, upper_dev, + br_dev); + } + break; + } + + return err; +} + +static int mlxsw_sp_netdevice_port_lower_event(struct net_device *dev, + unsigned long event, void *ptr) +{ + struct netdev_notifier_changelowerstate_info *info; + struct mlxsw_sp_port *mlxsw_sp_port; + int err; + + mlxsw_sp_port = netdev_priv(dev); + info = ptr; + + switch (event) { + case NETDEV_CHANGELOWERSTATE: + if (netif_is_lag_port(dev) && mlxsw_sp_port->lagged) { + err = mlxsw_sp_port_lag_changed(mlxsw_sp_port, + info->lower_state_info); + if (err) + netdev_err(dev, "Failed to reflect link aggregation lower state change\n"); + } + break; + } + + return 0; +} + +static int mlxsw_sp_netdevice_port_event(struct net_device *lower_dev, + struct net_device *port_dev, + unsigned long event, void *ptr) +{ + switch (event) { + case NETDEV_PRECHANGEUPPER: + case NETDEV_CHANGEUPPER: + return mlxsw_sp_netdevice_port_upper_event(lower_dev, port_dev, + event, ptr); + case NETDEV_CHANGELOWERSTATE: + return mlxsw_sp_netdevice_port_lower_event(port_dev, event, + ptr); + } + + return 0; +} + +static int mlxsw_sp_netdevice_lag_event(struct net_device *lag_dev, + unsigned long event, void *ptr) +{ + struct net_device *dev; + struct list_head *iter; + int ret; + + netdev_for_each_lower_dev(lag_dev, dev, iter) { + if (mlxsw_sp_port_dev_check(dev)) { + ret = mlxsw_sp_netdevice_port_event(lag_dev, dev, event, + ptr); + if (ret) + return ret; + } + } + + return 0; +} + +static int mlxsw_sp_netdevice_port_vlan_event(struct net_device *vlan_dev, + struct net_device *dev, + unsigned long event, void *ptr, + u16 vid) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct netdev_notifier_changeupper_info *info = ptr; + struct netlink_ext_ack *extack; + struct net_device *upper_dev; + int err = 0; + + extack = netdev_notifier_info_to_extack(&info->info); + + switch (event) { + case NETDEV_PRECHANGEUPPER: + upper_dev = info->upper_dev; + if (!netif_is_bridge_master(upper_dev) && + !netif_is_macvlan(upper_dev) && + !netif_is_l3_master(upper_dev)) { + NL_SET_ERR_MSG_MOD(extack, "Unknown upper device type"); + return -EINVAL; + } + if (!info->linking) + break; + if (netif_is_bridge_master(upper_dev) && + !mlxsw_sp_bridge_device_is_offloaded(mlxsw_sp, upper_dev) && + mlxsw_sp_bridge_has_vxlan(upper_dev) && + !mlxsw_sp_bridge_vxlan_is_valid(upper_dev, extack)) + return -EOPNOTSUPP; + if (netdev_has_any_upper_dev(upper_dev) && + (!netif_is_bridge_master(upper_dev) || + !mlxsw_sp_bridge_device_is_offloaded(mlxsw_sp, + upper_dev))) { + NL_SET_ERR_MSG_MOD(extack, "Enslaving a port to a device that already has an upper device is not supported"); + return -EINVAL; + } + if (netif_is_macvlan(upper_dev) && + !mlxsw_sp_rif_exists(mlxsw_sp, vlan_dev)) { + NL_SET_ERR_MSG_MOD(extack, "macvlan is only supported on top of router interfaces"); + return -EOPNOTSUPP; + } + break; + case NETDEV_CHANGEUPPER: + upper_dev = info->upper_dev; + if (netif_is_bridge_master(upper_dev)) { + if (info->linking) + err = mlxsw_sp_port_bridge_join(mlxsw_sp_port, + vlan_dev, + upper_dev, + extack); + else + mlxsw_sp_port_bridge_leave(mlxsw_sp_port, + vlan_dev, + upper_dev); + } else if (netif_is_macvlan(upper_dev)) { + if (!info->linking) + mlxsw_sp_rif_macvlan_del(mlxsw_sp, upper_dev); + } + break; + } + + return err; +} + +static int mlxsw_sp_netdevice_lag_port_vlan_event(struct net_device *vlan_dev, + struct net_device *lag_dev, + unsigned long event, + void *ptr, u16 vid) +{ + struct net_device *dev; + struct list_head *iter; + int ret; + + netdev_for_each_lower_dev(lag_dev, dev, iter) { + if (mlxsw_sp_port_dev_check(dev)) { + ret = mlxsw_sp_netdevice_port_vlan_event(vlan_dev, dev, + event, ptr, + vid); + if (ret) + return ret; + } + } + + return 0; +} + +static int mlxsw_sp_netdevice_bridge_vlan_event(struct net_device *vlan_dev, + struct net_device *br_dev, + unsigned long event, void *ptr, + u16 vid) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_lower_get(vlan_dev); + struct netdev_notifier_changeupper_info *info = ptr; + struct netlink_ext_ack *extack; + struct net_device *upper_dev; + + if (!mlxsw_sp) + return 0; + + extack = netdev_notifier_info_to_extack(&info->info); + + switch (event) { + case NETDEV_PRECHANGEUPPER: + upper_dev = info->upper_dev; + if (!netif_is_macvlan(upper_dev) && + !netif_is_l3_master(upper_dev)) { + NL_SET_ERR_MSG_MOD(extack, "Unknown upper device type"); + return -EOPNOTSUPP; + } + if (!info->linking) + break; + if (netif_is_macvlan(upper_dev) && + !mlxsw_sp_rif_exists(mlxsw_sp, vlan_dev)) { + NL_SET_ERR_MSG_MOD(extack, "macvlan is only supported on top of router interfaces"); + return -EOPNOTSUPP; + } + break; + case NETDEV_CHANGEUPPER: + upper_dev = info->upper_dev; + if (info->linking) + break; + if (netif_is_macvlan(upper_dev)) + mlxsw_sp_rif_macvlan_del(mlxsw_sp, upper_dev); + break; + } + + return 0; +} + +static int mlxsw_sp_netdevice_vlan_event(struct net_device *vlan_dev, + unsigned long event, void *ptr) +{ + struct net_device *real_dev = vlan_dev_real_dev(vlan_dev); + u16 vid = vlan_dev_vlan_id(vlan_dev); + + if (mlxsw_sp_port_dev_check(real_dev)) + return mlxsw_sp_netdevice_port_vlan_event(vlan_dev, real_dev, + event, ptr, vid); + else if (netif_is_lag_master(real_dev)) + return mlxsw_sp_netdevice_lag_port_vlan_event(vlan_dev, + real_dev, event, + ptr, vid); + else if (netif_is_bridge_master(real_dev)) + return mlxsw_sp_netdevice_bridge_vlan_event(vlan_dev, real_dev, + event, ptr, vid); + + return 0; +} + +static int mlxsw_sp_netdevice_bridge_event(struct net_device *br_dev, + unsigned long event, void *ptr) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_lower_get(br_dev); + struct netdev_notifier_changeupper_info *info = ptr; + struct netlink_ext_ack *extack; + struct net_device *upper_dev; + u16 proto; + + if (!mlxsw_sp) + return 0; + + extack = netdev_notifier_info_to_extack(&info->info); + + switch (event) { + case NETDEV_PRECHANGEUPPER: + upper_dev = info->upper_dev; + if (!is_vlan_dev(upper_dev) && + !netif_is_macvlan(upper_dev) && + !netif_is_l3_master(upper_dev)) { + NL_SET_ERR_MSG_MOD(extack, "Unknown upper device type"); + return -EOPNOTSUPP; + } + if (!info->linking) + break; + if (br_vlan_enabled(br_dev)) { + br_vlan_get_proto(br_dev, &proto); + if (proto == ETH_P_8021AD) { + NL_SET_ERR_MSG_MOD(extack, "Upper devices are not supported on top of an 802.1ad bridge"); + return -EOPNOTSUPP; + } + } + if (is_vlan_dev(upper_dev) && + ntohs(vlan_dev_vlan_proto(upper_dev)) != ETH_P_8021Q) { + NL_SET_ERR_MSG_MOD(extack, "VLAN uppers are only supported with 802.1q VLAN protocol"); + return -EOPNOTSUPP; + } + if (netif_is_macvlan(upper_dev) && + !mlxsw_sp_rif_exists(mlxsw_sp, br_dev)) { + NL_SET_ERR_MSG_MOD(extack, "macvlan is only supported on top of router interfaces"); + return -EOPNOTSUPP; + } + break; + case NETDEV_CHANGEUPPER: + upper_dev = info->upper_dev; + if (info->linking) + break; + if (is_vlan_dev(upper_dev)) + mlxsw_sp_rif_destroy_by_dev(mlxsw_sp, upper_dev); + if (netif_is_macvlan(upper_dev)) + mlxsw_sp_rif_macvlan_del(mlxsw_sp, upper_dev); + break; + } + + return 0; +} + +static int mlxsw_sp_netdevice_macvlan_event(struct net_device *macvlan_dev, + unsigned long event, void *ptr) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_lower_get(macvlan_dev); + struct netdev_notifier_changeupper_info *info = ptr; + struct netlink_ext_ack *extack; + struct net_device *upper_dev; + + if (!mlxsw_sp || event != NETDEV_PRECHANGEUPPER) + return 0; + + extack = netdev_notifier_info_to_extack(&info->info); + upper_dev = info->upper_dev; + + if (!netif_is_l3_master(upper_dev)) { + NL_SET_ERR_MSG_MOD(extack, "Unknown upper device type"); + return -EOPNOTSUPP; + } + + return 0; +} + +static int mlxsw_sp_netdevice_vxlan_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *dev, + unsigned long event, void *ptr) +{ + struct netdev_notifier_changeupper_info *cu_info; + struct netdev_notifier_info *info = ptr; + struct netlink_ext_ack *extack; + struct net_device *upper_dev; + + extack = netdev_notifier_info_to_extack(info); + + switch (event) { + case NETDEV_CHANGEUPPER: + cu_info = container_of(info, + struct netdev_notifier_changeupper_info, + info); + upper_dev = cu_info->upper_dev; + if (!netif_is_bridge_master(upper_dev)) + return 0; + if (!mlxsw_sp_lower_get(upper_dev)) + return 0; + if (!mlxsw_sp_bridge_vxlan_is_valid(upper_dev, extack)) + return -EOPNOTSUPP; + if (cu_info->linking) { + if (!netif_running(dev)) + return 0; + /* When the bridge is VLAN-aware, the VNI of the VxLAN + * device needs to be mapped to a VLAN, but at this + * point no VLANs are configured on the VxLAN device + */ + if (br_vlan_enabled(upper_dev)) + return 0; + return mlxsw_sp_bridge_vxlan_join(mlxsw_sp, upper_dev, + dev, 0, extack); + } else { + /* VLANs were already flushed, which triggered the + * necessary cleanup + */ + if (br_vlan_enabled(upper_dev)) + return 0; + mlxsw_sp_bridge_vxlan_leave(mlxsw_sp, dev); + } + break; + case NETDEV_PRE_UP: + upper_dev = netdev_master_upper_dev_get(dev); + if (!upper_dev) + return 0; + if (!netif_is_bridge_master(upper_dev)) + return 0; + if (!mlxsw_sp_lower_get(upper_dev)) + return 0; + return mlxsw_sp_bridge_vxlan_join(mlxsw_sp, upper_dev, dev, 0, + extack); + case NETDEV_DOWN: + upper_dev = netdev_master_upper_dev_get(dev); + if (!upper_dev) + return 0; + if (!netif_is_bridge_master(upper_dev)) + return 0; + if (!mlxsw_sp_lower_get(upper_dev)) + return 0; + mlxsw_sp_bridge_vxlan_leave(mlxsw_sp, dev); + break; + } + + return 0; +} + +static int mlxsw_sp_netdevice_event(struct notifier_block *nb, + unsigned long event, void *ptr) +{ + struct net_device *dev = netdev_notifier_info_to_dev(ptr); + struct mlxsw_sp_span_entry *span_entry; + struct mlxsw_sp *mlxsw_sp; + int err = 0; + + mlxsw_sp = container_of(nb, struct mlxsw_sp, netdevice_nb); + if (event == NETDEV_UNREGISTER) { + span_entry = mlxsw_sp_span_entry_find_by_port(mlxsw_sp, dev); + if (span_entry) + mlxsw_sp_span_entry_invalidate(mlxsw_sp, span_entry); + } + mlxsw_sp_span_respin(mlxsw_sp); + + if (netif_is_vxlan(dev)) + err = mlxsw_sp_netdevice_vxlan_event(mlxsw_sp, dev, event, ptr); + else if (mlxsw_sp_port_dev_check(dev)) + err = mlxsw_sp_netdevice_port_event(dev, dev, event, ptr); + else if (netif_is_lag_master(dev)) + err = mlxsw_sp_netdevice_lag_event(dev, event, ptr); + else if (is_vlan_dev(dev)) + err = mlxsw_sp_netdevice_vlan_event(dev, event, ptr); + else if (netif_is_bridge_master(dev)) + err = mlxsw_sp_netdevice_bridge_event(dev, event, ptr); + else if (netif_is_macvlan(dev)) + err = mlxsw_sp_netdevice_macvlan_event(dev, event, ptr); + + return notifier_from_errno(err); +} + +static struct notifier_block mlxsw_sp_inetaddr_valid_nb __read_mostly = { + .notifier_call = mlxsw_sp_inetaddr_valid_event, +}; + +static struct notifier_block mlxsw_sp_inet6addr_valid_nb __read_mostly = { + .notifier_call = mlxsw_sp_inet6addr_valid_event, +}; + +static const struct pci_device_id mlxsw_sp1_pci_id_table[] = { + {PCI_VDEVICE(MELLANOX, PCI_DEVICE_ID_MELLANOX_SPECTRUM), 0}, + {0, }, +}; + +static struct pci_driver mlxsw_sp1_pci_driver = { + .name = mlxsw_sp1_driver_name, + .id_table = mlxsw_sp1_pci_id_table, +}; + +static const struct pci_device_id mlxsw_sp2_pci_id_table[] = { + {PCI_VDEVICE(MELLANOX, PCI_DEVICE_ID_MELLANOX_SPECTRUM2), 0}, + {0, }, +}; + +static struct pci_driver mlxsw_sp2_pci_driver = { + .name = mlxsw_sp2_driver_name, + .id_table = mlxsw_sp2_pci_id_table, +}; + +static const struct pci_device_id mlxsw_sp3_pci_id_table[] = { + {PCI_VDEVICE(MELLANOX, PCI_DEVICE_ID_MELLANOX_SPECTRUM3), 0}, + {0, }, +}; + +static struct pci_driver mlxsw_sp3_pci_driver = { + .name = mlxsw_sp3_driver_name, + .id_table = mlxsw_sp3_pci_id_table, +}; + +static const struct pci_device_id mlxsw_sp4_pci_id_table[] = { + {PCI_VDEVICE(MELLANOX, PCI_DEVICE_ID_MELLANOX_SPECTRUM4), 0}, + {0, }, +}; + +static struct pci_driver mlxsw_sp4_pci_driver = { + .name = mlxsw_sp4_driver_name, + .id_table = mlxsw_sp4_pci_id_table, +}; + +static int __init mlxsw_sp_module_init(void) +{ + int err; + + register_inetaddr_validator_notifier(&mlxsw_sp_inetaddr_valid_nb); + register_inet6addr_validator_notifier(&mlxsw_sp_inet6addr_valid_nb); + + err = mlxsw_core_driver_register(&mlxsw_sp1_driver); + if (err) + goto err_sp1_core_driver_register; + + err = mlxsw_core_driver_register(&mlxsw_sp2_driver); + if (err) + goto err_sp2_core_driver_register; + + err = mlxsw_core_driver_register(&mlxsw_sp3_driver); + if (err) + goto err_sp3_core_driver_register; + + err = mlxsw_core_driver_register(&mlxsw_sp4_driver); + if (err) + goto err_sp4_core_driver_register; + + err = mlxsw_pci_driver_register(&mlxsw_sp1_pci_driver); + if (err) + goto err_sp1_pci_driver_register; + + err = mlxsw_pci_driver_register(&mlxsw_sp2_pci_driver); + if (err) + goto err_sp2_pci_driver_register; + + err = mlxsw_pci_driver_register(&mlxsw_sp3_pci_driver); + if (err) + goto err_sp3_pci_driver_register; + + err = mlxsw_pci_driver_register(&mlxsw_sp4_pci_driver); + if (err) + goto err_sp4_pci_driver_register; + + return 0; + +err_sp4_pci_driver_register: + mlxsw_pci_driver_unregister(&mlxsw_sp3_pci_driver); +err_sp3_pci_driver_register: + mlxsw_pci_driver_unregister(&mlxsw_sp2_pci_driver); +err_sp2_pci_driver_register: + mlxsw_pci_driver_unregister(&mlxsw_sp1_pci_driver); +err_sp1_pci_driver_register: + mlxsw_core_driver_unregister(&mlxsw_sp4_driver); +err_sp4_core_driver_register: + mlxsw_core_driver_unregister(&mlxsw_sp3_driver); +err_sp3_core_driver_register: + mlxsw_core_driver_unregister(&mlxsw_sp2_driver); +err_sp2_core_driver_register: + mlxsw_core_driver_unregister(&mlxsw_sp1_driver); +err_sp1_core_driver_register: + unregister_inet6addr_validator_notifier(&mlxsw_sp_inet6addr_valid_nb); + unregister_inetaddr_validator_notifier(&mlxsw_sp_inetaddr_valid_nb); + return err; +} + +static void __exit mlxsw_sp_module_exit(void) +{ + mlxsw_pci_driver_unregister(&mlxsw_sp4_pci_driver); + mlxsw_pci_driver_unregister(&mlxsw_sp3_pci_driver); + mlxsw_pci_driver_unregister(&mlxsw_sp2_pci_driver); + mlxsw_pci_driver_unregister(&mlxsw_sp1_pci_driver); + mlxsw_core_driver_unregister(&mlxsw_sp4_driver); + mlxsw_core_driver_unregister(&mlxsw_sp3_driver); + mlxsw_core_driver_unregister(&mlxsw_sp2_driver); + mlxsw_core_driver_unregister(&mlxsw_sp1_driver); + unregister_inet6addr_validator_notifier(&mlxsw_sp_inet6addr_valid_nb); + unregister_inetaddr_validator_notifier(&mlxsw_sp_inetaddr_valid_nb); +} + +module_init(mlxsw_sp_module_init); +module_exit(mlxsw_sp_module_exit); + +MODULE_LICENSE("Dual BSD/GPL"); +MODULE_AUTHOR("Jiri Pirko <jiri@mellanox.com>"); +MODULE_DESCRIPTION("Mellanox Spectrum driver"); +MODULE_DEVICE_TABLE(pci, mlxsw_sp1_pci_id_table); +MODULE_DEVICE_TABLE(pci, mlxsw_sp2_pci_id_table); +MODULE_DEVICE_TABLE(pci, mlxsw_sp3_pci_id_table); +MODULE_DEVICE_TABLE(pci, mlxsw_sp4_pci_id_table); +MODULE_FIRMWARE(MLXSW_SP1_FW_FILENAME); +MODULE_FIRMWARE(MLXSW_SP2_FW_FILENAME); +MODULE_FIRMWARE(MLXSW_SP3_FW_FILENAME); +MODULE_FIRMWARE(MLXSW_SP_LINECARDS_INI_BUNDLE_FILENAME); diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h new file mode 100644 index 000000000..c8ff2a6d7 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h @@ -0,0 +1,1493 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_SPECTRUM_H +#define _MLXSW_SPECTRUM_H + +#include <linux/ethtool.h> +#include <linux/types.h> +#include <linux/netdevice.h> +#include <linux/rhashtable.h> +#include <linux/bitops.h> +#include <linux/if_bridge.h> +#include <linux/if_vlan.h> +#include <linux/list.h> +#include <linux/dcbnl.h> +#include <linux/in6.h> +#include <linux/notifier.h> +#include <linux/net_namespace.h> +#include <linux/spinlock.h> +#include <net/psample.h> +#include <net/pkt_cls.h> +#include <net/red.h> +#include <net/vxlan.h> +#include <net/flow_offload.h> +#include <net/inet_ecn.h> + +#include "port.h" +#include "core.h" +#include "core_acl_flex_keys.h" +#include "core_acl_flex_actions.h" +#include "reg.h" + +#define MLXSW_SP_DEFAULT_VID (VLAN_N_VID - 1) + +#define MLXSW_SP_FID_8021D_MAX 1024 + +#define MLXSW_SP_MID_MAX 7000 + +#define MLXSW_SP_KVD_LINEAR_SIZE 98304 /* entries */ +#define MLXSW_SP_KVD_GRANULARITY 128 + +#define MLXSW_SP_RESOURCE_NAME_KVD "kvd" +#define MLXSW_SP_RESOURCE_NAME_KVD_LINEAR "linear" +#define MLXSW_SP_RESOURCE_NAME_KVD_HASH_SINGLE "hash_single" +#define MLXSW_SP_RESOURCE_NAME_KVD_HASH_DOUBLE "hash_double" +#define MLXSW_SP_RESOURCE_NAME_KVD_LINEAR_SINGLES "singles" +#define MLXSW_SP_RESOURCE_NAME_KVD_LINEAR_CHUNKS "chunks" +#define MLXSW_SP_RESOURCE_NAME_KVD_LINEAR_LARGE_CHUNKS "large_chunks" + +#define MLXSW_SP_RESOURCE_NAME_SPAN "span_agents" + +#define MLXSW_SP_RESOURCE_NAME_COUNTERS "counters" +#define MLXSW_SP_RESOURCE_NAME_COUNTERS_FLOW "flow" +#define MLXSW_SP_RESOURCE_NAME_COUNTERS_RIF "rif" + +enum mlxsw_sp_resource_id { + MLXSW_SP_RESOURCE_KVD = MLXSW_CORE_RESOURCE_MAX, + MLXSW_SP_RESOURCE_KVD_LINEAR, + MLXSW_SP_RESOURCE_KVD_HASH_SINGLE, + MLXSW_SP_RESOURCE_KVD_HASH_DOUBLE, + MLXSW_SP_RESOURCE_KVD_LINEAR_SINGLE, + MLXSW_SP_RESOURCE_KVD_LINEAR_CHUNKS, + MLXSW_SP_RESOURCE_KVD_LINEAR_LARGE_CHUNKS, + MLXSW_SP_RESOURCE_SPAN, + MLXSW_SP_RESOURCE_COUNTERS, + MLXSW_SP_RESOURCE_COUNTERS_FLOW, + MLXSW_SP_RESOURCE_COUNTERS_RIF, + MLXSW_SP_RESOURCE_GLOBAL_POLICERS, + MLXSW_SP_RESOURCE_SINGLE_RATE_POLICERS, + MLXSW_SP_RESOURCE_RIF_MAC_PROFILES, + MLXSW_SP_RESOURCE_RIFS, +}; + +struct mlxsw_sp_port; +struct mlxsw_sp_rif; +struct mlxsw_sp_span_entry; +enum mlxsw_sp_l3proto; +union mlxsw_sp_l3addr; + +struct mlxsw_sp_upper { + struct net_device *dev; + unsigned int ref_count; +}; + +enum mlxsw_sp_rif_type { + MLXSW_SP_RIF_TYPE_SUBPORT, + MLXSW_SP_RIF_TYPE_VLAN, + MLXSW_SP_RIF_TYPE_FID, + MLXSW_SP_RIF_TYPE_IPIP_LB, /* IP-in-IP loopback. */ + MLXSW_SP_RIF_TYPE_MAX, +}; + +struct mlxsw_sp_router_ops; + +extern const struct mlxsw_sp_router_ops mlxsw_sp1_router_ops; +extern const struct mlxsw_sp_router_ops mlxsw_sp2_router_ops; + +struct mlxsw_sp_switchdev_ops; + +extern const struct mlxsw_sp_switchdev_ops mlxsw_sp1_switchdev_ops; +extern const struct mlxsw_sp_switchdev_ops mlxsw_sp2_switchdev_ops; + +enum mlxsw_sp_fid_type { + MLXSW_SP_FID_TYPE_8021Q, + MLXSW_SP_FID_TYPE_8021D, + MLXSW_SP_FID_TYPE_RFID, + MLXSW_SP_FID_TYPE_DUMMY, + MLXSW_SP_FID_TYPE_MAX, +}; + +enum mlxsw_sp_nve_type { + MLXSW_SP_NVE_TYPE_VXLAN, +}; + +struct mlxsw_sp_sb; +struct mlxsw_sp_bridge; +struct mlxsw_sp_router; +struct mlxsw_sp_mr; +struct mlxsw_sp_acl; +struct mlxsw_sp_counter_pool; +struct mlxsw_sp_fid_core; +struct mlxsw_sp_kvdl; +struct mlxsw_sp_nve; +struct mlxsw_sp_kvdl_ops; +struct mlxsw_sp_mr_tcam_ops; +struct mlxsw_sp_acl_rulei_ops; +struct mlxsw_sp_acl_tcam_ops; +struct mlxsw_sp_nve_ops; +struct mlxsw_sp_sb_ops; +struct mlxsw_sp_sb_vals; +struct mlxsw_sp_port_type_speed_ops; +struct mlxsw_sp_ptp_state; +struct mlxsw_sp_ptp_ops; +struct mlxsw_sp_span_ops; +struct mlxsw_sp_qdisc_state; +struct mlxsw_sp_mall_entry; +struct mlxsw_sp_pgt; + +struct mlxsw_sp_port_mapping { + u8 module; + u8 slot_index; + u8 width; /* Number of lanes used by the port */ + u8 module_width; /* Number of lanes in the module (static) */ + u8 lane; +}; + +struct mlxsw_sp_port_mapping_events { + struct list_head queue; + spinlock_t queue_lock; /* protects queue */ + struct work_struct work; +}; + +struct mlxsw_sp_parsing { + refcount_t parsing_depth_ref; + u16 parsing_depth; + u16 vxlan_udp_dport; + struct mutex lock; /* Protects parsing configuration */ +}; + +struct mlxsw_sp { + struct mlxsw_sp_port **ports; + struct mlxsw_core *core; + const struct mlxsw_bus_info *bus_info; + unsigned char base_mac[ETH_ALEN]; + const unsigned char *mac_mask; + struct mlxsw_sp_upper *lags; + struct mlxsw_sp_port_mapping *port_mapping; + struct mlxsw_sp_port_mapping_events port_mapping_events; + struct rhashtable sample_trigger_ht; + struct mlxsw_sp_sb *sb; + struct mlxsw_sp_bridge *bridge; + struct mlxsw_sp_router *router; + struct mlxsw_sp_mr *mr; + struct mlxsw_afa *afa; + struct mlxsw_sp_acl *acl; + struct mlxsw_sp_fid_core *fid_core; + struct mlxsw_sp_policer_core *policer_core; + struct mlxsw_sp_kvdl *kvdl; + struct mlxsw_sp_nve *nve; + struct notifier_block netdevice_nb; + struct mlxsw_sp_ptp_clock *clock; + struct mlxsw_sp_ptp_state *ptp_state; + struct mlxsw_sp_counter_pool *counter_pool; + struct mlxsw_sp_span *span; + struct mlxsw_sp_trap *trap; + struct mlxsw_sp_parsing parsing; + const struct mlxsw_sp_switchdev_ops *switchdev_ops; + const struct mlxsw_sp_kvdl_ops *kvdl_ops; + const struct mlxsw_afa_ops *afa_ops; + const struct mlxsw_afk_ops *afk_ops; + const struct mlxsw_sp_mr_tcam_ops *mr_tcam_ops; + const struct mlxsw_sp_acl_rulei_ops *acl_rulei_ops; + const struct mlxsw_sp_acl_tcam_ops *acl_tcam_ops; + const struct mlxsw_sp_acl_bf_ops *acl_bf_ops; + const struct mlxsw_sp_nve_ops **nve_ops_arr; + const struct mlxsw_sp_sb_vals *sb_vals; + const struct mlxsw_sp_sb_ops *sb_ops; + const struct mlxsw_sp_port_type_speed_ops *port_type_speed_ops; + const struct mlxsw_sp_ptp_ops *ptp_ops; + const struct mlxsw_sp_span_ops *span_ops; + const struct mlxsw_sp_policer_core_ops *policer_core_ops; + const struct mlxsw_sp_trap_ops *trap_ops; + const struct mlxsw_sp_mall_ops *mall_ops; + const struct mlxsw_sp_router_ops *router_ops; + const struct mlxsw_listener *listeners; + const struct mlxsw_sp_fid_family **fid_family_arr; + size_t listeners_count; + u32 lowest_shaper_bs; + struct rhashtable ipv6_addr_ht; + struct mutex ipv6_addr_ht_lock; /* Protects ipv6_addr_ht */ + struct mlxsw_sp_pgt *pgt; + bool pgt_smpe_index_valid; +}; + +struct mlxsw_sp_ptp_ops { + struct mlxsw_sp_ptp_clock * + (*clock_init)(struct mlxsw_sp *mlxsw_sp, struct device *dev); + void (*clock_fini)(struct mlxsw_sp_ptp_clock *clock); + + struct mlxsw_sp_ptp_state *(*init)(struct mlxsw_sp *mlxsw_sp); + void (*fini)(struct mlxsw_sp_ptp_state *ptp_state); + + /* Notify a driver that a packet that might be PTP was received. Driver + * is responsible for freeing the passed-in SKB. + */ + void (*receive)(struct mlxsw_sp *mlxsw_sp, struct sk_buff *skb, + u16 local_port); + + /* Notify a driver that a timestamped packet was transmitted. Driver + * is responsible for freeing the passed-in SKB. + */ + void (*transmitted)(struct mlxsw_sp *mlxsw_sp, struct sk_buff *skb, + u16 local_port); + + int (*hwtstamp_get)(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config); + int (*hwtstamp_set)(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config); + void (*shaper_work)(struct work_struct *work); + int (*get_ts_info)(struct mlxsw_sp *mlxsw_sp, + struct ethtool_ts_info *info); + int (*get_stats_count)(void); + void (*get_stats_strings)(u8 **p); + void (*get_stats)(struct mlxsw_sp_port *mlxsw_sp_port, + u64 *data, int data_index); + int (*txhdr_construct)(struct mlxsw_core *mlxsw_core, + struct mlxsw_sp_port *mlxsw_sp_port, + struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info); +}; + +static inline struct mlxsw_sp_upper * +mlxsw_sp_lag_get(struct mlxsw_sp *mlxsw_sp, u16 lag_id) +{ + return &mlxsw_sp->lags[lag_id]; +} + +struct mlxsw_sp_port_pcpu_stats { + u64 rx_packets; + u64 rx_bytes; + u64 tx_packets; + u64 tx_bytes; + struct u64_stats_sync syncp; + u32 tx_dropped; +}; + +enum mlxsw_sp_sample_trigger_type { + MLXSW_SP_SAMPLE_TRIGGER_TYPE_INGRESS, + MLXSW_SP_SAMPLE_TRIGGER_TYPE_EGRESS, + MLXSW_SP_SAMPLE_TRIGGER_TYPE_POLICY_ENGINE, +}; + +struct mlxsw_sp_sample_trigger { + enum mlxsw_sp_sample_trigger_type type; + u16 local_port; /* Reserved when trigger type is not ingress / egress. */ +}; + +struct mlxsw_sp_sample_params { + struct psample_group *psample_group; + u32 trunc_size; + u32 rate; + bool truncate; +}; + +struct mlxsw_sp_bridge_port; +struct mlxsw_sp_fid; + +struct mlxsw_sp_port_vlan { + struct list_head list; + struct mlxsw_sp_port *mlxsw_sp_port; + struct mlxsw_sp_fid *fid; + u16 vid; + struct mlxsw_sp_bridge_port *bridge_port; + struct list_head bridge_vlan_node; +}; + +/* No need an internal lock; At worse - miss a single periodic iteration */ +struct mlxsw_sp_port_xstats { + u64 ecn; + u64 tc_ecn[TC_MAX_QUEUE]; + u64 wred_drop[TC_MAX_QUEUE]; + u64 tail_drop[TC_MAX_QUEUE]; + u64 backlog[TC_MAX_QUEUE]; + u64 tx_bytes[IEEE_8021QAZ_MAX_TCS]; + u64 tx_packets[IEEE_8021QAZ_MAX_TCS]; +}; + +struct mlxsw_sp_ptp_port_dir_stats { + u64 packets; + u64 timestamps; +}; + +struct mlxsw_sp_ptp_port_stats { + struct mlxsw_sp_ptp_port_dir_stats rx_gcd; + struct mlxsw_sp_ptp_port_dir_stats tx_gcd; +}; + +struct mlxsw_sp_port { + struct net_device *dev; + struct mlxsw_sp_port_pcpu_stats __percpu *pcpu_stats; + struct mlxsw_sp *mlxsw_sp; + u16 local_port; + u8 lagged:1, + split:1; + u16 pvid; + u16 lag_id; + struct { + u8 tx_pause:1, + rx_pause:1, + autoneg:1; + } link; + struct { + struct ieee_ets *ets; + struct ieee_maxrate *maxrate; + struct ieee_pfc *pfc; + enum mlxsw_reg_qpts_trust_state trust_state; + } dcb; + struct mlxsw_sp_port_mapping mapping; /* mapping is constant during the + * mlxsw_sp_port lifetime, however + * the same localport can have + * different mapping. + */ + struct { + #define MLXSW_HW_STATS_UPDATE_TIME HZ + struct rtnl_link_stats64 stats; + struct mlxsw_sp_port_xstats xstats; + struct delayed_work update_dw; + } periodic_hw_stats; + struct list_head vlans_list; + struct mlxsw_sp_port_vlan *default_vlan; + struct mlxsw_sp_qdisc_state *qdisc; + unsigned acl_rule_count; + struct mlxsw_sp_flow_block *ing_flow_block; + struct mlxsw_sp_flow_block *eg_flow_block; + struct { + struct delayed_work shaper_dw; + struct hwtstamp_config hwtstamp_config; + u16 ing_types; + u16 egr_types; + struct mlxsw_sp_ptp_port_stats stats; + } ptp; + int max_mtu; + u32 max_speed; + struct mlxsw_sp_hdroom *hdroom; + u64 module_overheat_initial_val; +}; + +struct mlxsw_sp_port_type_speed_ops { + void (*from_ptys_supported_port)(struct mlxsw_sp *mlxsw_sp, + u32 ptys_eth_proto, + struct ethtool_link_ksettings *cmd); + void (*from_ptys_link)(struct mlxsw_sp *mlxsw_sp, u32 ptys_eth_proto, + unsigned long *mode); + u32 (*from_ptys_speed)(struct mlxsw_sp *mlxsw_sp, u32 ptys_eth_proto); + void (*from_ptys_link_mode)(struct mlxsw_sp *mlxsw_sp, + bool carrier_ok, u32 ptys_eth_proto, + struct ethtool_link_ksettings *cmd); + int (*ptys_max_speed)(struct mlxsw_sp_port *mlxsw_sp_port, u32 *p_max_speed); + u32 (*to_ptys_advert_link)(struct mlxsw_sp *mlxsw_sp, + const struct ethtool_link_ksettings *cmd); + u32 (*to_ptys_speed_lanes)(struct mlxsw_sp *mlxsw_sp, u8 width, + const struct ethtool_link_ksettings *cmd); + void (*reg_ptys_eth_pack)(struct mlxsw_sp *mlxsw_sp, char *payload, + u16 local_port, u32 proto_admin, bool autoneg); + void (*reg_ptys_eth_unpack)(struct mlxsw_sp *mlxsw_sp, char *payload, + u32 *p_eth_proto_cap, + u32 *p_eth_proto_admin, + u32 *p_eth_proto_oper); + u32 (*ptys_proto_cap_masked_get)(u32 eth_proto_cap); +}; + +struct mlxsw_sp_ports_bitmap { + unsigned long *bitmap; + unsigned int nbits; +}; + +static inline int +mlxsw_sp_port_bitmap_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ports_bitmap *ports_bm) +{ + unsigned int nbits = mlxsw_core_max_ports(mlxsw_sp->core); + + ports_bm->nbits = nbits; + ports_bm->bitmap = bitmap_zalloc(nbits, GFP_KERNEL); + if (!ports_bm->bitmap) + return -ENOMEM; + + return 0; +} + +static inline void +mlxsw_sp_port_bitmap_fini(struct mlxsw_sp_ports_bitmap *ports_bm) +{ + bitmap_free(ports_bm->bitmap); +} + +static inline u8 mlxsw_sp_tunnel_ecn_decap(u8 outer_ecn, u8 inner_ecn, + bool *trap_en) +{ + bool set_ce = false; + + *trap_en = !!__INET_ECN_decapsulate(outer_ecn, inner_ecn, &set_ce); + if (set_ce) + return INET_ECN_CE; + else if (outer_ecn == INET_ECN_ECT_1 && inner_ecn == INET_ECN_ECT_0) + return INET_ECN_ECT_1; + else + return inner_ecn; +} + +static inline struct net_device * +mlxsw_sp_bridge_vxlan_dev_find(struct net_device *br_dev) +{ + struct net_device *dev; + struct list_head *iter; + + netdev_for_each_lower_dev(br_dev, dev, iter) { + if (netif_is_vxlan(dev)) + return dev; + } + + return NULL; +} + +static inline bool mlxsw_sp_bridge_has_vxlan(struct net_device *br_dev) +{ + return !!mlxsw_sp_bridge_vxlan_dev_find(br_dev); +} + +static inline int +mlxsw_sp_vxlan_mapped_vid(const struct net_device *vxlan_dev, u16 *p_vid) +{ + struct bridge_vlan_info vinfo; + u16 vid = 0; + int err; + + err = br_vlan_get_pvid(vxlan_dev, &vid); + if (err || !vid) + goto out; + + err = br_vlan_get_info(vxlan_dev, vid, &vinfo); + if (err || !(vinfo.flags & BRIDGE_VLAN_INFO_UNTAGGED)) + vid = 0; + +out: + *p_vid = vid; + return err; +} + +static inline bool +mlxsw_sp_port_is_pause_en(const struct mlxsw_sp_port *mlxsw_sp_port) +{ + return mlxsw_sp_port->link.tx_pause || mlxsw_sp_port->link.rx_pause; +} + +static inline struct mlxsw_sp_port * +mlxsw_sp_port_lagged_get(struct mlxsw_sp *mlxsw_sp, u16 lag_id, u8 port_index) +{ + struct mlxsw_sp_port *mlxsw_sp_port; + u16 local_port; + + local_port = mlxsw_core_lag_mapping_get(mlxsw_sp->core, + lag_id, port_index); + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + return mlxsw_sp_port && mlxsw_sp_port->lagged ? mlxsw_sp_port : NULL; +} + +static inline struct mlxsw_sp_port_vlan * +mlxsw_sp_port_vlan_find_by_vid(const struct mlxsw_sp_port *mlxsw_sp_port, + u16 vid) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + + list_for_each_entry(mlxsw_sp_port_vlan, &mlxsw_sp_port->vlans_list, + list) { + if (mlxsw_sp_port_vlan->vid == vid) + return mlxsw_sp_port_vlan; + } + + return NULL; +} + +enum mlxsw_sp_flood_type { + MLXSW_SP_FLOOD_TYPE_UC, + MLXSW_SP_FLOOD_TYPE_BC, + MLXSW_SP_FLOOD_TYPE_MC, +}; + +int mlxsw_sp_port_get_stats_raw(struct net_device *dev, int grp, + int prio, char *ppcnt_pl); +int mlxsw_sp_port_admin_status_set(struct mlxsw_sp_port *mlxsw_sp_port, + bool is_up); +int +mlxsw_sp_port_vlan_classification_set(struct mlxsw_sp_port *mlxsw_sp_port, + bool is_8021ad_tagged, + bool is_8021q_tagged); +static inline bool +mlxsw_sp_local_port_is_valid(struct mlxsw_sp *mlxsw_sp, u16 local_port) +{ + unsigned int max_ports = mlxsw_core_max_ports(mlxsw_sp->core); + + return local_port < max_ports && local_port; +} + +/* spectrum_buffers.c */ +struct mlxsw_sp_hdroom_prio { + /* Number of port buffer associated with this priority. This is the + * actually configured value. + */ + u8 buf_idx; + /* Value of buf_idx deduced from the DCB ETS configuration. */ + u8 ets_buf_idx; + /* Value of buf_idx taken from the dcbnl_setbuffer configuration. */ + u8 set_buf_idx; + bool lossy; +}; + +struct mlxsw_sp_hdroom_buf { + u32 thres_cells; + u32 size_cells; + /* Size requirement form dcbnl_setbuffer. */ + u32 set_size_cells; + bool lossy; +}; + +enum mlxsw_sp_hdroom_mode { + MLXSW_SP_HDROOM_MODE_DCB, + MLXSW_SP_HDROOM_MODE_TC, +}; + +#define MLXSW_SP_PB_COUNT 10 + +struct mlxsw_sp_hdroom { + enum mlxsw_sp_hdroom_mode mode; + + struct { + struct mlxsw_sp_hdroom_prio prio[IEEE_8021Q_MAX_PRIORITIES]; + } prios; + struct { + struct mlxsw_sp_hdroom_buf buf[MLXSW_SP_PB_COUNT]; + } bufs; + struct { + /* Size actually configured for the internal buffer. Equal to + * reserve when internal buffer is enabled. + */ + u32 size_cells; + /* Space reserved in the headroom for the internal buffer. Port + * buffers are not allowed to grow into this space. + */ + u32 reserve_cells; + bool enable; + } int_buf; + int delay_bytes; + int mtu; +}; + +int mlxsw_sp_buffers_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_buffers_fini(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_port_buffers_init(struct mlxsw_sp_port *mlxsw_sp_port); +void mlxsw_sp_port_buffers_fini(struct mlxsw_sp_port *mlxsw_sp_port); +int mlxsw_sp_sb_pool_get(struct mlxsw_core *mlxsw_core, + unsigned int sb_index, u16 pool_index, + struct devlink_sb_pool_info *pool_info); +int mlxsw_sp_sb_pool_set(struct mlxsw_core *mlxsw_core, + unsigned int sb_index, u16 pool_index, u32 size, + enum devlink_sb_threshold_type threshold_type, + struct netlink_ext_ack *extack); +int mlxsw_sp_sb_port_pool_get(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 pool_index, + u32 *p_threshold); +int mlxsw_sp_sb_port_pool_set(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 pool_index, + u32 threshold, struct netlink_ext_ack *extack); +int mlxsw_sp_sb_tc_pool_bind_get(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 tc_index, + enum devlink_sb_pool_type pool_type, + u16 *p_pool_index, u32 *p_threshold); +int mlxsw_sp_sb_tc_pool_bind_set(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 tc_index, + enum devlink_sb_pool_type pool_type, + u16 pool_index, u32 threshold, + struct netlink_ext_ack *extack); +int mlxsw_sp_sb_occ_snapshot(struct mlxsw_core *mlxsw_core, + unsigned int sb_index); +int mlxsw_sp_sb_occ_max_clear(struct mlxsw_core *mlxsw_core, + unsigned int sb_index); +int mlxsw_sp_sb_occ_port_pool_get(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 pool_index, + u32 *p_cur, u32 *p_max); +int mlxsw_sp_sb_occ_tc_port_bind_get(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 tc_index, + enum devlink_sb_pool_type pool_type, + u32 *p_cur, u32 *p_max); +u32 mlxsw_sp_cells_bytes(const struct mlxsw_sp *mlxsw_sp, u32 cells); +u32 mlxsw_sp_bytes_cells(const struct mlxsw_sp *mlxsw_sp, u32 bytes); +void mlxsw_sp_hdroom_prios_reset_buf_idx(struct mlxsw_sp_hdroom *hdroom); +void mlxsw_sp_hdroom_bufs_reset_lossiness(struct mlxsw_sp_hdroom *hdroom); +void mlxsw_sp_hdroom_bufs_reset_sizes(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_hdroom *hdroom); +int mlxsw_sp_hdroom_configure(struct mlxsw_sp_port *mlxsw_sp_port, + const struct mlxsw_sp_hdroom *hdroom); +struct mlxsw_sp_sample_params * +mlxsw_sp_sample_trigger_params_lookup(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_sample_trigger *trigger); +int +mlxsw_sp_sample_trigger_params_set(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_sample_trigger *trigger, + const struct mlxsw_sp_sample_params *params, + struct netlink_ext_ack *extack); +void +mlxsw_sp_sample_trigger_params_unset(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_sample_trigger *trigger); +int mlxsw_sp_ipv6_addr_kvdl_index_get(struct mlxsw_sp *mlxsw_sp, + const struct in6_addr *addr6, + u32 *p_kvdl_index); +void +mlxsw_sp_ipv6_addr_put(struct mlxsw_sp *mlxsw_sp, const struct in6_addr *addr6); + +extern const struct mlxsw_sp_sb_vals mlxsw_sp1_sb_vals; +extern const struct mlxsw_sp_sb_vals mlxsw_sp2_sb_vals; + +extern const struct mlxsw_sp_sb_ops mlxsw_sp1_sb_ops; +extern const struct mlxsw_sp_sb_ops mlxsw_sp2_sb_ops; +extern const struct mlxsw_sp_sb_ops mlxsw_sp3_sb_ops; + +/* spectrum_switchdev.c */ +int mlxsw_sp_switchdev_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_switchdev_fini(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_rif_fdb_op(struct mlxsw_sp *mlxsw_sp, const char *mac, u16 fid, + bool adding); +void +mlxsw_sp_port_vlan_bridge_leave(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan); +int mlxsw_sp_port_bridge_join(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *brport_dev, + struct net_device *br_dev, + struct netlink_ext_ack *extack); +void mlxsw_sp_port_bridge_leave(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *brport_dev, + struct net_device *br_dev); +bool mlxsw_sp_bridge_device_is_offloaded(const struct mlxsw_sp *mlxsw_sp, + const struct net_device *br_dev); +int mlxsw_sp_bridge_vxlan_join(struct mlxsw_sp *mlxsw_sp, + const struct net_device *br_dev, + const struct net_device *vxlan_dev, u16 vid, + struct netlink_ext_ack *extack); +void mlxsw_sp_bridge_vxlan_leave(struct mlxsw_sp *mlxsw_sp, + const struct net_device *vxlan_dev); +extern struct notifier_block mlxsw_sp_switchdev_notifier; + +/* spectrum.c */ +void mlxsw_sp_rx_listener_no_mark_func(struct sk_buff *skb, + u16 local_port, void *priv); +void mlxsw_sp_ptp_receive(struct mlxsw_sp *mlxsw_sp, struct sk_buff *skb, + u16 local_port); +int mlxsw_sp_port_speed_get(struct mlxsw_sp_port *mlxsw_sp_port, u32 *speed); +int mlxsw_sp_port_ets_set(struct mlxsw_sp_port *mlxsw_sp_port, + enum mlxsw_reg_qeec_hr hr, u8 index, u8 next_index, + bool dwrr, u8 dwrr_weight); +int mlxsw_sp_port_prio_tc_set(struct mlxsw_sp_port *mlxsw_sp_port, + u8 switch_prio, u8 tclass); +int mlxsw_sp_port_ets_maxrate_set(struct mlxsw_sp_port *mlxsw_sp_port, + enum mlxsw_reg_qeec_hr hr, u8 index, + u8 next_index, u32 maxrate, u8 burst_size); +enum mlxsw_reg_spms_state mlxsw_sp_stp_spms_state(u8 stp_state); +int mlxsw_sp_port_vid_stp_set(struct mlxsw_sp_port *mlxsw_sp_port, u16 vid, + u8 state); +int mlxsw_sp_port_vp_mode_set(struct mlxsw_sp_port *mlxsw_sp_port, bool enable); +int mlxsw_sp_port_vid_learning_set(struct mlxsw_sp_port *mlxsw_sp_port, u16 vid, + bool learn_enable); +int mlxsw_sp_ethtype_to_sver_type(u16 ethtype, u8 *p_sver_type); +int mlxsw_sp_port_egress_ethtype_set(struct mlxsw_sp_port *mlxsw_sp_port, + u16 ethtype); +int mlxsw_sp_port_pvid_set(struct mlxsw_sp_port *mlxsw_sp_port, u16 vid, + u16 ethtype); +struct mlxsw_sp_port_vlan * +mlxsw_sp_port_vlan_create(struct mlxsw_sp_port *mlxsw_sp_port, u16 vid); +void mlxsw_sp_port_vlan_destroy(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan); +int mlxsw_sp_port_vlan_set(struct mlxsw_sp_port *mlxsw_sp_port, u16 vid_begin, + u16 vid_end, bool is_member, bool untagged); +int mlxsw_sp_flow_counter_get(struct mlxsw_sp *mlxsw_sp, + unsigned int counter_index, u64 *packets, + u64 *bytes); +int mlxsw_sp_flow_counter_alloc(struct mlxsw_sp *mlxsw_sp, + unsigned int *p_counter_index); +void mlxsw_sp_flow_counter_free(struct mlxsw_sp *mlxsw_sp, + unsigned int counter_index); +void mlxsw_sp_txhdr_construct(struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info); +int mlxsw_sp_txhdr_ptp_data_construct(struct mlxsw_core *mlxsw_core, + struct mlxsw_sp_port *mlxsw_sp_port, + struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info); +bool mlxsw_sp_port_dev_check(const struct net_device *dev); +struct mlxsw_sp *mlxsw_sp_lower_get(struct net_device *dev); +struct mlxsw_sp_port *mlxsw_sp_port_dev_lower_find(struct net_device *dev); +struct mlxsw_sp_port *mlxsw_sp_port_lower_dev_hold(struct net_device *dev); +void mlxsw_sp_port_dev_put(struct mlxsw_sp_port *mlxsw_sp_port); +struct mlxsw_sp_port *mlxsw_sp_port_dev_lower_find_rcu(struct net_device *dev); +int mlxsw_sp_parsing_depth_inc(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_parsing_depth_dec(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_parsing_vxlan_udp_dport_set(struct mlxsw_sp *mlxsw_sp, + __be16 udp_dport); + +/* spectrum_dcb.c */ +#ifdef CONFIG_MLXSW_SPECTRUM_DCB +int mlxsw_sp_port_dcb_init(struct mlxsw_sp_port *mlxsw_sp_port); +void mlxsw_sp_port_dcb_fini(struct mlxsw_sp_port *mlxsw_sp_port); +#else +static inline int mlxsw_sp_port_dcb_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + return 0; +} +static inline void mlxsw_sp_port_dcb_fini(struct mlxsw_sp_port *mlxsw_sp_port) +{} +#endif + +/* spectrum_router.c */ +enum mlxsw_sp_l3proto { + MLXSW_SP_L3_PROTO_IPV4, + MLXSW_SP_L3_PROTO_IPV6, +#define MLXSW_SP_L3_PROTO_MAX (MLXSW_SP_L3_PROTO_IPV6 + 1) +}; + +union mlxsw_sp_l3addr { + __be32 addr4; + struct in6_addr addr6; +}; + +u16 mlxsw_sp_rif_index(const struct mlxsw_sp_rif *rif); +int mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp, + struct netlink_ext_ack *extack); +void mlxsw_sp_router_fini(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_rif_macvlan_del(struct mlxsw_sp *mlxsw_sp, + const struct net_device *macvlan_dev); +int mlxsw_sp_inetaddr_valid_event(struct notifier_block *unused, + unsigned long event, void *ptr); +int mlxsw_sp_inet6addr_valid_event(struct notifier_block *unused, + unsigned long event, void *ptr); +int +mlxsw_sp_port_vlan_router_join(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan, + struct net_device *l3_dev, + struct netlink_ext_ack *extack); +void +mlxsw_sp_port_vlan_router_leave(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan); +void mlxsw_sp_rif_destroy_by_dev(struct mlxsw_sp *mlxsw_sp, + struct net_device *dev); +bool mlxsw_sp_rif_exists(struct mlxsw_sp *mlxsw_sp, + const struct net_device *dev); +u16 mlxsw_sp_rif_vid(struct mlxsw_sp *mlxsw_sp, const struct net_device *dev); +u16 mlxsw_sp_router_port(const struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_router_nve_promote_decap(struct mlxsw_sp *mlxsw_sp, u32 ul_tb_id, + enum mlxsw_sp_l3proto ul_proto, + const union mlxsw_sp_l3addr *ul_sip, + u32 tunnel_index); +void mlxsw_sp_router_nve_demote_decap(struct mlxsw_sp *mlxsw_sp, u32 ul_tb_id, + enum mlxsw_sp_l3proto ul_proto, + const union mlxsw_sp_l3addr *ul_sip); +int mlxsw_sp_router_tb_id_vr_id(struct mlxsw_sp *mlxsw_sp, u32 tb_id, + u16 *vr_id); +int mlxsw_sp_router_ul_rif_get(struct mlxsw_sp *mlxsw_sp, u32 ul_tb_id, + u16 *ul_rif_index); +void mlxsw_sp_router_ul_rif_put(struct mlxsw_sp *mlxsw_sp, u16 ul_rif_index); + +/* spectrum_kvdl.c */ +enum mlxsw_sp_kvdl_entry_type { + MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, + MLXSW_SP_KVDL_ENTRY_TYPE_ACTSET, + MLXSW_SP_KVDL_ENTRY_TYPE_PBS, + MLXSW_SP_KVDL_ENTRY_TYPE_MCRIGR, + MLXSW_SP_KVDL_ENTRY_TYPE_IPV6_ADDRESS, + MLXSW_SP_KVDL_ENTRY_TYPE_TNUMT, +}; + +static inline unsigned int +mlxsw_sp_kvdl_entry_size(enum mlxsw_sp_kvdl_entry_type type) +{ + switch (type) { + case MLXSW_SP_KVDL_ENTRY_TYPE_ADJ: + case MLXSW_SP_KVDL_ENTRY_TYPE_ACTSET: + case MLXSW_SP_KVDL_ENTRY_TYPE_PBS: + case MLXSW_SP_KVDL_ENTRY_TYPE_MCRIGR: + case MLXSW_SP_KVDL_ENTRY_TYPE_IPV6_ADDRESS: + case MLXSW_SP_KVDL_ENTRY_TYPE_TNUMT: + default: + return 1; + } +} + +struct mlxsw_sp_kvdl_ops { + size_t priv_size; + int (*init)(struct mlxsw_sp *mlxsw_sp, void *priv); + void (*fini)(struct mlxsw_sp *mlxsw_sp, void *priv); + int (*alloc)(struct mlxsw_sp *mlxsw_sp, void *priv, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, u32 *p_entry_index); + void (*free)(struct mlxsw_sp *mlxsw_sp, void *priv, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, int entry_index); + int (*alloc_size_query)(struct mlxsw_sp *mlxsw_sp, void *priv, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, + unsigned int *p_alloc_count); + int (*resources_register)(struct mlxsw_sp *mlxsw_sp, void *priv); +}; + +int mlxsw_sp_kvdl_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_kvdl_fini(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_kvdl_alloc(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, u32 *p_entry_index); +void mlxsw_sp_kvdl_free(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, int entry_index); +int mlxsw_sp_kvdl_alloc_count_query(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, + unsigned int *p_alloc_count); + +/* spectrum1_kvdl.c */ +extern const struct mlxsw_sp_kvdl_ops mlxsw_sp1_kvdl_ops; +int mlxsw_sp1_kvdl_resources_register(struct mlxsw_core *mlxsw_core); + +/* spectrum2_kvdl.c */ +extern const struct mlxsw_sp_kvdl_ops mlxsw_sp2_kvdl_ops; + +enum mlxsw_sp_acl_mangle_field { + MLXSW_SP_ACL_MANGLE_FIELD_IP_DSFIELD, + MLXSW_SP_ACL_MANGLE_FIELD_IP_DSCP, + MLXSW_SP_ACL_MANGLE_FIELD_IP_ECN, + MLXSW_SP_ACL_MANGLE_FIELD_IP_SPORT, + MLXSW_SP_ACL_MANGLE_FIELD_IP_DPORT, + MLXSW_SP_ACL_MANGLE_FIELD_IP4_SIP, + MLXSW_SP_ACL_MANGLE_FIELD_IP4_DIP, + MLXSW_SP_ACL_MANGLE_FIELD_IP6_SIP_1, + MLXSW_SP_ACL_MANGLE_FIELD_IP6_SIP_2, + MLXSW_SP_ACL_MANGLE_FIELD_IP6_SIP_3, + MLXSW_SP_ACL_MANGLE_FIELD_IP6_SIP_4, + MLXSW_SP_ACL_MANGLE_FIELD_IP6_DIP_1, + MLXSW_SP_ACL_MANGLE_FIELD_IP6_DIP_2, + MLXSW_SP_ACL_MANGLE_FIELD_IP6_DIP_3, + MLXSW_SP_ACL_MANGLE_FIELD_IP6_DIP_4, +}; + +struct mlxsw_sp_acl_rule_info { + unsigned int priority; + struct mlxsw_afk_element_values values; + struct mlxsw_afa_block *act_block; + u8 action_created:1, + ingress_bind_blocker:1, + egress_bind_blocker:1, + counter_valid:1, + policer_index_valid:1, + ipv6_valid:1; + unsigned int counter_index; + u16 policer_index; + struct { + u32 prev_val; + enum mlxsw_sp_acl_mangle_field prev_field; + } ipv6; +}; + +/* spectrum_flow.c */ +struct mlxsw_sp_flow_block { + struct list_head binding_list; + struct { + struct list_head list; + unsigned int min_prio; + unsigned int max_prio; + } mall; + struct mlxsw_sp_acl_ruleset *ruleset_zero; + struct mlxsw_sp *mlxsw_sp; + unsigned int rule_count; + unsigned int disable_count; + unsigned int ingress_blocker_rule_count; + unsigned int egress_blocker_rule_count; + unsigned int ingress_binding_count; + unsigned int egress_binding_count; + struct net *net; +}; + +struct mlxsw_sp_flow_block_binding { + struct list_head list; + struct mlxsw_sp_port *mlxsw_sp_port; + bool ingress; +}; + +static inline struct mlxsw_sp * +mlxsw_sp_flow_block_mlxsw_sp(struct mlxsw_sp_flow_block *block) +{ + return block->mlxsw_sp; +} + +static inline unsigned int +mlxsw_sp_flow_block_rule_count(const struct mlxsw_sp_flow_block *block) +{ + return block ? block->rule_count : 0; +} + +static inline void +mlxsw_sp_flow_block_disable_inc(struct mlxsw_sp_flow_block *block) +{ + if (block) + block->disable_count++; +} + +static inline void +mlxsw_sp_flow_block_disable_dec(struct mlxsw_sp_flow_block *block) +{ + if (block) + block->disable_count--; +} + +static inline bool +mlxsw_sp_flow_block_disabled(const struct mlxsw_sp_flow_block *block) +{ + return block->disable_count; +} + +static inline bool +mlxsw_sp_flow_block_is_egress_bound(const struct mlxsw_sp_flow_block *block) +{ + return block->egress_binding_count; +} + +static inline bool +mlxsw_sp_flow_block_is_ingress_bound(const struct mlxsw_sp_flow_block *block) +{ + return block->ingress_binding_count; +} + +static inline bool +mlxsw_sp_flow_block_is_mixed_bound(const struct mlxsw_sp_flow_block *block) +{ + return block->ingress_binding_count && block->egress_binding_count; +} + +struct mlxsw_sp_flow_block *mlxsw_sp_flow_block_create(struct mlxsw_sp *mlxsw_sp, + struct net *net); +void mlxsw_sp_flow_block_destroy(struct mlxsw_sp_flow_block *block); +int mlxsw_sp_setup_tc_block_clsact(struct mlxsw_sp_port *mlxsw_sp_port, + struct flow_block_offload *f, + bool ingress); + +/* spectrum_acl.c */ +struct mlxsw_sp_acl_ruleset; + +enum mlxsw_sp_acl_profile { + MLXSW_SP_ACL_PROFILE_FLOWER, + MLXSW_SP_ACL_PROFILE_MR, +}; + +struct mlxsw_afk *mlxsw_sp_acl_afk(struct mlxsw_sp_acl *acl); + +int mlxsw_sp_acl_ruleset_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_flow_block_binding *binding); +void mlxsw_sp_acl_ruleset_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_flow_block_binding *binding); +struct mlxsw_sp_acl_ruleset * +mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, u32 chain_index, + enum mlxsw_sp_acl_profile profile); +struct mlxsw_sp_acl_ruleset * +mlxsw_sp_acl_ruleset_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, u32 chain_index, + enum mlxsw_sp_acl_profile profile, + struct mlxsw_afk_element_usage *tmplt_elusage); +void mlxsw_sp_acl_ruleset_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset); +u16 mlxsw_sp_acl_ruleset_group_id(struct mlxsw_sp_acl_ruleset *ruleset); +void mlxsw_sp_acl_ruleset_prio_get(struct mlxsw_sp_acl_ruleset *ruleset, + unsigned int *p_min_prio, + unsigned int *p_max_prio); + +struct mlxsw_sp_acl_rule_info * +mlxsw_sp_acl_rulei_create(struct mlxsw_sp_acl *acl, + struct mlxsw_afa_block *afa_block); +void mlxsw_sp_acl_rulei_destroy(struct mlxsw_sp_acl_rule_info *rulei); +int mlxsw_sp_acl_rulei_commit(struct mlxsw_sp_acl_rule_info *rulei); +void mlxsw_sp_acl_rulei_priority(struct mlxsw_sp_acl_rule_info *rulei, + unsigned int priority); +void mlxsw_sp_acl_rulei_keymask_u32(struct mlxsw_sp_acl_rule_info *rulei, + enum mlxsw_afk_element element, + u32 key_value, u32 mask_value); +void mlxsw_sp_acl_rulei_keymask_buf(struct mlxsw_sp_acl_rule_info *rulei, + enum mlxsw_afk_element element, + const char *key_value, + const char *mask_value, unsigned int len); +int mlxsw_sp_acl_rulei_act_continue(struct mlxsw_sp_acl_rule_info *rulei); +int mlxsw_sp_acl_rulei_act_jump(struct mlxsw_sp_acl_rule_info *rulei, + u16 group_id); +int mlxsw_sp_acl_rulei_act_terminate(struct mlxsw_sp_acl_rule_info *rulei); +int mlxsw_sp_acl_rulei_act_drop(struct mlxsw_sp_acl_rule_info *rulei, + bool ingress, + const struct flow_action_cookie *fa_cookie, + struct netlink_ext_ack *extack); +int mlxsw_sp_acl_rulei_act_trap(struct mlxsw_sp_acl_rule_info *rulei); +int mlxsw_sp_acl_rulei_act_mirror(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct mlxsw_sp_flow_block *block, + struct net_device *out_dev, + struct netlink_ext_ack *extack); +int mlxsw_sp_acl_rulei_act_fwd(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct net_device *out_dev, + struct netlink_ext_ack *extack); +int mlxsw_sp_acl_rulei_act_vlan(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + u32 action, u16 vid, u16 proto, u8 prio, + struct netlink_ext_ack *extack); +int mlxsw_sp_acl_rulei_act_priority(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + u32 prio, struct netlink_ext_ack *extack); +int mlxsw_sp_acl_rulei_act_mangle(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + enum flow_action_mangle_base htype, + u32 offset, u32 mask, u32 val, + struct netlink_ext_ack *extack); +int mlxsw_sp_acl_rulei_act_police(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + u32 index, u64 rate_bytes_ps, + u32 burst, struct netlink_ext_ack *extack); +int mlxsw_sp_acl_rulei_act_count(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct netlink_ext_ack *extack); +int mlxsw_sp_acl_rulei_act_fid_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + u16 fid, struct netlink_ext_ack *extack); +int mlxsw_sp_acl_rulei_act_sample(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct mlxsw_sp_flow_block *block, + struct psample_group *psample_group, u32 rate, + u32 trunc_size, bool truncate, + struct netlink_ext_ack *extack); + +struct mlxsw_sp_acl_rule; + +struct mlxsw_sp_acl_rule * +mlxsw_sp_acl_rule_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset, + unsigned long cookie, + struct mlxsw_afa_block *afa_block, + struct netlink_ext_ack *extack); +void mlxsw_sp_acl_rule_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule *rule); +int mlxsw_sp_acl_rule_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule *rule); +void mlxsw_sp_acl_rule_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule *rule); +int mlxsw_sp_acl_rule_action_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule *rule, + struct mlxsw_afa_block *afa_block); +struct mlxsw_sp_acl_rule * +mlxsw_sp_acl_rule_lookup(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset, + unsigned long cookie); +struct mlxsw_sp_acl_rule_info * +mlxsw_sp_acl_rule_rulei(struct mlxsw_sp_acl_rule *rule); +int mlxsw_sp_acl_rule_get_stats(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule *rule, + u64 *packets, u64 *bytes, u64 *drops, + u64 *last_use, + enum flow_action_hw_stats *used_hw_stats); + +struct mlxsw_sp_fid *mlxsw_sp_acl_dummy_fid(struct mlxsw_sp *mlxsw_sp); + +static inline const struct flow_action_cookie * +mlxsw_sp_acl_act_cookie_lookup(struct mlxsw_sp *mlxsw_sp, u32 cookie_index) +{ + return mlxsw_afa_cookie_lookup(mlxsw_sp->afa, cookie_index); +} + +int mlxsw_sp_acl_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_acl_fini(struct mlxsw_sp *mlxsw_sp); +u32 mlxsw_sp_acl_region_rehash_intrvl_get(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_acl_region_rehash_intrvl_set(struct mlxsw_sp *mlxsw_sp, u32 val); + +struct mlxsw_sp_acl_mangle_action; + +struct mlxsw_sp_acl_rulei_ops { + int (*act_mangle_field)(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_rule_info *rulei, + struct mlxsw_sp_acl_mangle_action *mact, u32 val, + struct netlink_ext_ack *extack); +}; + +extern struct mlxsw_sp_acl_rulei_ops mlxsw_sp1_acl_rulei_ops; +extern struct mlxsw_sp_acl_rulei_ops mlxsw_sp2_acl_rulei_ops; + +/* spectrum_acl_tcam.c */ +struct mlxsw_sp_acl_tcam; +struct mlxsw_sp_acl_tcam_region; + +struct mlxsw_sp_acl_tcam_ops { + enum mlxsw_reg_ptar_key_type key_type; + size_t priv_size; + int (*init)(struct mlxsw_sp *mlxsw_sp, void *priv, + struct mlxsw_sp_acl_tcam *tcam); + void (*fini)(struct mlxsw_sp *mlxsw_sp, void *priv); + size_t region_priv_size; + int (*region_init)(struct mlxsw_sp *mlxsw_sp, void *region_priv, + void *tcam_priv, + struct mlxsw_sp_acl_tcam_region *region, + void *hints_priv); + void (*region_fini)(struct mlxsw_sp *mlxsw_sp, void *region_priv); + int (*region_associate)(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *region); + void * (*region_rehash_hints_get)(void *region_priv); + void (*region_rehash_hints_put)(void *hints_priv); + size_t chunk_priv_size; + void (*chunk_init)(void *region_priv, void *chunk_priv, + unsigned int priority); + void (*chunk_fini)(void *chunk_priv); + size_t entry_priv_size; + int (*entry_add)(struct mlxsw_sp *mlxsw_sp, + void *region_priv, void *chunk_priv, + void *entry_priv, + struct mlxsw_sp_acl_rule_info *rulei); + void (*entry_del)(struct mlxsw_sp *mlxsw_sp, + void *region_priv, void *chunk_priv, + void *entry_priv); + int (*entry_action_replace)(struct mlxsw_sp *mlxsw_sp, + void *region_priv, void *entry_priv, + struct mlxsw_sp_acl_rule_info *rulei); + int (*entry_activity_get)(struct mlxsw_sp *mlxsw_sp, + void *region_priv, void *entry_priv, + bool *activity); +}; + +/* spectrum1_acl_tcam.c */ +extern const struct mlxsw_sp_acl_tcam_ops mlxsw_sp1_acl_tcam_ops; + +/* spectrum2_acl_tcam.c */ +extern const struct mlxsw_sp_acl_tcam_ops mlxsw_sp2_acl_tcam_ops; + +/* spectrum_acl_flex_actions.c */ +extern const struct mlxsw_afa_ops mlxsw_sp1_act_afa_ops; +extern const struct mlxsw_afa_ops mlxsw_sp2_act_afa_ops; + +/* spectrum_acl_flex_keys.c */ +extern const struct mlxsw_afk_ops mlxsw_sp1_afk_ops; +extern const struct mlxsw_afk_ops mlxsw_sp2_afk_ops; +extern const struct mlxsw_afk_ops mlxsw_sp4_afk_ops; + +/* spectrum_acl_bloom_filter.c */ +extern const struct mlxsw_sp_acl_bf_ops mlxsw_sp2_acl_bf_ops; +extern const struct mlxsw_sp_acl_bf_ops mlxsw_sp4_acl_bf_ops; + +/* spectrum_matchall.c */ +struct mlxsw_sp_mall_ops { + int (*sample_add)(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_mall_entry *mall_entry, + struct netlink_ext_ack *extack); + void (*sample_del)(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_mall_entry *mall_entry); +}; + +extern const struct mlxsw_sp_mall_ops mlxsw_sp1_mall_ops; +extern const struct mlxsw_sp_mall_ops mlxsw_sp2_mall_ops; + +enum mlxsw_sp_mall_action_type { + MLXSW_SP_MALL_ACTION_TYPE_MIRROR, + MLXSW_SP_MALL_ACTION_TYPE_SAMPLE, + MLXSW_SP_MALL_ACTION_TYPE_TRAP, +}; + +struct mlxsw_sp_mall_mirror_entry { + const struct net_device *to_dev; + int span_id; +}; + +struct mlxsw_sp_mall_trap_entry { + int span_id; +}; + +struct mlxsw_sp_mall_sample_entry { + struct mlxsw_sp_sample_params params; + int span_id; /* Relevant for Spectrum-2 onwards. */ +}; + +struct mlxsw_sp_mall_entry { + struct list_head list; + unsigned long cookie; + unsigned int priority; + enum mlxsw_sp_mall_action_type type; + bool ingress; + union { + struct mlxsw_sp_mall_mirror_entry mirror; + struct mlxsw_sp_mall_trap_entry trap; + struct mlxsw_sp_mall_sample_entry sample; + }; + struct rcu_head rcu; +}; + +int mlxsw_sp_mall_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct tc_cls_matchall_offload *f); +void mlxsw_sp_mall_destroy(struct mlxsw_sp_flow_block *block, + struct tc_cls_matchall_offload *f); +int mlxsw_sp_mall_port_bind(struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + struct netlink_ext_ack *extack); +void mlxsw_sp_mall_port_unbind(struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_port *mlxsw_sp_port); +int mlxsw_sp_mall_prio_get(struct mlxsw_sp_flow_block *block, u32 chain_index, + unsigned int *p_min_prio, unsigned int *p_max_prio); + +/* spectrum_flower.c */ +int mlxsw_sp_flower_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct flow_cls_offload *f); +void mlxsw_sp_flower_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct flow_cls_offload *f); +int mlxsw_sp_flower_stats(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct flow_cls_offload *f); +int mlxsw_sp_flower_tmplt_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct flow_cls_offload *f); +void mlxsw_sp_flower_tmplt_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct flow_cls_offload *f); +int mlxsw_sp_flower_prio_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + u32 chain_index, unsigned int *p_min_prio, + unsigned int *p_max_prio); + +/* spectrum_qdisc.c */ +int mlxsw_sp_tc_qdisc_init(struct mlxsw_sp_port *mlxsw_sp_port); +void mlxsw_sp_tc_qdisc_fini(struct mlxsw_sp_port *mlxsw_sp_port); +int mlxsw_sp_setup_tc_red(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_red_qopt_offload *p); +int mlxsw_sp_setup_tc_prio(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_prio_qopt_offload *p); +int mlxsw_sp_setup_tc_ets(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_ets_qopt_offload *p); +int mlxsw_sp_setup_tc_tbf(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_tbf_qopt_offload *p); +int mlxsw_sp_setup_tc_fifo(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_fifo_qopt_offload *p); +int mlxsw_sp_setup_tc_block_qevent_early_drop(struct mlxsw_sp_port *mlxsw_sp_port, + struct flow_block_offload *f); +int mlxsw_sp_setup_tc_block_qevent_mark(struct mlxsw_sp_port *mlxsw_sp_port, + struct flow_block_offload *f); + +/* spectrum_fid.c */ +bool mlxsw_sp_fid_is_dummy(struct mlxsw_sp *mlxsw_sp, u16 fid_index); +struct mlxsw_sp_fid *mlxsw_sp_fid_lookup_by_index(struct mlxsw_sp *mlxsw_sp, + u16 fid_index); +int mlxsw_sp_fid_nve_ifindex(const struct mlxsw_sp_fid *fid, int *nve_ifindex); +int mlxsw_sp_fid_nve_type(const struct mlxsw_sp_fid *fid, + enum mlxsw_sp_nve_type *p_type); +struct mlxsw_sp_fid *mlxsw_sp_fid_lookup_by_vni(struct mlxsw_sp *mlxsw_sp, + __be32 vni); +int mlxsw_sp_fid_vni(const struct mlxsw_sp_fid *fid, __be32 *vni); +int mlxsw_sp_fid_nve_flood_index_set(struct mlxsw_sp_fid *fid, + u32 nve_flood_index); +void mlxsw_sp_fid_nve_flood_index_clear(struct mlxsw_sp_fid *fid); +bool mlxsw_sp_fid_nve_flood_index_is_set(const struct mlxsw_sp_fid *fid); +int mlxsw_sp_fid_vni_set(struct mlxsw_sp_fid *fid, enum mlxsw_sp_nve_type type, + __be32 vni, int nve_ifindex); +void mlxsw_sp_fid_vni_clear(struct mlxsw_sp_fid *fid); +bool mlxsw_sp_fid_vni_is_set(const struct mlxsw_sp_fid *fid); +void mlxsw_sp_fid_fdb_clear_offload(const struct mlxsw_sp_fid *fid, + const struct net_device *nve_dev); +int mlxsw_sp_fid_flood_set(struct mlxsw_sp_fid *fid, + enum mlxsw_sp_flood_type packet_type, u16 local_port, + bool member); +int mlxsw_sp_fid_port_vid_map(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_port *mlxsw_sp_port, u16 vid); +void mlxsw_sp_fid_port_vid_unmap(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_port *mlxsw_sp_port, u16 vid); +u16 mlxsw_sp_fid_index(const struct mlxsw_sp_fid *fid); +enum mlxsw_sp_fid_type mlxsw_sp_fid_type(const struct mlxsw_sp_fid *fid); +int mlxsw_sp_fid_rif_set(struct mlxsw_sp_fid *fid, struct mlxsw_sp_rif *rif); +void mlxsw_sp_fid_rif_unset(struct mlxsw_sp_fid *fid); +struct mlxsw_sp_rif *mlxsw_sp_fid_rif(const struct mlxsw_sp_fid *fid); +enum mlxsw_sp_rif_type +mlxsw_sp_fid_type_rif_type(const struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_fid_type type); +u16 mlxsw_sp_fid_8021q_vid(const struct mlxsw_sp_fid *fid); +struct mlxsw_sp_fid *mlxsw_sp_fid_8021q_get(struct mlxsw_sp *mlxsw_sp, u16 vid); +struct mlxsw_sp_fid *mlxsw_sp_fid_8021d_get(struct mlxsw_sp *mlxsw_sp, + int br_ifindex); +struct mlxsw_sp_fid *mlxsw_sp_fid_8021q_lookup(struct mlxsw_sp *mlxsw_sp, + u16 vid); +struct mlxsw_sp_fid *mlxsw_sp_fid_8021d_lookup(struct mlxsw_sp *mlxsw_sp, + int br_ifindex); +struct mlxsw_sp_fid *mlxsw_sp_fid_rfid_get(struct mlxsw_sp *mlxsw_sp, + u16 rif_index); +struct mlxsw_sp_fid *mlxsw_sp_fid_dummy_get(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_fid_put(struct mlxsw_sp_fid *fid); +int mlxsw_sp_port_fids_init(struct mlxsw_sp_port *mlxsw_sp_port); +void mlxsw_sp_port_fids_fini(struct mlxsw_sp_port *mlxsw_sp_port); +int mlxsw_sp_fids_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_fids_fini(struct mlxsw_sp *mlxsw_sp); + +extern const struct mlxsw_sp_fid_family *mlxsw_sp1_fid_family_arr[]; +extern const struct mlxsw_sp_fid_family *mlxsw_sp2_fid_family_arr[]; + +/* spectrum_mr.c */ +enum mlxsw_sp_mr_route_prio { + MLXSW_SP_MR_ROUTE_PRIO_SG, + MLXSW_SP_MR_ROUTE_PRIO_STARG, + MLXSW_SP_MR_ROUTE_PRIO_CATCHALL, + __MLXSW_SP_MR_ROUTE_PRIO_MAX +}; + +#define MLXSW_SP_MR_ROUTE_PRIO_MAX (__MLXSW_SP_MR_ROUTE_PRIO_MAX - 1) + +struct mlxsw_sp_mr_route_key; + +struct mlxsw_sp_mr_tcam_ops { + size_t priv_size; + int (*init)(struct mlxsw_sp *mlxsw_sp, void *priv); + void (*fini)(void *priv); + size_t route_priv_size; + int (*route_create)(struct mlxsw_sp *mlxsw_sp, void *priv, + void *route_priv, + struct mlxsw_sp_mr_route_key *key, + struct mlxsw_afa_block *afa_block, + enum mlxsw_sp_mr_route_prio prio); + void (*route_destroy)(struct mlxsw_sp *mlxsw_sp, void *priv, + void *route_priv, + struct mlxsw_sp_mr_route_key *key); + int (*route_update)(struct mlxsw_sp *mlxsw_sp, void *route_priv, + struct mlxsw_sp_mr_route_key *key, + struct mlxsw_afa_block *afa_block); +}; + +/* spectrum1_mr_tcam.c */ +extern const struct mlxsw_sp_mr_tcam_ops mlxsw_sp1_mr_tcam_ops; + +/* spectrum2_mr_tcam.c */ +extern const struct mlxsw_sp_mr_tcam_ops mlxsw_sp2_mr_tcam_ops; + +/* spectrum_nve.c */ +struct mlxsw_sp_nve_params { + enum mlxsw_sp_nve_type type; + __be32 vni; + const struct net_device *dev; + u16 ethertype; +}; + +extern const struct mlxsw_sp_nve_ops *mlxsw_sp1_nve_ops_arr[]; +extern const struct mlxsw_sp_nve_ops *mlxsw_sp2_nve_ops_arr[]; + +int mlxsw_sp_nve_learned_ip_resolve(struct mlxsw_sp *mlxsw_sp, u32 uip, + enum mlxsw_sp_l3proto proto, + union mlxsw_sp_l3addr *addr); +int mlxsw_sp_nve_flood_ip_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fid *fid, + enum mlxsw_sp_l3proto proto, + union mlxsw_sp_l3addr *addr); +void mlxsw_sp_nve_flood_ip_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fid *fid, + enum mlxsw_sp_l3proto proto, + union mlxsw_sp_l3addr *addr); +int mlxsw_sp_nve_ipv6_addr_kvdl_set(struct mlxsw_sp *mlxsw_sp, + const struct in6_addr *addr6, + u32 *p_kvdl_index); +void mlxsw_sp_nve_ipv6_addr_kvdl_unset(struct mlxsw_sp *mlxsw_sp, + const struct in6_addr *addr6); +int +mlxsw_sp_nve_ipv6_addr_map_replace(struct mlxsw_sp *mlxsw_sp, const char *mac, + u16 fid_index, + const struct in6_addr *new_addr6); +void mlxsw_sp_nve_ipv6_addr_map_del(struct mlxsw_sp *mlxsw_sp, const char *mac, + u16 fid_index); +int mlxsw_sp_nve_fid_enable(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fid *fid, + struct mlxsw_sp_nve_params *params, + struct netlink_ext_ack *extack); +void mlxsw_sp_nve_fid_disable(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fid *fid); +int mlxsw_sp_port_nve_init(struct mlxsw_sp_port *mlxsw_sp_port); +void mlxsw_sp_port_nve_fini(struct mlxsw_sp_port *mlxsw_sp_port); +int mlxsw_sp_nve_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_nve_fini(struct mlxsw_sp *mlxsw_sp); + +/* spectrum_nve_vxlan.c */ +int mlxsw_sp_nve_inc_parsing_depth_get(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_nve_inc_parsing_depth_put(struct mlxsw_sp *mlxsw_sp); + +/* spectrum_trap.c */ +int mlxsw_sp_devlink_traps_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_devlink_traps_fini(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_trap_init(struct mlxsw_core *mlxsw_core, + const struct devlink_trap *trap, void *trap_ctx); +void mlxsw_sp_trap_fini(struct mlxsw_core *mlxsw_core, + const struct devlink_trap *trap, void *trap_ctx); +int mlxsw_sp_trap_action_set(struct mlxsw_core *mlxsw_core, + const struct devlink_trap *trap, + enum devlink_trap_action action, + struct netlink_ext_ack *extack); +int mlxsw_sp_trap_group_init(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_group *group); +int mlxsw_sp_trap_group_set(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_group *group, + const struct devlink_trap_policer *policer, + struct netlink_ext_ack *extack); +int +mlxsw_sp_trap_policer_init(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_policer *policer); +void mlxsw_sp_trap_policer_fini(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_policer *policer); +int +mlxsw_sp_trap_policer_set(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_policer *policer, + u64 rate, u64 burst, struct netlink_ext_ack *extack); +int +mlxsw_sp_trap_policer_counter_get(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_policer *policer, + u64 *p_drops); +int mlxsw_sp_trap_group_policer_hw_id_get(struct mlxsw_sp *mlxsw_sp, u16 id, + bool *p_enabled, u16 *p_hw_id); + +static inline struct net *mlxsw_sp_net(struct mlxsw_sp *mlxsw_sp) +{ + return mlxsw_core_net(mlxsw_sp->core); +} + +/* spectrum_ethtool.c */ +extern const struct ethtool_ops mlxsw_sp_port_ethtool_ops; +extern const struct mlxsw_sp_port_type_speed_ops mlxsw_sp1_port_type_speed_ops; +extern const struct mlxsw_sp_port_type_speed_ops mlxsw_sp2_port_type_speed_ops; + +/* spectrum_policer.c */ +extern const struct mlxsw_sp_policer_core_ops mlxsw_sp1_policer_core_ops; +extern const struct mlxsw_sp_policer_core_ops mlxsw_sp2_policer_core_ops; + +enum mlxsw_sp_policer_type { + MLXSW_SP_POLICER_TYPE_SINGLE_RATE, + + __MLXSW_SP_POLICER_TYPE_MAX, + MLXSW_SP_POLICER_TYPE_MAX = __MLXSW_SP_POLICER_TYPE_MAX - 1, +}; + +struct mlxsw_sp_policer_params { + u64 rate; + u64 burst; + bool bytes; +}; + +int mlxsw_sp_policer_add(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_policer_type type, + const struct mlxsw_sp_policer_params *params, + struct netlink_ext_ack *extack, u16 *p_policer_index); +void mlxsw_sp_policer_del(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_policer_type type, + u16 policer_index); +int mlxsw_sp_policer_drops_counter_get(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_policer_type type, + u16 policer_index, u64 *p_drops); +int mlxsw_sp_policers_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_policers_fini(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_policer_resources_register(struct mlxsw_core *mlxsw_core); + +/* spectrum_pgt.c */ +int mlxsw_sp_pgt_mid_alloc(struct mlxsw_sp *mlxsw_sp, u16 *p_mid); +void mlxsw_sp_pgt_mid_free(struct mlxsw_sp *mlxsw_sp, u16 mid_base); +int mlxsw_sp_pgt_mid_alloc_range(struct mlxsw_sp *mlxsw_sp, u16 mid_base, + u16 count); +void mlxsw_sp_pgt_mid_free_range(struct mlxsw_sp *mlxsw_sp, u16 mid_base, + u16 count); +int mlxsw_sp_pgt_entry_port_set(struct mlxsw_sp *mlxsw_sp, u16 mid, + u16 smpe, u16 local_port, bool member); +int mlxsw_sp_pgt_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_pgt_fini(struct mlxsw_sp *mlxsw_sp); + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum1_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum1_acl_tcam.c new file mode 100644 index 000000000..3a636f753 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum1_acl_tcam.c @@ -0,0 +1,254 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/slab.h> + +#include "reg.h" +#include "core.h" +#include "spectrum.h" +#include "spectrum_acl_tcam.h" + +struct mlxsw_sp1_acl_tcam_region { + struct mlxsw_sp_acl_ctcam_region cregion; + struct mlxsw_sp_acl_tcam_region *region; + struct { + struct mlxsw_sp_acl_ctcam_chunk cchunk; + struct mlxsw_sp_acl_ctcam_entry centry; + struct mlxsw_sp_acl_rule_info *rulei; + } catchall; +}; + +struct mlxsw_sp1_acl_tcam_chunk { + struct mlxsw_sp_acl_ctcam_chunk cchunk; +}; + +struct mlxsw_sp1_acl_tcam_entry { + struct mlxsw_sp_acl_ctcam_entry centry; +}; + +static int +mlxsw_sp1_acl_ctcam_region_entry_insert(struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_entry *centry, + const char *mask) +{ + return 0; +} + +static void +mlxsw_sp1_acl_ctcam_region_entry_remove(struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_entry *centry) +{ +} + +static const struct mlxsw_sp_acl_ctcam_region_ops +mlxsw_sp1_acl_ctcam_region_ops = { + .entry_insert = mlxsw_sp1_acl_ctcam_region_entry_insert, + .entry_remove = mlxsw_sp1_acl_ctcam_region_entry_remove, +}; + +static int mlxsw_sp1_acl_tcam_init(struct mlxsw_sp *mlxsw_sp, void *priv, + struct mlxsw_sp_acl_tcam *tcam) +{ + return 0; +} + +static void mlxsw_sp1_acl_tcam_fini(struct mlxsw_sp *mlxsw_sp, void *priv) +{ +} + +static int +mlxsw_sp1_acl_ctcam_region_catchall_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp1_acl_tcam_region *region) +{ + struct mlxsw_sp_acl_rule_info *rulei; + int err; + + mlxsw_sp_acl_ctcam_chunk_init(®ion->cregion, + ®ion->catchall.cchunk, + MLXSW_SP_ACL_TCAM_CATCHALL_PRIO); + rulei = mlxsw_sp_acl_rulei_create(mlxsw_sp->acl, NULL); + if (IS_ERR(rulei)) { + err = PTR_ERR(rulei); + goto err_rulei_create; + } + err = mlxsw_sp_acl_rulei_act_continue(rulei); + if (WARN_ON(err)) + goto err_rulei_act_continue; + err = mlxsw_sp_acl_rulei_commit(rulei); + if (err) + goto err_rulei_commit; + err = mlxsw_sp_acl_ctcam_entry_add(mlxsw_sp, ®ion->cregion, + ®ion->catchall.cchunk, + ®ion->catchall.centry, + rulei, false); + if (err) + goto err_entry_add; + region->catchall.rulei = rulei; + return 0; + +err_entry_add: +err_rulei_commit: +err_rulei_act_continue: + mlxsw_sp_acl_rulei_destroy(rulei); +err_rulei_create: + mlxsw_sp_acl_ctcam_chunk_fini(®ion->catchall.cchunk); + return err; +} + +static void +mlxsw_sp1_acl_ctcam_region_catchall_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp1_acl_tcam_region *region) +{ + struct mlxsw_sp_acl_rule_info *rulei = region->catchall.rulei; + + mlxsw_sp_acl_ctcam_entry_del(mlxsw_sp, ®ion->cregion, + ®ion->catchall.cchunk, + ®ion->catchall.centry); + mlxsw_sp_acl_rulei_destroy(rulei); + mlxsw_sp_acl_ctcam_chunk_fini(®ion->catchall.cchunk); +} + +static int +mlxsw_sp1_acl_tcam_region_init(struct mlxsw_sp *mlxsw_sp, void *region_priv, + void *tcam_priv, + struct mlxsw_sp_acl_tcam_region *_region, + void *hints_priv) +{ + struct mlxsw_sp1_acl_tcam_region *region = region_priv; + int err; + + err = mlxsw_sp_acl_ctcam_region_init(mlxsw_sp, ®ion->cregion, + _region, + &mlxsw_sp1_acl_ctcam_region_ops); + if (err) + return err; + err = mlxsw_sp1_acl_ctcam_region_catchall_add(mlxsw_sp, region); + if (err) + goto err_catchall_add; + region->region = _region; + return 0; + +err_catchall_add: + mlxsw_sp_acl_ctcam_region_fini(®ion->cregion); + return err; +} + +static void +mlxsw_sp1_acl_tcam_region_fini(struct mlxsw_sp *mlxsw_sp, void *region_priv) +{ + struct mlxsw_sp1_acl_tcam_region *region = region_priv; + + mlxsw_sp1_acl_ctcam_region_catchall_del(mlxsw_sp, region); + mlxsw_sp_acl_ctcam_region_fini(®ion->cregion); +} + +static int +mlxsw_sp1_acl_tcam_region_associate(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *region) +{ + return 0; +} + +static void mlxsw_sp1_acl_tcam_chunk_init(void *region_priv, void *chunk_priv, + unsigned int priority) +{ + struct mlxsw_sp1_acl_tcam_region *region = region_priv; + struct mlxsw_sp1_acl_tcam_chunk *chunk = chunk_priv; + + mlxsw_sp_acl_ctcam_chunk_init(®ion->cregion, &chunk->cchunk, + priority); +} + +static void mlxsw_sp1_acl_tcam_chunk_fini(void *chunk_priv) +{ + struct mlxsw_sp1_acl_tcam_chunk *chunk = chunk_priv; + + mlxsw_sp_acl_ctcam_chunk_fini(&chunk->cchunk); +} + +static int mlxsw_sp1_acl_tcam_entry_add(struct mlxsw_sp *mlxsw_sp, + void *region_priv, void *chunk_priv, + void *entry_priv, + struct mlxsw_sp_acl_rule_info *rulei) +{ + struct mlxsw_sp1_acl_tcam_region *region = region_priv; + struct mlxsw_sp1_acl_tcam_chunk *chunk = chunk_priv; + struct mlxsw_sp1_acl_tcam_entry *entry = entry_priv; + + return mlxsw_sp_acl_ctcam_entry_add(mlxsw_sp, ®ion->cregion, + &chunk->cchunk, &entry->centry, + rulei, false); +} + +static void mlxsw_sp1_acl_tcam_entry_del(struct mlxsw_sp *mlxsw_sp, + void *region_priv, void *chunk_priv, + void *entry_priv) +{ + struct mlxsw_sp1_acl_tcam_region *region = region_priv; + struct mlxsw_sp1_acl_tcam_chunk *chunk = chunk_priv; + struct mlxsw_sp1_acl_tcam_entry *entry = entry_priv; + + mlxsw_sp_acl_ctcam_entry_del(mlxsw_sp, ®ion->cregion, + &chunk->cchunk, &entry->centry); +} + +static int +mlxsw_sp1_acl_tcam_entry_action_replace(struct mlxsw_sp *mlxsw_sp, + void *region_priv, void *entry_priv, + struct mlxsw_sp_acl_rule_info *rulei) +{ + return -EOPNOTSUPP; +} + +static int +mlxsw_sp1_acl_tcam_region_entry_activity_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *_region, + unsigned int offset, + bool *activity) +{ + char ptce2_pl[MLXSW_REG_PTCE2_LEN]; + int err; + + mlxsw_reg_ptce2_pack(ptce2_pl, true, MLXSW_REG_PTCE2_OP_QUERY_CLEAR_ON_READ, + _region->tcam_region_info, offset, 0); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ptce2), ptce2_pl); + if (err) + return err; + *activity = mlxsw_reg_ptce2_a_get(ptce2_pl); + return 0; +} + +static int +mlxsw_sp1_acl_tcam_entry_activity_get(struct mlxsw_sp *mlxsw_sp, + void *region_priv, void *entry_priv, + bool *activity) +{ + struct mlxsw_sp1_acl_tcam_region *region = region_priv; + struct mlxsw_sp1_acl_tcam_entry *entry = entry_priv; + unsigned int offset; + + offset = mlxsw_sp_acl_ctcam_entry_offset(&entry->centry); + return mlxsw_sp1_acl_tcam_region_entry_activity_get(mlxsw_sp, + region->region, + offset, activity); +} + +const struct mlxsw_sp_acl_tcam_ops mlxsw_sp1_acl_tcam_ops = { + .key_type = MLXSW_REG_PTAR_KEY_TYPE_FLEX, + .priv_size = 0, + .init = mlxsw_sp1_acl_tcam_init, + .fini = mlxsw_sp1_acl_tcam_fini, + .region_priv_size = sizeof(struct mlxsw_sp1_acl_tcam_region), + .region_init = mlxsw_sp1_acl_tcam_region_init, + .region_fini = mlxsw_sp1_acl_tcam_region_fini, + .region_associate = mlxsw_sp1_acl_tcam_region_associate, + .chunk_priv_size = sizeof(struct mlxsw_sp1_acl_tcam_chunk), + .chunk_init = mlxsw_sp1_acl_tcam_chunk_init, + .chunk_fini = mlxsw_sp1_acl_tcam_chunk_fini, + .entry_priv_size = sizeof(struct mlxsw_sp1_acl_tcam_entry), + .entry_add = mlxsw_sp1_acl_tcam_entry_add, + .entry_del = mlxsw_sp1_acl_tcam_entry_del, + .entry_action_replace = mlxsw_sp1_acl_tcam_entry_action_replace, + .entry_activity_get = mlxsw_sp1_acl_tcam_entry_activity_get, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum1_kvdl.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum1_kvdl.c new file mode 100644 index 000000000..1e3fc9893 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum1_kvdl.c @@ -0,0 +1,427 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/bitops.h> + +#include "spectrum.h" + +#define MLXSW_SP1_KVDL_SINGLE_BASE 0 +#define MLXSW_SP1_KVDL_SINGLE_SIZE 16384 +#define MLXSW_SP1_KVDL_SINGLE_END \ + (MLXSW_SP1_KVDL_SINGLE_SIZE + MLXSW_SP1_KVDL_SINGLE_BASE - 1) + +#define MLXSW_SP1_KVDL_CHUNKS_BASE \ + (MLXSW_SP1_KVDL_SINGLE_BASE + MLXSW_SP1_KVDL_SINGLE_SIZE) +#define MLXSW_SP1_KVDL_CHUNKS_SIZE 49152 +#define MLXSW_SP1_KVDL_CHUNKS_END \ + (MLXSW_SP1_KVDL_CHUNKS_SIZE + MLXSW_SP1_KVDL_CHUNKS_BASE - 1) + +#define MLXSW_SP1_KVDL_LARGE_CHUNKS_BASE \ + (MLXSW_SP1_KVDL_CHUNKS_BASE + MLXSW_SP1_KVDL_CHUNKS_SIZE) +#define MLXSW_SP1_KVDL_LARGE_CHUNKS_SIZE \ + (MLXSW_SP_KVD_LINEAR_SIZE - MLXSW_SP1_KVDL_LARGE_CHUNKS_BASE) +#define MLXSW_SP1_KVDL_LARGE_CHUNKS_END \ + (MLXSW_SP1_KVDL_LARGE_CHUNKS_SIZE + MLXSW_SP1_KVDL_LARGE_CHUNKS_BASE - 1) + +#define MLXSW_SP1_KVDL_SINGLE_ALLOC_SIZE 1 +#define MLXSW_SP1_KVDL_CHUNKS_ALLOC_SIZE 32 +#define MLXSW_SP1_KVDL_LARGE_CHUNKS_ALLOC_SIZE 512 + +struct mlxsw_sp1_kvdl_part_info { + unsigned int part_index; + unsigned int start_index; + unsigned int end_index; + unsigned int alloc_size; + enum mlxsw_sp_resource_id resource_id; +}; + +enum mlxsw_sp1_kvdl_part_id { + MLXSW_SP1_KVDL_PART_ID_SINGLE, + MLXSW_SP1_KVDL_PART_ID_CHUNKS, + MLXSW_SP1_KVDL_PART_ID_LARGE_CHUNKS, +}; + +#define MLXSW_SP1_KVDL_PART_INFO(id) \ +[MLXSW_SP1_KVDL_PART_ID_##id] = { \ + .start_index = MLXSW_SP1_KVDL_##id##_BASE, \ + .end_index = MLXSW_SP1_KVDL_##id##_END, \ + .alloc_size = MLXSW_SP1_KVDL_##id##_ALLOC_SIZE, \ + .resource_id = MLXSW_SP_RESOURCE_KVD_LINEAR_##id, \ +} + +static const struct mlxsw_sp1_kvdl_part_info mlxsw_sp1_kvdl_parts_info[] = { + MLXSW_SP1_KVDL_PART_INFO(SINGLE), + MLXSW_SP1_KVDL_PART_INFO(CHUNKS), + MLXSW_SP1_KVDL_PART_INFO(LARGE_CHUNKS), +}; + +#define MLXSW_SP1_KVDL_PARTS_INFO_LEN ARRAY_SIZE(mlxsw_sp1_kvdl_parts_info) + +struct mlxsw_sp1_kvdl_part { + struct mlxsw_sp1_kvdl_part_info info; + unsigned long usage[]; /* Entries */ +}; + +struct mlxsw_sp1_kvdl { + struct mlxsw_sp1_kvdl_part *parts[MLXSW_SP1_KVDL_PARTS_INFO_LEN]; +}; + +static struct mlxsw_sp1_kvdl_part * +mlxsw_sp1_kvdl_alloc_size_part(struct mlxsw_sp1_kvdl *kvdl, + unsigned int alloc_size) +{ + struct mlxsw_sp1_kvdl_part *part, *min_part = NULL; + int i; + + for (i = 0; i < MLXSW_SP1_KVDL_PARTS_INFO_LEN; i++) { + part = kvdl->parts[i]; + if (alloc_size <= part->info.alloc_size && + (!min_part || + part->info.alloc_size <= min_part->info.alloc_size)) + min_part = part; + } + + return min_part ?: ERR_PTR(-ENOBUFS); +} + +static struct mlxsw_sp1_kvdl_part * +mlxsw_sp1_kvdl_index_part(struct mlxsw_sp1_kvdl *kvdl, u32 kvdl_index) +{ + struct mlxsw_sp1_kvdl_part *part; + int i; + + for (i = 0; i < MLXSW_SP1_KVDL_PARTS_INFO_LEN; i++) { + part = kvdl->parts[i]; + if (kvdl_index >= part->info.start_index && + kvdl_index <= part->info.end_index) + return part; + } + + return ERR_PTR(-EINVAL); +} + +static u32 +mlxsw_sp1_kvdl_to_kvdl_index(const struct mlxsw_sp1_kvdl_part_info *info, + unsigned int entry_index) +{ + return info->start_index + entry_index * info->alloc_size; +} + +static unsigned int +mlxsw_sp1_kvdl_to_entry_index(const struct mlxsw_sp1_kvdl_part_info *info, + u32 kvdl_index) +{ + return (kvdl_index - info->start_index) / info->alloc_size; +} + +static int mlxsw_sp1_kvdl_part_alloc(struct mlxsw_sp1_kvdl_part *part, + u32 *p_kvdl_index) +{ + const struct mlxsw_sp1_kvdl_part_info *info = &part->info; + unsigned int entry_index, nr_entries; + + nr_entries = (info->end_index - info->start_index + 1) / + info->alloc_size; + entry_index = find_first_zero_bit(part->usage, nr_entries); + if (entry_index == nr_entries) + return -ENOBUFS; + __set_bit(entry_index, part->usage); + + *p_kvdl_index = mlxsw_sp1_kvdl_to_kvdl_index(info, entry_index); + + return 0; +} + +static void mlxsw_sp1_kvdl_part_free(struct mlxsw_sp1_kvdl_part *part, + u32 kvdl_index) +{ + const struct mlxsw_sp1_kvdl_part_info *info = &part->info; + unsigned int entry_index; + + entry_index = mlxsw_sp1_kvdl_to_entry_index(info, kvdl_index); + __clear_bit(entry_index, part->usage); +} + +static int mlxsw_sp1_kvdl_alloc(struct mlxsw_sp *mlxsw_sp, void *priv, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, + u32 *p_entry_index) +{ + struct mlxsw_sp1_kvdl *kvdl = priv; + struct mlxsw_sp1_kvdl_part *part; + + /* Find partition with smallest allocation size satisfying the + * requested size. + */ + part = mlxsw_sp1_kvdl_alloc_size_part(kvdl, entry_count); + if (IS_ERR(part)) + return PTR_ERR(part); + + return mlxsw_sp1_kvdl_part_alloc(part, p_entry_index); +} + +static void mlxsw_sp1_kvdl_free(struct mlxsw_sp *mlxsw_sp, void *priv, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, int entry_index) +{ + struct mlxsw_sp1_kvdl *kvdl = priv; + struct mlxsw_sp1_kvdl_part *part; + + part = mlxsw_sp1_kvdl_index_part(kvdl, entry_index); + if (IS_ERR(part)) + return; + mlxsw_sp1_kvdl_part_free(part, entry_index); +} + +static int mlxsw_sp1_kvdl_alloc_size_query(struct mlxsw_sp *mlxsw_sp, + void *priv, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, + unsigned int *p_alloc_size) +{ + struct mlxsw_sp1_kvdl *kvdl = priv; + struct mlxsw_sp1_kvdl_part *part; + + part = mlxsw_sp1_kvdl_alloc_size_part(kvdl, entry_count); + if (IS_ERR(part)) + return PTR_ERR(part); + + *p_alloc_size = part->info.alloc_size; + + return 0; +} + +static void mlxsw_sp1_kvdl_part_update(struct mlxsw_sp1_kvdl_part *part, + struct mlxsw_sp1_kvdl_part *part_prev, + unsigned int size) +{ + if (!part_prev) { + part->info.end_index = size - 1; + } else { + part->info.start_index = part_prev->info.end_index + 1; + part->info.end_index = part->info.start_index + size - 1; + } +} + +static struct mlxsw_sp1_kvdl_part * +mlxsw_sp1_kvdl_part_init(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp1_kvdl_part_info *info, + struct mlxsw_sp1_kvdl_part *part_prev) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + struct mlxsw_sp1_kvdl_part *part; + bool need_update = true; + unsigned int nr_entries; + u64 resource_size; + int err; + + err = devl_resource_size_get(devlink, info->resource_id, + &resource_size); + if (err) { + need_update = false; + resource_size = info->end_index - info->start_index + 1; + } + + nr_entries = div_u64(resource_size, info->alloc_size); + part = kzalloc(struct_size(part, usage, BITS_TO_LONGS(nr_entries)), + GFP_KERNEL); + if (!part) + return ERR_PTR(-ENOMEM); + + memcpy(&part->info, info, sizeof(part->info)); + + if (need_update) + mlxsw_sp1_kvdl_part_update(part, part_prev, resource_size); + return part; +} + +static void mlxsw_sp1_kvdl_part_fini(struct mlxsw_sp1_kvdl_part *part) +{ + kfree(part); +} + +static int mlxsw_sp1_kvdl_parts_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp1_kvdl *kvdl) +{ + const struct mlxsw_sp1_kvdl_part_info *info; + struct mlxsw_sp1_kvdl_part *part_prev = NULL; + int err, i; + + for (i = 0; i < MLXSW_SP1_KVDL_PARTS_INFO_LEN; i++) { + info = &mlxsw_sp1_kvdl_parts_info[i]; + kvdl->parts[i] = mlxsw_sp1_kvdl_part_init(mlxsw_sp, info, + part_prev); + if (IS_ERR(kvdl->parts[i])) { + err = PTR_ERR(kvdl->parts[i]); + goto err_kvdl_part_init; + } + part_prev = kvdl->parts[i]; + } + return 0; + +err_kvdl_part_init: + for (i--; i >= 0; i--) + mlxsw_sp1_kvdl_part_fini(kvdl->parts[i]); + return err; +} + +static void mlxsw_sp1_kvdl_parts_fini(struct mlxsw_sp1_kvdl *kvdl) +{ + int i; + + for (i = 0; i < MLXSW_SP1_KVDL_PARTS_INFO_LEN; i++) + mlxsw_sp1_kvdl_part_fini(kvdl->parts[i]); +} + +static u64 mlxsw_sp1_kvdl_part_occ(struct mlxsw_sp1_kvdl_part *part) +{ + const struct mlxsw_sp1_kvdl_part_info *info = &part->info; + unsigned int nr_entries; + int bit = -1; + u64 occ = 0; + + nr_entries = (info->end_index - + info->start_index + 1) / + info->alloc_size; + while ((bit = find_next_bit(part->usage, nr_entries, bit + 1)) + < nr_entries) + occ += info->alloc_size; + return occ; +} + +static u64 mlxsw_sp1_kvdl_occ_get(void *priv) +{ + const struct mlxsw_sp1_kvdl *kvdl = priv; + u64 occ = 0; + int i; + + for (i = 0; i < MLXSW_SP1_KVDL_PARTS_INFO_LEN; i++) + occ += mlxsw_sp1_kvdl_part_occ(kvdl->parts[i]); + + return occ; +} + +static u64 mlxsw_sp1_kvdl_single_occ_get(void *priv) +{ + const struct mlxsw_sp1_kvdl *kvdl = priv; + struct mlxsw_sp1_kvdl_part *part; + + part = kvdl->parts[MLXSW_SP1_KVDL_PART_ID_SINGLE]; + return mlxsw_sp1_kvdl_part_occ(part); +} + +static u64 mlxsw_sp1_kvdl_chunks_occ_get(void *priv) +{ + const struct mlxsw_sp1_kvdl *kvdl = priv; + struct mlxsw_sp1_kvdl_part *part; + + part = kvdl->parts[MLXSW_SP1_KVDL_PART_ID_CHUNKS]; + return mlxsw_sp1_kvdl_part_occ(part); +} + +static u64 mlxsw_sp1_kvdl_large_chunks_occ_get(void *priv) +{ + const struct mlxsw_sp1_kvdl *kvdl = priv; + struct mlxsw_sp1_kvdl_part *part; + + part = kvdl->parts[MLXSW_SP1_KVDL_PART_ID_LARGE_CHUNKS]; + return mlxsw_sp1_kvdl_part_occ(part); +} + +static int mlxsw_sp1_kvdl_init(struct mlxsw_sp *mlxsw_sp, void *priv) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + struct mlxsw_sp1_kvdl *kvdl = priv; + int err; + + err = mlxsw_sp1_kvdl_parts_init(mlxsw_sp, kvdl); + if (err) + return err; + devl_resource_occ_get_register(devlink, + MLXSW_SP_RESOURCE_KVD_LINEAR, + mlxsw_sp1_kvdl_occ_get, + kvdl); + devl_resource_occ_get_register(devlink, + MLXSW_SP_RESOURCE_KVD_LINEAR_SINGLE, + mlxsw_sp1_kvdl_single_occ_get, + kvdl); + devl_resource_occ_get_register(devlink, + MLXSW_SP_RESOURCE_KVD_LINEAR_CHUNKS, + mlxsw_sp1_kvdl_chunks_occ_get, + kvdl); + devl_resource_occ_get_register(devlink, + MLXSW_SP_RESOURCE_KVD_LINEAR_LARGE_CHUNKS, + mlxsw_sp1_kvdl_large_chunks_occ_get, + kvdl); + return 0; +} + +static void mlxsw_sp1_kvdl_fini(struct mlxsw_sp *mlxsw_sp, void *priv) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + struct mlxsw_sp1_kvdl *kvdl = priv; + + devl_resource_occ_get_unregister(devlink, + MLXSW_SP_RESOURCE_KVD_LINEAR_LARGE_CHUNKS); + devl_resource_occ_get_unregister(devlink, + MLXSW_SP_RESOURCE_KVD_LINEAR_CHUNKS); + devl_resource_occ_get_unregister(devlink, + MLXSW_SP_RESOURCE_KVD_LINEAR_SINGLE); + devl_resource_occ_get_unregister(devlink, + MLXSW_SP_RESOURCE_KVD_LINEAR); + mlxsw_sp1_kvdl_parts_fini(kvdl); +} + +const struct mlxsw_sp_kvdl_ops mlxsw_sp1_kvdl_ops = { + .priv_size = sizeof(struct mlxsw_sp1_kvdl), + .init = mlxsw_sp1_kvdl_init, + .fini = mlxsw_sp1_kvdl_fini, + .alloc = mlxsw_sp1_kvdl_alloc, + .free = mlxsw_sp1_kvdl_free, + .alloc_size_query = mlxsw_sp1_kvdl_alloc_size_query, +}; + +int mlxsw_sp1_kvdl_resources_register(struct mlxsw_core *mlxsw_core) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + static struct devlink_resource_size_params size_params; + u32 kvdl_max_size; + int err; + + kvdl_max_size = MLXSW_CORE_RES_GET(mlxsw_core, KVD_SIZE) - + MLXSW_CORE_RES_GET(mlxsw_core, KVD_SINGLE_MIN_SIZE) - + MLXSW_CORE_RES_GET(mlxsw_core, KVD_DOUBLE_MIN_SIZE); + + devlink_resource_size_params_init(&size_params, 0, kvdl_max_size, + MLXSW_SP1_KVDL_SINGLE_ALLOC_SIZE, + DEVLINK_RESOURCE_UNIT_ENTRY); + err = devl_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_KVD_LINEAR_SINGLES, + MLXSW_SP1_KVDL_SINGLE_SIZE, + MLXSW_SP_RESOURCE_KVD_LINEAR_SINGLE, + MLXSW_SP_RESOURCE_KVD_LINEAR, + &size_params); + if (err) + return err; + + devlink_resource_size_params_init(&size_params, 0, kvdl_max_size, + MLXSW_SP1_KVDL_CHUNKS_ALLOC_SIZE, + DEVLINK_RESOURCE_UNIT_ENTRY); + err = devl_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_KVD_LINEAR_CHUNKS, + MLXSW_SP1_KVDL_CHUNKS_SIZE, + MLXSW_SP_RESOURCE_KVD_LINEAR_CHUNKS, + MLXSW_SP_RESOURCE_KVD_LINEAR, + &size_params); + if (err) + return err; + + devlink_resource_size_params_init(&size_params, 0, kvdl_max_size, + MLXSW_SP1_KVDL_LARGE_CHUNKS_ALLOC_SIZE, + DEVLINK_RESOURCE_UNIT_ENTRY); + err = devl_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_KVD_LINEAR_LARGE_CHUNKS, + MLXSW_SP1_KVDL_LARGE_CHUNKS_SIZE, + MLXSW_SP_RESOURCE_KVD_LINEAR_LARGE_CHUNKS, + MLXSW_SP_RESOURCE_KVD_LINEAR, + &size_params); + return err; +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum1_mr_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum1_mr_tcam.c new file mode 100644 index 000000000..c8c675369 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum1_mr_tcam.c @@ -0,0 +1,342 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/parman.h> + +#include "reg.h" +#include "spectrum.h" +#include "core_acl_flex_actions.h" +#include "spectrum_mr.h" + +struct mlxsw_sp1_mr_tcam_region { + struct mlxsw_sp *mlxsw_sp; + enum mlxsw_reg_rtar_key_type rtar_key_type; + struct parman *parman; + struct parman_prio *parman_prios; +}; + +struct mlxsw_sp1_mr_tcam { + struct mlxsw_sp1_mr_tcam_region tcam_regions[MLXSW_SP_L3_PROTO_MAX]; +}; + +struct mlxsw_sp1_mr_tcam_route { + struct parman_item parman_item; + struct parman_prio *parman_prio; +}; + +static int mlxsw_sp1_mr_tcam_route_replace(struct mlxsw_sp *mlxsw_sp, + struct parman_item *parman_item, + struct mlxsw_sp_mr_route_key *key, + struct mlxsw_afa_block *afa_block) +{ + char rmft2_pl[MLXSW_REG_RMFT2_LEN]; + + switch (key->proto) { + case MLXSW_SP_L3_PROTO_IPV4: + mlxsw_reg_rmft2_ipv4_pack(rmft2_pl, true, parman_item->index, + key->vrid, + MLXSW_REG_RMFT2_IRIF_MASK_IGNORE, 0, + ntohl(key->group.addr4), + ntohl(key->group_mask.addr4), + ntohl(key->source.addr4), + ntohl(key->source_mask.addr4), + mlxsw_afa_block_first_set(afa_block)); + break; + case MLXSW_SP_L3_PROTO_IPV6: + mlxsw_reg_rmft2_ipv6_pack(rmft2_pl, true, parman_item->index, + key->vrid, + MLXSW_REG_RMFT2_IRIF_MASK_IGNORE, 0, + key->group.addr6, + key->group_mask.addr6, + key->source.addr6, + key->source_mask.addr6, + mlxsw_afa_block_first_set(afa_block)); + } + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rmft2), rmft2_pl); +} + +static int mlxsw_sp1_mr_tcam_route_remove(struct mlxsw_sp *mlxsw_sp, + struct parman_item *parman_item, + struct mlxsw_sp_mr_route_key *key) +{ + struct in6_addr zero_addr = IN6ADDR_ANY_INIT; + char rmft2_pl[MLXSW_REG_RMFT2_LEN]; + + switch (key->proto) { + case MLXSW_SP_L3_PROTO_IPV4: + mlxsw_reg_rmft2_ipv4_pack(rmft2_pl, false, parman_item->index, + key->vrid, 0, 0, 0, 0, 0, 0, NULL); + break; + case MLXSW_SP_L3_PROTO_IPV6: + mlxsw_reg_rmft2_ipv6_pack(rmft2_pl, false, parman_item->index, + key->vrid, 0, 0, zero_addr, zero_addr, + zero_addr, zero_addr, NULL); + break; + } + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rmft2), rmft2_pl); +} + +static struct mlxsw_sp1_mr_tcam_region * +mlxsw_sp1_mr_tcam_protocol_region(struct mlxsw_sp1_mr_tcam *mr_tcam, + enum mlxsw_sp_l3proto proto) +{ + return &mr_tcam->tcam_regions[proto]; +} + +static int +mlxsw_sp1_mr_tcam_route_parman_item_add(struct mlxsw_sp1_mr_tcam *mr_tcam, + struct mlxsw_sp1_mr_tcam_route *route, + struct mlxsw_sp_mr_route_key *key, + enum mlxsw_sp_mr_route_prio prio) +{ + struct mlxsw_sp1_mr_tcam_region *tcam_region; + int err; + + tcam_region = mlxsw_sp1_mr_tcam_protocol_region(mr_tcam, key->proto); + err = parman_item_add(tcam_region->parman, + &tcam_region->parman_prios[prio], + &route->parman_item); + if (err) + return err; + + route->parman_prio = &tcam_region->parman_prios[prio]; + return 0; +} + +static void +mlxsw_sp1_mr_tcam_route_parman_item_remove(struct mlxsw_sp1_mr_tcam *mr_tcam, + struct mlxsw_sp1_mr_tcam_route *route, + struct mlxsw_sp_mr_route_key *key) +{ + struct mlxsw_sp1_mr_tcam_region *tcam_region; + + tcam_region = mlxsw_sp1_mr_tcam_protocol_region(mr_tcam, key->proto); + parman_item_remove(tcam_region->parman, + route->parman_prio, &route->parman_item); +} + +static int +mlxsw_sp1_mr_tcam_route_create(struct mlxsw_sp *mlxsw_sp, void *priv, + void *route_priv, + struct mlxsw_sp_mr_route_key *key, + struct mlxsw_afa_block *afa_block, + enum mlxsw_sp_mr_route_prio prio) +{ + struct mlxsw_sp1_mr_tcam_route *route = route_priv; + struct mlxsw_sp1_mr_tcam *mr_tcam = priv; + int err; + + err = mlxsw_sp1_mr_tcam_route_parman_item_add(mr_tcam, route, + key, prio); + if (err) + return err; + + err = mlxsw_sp1_mr_tcam_route_replace(mlxsw_sp, &route->parman_item, + key, afa_block); + if (err) + goto err_route_replace; + return 0; + +err_route_replace: + mlxsw_sp1_mr_tcam_route_parman_item_remove(mr_tcam, route, key); + return err; +} + +static void +mlxsw_sp1_mr_tcam_route_destroy(struct mlxsw_sp *mlxsw_sp, void *priv, + void *route_priv, + struct mlxsw_sp_mr_route_key *key) +{ + struct mlxsw_sp1_mr_tcam_route *route = route_priv; + struct mlxsw_sp1_mr_tcam *mr_tcam = priv; + + mlxsw_sp1_mr_tcam_route_remove(mlxsw_sp, &route->parman_item, key); + mlxsw_sp1_mr_tcam_route_parman_item_remove(mr_tcam, route, key); +} + +static int +mlxsw_sp1_mr_tcam_route_update(struct mlxsw_sp *mlxsw_sp, + void *route_priv, + struct mlxsw_sp_mr_route_key *key, + struct mlxsw_afa_block *afa_block) +{ + struct mlxsw_sp1_mr_tcam_route *route = route_priv; + + return mlxsw_sp1_mr_tcam_route_replace(mlxsw_sp, &route->parman_item, + key, afa_block); +} + +#define MLXSW_SP1_MR_TCAM_REGION_BASE_COUNT 16 +#define MLXSW_SP1_MR_TCAM_REGION_RESIZE_STEP 16 + +static int +mlxsw_sp1_mr_tcam_region_alloc(struct mlxsw_sp1_mr_tcam_region *mr_tcam_region) +{ + struct mlxsw_sp *mlxsw_sp = mr_tcam_region->mlxsw_sp; + char rtar_pl[MLXSW_REG_RTAR_LEN]; + + mlxsw_reg_rtar_pack(rtar_pl, MLXSW_REG_RTAR_OP_ALLOCATE, + mr_tcam_region->rtar_key_type, + MLXSW_SP1_MR_TCAM_REGION_BASE_COUNT); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rtar), rtar_pl); +} + +static void +mlxsw_sp1_mr_tcam_region_free(struct mlxsw_sp1_mr_tcam_region *mr_tcam_region) +{ + struct mlxsw_sp *mlxsw_sp = mr_tcam_region->mlxsw_sp; + char rtar_pl[MLXSW_REG_RTAR_LEN]; + + mlxsw_reg_rtar_pack(rtar_pl, MLXSW_REG_RTAR_OP_DEALLOCATE, + mr_tcam_region->rtar_key_type, 0); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rtar), rtar_pl); +} + +static int mlxsw_sp1_mr_tcam_region_parman_resize(void *priv, + unsigned long new_count) +{ + struct mlxsw_sp1_mr_tcam_region *mr_tcam_region = priv; + struct mlxsw_sp *mlxsw_sp = mr_tcam_region->mlxsw_sp; + char rtar_pl[MLXSW_REG_RTAR_LEN]; + u64 max_tcam_rules; + + max_tcam_rules = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_TCAM_RULES); + if (new_count > max_tcam_rules) + return -EINVAL; + mlxsw_reg_rtar_pack(rtar_pl, MLXSW_REG_RTAR_OP_RESIZE, + mr_tcam_region->rtar_key_type, new_count); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rtar), rtar_pl); +} + +static void mlxsw_sp1_mr_tcam_region_parman_move(void *priv, + unsigned long from_index, + unsigned long to_index, + unsigned long count) +{ + struct mlxsw_sp1_mr_tcam_region *mr_tcam_region = priv; + struct mlxsw_sp *mlxsw_sp = mr_tcam_region->mlxsw_sp; + char rrcr_pl[MLXSW_REG_RRCR_LEN]; + + mlxsw_reg_rrcr_pack(rrcr_pl, MLXSW_REG_RRCR_OP_MOVE, + from_index, count, + mr_tcam_region->rtar_key_type, to_index); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rrcr), rrcr_pl); +} + +static const struct parman_ops mlxsw_sp1_mr_tcam_region_parman_ops = { + .base_count = MLXSW_SP1_MR_TCAM_REGION_BASE_COUNT, + .resize_step = MLXSW_SP1_MR_TCAM_REGION_RESIZE_STEP, + .resize = mlxsw_sp1_mr_tcam_region_parman_resize, + .move = mlxsw_sp1_mr_tcam_region_parman_move, + .algo = PARMAN_ALGO_TYPE_LSORT, +}; + +static int +mlxsw_sp1_mr_tcam_region_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp1_mr_tcam_region *mr_tcam_region, + enum mlxsw_reg_rtar_key_type rtar_key_type) +{ + struct parman_prio *parman_prios; + struct parman *parman; + int err; + int i; + + mr_tcam_region->rtar_key_type = rtar_key_type; + mr_tcam_region->mlxsw_sp = mlxsw_sp; + + err = mlxsw_sp1_mr_tcam_region_alloc(mr_tcam_region); + if (err) + return err; + + parman = parman_create(&mlxsw_sp1_mr_tcam_region_parman_ops, + mr_tcam_region); + if (!parman) { + err = -ENOMEM; + goto err_parman_create; + } + mr_tcam_region->parman = parman; + + parman_prios = kmalloc_array(MLXSW_SP_MR_ROUTE_PRIO_MAX + 1, + sizeof(*parman_prios), GFP_KERNEL); + if (!parman_prios) { + err = -ENOMEM; + goto err_parman_prios_alloc; + } + mr_tcam_region->parman_prios = parman_prios; + + for (i = 0; i < MLXSW_SP_MR_ROUTE_PRIO_MAX + 1; i++) + parman_prio_init(mr_tcam_region->parman, + &mr_tcam_region->parman_prios[i], i); + return 0; + +err_parman_prios_alloc: + parman_destroy(parman); +err_parman_create: + mlxsw_sp1_mr_tcam_region_free(mr_tcam_region); + return err; +} + +static void +mlxsw_sp1_mr_tcam_region_fini(struct mlxsw_sp1_mr_tcam_region *mr_tcam_region) +{ + int i; + + for (i = 0; i < MLXSW_SP_MR_ROUTE_PRIO_MAX + 1; i++) + parman_prio_fini(&mr_tcam_region->parman_prios[i]); + kfree(mr_tcam_region->parman_prios); + parman_destroy(mr_tcam_region->parman); + mlxsw_sp1_mr_tcam_region_free(mr_tcam_region); +} + +static int mlxsw_sp1_mr_tcam_init(struct mlxsw_sp *mlxsw_sp, void *priv) +{ + struct mlxsw_sp1_mr_tcam *mr_tcam = priv; + struct mlxsw_sp1_mr_tcam_region *region = &mr_tcam->tcam_regions[0]; + u32 rtar_key; + int err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, ACL_MAX_TCAM_RULES)) + return -EIO; + + rtar_key = MLXSW_REG_RTAR_KEY_TYPE_IPV4_MULTICAST; + err = mlxsw_sp1_mr_tcam_region_init(mlxsw_sp, + ®ion[MLXSW_SP_L3_PROTO_IPV4], + rtar_key); + if (err) + return err; + + rtar_key = MLXSW_REG_RTAR_KEY_TYPE_IPV6_MULTICAST; + err = mlxsw_sp1_mr_tcam_region_init(mlxsw_sp, + ®ion[MLXSW_SP_L3_PROTO_IPV6], + rtar_key); + if (err) + goto err_ipv6_region_init; + + return 0; + +err_ipv6_region_init: + mlxsw_sp1_mr_tcam_region_fini(®ion[MLXSW_SP_L3_PROTO_IPV4]); + return err; +} + +static void mlxsw_sp1_mr_tcam_fini(void *priv) +{ + struct mlxsw_sp1_mr_tcam *mr_tcam = priv; + struct mlxsw_sp1_mr_tcam_region *region = &mr_tcam->tcam_regions[0]; + + mlxsw_sp1_mr_tcam_region_fini(®ion[MLXSW_SP_L3_PROTO_IPV6]); + mlxsw_sp1_mr_tcam_region_fini(®ion[MLXSW_SP_L3_PROTO_IPV4]); +} + +const struct mlxsw_sp_mr_tcam_ops mlxsw_sp1_mr_tcam_ops = { + .priv_size = sizeof(struct mlxsw_sp1_mr_tcam), + .init = mlxsw_sp1_mr_tcam_init, + .fini = mlxsw_sp1_mr_tcam_fini, + .route_priv_size = sizeof(struct mlxsw_sp1_mr_tcam_route), + .route_create = mlxsw_sp1_mr_tcam_route_create, + .route_destroy = mlxsw_sp1_mr_tcam_route_destroy, + .route_update = mlxsw_sp1_mr_tcam_route_update, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum2_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum2_acl_tcam.c new file mode 100644 index 000000000..5b0210862 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum2_acl_tcam.c @@ -0,0 +1,280 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> + +#include "spectrum.h" +#include "spectrum_acl_tcam.h" +#include "core_acl_flex_actions.h" + +struct mlxsw_sp2_acl_tcam { + struct mlxsw_sp_acl_atcam atcam; + u32 kvdl_index; + unsigned int kvdl_count; +}; + +struct mlxsw_sp2_acl_tcam_region { + struct mlxsw_sp_acl_atcam_region aregion; + struct mlxsw_sp_acl_tcam_region *region; +}; + +struct mlxsw_sp2_acl_tcam_chunk { + struct mlxsw_sp_acl_atcam_chunk achunk; +}; + +struct mlxsw_sp2_acl_tcam_entry { + struct mlxsw_sp_acl_atcam_entry aentry; + struct mlxsw_afa_block *act_block; +}; + +static int +mlxsw_sp2_acl_ctcam_region_entry_insert(struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_entry *centry, + const char *mask) +{ + struct mlxsw_sp_acl_atcam_region *aregion; + struct mlxsw_sp_acl_atcam_entry *aentry; + struct mlxsw_sp_acl_erp_mask *erp_mask; + + aregion = mlxsw_sp_acl_tcam_cregion_aregion(cregion); + aentry = mlxsw_sp_acl_tcam_centry_aentry(centry); + + erp_mask = mlxsw_sp_acl_erp_mask_get(aregion, mask, true); + if (IS_ERR(erp_mask)) + return PTR_ERR(erp_mask); + aentry->erp_mask = erp_mask; + + return 0; +} + +static void +mlxsw_sp2_acl_ctcam_region_entry_remove(struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_entry *centry) +{ + struct mlxsw_sp_acl_atcam_region *aregion; + struct mlxsw_sp_acl_atcam_entry *aentry; + + aregion = mlxsw_sp_acl_tcam_cregion_aregion(cregion); + aentry = mlxsw_sp_acl_tcam_centry_aentry(centry); + + mlxsw_sp_acl_erp_mask_put(aregion, aentry->erp_mask); +} + +static const struct mlxsw_sp_acl_ctcam_region_ops +mlxsw_sp2_acl_ctcam_region_ops = { + .entry_insert = mlxsw_sp2_acl_ctcam_region_entry_insert, + .entry_remove = mlxsw_sp2_acl_ctcam_region_entry_remove, +}; + +static int mlxsw_sp2_acl_tcam_init(struct mlxsw_sp *mlxsw_sp, void *priv, + struct mlxsw_sp_acl_tcam *_tcam) +{ + struct mlxsw_sp2_acl_tcam *tcam = priv; + struct mlxsw_afa_block *afa_block; + char pefa_pl[MLXSW_REG_PEFA_LEN]; + char pgcr_pl[MLXSW_REG_PGCR_LEN]; + char *enc_actions; + int i; + int err; + + /* Some TCAM regions are not exposed to the host and used internally + * by the device. Allocate KVDL entries for the default actions of + * these regions to avoid the host from overwriting them. + */ + tcam->kvdl_count = _tcam->max_regions; + if (MLXSW_CORE_RES_VALID(mlxsw_sp->core, ACL_MAX_DEFAULT_ACTIONS)) + tcam->kvdl_count = MLXSW_CORE_RES_GET(mlxsw_sp->core, + ACL_MAX_DEFAULT_ACTIONS); + err = mlxsw_sp_kvdl_alloc(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ACTSET, + tcam->kvdl_count, &tcam->kvdl_index); + if (err) + return err; + + /* Create flex action block, set default action (continue) + * but don't commit. We need just the current set encoding + * to be written using PEFA register to all indexes for all regions. + */ + afa_block = mlxsw_afa_block_create(mlxsw_sp->afa); + if (IS_ERR(afa_block)) { + err = PTR_ERR(afa_block); + goto err_afa_block; + } + err = mlxsw_afa_block_continue(afa_block); + if (WARN_ON(err)) + goto err_afa_block_continue; + enc_actions = mlxsw_afa_block_cur_set(afa_block); + + /* Only write to KVDL entries used by TCAM regions exposed to the + * host. + */ + for (i = 0; i < _tcam->max_regions; i++) { + mlxsw_reg_pefa_pack(pefa_pl, tcam->kvdl_index + i, + true, enc_actions); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pefa), pefa_pl); + if (err) + goto err_pefa_write; + } + mlxsw_reg_pgcr_pack(pgcr_pl, tcam->kvdl_index); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pgcr), pgcr_pl); + if (err) + goto err_pgcr_write; + + err = mlxsw_sp_acl_atcam_init(mlxsw_sp, &tcam->atcam); + if (err) + goto err_atcam_init; + + mlxsw_afa_block_destroy(afa_block); + return 0; + +err_atcam_init: +err_pgcr_write: +err_pefa_write: +err_afa_block_continue: + mlxsw_afa_block_destroy(afa_block); +err_afa_block: + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ACTSET, + tcam->kvdl_count, tcam->kvdl_index); + return err; +} + +static void mlxsw_sp2_acl_tcam_fini(struct mlxsw_sp *mlxsw_sp, void *priv) +{ + struct mlxsw_sp2_acl_tcam *tcam = priv; + + mlxsw_sp_acl_atcam_fini(mlxsw_sp, &tcam->atcam); + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ACTSET, + tcam->kvdl_count, tcam->kvdl_index); +} + +static int +mlxsw_sp2_acl_tcam_region_init(struct mlxsw_sp *mlxsw_sp, void *region_priv, + void *tcam_priv, + struct mlxsw_sp_acl_tcam_region *_region, + void *hints_priv) +{ + struct mlxsw_sp2_acl_tcam_region *region = region_priv; + struct mlxsw_sp2_acl_tcam *tcam = tcam_priv; + + region->region = _region; + + return mlxsw_sp_acl_atcam_region_init(mlxsw_sp, &tcam->atcam, + ®ion->aregion, + _region, hints_priv, + &mlxsw_sp2_acl_ctcam_region_ops); +} + +static void +mlxsw_sp2_acl_tcam_region_fini(struct mlxsw_sp *mlxsw_sp, void *region_priv) +{ + struct mlxsw_sp2_acl_tcam_region *region = region_priv; + + mlxsw_sp_acl_atcam_region_fini(®ion->aregion); +} + +static int +mlxsw_sp2_acl_tcam_region_associate(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *region) +{ + return mlxsw_sp_acl_atcam_region_associate(mlxsw_sp, region->id); +} + +static void *mlxsw_sp2_acl_tcam_region_rehash_hints_get(void *region_priv) +{ + struct mlxsw_sp2_acl_tcam_region *region = region_priv; + + return mlxsw_sp_acl_atcam_rehash_hints_get(®ion->aregion); +} + +static void mlxsw_sp2_acl_tcam_region_rehash_hints_put(void *hints_priv) +{ + mlxsw_sp_acl_atcam_rehash_hints_put(hints_priv); +} + +static void mlxsw_sp2_acl_tcam_chunk_init(void *region_priv, void *chunk_priv, + unsigned int priority) +{ + struct mlxsw_sp2_acl_tcam_region *region = region_priv; + struct mlxsw_sp2_acl_tcam_chunk *chunk = chunk_priv; + + mlxsw_sp_acl_atcam_chunk_init(®ion->aregion, &chunk->achunk, + priority); +} + +static void mlxsw_sp2_acl_tcam_chunk_fini(void *chunk_priv) +{ + struct mlxsw_sp2_acl_tcam_chunk *chunk = chunk_priv; + + mlxsw_sp_acl_atcam_chunk_fini(&chunk->achunk); +} + +static int mlxsw_sp2_acl_tcam_entry_add(struct mlxsw_sp *mlxsw_sp, + void *region_priv, void *chunk_priv, + void *entry_priv, + struct mlxsw_sp_acl_rule_info *rulei) +{ + struct mlxsw_sp2_acl_tcam_region *region = region_priv; + struct mlxsw_sp2_acl_tcam_chunk *chunk = chunk_priv; + struct mlxsw_sp2_acl_tcam_entry *entry = entry_priv; + + entry->act_block = rulei->act_block; + return mlxsw_sp_acl_atcam_entry_add(mlxsw_sp, ®ion->aregion, + &chunk->achunk, &entry->aentry, + rulei); +} + +static void mlxsw_sp2_acl_tcam_entry_del(struct mlxsw_sp *mlxsw_sp, + void *region_priv, void *chunk_priv, + void *entry_priv) +{ + struct mlxsw_sp2_acl_tcam_region *region = region_priv; + struct mlxsw_sp2_acl_tcam_chunk *chunk = chunk_priv; + struct mlxsw_sp2_acl_tcam_entry *entry = entry_priv; + + mlxsw_sp_acl_atcam_entry_del(mlxsw_sp, ®ion->aregion, &chunk->achunk, + &entry->aentry); +} + +static int +mlxsw_sp2_acl_tcam_entry_action_replace(struct mlxsw_sp *mlxsw_sp, + void *region_priv, void *entry_priv, + struct mlxsw_sp_acl_rule_info *rulei) +{ + struct mlxsw_sp2_acl_tcam_region *region = region_priv; + struct mlxsw_sp2_acl_tcam_entry *entry = entry_priv; + + entry->act_block = rulei->act_block; + return mlxsw_sp_acl_atcam_entry_action_replace(mlxsw_sp, + ®ion->aregion, + &entry->aentry, rulei); +} + +static int +mlxsw_sp2_acl_tcam_entry_activity_get(struct mlxsw_sp *mlxsw_sp, + void *region_priv, void *entry_priv, + bool *activity) +{ + struct mlxsw_sp2_acl_tcam_entry *entry = entry_priv; + + return mlxsw_afa_block_activity_get(entry->act_block, activity); +} + +const struct mlxsw_sp_acl_tcam_ops mlxsw_sp2_acl_tcam_ops = { + .key_type = MLXSW_REG_PTAR_KEY_TYPE_FLEX2, + .priv_size = sizeof(struct mlxsw_sp2_acl_tcam), + .init = mlxsw_sp2_acl_tcam_init, + .fini = mlxsw_sp2_acl_tcam_fini, + .region_priv_size = sizeof(struct mlxsw_sp2_acl_tcam_region), + .region_init = mlxsw_sp2_acl_tcam_region_init, + .region_fini = mlxsw_sp2_acl_tcam_region_fini, + .region_associate = mlxsw_sp2_acl_tcam_region_associate, + .region_rehash_hints_get = mlxsw_sp2_acl_tcam_region_rehash_hints_get, + .region_rehash_hints_put = mlxsw_sp2_acl_tcam_region_rehash_hints_put, + .chunk_priv_size = sizeof(struct mlxsw_sp2_acl_tcam_chunk), + .chunk_init = mlxsw_sp2_acl_tcam_chunk_init, + .chunk_fini = mlxsw_sp2_acl_tcam_chunk_fini, + .entry_priv_size = sizeof(struct mlxsw_sp2_acl_tcam_entry), + .entry_add = mlxsw_sp2_acl_tcam_entry_add, + .entry_del = mlxsw_sp2_acl_tcam_entry_del, + .entry_action_replace = mlxsw_sp2_acl_tcam_entry_action_replace, + .entry_activity_get = mlxsw_sp2_acl_tcam_entry_activity_get, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum2_kvdl.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum2_kvdl.c new file mode 100644 index 000000000..24ff305a2 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum2_kvdl.c @@ -0,0 +1,273 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/bitops.h> + +#include "spectrum.h" +#include "core.h" +#include "reg.h" +#include "resources.h" + +struct mlxsw_sp2_kvdl_part_info { + u8 res_type; + /* For each defined partititon we need to know how many + * usage bits we need and how many indexes there are + * represented by a single bit. This could be got from FW + * querying appropriate resources. So have the resource + * ids for this purpose in partition definition. + */ + enum mlxsw_res_id usage_bit_count_res_id; + enum mlxsw_res_id index_range_res_id; +}; + +#define MLXSW_SP2_KVDL_PART_INFO(_entry_type, _res_type, \ + _usage_bit_count_res_id, _index_range_res_id) \ +[MLXSW_SP_KVDL_ENTRY_TYPE_##_entry_type] = { \ + .res_type = _res_type, \ + .usage_bit_count_res_id = MLXSW_RES_ID_##_usage_bit_count_res_id, \ + .index_range_res_id = MLXSW_RES_ID_##_index_range_res_id, \ +} + +static const struct mlxsw_sp2_kvdl_part_info mlxsw_sp2_kvdl_parts_info[] = { + MLXSW_SP2_KVDL_PART_INFO(ADJ, 0x21, KVD_SIZE, MAX_KVD_LINEAR_RANGE), + MLXSW_SP2_KVDL_PART_INFO(ACTSET, 0x23, MAX_KVD_ACTION_SETS, + MAX_KVD_ACTION_SETS), + MLXSW_SP2_KVDL_PART_INFO(PBS, 0x24, KVD_SIZE, KVD_SIZE), + MLXSW_SP2_KVDL_PART_INFO(MCRIGR, 0x26, KVD_SIZE, KVD_SIZE), + MLXSW_SP2_KVDL_PART_INFO(IPV6_ADDRESS, 0x28, KVD_SIZE, KVD_SIZE), + MLXSW_SP2_KVDL_PART_INFO(TNUMT, 0x29, KVD_SIZE, KVD_SIZE), +}; + +#define MLXSW_SP2_KVDL_PARTS_INFO_LEN ARRAY_SIZE(mlxsw_sp2_kvdl_parts_info) + +struct mlxsw_sp2_kvdl_part { + const struct mlxsw_sp2_kvdl_part_info *info; + unsigned int usage_bit_count; + unsigned int indexes_per_usage_bit; + unsigned int last_allocated_bit; + unsigned long usage[]; /* Usage bits */ +}; + +struct mlxsw_sp2_kvdl { + struct mlxsw_sp2_kvdl_part *parts[MLXSW_SP2_KVDL_PARTS_INFO_LEN]; +}; + +static int mlxsw_sp2_kvdl_part_find_zero_bits(struct mlxsw_sp2_kvdl_part *part, + unsigned int bit_count, + unsigned int *p_bit) +{ + unsigned int start_bit; + unsigned int bit; + unsigned int i; + bool wrap = false; + + start_bit = part->last_allocated_bit + 1; + if (start_bit == part->usage_bit_count) + start_bit = 0; + bit = start_bit; +again: + bit = find_next_zero_bit(part->usage, part->usage_bit_count, bit); + if (!wrap && bit + bit_count >= part->usage_bit_count) { + wrap = true; + bit = 0; + goto again; + } + if (wrap && bit + bit_count >= start_bit) + return -ENOBUFS; + for (i = 0; i < bit_count; i++) { + if (test_bit(bit + i, part->usage)) { + bit += bit_count; + goto again; + } + } + *p_bit = bit; + return 0; +} + +static int mlxsw_sp2_kvdl_part_alloc(struct mlxsw_sp2_kvdl_part *part, + unsigned int size, + u32 *p_kvdl_index) +{ + unsigned int bit_count; + unsigned int bit; + unsigned int i; + int err; + + bit_count = DIV_ROUND_UP(size, part->indexes_per_usage_bit); + err = mlxsw_sp2_kvdl_part_find_zero_bits(part, bit_count, &bit); + if (err) + return err; + for (i = 0; i < bit_count; i++) + __set_bit(bit + i, part->usage); + *p_kvdl_index = bit * part->indexes_per_usage_bit; + return 0; +} + +static int mlxsw_sp2_kvdl_rec_del(struct mlxsw_sp *mlxsw_sp, u8 res_type, + u16 size, u32 kvdl_index) +{ + char *iedr_pl; + int err; + + iedr_pl = kmalloc(MLXSW_REG_IEDR_LEN, GFP_KERNEL); + if (!iedr_pl) + return -ENOMEM; + + mlxsw_reg_iedr_pack(iedr_pl); + mlxsw_reg_iedr_rec_pack(iedr_pl, 0, res_type, size, kvdl_index); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(iedr), iedr_pl); + kfree(iedr_pl); + return err; +} + +static void mlxsw_sp2_kvdl_part_free(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp2_kvdl_part *part, + unsigned int size, u32 kvdl_index) +{ + unsigned int bit_count; + unsigned int bit; + unsigned int i; + int err; + + /* We need to ask FW to delete previously used KVD linear index */ + err = mlxsw_sp2_kvdl_rec_del(mlxsw_sp, part->info->res_type, + size, kvdl_index); + if (err) + return; + + bit_count = DIV_ROUND_UP(size, part->indexes_per_usage_bit); + bit = kvdl_index / part->indexes_per_usage_bit; + for (i = 0; i < bit_count; i++) + __clear_bit(bit + i, part->usage); +} + +static int mlxsw_sp2_kvdl_alloc(struct mlxsw_sp *mlxsw_sp, void *priv, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, + u32 *p_entry_index) +{ + unsigned int size = entry_count * mlxsw_sp_kvdl_entry_size(type); + struct mlxsw_sp2_kvdl *kvdl = priv; + struct mlxsw_sp2_kvdl_part *part = kvdl->parts[type]; + + return mlxsw_sp2_kvdl_part_alloc(part, size, p_entry_index); +} + +static void mlxsw_sp2_kvdl_free(struct mlxsw_sp *mlxsw_sp, void *priv, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, + int entry_index) +{ + unsigned int size = entry_count * mlxsw_sp_kvdl_entry_size(type); + struct mlxsw_sp2_kvdl *kvdl = priv; + struct mlxsw_sp2_kvdl_part *part = kvdl->parts[type]; + + return mlxsw_sp2_kvdl_part_free(mlxsw_sp, part, size, entry_index); +} + +static int mlxsw_sp2_kvdl_alloc_size_query(struct mlxsw_sp *mlxsw_sp, + void *priv, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, + unsigned int *p_alloc_count) +{ + *p_alloc_count = entry_count; + return 0; +} + +static struct mlxsw_sp2_kvdl_part * +mlxsw_sp2_kvdl_part_init(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp2_kvdl_part_info *info) +{ + unsigned int indexes_per_usage_bit; + struct mlxsw_sp2_kvdl_part *part; + unsigned int index_range; + unsigned int usage_bit_count; + size_t usage_size; + + if (!mlxsw_core_res_valid(mlxsw_sp->core, + info->usage_bit_count_res_id) || + !mlxsw_core_res_valid(mlxsw_sp->core, + info->index_range_res_id)) + return ERR_PTR(-EIO); + usage_bit_count = mlxsw_core_res_get(mlxsw_sp->core, + info->usage_bit_count_res_id); + index_range = mlxsw_core_res_get(mlxsw_sp->core, + info->index_range_res_id); + + /* For some partitions, one usage bit represents a group of indexes. + * That's why we compute the number of indexes per usage bit here, + * according to queried resources. + */ + indexes_per_usage_bit = index_range / usage_bit_count; + + usage_size = BITS_TO_LONGS(usage_bit_count) * sizeof(unsigned long); + part = kzalloc(sizeof(*part) + usage_size, GFP_KERNEL); + if (!part) + return ERR_PTR(-ENOMEM); + part->info = info; + part->usage_bit_count = usage_bit_count; + part->indexes_per_usage_bit = indexes_per_usage_bit; + part->last_allocated_bit = usage_bit_count - 1; + return part; +} + +static void mlxsw_sp2_kvdl_part_fini(struct mlxsw_sp2_kvdl_part *part) +{ + kfree(part); +} + +static int mlxsw_sp2_kvdl_parts_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp2_kvdl *kvdl) +{ + const struct mlxsw_sp2_kvdl_part_info *info; + int i; + int err; + + for (i = 0; i < MLXSW_SP2_KVDL_PARTS_INFO_LEN; i++) { + info = &mlxsw_sp2_kvdl_parts_info[i]; + kvdl->parts[i] = mlxsw_sp2_kvdl_part_init(mlxsw_sp, info); + if (IS_ERR(kvdl->parts[i])) { + err = PTR_ERR(kvdl->parts[i]); + goto err_kvdl_part_init; + } + } + return 0; + +err_kvdl_part_init: + for (i--; i >= 0; i--) + mlxsw_sp2_kvdl_part_fini(kvdl->parts[i]); + return err; +} + +static void mlxsw_sp2_kvdl_parts_fini(struct mlxsw_sp2_kvdl *kvdl) +{ + int i; + + for (i = 0; i < MLXSW_SP2_KVDL_PARTS_INFO_LEN; i++) + mlxsw_sp2_kvdl_part_fini(kvdl->parts[i]); +} + +static int mlxsw_sp2_kvdl_init(struct mlxsw_sp *mlxsw_sp, void *priv) +{ + struct mlxsw_sp2_kvdl *kvdl = priv; + + return mlxsw_sp2_kvdl_parts_init(mlxsw_sp, kvdl); +} + +static void mlxsw_sp2_kvdl_fini(struct mlxsw_sp *mlxsw_sp, void *priv) +{ + struct mlxsw_sp2_kvdl *kvdl = priv; + + mlxsw_sp2_kvdl_parts_fini(kvdl); +} + +const struct mlxsw_sp_kvdl_ops mlxsw_sp2_kvdl_ops = { + .priv_size = sizeof(struct mlxsw_sp2_kvdl), + .init = mlxsw_sp2_kvdl_init, + .fini = mlxsw_sp2_kvdl_fini, + .alloc = mlxsw_sp2_kvdl_alloc, + .free = mlxsw_sp2_kvdl_free, + .alloc_size_query = mlxsw_sp2_kvdl_alloc_size_query, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum2_mr_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum2_mr_tcam.c new file mode 100644 index 000000000..b1178b7a7 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum2_mr_tcam.c @@ -0,0 +1,330 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> + +#include "core_acl_flex_actions.h" +#include "spectrum.h" +#include "spectrum_mr.h" + +struct mlxsw_sp2_mr_tcam { + struct mlxsw_sp *mlxsw_sp; + struct mlxsw_sp_flow_block *flow_block; + struct mlxsw_sp_acl_ruleset *ruleset4; + struct mlxsw_sp_acl_ruleset *ruleset6; +}; + +struct mlxsw_sp2_mr_route { + struct mlxsw_sp2_mr_tcam *mr_tcam; +}; + +static struct mlxsw_sp_acl_ruleset * +mlxsw_sp2_mr_tcam_proto_ruleset(struct mlxsw_sp2_mr_tcam *mr_tcam, + enum mlxsw_sp_l3proto proto) +{ + switch (proto) { + case MLXSW_SP_L3_PROTO_IPV4: + return mr_tcam->ruleset4; + case MLXSW_SP_L3_PROTO_IPV6: + return mr_tcam->ruleset6; + } + return NULL; +} + +static int mlxsw_sp2_mr_tcam_bind_group(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_reg_pemrbt_protocol protocol, + struct mlxsw_sp_acl_ruleset *ruleset) +{ + char pemrbt_pl[MLXSW_REG_PEMRBT_LEN]; + u16 group_id; + + group_id = mlxsw_sp_acl_ruleset_group_id(ruleset); + + mlxsw_reg_pemrbt_pack(pemrbt_pl, protocol, group_id); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pemrbt), pemrbt_pl); +} + +static const enum mlxsw_afk_element mlxsw_sp2_mr_tcam_usage_ipv4[] = { + MLXSW_AFK_ELEMENT_VIRT_ROUTER_MSB, + MLXSW_AFK_ELEMENT_VIRT_ROUTER_LSB, + MLXSW_AFK_ELEMENT_SRC_IP_0_31, + MLXSW_AFK_ELEMENT_DST_IP_0_31, +}; + +static int mlxsw_sp2_mr_tcam_ipv4_init(struct mlxsw_sp2_mr_tcam *mr_tcam) +{ + struct mlxsw_afk_element_usage elusage; + int err; + + /* Initialize IPv4 ACL group. */ + mlxsw_afk_element_usage_fill(&elusage, + mlxsw_sp2_mr_tcam_usage_ipv4, + ARRAY_SIZE(mlxsw_sp2_mr_tcam_usage_ipv4)); + mr_tcam->ruleset4 = mlxsw_sp_acl_ruleset_get(mr_tcam->mlxsw_sp, + mr_tcam->flow_block, + MLXSW_SP_L3_PROTO_IPV4, + MLXSW_SP_ACL_PROFILE_MR, + &elusage); + + if (IS_ERR(mr_tcam->ruleset4)) + return PTR_ERR(mr_tcam->ruleset4); + + /* MC Router groups should be bound before routes are inserted. */ + err = mlxsw_sp2_mr_tcam_bind_group(mr_tcam->mlxsw_sp, + MLXSW_REG_PEMRBT_PROTO_IPV4, + mr_tcam->ruleset4); + if (err) + goto err_bind_group; + + return 0; + +err_bind_group: + mlxsw_sp_acl_ruleset_put(mr_tcam->mlxsw_sp, mr_tcam->ruleset4); + return err; +} + +static void mlxsw_sp2_mr_tcam_ipv4_fini(struct mlxsw_sp2_mr_tcam *mr_tcam) +{ + mlxsw_sp_acl_ruleset_put(mr_tcam->mlxsw_sp, mr_tcam->ruleset4); +} + +static const enum mlxsw_afk_element mlxsw_sp2_mr_tcam_usage_ipv6[] = { + MLXSW_AFK_ELEMENT_VIRT_ROUTER_MSB, + MLXSW_AFK_ELEMENT_VIRT_ROUTER_LSB, + MLXSW_AFK_ELEMENT_SRC_IP_96_127, + MLXSW_AFK_ELEMENT_SRC_IP_64_95, + MLXSW_AFK_ELEMENT_SRC_IP_32_63, + MLXSW_AFK_ELEMENT_SRC_IP_0_31, + MLXSW_AFK_ELEMENT_DST_IP_96_127, + MLXSW_AFK_ELEMENT_DST_IP_64_95, + MLXSW_AFK_ELEMENT_DST_IP_32_63, + MLXSW_AFK_ELEMENT_DST_IP_0_31, +}; + +static int mlxsw_sp2_mr_tcam_ipv6_init(struct mlxsw_sp2_mr_tcam *mr_tcam) +{ + struct mlxsw_afk_element_usage elusage; + int err; + + /* Initialize IPv6 ACL group */ + mlxsw_afk_element_usage_fill(&elusage, + mlxsw_sp2_mr_tcam_usage_ipv6, + ARRAY_SIZE(mlxsw_sp2_mr_tcam_usage_ipv6)); + mr_tcam->ruleset6 = mlxsw_sp_acl_ruleset_get(mr_tcam->mlxsw_sp, + mr_tcam->flow_block, + MLXSW_SP_L3_PROTO_IPV6, + MLXSW_SP_ACL_PROFILE_MR, + &elusage); + + if (IS_ERR(mr_tcam->ruleset6)) + return PTR_ERR(mr_tcam->ruleset6); + + /* MC Router groups should be bound before routes are inserted. */ + err = mlxsw_sp2_mr_tcam_bind_group(mr_tcam->mlxsw_sp, + MLXSW_REG_PEMRBT_PROTO_IPV6, + mr_tcam->ruleset6); + if (err) + goto err_bind_group; + + return 0; + +err_bind_group: + mlxsw_sp_acl_ruleset_put(mr_tcam->mlxsw_sp, mr_tcam->ruleset6); + return err; +} + +static void mlxsw_sp2_mr_tcam_ipv6_fini(struct mlxsw_sp2_mr_tcam *mr_tcam) +{ + mlxsw_sp_acl_ruleset_put(mr_tcam->mlxsw_sp, mr_tcam->ruleset6); +} + +static void +mlxsw_sp2_mr_tcam_rule_parse4(struct mlxsw_sp_acl_rule_info *rulei, + struct mlxsw_sp_mr_route_key *key) +{ + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_SRC_IP_0_31, + (char *) &key->source.addr4, + (char *) &key->source_mask.addr4, 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_DST_IP_0_31, + (char *) &key->group.addr4, + (char *) &key->group_mask.addr4, 4); +} + +static void +mlxsw_sp2_mr_tcam_rule_parse6(struct mlxsw_sp_acl_rule_info *rulei, + struct mlxsw_sp_mr_route_key *key) +{ + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_SRC_IP_96_127, + &key->source.addr6.s6_addr[0x0], + &key->source_mask.addr6.s6_addr[0x0], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_SRC_IP_64_95, + &key->source.addr6.s6_addr[0x4], + &key->source_mask.addr6.s6_addr[0x4], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_SRC_IP_32_63, + &key->source.addr6.s6_addr[0x8], + &key->source_mask.addr6.s6_addr[0x8], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_SRC_IP_0_31, + &key->source.addr6.s6_addr[0xc], + &key->source_mask.addr6.s6_addr[0xc], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_DST_IP_96_127, + &key->group.addr6.s6_addr[0x0], + &key->group_mask.addr6.s6_addr[0x0], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_DST_IP_64_95, + &key->group.addr6.s6_addr[0x4], + &key->group_mask.addr6.s6_addr[0x4], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_DST_IP_32_63, + &key->group.addr6.s6_addr[0x8], + &key->group_mask.addr6.s6_addr[0x8], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_DST_IP_0_31, + &key->group.addr6.s6_addr[0xc], + &key->group_mask.addr6.s6_addr[0xc], 4); +} + +static void +mlxsw_sp2_mr_tcam_rule_parse(struct mlxsw_sp_acl_rule *rule, + struct mlxsw_sp_mr_route_key *key, + unsigned int priority) +{ + struct mlxsw_sp_acl_rule_info *rulei; + + rulei = mlxsw_sp_acl_rule_rulei(rule); + rulei->priority = priority; + mlxsw_sp_acl_rulei_keymask_u32(rulei, MLXSW_AFK_ELEMENT_VIRT_ROUTER_LSB, + key->vrid, GENMASK(7, 0)); + mlxsw_sp_acl_rulei_keymask_u32(rulei, + MLXSW_AFK_ELEMENT_VIRT_ROUTER_MSB, + key->vrid >> 8, GENMASK(3, 0)); + switch (key->proto) { + case MLXSW_SP_L3_PROTO_IPV4: + return mlxsw_sp2_mr_tcam_rule_parse4(rulei, key); + case MLXSW_SP_L3_PROTO_IPV6: + return mlxsw_sp2_mr_tcam_rule_parse6(rulei, key); + } +} + +static int +mlxsw_sp2_mr_tcam_route_create(struct mlxsw_sp *mlxsw_sp, void *priv, + void *route_priv, + struct mlxsw_sp_mr_route_key *key, + struct mlxsw_afa_block *afa_block, + enum mlxsw_sp_mr_route_prio prio) +{ + struct mlxsw_sp2_mr_route *mr_route = route_priv; + struct mlxsw_sp2_mr_tcam *mr_tcam = priv; + struct mlxsw_sp_acl_ruleset *ruleset; + struct mlxsw_sp_acl_rule *rule; + int err; + + mr_route->mr_tcam = mr_tcam; + ruleset = mlxsw_sp2_mr_tcam_proto_ruleset(mr_tcam, key->proto); + if (WARN_ON(!ruleset)) + return -EINVAL; + + rule = mlxsw_sp_acl_rule_create(mlxsw_sp, ruleset, + (unsigned long) route_priv, afa_block, + NULL); + if (IS_ERR(rule)) + return PTR_ERR(rule); + + mlxsw_sp2_mr_tcam_rule_parse(rule, key, prio); + err = mlxsw_sp_acl_rule_add(mlxsw_sp, rule); + if (err) + goto err_rule_add; + + return 0; + +err_rule_add: + mlxsw_sp_acl_rule_destroy(mlxsw_sp, rule); + return err; +} + +static void +mlxsw_sp2_mr_tcam_route_destroy(struct mlxsw_sp *mlxsw_sp, void *priv, + void *route_priv, + struct mlxsw_sp_mr_route_key *key) +{ + struct mlxsw_sp2_mr_tcam *mr_tcam = priv; + struct mlxsw_sp_acl_ruleset *ruleset; + struct mlxsw_sp_acl_rule *rule; + + ruleset = mlxsw_sp2_mr_tcam_proto_ruleset(mr_tcam, key->proto); + if (WARN_ON(!ruleset)) + return; + + rule = mlxsw_sp_acl_rule_lookup(mlxsw_sp, ruleset, + (unsigned long) route_priv); + if (WARN_ON(!rule)) + return; + + mlxsw_sp_acl_rule_del(mlxsw_sp, rule); + mlxsw_sp_acl_rule_destroy(mlxsw_sp, rule); +} + +static int +mlxsw_sp2_mr_tcam_route_update(struct mlxsw_sp *mlxsw_sp, + void *route_priv, + struct mlxsw_sp_mr_route_key *key, + struct mlxsw_afa_block *afa_block) +{ + struct mlxsw_sp2_mr_route *mr_route = route_priv; + struct mlxsw_sp2_mr_tcam *mr_tcam = mr_route->mr_tcam; + struct mlxsw_sp_acl_ruleset *ruleset; + struct mlxsw_sp_acl_rule *rule; + + ruleset = mlxsw_sp2_mr_tcam_proto_ruleset(mr_tcam, key->proto); + if (WARN_ON(!ruleset)) + return -EINVAL; + + rule = mlxsw_sp_acl_rule_lookup(mlxsw_sp, ruleset, + (unsigned long) route_priv); + if (WARN_ON(!rule)) + return -EINVAL; + + return mlxsw_sp_acl_rule_action_replace(mlxsw_sp, rule, afa_block); +} + +static int mlxsw_sp2_mr_tcam_init(struct mlxsw_sp *mlxsw_sp, void *priv) +{ + struct mlxsw_sp2_mr_tcam *mr_tcam = priv; + int err; + + mr_tcam->mlxsw_sp = mlxsw_sp; + mr_tcam->flow_block = mlxsw_sp_flow_block_create(mlxsw_sp, NULL); + if (!mr_tcam->flow_block) + return -ENOMEM; + + err = mlxsw_sp2_mr_tcam_ipv4_init(mr_tcam); + if (err) + goto err_ipv4_init; + + err = mlxsw_sp2_mr_tcam_ipv6_init(mr_tcam); + if (err) + goto err_ipv6_init; + + return 0; + +err_ipv6_init: + mlxsw_sp2_mr_tcam_ipv4_fini(mr_tcam); +err_ipv4_init: + mlxsw_sp_flow_block_destroy(mr_tcam->flow_block); + return err; +} + +static void mlxsw_sp2_mr_tcam_fini(void *priv) +{ + struct mlxsw_sp2_mr_tcam *mr_tcam = priv; + + mlxsw_sp2_mr_tcam_ipv6_fini(mr_tcam); + mlxsw_sp2_mr_tcam_ipv4_fini(mr_tcam); + mlxsw_sp_flow_block_destroy(mr_tcam->flow_block); +} + +const struct mlxsw_sp_mr_tcam_ops mlxsw_sp2_mr_tcam_ops = { + .priv_size = sizeof(struct mlxsw_sp2_mr_tcam), + .init = mlxsw_sp2_mr_tcam_init, + .fini = mlxsw_sp2_mr_tcam_fini, + .route_priv_size = sizeof(struct mlxsw_sp2_mr_route), + .route_create = mlxsw_sp2_mr_tcam_route_create, + .route_destroy = mlxsw_sp2_mr_tcam_route_destroy, + .route_update = mlxsw_sp2_mr_tcam_route_update, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c new file mode 100644 index 000000000..6c5af0185 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c @@ -0,0 +1,1124 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/slab.h> +#include <linux/errno.h> +#include <linux/list.h> +#include <linux/string.h> +#include <linux/rhashtable.h> +#include <linux/netdevice.h> +#include <linux/mutex.h> +#include <net/net_namespace.h> +#include <net/tc_act/tc_vlan.h> + +#include "reg.h" +#include "core.h" +#include "resources.h" +#include "spectrum.h" +#include "core_acl_flex_keys.h" +#include "core_acl_flex_actions.h" +#include "spectrum_acl_tcam.h" + +struct mlxsw_sp_acl { + struct mlxsw_sp *mlxsw_sp; + struct mlxsw_afk *afk; + struct mlxsw_sp_fid *dummy_fid; + struct rhashtable ruleset_ht; + struct list_head rules; + struct mutex rules_lock; /* Protects rules list */ + struct { + struct delayed_work dw; + unsigned long interval; /* ms */ +#define MLXSW_SP_ACL_RULE_ACTIVITY_UPDATE_PERIOD_MS 1000 + } rule_activity_update; + struct mlxsw_sp_acl_tcam tcam; +}; + +struct mlxsw_afk *mlxsw_sp_acl_afk(struct mlxsw_sp_acl *acl) +{ + return acl->afk; +} + +struct mlxsw_sp_acl_ruleset_ht_key { + struct mlxsw_sp_flow_block *block; + u32 chain_index; + const struct mlxsw_sp_acl_profile_ops *ops; +}; + +struct mlxsw_sp_acl_ruleset { + struct rhash_head ht_node; /* Member of acl HT */ + struct mlxsw_sp_acl_ruleset_ht_key ht_key; + struct rhashtable rule_ht; + unsigned int ref_count; + unsigned int min_prio; + unsigned int max_prio; + unsigned long priv[]; + /* priv has to be always the last item */ +}; + +struct mlxsw_sp_acl_rule { + struct rhash_head ht_node; /* Member of rule HT */ + struct list_head list; + unsigned long cookie; /* HT key */ + struct mlxsw_sp_acl_ruleset *ruleset; + struct mlxsw_sp_acl_rule_info *rulei; + u64 last_used; + u64 last_packets; + u64 last_bytes; + u64 last_drops; + unsigned long priv[]; + /* priv has to be always the last item */ +}; + +static const struct rhashtable_params mlxsw_sp_acl_ruleset_ht_params = { + .key_len = sizeof(struct mlxsw_sp_acl_ruleset_ht_key), + .key_offset = offsetof(struct mlxsw_sp_acl_ruleset, ht_key), + .head_offset = offsetof(struct mlxsw_sp_acl_ruleset, ht_node), + .automatic_shrinking = true, +}; + +static const struct rhashtable_params mlxsw_sp_acl_rule_ht_params = { + .key_len = sizeof(unsigned long), + .key_offset = offsetof(struct mlxsw_sp_acl_rule, cookie), + .head_offset = offsetof(struct mlxsw_sp_acl_rule, ht_node), + .automatic_shrinking = true, +}; + +struct mlxsw_sp_fid *mlxsw_sp_acl_dummy_fid(struct mlxsw_sp *mlxsw_sp) +{ + return mlxsw_sp->acl->dummy_fid; +} + +static bool +mlxsw_sp_acl_ruleset_is_singular(const struct mlxsw_sp_acl_ruleset *ruleset) +{ + /* We hold a reference on ruleset ourselves */ + return ruleset->ref_count == 2; +} + +int mlxsw_sp_acl_ruleset_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_flow_block_binding *binding) +{ + struct mlxsw_sp_acl_ruleset *ruleset = block->ruleset_zero; + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + + return ops->ruleset_bind(mlxsw_sp, ruleset->priv, + binding->mlxsw_sp_port, binding->ingress); +} + +void mlxsw_sp_acl_ruleset_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_flow_block_binding *binding) +{ + struct mlxsw_sp_acl_ruleset *ruleset = block->ruleset_zero; + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + + ops->ruleset_unbind(mlxsw_sp, ruleset->priv, + binding->mlxsw_sp_port, binding->ingress); +} + +static int +mlxsw_sp_acl_ruleset_block_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset, + struct mlxsw_sp_flow_block *block) +{ + struct mlxsw_sp_flow_block_binding *binding; + int err; + + block->ruleset_zero = ruleset; + list_for_each_entry(binding, &block->binding_list, list) { + err = mlxsw_sp_acl_ruleset_bind(mlxsw_sp, block, binding); + if (err) + goto rollback; + } + return 0; + +rollback: + list_for_each_entry_continue_reverse(binding, &block->binding_list, + list) + mlxsw_sp_acl_ruleset_unbind(mlxsw_sp, block, binding); + block->ruleset_zero = NULL; + + return err; +} + +static void +mlxsw_sp_acl_ruleset_block_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset, + struct mlxsw_sp_flow_block *block) +{ + struct mlxsw_sp_flow_block_binding *binding; + + list_for_each_entry(binding, &block->binding_list, list) + mlxsw_sp_acl_ruleset_unbind(mlxsw_sp, block, binding); + block->ruleset_zero = NULL; +} + +static struct mlxsw_sp_acl_ruleset * +mlxsw_sp_acl_ruleset_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, u32 chain_index, + const struct mlxsw_sp_acl_profile_ops *ops, + struct mlxsw_afk_element_usage *tmplt_elusage) +{ + struct mlxsw_sp_acl *acl = mlxsw_sp->acl; + struct mlxsw_sp_acl_ruleset *ruleset; + size_t alloc_size; + int err; + + alloc_size = sizeof(*ruleset) + ops->ruleset_priv_size; + ruleset = kzalloc(alloc_size, GFP_KERNEL); + if (!ruleset) + return ERR_PTR(-ENOMEM); + ruleset->ref_count = 1; + ruleset->ht_key.block = block; + ruleset->ht_key.chain_index = chain_index; + ruleset->ht_key.ops = ops; + + err = rhashtable_init(&ruleset->rule_ht, &mlxsw_sp_acl_rule_ht_params); + if (err) + goto err_rhashtable_init; + + err = ops->ruleset_add(mlxsw_sp, &acl->tcam, ruleset->priv, + tmplt_elusage, &ruleset->min_prio, + &ruleset->max_prio); + if (err) + goto err_ops_ruleset_add; + + err = rhashtable_insert_fast(&acl->ruleset_ht, &ruleset->ht_node, + mlxsw_sp_acl_ruleset_ht_params); + if (err) + goto err_ht_insert; + + return ruleset; + +err_ht_insert: + ops->ruleset_del(mlxsw_sp, ruleset->priv); +err_ops_ruleset_add: + rhashtable_destroy(&ruleset->rule_ht); +err_rhashtable_init: + kfree(ruleset); + return ERR_PTR(err); +} + +static void mlxsw_sp_acl_ruleset_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset) +{ + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + struct mlxsw_sp_acl *acl = mlxsw_sp->acl; + + rhashtable_remove_fast(&acl->ruleset_ht, &ruleset->ht_node, + mlxsw_sp_acl_ruleset_ht_params); + ops->ruleset_del(mlxsw_sp, ruleset->priv); + rhashtable_destroy(&ruleset->rule_ht); + kfree(ruleset); +} + +static void mlxsw_sp_acl_ruleset_ref_inc(struct mlxsw_sp_acl_ruleset *ruleset) +{ + ruleset->ref_count++; +} + +static void mlxsw_sp_acl_ruleset_ref_dec(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset) +{ + if (--ruleset->ref_count) + return; + mlxsw_sp_acl_ruleset_destroy(mlxsw_sp, ruleset); +} + +static struct mlxsw_sp_acl_ruleset * +__mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp_acl *acl, + struct mlxsw_sp_flow_block *block, u32 chain_index, + const struct mlxsw_sp_acl_profile_ops *ops) +{ + struct mlxsw_sp_acl_ruleset_ht_key ht_key; + + memset(&ht_key, 0, sizeof(ht_key)); + ht_key.block = block; + ht_key.chain_index = chain_index; + ht_key.ops = ops; + return rhashtable_lookup_fast(&acl->ruleset_ht, &ht_key, + mlxsw_sp_acl_ruleset_ht_params); +} + +struct mlxsw_sp_acl_ruleset * +mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, u32 chain_index, + enum mlxsw_sp_acl_profile profile) +{ + const struct mlxsw_sp_acl_profile_ops *ops; + struct mlxsw_sp_acl *acl = mlxsw_sp->acl; + struct mlxsw_sp_acl_ruleset *ruleset; + + ops = mlxsw_sp_acl_tcam_profile_ops(mlxsw_sp, profile); + if (!ops) + return ERR_PTR(-EINVAL); + ruleset = __mlxsw_sp_acl_ruleset_lookup(acl, block, chain_index, ops); + if (!ruleset) + return ERR_PTR(-ENOENT); + return ruleset; +} + +struct mlxsw_sp_acl_ruleset * +mlxsw_sp_acl_ruleset_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, u32 chain_index, + enum mlxsw_sp_acl_profile profile, + struct mlxsw_afk_element_usage *tmplt_elusage) +{ + const struct mlxsw_sp_acl_profile_ops *ops; + struct mlxsw_sp_acl *acl = mlxsw_sp->acl; + struct mlxsw_sp_acl_ruleset *ruleset; + + ops = mlxsw_sp_acl_tcam_profile_ops(mlxsw_sp, profile); + if (!ops) + return ERR_PTR(-EINVAL); + + ruleset = __mlxsw_sp_acl_ruleset_lookup(acl, block, chain_index, ops); + if (ruleset) { + mlxsw_sp_acl_ruleset_ref_inc(ruleset); + return ruleset; + } + return mlxsw_sp_acl_ruleset_create(mlxsw_sp, block, chain_index, ops, + tmplt_elusage); +} + +void mlxsw_sp_acl_ruleset_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset) +{ + mlxsw_sp_acl_ruleset_ref_dec(mlxsw_sp, ruleset); +} + +u16 mlxsw_sp_acl_ruleset_group_id(struct mlxsw_sp_acl_ruleset *ruleset) +{ + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + + return ops->ruleset_group_id(ruleset->priv); +} + +void mlxsw_sp_acl_ruleset_prio_get(struct mlxsw_sp_acl_ruleset *ruleset, + unsigned int *p_min_prio, + unsigned int *p_max_prio) +{ + *p_min_prio = ruleset->min_prio; + *p_max_prio = ruleset->max_prio; +} + +struct mlxsw_sp_acl_rule_info * +mlxsw_sp_acl_rulei_create(struct mlxsw_sp_acl *acl, + struct mlxsw_afa_block *afa_block) +{ + struct mlxsw_sp_acl_rule_info *rulei; + int err; + + rulei = kzalloc(sizeof(*rulei), GFP_KERNEL); + if (!rulei) + return ERR_PTR(-ENOMEM); + + if (afa_block) { + rulei->act_block = afa_block; + return rulei; + } + + rulei->act_block = mlxsw_afa_block_create(acl->mlxsw_sp->afa); + if (IS_ERR(rulei->act_block)) { + err = PTR_ERR(rulei->act_block); + goto err_afa_block_create; + } + rulei->action_created = 1; + return rulei; + +err_afa_block_create: + kfree(rulei); + return ERR_PTR(err); +} + +void mlxsw_sp_acl_rulei_destroy(struct mlxsw_sp_acl_rule_info *rulei) +{ + if (rulei->action_created) + mlxsw_afa_block_destroy(rulei->act_block); + kfree(rulei); +} + +int mlxsw_sp_acl_rulei_commit(struct mlxsw_sp_acl_rule_info *rulei) +{ + return mlxsw_afa_block_commit(rulei->act_block); +} + +void mlxsw_sp_acl_rulei_priority(struct mlxsw_sp_acl_rule_info *rulei, + unsigned int priority) +{ + rulei->priority = priority; +} + +void mlxsw_sp_acl_rulei_keymask_u32(struct mlxsw_sp_acl_rule_info *rulei, + enum mlxsw_afk_element element, + u32 key_value, u32 mask_value) +{ + mlxsw_afk_values_add_u32(&rulei->values, element, + key_value, mask_value); +} + +void mlxsw_sp_acl_rulei_keymask_buf(struct mlxsw_sp_acl_rule_info *rulei, + enum mlxsw_afk_element element, + const char *key_value, + const char *mask_value, unsigned int len) +{ + mlxsw_afk_values_add_buf(&rulei->values, element, + key_value, mask_value, len); +} + +int mlxsw_sp_acl_rulei_act_continue(struct mlxsw_sp_acl_rule_info *rulei) +{ + return mlxsw_afa_block_continue(rulei->act_block); +} + +int mlxsw_sp_acl_rulei_act_jump(struct mlxsw_sp_acl_rule_info *rulei, + u16 group_id) +{ + return mlxsw_afa_block_jump(rulei->act_block, group_id); +} + +int mlxsw_sp_acl_rulei_act_terminate(struct mlxsw_sp_acl_rule_info *rulei) +{ + return mlxsw_afa_block_terminate(rulei->act_block); +} + +int mlxsw_sp_acl_rulei_act_drop(struct mlxsw_sp_acl_rule_info *rulei, + bool ingress, + const struct flow_action_cookie *fa_cookie, + struct netlink_ext_ack *extack) +{ + return mlxsw_afa_block_append_drop(rulei->act_block, ingress, + fa_cookie, extack); +} + +int mlxsw_sp_acl_rulei_act_trap(struct mlxsw_sp_acl_rule_info *rulei) +{ + return mlxsw_afa_block_append_trap(rulei->act_block, + MLXSW_TRAP_ID_ACL0); +} + +int mlxsw_sp_acl_rulei_act_fwd(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct net_device *out_dev, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port *mlxsw_sp_port; + u16 local_port; + bool in_port; + + if (out_dev) { + if (!mlxsw_sp_port_dev_check(out_dev)) { + NL_SET_ERR_MSG_MOD(extack, "Invalid output device"); + return -EINVAL; + } + mlxsw_sp_port = netdev_priv(out_dev); + if (mlxsw_sp_port->mlxsw_sp != mlxsw_sp) { + NL_SET_ERR_MSG_MOD(extack, "Invalid output device"); + return -EINVAL; + } + local_port = mlxsw_sp_port->local_port; + in_port = false; + } else { + /* If out_dev is NULL, the caller wants to + * set forward to ingress port. + */ + local_port = 0; + in_port = true; + } + return mlxsw_afa_block_append_fwd(rulei->act_block, + local_port, in_port, extack); +} + +int mlxsw_sp_acl_rulei_act_mirror(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct mlxsw_sp_flow_block *block, + struct net_device *out_dev, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_flow_block_binding *binding; + struct mlxsw_sp_port *in_port; + + if (!list_is_singular(&block->binding_list)) { + NL_SET_ERR_MSG_MOD(extack, "Only a single mirror source is allowed"); + return -EOPNOTSUPP; + } + binding = list_first_entry(&block->binding_list, + struct mlxsw_sp_flow_block_binding, list); + in_port = binding->mlxsw_sp_port; + + return mlxsw_afa_block_append_mirror(rulei->act_block, + in_port->local_port, + out_dev, + binding->ingress, + extack); +} + +int mlxsw_sp_acl_rulei_act_vlan(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + u32 action, u16 vid, u16 proto, u8 prio, + struct netlink_ext_ack *extack) +{ + u8 ethertype; + + if (action == FLOW_ACTION_VLAN_MANGLE) { + switch (proto) { + case ETH_P_8021Q: + ethertype = 0; + break; + case ETH_P_8021AD: + ethertype = 1; + break; + default: + NL_SET_ERR_MSG_MOD(extack, "Unsupported VLAN protocol"); + dev_err(mlxsw_sp->bus_info->dev, "Unsupported VLAN protocol %#04x\n", + proto); + return -EINVAL; + } + + return mlxsw_afa_block_append_vlan_modify(rulei->act_block, + vid, prio, ethertype, + extack); + } else { + NL_SET_ERR_MSG_MOD(extack, "Unsupported VLAN action"); + dev_err(mlxsw_sp->bus_info->dev, "Unsupported VLAN action\n"); + return -EINVAL; + } +} + +int mlxsw_sp_acl_rulei_act_priority(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + u32 prio, struct netlink_ext_ack *extack) +{ + /* Even though both Linux and Spectrum switches support 16 priorities, + * spectrum_qdisc only processes the first eight priomap elements, and + * the DCB and PFC features are tied to 8 priorities as well. Therefore + * bounce attempts to prioritize packets to higher priorities. + */ + if (prio >= IEEE_8021QAZ_MAX_TCS) { + NL_SET_ERR_MSG_MOD(extack, "Only priorities 0..7 are supported"); + return -EINVAL; + } + return mlxsw_afa_block_append_qos_switch_prio(rulei->act_block, prio, + extack); +} + +struct mlxsw_sp_acl_mangle_action { + enum flow_action_mangle_base htype; + /* Offset is u32-aligned. */ + u32 offset; + /* Mask bits are unset for the modified field. */ + u32 mask; + /* Shift required to extract the set value. */ + u32 shift; + enum mlxsw_sp_acl_mangle_field field; +}; + +#define MLXSW_SP_ACL_MANGLE_ACTION(_htype, _offset, _mask, _shift, _field) \ + { \ + .htype = _htype, \ + .offset = _offset, \ + .mask = _mask, \ + .shift = _shift, \ + .field = MLXSW_SP_ACL_MANGLE_FIELD_##_field, \ + } + +#define MLXSW_SP_ACL_MANGLE_ACTION_IP4(_offset, _mask, _shift, _field) \ + MLXSW_SP_ACL_MANGLE_ACTION(FLOW_ACT_MANGLE_HDR_TYPE_IP4, \ + _offset, _mask, _shift, _field) + +#define MLXSW_SP_ACL_MANGLE_ACTION_IP6(_offset, _mask, _shift, _field) \ + MLXSW_SP_ACL_MANGLE_ACTION(FLOW_ACT_MANGLE_HDR_TYPE_IP6, \ + _offset, _mask, _shift, _field) + +#define MLXSW_SP_ACL_MANGLE_ACTION_TCP(_offset, _mask, _shift, _field) \ + MLXSW_SP_ACL_MANGLE_ACTION(FLOW_ACT_MANGLE_HDR_TYPE_TCP, _offset, _mask, _shift, _field) + +#define MLXSW_SP_ACL_MANGLE_ACTION_UDP(_offset, _mask, _shift, _field) \ + MLXSW_SP_ACL_MANGLE_ACTION(FLOW_ACT_MANGLE_HDR_TYPE_UDP, _offset, _mask, _shift, _field) + +static struct mlxsw_sp_acl_mangle_action mlxsw_sp_acl_mangle_actions[] = { + MLXSW_SP_ACL_MANGLE_ACTION_IP4(0, 0xff00ffff, 16, IP_DSFIELD), + MLXSW_SP_ACL_MANGLE_ACTION_IP4(0, 0xff03ffff, 18, IP_DSCP), + MLXSW_SP_ACL_MANGLE_ACTION_IP4(0, 0xfffcffff, 16, IP_ECN), + + MLXSW_SP_ACL_MANGLE_ACTION_IP6(0, 0xf00fffff, 20, IP_DSFIELD), + MLXSW_SP_ACL_MANGLE_ACTION_IP6(0, 0xf03fffff, 22, IP_DSCP), + MLXSW_SP_ACL_MANGLE_ACTION_IP6(0, 0xffcfffff, 20, IP_ECN), + + MLXSW_SP_ACL_MANGLE_ACTION_TCP(0, 0x0000ffff, 16, IP_SPORT), + MLXSW_SP_ACL_MANGLE_ACTION_TCP(0, 0xffff0000, 0, IP_DPORT), + + MLXSW_SP_ACL_MANGLE_ACTION_UDP(0, 0x0000ffff, 16, IP_SPORT), + MLXSW_SP_ACL_MANGLE_ACTION_UDP(0, 0xffff0000, 0, IP_DPORT), + + MLXSW_SP_ACL_MANGLE_ACTION_IP4(12, 0x00000000, 0, IP4_SIP), + MLXSW_SP_ACL_MANGLE_ACTION_IP4(16, 0x00000000, 0, IP4_DIP), + + MLXSW_SP_ACL_MANGLE_ACTION_IP6(8, 0x00000000, 0, IP6_SIP_1), + MLXSW_SP_ACL_MANGLE_ACTION_IP6(12, 0x00000000, 0, IP6_SIP_2), + MLXSW_SP_ACL_MANGLE_ACTION_IP6(16, 0x00000000, 0, IP6_SIP_3), + MLXSW_SP_ACL_MANGLE_ACTION_IP6(20, 0x00000000, 0, IP6_SIP_4), + MLXSW_SP_ACL_MANGLE_ACTION_IP6(24, 0x00000000, 0, IP6_DIP_1), + MLXSW_SP_ACL_MANGLE_ACTION_IP6(28, 0x00000000, 0, IP6_DIP_2), + MLXSW_SP_ACL_MANGLE_ACTION_IP6(32, 0x00000000, 0, IP6_DIP_3), + MLXSW_SP_ACL_MANGLE_ACTION_IP6(36, 0x00000000, 0, IP6_DIP_4), +}; + +static int +mlxsw_sp_acl_rulei_act_mangle_field(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct mlxsw_sp_acl_mangle_action *mact, + u32 val, struct netlink_ext_ack *extack) +{ + switch (mact->field) { + case MLXSW_SP_ACL_MANGLE_FIELD_IP_DSFIELD: + return mlxsw_afa_block_append_qos_dsfield(rulei->act_block, + val, extack); + case MLXSW_SP_ACL_MANGLE_FIELD_IP_DSCP: + return mlxsw_afa_block_append_qos_dscp(rulei->act_block, + val, extack); + case MLXSW_SP_ACL_MANGLE_FIELD_IP_ECN: + return mlxsw_afa_block_append_qos_ecn(rulei->act_block, + val, extack); + default: + return -EOPNOTSUPP; + } +} + +static int mlxsw_sp1_acl_rulei_act_mangle_field(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct mlxsw_sp_acl_mangle_action *mact, + u32 val, struct netlink_ext_ack *extack) +{ + int err; + + err = mlxsw_sp_acl_rulei_act_mangle_field(mlxsw_sp, rulei, mact, val, extack); + if (err != -EOPNOTSUPP) + return err; + + NL_SET_ERR_MSG_MOD(extack, "Unsupported mangle field"); + return err; +} + +static int +mlxsw_sp2_acl_rulei_act_mangle_field_ip_odd(struct mlxsw_sp_acl_rule_info *rulei, + enum mlxsw_sp_acl_mangle_field field, + u32 val, struct netlink_ext_ack *extack) +{ + if (!rulei->ipv6_valid) { + rulei->ipv6.prev_val = val; + rulei->ipv6_valid = true; + rulei->ipv6.prev_field = field; + return 0; + } + + NL_SET_ERR_MSG_MOD(extack, "Unsupported mangle field order"); + return -EOPNOTSUPP; +} + +static int mlxsw_sp2_acl_rulei_act_mangle_field(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct mlxsw_sp_acl_mangle_action *mact, + u32 val, struct netlink_ext_ack *extack) +{ + int err; + + err = mlxsw_sp_acl_rulei_act_mangle_field(mlxsw_sp, rulei, mact, val, extack); + if (err != -EOPNOTSUPP) + return err; + + switch (mact->field) { + case MLXSW_SP_ACL_MANGLE_FIELD_IP_SPORT: + return mlxsw_afa_block_append_l4port(rulei->act_block, false, val, extack); + case MLXSW_SP_ACL_MANGLE_FIELD_IP_DPORT: + return mlxsw_afa_block_append_l4port(rulei->act_block, true, val, extack); + /* IPv4 fields */ + case MLXSW_SP_ACL_MANGLE_FIELD_IP4_SIP: + return mlxsw_afa_block_append_ip(rulei->act_block, false, + true, val, 0, extack); + case MLXSW_SP_ACL_MANGLE_FIELD_IP4_DIP: + return mlxsw_afa_block_append_ip(rulei->act_block, true, + true, val, 0, extack); + /* IPv6 fields */ + case MLXSW_SP_ACL_MANGLE_FIELD_IP6_SIP_1: + case MLXSW_SP_ACL_MANGLE_FIELD_IP6_SIP_3: + case MLXSW_SP_ACL_MANGLE_FIELD_IP6_DIP_1: + case MLXSW_SP_ACL_MANGLE_FIELD_IP6_DIP_3: + return mlxsw_sp2_acl_rulei_act_mangle_field_ip_odd(rulei, + mact->field, + val, extack); + case MLXSW_SP_ACL_MANGLE_FIELD_IP6_SIP_2: + if (rulei->ipv6_valid && + rulei->ipv6.prev_field == MLXSW_SP_ACL_MANGLE_FIELD_IP6_SIP_1) { + rulei->ipv6_valid = false; + return mlxsw_afa_block_append_ip(rulei->act_block, + false, false, val, + rulei->ipv6.prev_val, + extack); + } + break; + case MLXSW_SP_ACL_MANGLE_FIELD_IP6_SIP_4: + if (rulei->ipv6_valid && + rulei->ipv6.prev_field == MLXSW_SP_ACL_MANGLE_FIELD_IP6_SIP_3) { + rulei->ipv6_valid = false; + return mlxsw_afa_block_append_ip(rulei->act_block, + false, true, val, + rulei->ipv6.prev_val, + extack); + } + break; + case MLXSW_SP_ACL_MANGLE_FIELD_IP6_DIP_2: + if (rulei->ipv6_valid && + rulei->ipv6.prev_field == MLXSW_SP_ACL_MANGLE_FIELD_IP6_DIP_1) { + rulei->ipv6_valid = false; + return mlxsw_afa_block_append_ip(rulei->act_block, + true, false, val, + rulei->ipv6.prev_val, + extack); + } + break; + case MLXSW_SP_ACL_MANGLE_FIELD_IP6_DIP_4: + if (rulei->ipv6_valid && + rulei->ipv6.prev_field == MLXSW_SP_ACL_MANGLE_FIELD_IP6_DIP_3) { + rulei->ipv6_valid = false; + return mlxsw_afa_block_append_ip(rulei->act_block, + true, true, val, + rulei->ipv6.prev_val, + extack); + } + break; + default: + break; + } + + NL_SET_ERR_MSG_MOD(extack, "Unsupported mangle field"); + return err; +} + +int mlxsw_sp_acl_rulei_act_mangle(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + enum flow_action_mangle_base htype, + u32 offset, u32 mask, u32 val, + struct netlink_ext_ack *extack) +{ + const struct mlxsw_sp_acl_rulei_ops *acl_rulei_ops = mlxsw_sp->acl_rulei_ops; + struct mlxsw_sp_acl_mangle_action *mact; + size_t i; + + for (i = 0; i < ARRAY_SIZE(mlxsw_sp_acl_mangle_actions); ++i) { + mact = &mlxsw_sp_acl_mangle_actions[i]; + if (mact->htype == htype && + mact->offset == offset && + mact->mask == mask) { + val >>= mact->shift; + return acl_rulei_ops->act_mangle_field(mlxsw_sp, + rulei, mact, + val, extack); + } + } + + NL_SET_ERR_MSG_MOD(extack, "Unknown mangle field"); + return -EINVAL; +} + +int mlxsw_sp_acl_rulei_act_police(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + u32 index, u64 rate_bytes_ps, + u32 burst, struct netlink_ext_ack *extack) +{ + int err; + + err = mlxsw_afa_block_append_police(rulei->act_block, index, + rate_bytes_ps, burst, + &rulei->policer_index, extack); + if (err) + return err; + + rulei->policer_index_valid = true; + + return 0; +} + +int mlxsw_sp_acl_rulei_act_count(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct netlink_ext_ack *extack) +{ + int err; + + err = mlxsw_afa_block_append_counter(rulei->act_block, + &rulei->counter_index, extack); + if (err) + return err; + rulei->counter_valid = true; + return 0; +} + +int mlxsw_sp_acl_rulei_act_fid_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + u16 fid, struct netlink_ext_ack *extack) +{ + return mlxsw_afa_block_append_fid_set(rulei->act_block, fid, extack); +} + +int mlxsw_sp_acl_rulei_act_sample(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct mlxsw_sp_flow_block *block, + struct psample_group *psample_group, u32 rate, + u32 trunc_size, bool truncate, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_flow_block_binding *binding; + struct mlxsw_sp_port *mlxsw_sp_port; + + if (!list_is_singular(&block->binding_list)) { + NL_SET_ERR_MSG_MOD(extack, "Only a single sampling source is allowed"); + return -EOPNOTSUPP; + } + binding = list_first_entry(&block->binding_list, + struct mlxsw_sp_flow_block_binding, list); + mlxsw_sp_port = binding->mlxsw_sp_port; + + return mlxsw_afa_block_append_sampler(rulei->act_block, + mlxsw_sp_port->local_port, + psample_group, rate, trunc_size, + truncate, binding->ingress, + extack); +} + +struct mlxsw_sp_acl_rule * +mlxsw_sp_acl_rule_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset, + unsigned long cookie, + struct mlxsw_afa_block *afa_block, + struct netlink_ext_ack *extack) +{ + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + struct mlxsw_sp_acl_rule *rule; + int err; + + mlxsw_sp_acl_ruleset_ref_inc(ruleset); + rule = kzalloc(sizeof(*rule) + ops->rule_priv_size, + GFP_KERNEL); + if (!rule) { + err = -ENOMEM; + goto err_alloc; + } + rule->cookie = cookie; + rule->ruleset = ruleset; + + rule->rulei = mlxsw_sp_acl_rulei_create(mlxsw_sp->acl, afa_block); + if (IS_ERR(rule->rulei)) { + err = PTR_ERR(rule->rulei); + goto err_rulei_create; + } + + return rule; + +err_rulei_create: + kfree(rule); +err_alloc: + mlxsw_sp_acl_ruleset_ref_dec(mlxsw_sp, ruleset); + return ERR_PTR(err); +} + +void mlxsw_sp_acl_rule_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule *rule) +{ + struct mlxsw_sp_acl_ruleset *ruleset = rule->ruleset; + + mlxsw_sp_acl_rulei_destroy(rule->rulei); + kfree(rule); + mlxsw_sp_acl_ruleset_ref_dec(mlxsw_sp, ruleset); +} + +int mlxsw_sp_acl_rule_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule *rule) +{ + struct mlxsw_sp_acl_ruleset *ruleset = rule->ruleset; + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + struct mlxsw_sp_flow_block *block = ruleset->ht_key.block; + int err; + + err = ops->rule_add(mlxsw_sp, ruleset->priv, rule->priv, rule->rulei); + if (err) + return err; + + err = rhashtable_insert_fast(&ruleset->rule_ht, &rule->ht_node, + mlxsw_sp_acl_rule_ht_params); + if (err) + goto err_rhashtable_insert; + + if (!ruleset->ht_key.chain_index && + mlxsw_sp_acl_ruleset_is_singular(ruleset)) { + /* We only need ruleset with chain index 0, the implicit + * one, to be directly bound to device. The rest of the + * rulesets are bound by "Goto action set". + */ + err = mlxsw_sp_acl_ruleset_block_bind(mlxsw_sp, ruleset, block); + if (err) + goto err_ruleset_block_bind; + } + + mutex_lock(&mlxsw_sp->acl->rules_lock); + list_add_tail(&rule->list, &mlxsw_sp->acl->rules); + mutex_unlock(&mlxsw_sp->acl->rules_lock); + block->rule_count++; + block->ingress_blocker_rule_count += rule->rulei->ingress_bind_blocker; + block->egress_blocker_rule_count += rule->rulei->egress_bind_blocker; + return 0; + +err_ruleset_block_bind: + rhashtable_remove_fast(&ruleset->rule_ht, &rule->ht_node, + mlxsw_sp_acl_rule_ht_params); +err_rhashtable_insert: + ops->rule_del(mlxsw_sp, rule->priv); + return err; +} + +void mlxsw_sp_acl_rule_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule *rule) +{ + struct mlxsw_sp_acl_ruleset *ruleset = rule->ruleset; + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + struct mlxsw_sp_flow_block *block = ruleset->ht_key.block; + + block->egress_blocker_rule_count -= rule->rulei->egress_bind_blocker; + block->ingress_blocker_rule_count -= rule->rulei->ingress_bind_blocker; + block->rule_count--; + mutex_lock(&mlxsw_sp->acl->rules_lock); + list_del(&rule->list); + mutex_unlock(&mlxsw_sp->acl->rules_lock); + if (!ruleset->ht_key.chain_index && + mlxsw_sp_acl_ruleset_is_singular(ruleset)) + mlxsw_sp_acl_ruleset_block_unbind(mlxsw_sp, ruleset, block); + rhashtable_remove_fast(&ruleset->rule_ht, &rule->ht_node, + mlxsw_sp_acl_rule_ht_params); + ops->rule_del(mlxsw_sp, rule->priv); +} + +int mlxsw_sp_acl_rule_action_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule *rule, + struct mlxsw_afa_block *afa_block) +{ + struct mlxsw_sp_acl_ruleset *ruleset = rule->ruleset; + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + struct mlxsw_sp_acl_rule_info *rulei; + + rulei = mlxsw_sp_acl_rule_rulei(rule); + rulei->act_block = afa_block; + + return ops->rule_action_replace(mlxsw_sp, rule->priv, rule->rulei); +} + +struct mlxsw_sp_acl_rule * +mlxsw_sp_acl_rule_lookup(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset, + unsigned long cookie) +{ + return rhashtable_lookup_fast(&ruleset->rule_ht, &cookie, + mlxsw_sp_acl_rule_ht_params); +} + +struct mlxsw_sp_acl_rule_info * +mlxsw_sp_acl_rule_rulei(struct mlxsw_sp_acl_rule *rule) +{ + return rule->rulei; +} + +static int mlxsw_sp_acl_rule_activity_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule *rule) +{ + struct mlxsw_sp_acl_ruleset *ruleset = rule->ruleset; + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + bool active; + int err; + + err = ops->rule_activity_get(mlxsw_sp, rule->priv, &active); + if (err) + return err; + if (active) + rule->last_used = jiffies; + return 0; +} + +static int mlxsw_sp_acl_rules_activity_update(struct mlxsw_sp_acl *acl) +{ + struct mlxsw_sp_acl_rule *rule; + int err; + + mutex_lock(&acl->rules_lock); + list_for_each_entry(rule, &acl->rules, list) { + err = mlxsw_sp_acl_rule_activity_update(acl->mlxsw_sp, + rule); + if (err) + goto err_rule_update; + } + mutex_unlock(&acl->rules_lock); + return 0; + +err_rule_update: + mutex_unlock(&acl->rules_lock); + return err; +} + +static void mlxsw_sp_acl_rule_activity_work_schedule(struct mlxsw_sp_acl *acl) +{ + unsigned long interval = acl->rule_activity_update.interval; + + mlxsw_core_schedule_dw(&acl->rule_activity_update.dw, + msecs_to_jiffies(interval)); +} + +static void mlxsw_sp_acl_rule_activity_update_work(struct work_struct *work) +{ + struct mlxsw_sp_acl *acl = container_of(work, struct mlxsw_sp_acl, + rule_activity_update.dw.work); + int err; + + err = mlxsw_sp_acl_rules_activity_update(acl); + if (err) + dev_err(acl->mlxsw_sp->bus_info->dev, "Could not update acl activity"); + + mlxsw_sp_acl_rule_activity_work_schedule(acl); +} + +int mlxsw_sp_acl_rule_get_stats(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule *rule, + u64 *packets, u64 *bytes, u64 *drops, + u64 *last_use, + enum flow_action_hw_stats *used_hw_stats) + +{ + enum mlxsw_sp_policer_type type = MLXSW_SP_POLICER_TYPE_SINGLE_RATE; + struct mlxsw_sp_acl_rule_info *rulei; + u64 current_packets = 0; + u64 current_bytes = 0; + u64 current_drops = 0; + int err; + + rulei = mlxsw_sp_acl_rule_rulei(rule); + if (rulei->counter_valid) { + err = mlxsw_sp_flow_counter_get(mlxsw_sp, rulei->counter_index, + ¤t_packets, + ¤t_bytes); + if (err) + return err; + *used_hw_stats = FLOW_ACTION_HW_STATS_IMMEDIATE; + } + if (rulei->policer_index_valid) { + err = mlxsw_sp_policer_drops_counter_get(mlxsw_sp, type, + rulei->policer_index, + ¤t_drops); + if (err) + return err; + } + *packets = current_packets - rule->last_packets; + *bytes = current_bytes - rule->last_bytes; + *drops = current_drops - rule->last_drops; + *last_use = rule->last_used; + + rule->last_bytes = current_bytes; + rule->last_packets = current_packets; + rule->last_drops = current_drops; + + return 0; +} + +int mlxsw_sp_acl_init(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_fid *fid; + struct mlxsw_sp_acl *acl; + size_t alloc_size; + int err; + + alloc_size = sizeof(*acl) + mlxsw_sp_acl_tcam_priv_size(mlxsw_sp); + acl = kzalloc(alloc_size, GFP_KERNEL); + if (!acl) + return -ENOMEM; + mlxsw_sp->acl = acl; + acl->mlxsw_sp = mlxsw_sp; + acl->afk = mlxsw_afk_create(MLXSW_CORE_RES_GET(mlxsw_sp->core, + ACL_FLEX_KEYS), + mlxsw_sp->afk_ops); + if (!acl->afk) { + err = -ENOMEM; + goto err_afk_create; + } + + err = rhashtable_init(&acl->ruleset_ht, + &mlxsw_sp_acl_ruleset_ht_params); + if (err) + goto err_rhashtable_init; + + fid = mlxsw_sp_fid_dummy_get(mlxsw_sp); + if (IS_ERR(fid)) { + err = PTR_ERR(fid); + goto err_fid_get; + } + acl->dummy_fid = fid; + + INIT_LIST_HEAD(&acl->rules); + mutex_init(&acl->rules_lock); + err = mlxsw_sp_acl_tcam_init(mlxsw_sp, &acl->tcam); + if (err) + goto err_acl_ops_init; + + /* Create the delayed work for the rule activity_update */ + INIT_DELAYED_WORK(&acl->rule_activity_update.dw, + mlxsw_sp_acl_rule_activity_update_work); + acl->rule_activity_update.interval = MLXSW_SP_ACL_RULE_ACTIVITY_UPDATE_PERIOD_MS; + mlxsw_core_schedule_dw(&acl->rule_activity_update.dw, 0); + return 0; + +err_acl_ops_init: + mutex_destroy(&acl->rules_lock); + mlxsw_sp_fid_put(fid); +err_fid_get: + rhashtable_destroy(&acl->ruleset_ht); +err_rhashtable_init: + mlxsw_afk_destroy(acl->afk); +err_afk_create: + kfree(acl); + return err; +} + +void mlxsw_sp_acl_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_acl *acl = mlxsw_sp->acl; + + cancel_delayed_work_sync(&mlxsw_sp->acl->rule_activity_update.dw); + mlxsw_sp_acl_tcam_fini(mlxsw_sp, &acl->tcam); + mutex_destroy(&acl->rules_lock); + WARN_ON(!list_empty(&acl->rules)); + mlxsw_sp_fid_put(acl->dummy_fid); + rhashtable_destroy(&acl->ruleset_ht); + mlxsw_afk_destroy(acl->afk); + kfree(acl); +} + +u32 mlxsw_sp_acl_region_rehash_intrvl_get(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_acl *acl = mlxsw_sp->acl; + + return mlxsw_sp_acl_tcam_vregion_rehash_intrvl_get(mlxsw_sp, + &acl->tcam); +} + +int mlxsw_sp_acl_region_rehash_intrvl_set(struct mlxsw_sp *mlxsw_sp, u32 val) +{ + struct mlxsw_sp_acl *acl = mlxsw_sp->acl; + + return mlxsw_sp_acl_tcam_vregion_rehash_intrvl_set(mlxsw_sp, + &acl->tcam, val); +} + +struct mlxsw_sp_acl_rulei_ops mlxsw_sp1_acl_rulei_ops = { + .act_mangle_field = mlxsw_sp1_acl_rulei_act_mangle_field, +}; + +struct mlxsw_sp_acl_rulei_ops mlxsw_sp2_acl_rulei_ops = { + .act_mangle_field = mlxsw_sp2_acl_rulei_act_mangle_field, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_atcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_atcam.c new file mode 100644 index 000000000..4b713832f --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_atcam.c @@ -0,0 +1,646 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/err.h> +#include <linux/errno.h> +#include <linux/gfp.h> +#include <linux/refcount.h> +#include <linux/rhashtable.h> +#define CREATE_TRACE_POINTS +#include <trace/events/mlxsw.h> + +#include "reg.h" +#include "core.h" +#include "spectrum.h" +#include "spectrum_acl_tcam.h" +#include "core_acl_flex_keys.h" + +#define MLXSW_SP_ACL_ATCAM_LKEY_ID_BLOCK_CLEAR_START 0 +#define MLXSW_SP_ACL_ATCAM_LKEY_ID_BLOCK_CLEAR_END 5 + +struct mlxsw_sp_acl_atcam_lkey_id_ht_key { + char enc_key[MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN]; /* MSB blocks */ + u8 erp_id; +}; + +struct mlxsw_sp_acl_atcam_lkey_id { + struct rhash_head ht_node; + struct mlxsw_sp_acl_atcam_lkey_id_ht_key ht_key; + refcount_t refcnt; + u32 id; +}; + +struct mlxsw_sp_acl_atcam_region_ops { + int (*init)(struct mlxsw_sp_acl_atcam_region *aregion); + void (*fini)(struct mlxsw_sp_acl_atcam_region *aregion); + struct mlxsw_sp_acl_atcam_lkey_id * + (*lkey_id_get)(struct mlxsw_sp_acl_atcam_region *aregion, + char *enc_key, u8 erp_id); + void (*lkey_id_put)(struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_lkey_id *lkey_id); +}; + +struct mlxsw_sp_acl_atcam_region_generic { + struct mlxsw_sp_acl_atcam_lkey_id dummy_lkey_id; +}; + +struct mlxsw_sp_acl_atcam_region_12kb { + struct rhashtable lkey_ht; + unsigned int max_lkey_id; + unsigned long *used_lkey_id; +}; + +static const struct rhashtable_params mlxsw_sp_acl_atcam_lkey_id_ht_params = { + .key_len = sizeof(struct mlxsw_sp_acl_atcam_lkey_id_ht_key), + .key_offset = offsetof(struct mlxsw_sp_acl_atcam_lkey_id, ht_key), + .head_offset = offsetof(struct mlxsw_sp_acl_atcam_lkey_id, ht_node), +}; + +static const struct rhashtable_params mlxsw_sp_acl_atcam_entries_ht_params = { + .key_len = sizeof(struct mlxsw_sp_acl_atcam_entry_ht_key), + .key_offset = offsetof(struct mlxsw_sp_acl_atcam_entry, ht_key), + .head_offset = offsetof(struct mlxsw_sp_acl_atcam_entry, ht_node), +}; + +static bool +mlxsw_sp_acl_atcam_is_centry(const struct mlxsw_sp_acl_atcam_entry *aentry) +{ + return mlxsw_sp_acl_erp_mask_is_ctcam(aentry->erp_mask); +} + +static int +mlxsw_sp_acl_atcam_region_generic_init(struct mlxsw_sp_acl_atcam_region *aregion) +{ + struct mlxsw_sp_acl_atcam_region_generic *region_generic; + + region_generic = kzalloc(sizeof(*region_generic), GFP_KERNEL); + if (!region_generic) + return -ENOMEM; + + refcount_set(®ion_generic->dummy_lkey_id.refcnt, 1); + aregion->priv = region_generic; + + return 0; +} + +static void +mlxsw_sp_acl_atcam_region_generic_fini(struct mlxsw_sp_acl_atcam_region *aregion) +{ + kfree(aregion->priv); +} + +static struct mlxsw_sp_acl_atcam_lkey_id * +mlxsw_sp_acl_atcam_generic_lkey_id_get(struct mlxsw_sp_acl_atcam_region *aregion, + char *enc_key, u8 erp_id) +{ + struct mlxsw_sp_acl_atcam_region_generic *region_generic; + + region_generic = aregion->priv; + return ®ion_generic->dummy_lkey_id; +} + +static void +mlxsw_sp_acl_atcam_generic_lkey_id_put(struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_lkey_id *lkey_id) +{ +} + +static const struct mlxsw_sp_acl_atcam_region_ops +mlxsw_sp_acl_atcam_region_generic_ops = { + .init = mlxsw_sp_acl_atcam_region_generic_init, + .fini = mlxsw_sp_acl_atcam_region_generic_fini, + .lkey_id_get = mlxsw_sp_acl_atcam_generic_lkey_id_get, + .lkey_id_put = mlxsw_sp_acl_atcam_generic_lkey_id_put, +}; + +static int +mlxsw_sp_acl_atcam_region_12kb_init(struct mlxsw_sp_acl_atcam_region *aregion) +{ + struct mlxsw_sp *mlxsw_sp = aregion->region->mlxsw_sp; + struct mlxsw_sp_acl_atcam_region_12kb *region_12kb; + u64 max_lkey_id; + int err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, ACL_MAX_LARGE_KEY_ID)) + return -EIO; + + max_lkey_id = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_LARGE_KEY_ID); + region_12kb = kzalloc(sizeof(*region_12kb), GFP_KERNEL); + if (!region_12kb) + return -ENOMEM; + + region_12kb->used_lkey_id = bitmap_zalloc(max_lkey_id, GFP_KERNEL); + if (!region_12kb->used_lkey_id) { + err = -ENOMEM; + goto err_used_lkey_id_alloc; + } + + err = rhashtable_init(®ion_12kb->lkey_ht, + &mlxsw_sp_acl_atcam_lkey_id_ht_params); + if (err) + goto err_rhashtable_init; + + region_12kb->max_lkey_id = max_lkey_id; + aregion->priv = region_12kb; + + return 0; + +err_rhashtable_init: + bitmap_free(region_12kb->used_lkey_id); +err_used_lkey_id_alloc: + kfree(region_12kb); + return err; +} + +static void +mlxsw_sp_acl_atcam_region_12kb_fini(struct mlxsw_sp_acl_atcam_region *aregion) +{ + struct mlxsw_sp_acl_atcam_region_12kb *region_12kb = aregion->priv; + + rhashtable_destroy(®ion_12kb->lkey_ht); + bitmap_free(region_12kb->used_lkey_id); + kfree(region_12kb); +} + +static struct mlxsw_sp_acl_atcam_lkey_id * +mlxsw_sp_acl_atcam_lkey_id_create(struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_lkey_id_ht_key *ht_key) +{ + struct mlxsw_sp_acl_atcam_region_12kb *region_12kb = aregion->priv; + struct mlxsw_sp_acl_atcam_lkey_id *lkey_id; + u32 id; + int err; + + id = find_first_zero_bit(region_12kb->used_lkey_id, + region_12kb->max_lkey_id); + if (id < region_12kb->max_lkey_id) + __set_bit(id, region_12kb->used_lkey_id); + else + return ERR_PTR(-ENOBUFS); + + lkey_id = kzalloc(sizeof(*lkey_id), GFP_KERNEL); + if (!lkey_id) { + err = -ENOMEM; + goto err_lkey_id_alloc; + } + + lkey_id->id = id; + memcpy(&lkey_id->ht_key, ht_key, sizeof(*ht_key)); + refcount_set(&lkey_id->refcnt, 1); + + err = rhashtable_insert_fast(®ion_12kb->lkey_ht, + &lkey_id->ht_node, + mlxsw_sp_acl_atcam_lkey_id_ht_params); + if (err) + goto err_rhashtable_insert; + + return lkey_id; + +err_rhashtable_insert: + kfree(lkey_id); +err_lkey_id_alloc: + __clear_bit(id, region_12kb->used_lkey_id); + return ERR_PTR(err); +} + +static void +mlxsw_sp_acl_atcam_lkey_id_destroy(struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_lkey_id *lkey_id) +{ + struct mlxsw_sp_acl_atcam_region_12kb *region_12kb = aregion->priv; + u32 id = lkey_id->id; + + rhashtable_remove_fast(®ion_12kb->lkey_ht, &lkey_id->ht_node, + mlxsw_sp_acl_atcam_lkey_id_ht_params); + kfree(lkey_id); + __clear_bit(id, region_12kb->used_lkey_id); +} + +static struct mlxsw_sp_acl_atcam_lkey_id * +mlxsw_sp_acl_atcam_12kb_lkey_id_get(struct mlxsw_sp_acl_atcam_region *aregion, + char *enc_key, u8 erp_id) +{ + struct mlxsw_sp_acl_atcam_region_12kb *region_12kb = aregion->priv; + struct mlxsw_sp_acl_tcam_region *region = aregion->region; + struct mlxsw_sp_acl_atcam_lkey_id_ht_key ht_key = {{ 0 } }; + struct mlxsw_sp *mlxsw_sp = region->mlxsw_sp; + struct mlxsw_afk *afk = mlxsw_sp_acl_afk(mlxsw_sp->acl); + struct mlxsw_sp_acl_atcam_lkey_id *lkey_id; + + memcpy(ht_key.enc_key, enc_key, sizeof(ht_key.enc_key)); + mlxsw_afk_clear(afk, ht_key.enc_key, + MLXSW_SP_ACL_ATCAM_LKEY_ID_BLOCK_CLEAR_START, + MLXSW_SP_ACL_ATCAM_LKEY_ID_BLOCK_CLEAR_END); + ht_key.erp_id = erp_id; + lkey_id = rhashtable_lookup_fast(®ion_12kb->lkey_ht, &ht_key, + mlxsw_sp_acl_atcam_lkey_id_ht_params); + if (lkey_id) { + refcount_inc(&lkey_id->refcnt); + return lkey_id; + } + + return mlxsw_sp_acl_atcam_lkey_id_create(aregion, &ht_key); +} + +static void +mlxsw_sp_acl_atcam_12kb_lkey_id_put(struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_lkey_id *lkey_id) +{ + if (refcount_dec_and_test(&lkey_id->refcnt)) + mlxsw_sp_acl_atcam_lkey_id_destroy(aregion, lkey_id); +} + +static const struct mlxsw_sp_acl_atcam_region_ops +mlxsw_sp_acl_atcam_region_12kb_ops = { + .init = mlxsw_sp_acl_atcam_region_12kb_init, + .fini = mlxsw_sp_acl_atcam_region_12kb_fini, + .lkey_id_get = mlxsw_sp_acl_atcam_12kb_lkey_id_get, + .lkey_id_put = mlxsw_sp_acl_atcam_12kb_lkey_id_put, +}; + +static const struct mlxsw_sp_acl_atcam_region_ops * +mlxsw_sp_acl_atcam_region_ops_arr[] = { + [MLXSW_SP_ACL_ATCAM_REGION_TYPE_2KB] = + &mlxsw_sp_acl_atcam_region_generic_ops, + [MLXSW_SP_ACL_ATCAM_REGION_TYPE_4KB] = + &mlxsw_sp_acl_atcam_region_generic_ops, + [MLXSW_SP_ACL_ATCAM_REGION_TYPE_8KB] = + &mlxsw_sp_acl_atcam_region_generic_ops, + [MLXSW_SP_ACL_ATCAM_REGION_TYPE_12KB] = + &mlxsw_sp_acl_atcam_region_12kb_ops, +}; + +int mlxsw_sp_acl_atcam_region_associate(struct mlxsw_sp *mlxsw_sp, + u16 region_id) +{ + char perar_pl[MLXSW_REG_PERAR_LEN]; + /* For now, just assume that every region has 12 key blocks */ + u16 hw_region = region_id * 3; + u64 max_regions; + + max_regions = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_REGIONS); + if (hw_region >= max_regions) + return -ENOBUFS; + + mlxsw_reg_perar_pack(perar_pl, region_id, hw_region); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(perar), perar_pl); +} + +static void +mlxsw_sp_acl_atcam_region_type_init(struct mlxsw_sp_acl_atcam_region *aregion) +{ + struct mlxsw_sp_acl_tcam_region *region = aregion->region; + enum mlxsw_sp_acl_atcam_region_type region_type; + unsigned int blocks_count; + + /* We already know the blocks count can not exceed the maximum + * blocks count. + */ + blocks_count = mlxsw_afk_key_info_blocks_count_get(region->key_info); + if (blocks_count <= 2) + region_type = MLXSW_SP_ACL_ATCAM_REGION_TYPE_2KB; + else if (blocks_count <= 4) + region_type = MLXSW_SP_ACL_ATCAM_REGION_TYPE_4KB; + else if (blocks_count <= 8) + region_type = MLXSW_SP_ACL_ATCAM_REGION_TYPE_8KB; + else + region_type = MLXSW_SP_ACL_ATCAM_REGION_TYPE_12KB; + + aregion->type = region_type; + aregion->ops = mlxsw_sp_acl_atcam_region_ops_arr[region_type]; +} + +int +mlxsw_sp_acl_atcam_region_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam *atcam, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_tcam_region *region, + void *hints_priv, + const struct mlxsw_sp_acl_ctcam_region_ops *ops) +{ + int err; + + aregion->region = region; + aregion->atcam = atcam; + mlxsw_sp_acl_atcam_region_type_init(aregion); + INIT_LIST_HEAD(&aregion->entries_list); + + err = rhashtable_init(&aregion->entries_ht, + &mlxsw_sp_acl_atcam_entries_ht_params); + if (err) + return err; + err = aregion->ops->init(aregion); + if (err) + goto err_ops_init; + err = mlxsw_sp_acl_erp_region_init(aregion, hints_priv); + if (err) + goto err_erp_region_init; + err = mlxsw_sp_acl_ctcam_region_init(mlxsw_sp, &aregion->cregion, + region, ops); + if (err) + goto err_ctcam_region_init; + + return 0; + +err_ctcam_region_init: + mlxsw_sp_acl_erp_region_fini(aregion); +err_erp_region_init: + aregion->ops->fini(aregion); +err_ops_init: + rhashtable_destroy(&aregion->entries_ht); + return err; +} + +void mlxsw_sp_acl_atcam_region_fini(struct mlxsw_sp_acl_atcam_region *aregion) +{ + mlxsw_sp_acl_ctcam_region_fini(&aregion->cregion); + mlxsw_sp_acl_erp_region_fini(aregion); + aregion->ops->fini(aregion); + rhashtable_destroy(&aregion->entries_ht); + WARN_ON(!list_empty(&aregion->entries_list)); +} + +void mlxsw_sp_acl_atcam_chunk_init(struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_chunk *achunk, + unsigned int priority) +{ + mlxsw_sp_acl_ctcam_chunk_init(&aregion->cregion, &achunk->cchunk, + priority); +} + +void mlxsw_sp_acl_atcam_chunk_fini(struct mlxsw_sp_acl_atcam_chunk *achunk) +{ + mlxsw_sp_acl_ctcam_chunk_fini(&achunk->cchunk); +} + +static int +mlxsw_sp_acl_atcam_region_entry_insert(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry, + struct mlxsw_sp_acl_rule_info *rulei) +{ + struct mlxsw_sp_acl_tcam_region *region = aregion->region; + u8 erp_id = mlxsw_sp_acl_erp_mask_erp_id(aentry->erp_mask); + struct mlxsw_sp_acl_atcam_lkey_id *lkey_id; + char ptce3_pl[MLXSW_REG_PTCE3_LEN]; + u32 kvdl_index, priority; + int err; + + err = mlxsw_sp_acl_tcam_priority_get(mlxsw_sp, rulei, &priority, true); + if (err) + return err; + + lkey_id = aregion->ops->lkey_id_get(aregion, aentry->enc_key, erp_id); + if (IS_ERR(lkey_id)) + return PTR_ERR(lkey_id); + aentry->lkey_id = lkey_id; + + kvdl_index = mlxsw_afa_block_first_kvdl_index(rulei->act_block); + mlxsw_reg_ptce3_pack(ptce3_pl, true, MLXSW_REG_PTCE3_OP_WRITE_WRITE, + priority, region->tcam_region_info, + aentry->enc_key, erp_id, + aentry->delta_info.start, + aentry->delta_info.mask, + aentry->delta_info.value, + refcount_read(&lkey_id->refcnt) != 1, lkey_id->id, + kvdl_index); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptce3), ptce3_pl); + if (err) + goto err_ptce3_write; + + return 0; + +err_ptce3_write: + aregion->ops->lkey_id_put(aregion, lkey_id); + return err; +} + +static void +mlxsw_sp_acl_atcam_region_entry_remove(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry) +{ + struct mlxsw_sp_acl_atcam_lkey_id *lkey_id = aentry->lkey_id; + struct mlxsw_sp_acl_tcam_region *region = aregion->region; + u8 erp_id = mlxsw_sp_acl_erp_mask_erp_id(aentry->erp_mask); + char ptce3_pl[MLXSW_REG_PTCE3_LEN]; + + mlxsw_reg_ptce3_pack(ptce3_pl, false, MLXSW_REG_PTCE3_OP_WRITE_WRITE, 0, + region->tcam_region_info, + aentry->enc_key, erp_id, + aentry->delta_info.start, + aentry->delta_info.mask, + aentry->delta_info.value, + refcount_read(&lkey_id->refcnt) != 1, + lkey_id->id, 0); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptce3), ptce3_pl); + aregion->ops->lkey_id_put(aregion, lkey_id); +} + +static int +mlxsw_sp_acl_atcam_region_entry_action_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry, + struct mlxsw_sp_acl_rule_info *rulei) +{ + struct mlxsw_sp_acl_atcam_lkey_id *lkey_id = aentry->lkey_id; + u8 erp_id = mlxsw_sp_acl_erp_mask_erp_id(aentry->erp_mask); + struct mlxsw_sp_acl_tcam_region *region = aregion->region; + char ptce3_pl[MLXSW_REG_PTCE3_LEN]; + u32 kvdl_index, priority; + int err; + + err = mlxsw_sp_acl_tcam_priority_get(mlxsw_sp, rulei, &priority, true); + if (err) + return err; + kvdl_index = mlxsw_afa_block_first_kvdl_index(rulei->act_block); + mlxsw_reg_ptce3_pack(ptce3_pl, true, MLXSW_REG_PTCE3_OP_WRITE_UPDATE, + priority, region->tcam_region_info, + aentry->enc_key, erp_id, + aentry->delta_info.start, + aentry->delta_info.mask, + aentry->delta_info.value, + refcount_read(&lkey_id->refcnt) != 1, lkey_id->id, + kvdl_index); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptce3), ptce3_pl); +} + +static int +__mlxsw_sp_acl_atcam_entry_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry, + struct mlxsw_sp_acl_rule_info *rulei) +{ + struct mlxsw_sp_acl_tcam_region *region = aregion->region; + char mask[MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN] = { 0 }; + struct mlxsw_afk *afk = mlxsw_sp_acl_afk(mlxsw_sp->acl); + const struct mlxsw_sp_acl_erp_delta *delta; + struct mlxsw_sp_acl_erp_mask *erp_mask; + int err; + + mlxsw_afk_encode(afk, region->key_info, &rulei->values, + aentry->ht_key.full_enc_key, mask); + + erp_mask = mlxsw_sp_acl_erp_mask_get(aregion, mask, false); + if (IS_ERR(erp_mask)) + return PTR_ERR(erp_mask); + aentry->erp_mask = erp_mask; + aentry->ht_key.erp_id = mlxsw_sp_acl_erp_mask_erp_id(erp_mask); + memcpy(aentry->enc_key, aentry->ht_key.full_enc_key, + sizeof(aentry->enc_key)); + + /* Compute all needed delta information and clear the delta bits + * from the encrypted key. + */ + delta = mlxsw_sp_acl_erp_delta(aentry->erp_mask); + aentry->delta_info.start = mlxsw_sp_acl_erp_delta_start(delta); + aentry->delta_info.mask = mlxsw_sp_acl_erp_delta_mask(delta); + aentry->delta_info.value = + mlxsw_sp_acl_erp_delta_value(delta, + aentry->ht_key.full_enc_key); + mlxsw_sp_acl_erp_delta_clear(delta, aentry->enc_key); + + /* Add rule to the list of A-TCAM rules, assuming this + * rule is intended to A-TCAM. In case this rule does + * not fit into A-TCAM it will be removed from the list. + */ + list_add(&aentry->list, &aregion->entries_list); + + /* We can't insert identical rules into the A-TCAM, so fail and + * let the rule spill into C-TCAM + */ + err = rhashtable_lookup_insert_fast(&aregion->entries_ht, + &aentry->ht_node, + mlxsw_sp_acl_atcam_entries_ht_params); + if (err) + goto err_rhashtable_insert; + + /* Bloom filter must be updated here, before inserting the rule into + * the A-TCAM. + */ + err = mlxsw_sp_acl_erp_bf_insert(mlxsw_sp, aregion, erp_mask, aentry); + if (err) + goto err_bf_insert; + + err = mlxsw_sp_acl_atcam_region_entry_insert(mlxsw_sp, aregion, aentry, + rulei); + if (err) + goto err_rule_insert; + + return 0; + +err_rule_insert: + mlxsw_sp_acl_erp_bf_remove(mlxsw_sp, aregion, erp_mask, aentry); +err_bf_insert: + rhashtable_remove_fast(&aregion->entries_ht, &aentry->ht_node, + mlxsw_sp_acl_atcam_entries_ht_params); +err_rhashtable_insert: + list_del(&aentry->list); + mlxsw_sp_acl_erp_mask_put(aregion, erp_mask); + return err; +} + +static void +__mlxsw_sp_acl_atcam_entry_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry) +{ + mlxsw_sp_acl_atcam_region_entry_remove(mlxsw_sp, aregion, aentry); + mlxsw_sp_acl_erp_bf_remove(mlxsw_sp, aregion, aentry->erp_mask, aentry); + rhashtable_remove_fast(&aregion->entries_ht, &aentry->ht_node, + mlxsw_sp_acl_atcam_entries_ht_params); + list_del(&aentry->list); + mlxsw_sp_acl_erp_mask_put(aregion, aentry->erp_mask); +} + +static int +__mlxsw_sp_acl_atcam_entry_action_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry, + struct mlxsw_sp_acl_rule_info *rulei) +{ + return mlxsw_sp_acl_atcam_region_entry_action_replace(mlxsw_sp, aregion, + aentry, rulei); +} + +int mlxsw_sp_acl_atcam_entry_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_chunk *achunk, + struct mlxsw_sp_acl_atcam_entry *aentry, + struct mlxsw_sp_acl_rule_info *rulei) +{ + int err; + + err = __mlxsw_sp_acl_atcam_entry_add(mlxsw_sp, aregion, aentry, rulei); + if (!err) + return 0; + + /* It is possible we failed to add the rule to the A-TCAM due to + * exceeded number of masks. Try to spill into C-TCAM. + */ + trace_mlxsw_sp_acl_atcam_entry_add_ctcam_spill(mlxsw_sp, aregion); + err = mlxsw_sp_acl_ctcam_entry_add(mlxsw_sp, &aregion->cregion, + &achunk->cchunk, &aentry->centry, + rulei, true); + if (!err) + return 0; + + return err; +} + +void mlxsw_sp_acl_atcam_entry_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_chunk *achunk, + struct mlxsw_sp_acl_atcam_entry *aentry) +{ + if (mlxsw_sp_acl_atcam_is_centry(aentry)) + mlxsw_sp_acl_ctcam_entry_del(mlxsw_sp, &aregion->cregion, + &achunk->cchunk, &aentry->centry); + else + __mlxsw_sp_acl_atcam_entry_del(mlxsw_sp, aregion, aentry); +} + +int +mlxsw_sp_acl_atcam_entry_action_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry, + struct mlxsw_sp_acl_rule_info *rulei) +{ + int err; + + if (mlxsw_sp_acl_atcam_is_centry(aentry)) + err = mlxsw_sp_acl_ctcam_entry_action_replace(mlxsw_sp, + &aregion->cregion, + &aentry->centry, + rulei); + else + err = __mlxsw_sp_acl_atcam_entry_action_replace(mlxsw_sp, + aregion, aentry, + rulei); + + return err; +} + +int mlxsw_sp_acl_atcam_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam *atcam) +{ + return mlxsw_sp_acl_erps_init(mlxsw_sp, atcam); +} + +void mlxsw_sp_acl_atcam_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam *atcam) +{ + mlxsw_sp_acl_erps_fini(mlxsw_sp, atcam); +} + +void * +mlxsw_sp_acl_atcam_rehash_hints_get(struct mlxsw_sp_acl_atcam_region *aregion) +{ + return mlxsw_sp_acl_erp_rehash_hints_get(aregion); +} + +void mlxsw_sp_acl_atcam_rehash_hints_put(void *hints_priv) +{ + mlxsw_sp_acl_erp_rehash_hints_put(hints_priv); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_bloom_filter.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_bloom_filter.c new file mode 100644 index 000000000..95f63fcf4 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_bloom_filter.c @@ -0,0 +1,522 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#include <linux/errno.h> +#include <linux/gfp.h> +#include <linux/kernel.h> +#include <linux/refcount.h> +#include <linux/mutex.h> + +#include "spectrum.h" +#include "spectrum_acl_tcam.h" + +struct mlxsw_sp_acl_bf { + struct mutex lock; /* Protects Bloom Filter updates. */ + unsigned int bank_size; + refcount_t refcnt[]; +}; + +/* Bloom filter uses a crc-16 hash over chunks of data which contain 4 key + * blocks, eRP ID and region ID. In Spectrum-2 and above, region key is combined + * of up to 12 key blocks, so there can be up to 3 chunks in the Bloom filter + * key, depending on the actual number of key blocks used in the region. + * The layout of the Bloom filter key is as follows: + * + * +-------------------------+------------------------+------------------------+ + * | Chunk 2 Key blocks 11-8 | Chunk 1 Key blocks 7-4 | Chunk 0 Key blocks 3-0 | + * +-------------------------+------------------------+------------------------+ + */ +#define MLXSW_BLOOM_KEY_CHUNKS 3 + +/* Spectrum-2 and Spectrum-3 chunks */ +#define MLXSW_SP2_BLOOM_KEY_LEN 69 + +/* Each chunk size is 23 bytes. 18 bytes of it contain 4 key blocks, each is + * 36 bits, 2 bytes which hold eRP ID and region ID, and 3 bytes of zero + * padding. + * The layout of each chunk is as follows: + * + * +---------+----------------------+-----------------------------------+ + * | 3 bytes | 2 bytes | 18 bytes | + * +---------+-----------+----------+-----------------------------------+ + * | 183:158 | 157:148 | 147:144 | 143:0 | + * +---------+-----------+----------+-----------------------------------+ + * | 0 | region ID | eRP ID | 4 Key blocks (18 Bytes) | + * +---------+-----------+----------+-----------------------------------+ + */ +#define MLXSW_SP2_BLOOM_CHUNK_PAD_BYTES 3 +#define MLXSW_SP2_BLOOM_CHUNK_KEY_BYTES 18 +#define MLXSW_SP2_BLOOM_KEY_CHUNK_BYTES 23 + +/* The offset of the key block within a chunk is 5 bytes as it comes after + * 3 bytes of zero padding and 16 bits of region ID and eRP ID. + */ +#define MLXSW_SP2_BLOOM_CHUNK_KEY_OFFSET 5 + +/* This table is just the CRC of each possible byte which is used for + * Spectrum-{2-3}. It is computed, Msbit first, for the Bloom filter + * polynomial which is 0x8529 (1 + x^3 + x^5 + x^8 + x^10 + x^15 and + * the implicit x^16). + */ +static const u16 mlxsw_sp2_acl_bf_crc16_tab[256] = { +0x0000, 0x8529, 0x8f7b, 0x0a52, 0x9bdf, 0x1ef6, 0x14a4, 0x918d, +0xb297, 0x37be, 0x3dec, 0xb8c5, 0x2948, 0xac61, 0xa633, 0x231a, +0xe007, 0x652e, 0x6f7c, 0xea55, 0x7bd8, 0xfef1, 0xf4a3, 0x718a, +0x5290, 0xd7b9, 0xddeb, 0x58c2, 0xc94f, 0x4c66, 0x4634, 0xc31d, +0x4527, 0xc00e, 0xca5c, 0x4f75, 0xdef8, 0x5bd1, 0x5183, 0xd4aa, +0xf7b0, 0x7299, 0x78cb, 0xfde2, 0x6c6f, 0xe946, 0xe314, 0x663d, +0xa520, 0x2009, 0x2a5b, 0xaf72, 0x3eff, 0xbbd6, 0xb184, 0x34ad, +0x17b7, 0x929e, 0x98cc, 0x1de5, 0x8c68, 0x0941, 0x0313, 0x863a, +0x8a4e, 0x0f67, 0x0535, 0x801c, 0x1191, 0x94b8, 0x9eea, 0x1bc3, +0x38d9, 0xbdf0, 0xb7a2, 0x328b, 0xa306, 0x262f, 0x2c7d, 0xa954, +0x6a49, 0xef60, 0xe532, 0x601b, 0xf196, 0x74bf, 0x7eed, 0xfbc4, +0xd8de, 0x5df7, 0x57a5, 0xd28c, 0x4301, 0xc628, 0xcc7a, 0x4953, +0xcf69, 0x4a40, 0x4012, 0xc53b, 0x54b6, 0xd19f, 0xdbcd, 0x5ee4, +0x7dfe, 0xf8d7, 0xf285, 0x77ac, 0xe621, 0x6308, 0x695a, 0xec73, +0x2f6e, 0xaa47, 0xa015, 0x253c, 0xb4b1, 0x3198, 0x3bca, 0xbee3, +0x9df9, 0x18d0, 0x1282, 0x97ab, 0x0626, 0x830f, 0x895d, 0x0c74, +0x91b5, 0x149c, 0x1ece, 0x9be7, 0x0a6a, 0x8f43, 0x8511, 0x0038, +0x2322, 0xa60b, 0xac59, 0x2970, 0xb8fd, 0x3dd4, 0x3786, 0xb2af, +0x71b2, 0xf49b, 0xfec9, 0x7be0, 0xea6d, 0x6f44, 0x6516, 0xe03f, +0xc325, 0x460c, 0x4c5e, 0xc977, 0x58fa, 0xddd3, 0xd781, 0x52a8, +0xd492, 0x51bb, 0x5be9, 0xdec0, 0x4f4d, 0xca64, 0xc036, 0x451f, +0x6605, 0xe32c, 0xe97e, 0x6c57, 0xfdda, 0x78f3, 0x72a1, 0xf788, +0x3495, 0xb1bc, 0xbbee, 0x3ec7, 0xaf4a, 0x2a63, 0x2031, 0xa518, +0x8602, 0x032b, 0x0979, 0x8c50, 0x1ddd, 0x98f4, 0x92a6, 0x178f, +0x1bfb, 0x9ed2, 0x9480, 0x11a9, 0x8024, 0x050d, 0x0f5f, 0x8a76, +0xa96c, 0x2c45, 0x2617, 0xa33e, 0x32b3, 0xb79a, 0xbdc8, 0x38e1, +0xfbfc, 0x7ed5, 0x7487, 0xf1ae, 0x6023, 0xe50a, 0xef58, 0x6a71, +0x496b, 0xcc42, 0xc610, 0x4339, 0xd2b4, 0x579d, 0x5dcf, 0xd8e6, +0x5edc, 0xdbf5, 0xd1a7, 0x548e, 0xc503, 0x402a, 0x4a78, 0xcf51, +0xec4b, 0x6962, 0x6330, 0xe619, 0x7794, 0xf2bd, 0xf8ef, 0x7dc6, +0xbedb, 0x3bf2, 0x31a0, 0xb489, 0x2504, 0xa02d, 0xaa7f, 0x2f56, +0x0c4c, 0x8965, 0x8337, 0x061e, 0x9793, 0x12ba, 0x18e8, 0x9dc1, +}; + +/* Spectrum-4 chunks */ +#define MLXSW_SP4_BLOOM_KEY_LEN 60 + +/* In Spectrum-4, there is no padding. Each chunk size is 20 bytes. + * 18 bytes of it contain 4 key blocks, each is 36 bits, and 2 bytes which hold + * eRP ID and region ID. + * The layout of each chunk is as follows: + * + * +----------------------+-----------------------------------+ + * | 2 bytes | 18 bytes | + * +-----------+----------+-----------------------------------+ + * | 157:148 | 147:144 | 143:0 | + * +---------+-----------+----------+-------------------------+ + * | region ID | eRP ID | 4 Key blocks (18 Bytes) | + * +-----------+----------+-----------------------------------+ + */ + +#define MLXSW_SP4_BLOOM_CHUNK_PAD_BYTES 0 +#define MLXSW_SP4_BLOOM_CHUNK_KEY_BYTES 18 +#define MLXSW_SP4_BLOOM_KEY_CHUNK_BYTES 20 + +/* The offset of the key block within a chunk is 2 bytes as it comes after + * 16 bits of region ID and eRP ID. + */ +#define MLXSW_SP4_BLOOM_CHUNK_KEY_OFFSET 2 + +/* For Spectrum-4, two hash functions are used, CRC-10 and CRC-6 based. + * The result is combination of the two calculations - + * 6 bit column are MSB (result of CRC-6), + * 10 bit row are LSB (result of CRC-10). + */ + +/* This table is just the CRC of each possible byte which is used for + * Spectrum-4. It is computed, Msbit first, for the Bloom filter + * polynomial which is 0x1b (1 + x^1 + x^3 + x^4 and the implicit x^10). + */ +static const u16 mlxsw_sp4_acl_bf_crc10_tab[256] = { +0x0000, 0x001b, 0x0036, 0x002d, 0x006c, 0x0077, 0x005a, 0x0041, +0x00d8, 0x00c3, 0x00ee, 0x00f5, 0x00b4, 0x00af, 0x0082, 0x0099, +0x01b0, 0x01ab, 0x0186, 0x019d, 0x01dc, 0x01c7, 0x01ea, 0x01f1, +0x0168, 0x0173, 0x015e, 0x0145, 0x0104, 0x011f, 0x0132, 0x0129, +0x0360, 0x037b, 0x0356, 0x034d, 0x030c, 0x0317, 0x033a, 0x0321, +0x03b8, 0x03a3, 0x038e, 0x0395, 0x03d4, 0x03cf, 0x03e2, 0x03f9, +0x02d0, 0x02cb, 0x02e6, 0x02fd, 0x02bc, 0x02a7, 0x028a, 0x0291, +0x0208, 0x0213, 0x023e, 0x0225, 0x0264, 0x027f, 0x0252, 0x0249, +0x02db, 0x02c0, 0x02ed, 0x02f6, 0x02b7, 0x02ac, 0x0281, 0x029a, +0x0203, 0x0218, 0x0235, 0x022e, 0x026f, 0x0274, 0x0259, 0x0242, +0x036b, 0x0370, 0x035d, 0x0346, 0x0307, 0x031c, 0x0331, 0x032a, +0x03b3, 0x03a8, 0x0385, 0x039e, 0x03df, 0x03c4, 0x03e9, 0x03f2, +0x01bb, 0x01a0, 0x018d, 0x0196, 0x01d7, 0x01cc, 0x01e1, 0x01fa, +0x0163, 0x0178, 0x0155, 0x014e, 0x010f, 0x0114, 0x0139, 0x0122, +0x000b, 0x0010, 0x003d, 0x0026, 0x0067, 0x007c, 0x0051, 0x004a, +0x00d3, 0x00c8, 0x00e5, 0x00fe, 0x00bf, 0x00a4, 0x0089, 0x0092, +0x01ad, 0x01b6, 0x019b, 0x0180, 0x01c1, 0x01da, 0x01f7, 0x01ec, +0x0175, 0x016e, 0x0143, 0x0158, 0x0119, 0x0102, 0x012f, 0x0134, +0x001d, 0x0006, 0x002b, 0x0030, 0x0071, 0x006a, 0x0047, 0x005c, +0x00c5, 0x00de, 0x00f3, 0x00e8, 0x00a9, 0x00b2, 0x009f, 0x0084, +0x02cd, 0x02d6, 0x02fb, 0x02e0, 0x02a1, 0x02ba, 0x0297, 0x028c, +0x0215, 0x020e, 0x0223, 0x0238, 0x0279, 0x0262, 0x024f, 0x0254, +0x037d, 0x0366, 0x034b, 0x0350, 0x0311, 0x030a, 0x0327, 0x033c, +0x03a5, 0x03be, 0x0393, 0x0388, 0x03c9, 0x03d2, 0x03ff, 0x03e4, +0x0376, 0x036d, 0x0340, 0x035b, 0x031a, 0x0301, 0x032c, 0x0337, +0x03ae, 0x03b5, 0x0398, 0x0383, 0x03c2, 0x03d9, 0x03f4, 0x03ef, +0x02c6, 0x02dd, 0x02f0, 0x02eb, 0x02aa, 0x02b1, 0x029c, 0x0287, +0x021e, 0x0205, 0x0228, 0x0233, 0x0272, 0x0269, 0x0244, 0x025f, +0x0016, 0x000d, 0x0020, 0x003b, 0x007a, 0x0061, 0x004c, 0x0057, +0x00ce, 0x00d5, 0x00f8, 0x00e3, 0x00a2, 0x00b9, 0x0094, 0x008f, +0x01a6, 0x01bd, 0x0190, 0x018b, 0x01ca, 0x01d1, 0x01fc, 0x01e7, +0x017e, 0x0165, 0x0148, 0x0153, 0x0112, 0x0109, 0x0124, 0x013f, +}; + +/* This table is just the CRC of each possible byte which is used for + * Spectrum-4. It is computed, Msbit first, for the Bloom filter + * polynomial which is 0x2d (1 + x^2+ x^3 + x^5 and the implicit x^6). + */ +static const u8 mlxsw_sp4_acl_bf_crc6_tab[256] = { +0x00, 0x2d, 0x37, 0x1a, 0x03, 0x2e, 0x34, 0x19, +0x06, 0x2b, 0x31, 0x1c, 0x05, 0x28, 0x32, 0x1f, +0x0c, 0x21, 0x3b, 0x16, 0x0f, 0x22, 0x38, 0x15, +0x0a, 0x27, 0x3d, 0x10, 0x09, 0x24, 0x3e, 0x13, +0x18, 0x35, 0x2f, 0x02, 0x1b, 0x36, 0x2c, 0x01, +0x1e, 0x33, 0x29, 0x04, 0x1d, 0x30, 0x2a, 0x07, +0x14, 0x39, 0x23, 0x0e, 0x17, 0x3a, 0x20, 0x0d, +0x12, 0x3f, 0x25, 0x08, 0x11, 0x3c, 0x26, 0x0b, +0x30, 0x1d, 0x07, 0x2a, 0x33, 0x1e, 0x04, 0x29, +0x36, 0x1b, 0x01, 0x2c, 0x35, 0x18, 0x02, 0x2f, +0x3c, 0x11, 0x0b, 0x26, 0x3f, 0x12, 0x08, 0x25, +0x3a, 0x17, 0x0d, 0x20, 0x39, 0x14, 0x0e, 0x23, +0x28, 0x05, 0x1f, 0x32, 0x2b, 0x06, 0x1c, 0x31, +0x2e, 0x03, 0x19, 0x34, 0x2d, 0x00, 0x1a, 0x37, +0x24, 0x09, 0x13, 0x3e, 0x27, 0x0a, 0x10, 0x3d, +0x22, 0x0f, 0x15, 0x38, 0x21, 0x0c, 0x16, 0x3b, +0x0d, 0x20, 0x3a, 0x17, 0x0e, 0x23, 0x39, 0x14, +0x0b, 0x26, 0x3c, 0x11, 0x08, 0x25, 0x3f, 0x12, +0x01, 0x2c, 0x36, 0x1b, 0x02, 0x2f, 0x35, 0x18, +0x07, 0x2a, 0x30, 0x1d, 0x04, 0x29, 0x33, 0x1e, +0x15, 0x38, 0x22, 0x0f, 0x16, 0x3b, 0x21, 0x0c, +0x13, 0x3e, 0x24, 0x09, 0x10, 0x3d, 0x27, 0x0a, +0x19, 0x34, 0x2e, 0x03, 0x1a, 0x37, 0x2d, 0x00, +0x1f, 0x32, 0x28, 0x05, 0x1c, 0x31, 0x2b, 0x06, +0x3d, 0x10, 0x0a, 0x27, 0x3e, 0x13, 0x09, 0x24, +0x3b, 0x16, 0x0c, 0x21, 0x38, 0x15, 0x0f, 0x22, +0x31, 0x1c, 0x06, 0x2b, 0x32, 0x1f, 0x05, 0x28, +0x37, 0x1a, 0x00, 0x2d, 0x34, 0x19, 0x03, 0x2e, +0x25, 0x08, 0x12, 0x3f, 0x26, 0x0b, 0x11, 0x3c, +0x23, 0x0e, 0x14, 0x39, 0x20, 0x0d, 0x17, 0x3a, +0x29, 0x04, 0x1e, 0x33, 0x2a, 0x07, 0x1d, 0x30, +0x2f, 0x02, 0x18, 0x35, 0x2c, 0x01, 0x1b, 0x36, +}; + +/* Each chunk contains 4 key blocks. Chunk 2 uses key blocks 11-8, + * and we need to populate it with 4 key blocks copied from the entry encoded + * key. The original keys layout is same for Spectrum-{2,3,4}. + * Since the encoded key contains a 2 bytes padding, key block 11 starts at + * offset 2. block 7 that is used in chunk 1 starts at offset 20 as 4 key blocks + * take 18 bytes. See 'MLXSW_SP2_AFK_BLOCK_LAYOUT' for more details. + * This array defines key offsets for easy access when copying key blocks from + * entry key to Bloom filter chunk. + */ +static const u8 chunk_key_offsets[MLXSW_BLOOM_KEY_CHUNKS] = {2, 20, 38}; + +static u16 mlxsw_sp2_acl_bf_crc16_byte(u16 crc, u8 c) +{ + return (crc << 8) ^ mlxsw_sp2_acl_bf_crc16_tab[(crc >> 8) ^ c]; +} + +static u16 mlxsw_sp2_acl_bf_crc(const u8 *buffer, size_t len) +{ + u16 crc = 0; + + while (len--) + crc = mlxsw_sp2_acl_bf_crc16_byte(crc, *buffer++); + return crc; +} + +static void +__mlxsw_sp_acl_bf_key_encode(struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry, + char *output, u8 *len, u8 max_chunks, u8 pad_bytes, + u8 key_offset, u8 chunk_key_len, u8 chunk_len) +{ + struct mlxsw_afk_key_info *key_info = aregion->region->key_info; + u8 chunk_index, chunk_count, block_count; + char *chunk = output; + __be16 erp_region_id; + + block_count = mlxsw_afk_key_info_blocks_count_get(key_info); + chunk_count = 1 + ((block_count - 1) >> 2); + erp_region_id = cpu_to_be16(aentry->ht_key.erp_id | + (aregion->region->id << 4)); + for (chunk_index = max_chunks - chunk_count; chunk_index < max_chunks; + chunk_index++) { + memset(chunk, 0, pad_bytes); + memcpy(chunk + pad_bytes, &erp_region_id, + sizeof(erp_region_id)); + memcpy(chunk + key_offset, + &aentry->enc_key[chunk_key_offsets[chunk_index]], + chunk_key_len); + chunk += chunk_len; + } + *len = chunk_count * chunk_len; +} + +static void +mlxsw_sp2_acl_bf_key_encode(struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry, + char *output, u8 *len) +{ + __mlxsw_sp_acl_bf_key_encode(aregion, aentry, output, len, + MLXSW_BLOOM_KEY_CHUNKS, + MLXSW_SP2_BLOOM_CHUNK_PAD_BYTES, + MLXSW_SP2_BLOOM_CHUNK_KEY_OFFSET, + MLXSW_SP2_BLOOM_CHUNK_KEY_BYTES, + MLXSW_SP2_BLOOM_KEY_CHUNK_BYTES); +} + +static unsigned int +mlxsw_sp2_acl_bf_index_get(struct mlxsw_sp_acl_bf *bf, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry) +{ + char bf_key[MLXSW_SP2_BLOOM_KEY_LEN]; + u8 bf_size; + + mlxsw_sp2_acl_bf_key_encode(aregion, aentry, bf_key, &bf_size); + return mlxsw_sp2_acl_bf_crc(bf_key, bf_size); +} + +static u16 mlxsw_sp4_acl_bf_crc10_byte(u16 crc, u8 c) +{ + u8 index = ((crc >> 2) ^ c) & 0xff; + + return ((crc << 8) ^ mlxsw_sp4_acl_bf_crc10_tab[index]) & 0x3ff; +} + +static u16 mlxsw_sp4_acl_bf_crc6_byte(u16 crc, u8 c) +{ + u8 index = (crc ^ c) & 0xff; + + return ((crc << 6) ^ (mlxsw_sp4_acl_bf_crc6_tab[index] << 2)) & 0xfc; +} + +static u16 mlxsw_sp4_acl_bf_crc(const u8 *buffer, size_t len) +{ + u16 crc_row = 0, crc_col = 0; + + while (len--) { + crc_row = mlxsw_sp4_acl_bf_crc10_byte(crc_row, *buffer); + crc_col = mlxsw_sp4_acl_bf_crc6_byte(crc_col, *buffer); + buffer++; + } + + crc_col >>= 2; + + /* 6 bit column are MSB, 10 bit row are LSB */ + return (crc_col << 10) | crc_row; +} + +static void right_shift_array(char *arr, u8 len, u8 shift_bits) +{ + u8 byte_mask = 0xff >> shift_bits; + int i; + + if (WARN_ON(!shift_bits || shift_bits >= 8)) + return; + + for (i = len - 1; i >= 0; i--) { + /* The first iteration looks like out-of-bounds access, + * but actually references a buffer that the array is shifted + * into. This move is legal as we never send the last chunk to + * this function. + */ + arr[i + 1] &= byte_mask; + arr[i + 1] |= arr[i] << (8 - shift_bits); + arr[i] = arr[i] >> shift_bits; + } +} + +static void mlxsw_sp4_bf_key_shift_chunks(u8 chunk_count, char *output) +{ + /* The chunks are suppoosed to be continuous, with no padding. + * Since region ID and eRP ID use 14 bits, and not fully 2 bytes, + * and in Spectrum-4 there is no padding, it is necessary to shift some + * chunks 2 bits right. + */ + switch (chunk_count) { + case 2: + /* The chunks are copied as follow: + * +-------------+-----------------+ + * | Chunk 0 | Chunk 1 | + * | IDs | keys |(**) IDs | keys | + * +-------------+-----------------+ + * In (**), there are two unused bits, therefore, chunk 0 needs + * to be shifted two bits right. + */ + right_shift_array(output, MLXSW_SP4_BLOOM_KEY_CHUNK_BYTES, 2); + break; + case 3: + /* The chunks are copied as follow: + * +-------------+-----------------+-----------------+ + * | Chunk 0 | Chunk 1 | Chunk 2 | + * | IDs | keys |(**) IDs | keys |(**) IDs | keys | + * +-------------+-----------------+-----------------+ + * In (**), there are two unused bits, therefore, chunk 1 needs + * to be shifted two bits right and chunk 0 needs to be shifted + * four bits right. + */ + right_shift_array(output + MLXSW_SP4_BLOOM_KEY_CHUNK_BYTES, + MLXSW_SP4_BLOOM_KEY_CHUNK_BYTES, 2); + right_shift_array(output, MLXSW_SP4_BLOOM_KEY_CHUNK_BYTES, 4); + break; + default: + WARN_ON(chunk_count > MLXSW_BLOOM_KEY_CHUNKS); + } +} + +static void +mlxsw_sp4_acl_bf_key_encode(struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry, + char *output, u8 *len) +{ + struct mlxsw_afk_key_info *key_info = aregion->region->key_info; + u8 block_count = mlxsw_afk_key_info_blocks_count_get(key_info); + u8 chunk_count = 1 + ((block_count - 1) >> 2); + + __mlxsw_sp_acl_bf_key_encode(aregion, aentry, output, len, + MLXSW_BLOOM_KEY_CHUNKS, + MLXSW_SP4_BLOOM_CHUNK_PAD_BYTES, + MLXSW_SP4_BLOOM_CHUNK_KEY_OFFSET, + MLXSW_SP4_BLOOM_CHUNK_KEY_BYTES, + MLXSW_SP4_BLOOM_KEY_CHUNK_BYTES); + mlxsw_sp4_bf_key_shift_chunks(chunk_count, output); +} + +static unsigned int +mlxsw_sp4_acl_bf_index_get(struct mlxsw_sp_acl_bf *bf, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry) +{ + char bf_key[MLXSW_SP4_BLOOM_KEY_LEN] = {}; + u8 bf_size; + + mlxsw_sp4_acl_bf_key_encode(aregion, aentry, bf_key, &bf_size); + return mlxsw_sp4_acl_bf_crc(bf_key, bf_size); +} + +static unsigned int +mlxsw_sp_acl_bf_rule_count_index_get(struct mlxsw_sp_acl_bf *bf, + unsigned int erp_bank, + unsigned int bf_index) +{ + return erp_bank * bf->bank_size + bf_index; +} + +int +mlxsw_sp_acl_bf_entry_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_bf *bf, + struct mlxsw_sp_acl_atcam_region *aregion, + unsigned int erp_bank, + struct mlxsw_sp_acl_atcam_entry *aentry) +{ + unsigned int rule_index; + char *peabfe_pl; + u16 bf_index; + int err; + + mutex_lock(&bf->lock); + + bf_index = mlxsw_sp->acl_bf_ops->index_get(bf, aregion, aentry); + rule_index = mlxsw_sp_acl_bf_rule_count_index_get(bf, erp_bank, + bf_index); + + if (refcount_inc_not_zero(&bf->refcnt[rule_index])) { + err = 0; + goto unlock; + } + + peabfe_pl = kmalloc(MLXSW_REG_PEABFE_LEN, GFP_KERNEL); + if (!peabfe_pl) { + err = -ENOMEM; + goto unlock; + } + + mlxsw_reg_peabfe_pack(peabfe_pl); + mlxsw_reg_peabfe_rec_pack(peabfe_pl, 0, 1, erp_bank, bf_index); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(peabfe), peabfe_pl); + kfree(peabfe_pl); + if (err) + goto unlock; + + refcount_set(&bf->refcnt[rule_index], 1); + err = 0; + +unlock: + mutex_unlock(&bf->lock); + return err; +} + +void +mlxsw_sp_acl_bf_entry_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_bf *bf, + struct mlxsw_sp_acl_atcam_region *aregion, + unsigned int erp_bank, + struct mlxsw_sp_acl_atcam_entry *aentry) +{ + unsigned int rule_index; + char *peabfe_pl; + u16 bf_index; + + mutex_lock(&bf->lock); + + bf_index = mlxsw_sp->acl_bf_ops->index_get(bf, aregion, aentry); + rule_index = mlxsw_sp_acl_bf_rule_count_index_get(bf, erp_bank, + bf_index); + + if (refcount_dec_and_test(&bf->refcnt[rule_index])) { + peabfe_pl = kmalloc(MLXSW_REG_PEABFE_LEN, GFP_KERNEL); + if (!peabfe_pl) + goto unlock; + + mlxsw_reg_peabfe_pack(peabfe_pl); + mlxsw_reg_peabfe_rec_pack(peabfe_pl, 0, 0, erp_bank, bf_index); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(peabfe), peabfe_pl); + kfree(peabfe_pl); + } + +unlock: + mutex_unlock(&bf->lock); +} + +struct mlxsw_sp_acl_bf * +mlxsw_sp_acl_bf_init(struct mlxsw_sp *mlxsw_sp, unsigned int num_erp_banks) +{ + struct mlxsw_sp_acl_bf *bf; + unsigned int bf_bank_size; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, ACL_MAX_BF_LOG)) + return ERR_PTR(-EIO); + + /* Bloom filter size per erp_table_bank + * is 2^ACL_MAX_BF_LOG + */ + bf_bank_size = 1 << MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_BF_LOG); + bf = kzalloc(struct_size(bf, refcnt, size_mul(bf_bank_size, num_erp_banks)), + GFP_KERNEL); + if (!bf) + return ERR_PTR(-ENOMEM); + + bf->bank_size = bf_bank_size; + mutex_init(&bf->lock); + + return bf; +} + +void mlxsw_sp_acl_bf_fini(struct mlxsw_sp_acl_bf *bf) +{ + mutex_destroy(&bf->lock); + kfree(bf); +} + +const struct mlxsw_sp_acl_bf_ops mlxsw_sp2_acl_bf_ops = { + .index_get = mlxsw_sp2_acl_bf_index_get, +}; + +const struct mlxsw_sp_acl_bf_ops mlxsw_sp4_acl_bf_ops = { + .index_get = mlxsw_sp4_acl_bf_index_get, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_ctcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_ctcam.c new file mode 100644 index 000000000..05680a7e6 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_ctcam.c @@ -0,0 +1,233 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/parman.h> + +#include "reg.h" +#include "core.h" +#include "spectrum.h" +#include "spectrum_acl_tcam.h" + +static int +mlxsw_sp_acl_ctcam_region_resize(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *region, + u16 new_size) +{ + char ptar_pl[MLXSW_REG_PTAR_LEN]; + + mlxsw_reg_ptar_pack(ptar_pl, MLXSW_REG_PTAR_OP_RESIZE, + region->key_type, new_size, region->id, + region->tcam_region_info); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptar), ptar_pl); +} + +static void +mlxsw_sp_acl_ctcam_region_move(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *region, + u16 src_offset, u16 dst_offset, u16 size) +{ + char prcr_pl[MLXSW_REG_PRCR_LEN]; + + mlxsw_reg_prcr_pack(prcr_pl, MLXSW_REG_PRCR_OP_MOVE, + region->tcam_region_info, src_offset, + region->tcam_region_info, dst_offset, size); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(prcr), prcr_pl); +} + +static int +mlxsw_sp_acl_ctcam_region_entry_insert(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_entry *centry, + struct mlxsw_sp_acl_rule_info *rulei, + bool fillup_priority) +{ + struct mlxsw_sp_acl_tcam_region *region = cregion->region; + struct mlxsw_afk *afk = mlxsw_sp_acl_afk(mlxsw_sp->acl); + char ptce2_pl[MLXSW_REG_PTCE2_LEN]; + char *act_set; + u32 priority; + char *mask; + char *key; + int err; + + err = mlxsw_sp_acl_tcam_priority_get(mlxsw_sp, rulei, &priority, + fillup_priority); + if (err) + return err; + + mlxsw_reg_ptce2_pack(ptce2_pl, true, MLXSW_REG_PTCE2_OP_WRITE_WRITE, + region->tcam_region_info, + centry->parman_item.index, priority); + key = mlxsw_reg_ptce2_flex_key_blocks_data(ptce2_pl); + mask = mlxsw_reg_ptce2_mask_data(ptce2_pl); + mlxsw_afk_encode(afk, region->key_info, &rulei->values, key, mask); + + err = cregion->ops->entry_insert(cregion, centry, mask); + if (err) + return err; + + /* Only the first action set belongs here, the rest is in KVD */ + act_set = mlxsw_afa_block_first_set(rulei->act_block); + mlxsw_reg_ptce2_flex_action_set_memcpy_to(ptce2_pl, act_set); + + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptce2), ptce2_pl); + if (err) + goto err_ptce2_write; + + return 0; + +err_ptce2_write: + cregion->ops->entry_remove(cregion, centry); + return err; +} + +static void +mlxsw_sp_acl_ctcam_region_entry_remove(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_entry *centry) +{ + char ptce2_pl[MLXSW_REG_PTCE2_LEN]; + + mlxsw_reg_ptce2_pack(ptce2_pl, false, MLXSW_REG_PTCE2_OP_WRITE_WRITE, + cregion->region->tcam_region_info, + centry->parman_item.index, 0); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptce2), ptce2_pl); + cregion->ops->entry_remove(cregion, centry); +} + +static int +mlxsw_sp_acl_ctcam_region_entry_action_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_entry *centry, + struct mlxsw_afa_block *afa_block, + unsigned int priority) +{ + char ptce2_pl[MLXSW_REG_PTCE2_LEN]; + char *act_set; + + mlxsw_reg_ptce2_pack(ptce2_pl, true, MLXSW_REG_PTCE2_OP_WRITE_UPDATE, + cregion->region->tcam_region_info, + centry->parman_item.index, priority); + + act_set = mlxsw_afa_block_first_set(afa_block); + mlxsw_reg_ptce2_flex_action_set_memcpy_to(ptce2_pl, act_set); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptce2), ptce2_pl); +} + + +static int mlxsw_sp_acl_ctcam_region_parman_resize(void *priv, + unsigned long new_count) +{ + struct mlxsw_sp_acl_ctcam_region *cregion = priv; + struct mlxsw_sp_acl_tcam_region *region = cregion->region; + struct mlxsw_sp *mlxsw_sp = region->mlxsw_sp; + u64 max_tcam_rules; + + max_tcam_rules = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_TCAM_RULES); + if (new_count > max_tcam_rules) + return -EINVAL; + return mlxsw_sp_acl_ctcam_region_resize(mlxsw_sp, region, new_count); +} + +static void mlxsw_sp_acl_ctcam_region_parman_move(void *priv, + unsigned long from_index, + unsigned long to_index, + unsigned long count) +{ + struct mlxsw_sp_acl_ctcam_region *cregion = priv; + struct mlxsw_sp_acl_tcam_region *region = cregion->region; + struct mlxsw_sp *mlxsw_sp = region->mlxsw_sp; + + mlxsw_sp_acl_ctcam_region_move(mlxsw_sp, region, + from_index, to_index, count); +} + +static const struct parman_ops mlxsw_sp_acl_ctcam_region_parman_ops = { + .base_count = MLXSW_SP_ACL_TCAM_REGION_BASE_COUNT, + .resize_step = MLXSW_SP_ACL_TCAM_REGION_RESIZE_STEP, + .resize = mlxsw_sp_acl_ctcam_region_parman_resize, + .move = mlxsw_sp_acl_ctcam_region_parman_move, + .algo = PARMAN_ALGO_TYPE_LSORT, +}; + +int +mlxsw_sp_acl_ctcam_region_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_tcam_region *region, + const struct mlxsw_sp_acl_ctcam_region_ops *ops) +{ + cregion->region = region; + cregion->ops = ops; + cregion->parman = parman_create(&mlxsw_sp_acl_ctcam_region_parman_ops, + cregion); + if (!cregion->parman) + return -ENOMEM; + return 0; +} + +void mlxsw_sp_acl_ctcam_region_fini(struct mlxsw_sp_acl_ctcam_region *cregion) +{ + parman_destroy(cregion->parman); +} + +void mlxsw_sp_acl_ctcam_chunk_init(struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_chunk *cchunk, + unsigned int priority) +{ + parman_prio_init(cregion->parman, &cchunk->parman_prio, priority); +} + +void mlxsw_sp_acl_ctcam_chunk_fini(struct mlxsw_sp_acl_ctcam_chunk *cchunk) +{ + parman_prio_fini(&cchunk->parman_prio); +} + +int mlxsw_sp_acl_ctcam_entry_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_chunk *cchunk, + struct mlxsw_sp_acl_ctcam_entry *centry, + struct mlxsw_sp_acl_rule_info *rulei, + bool fillup_priority) +{ + int err; + + err = parman_item_add(cregion->parman, &cchunk->parman_prio, + ¢ry->parman_item); + if (err) + return err; + + err = mlxsw_sp_acl_ctcam_region_entry_insert(mlxsw_sp, cregion, centry, + rulei, fillup_priority); + if (err) + goto err_rule_insert; + return 0; + +err_rule_insert: + parman_item_remove(cregion->parman, &cchunk->parman_prio, + ¢ry->parman_item); + return err; +} + +void mlxsw_sp_acl_ctcam_entry_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_chunk *cchunk, + struct mlxsw_sp_acl_ctcam_entry *centry) +{ + mlxsw_sp_acl_ctcam_region_entry_remove(mlxsw_sp, cregion, centry); + parman_item_remove(cregion->parman, &cchunk->parman_prio, + ¢ry->parman_item); +} + +int mlxsw_sp_acl_ctcam_entry_action_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_entry *centry, + struct mlxsw_sp_acl_rule_info *rulei) +{ + return mlxsw_sp_acl_ctcam_region_entry_action_replace(mlxsw_sp, cregion, + centry, + rulei->act_block, + rulei->priority); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_erp.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_erp.c new file mode 100644 index 000000000..d231f4d28 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_erp.c @@ -0,0 +1,1601 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#include <linux/bitmap.h> +#include <linux/errno.h> +#include <linux/genalloc.h> +#include <linux/gfp.h> +#include <linux/kernel.h> +#include <linux/list.h> +#include <linux/mutex.h> +#include <linux/objagg.h> +#include <linux/rtnetlink.h> +#include <linux/slab.h> + +#include "core.h" +#include "reg.h" +#include "spectrum.h" +#include "spectrum_acl_tcam.h" + +/* gen_pool_alloc() returns 0 when allocation fails, so use an offset */ +#define MLXSW_SP_ACL_ERP_GENALLOC_OFFSET 0x100 +#define MLXSW_SP_ACL_ERP_MAX_PER_REGION 16 + +struct mlxsw_sp_acl_erp_core { + unsigned int erpt_entries_size[MLXSW_SP_ACL_ATCAM_REGION_TYPE_MAX + 1]; + struct gen_pool *erp_tables; + struct mlxsw_sp *mlxsw_sp; + struct mlxsw_sp_acl_bf *bf; + unsigned int num_erp_banks; +}; + +struct mlxsw_sp_acl_erp_key { + char mask[MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN]; +#define __MASK_LEN 0x38 +#define __MASK_IDX(i) (__MASK_LEN - (i) - 1) + bool ctcam; +}; + +struct mlxsw_sp_acl_erp { + struct mlxsw_sp_acl_erp_key key; + u8 id; + u8 index; + DECLARE_BITMAP(mask_bitmap, MLXSW_SP_ACL_TCAM_MASK_LEN); + struct list_head list; + struct mlxsw_sp_acl_erp_table *erp_table; +}; + +struct mlxsw_sp_acl_erp_master_mask { + DECLARE_BITMAP(bitmap, MLXSW_SP_ACL_TCAM_MASK_LEN); + unsigned int count[MLXSW_SP_ACL_TCAM_MASK_LEN]; +}; + +struct mlxsw_sp_acl_erp_table { + struct mlxsw_sp_acl_erp_master_mask master_mask; + DECLARE_BITMAP(erp_id_bitmap, MLXSW_SP_ACL_ERP_MAX_PER_REGION); + DECLARE_BITMAP(erp_index_bitmap, MLXSW_SP_ACL_ERP_MAX_PER_REGION); + struct list_head atcam_erps_list; + struct mlxsw_sp_acl_erp_core *erp_core; + struct mlxsw_sp_acl_atcam_region *aregion; + const struct mlxsw_sp_acl_erp_table_ops *ops; + unsigned long base_index; + unsigned int num_atcam_erps; + unsigned int num_max_atcam_erps; + unsigned int num_ctcam_erps; + unsigned int num_deltas; + struct objagg *objagg; + struct mutex objagg_lock; /* guards objagg manipulation */ +}; + +struct mlxsw_sp_acl_erp_table_ops { + struct mlxsw_sp_acl_erp * + (*erp_create)(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp_key *key); + void (*erp_destroy)(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp); +}; + +static struct mlxsw_sp_acl_erp * +mlxsw_sp_acl_erp_mask_create(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp_key *key); +static void +mlxsw_sp_acl_erp_mask_destroy(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp); +static struct mlxsw_sp_acl_erp * +mlxsw_sp_acl_erp_second_mask_create(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp_key *key); +static void +mlxsw_sp_acl_erp_second_mask_destroy(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp); +static struct mlxsw_sp_acl_erp * +mlxsw_sp_acl_erp_first_mask_create(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp_key *key); +static void +mlxsw_sp_acl_erp_first_mask_destroy(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp); +static void +mlxsw_sp_acl_erp_no_mask_destroy(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp); + +static const struct mlxsw_sp_acl_erp_table_ops erp_multiple_masks_ops = { + .erp_create = mlxsw_sp_acl_erp_mask_create, + .erp_destroy = mlxsw_sp_acl_erp_mask_destroy, +}; + +static const struct mlxsw_sp_acl_erp_table_ops erp_two_masks_ops = { + .erp_create = mlxsw_sp_acl_erp_mask_create, + .erp_destroy = mlxsw_sp_acl_erp_second_mask_destroy, +}; + +static const struct mlxsw_sp_acl_erp_table_ops erp_single_mask_ops = { + .erp_create = mlxsw_sp_acl_erp_second_mask_create, + .erp_destroy = mlxsw_sp_acl_erp_first_mask_destroy, +}; + +static const struct mlxsw_sp_acl_erp_table_ops erp_no_mask_ops = { + .erp_create = mlxsw_sp_acl_erp_first_mask_create, + .erp_destroy = mlxsw_sp_acl_erp_no_mask_destroy, +}; + +static bool +mlxsw_sp_acl_erp_table_is_used(const struct mlxsw_sp_acl_erp_table *erp_table) +{ + return erp_table->ops != &erp_single_mask_ops && + erp_table->ops != &erp_no_mask_ops; +} + +static unsigned int +mlxsw_sp_acl_erp_bank_get(const struct mlxsw_sp_acl_erp *erp) +{ + return erp->index % erp->erp_table->erp_core->num_erp_banks; +} + +static unsigned int +mlxsw_sp_acl_erp_table_entry_size(const struct mlxsw_sp_acl_erp_table *erp_table) +{ + struct mlxsw_sp_acl_atcam_region *aregion = erp_table->aregion; + struct mlxsw_sp_acl_erp_core *erp_core = erp_table->erp_core; + + return erp_core->erpt_entries_size[aregion->type]; +} + +static int mlxsw_sp_acl_erp_id_get(struct mlxsw_sp_acl_erp_table *erp_table, + u8 *p_id) +{ + u8 id; + + id = find_first_zero_bit(erp_table->erp_id_bitmap, + MLXSW_SP_ACL_ERP_MAX_PER_REGION); + if (id < MLXSW_SP_ACL_ERP_MAX_PER_REGION) { + __set_bit(id, erp_table->erp_id_bitmap); + *p_id = id; + return 0; + } + + return -ENOBUFS; +} + +static void mlxsw_sp_acl_erp_id_put(struct mlxsw_sp_acl_erp_table *erp_table, + u8 id) +{ + __clear_bit(id, erp_table->erp_id_bitmap); +} + +static void +mlxsw_sp_acl_erp_master_mask_bit_set(unsigned long bit, + struct mlxsw_sp_acl_erp_master_mask *mask) +{ + if (mask->count[bit]++ == 0) + __set_bit(bit, mask->bitmap); +} + +static void +mlxsw_sp_acl_erp_master_mask_bit_clear(unsigned long bit, + struct mlxsw_sp_acl_erp_master_mask *mask) +{ + if (--mask->count[bit] == 0) + __clear_bit(bit, mask->bitmap); +} + +static int +mlxsw_sp_acl_erp_master_mask_update(struct mlxsw_sp_acl_erp_table *erp_table) +{ + struct mlxsw_sp_acl_tcam_region *region = erp_table->aregion->region; + struct mlxsw_sp *mlxsw_sp = region->mlxsw_sp; + char percr_pl[MLXSW_REG_PERCR_LEN]; + char *master_mask; + + mlxsw_reg_percr_pack(percr_pl, region->id); + master_mask = mlxsw_reg_percr_master_mask_data(percr_pl); + bitmap_to_arr32((u32 *) master_mask, erp_table->master_mask.bitmap, + MLXSW_SP_ACL_TCAM_MASK_LEN); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(percr), percr_pl); +} + +static int +mlxsw_sp_acl_erp_master_mask_set(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp_key *key) +{ + DECLARE_BITMAP(mask_bitmap, MLXSW_SP_ACL_TCAM_MASK_LEN); + unsigned long bit; + int err; + + bitmap_from_arr32(mask_bitmap, (u32 *) key->mask, + MLXSW_SP_ACL_TCAM_MASK_LEN); + for_each_set_bit(bit, mask_bitmap, MLXSW_SP_ACL_TCAM_MASK_LEN) + mlxsw_sp_acl_erp_master_mask_bit_set(bit, + &erp_table->master_mask); + + err = mlxsw_sp_acl_erp_master_mask_update(erp_table); + if (err) + goto err_master_mask_update; + + return 0; + +err_master_mask_update: + for_each_set_bit(bit, mask_bitmap, MLXSW_SP_ACL_TCAM_MASK_LEN) + mlxsw_sp_acl_erp_master_mask_bit_clear(bit, + &erp_table->master_mask); + return err; +} + +static int +mlxsw_sp_acl_erp_master_mask_clear(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp_key *key) +{ + DECLARE_BITMAP(mask_bitmap, MLXSW_SP_ACL_TCAM_MASK_LEN); + unsigned long bit; + int err; + + bitmap_from_arr32(mask_bitmap, (u32 *) key->mask, + MLXSW_SP_ACL_TCAM_MASK_LEN); + for_each_set_bit(bit, mask_bitmap, MLXSW_SP_ACL_TCAM_MASK_LEN) + mlxsw_sp_acl_erp_master_mask_bit_clear(bit, + &erp_table->master_mask); + + err = mlxsw_sp_acl_erp_master_mask_update(erp_table); + if (err) + goto err_master_mask_update; + + return 0; + +err_master_mask_update: + for_each_set_bit(bit, mask_bitmap, MLXSW_SP_ACL_TCAM_MASK_LEN) + mlxsw_sp_acl_erp_master_mask_bit_set(bit, + &erp_table->master_mask); + return err; +} + +static struct mlxsw_sp_acl_erp * +mlxsw_sp_acl_erp_generic_create(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp_key *key) +{ + struct mlxsw_sp_acl_erp *erp; + int err; + + erp = kzalloc(sizeof(*erp), GFP_KERNEL); + if (!erp) + return ERR_PTR(-ENOMEM); + + err = mlxsw_sp_acl_erp_id_get(erp_table, &erp->id); + if (err) + goto err_erp_id_get; + + memcpy(&erp->key, key, sizeof(*key)); + list_add(&erp->list, &erp_table->atcam_erps_list); + erp_table->num_atcam_erps++; + erp->erp_table = erp_table; + + err = mlxsw_sp_acl_erp_master_mask_set(erp_table, &erp->key); + if (err) + goto err_master_mask_set; + + return erp; + +err_master_mask_set: + erp_table->num_atcam_erps--; + list_del(&erp->list); + mlxsw_sp_acl_erp_id_put(erp_table, erp->id); +err_erp_id_get: + kfree(erp); + return ERR_PTR(err); +} + +static void +mlxsw_sp_acl_erp_generic_destroy(struct mlxsw_sp_acl_erp *erp) +{ + struct mlxsw_sp_acl_erp_table *erp_table = erp->erp_table; + + mlxsw_sp_acl_erp_master_mask_clear(erp_table, &erp->key); + erp_table->num_atcam_erps--; + list_del(&erp->list); + mlxsw_sp_acl_erp_id_put(erp_table, erp->id); + kfree(erp); +} + +static int +mlxsw_sp_acl_erp_table_alloc(struct mlxsw_sp_acl_erp_core *erp_core, + unsigned int num_erps, + enum mlxsw_sp_acl_atcam_region_type region_type, + unsigned long *p_index) +{ + unsigned int num_rows, entry_size; + unsigned long index; + + /* We only allow allocations of entire rows */ + if (num_erps % erp_core->num_erp_banks != 0) + return -EINVAL; + + entry_size = erp_core->erpt_entries_size[region_type]; + num_rows = num_erps / erp_core->num_erp_banks; + + index = gen_pool_alloc(erp_core->erp_tables, num_rows * entry_size); + if (!index) + return -ENOBUFS; + + *p_index = index - MLXSW_SP_ACL_ERP_GENALLOC_OFFSET; + + return 0; +} + +static void +mlxsw_sp_acl_erp_table_free(struct mlxsw_sp_acl_erp_core *erp_core, + unsigned int num_erps, + enum mlxsw_sp_acl_atcam_region_type region_type, + unsigned long index) +{ + unsigned long base_index; + unsigned int entry_size; + size_t size; + + entry_size = erp_core->erpt_entries_size[region_type]; + base_index = index + MLXSW_SP_ACL_ERP_GENALLOC_OFFSET; + size = num_erps / erp_core->num_erp_banks * entry_size; + gen_pool_free(erp_core->erp_tables, base_index, size); +} + +static struct mlxsw_sp_acl_erp * +mlxsw_sp_acl_erp_table_master_rp(struct mlxsw_sp_acl_erp_table *erp_table) +{ + if (!list_is_singular(&erp_table->atcam_erps_list)) + return NULL; + + return list_first_entry(&erp_table->atcam_erps_list, + struct mlxsw_sp_acl_erp, list); +} + +static int mlxsw_sp_acl_erp_index_get(struct mlxsw_sp_acl_erp_table *erp_table, + u8 *p_index) +{ + u8 index; + + index = find_first_zero_bit(erp_table->erp_index_bitmap, + erp_table->num_max_atcam_erps); + if (index < erp_table->num_max_atcam_erps) { + __set_bit(index, erp_table->erp_index_bitmap); + *p_index = index; + return 0; + } + + return -ENOBUFS; +} + +static void mlxsw_sp_acl_erp_index_put(struct mlxsw_sp_acl_erp_table *erp_table, + u8 index) +{ + __clear_bit(index, erp_table->erp_index_bitmap); +} + +static void +mlxsw_sp_acl_erp_table_locate(const struct mlxsw_sp_acl_erp_table *erp_table, + const struct mlxsw_sp_acl_erp *erp, + u8 *p_erpt_bank, u8 *p_erpt_index) +{ + unsigned int entry_size = mlxsw_sp_acl_erp_table_entry_size(erp_table); + struct mlxsw_sp_acl_erp_core *erp_core = erp_table->erp_core; + unsigned int row; + + *p_erpt_bank = erp->index % erp_core->num_erp_banks; + row = erp->index / erp_core->num_erp_banks; + *p_erpt_index = erp_table->base_index + row * entry_size; +} + +static int +mlxsw_sp_acl_erp_table_erp_add(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp) +{ + struct mlxsw_sp *mlxsw_sp = erp_table->erp_core->mlxsw_sp; + enum mlxsw_reg_perpt_key_size key_size; + char perpt_pl[MLXSW_REG_PERPT_LEN]; + u8 erpt_bank, erpt_index; + + mlxsw_sp_acl_erp_table_locate(erp_table, erp, &erpt_bank, &erpt_index); + key_size = (enum mlxsw_reg_perpt_key_size) erp_table->aregion->type; + mlxsw_reg_perpt_pack(perpt_pl, erpt_bank, erpt_index, key_size, erp->id, + 0, erp_table->base_index, erp->index, + erp->key.mask); + mlxsw_reg_perpt_erp_vector_pack(perpt_pl, erp_table->erp_index_bitmap, + MLXSW_SP_ACL_ERP_MAX_PER_REGION); + mlxsw_reg_perpt_erp_vector_set(perpt_pl, erp->index, true); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(perpt), perpt_pl); +} + +static void mlxsw_sp_acl_erp_table_erp_del(struct mlxsw_sp_acl_erp *erp) +{ + char empty_mask[MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN] = { 0 }; + struct mlxsw_sp_acl_erp_table *erp_table = erp->erp_table; + struct mlxsw_sp *mlxsw_sp = erp_table->erp_core->mlxsw_sp; + enum mlxsw_reg_perpt_key_size key_size; + char perpt_pl[MLXSW_REG_PERPT_LEN]; + u8 erpt_bank, erpt_index; + + mlxsw_sp_acl_erp_table_locate(erp_table, erp, &erpt_bank, &erpt_index); + key_size = (enum mlxsw_reg_perpt_key_size) erp_table->aregion->type; + mlxsw_reg_perpt_pack(perpt_pl, erpt_bank, erpt_index, key_size, erp->id, + 0, erp_table->base_index, erp->index, empty_mask); + mlxsw_reg_perpt_erp_vector_pack(perpt_pl, erp_table->erp_index_bitmap, + MLXSW_SP_ACL_ERP_MAX_PER_REGION); + mlxsw_reg_perpt_erp_vector_set(perpt_pl, erp->index, false); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(perpt), perpt_pl); +} + +static int +mlxsw_sp_acl_erp_table_enable(struct mlxsw_sp_acl_erp_table *erp_table, + bool ctcam_le) +{ + struct mlxsw_sp_acl_tcam_region *region = erp_table->aregion->region; + struct mlxsw_sp *mlxsw_sp = erp_table->erp_core->mlxsw_sp; + char pererp_pl[MLXSW_REG_PERERP_LEN]; + + mlxsw_reg_pererp_pack(pererp_pl, region->id, ctcam_le, true, 0, + erp_table->base_index, 0); + mlxsw_reg_pererp_erp_vector_pack(pererp_pl, erp_table->erp_index_bitmap, + MLXSW_SP_ACL_ERP_MAX_PER_REGION); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pererp), pererp_pl); +} + +static void +mlxsw_sp_acl_erp_table_disable(struct mlxsw_sp_acl_erp_table *erp_table) +{ + struct mlxsw_sp_acl_tcam_region *region = erp_table->aregion->region; + struct mlxsw_sp *mlxsw_sp = erp_table->erp_core->mlxsw_sp; + char pererp_pl[MLXSW_REG_PERERP_LEN]; + struct mlxsw_sp_acl_erp *master_rp; + + master_rp = mlxsw_sp_acl_erp_table_master_rp(erp_table); + /* It is possible we do not have a master RP when we disable the + * table when there are no rules in the A-TCAM and the last C-TCAM + * rule is deleted + */ + mlxsw_reg_pererp_pack(pererp_pl, region->id, false, false, 0, 0, + master_rp ? master_rp->id : 0); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pererp), pererp_pl); +} + +static int +mlxsw_sp_acl_erp_table_relocate(struct mlxsw_sp_acl_erp_table *erp_table) +{ + struct mlxsw_sp_acl_erp *erp; + int err; + + list_for_each_entry(erp, &erp_table->atcam_erps_list, list) { + err = mlxsw_sp_acl_erp_table_erp_add(erp_table, erp); + if (err) + goto err_table_erp_add; + } + + return 0; + +err_table_erp_add: + list_for_each_entry_continue_reverse(erp, &erp_table->atcam_erps_list, + list) + mlxsw_sp_acl_erp_table_erp_del(erp); + return err; +} + +static int +mlxsw_sp_acl_erp_table_expand(struct mlxsw_sp_acl_erp_table *erp_table) +{ + unsigned int num_erps, old_num_erps = erp_table->num_max_atcam_erps; + struct mlxsw_sp_acl_erp_core *erp_core = erp_table->erp_core; + unsigned long old_base_index = erp_table->base_index; + bool ctcam_le = erp_table->num_ctcam_erps > 0; + int err; + + if (erp_table->num_atcam_erps < erp_table->num_max_atcam_erps) + return 0; + + if (erp_table->num_max_atcam_erps == MLXSW_SP_ACL_ERP_MAX_PER_REGION) + return -ENOBUFS; + + num_erps = old_num_erps + erp_core->num_erp_banks; + err = mlxsw_sp_acl_erp_table_alloc(erp_core, num_erps, + erp_table->aregion->type, + &erp_table->base_index); + if (err) + return err; + erp_table->num_max_atcam_erps = num_erps; + + err = mlxsw_sp_acl_erp_table_relocate(erp_table); + if (err) + goto err_table_relocate; + + err = mlxsw_sp_acl_erp_table_enable(erp_table, ctcam_le); + if (err) + goto err_table_enable; + + mlxsw_sp_acl_erp_table_free(erp_core, old_num_erps, + erp_table->aregion->type, old_base_index); + + return 0; + +err_table_enable: +err_table_relocate: + erp_table->num_max_atcam_erps = old_num_erps; + mlxsw_sp_acl_erp_table_free(erp_core, num_erps, + erp_table->aregion->type, + erp_table->base_index); + erp_table->base_index = old_base_index; + return err; +} + +static int +mlxsw_acl_erp_table_bf_add(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp) +{ + struct mlxsw_sp_acl_atcam_region *aregion = erp_table->aregion; + unsigned int erp_bank = mlxsw_sp_acl_erp_bank_get(erp); + struct mlxsw_sp_acl_atcam_entry *aentry; + int err; + + list_for_each_entry(aentry, &aregion->entries_list, list) { + err = mlxsw_sp_acl_bf_entry_add(aregion->region->mlxsw_sp, + erp_table->erp_core->bf, + aregion, erp_bank, aentry); + if (err) + goto bf_entry_add_err; + } + + return 0; + +bf_entry_add_err: + list_for_each_entry_continue_reverse(aentry, &aregion->entries_list, + list) + mlxsw_sp_acl_bf_entry_del(aregion->region->mlxsw_sp, + erp_table->erp_core->bf, + aregion, erp_bank, aentry); + return err; +} + +static void +mlxsw_acl_erp_table_bf_del(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp) +{ + struct mlxsw_sp_acl_atcam_region *aregion = erp_table->aregion; + unsigned int erp_bank = mlxsw_sp_acl_erp_bank_get(erp); + struct mlxsw_sp_acl_atcam_entry *aentry; + + list_for_each_entry_reverse(aentry, &aregion->entries_list, list) + mlxsw_sp_acl_bf_entry_del(aregion->region->mlxsw_sp, + erp_table->erp_core->bf, + aregion, erp_bank, aentry); +} + +static int +mlxsw_sp_acl_erp_region_table_trans(struct mlxsw_sp_acl_erp_table *erp_table) +{ + struct mlxsw_sp_acl_erp_core *erp_core = erp_table->erp_core; + struct mlxsw_sp_acl_erp *master_rp; + int err; + + /* Initially, allocate a single eRP row. Expand later as needed */ + err = mlxsw_sp_acl_erp_table_alloc(erp_core, erp_core->num_erp_banks, + erp_table->aregion->type, + &erp_table->base_index); + if (err) + return err; + erp_table->num_max_atcam_erps = erp_core->num_erp_banks; + + /* Transition the sole RP currently configured (the master RP) + * to the eRP table + */ + master_rp = mlxsw_sp_acl_erp_table_master_rp(erp_table); + if (!master_rp) { + err = -EINVAL; + goto err_table_master_rp; + } + + /* Make sure the master RP is using a valid index, as + * only a single eRP row is currently allocated. + */ + master_rp->index = 0; + __set_bit(master_rp->index, erp_table->erp_index_bitmap); + + err = mlxsw_sp_acl_erp_table_erp_add(erp_table, master_rp); + if (err) + goto err_table_master_rp_add; + + /* Update Bloom filter before enabling eRP table, as rules + * on the master RP were not set to Bloom filter up to this + * point. + */ + err = mlxsw_acl_erp_table_bf_add(erp_table, master_rp); + if (err) + goto err_table_bf_add; + + err = mlxsw_sp_acl_erp_table_enable(erp_table, false); + if (err) + goto err_table_enable; + + return 0; + +err_table_enable: + mlxsw_acl_erp_table_bf_del(erp_table, master_rp); +err_table_bf_add: + mlxsw_sp_acl_erp_table_erp_del(master_rp); +err_table_master_rp_add: + __clear_bit(master_rp->index, erp_table->erp_index_bitmap); +err_table_master_rp: + mlxsw_sp_acl_erp_table_free(erp_core, erp_table->num_max_atcam_erps, + erp_table->aregion->type, + erp_table->base_index); + return err; +} + +static void +mlxsw_sp_acl_erp_region_master_mask_trans(struct mlxsw_sp_acl_erp_table *erp_table) +{ + struct mlxsw_sp_acl_erp_core *erp_core = erp_table->erp_core; + struct mlxsw_sp_acl_erp *master_rp; + + mlxsw_sp_acl_erp_table_disable(erp_table); + master_rp = mlxsw_sp_acl_erp_table_master_rp(erp_table); + if (!master_rp) + return; + mlxsw_acl_erp_table_bf_del(erp_table, master_rp); + mlxsw_sp_acl_erp_table_erp_del(master_rp); + __clear_bit(master_rp->index, erp_table->erp_index_bitmap); + mlxsw_sp_acl_erp_table_free(erp_core, erp_table->num_max_atcam_erps, + erp_table->aregion->type, + erp_table->base_index); +} + +static int +mlxsw_sp_acl_erp_region_erp_add(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp) +{ + struct mlxsw_sp_acl_tcam_region *region = erp_table->aregion->region; + struct mlxsw_sp *mlxsw_sp = erp_table->erp_core->mlxsw_sp; + bool ctcam_le = erp_table->num_ctcam_erps > 0; + char pererp_pl[MLXSW_REG_PERERP_LEN]; + + mlxsw_reg_pererp_pack(pererp_pl, region->id, ctcam_le, true, 0, + erp_table->base_index, 0); + mlxsw_reg_pererp_erp_vector_pack(pererp_pl, erp_table->erp_index_bitmap, + MLXSW_SP_ACL_ERP_MAX_PER_REGION); + mlxsw_reg_pererp_erpt_vector_set(pererp_pl, erp->index, true); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pererp), pererp_pl); +} + +static void mlxsw_sp_acl_erp_region_erp_del(struct mlxsw_sp_acl_erp *erp) +{ + struct mlxsw_sp_acl_erp_table *erp_table = erp->erp_table; + struct mlxsw_sp_acl_tcam_region *region = erp_table->aregion->region; + struct mlxsw_sp *mlxsw_sp = erp_table->erp_core->mlxsw_sp; + bool ctcam_le = erp_table->num_ctcam_erps > 0; + char pererp_pl[MLXSW_REG_PERERP_LEN]; + + mlxsw_reg_pererp_pack(pererp_pl, region->id, ctcam_le, true, 0, + erp_table->base_index, 0); + mlxsw_reg_pererp_erp_vector_pack(pererp_pl, erp_table->erp_index_bitmap, + MLXSW_SP_ACL_ERP_MAX_PER_REGION); + mlxsw_reg_pererp_erpt_vector_set(pererp_pl, erp->index, false); + + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pererp), pererp_pl); +} + +static int +mlxsw_sp_acl_erp_region_ctcam_enable(struct mlxsw_sp_acl_erp_table *erp_table) +{ + /* No need to re-enable lookup in the C-TCAM */ + if (erp_table->num_ctcam_erps > 1) + return 0; + + return mlxsw_sp_acl_erp_table_enable(erp_table, true); +} + +static void +mlxsw_sp_acl_erp_region_ctcam_disable(struct mlxsw_sp_acl_erp_table *erp_table) +{ + /* Only disable C-TCAM lookup when last C-TCAM eRP is deleted */ + if (erp_table->num_ctcam_erps > 1) + return; + + mlxsw_sp_acl_erp_table_enable(erp_table, false); +} + +static int +__mlxsw_sp_acl_erp_table_other_inc(struct mlxsw_sp_acl_erp_table *erp_table, + unsigned int *inc_num) +{ + int err; + + /* If there are C-TCAM eRP or deltas in use we need to transition + * the region to use eRP table, if it is not already done + */ + if (!mlxsw_sp_acl_erp_table_is_used(erp_table)) { + err = mlxsw_sp_acl_erp_region_table_trans(erp_table); + if (err) + return err; + } + + /* When C-TCAM or deltas are used, the eRP table must be used */ + if (erp_table->ops != &erp_multiple_masks_ops) + erp_table->ops = &erp_multiple_masks_ops; + + (*inc_num)++; + + return 0; +} + +static int mlxsw_sp_acl_erp_ctcam_inc(struct mlxsw_sp_acl_erp_table *erp_table) +{ + return __mlxsw_sp_acl_erp_table_other_inc(erp_table, + &erp_table->num_ctcam_erps); +} + +static int mlxsw_sp_acl_erp_delta_inc(struct mlxsw_sp_acl_erp_table *erp_table) +{ + return __mlxsw_sp_acl_erp_table_other_inc(erp_table, + &erp_table->num_deltas); +} + +static void +__mlxsw_sp_acl_erp_table_other_dec(struct mlxsw_sp_acl_erp_table *erp_table, + unsigned int *dec_num) +{ + (*dec_num)--; + + /* If there are no C-TCAM eRP or deltas in use, the state we + * transition to depends on the number of A-TCAM eRPs currently + * in use. + */ + if (erp_table->num_ctcam_erps > 0 || erp_table->num_deltas > 0) + return; + + switch (erp_table->num_atcam_erps) { + case 2: + /* Keep using the eRP table, but correctly set the + * operations pointer so that when an A-TCAM eRP is + * deleted we will transition to use the master mask + */ + erp_table->ops = &erp_two_masks_ops; + break; + case 1: + /* We only kept the eRP table because we had C-TCAM + * eRPs in use. Now that the last C-TCAM eRP is gone we + * can stop using the table and transition to use the + * master mask + */ + mlxsw_sp_acl_erp_region_master_mask_trans(erp_table); + erp_table->ops = &erp_single_mask_ops; + break; + case 0: + /* There are no more eRPs of any kind used by the region + * so free its eRP table and transition to initial state + */ + mlxsw_sp_acl_erp_table_disable(erp_table); + mlxsw_sp_acl_erp_table_free(erp_table->erp_core, + erp_table->num_max_atcam_erps, + erp_table->aregion->type, + erp_table->base_index); + erp_table->ops = &erp_no_mask_ops; + break; + default: + break; + } +} + +static void mlxsw_sp_acl_erp_ctcam_dec(struct mlxsw_sp_acl_erp_table *erp_table) +{ + __mlxsw_sp_acl_erp_table_other_dec(erp_table, + &erp_table->num_ctcam_erps); +} + +static void mlxsw_sp_acl_erp_delta_dec(struct mlxsw_sp_acl_erp_table *erp_table) +{ + __mlxsw_sp_acl_erp_table_other_dec(erp_table, + &erp_table->num_deltas); +} + +static struct mlxsw_sp_acl_erp * +mlxsw_sp_acl_erp_ctcam_mask_create(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp_key *key) +{ + struct mlxsw_sp_acl_erp *erp; + int err; + + erp = kzalloc(sizeof(*erp), GFP_KERNEL); + if (!erp) + return ERR_PTR(-ENOMEM); + + memcpy(&erp->key, key, sizeof(*key)); + bitmap_from_arr32(erp->mask_bitmap, (u32 *) key->mask, + MLXSW_SP_ACL_TCAM_MASK_LEN); + + err = mlxsw_sp_acl_erp_ctcam_inc(erp_table); + if (err) + goto err_erp_ctcam_inc; + + erp->erp_table = erp_table; + + err = mlxsw_sp_acl_erp_master_mask_set(erp_table, &erp->key); + if (err) + goto err_master_mask_set; + + err = mlxsw_sp_acl_erp_region_ctcam_enable(erp_table); + if (err) + goto err_erp_region_ctcam_enable; + + return erp; + +err_erp_region_ctcam_enable: + mlxsw_sp_acl_erp_master_mask_clear(erp_table, &erp->key); +err_master_mask_set: + mlxsw_sp_acl_erp_ctcam_dec(erp_table); +err_erp_ctcam_inc: + kfree(erp); + return ERR_PTR(err); +} + +static void +mlxsw_sp_acl_erp_ctcam_mask_destroy(struct mlxsw_sp_acl_erp *erp) +{ + struct mlxsw_sp_acl_erp_table *erp_table = erp->erp_table; + + mlxsw_sp_acl_erp_region_ctcam_disable(erp_table); + mlxsw_sp_acl_erp_master_mask_clear(erp_table, &erp->key); + mlxsw_sp_acl_erp_ctcam_dec(erp_table); + kfree(erp); +} + +static struct mlxsw_sp_acl_erp * +mlxsw_sp_acl_erp_mask_create(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp_key *key) +{ + struct mlxsw_sp_acl_erp *erp; + int err; + + if (key->ctcam) + return mlxsw_sp_acl_erp_ctcam_mask_create(erp_table, key); + + /* Expand the eRP table for the new eRP, if needed */ + err = mlxsw_sp_acl_erp_table_expand(erp_table); + if (err) + return ERR_PTR(err); + + erp = mlxsw_sp_acl_erp_generic_create(erp_table, key); + if (IS_ERR(erp)) + return erp; + + err = mlxsw_sp_acl_erp_index_get(erp_table, &erp->index); + if (err) + goto err_erp_index_get; + + err = mlxsw_sp_acl_erp_table_erp_add(erp_table, erp); + if (err) + goto err_table_erp_add; + + err = mlxsw_sp_acl_erp_region_erp_add(erp_table, erp); + if (err) + goto err_region_erp_add; + + erp_table->ops = &erp_multiple_masks_ops; + + return erp; + +err_region_erp_add: + mlxsw_sp_acl_erp_table_erp_del(erp); +err_table_erp_add: + mlxsw_sp_acl_erp_index_put(erp_table, erp->index); +err_erp_index_get: + mlxsw_sp_acl_erp_generic_destroy(erp); + return ERR_PTR(err); +} + +static void +mlxsw_sp_acl_erp_mask_destroy(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp) +{ + if (erp->key.ctcam) + return mlxsw_sp_acl_erp_ctcam_mask_destroy(erp); + + mlxsw_sp_acl_erp_region_erp_del(erp); + mlxsw_sp_acl_erp_table_erp_del(erp); + mlxsw_sp_acl_erp_index_put(erp_table, erp->index); + mlxsw_sp_acl_erp_generic_destroy(erp); + + if (erp_table->num_atcam_erps == 2 && erp_table->num_ctcam_erps == 0 && + erp_table->num_deltas == 0) + erp_table->ops = &erp_two_masks_ops; +} + +static struct mlxsw_sp_acl_erp * +mlxsw_sp_acl_erp_second_mask_create(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp_key *key) +{ + struct mlxsw_sp_acl_erp *erp; + int err; + + if (key->ctcam) + return mlxsw_sp_acl_erp_ctcam_mask_create(erp_table, key); + + /* Transition to use eRP table instead of master mask */ + err = mlxsw_sp_acl_erp_region_table_trans(erp_table); + if (err) + return ERR_PTR(err); + + erp = mlxsw_sp_acl_erp_generic_create(erp_table, key); + if (IS_ERR(erp)) { + err = PTR_ERR(erp); + goto err_erp_create; + } + + err = mlxsw_sp_acl_erp_index_get(erp_table, &erp->index); + if (err) + goto err_erp_index_get; + + err = mlxsw_sp_acl_erp_table_erp_add(erp_table, erp); + if (err) + goto err_table_erp_add; + + err = mlxsw_sp_acl_erp_region_erp_add(erp_table, erp); + if (err) + goto err_region_erp_add; + + erp_table->ops = &erp_two_masks_ops; + + return erp; + +err_region_erp_add: + mlxsw_sp_acl_erp_table_erp_del(erp); +err_table_erp_add: + mlxsw_sp_acl_erp_index_put(erp_table, erp->index); +err_erp_index_get: + mlxsw_sp_acl_erp_generic_destroy(erp); +err_erp_create: + mlxsw_sp_acl_erp_region_master_mask_trans(erp_table); + return ERR_PTR(err); +} + +static void +mlxsw_sp_acl_erp_second_mask_destroy(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp) +{ + if (erp->key.ctcam) + return mlxsw_sp_acl_erp_ctcam_mask_destroy(erp); + + mlxsw_sp_acl_erp_region_erp_del(erp); + mlxsw_sp_acl_erp_table_erp_del(erp); + mlxsw_sp_acl_erp_index_put(erp_table, erp->index); + mlxsw_sp_acl_erp_generic_destroy(erp); + /* Transition to use master mask instead of eRP table */ + mlxsw_sp_acl_erp_region_master_mask_trans(erp_table); + + erp_table->ops = &erp_single_mask_ops; +} + +static struct mlxsw_sp_acl_erp * +mlxsw_sp_acl_erp_first_mask_create(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp_key *key) +{ + struct mlxsw_sp_acl_erp *erp; + + if (key->ctcam) + return ERR_PTR(-EINVAL); + + erp = mlxsw_sp_acl_erp_generic_create(erp_table, key); + if (IS_ERR(erp)) + return erp; + + erp_table->ops = &erp_single_mask_ops; + + return erp; +} + +static void +mlxsw_sp_acl_erp_first_mask_destroy(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp) +{ + mlxsw_sp_acl_erp_generic_destroy(erp); + erp_table->ops = &erp_no_mask_ops; +} + +static void +mlxsw_sp_acl_erp_no_mask_destroy(struct mlxsw_sp_acl_erp_table *erp_table, + struct mlxsw_sp_acl_erp *erp) +{ + WARN_ON(1); +} + +struct mlxsw_sp_acl_erp_mask * +mlxsw_sp_acl_erp_mask_get(struct mlxsw_sp_acl_atcam_region *aregion, + const char *mask, bool ctcam) +{ + struct mlxsw_sp_acl_erp_table *erp_table = aregion->erp_table; + struct mlxsw_sp_acl_erp_key key; + struct objagg_obj *objagg_obj; + + memcpy(key.mask, mask, MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN); + key.ctcam = ctcam; + mutex_lock(&erp_table->objagg_lock); + objagg_obj = objagg_obj_get(erp_table->objagg, &key); + mutex_unlock(&erp_table->objagg_lock); + if (IS_ERR(objagg_obj)) + return ERR_CAST(objagg_obj); + return (struct mlxsw_sp_acl_erp_mask *) objagg_obj; +} + +void mlxsw_sp_acl_erp_mask_put(struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_erp_mask *erp_mask) +{ + struct objagg_obj *objagg_obj = (struct objagg_obj *) erp_mask; + struct mlxsw_sp_acl_erp_table *erp_table = aregion->erp_table; + + mutex_lock(&erp_table->objagg_lock); + objagg_obj_put(erp_table->objagg, objagg_obj); + mutex_unlock(&erp_table->objagg_lock); +} + +int mlxsw_sp_acl_erp_bf_insert(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_erp_mask *erp_mask, + struct mlxsw_sp_acl_atcam_entry *aentry) +{ + struct objagg_obj *objagg_obj = (struct objagg_obj *) erp_mask; + const struct mlxsw_sp_acl_erp *erp = objagg_obj_root_priv(objagg_obj); + unsigned int erp_bank; + + if (!mlxsw_sp_acl_erp_table_is_used(erp->erp_table)) + return 0; + + erp_bank = mlxsw_sp_acl_erp_bank_get(erp); + return mlxsw_sp_acl_bf_entry_add(mlxsw_sp, + erp->erp_table->erp_core->bf, + aregion, erp_bank, aentry); +} + +void mlxsw_sp_acl_erp_bf_remove(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_erp_mask *erp_mask, + struct mlxsw_sp_acl_atcam_entry *aentry) +{ + struct objagg_obj *objagg_obj = (struct objagg_obj *) erp_mask; + const struct mlxsw_sp_acl_erp *erp = objagg_obj_root_priv(objagg_obj); + unsigned int erp_bank; + + if (!mlxsw_sp_acl_erp_table_is_used(erp->erp_table)) + return; + + erp_bank = mlxsw_sp_acl_erp_bank_get(erp); + mlxsw_sp_acl_bf_entry_del(mlxsw_sp, + erp->erp_table->erp_core->bf, + aregion, erp_bank, aentry); +} + +bool +mlxsw_sp_acl_erp_mask_is_ctcam(const struct mlxsw_sp_acl_erp_mask *erp_mask) +{ + struct objagg_obj *objagg_obj = (struct objagg_obj *) erp_mask; + const struct mlxsw_sp_acl_erp_key *key = objagg_obj_raw(objagg_obj); + + return key->ctcam; +} + +u8 mlxsw_sp_acl_erp_mask_erp_id(const struct mlxsw_sp_acl_erp_mask *erp_mask) +{ + struct objagg_obj *objagg_obj = (struct objagg_obj *) erp_mask; + const struct mlxsw_sp_acl_erp *erp = objagg_obj_root_priv(objagg_obj); + + return erp->id; +} + +struct mlxsw_sp_acl_erp_delta { + struct mlxsw_sp_acl_erp_key key; + u16 start; + u8 mask; +}; + +u16 mlxsw_sp_acl_erp_delta_start(const struct mlxsw_sp_acl_erp_delta *delta) +{ + return delta->start; +} + +u8 mlxsw_sp_acl_erp_delta_mask(const struct mlxsw_sp_acl_erp_delta *delta) +{ + return delta->mask; +} + +u8 mlxsw_sp_acl_erp_delta_value(const struct mlxsw_sp_acl_erp_delta *delta, + const char *enc_key) +{ + u16 start = delta->start; + u8 mask = delta->mask; + u16 tmp; + + if (!mask) + return 0; + + tmp = (unsigned char) enc_key[__MASK_IDX(start / 8)]; + if (start / 8 + 1 < __MASK_LEN) + tmp |= (unsigned char) enc_key[__MASK_IDX(start / 8 + 1)] << 8; + tmp >>= start % 8; + tmp &= mask; + return tmp; +} + +void mlxsw_sp_acl_erp_delta_clear(const struct mlxsw_sp_acl_erp_delta *delta, + const char *enc_key) +{ + u16 start = delta->start; + u8 mask = delta->mask; + unsigned char *byte; + u16 tmp; + + tmp = mask; + tmp <<= start % 8; + tmp = ~tmp; + + byte = (unsigned char *) &enc_key[__MASK_IDX(start / 8)]; + *byte &= tmp & 0xff; + if (start / 8 + 1 < __MASK_LEN) { + byte = (unsigned char *) &enc_key[__MASK_IDX(start / 8 + 1)]; + *byte &= (tmp >> 8) & 0xff; + } +} + +static const struct mlxsw_sp_acl_erp_delta +mlxsw_sp_acl_erp_delta_default = {}; + +const struct mlxsw_sp_acl_erp_delta * +mlxsw_sp_acl_erp_delta(const struct mlxsw_sp_acl_erp_mask *erp_mask) +{ + struct objagg_obj *objagg_obj = (struct objagg_obj *) erp_mask; + const struct mlxsw_sp_acl_erp_delta *delta; + + delta = objagg_obj_delta_priv(objagg_obj); + if (!delta) + delta = &mlxsw_sp_acl_erp_delta_default; + return delta; +} + +static int +mlxsw_sp_acl_erp_delta_fill(const struct mlxsw_sp_acl_erp_key *parent_key, + const struct mlxsw_sp_acl_erp_key *key, + u16 *delta_start, u8 *delta_mask) +{ + int offset = 0; + int si = -1; + u16 pmask; + u16 mask; + int i; + + /* The difference between 2 masks can be up to 8 consecutive bits. */ + for (i = 0; i < __MASK_LEN; i++) { + if (parent_key->mask[__MASK_IDX(i)] == key->mask[__MASK_IDX(i)]) + continue; + if (si == -1) + si = i; + else if (si != i - 1) + return -EINVAL; + } + if (si == -1) { + /* The masks are the same, this can happen in case eRPs with + * the same mask were created in both A-TCAM and C-TCAM. + * The only possible condition under which this can happen + * is identical rule insertion. Delta is not possible here. + */ + return -EINVAL; + } + pmask = (unsigned char) parent_key->mask[__MASK_IDX(si)]; + mask = (unsigned char) key->mask[__MASK_IDX(si)]; + if (si + 1 < __MASK_LEN) { + pmask |= (unsigned char) parent_key->mask[__MASK_IDX(si + 1)] << 8; + mask |= (unsigned char) key->mask[__MASK_IDX(si + 1)] << 8; + } + + if ((pmask ^ mask) & pmask) + return -EINVAL; + mask &= ~pmask; + while (!(mask & (1 << offset))) + offset++; + while (!(mask & 1)) + mask >>= 1; + if (mask & 0xff00) + return -EINVAL; + + *delta_start = si * 8 + offset; + *delta_mask = mask; + + return 0; +} + +static bool mlxsw_sp_acl_erp_delta_check(void *priv, const void *parent_obj, + const void *obj) +{ + const struct mlxsw_sp_acl_erp_key *parent_key = parent_obj; + const struct mlxsw_sp_acl_erp_key *key = obj; + u16 delta_start; + u8 delta_mask; + int err; + + err = mlxsw_sp_acl_erp_delta_fill(parent_key, key, + &delta_start, &delta_mask); + return err ? false : true; +} + +static int mlxsw_sp_acl_erp_hints_obj_cmp(const void *obj1, const void *obj2) +{ + const struct mlxsw_sp_acl_erp_key *key1 = obj1; + const struct mlxsw_sp_acl_erp_key *key2 = obj2; + + /* For hints purposes, two objects are considered equal + * in case the masks are the same. Does not matter what + * the "ctcam" value is. + */ + return memcmp(key1->mask, key2->mask, sizeof(key1->mask)); +} + +static void *mlxsw_sp_acl_erp_delta_create(void *priv, void *parent_obj, + void *obj) +{ + struct mlxsw_sp_acl_erp_key *parent_key = parent_obj; + struct mlxsw_sp_acl_atcam_region *aregion = priv; + struct mlxsw_sp_acl_erp_table *erp_table = aregion->erp_table; + struct mlxsw_sp_acl_erp_key *key = obj; + struct mlxsw_sp_acl_erp_delta *delta; + u16 delta_start; + u8 delta_mask; + int err; + + if (parent_key->ctcam || key->ctcam) + return ERR_PTR(-EINVAL); + err = mlxsw_sp_acl_erp_delta_fill(parent_key, key, + &delta_start, &delta_mask); + if (err) + return ERR_PTR(-EINVAL); + + delta = kzalloc(sizeof(*delta), GFP_KERNEL); + if (!delta) + return ERR_PTR(-ENOMEM); + delta->start = delta_start; + delta->mask = delta_mask; + + err = mlxsw_sp_acl_erp_delta_inc(erp_table); + if (err) + goto err_erp_delta_inc; + + memcpy(&delta->key, key, sizeof(*key)); + err = mlxsw_sp_acl_erp_master_mask_set(erp_table, &delta->key); + if (err) + goto err_master_mask_set; + + return delta; + +err_master_mask_set: + mlxsw_sp_acl_erp_delta_dec(erp_table); +err_erp_delta_inc: + kfree(delta); + return ERR_PTR(err); +} + +static void mlxsw_sp_acl_erp_delta_destroy(void *priv, void *delta_priv) +{ + struct mlxsw_sp_acl_erp_delta *delta = delta_priv; + struct mlxsw_sp_acl_atcam_region *aregion = priv; + struct mlxsw_sp_acl_erp_table *erp_table = aregion->erp_table; + + mlxsw_sp_acl_erp_master_mask_clear(erp_table, &delta->key); + mlxsw_sp_acl_erp_delta_dec(erp_table); + kfree(delta); +} + +static void *mlxsw_sp_acl_erp_root_create(void *priv, void *obj, + unsigned int root_id) +{ + struct mlxsw_sp_acl_atcam_region *aregion = priv; + struct mlxsw_sp_acl_erp_table *erp_table = aregion->erp_table; + struct mlxsw_sp_acl_erp_key *key = obj; + + if (!key->ctcam && + root_id != OBJAGG_OBJ_ROOT_ID_INVALID && + root_id >= MLXSW_SP_ACL_ERP_MAX_PER_REGION) + return ERR_PTR(-ENOBUFS); + return erp_table->ops->erp_create(erp_table, key); +} + +static void mlxsw_sp_acl_erp_root_destroy(void *priv, void *root_priv) +{ + struct mlxsw_sp_acl_atcam_region *aregion = priv; + struct mlxsw_sp_acl_erp_table *erp_table = aregion->erp_table; + + erp_table->ops->erp_destroy(erp_table, root_priv); +} + +static const struct objagg_ops mlxsw_sp_acl_erp_objagg_ops = { + .obj_size = sizeof(struct mlxsw_sp_acl_erp_key), + .delta_check = mlxsw_sp_acl_erp_delta_check, + .hints_obj_cmp = mlxsw_sp_acl_erp_hints_obj_cmp, + .delta_create = mlxsw_sp_acl_erp_delta_create, + .delta_destroy = mlxsw_sp_acl_erp_delta_destroy, + .root_create = mlxsw_sp_acl_erp_root_create, + .root_destroy = mlxsw_sp_acl_erp_root_destroy, +}; + +static struct mlxsw_sp_acl_erp_table * +mlxsw_sp_acl_erp_table_create(struct mlxsw_sp_acl_atcam_region *aregion, + struct objagg_hints *hints) +{ + struct mlxsw_sp_acl_erp_table *erp_table; + int err; + + erp_table = kzalloc(sizeof(*erp_table), GFP_KERNEL); + if (!erp_table) + return ERR_PTR(-ENOMEM); + + erp_table->objagg = objagg_create(&mlxsw_sp_acl_erp_objagg_ops, + hints, aregion); + if (IS_ERR(erp_table->objagg)) { + err = PTR_ERR(erp_table->objagg); + goto err_objagg_create; + } + + erp_table->erp_core = aregion->atcam->erp_core; + erp_table->ops = &erp_no_mask_ops; + INIT_LIST_HEAD(&erp_table->atcam_erps_list); + erp_table->aregion = aregion; + mutex_init(&erp_table->objagg_lock); + + return erp_table; + +err_objagg_create: + kfree(erp_table); + return ERR_PTR(err); +} + +static void +mlxsw_sp_acl_erp_table_destroy(struct mlxsw_sp_acl_erp_table *erp_table) +{ + WARN_ON(!list_empty(&erp_table->atcam_erps_list)); + mutex_destroy(&erp_table->objagg_lock); + objagg_destroy(erp_table->objagg); + kfree(erp_table); +} + +static int +mlxsw_sp_acl_erp_master_mask_init(struct mlxsw_sp_acl_atcam_region *aregion) +{ + struct mlxsw_sp *mlxsw_sp = aregion->region->mlxsw_sp; + char percr_pl[MLXSW_REG_PERCR_LEN]; + + mlxsw_reg_percr_pack(percr_pl, aregion->region->id); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(percr), percr_pl); +} + +static int +mlxsw_sp_acl_erp_region_param_init(struct mlxsw_sp_acl_atcam_region *aregion) +{ + struct mlxsw_sp *mlxsw_sp = aregion->region->mlxsw_sp; + char pererp_pl[MLXSW_REG_PERERP_LEN]; + + mlxsw_reg_pererp_pack(pererp_pl, aregion->region->id, false, false, 0, + 0, 0); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pererp), pererp_pl); +} + +static int +mlxsw_sp_acl_erp_hints_check(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct objagg_hints *hints, bool *p_rehash_needed) +{ + struct mlxsw_sp_acl_erp_table *erp_table = aregion->erp_table; + const struct objagg_stats *ostats; + const struct objagg_stats *hstats; + int err; + + *p_rehash_needed = false; + + mutex_lock(&erp_table->objagg_lock); + ostats = objagg_stats_get(erp_table->objagg); + mutex_unlock(&erp_table->objagg_lock); + if (IS_ERR(ostats)) { + dev_err_ratelimited(mlxsw_sp->bus_info->dev, "Failed to get ERP stats\n"); + return PTR_ERR(ostats); + } + + hstats = objagg_hints_stats_get(hints); + if (IS_ERR(hstats)) { + dev_err_ratelimited(mlxsw_sp->bus_info->dev, "Failed to get ERP hints stats\n"); + err = PTR_ERR(hstats); + goto err_hints_stats_get; + } + + /* Very basic criterion for now. */ + if (hstats->root_count < ostats->root_count) + *p_rehash_needed = true; + + err = 0; + + objagg_stats_put(hstats); +err_hints_stats_get: + objagg_stats_put(ostats); + return err; +} + +void * +mlxsw_sp_acl_erp_rehash_hints_get(struct mlxsw_sp_acl_atcam_region *aregion) +{ + struct mlxsw_sp_acl_erp_table *erp_table = aregion->erp_table; + struct mlxsw_sp *mlxsw_sp = aregion->region->mlxsw_sp; + struct objagg_hints *hints; + bool rehash_needed; + int err; + + mutex_lock(&erp_table->objagg_lock); + hints = objagg_hints_get(erp_table->objagg, + OBJAGG_OPT_ALGO_SIMPLE_GREEDY); + mutex_unlock(&erp_table->objagg_lock); + if (IS_ERR(hints)) { + dev_err_ratelimited(mlxsw_sp->bus_info->dev, "Failed to create ERP hints\n"); + return ERR_CAST(hints); + } + err = mlxsw_sp_acl_erp_hints_check(mlxsw_sp, aregion, hints, + &rehash_needed); + if (err) + goto errout; + + if (!rehash_needed) { + err = -EAGAIN; + goto errout; + } + return hints; + +errout: + objagg_hints_put(hints); + return ERR_PTR(err); +} + +void mlxsw_sp_acl_erp_rehash_hints_put(void *hints_priv) +{ + struct objagg_hints *hints = hints_priv; + + objagg_hints_put(hints); +} + +int mlxsw_sp_acl_erp_region_init(struct mlxsw_sp_acl_atcam_region *aregion, + void *hints_priv) +{ + struct mlxsw_sp_acl_erp_table *erp_table; + struct objagg_hints *hints = hints_priv; + int err; + + erp_table = mlxsw_sp_acl_erp_table_create(aregion, hints); + if (IS_ERR(erp_table)) + return PTR_ERR(erp_table); + aregion->erp_table = erp_table; + + /* Initialize the region's master mask to all zeroes */ + err = mlxsw_sp_acl_erp_master_mask_init(aregion); + if (err) + goto err_erp_master_mask_init; + + /* Initialize the region to not use the eRP table */ + err = mlxsw_sp_acl_erp_region_param_init(aregion); + if (err) + goto err_erp_region_param_init; + + return 0; + +err_erp_region_param_init: +err_erp_master_mask_init: + mlxsw_sp_acl_erp_table_destroy(erp_table); + return err; +} + +void mlxsw_sp_acl_erp_region_fini(struct mlxsw_sp_acl_atcam_region *aregion) +{ + mlxsw_sp_acl_erp_table_destroy(aregion->erp_table); +} + +static int +mlxsw_sp_acl_erp_tables_sizes_query(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_erp_core *erp_core) +{ + unsigned int size; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, ACL_ERPT_ENTRIES_2KB) || + !MLXSW_CORE_RES_VALID(mlxsw_sp->core, ACL_ERPT_ENTRIES_4KB) || + !MLXSW_CORE_RES_VALID(mlxsw_sp->core, ACL_ERPT_ENTRIES_8KB) || + !MLXSW_CORE_RES_VALID(mlxsw_sp->core, ACL_ERPT_ENTRIES_12KB)) + return -EIO; + + size = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_ERPT_ENTRIES_2KB); + erp_core->erpt_entries_size[MLXSW_SP_ACL_ATCAM_REGION_TYPE_2KB] = size; + + size = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_ERPT_ENTRIES_4KB); + erp_core->erpt_entries_size[MLXSW_SP_ACL_ATCAM_REGION_TYPE_4KB] = size; + + size = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_ERPT_ENTRIES_8KB); + erp_core->erpt_entries_size[MLXSW_SP_ACL_ATCAM_REGION_TYPE_8KB] = size; + + size = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_ERPT_ENTRIES_12KB); + erp_core->erpt_entries_size[MLXSW_SP_ACL_ATCAM_REGION_TYPE_12KB] = size; + + return 0; +} + +static int mlxsw_sp_acl_erp_tables_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_erp_core *erp_core) +{ + unsigned int erpt_bank_size; + int err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, ACL_MAX_ERPT_BANK_SIZE) || + !MLXSW_CORE_RES_VALID(mlxsw_sp->core, ACL_MAX_ERPT_BANKS)) + return -EIO; + erpt_bank_size = MLXSW_CORE_RES_GET(mlxsw_sp->core, + ACL_MAX_ERPT_BANK_SIZE); + erp_core->num_erp_banks = MLXSW_CORE_RES_GET(mlxsw_sp->core, + ACL_MAX_ERPT_BANKS); + + erp_core->erp_tables = gen_pool_create(0, -1); + if (!erp_core->erp_tables) + return -ENOMEM; + gen_pool_set_algo(erp_core->erp_tables, gen_pool_best_fit, NULL); + + err = gen_pool_add(erp_core->erp_tables, + MLXSW_SP_ACL_ERP_GENALLOC_OFFSET, erpt_bank_size, + -1); + if (err) + goto err_gen_pool_add; + + erp_core->bf = mlxsw_sp_acl_bf_init(mlxsw_sp, erp_core->num_erp_banks); + if (IS_ERR(erp_core->bf)) { + err = PTR_ERR(erp_core->bf); + goto err_bf_init; + } + + /* Different regions require masks of different sizes */ + err = mlxsw_sp_acl_erp_tables_sizes_query(mlxsw_sp, erp_core); + if (err) + goto err_erp_tables_sizes_query; + + return 0; + +err_erp_tables_sizes_query: + mlxsw_sp_acl_bf_fini(erp_core->bf); +err_bf_init: +err_gen_pool_add: + gen_pool_destroy(erp_core->erp_tables); + return err; +} + +static void mlxsw_sp_acl_erp_tables_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_erp_core *erp_core) +{ + mlxsw_sp_acl_bf_fini(erp_core->bf); + gen_pool_destroy(erp_core->erp_tables); +} + +int mlxsw_sp_acl_erps_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam *atcam) +{ + struct mlxsw_sp_acl_erp_core *erp_core; + int err; + + erp_core = kzalloc(sizeof(*erp_core), GFP_KERNEL); + if (!erp_core) + return -ENOMEM; + erp_core->mlxsw_sp = mlxsw_sp; + atcam->erp_core = erp_core; + + err = mlxsw_sp_acl_erp_tables_init(mlxsw_sp, erp_core); + if (err) + goto err_erp_tables_init; + + return 0; + +err_erp_tables_init: + kfree(erp_core); + return err; +} + +void mlxsw_sp_acl_erps_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam *atcam) +{ + mlxsw_sp_acl_erp_tables_fini(mlxsw_sp, atcam->erp_core); + kfree(atcam->erp_core); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_actions.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_actions.c new file mode 100644 index 000000000..50806594d --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_actions.c @@ -0,0 +1,318 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include "spectrum_acl_flex_actions.h" +#include "core_acl_flex_actions.h" +#include "spectrum_span.h" + +static int mlxsw_sp_act_kvdl_set_add(void *priv, u32 *p_kvdl_index, + char *enc_actions, bool is_first, bool ca) +{ + struct mlxsw_sp *mlxsw_sp = priv; + char pefa_pl[MLXSW_REG_PEFA_LEN]; + u32 kvdl_index; + int err; + + /* The first action set of a TCAM entry is stored directly in TCAM, + * not KVD linear area. + */ + if (is_first) + return 0; + + err = mlxsw_sp_kvdl_alloc(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ACTSET, + 1, &kvdl_index); + if (err) + return err; + mlxsw_reg_pefa_pack(pefa_pl, kvdl_index, ca, enc_actions); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pefa), pefa_pl); + if (err) + goto err_pefa_write; + *p_kvdl_index = kvdl_index; + return 0; + +err_pefa_write: + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ACTSET, + 1, kvdl_index); + return err; +} + +static int mlxsw_sp1_act_kvdl_set_add(void *priv, u32 *p_kvdl_index, + char *enc_actions, bool is_first) +{ + return mlxsw_sp_act_kvdl_set_add(priv, p_kvdl_index, enc_actions, + is_first, false); +} + +static int mlxsw_sp2_act_kvdl_set_add(void *priv, u32 *p_kvdl_index, + char *enc_actions, bool is_first) +{ + return mlxsw_sp_act_kvdl_set_add(priv, p_kvdl_index, enc_actions, + is_first, true); +} + +static void mlxsw_sp_act_kvdl_set_del(void *priv, u32 kvdl_index, + bool is_first) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + if (is_first) + return; + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ACTSET, + 1, kvdl_index); +} + +static int mlxsw_sp1_act_kvdl_set_activity_get(void *priv, u32 kvdl_index, + bool *activity) +{ + return -EOPNOTSUPP; +} + +static int mlxsw_sp2_act_kvdl_set_activity_get(void *priv, u32 kvdl_index, + bool *activity) +{ + struct mlxsw_sp *mlxsw_sp = priv; + char pefa_pl[MLXSW_REG_PEFA_LEN]; + int err; + + mlxsw_reg_pefa_pack(pefa_pl, kvdl_index, true, NULL); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(pefa), pefa_pl); + if (err) + return err; + mlxsw_reg_pefa_unpack(pefa_pl, activity); + return 0; +} + +static int mlxsw_sp_act_kvdl_fwd_entry_add(void *priv, u32 *p_kvdl_index, + u16 local_port) +{ + struct mlxsw_sp *mlxsw_sp = priv; + char ppbs_pl[MLXSW_REG_PPBS_LEN]; + u32 kvdl_index; + int err; + + err = mlxsw_sp_kvdl_alloc(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_PBS, + 1, &kvdl_index); + if (err) + return err; + mlxsw_reg_ppbs_pack(ppbs_pl, kvdl_index, local_port); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ppbs), ppbs_pl); + if (err) + goto err_ppbs_write; + *p_kvdl_index = kvdl_index; + return 0; + +err_ppbs_write: + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_PBS, + 1, kvdl_index); + return err; +} + +static void mlxsw_sp_act_kvdl_fwd_entry_del(void *priv, u32 kvdl_index) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_PBS, + 1, kvdl_index); +} + +static int +mlxsw_sp_act_counter_index_get(void *priv, unsigned int *p_counter_index) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + return mlxsw_sp_flow_counter_alloc(mlxsw_sp, p_counter_index); +} + +static void +mlxsw_sp_act_counter_index_put(void *priv, unsigned int counter_index) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + mlxsw_sp_flow_counter_free(mlxsw_sp, counter_index); +} + +static int +mlxsw_sp_act_mirror_add(void *priv, u16 local_in_port, + const struct net_device *out_dev, + bool ingress, int *p_span_id) +{ + struct mlxsw_sp_span_agent_parms agent_parms = {}; + struct mlxsw_sp_port *mlxsw_sp_port; + struct mlxsw_sp *mlxsw_sp = priv; + int err; + + agent_parms.to_dev = out_dev; + err = mlxsw_sp_span_agent_get(mlxsw_sp, p_span_id, &agent_parms); + if (err) + return err; + + mlxsw_sp_port = mlxsw_sp->ports[local_in_port]; + err = mlxsw_sp_span_analyzed_port_get(mlxsw_sp_port, ingress); + if (err) + goto err_analyzed_port_get; + + return 0; + +err_analyzed_port_get: + mlxsw_sp_span_agent_put(mlxsw_sp, *p_span_id); + return err; +} + +static void +mlxsw_sp_act_mirror_del(void *priv, u16 local_in_port, int span_id, bool ingress) +{ + struct mlxsw_sp_port *mlxsw_sp_port; + struct mlxsw_sp *mlxsw_sp = priv; + + mlxsw_sp_port = mlxsw_sp->ports[local_in_port]; + mlxsw_sp_span_analyzed_port_put(mlxsw_sp_port, ingress); + mlxsw_sp_span_agent_put(mlxsw_sp, span_id); +} + +static int mlxsw_sp_act_policer_add(void *priv, u64 rate_bytes_ps, u32 burst, + u16 *p_policer_index, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_policer_params params; + struct mlxsw_sp *mlxsw_sp = priv; + + params.rate = rate_bytes_ps; + params.burst = burst; + params.bytes = true; + return mlxsw_sp_policer_add(mlxsw_sp, + MLXSW_SP_POLICER_TYPE_SINGLE_RATE, + ¶ms, extack, p_policer_index); +} + +static void mlxsw_sp_act_policer_del(void *priv, u16 policer_index) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + mlxsw_sp_policer_del(mlxsw_sp, MLXSW_SP_POLICER_TYPE_SINGLE_RATE, + policer_index); +} + +static int mlxsw_sp1_act_sampler_add(void *priv, u16 local_port, + struct psample_group *psample_group, + u32 rate, u32 trunc_size, bool truncate, + bool ingress, int *p_span_id, + struct netlink_ext_ack *extack) +{ + NL_SET_ERR_MSG_MOD(extack, "Sampling action is not supported on Spectrum-1"); + return -EOPNOTSUPP; +} + +static void mlxsw_sp1_act_sampler_del(void *priv, u16 local_port, int span_id, + bool ingress) +{ + WARN_ON_ONCE(1); +} + +const struct mlxsw_afa_ops mlxsw_sp1_act_afa_ops = { + .kvdl_set_add = mlxsw_sp1_act_kvdl_set_add, + .kvdl_set_del = mlxsw_sp_act_kvdl_set_del, + .kvdl_set_activity_get = mlxsw_sp1_act_kvdl_set_activity_get, + .kvdl_fwd_entry_add = mlxsw_sp_act_kvdl_fwd_entry_add, + .kvdl_fwd_entry_del = mlxsw_sp_act_kvdl_fwd_entry_del, + .counter_index_get = mlxsw_sp_act_counter_index_get, + .counter_index_put = mlxsw_sp_act_counter_index_put, + .mirror_add = mlxsw_sp_act_mirror_add, + .mirror_del = mlxsw_sp_act_mirror_del, + .policer_add = mlxsw_sp_act_policer_add, + .policer_del = mlxsw_sp_act_policer_del, + .sampler_add = mlxsw_sp1_act_sampler_add, + .sampler_del = mlxsw_sp1_act_sampler_del, +}; + +static int mlxsw_sp2_act_sampler_add(void *priv, u16 local_port, + struct psample_group *psample_group, + u32 rate, u32 trunc_size, bool truncate, + bool ingress, int *p_span_id, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_span_agent_parms agent_parms = { + .session_id = MLXSW_SP_SPAN_SESSION_ID_SAMPLING, + }; + struct mlxsw_sp_sample_trigger trigger = { + .type = MLXSW_SP_SAMPLE_TRIGGER_TYPE_POLICY_ENGINE, + }; + struct mlxsw_sp_sample_params params; + struct mlxsw_sp_port *mlxsw_sp_port; + struct mlxsw_sp *mlxsw_sp = priv; + int err; + + params.psample_group = psample_group; + params.trunc_size = trunc_size; + params.rate = rate; + params.truncate = truncate; + err = mlxsw_sp_sample_trigger_params_set(mlxsw_sp, &trigger, ¶ms, + extack); + if (err) + return err; + + err = mlxsw_sp_span_agent_get(mlxsw_sp, p_span_id, &agent_parms); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to get SPAN agent"); + goto err_span_agent_get; + } + + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + err = mlxsw_sp_span_analyzed_port_get(mlxsw_sp_port, ingress); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to get analyzed port"); + goto err_analyzed_port_get; + } + + return 0; + +err_analyzed_port_get: + mlxsw_sp_span_agent_put(mlxsw_sp, *p_span_id); +err_span_agent_get: + mlxsw_sp_sample_trigger_params_unset(mlxsw_sp, &trigger); + return err; +} + +static void mlxsw_sp2_act_sampler_del(void *priv, u16 local_port, int span_id, + bool ingress) +{ + struct mlxsw_sp_sample_trigger trigger = { + .type = MLXSW_SP_SAMPLE_TRIGGER_TYPE_POLICY_ENGINE, + }; + struct mlxsw_sp_port *mlxsw_sp_port; + struct mlxsw_sp *mlxsw_sp = priv; + + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + mlxsw_sp_span_analyzed_port_put(mlxsw_sp_port, ingress); + mlxsw_sp_span_agent_put(mlxsw_sp, span_id); + mlxsw_sp_sample_trigger_params_unset(mlxsw_sp, &trigger); +} + +const struct mlxsw_afa_ops mlxsw_sp2_act_afa_ops = { + .kvdl_set_add = mlxsw_sp2_act_kvdl_set_add, + .kvdl_set_del = mlxsw_sp_act_kvdl_set_del, + .kvdl_set_activity_get = mlxsw_sp2_act_kvdl_set_activity_get, + .kvdl_fwd_entry_add = mlxsw_sp_act_kvdl_fwd_entry_add, + .kvdl_fwd_entry_del = mlxsw_sp_act_kvdl_fwd_entry_del, + .counter_index_get = mlxsw_sp_act_counter_index_get, + .counter_index_put = mlxsw_sp_act_counter_index_put, + .mirror_add = mlxsw_sp_act_mirror_add, + .mirror_del = mlxsw_sp_act_mirror_del, + .policer_add = mlxsw_sp_act_policer_add, + .policer_del = mlxsw_sp_act_policer_del, + .sampler_add = mlxsw_sp2_act_sampler_add, + .sampler_del = mlxsw_sp2_act_sampler_del, + .dummy_first_set = true, +}; + +int mlxsw_sp_afa_init(struct mlxsw_sp *mlxsw_sp) +{ + mlxsw_sp->afa = mlxsw_afa_create(MLXSW_CORE_RES_GET(mlxsw_sp->core, + ACL_ACTIONS_PER_SET), + mlxsw_sp->afa_ops, mlxsw_sp); + return PTR_ERR_OR_ZERO(mlxsw_sp->afa); +} + +void mlxsw_sp_afa_fini(struct mlxsw_sp *mlxsw_sp) +{ + mlxsw_afa_destroy(mlxsw_sp->afa); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_actions.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_actions.h new file mode 100644 index 000000000..fe436d816 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_actions.h @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_SPECTRUM_ACL_FLEX_ACTIONS_H +#define _MLXSW_SPECTRUM_ACL_FLEX_ACTIONS_H + +#include "spectrum.h" + +int mlxsw_sp_afa_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_afa_fini(struct mlxsw_sp *mlxsw_sp); + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_keys.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_keys.c new file mode 100644 index 000000000..173808c09 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_keys.c @@ -0,0 +1,355 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/module.h> +#include "spectrum.h" +#include "item.h" +#include "core_acl_flex_keys.h" + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_l2_dmac[] = { + MLXSW_AFK_ELEMENT_INST_BUF(DMAC_32_47, 0x00, 2), + MLXSW_AFK_ELEMENT_INST_BUF(DMAC_0_31, 0x02, 4), + MLXSW_AFK_ELEMENT_INST_U32(PCP, 0x08, 13, 3), + MLXSW_AFK_ELEMENT_INST_U32(VID, 0x08, 0, 12), + MLXSW_AFK_ELEMENT_INST_U32(SRC_SYS_PORT, 0x0C, 0, 16), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_l2_smac[] = { + MLXSW_AFK_ELEMENT_INST_BUF(SMAC_32_47, 0x00, 2), + MLXSW_AFK_ELEMENT_INST_BUF(SMAC_0_31, 0x02, 4), + MLXSW_AFK_ELEMENT_INST_U32(PCP, 0x08, 13, 3), + MLXSW_AFK_ELEMENT_INST_U32(VID, 0x08, 0, 12), + MLXSW_AFK_ELEMENT_INST_U32(SRC_SYS_PORT, 0x0C, 0, 16), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_l2_smac_ex[] = { + MLXSW_AFK_ELEMENT_INST_BUF(SMAC_32_47, 0x02, 2), + MLXSW_AFK_ELEMENT_INST_BUF(SMAC_0_31, 0x04, 4), + MLXSW_AFK_ELEMENT_INST_U32(ETHERTYPE, 0x0C, 0, 16), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv4_sip[] = { + MLXSW_AFK_ELEMENT_INST_BUF(SRC_IP_0_31, 0x00, 4), + MLXSW_AFK_ELEMENT_INST_U32(IP_PROTO, 0x08, 0, 8), + MLXSW_AFK_ELEMENT_INST_U32(SRC_SYS_PORT, 0x0C, 0, 16), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv4_dip[] = { + MLXSW_AFK_ELEMENT_INST_BUF(DST_IP_0_31, 0x00, 4), + MLXSW_AFK_ELEMENT_INST_U32(IP_PROTO, 0x08, 0, 8), + MLXSW_AFK_ELEMENT_INST_U32(SRC_SYS_PORT, 0x0C, 0, 16), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv4[] = { + MLXSW_AFK_ELEMENT_INST_BUF(SRC_IP_0_31, 0x00, 4), + MLXSW_AFK_ELEMENT_INST_U32(IP_ECN, 0x04, 4, 2), + MLXSW_AFK_ELEMENT_INST_U32(IP_TTL_, 0x04, 24, 8), + MLXSW_AFK_ELEMENT_INST_U32(IP_DSCP, 0x08, 0, 6), + MLXSW_AFK_ELEMENT_INST_U32(TCP_FLAGS, 0x08, 8, 9), /* TCP_CONTROL+TCP_ECN */ +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv4_ex[] = { + MLXSW_AFK_ELEMENT_INST_U32(VID, 0x00, 0, 12), + MLXSW_AFK_ELEMENT_INST_U32(PCP, 0x08, 29, 3), + MLXSW_AFK_ELEMENT_INST_U32(SRC_L4_PORT, 0x08, 0, 16), + MLXSW_AFK_ELEMENT_INST_U32(DST_L4_PORT, 0x0C, 0, 16), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv6_dip[] = { + MLXSW_AFK_ELEMENT_INST_BUF(DST_IP_32_63, 0x00, 4), + MLXSW_AFK_ELEMENT_INST_BUF(DST_IP_0_31, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv6_ex1[] = { + MLXSW_AFK_ELEMENT_INST_BUF(DST_IP_96_127, 0x00, 4), + MLXSW_AFK_ELEMENT_INST_BUF(DST_IP_64_95, 0x04, 4), + MLXSW_AFK_ELEMENT_INST_U32(IP_PROTO, 0x08, 0, 8), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv6_sip[] = { + MLXSW_AFK_ELEMENT_INST_BUF(SRC_IP_32_63, 0x00, 4), + MLXSW_AFK_ELEMENT_INST_BUF(SRC_IP_0_31, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv6_sip_ex[] = { + MLXSW_AFK_ELEMENT_INST_BUF(SRC_IP_96_127, 0x00, 4), + MLXSW_AFK_ELEMENT_INST_BUF(SRC_IP_64_95, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_packet_type[] = { + MLXSW_AFK_ELEMENT_INST_U32(ETHERTYPE, 0x00, 0, 16), +}; + +static const struct mlxsw_afk_block mlxsw_sp1_afk_blocks[] = { + MLXSW_AFK_BLOCK(0x10, mlxsw_sp_afk_element_info_l2_dmac), + MLXSW_AFK_BLOCK(0x11, mlxsw_sp_afk_element_info_l2_smac), + MLXSW_AFK_BLOCK(0x12, mlxsw_sp_afk_element_info_l2_smac_ex), + MLXSW_AFK_BLOCK(0x30, mlxsw_sp_afk_element_info_ipv4_sip), + MLXSW_AFK_BLOCK(0x31, mlxsw_sp_afk_element_info_ipv4_dip), + MLXSW_AFK_BLOCK(0x32, mlxsw_sp_afk_element_info_ipv4), + MLXSW_AFK_BLOCK(0x33, mlxsw_sp_afk_element_info_ipv4_ex), + MLXSW_AFK_BLOCK(0x60, mlxsw_sp_afk_element_info_ipv6_dip), + MLXSW_AFK_BLOCK(0x65, mlxsw_sp_afk_element_info_ipv6_ex1), + MLXSW_AFK_BLOCK(0x62, mlxsw_sp_afk_element_info_ipv6_sip), + MLXSW_AFK_BLOCK(0x63, mlxsw_sp_afk_element_info_ipv6_sip_ex), + MLXSW_AFK_BLOCK(0xB0, mlxsw_sp_afk_element_info_packet_type), +}; + +#define MLXSW_SP1_AFK_KEY_BLOCK_SIZE 16 + +static void mlxsw_sp1_afk_encode_block(char *output, int block_index, + char *block) +{ + unsigned int offset = block_index * MLXSW_SP1_AFK_KEY_BLOCK_SIZE; + char *output_indexed = output + offset; + + memcpy(output_indexed, block, MLXSW_SP1_AFK_KEY_BLOCK_SIZE); +} + +static void mlxsw_sp1_afk_clear_block(char *output, int block_index) +{ + unsigned int offset = block_index * MLXSW_SP1_AFK_KEY_BLOCK_SIZE; + char *output_indexed = output + offset; + + memset(output_indexed, 0, MLXSW_SP1_AFK_KEY_BLOCK_SIZE); +} + +const struct mlxsw_afk_ops mlxsw_sp1_afk_ops = { + .blocks = mlxsw_sp1_afk_blocks, + .blocks_count = ARRAY_SIZE(mlxsw_sp1_afk_blocks), + .encode_block = mlxsw_sp1_afk_encode_block, + .clear_block = mlxsw_sp1_afk_clear_block, +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_mac_0[] = { + MLXSW_AFK_ELEMENT_INST_BUF(DMAC_0_31, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_mac_1[] = { + MLXSW_AFK_ELEMENT_INST_BUF(SMAC_0_31, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_mac_2[] = { + MLXSW_AFK_ELEMENT_INST_BUF(SMAC_32_47, 0x04, 2), + MLXSW_AFK_ELEMENT_INST_BUF(DMAC_32_47, 0x06, 2), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_mac_3[] = { + MLXSW_AFK_ELEMENT_INST_U32(PCP, 0x00, 0, 3), + MLXSW_AFK_ELEMENT_INST_U32(VID, 0x04, 16, 12), + MLXSW_AFK_ELEMENT_INST_BUF(DMAC_32_47, 0x06, 2), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_mac_4[] = { + MLXSW_AFK_ELEMENT_INST_U32(PCP, 0x00, 0, 3), + MLXSW_AFK_ELEMENT_INST_U32(VID, 0x04, 16, 12), + MLXSW_AFK_ELEMENT_INST_U32(ETHERTYPE, 0x04, 0, 16), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_mac_5[] = { + MLXSW_AFK_ELEMENT_INST_U32(VID, 0x04, 16, 12), + MLXSW_AFK_ELEMENT_INST_EXT_U32(SRC_SYS_PORT, 0x04, 0, 8, -1, true), /* RX_ACL_SYSTEM_PORT */ +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv4_0[] = { + MLXSW_AFK_ELEMENT_INST_BUF(DST_IP_0_31, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv4_1[] = { + MLXSW_AFK_ELEMENT_INST_BUF(SRC_IP_0_31, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv4_2[] = { + MLXSW_AFK_ELEMENT_INST_U32(IP_DSCP, 0x04, 0, 6), + MLXSW_AFK_ELEMENT_INST_U32(IP_ECN, 0x04, 6, 2), + MLXSW_AFK_ELEMENT_INST_U32(IP_TTL_, 0x04, 8, 8), + MLXSW_AFK_ELEMENT_INST_U32(IP_PROTO, 0x04, 16, 8), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv4_4[] = { + MLXSW_AFK_ELEMENT_INST_U32(VIRT_ROUTER_LSB, 0x04, 24, 8), + MLXSW_AFK_ELEMENT_INST_EXT_U32(VIRT_ROUTER_MSB, 0x00, 0, 3, 0, true), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv6_0[] = { + MLXSW_AFK_ELEMENT_INST_BUF(DST_IP_32_63, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv6_1[] = { + MLXSW_AFK_ELEMENT_INST_BUF(DST_IP_64_95, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv6_2[] = { + MLXSW_AFK_ELEMENT_INST_BUF(DST_IP_96_127, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv6_3[] = { + MLXSW_AFK_ELEMENT_INST_BUF(SRC_IP_32_63, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv6_4[] = { + MLXSW_AFK_ELEMENT_INST_BUF(SRC_IP_64_95, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv6_5[] = { + MLXSW_AFK_ELEMENT_INST_BUF(SRC_IP_96_127, 0x04, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_l4_0[] = { + MLXSW_AFK_ELEMENT_INST_U32(SRC_L4_PORT, 0x04, 16, 16), + MLXSW_AFK_ELEMENT_INST_U32(DST_L4_PORT, 0x04, 0, 16), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_l4_2[] = { + MLXSW_AFK_ELEMENT_INST_U32(TCP_FLAGS, 0x04, 16, 9), /* TCP_CONTROL + TCP_ECN */ +}; + +static const struct mlxsw_afk_block mlxsw_sp2_afk_blocks[] = { + MLXSW_AFK_BLOCK(0x10, mlxsw_sp_afk_element_info_mac_0), + MLXSW_AFK_BLOCK(0x11, mlxsw_sp_afk_element_info_mac_1), + MLXSW_AFK_BLOCK(0x12, mlxsw_sp_afk_element_info_mac_2), + MLXSW_AFK_BLOCK(0x13, mlxsw_sp_afk_element_info_mac_3), + MLXSW_AFK_BLOCK(0x14, mlxsw_sp_afk_element_info_mac_4), + MLXSW_AFK_BLOCK(0x15, mlxsw_sp_afk_element_info_mac_5), + MLXSW_AFK_BLOCK(0x38, mlxsw_sp_afk_element_info_ipv4_0), + MLXSW_AFK_BLOCK(0x39, mlxsw_sp_afk_element_info_ipv4_1), + MLXSW_AFK_BLOCK(0x3A, mlxsw_sp_afk_element_info_ipv4_2), + MLXSW_AFK_BLOCK(0x3C, mlxsw_sp_afk_element_info_ipv4_4), + MLXSW_AFK_BLOCK(0x40, mlxsw_sp_afk_element_info_ipv6_0), + MLXSW_AFK_BLOCK(0x41, mlxsw_sp_afk_element_info_ipv6_1), + MLXSW_AFK_BLOCK(0x42, mlxsw_sp_afk_element_info_ipv6_2), + MLXSW_AFK_BLOCK(0x43, mlxsw_sp_afk_element_info_ipv6_3), + MLXSW_AFK_BLOCK(0x44, mlxsw_sp_afk_element_info_ipv6_4), + MLXSW_AFK_BLOCK(0x45, mlxsw_sp_afk_element_info_ipv6_5), + MLXSW_AFK_BLOCK(0x90, mlxsw_sp_afk_element_info_l4_0), + MLXSW_AFK_BLOCK(0x92, mlxsw_sp_afk_element_info_l4_2), +}; + +#define MLXSW_SP2_AFK_BITS_PER_BLOCK 36 + +/* A block in Spectrum-2 is of the following form: + * + * +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ + * | | | | | | | | | | | | | | | | | | | | | | | | | | | | |35|34|33|32| + * +-----------------------------------------------------------------------------------------------+ + * |31|30|29|28|27|26|25|24|23|22|21|20|19|18|17|16|15|14|13|12|11|10| 9| 8| 7| 6| 5| 4| 3| 2| 1| 0| + * +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ + */ +MLXSW_ITEM64(sp2_afk, block, value, 0x00, 0, MLXSW_SP2_AFK_BITS_PER_BLOCK); + +/* The key / mask block layout in Spectrum-2 is of the following form: + * + * +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ + * | | | | | | | | | | | | | | | | | block11_high | + * +-----------------------------------------------------------------------------------------------+ + * | block11_low | block10_high | + * +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ + * ... + */ + +struct mlxsw_sp2_afk_block_layout { + unsigned short offset; + struct mlxsw_item item; +}; + +#define MLXSW_SP2_AFK_BLOCK_LAYOUT(_block, _offset, _shift) \ + { \ + .offset = _offset, \ + { \ + .shift = _shift, \ + .size = {.bits = MLXSW_SP2_AFK_BITS_PER_BLOCK}, \ + .name = #_block, \ + } \ + } \ + +static const struct mlxsw_sp2_afk_block_layout mlxsw_sp2_afk_blocks_layout[] = { + MLXSW_SP2_AFK_BLOCK_LAYOUT(block0, 0x30, 0), + MLXSW_SP2_AFK_BLOCK_LAYOUT(block1, 0x2C, 4), + MLXSW_SP2_AFK_BLOCK_LAYOUT(block2, 0x28, 8), + MLXSW_SP2_AFK_BLOCK_LAYOUT(block3, 0x24, 12), + MLXSW_SP2_AFK_BLOCK_LAYOUT(block4, 0x20, 16), + MLXSW_SP2_AFK_BLOCK_LAYOUT(block5, 0x1C, 20), + MLXSW_SP2_AFK_BLOCK_LAYOUT(block6, 0x18, 24), + MLXSW_SP2_AFK_BLOCK_LAYOUT(block7, 0x14, 28), + MLXSW_SP2_AFK_BLOCK_LAYOUT(block8, 0x0C, 0), + MLXSW_SP2_AFK_BLOCK_LAYOUT(block9, 0x08, 4), + MLXSW_SP2_AFK_BLOCK_LAYOUT(block10, 0x04, 8), + MLXSW_SP2_AFK_BLOCK_LAYOUT(block11, 0x00, 12), +}; + +static void __mlxsw_sp2_afk_block_value_set(char *output, int block_index, + u64 block_value) +{ + const struct mlxsw_sp2_afk_block_layout *block_layout; + + if (WARN_ON(block_index < 0 || + block_index >= ARRAY_SIZE(mlxsw_sp2_afk_blocks_layout))) + return; + + block_layout = &mlxsw_sp2_afk_blocks_layout[block_index]; + __mlxsw_item_set64(output + block_layout->offset, + &block_layout->item, 0, block_value); +} + +static void mlxsw_sp2_afk_encode_block(char *output, int block_index, + char *block) +{ + u64 block_value = mlxsw_sp2_afk_block_value_get(block); + + __mlxsw_sp2_afk_block_value_set(output, block_index, block_value); +} + +static void mlxsw_sp2_afk_clear_block(char *output, int block_index) +{ + __mlxsw_sp2_afk_block_value_set(output, block_index, 0); +} + +const struct mlxsw_afk_ops mlxsw_sp2_afk_ops = { + .blocks = mlxsw_sp2_afk_blocks, + .blocks_count = ARRAY_SIZE(mlxsw_sp2_afk_blocks), + .encode_block = mlxsw_sp2_afk_encode_block, + .clear_block = mlxsw_sp2_afk_clear_block, +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_mac_5b[] = { + MLXSW_AFK_ELEMENT_INST_U32(VID, 0x04, 18, 12), + MLXSW_AFK_ELEMENT_INST_EXT_U32(SRC_SYS_PORT, 0x04, 0, 9, -1, true), /* RX_ACL_SYSTEM_PORT */ +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv4_4b[] = { + MLXSW_AFK_ELEMENT_INST_U32(VIRT_ROUTER_LSB, 0x04, 13, 8), + MLXSW_AFK_ELEMENT_INST_U32(VIRT_ROUTER_MSB, 0x04, 21, 4), +}; + +static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv6_2b[] = { + MLXSW_AFK_ELEMENT_INST_BUF(DST_IP_96_127, 0x04, 4), +}; + +static const struct mlxsw_afk_block mlxsw_sp4_afk_blocks[] = { + MLXSW_AFK_BLOCK(0x10, mlxsw_sp_afk_element_info_mac_0), + MLXSW_AFK_BLOCK(0x11, mlxsw_sp_afk_element_info_mac_1), + MLXSW_AFK_BLOCK(0x12, mlxsw_sp_afk_element_info_mac_2), + MLXSW_AFK_BLOCK(0x13, mlxsw_sp_afk_element_info_mac_3), + MLXSW_AFK_BLOCK(0x14, mlxsw_sp_afk_element_info_mac_4), + MLXSW_AFK_BLOCK(0x1A, mlxsw_sp_afk_element_info_mac_5b), + MLXSW_AFK_BLOCK(0x38, mlxsw_sp_afk_element_info_ipv4_0), + MLXSW_AFK_BLOCK(0x39, mlxsw_sp_afk_element_info_ipv4_1), + MLXSW_AFK_BLOCK(0x3A, mlxsw_sp_afk_element_info_ipv4_2), + MLXSW_AFK_BLOCK(0x35, mlxsw_sp_afk_element_info_ipv4_4b), + MLXSW_AFK_BLOCK(0x40, mlxsw_sp_afk_element_info_ipv6_0), + MLXSW_AFK_BLOCK(0x41, mlxsw_sp_afk_element_info_ipv6_1), + MLXSW_AFK_BLOCK(0x47, mlxsw_sp_afk_element_info_ipv6_2b), + MLXSW_AFK_BLOCK(0x43, mlxsw_sp_afk_element_info_ipv6_3), + MLXSW_AFK_BLOCK(0x44, mlxsw_sp_afk_element_info_ipv6_4), + MLXSW_AFK_BLOCK(0x45, mlxsw_sp_afk_element_info_ipv6_5), + MLXSW_AFK_BLOCK(0x90, mlxsw_sp_afk_element_info_l4_0), + MLXSW_AFK_BLOCK(0x92, mlxsw_sp_afk_element_info_l4_2), +}; + +const struct mlxsw_afk_ops mlxsw_sp4_afk_ops = { + .blocks = mlxsw_sp4_afk_blocks, + .blocks_count = ARRAY_SIZE(mlxsw_sp4_afk_blocks), + .encode_block = mlxsw_sp2_afk_encode_block, + .clear_block = mlxsw_sp2_afk_clear_block, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c new file mode 100644 index 000000000..3b9ba8fa2 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c @@ -0,0 +1,1867 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/slab.h> +#include <linux/errno.h> +#include <linux/bitops.h> +#include <linux/list.h> +#include <linux/rhashtable.h> +#include <linux/netdevice.h> +#include <linux/mutex.h> +#include <trace/events/mlxsw.h> + +#include "reg.h" +#include "core.h" +#include "resources.h" +#include "spectrum.h" +#include "spectrum_acl_tcam.h" +#include "core_acl_flex_keys.h" + +size_t mlxsw_sp_acl_tcam_priv_size(struct mlxsw_sp *mlxsw_sp) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + + return ops->priv_size; +} + +#define MLXSW_SP_ACL_TCAM_VREGION_REHASH_INTRVL_DFLT 5000 /* ms */ +#define MLXSW_SP_ACL_TCAM_VREGION_REHASH_INTRVL_MIN 3000 /* ms */ +#define MLXSW_SP_ACL_TCAM_VREGION_REHASH_CREDITS 100 /* number of entries */ + +int mlxsw_sp_acl_tcam_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + u64 max_tcam_regions; + u64 max_regions; + u64 max_groups; + int err; + + mutex_init(&tcam->lock); + tcam->vregion_rehash_intrvl = + MLXSW_SP_ACL_TCAM_VREGION_REHASH_INTRVL_DFLT; + INIT_LIST_HEAD(&tcam->vregion_list); + + max_tcam_regions = MLXSW_CORE_RES_GET(mlxsw_sp->core, + ACL_MAX_TCAM_REGIONS); + max_regions = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_REGIONS); + + /* Use 1:1 mapping between ACL region and TCAM region */ + if (max_tcam_regions < max_regions) + max_regions = max_tcam_regions; + + tcam->used_regions = bitmap_zalloc(max_regions, GFP_KERNEL); + if (!tcam->used_regions) + return -ENOMEM; + tcam->max_regions = max_regions; + + max_groups = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_GROUPS); + tcam->used_groups = bitmap_zalloc(max_groups, GFP_KERNEL); + if (!tcam->used_groups) { + err = -ENOMEM; + goto err_alloc_used_groups; + } + tcam->max_groups = max_groups; + tcam->max_group_size = MLXSW_CORE_RES_GET(mlxsw_sp->core, + ACL_MAX_GROUP_SIZE); + + err = ops->init(mlxsw_sp, tcam->priv, tcam); + if (err) + goto err_tcam_init; + + return 0; + +err_tcam_init: + bitmap_free(tcam->used_groups); +err_alloc_used_groups: + bitmap_free(tcam->used_regions); + return err; +} + +void mlxsw_sp_acl_tcam_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + + mutex_destroy(&tcam->lock); + ops->fini(mlxsw_sp, tcam->priv); + bitmap_free(tcam->used_groups); + bitmap_free(tcam->used_regions); +} + +int mlxsw_sp_acl_tcam_priority_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + u32 *priority, bool fillup_priority) +{ + u64 max_priority; + + if (!fillup_priority) { + *priority = 0; + return 0; + } + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, KVD_SIZE)) + return -EIO; + + /* Priority range is 1..cap_kvd_size-1. */ + max_priority = MLXSW_CORE_RES_GET(mlxsw_sp->core, KVD_SIZE) - 1; + if (rulei->priority >= max_priority) + return -EINVAL; + + /* Unlike in TC, in HW, higher number means higher priority. */ + *priority = max_priority - rulei->priority; + return 0; +} + +static int mlxsw_sp_acl_tcam_region_id_get(struct mlxsw_sp_acl_tcam *tcam, + u16 *p_id) +{ + u16 id; + + id = find_first_zero_bit(tcam->used_regions, tcam->max_regions); + if (id < tcam->max_regions) { + __set_bit(id, tcam->used_regions); + *p_id = id; + return 0; + } + return -ENOBUFS; +} + +static void mlxsw_sp_acl_tcam_region_id_put(struct mlxsw_sp_acl_tcam *tcam, + u16 id) +{ + __clear_bit(id, tcam->used_regions); +} + +static int mlxsw_sp_acl_tcam_group_id_get(struct mlxsw_sp_acl_tcam *tcam, + u16 *p_id) +{ + u16 id; + + id = find_first_zero_bit(tcam->used_groups, tcam->max_groups); + if (id < tcam->max_groups) { + __set_bit(id, tcam->used_groups); + *p_id = id; + return 0; + } + return -ENOBUFS; +} + +static void mlxsw_sp_acl_tcam_group_id_put(struct mlxsw_sp_acl_tcam *tcam, + u16 id) +{ + __clear_bit(id, tcam->used_groups); +} + +struct mlxsw_sp_acl_tcam_pattern { + const enum mlxsw_afk_element *elements; + unsigned int elements_count; +}; + +struct mlxsw_sp_acl_tcam_group { + struct mlxsw_sp_acl_tcam *tcam; + u16 id; + struct mutex lock; /* guards region list updates */ + struct list_head region_list; + unsigned int region_count; +}; + +struct mlxsw_sp_acl_tcam_vgroup { + struct mlxsw_sp_acl_tcam_group group; + struct list_head vregion_list; + struct rhashtable vchunk_ht; + const struct mlxsw_sp_acl_tcam_pattern *patterns; + unsigned int patterns_count; + bool tmplt_elusage_set; + struct mlxsw_afk_element_usage tmplt_elusage; + bool vregion_rehash_enabled; + unsigned int *p_min_prio; + unsigned int *p_max_prio; +}; + +struct mlxsw_sp_acl_tcam_rehash_ctx { + void *hints_priv; + bool this_is_rollback; + struct mlxsw_sp_acl_tcam_vchunk *current_vchunk; /* vchunk being + * currently migrated. + */ + struct mlxsw_sp_acl_tcam_ventry *start_ventry; /* ventry to start + * migration from in + * a vchunk being + * currently migrated. + */ + struct mlxsw_sp_acl_tcam_ventry *stop_ventry; /* ventry to stop + * migration at + * a vchunk being + * currently migrated. + */ +}; + +struct mlxsw_sp_acl_tcam_vregion { + struct mutex lock; /* Protects consistency of region, region2 pointers + * and vchunk_list. + */ + struct mlxsw_sp_acl_tcam_region *region; + struct mlxsw_sp_acl_tcam_region *region2; /* Used during migration */ + struct list_head list; /* Member of a TCAM group */ + struct list_head tlist; /* Member of a TCAM */ + struct list_head vchunk_list; /* List of vchunks under this vregion */ + struct mlxsw_afk_key_info *key_info; + struct mlxsw_sp_acl_tcam *tcam; + struct mlxsw_sp_acl_tcam_vgroup *vgroup; + struct { + struct delayed_work dw; + struct mlxsw_sp_acl_tcam_rehash_ctx ctx; + } rehash; + struct mlxsw_sp *mlxsw_sp; + unsigned int ref_count; +}; + +struct mlxsw_sp_acl_tcam_vchunk; + +struct mlxsw_sp_acl_tcam_chunk { + struct mlxsw_sp_acl_tcam_vchunk *vchunk; + struct mlxsw_sp_acl_tcam_region *region; + unsigned long priv[]; + /* priv has to be always the last item */ +}; + +struct mlxsw_sp_acl_tcam_vchunk { + struct mlxsw_sp_acl_tcam_chunk *chunk; + struct mlxsw_sp_acl_tcam_chunk *chunk2; /* Used during migration */ + struct list_head list; /* Member of a TCAM vregion */ + struct rhash_head ht_node; /* Member of a chunk HT */ + struct list_head ventry_list; + unsigned int priority; /* Priority within the vregion and group */ + struct mlxsw_sp_acl_tcam_vgroup *vgroup; + struct mlxsw_sp_acl_tcam_vregion *vregion; + unsigned int ref_count; +}; + +struct mlxsw_sp_acl_tcam_entry { + struct mlxsw_sp_acl_tcam_ventry *ventry; + struct mlxsw_sp_acl_tcam_chunk *chunk; + unsigned long priv[]; + /* priv has to be always the last item */ +}; + +struct mlxsw_sp_acl_tcam_ventry { + struct mlxsw_sp_acl_tcam_entry *entry; + struct list_head list; /* Member of a TCAM vchunk */ + struct mlxsw_sp_acl_tcam_vchunk *vchunk; + struct mlxsw_sp_acl_rule_info *rulei; +}; + +static const struct rhashtable_params mlxsw_sp_acl_tcam_vchunk_ht_params = { + .key_len = sizeof(unsigned int), + .key_offset = offsetof(struct mlxsw_sp_acl_tcam_vchunk, priority), + .head_offset = offsetof(struct mlxsw_sp_acl_tcam_vchunk, ht_node), + .automatic_shrinking = true, +}; + +static int mlxsw_sp_acl_tcam_group_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_group *group) +{ + struct mlxsw_sp_acl_tcam_region *region; + char pagt_pl[MLXSW_REG_PAGT_LEN]; + int acl_index = 0; + + mlxsw_reg_pagt_pack(pagt_pl, group->id); + list_for_each_entry(region, &group->region_list, list) { + bool multi = false; + + /* Check if the next entry in the list has the same vregion. */ + if (region->list.next != &group->region_list && + list_next_entry(region, list)->vregion == region->vregion) + multi = true; + mlxsw_reg_pagt_acl_id_pack(pagt_pl, acl_index++, + region->id, multi); + } + mlxsw_reg_pagt_size_set(pagt_pl, acl_index); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pagt), pagt_pl); +} + +static int +mlxsw_sp_acl_tcam_group_add(struct mlxsw_sp_acl_tcam *tcam, + struct mlxsw_sp_acl_tcam_group *group) +{ + int err; + + group->tcam = tcam; + INIT_LIST_HEAD(&group->region_list); + + err = mlxsw_sp_acl_tcam_group_id_get(tcam, &group->id); + if (err) + return err; + + mutex_init(&group->lock); + + return 0; +} + +static void mlxsw_sp_acl_tcam_group_del(struct mlxsw_sp_acl_tcam_group *group) +{ + struct mlxsw_sp_acl_tcam *tcam = group->tcam; + + mutex_destroy(&group->lock); + mlxsw_sp_acl_tcam_group_id_put(tcam, group->id); + WARN_ON(!list_empty(&group->region_list)); +} + +static int +mlxsw_sp_acl_tcam_vgroup_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam, + struct mlxsw_sp_acl_tcam_vgroup *vgroup, + const struct mlxsw_sp_acl_tcam_pattern *patterns, + unsigned int patterns_count, + struct mlxsw_afk_element_usage *tmplt_elusage, + bool vregion_rehash_enabled, + unsigned int *p_min_prio, + unsigned int *p_max_prio) +{ + int err; + + vgroup->patterns = patterns; + vgroup->patterns_count = patterns_count; + vgroup->vregion_rehash_enabled = vregion_rehash_enabled; + vgroup->p_min_prio = p_min_prio; + vgroup->p_max_prio = p_max_prio; + + if (tmplt_elusage) { + vgroup->tmplt_elusage_set = true; + memcpy(&vgroup->tmplt_elusage, tmplt_elusage, + sizeof(vgroup->tmplt_elusage)); + } + INIT_LIST_HEAD(&vgroup->vregion_list); + + err = mlxsw_sp_acl_tcam_group_add(tcam, &vgroup->group); + if (err) + return err; + + err = rhashtable_init(&vgroup->vchunk_ht, + &mlxsw_sp_acl_tcam_vchunk_ht_params); + if (err) + goto err_rhashtable_init; + + return 0; + +err_rhashtable_init: + mlxsw_sp_acl_tcam_group_del(&vgroup->group); + return err; +} + +static void +mlxsw_sp_acl_tcam_vgroup_del(struct mlxsw_sp_acl_tcam_vgroup *vgroup) +{ + rhashtable_destroy(&vgroup->vchunk_ht); + mlxsw_sp_acl_tcam_group_del(&vgroup->group); + WARN_ON(!list_empty(&vgroup->vregion_list)); +} + +static int +mlxsw_sp_acl_tcam_group_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_group *group, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + char ppbt_pl[MLXSW_REG_PPBT_LEN]; + + mlxsw_reg_ppbt_pack(ppbt_pl, ingress ? MLXSW_REG_PXBT_E_IACL : + MLXSW_REG_PXBT_E_EACL, + MLXSW_REG_PXBT_OP_BIND, mlxsw_sp_port->local_port, + group->id); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ppbt), ppbt_pl); +} + +static void +mlxsw_sp_acl_tcam_group_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_group *group, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + char ppbt_pl[MLXSW_REG_PPBT_LEN]; + + mlxsw_reg_ppbt_pack(ppbt_pl, ingress ? MLXSW_REG_PXBT_E_IACL : + MLXSW_REG_PXBT_E_EACL, + MLXSW_REG_PXBT_OP_UNBIND, mlxsw_sp_port->local_port, + group->id); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ppbt), ppbt_pl); +} + +static u16 +mlxsw_sp_acl_tcam_group_id(struct mlxsw_sp_acl_tcam_group *group) +{ + return group->id; +} + +static unsigned int +mlxsw_sp_acl_tcam_vregion_prio(struct mlxsw_sp_acl_tcam_vregion *vregion) +{ + struct mlxsw_sp_acl_tcam_vchunk *vchunk; + + if (list_empty(&vregion->vchunk_list)) + return 0; + /* As a priority of a vregion, return priority of the first vchunk */ + vchunk = list_first_entry(&vregion->vchunk_list, + typeof(*vchunk), list); + return vchunk->priority; +} + +static unsigned int +mlxsw_sp_acl_tcam_vregion_max_prio(struct mlxsw_sp_acl_tcam_vregion *vregion) +{ + struct mlxsw_sp_acl_tcam_vchunk *vchunk; + + if (list_empty(&vregion->vchunk_list)) + return 0; + vchunk = list_last_entry(&vregion->vchunk_list, + typeof(*vchunk), list); + return vchunk->priority; +} + +static void +mlxsw_sp_acl_tcam_vgroup_prio_update(struct mlxsw_sp_acl_tcam_vgroup *vgroup) +{ + struct mlxsw_sp_acl_tcam_vregion *vregion; + + if (list_empty(&vgroup->vregion_list)) + return; + vregion = list_first_entry(&vgroup->vregion_list, + typeof(*vregion), list); + *vgroup->p_min_prio = mlxsw_sp_acl_tcam_vregion_prio(vregion); + vregion = list_last_entry(&vgroup->vregion_list, + typeof(*vregion), list); + *vgroup->p_max_prio = mlxsw_sp_acl_tcam_vregion_max_prio(vregion); +} + +static int +mlxsw_sp_acl_tcam_group_region_attach(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_group *group, + struct mlxsw_sp_acl_tcam_region *region, + unsigned int priority, + struct mlxsw_sp_acl_tcam_region *next_region) +{ + struct mlxsw_sp_acl_tcam_region *region2; + struct list_head *pos; + int err; + + mutex_lock(&group->lock); + if (group->region_count == group->tcam->max_group_size) { + err = -ENOBUFS; + goto err_region_count_check; + } + + if (next_region) { + /* If the next region is defined, place the new one + * before it. The next one is a sibling. + */ + pos = &next_region->list; + } else { + /* Position the region inside the list according to priority */ + list_for_each(pos, &group->region_list) { + region2 = list_entry(pos, typeof(*region2), list); + if (mlxsw_sp_acl_tcam_vregion_prio(region2->vregion) > + priority) + break; + } + } + list_add_tail(®ion->list, pos); + region->group = group; + + err = mlxsw_sp_acl_tcam_group_update(mlxsw_sp, group); + if (err) + goto err_group_update; + + group->region_count++; + mutex_unlock(&group->lock); + return 0; + +err_group_update: + list_del(®ion->list); +err_region_count_check: + mutex_unlock(&group->lock); + return err; +} + +static void +mlxsw_sp_acl_tcam_group_region_detach(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *region) +{ + struct mlxsw_sp_acl_tcam_group *group = region->group; + + mutex_lock(&group->lock); + list_del(®ion->list); + group->region_count--; + mlxsw_sp_acl_tcam_group_update(mlxsw_sp, group); + mutex_unlock(&group->lock); +} + +static int +mlxsw_sp_acl_tcam_vgroup_vregion_attach(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vgroup *vgroup, + struct mlxsw_sp_acl_tcam_vregion *vregion, + unsigned int priority) +{ + struct mlxsw_sp_acl_tcam_vregion *vregion2; + struct list_head *pos; + int err; + + /* Position the vregion inside the list according to priority */ + list_for_each(pos, &vgroup->vregion_list) { + vregion2 = list_entry(pos, typeof(*vregion2), list); + if (mlxsw_sp_acl_tcam_vregion_prio(vregion2) > priority) + break; + } + list_add_tail(&vregion->list, pos); + + err = mlxsw_sp_acl_tcam_group_region_attach(mlxsw_sp, &vgroup->group, + vregion->region, + priority, NULL); + if (err) + goto err_region_attach; + + return 0; + +err_region_attach: + list_del(&vregion->list); + return err; +} + +static void +mlxsw_sp_acl_tcam_vgroup_vregion_detach(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vregion *vregion) +{ + list_del(&vregion->list); + if (vregion->region2) + mlxsw_sp_acl_tcam_group_region_detach(mlxsw_sp, + vregion->region2); + mlxsw_sp_acl_tcam_group_region_detach(mlxsw_sp, vregion->region); +} + +static struct mlxsw_sp_acl_tcam_vregion * +mlxsw_sp_acl_tcam_vgroup_vregion_find(struct mlxsw_sp_acl_tcam_vgroup *vgroup, + unsigned int priority, + struct mlxsw_afk_element_usage *elusage, + bool *p_need_split) +{ + struct mlxsw_sp_acl_tcam_vregion *vregion, *vregion2; + struct list_head *pos; + bool issubset; + + list_for_each(pos, &vgroup->vregion_list) { + vregion = list_entry(pos, typeof(*vregion), list); + + /* First, check if the requested priority does not rather belong + * under some of the next vregions. + */ + if (pos->next != &vgroup->vregion_list) { /* not last */ + vregion2 = list_entry(pos->next, typeof(*vregion2), + list); + if (priority >= + mlxsw_sp_acl_tcam_vregion_prio(vregion2)) + continue; + } + + issubset = mlxsw_afk_key_info_subset(vregion->key_info, + elusage); + + /* If requested element usage would not fit and the priority + * is lower than the currently inspected vregion we cannot + * use this region, so return NULL to indicate new vregion has + * to be created. + */ + if (!issubset && + priority < mlxsw_sp_acl_tcam_vregion_prio(vregion)) + return NULL; + + /* If requested element usage would not fit and the priority + * is higher than the currently inspected vregion we cannot + * use this vregion. There is still some hope that the next + * vregion would be the fit. So let it be processed and + * eventually break at the check right above this. + */ + if (!issubset && + priority > mlxsw_sp_acl_tcam_vregion_max_prio(vregion)) + continue; + + /* Indicate if the vregion needs to be split in order to add + * the requested priority. Split is needed when requested + * element usage won't fit into the found vregion. + */ + *p_need_split = !issubset; + return vregion; + } + return NULL; /* New vregion has to be created. */ +} + +static void +mlxsw_sp_acl_tcam_vgroup_use_patterns(struct mlxsw_sp_acl_tcam_vgroup *vgroup, + struct mlxsw_afk_element_usage *elusage, + struct mlxsw_afk_element_usage *out) +{ + const struct mlxsw_sp_acl_tcam_pattern *pattern; + int i; + + /* In case the template is set, we don't have to look up the pattern + * and just use the template. + */ + if (vgroup->tmplt_elusage_set) { + memcpy(out, &vgroup->tmplt_elusage, sizeof(*out)); + WARN_ON(!mlxsw_afk_element_usage_subset(elusage, out)); + return; + } + + for (i = 0; i < vgroup->patterns_count; i++) { + pattern = &vgroup->patterns[i]; + mlxsw_afk_element_usage_fill(out, pattern->elements, + pattern->elements_count); + if (mlxsw_afk_element_usage_subset(elusage, out)) + return; + } + memcpy(out, elusage, sizeof(*out)); +} + +static int +mlxsw_sp_acl_tcam_region_alloc(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *region) +{ + struct mlxsw_afk_key_info *key_info = region->key_info; + char ptar_pl[MLXSW_REG_PTAR_LEN]; + unsigned int encodings_count; + int i; + int err; + + mlxsw_reg_ptar_pack(ptar_pl, MLXSW_REG_PTAR_OP_ALLOC, + region->key_type, + MLXSW_SP_ACL_TCAM_REGION_BASE_COUNT, + region->id, region->tcam_region_info); + encodings_count = mlxsw_afk_key_info_blocks_count_get(key_info); + for (i = 0; i < encodings_count; i++) { + u16 encoding; + + encoding = mlxsw_afk_key_info_block_encoding_get(key_info, i); + mlxsw_reg_ptar_key_id_pack(ptar_pl, i, encoding); + } + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptar), ptar_pl); + if (err) + return err; + mlxsw_reg_ptar_unpack(ptar_pl, region->tcam_region_info); + return 0; +} + +static void +mlxsw_sp_acl_tcam_region_free(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *region) +{ + char ptar_pl[MLXSW_REG_PTAR_LEN]; + + mlxsw_reg_ptar_pack(ptar_pl, MLXSW_REG_PTAR_OP_FREE, + region->key_type, 0, region->id, + region->tcam_region_info); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptar), ptar_pl); +} + +static int +mlxsw_sp_acl_tcam_region_enable(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *region) +{ + char pacl_pl[MLXSW_REG_PACL_LEN]; + + mlxsw_reg_pacl_pack(pacl_pl, region->id, true, + region->tcam_region_info); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pacl), pacl_pl); +} + +static void +mlxsw_sp_acl_tcam_region_disable(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *region) +{ + char pacl_pl[MLXSW_REG_PACL_LEN]; + + mlxsw_reg_pacl_pack(pacl_pl, region->id, false, + region->tcam_region_info); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pacl), pacl_pl); +} + +static struct mlxsw_sp_acl_tcam_region * +mlxsw_sp_acl_tcam_region_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam, + struct mlxsw_sp_acl_tcam_vregion *vregion, + void *hints_priv) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + struct mlxsw_sp_acl_tcam_region *region; + int err; + + region = kzalloc(sizeof(*region) + ops->region_priv_size, GFP_KERNEL); + if (!region) + return ERR_PTR(-ENOMEM); + region->mlxsw_sp = mlxsw_sp; + region->vregion = vregion; + region->key_info = vregion->key_info; + + err = mlxsw_sp_acl_tcam_region_id_get(tcam, ®ion->id); + if (err) + goto err_region_id_get; + + err = ops->region_associate(mlxsw_sp, region); + if (err) + goto err_tcam_region_associate; + + region->key_type = ops->key_type; + err = mlxsw_sp_acl_tcam_region_alloc(mlxsw_sp, region); + if (err) + goto err_tcam_region_alloc; + + err = mlxsw_sp_acl_tcam_region_enable(mlxsw_sp, region); + if (err) + goto err_tcam_region_enable; + + err = ops->region_init(mlxsw_sp, region->priv, tcam->priv, + region, hints_priv); + if (err) + goto err_tcam_region_init; + + return region; + +err_tcam_region_init: + mlxsw_sp_acl_tcam_region_disable(mlxsw_sp, region); +err_tcam_region_enable: + mlxsw_sp_acl_tcam_region_free(mlxsw_sp, region); +err_tcam_region_alloc: +err_tcam_region_associate: + mlxsw_sp_acl_tcam_region_id_put(tcam, region->id); +err_region_id_get: + kfree(region); + return ERR_PTR(err); +} + +static void +mlxsw_sp_acl_tcam_region_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *region) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + + ops->region_fini(mlxsw_sp, region->priv); + mlxsw_sp_acl_tcam_region_disable(mlxsw_sp, region); + mlxsw_sp_acl_tcam_region_free(mlxsw_sp, region); + mlxsw_sp_acl_tcam_region_id_put(region->group->tcam, + region->id); + kfree(region); +} + +static void +mlxsw_sp_acl_tcam_vregion_rehash_work_schedule(struct mlxsw_sp_acl_tcam_vregion *vregion) +{ + unsigned long interval = vregion->tcam->vregion_rehash_intrvl; + + if (!interval) + return; + mlxsw_core_schedule_dw(&vregion->rehash.dw, + msecs_to_jiffies(interval)); +} + +static void +mlxsw_sp_acl_tcam_vregion_rehash(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vregion *vregion, + int *credits); + +static void mlxsw_sp_acl_tcam_vregion_rehash_work(struct work_struct *work) +{ + struct mlxsw_sp_acl_tcam_vregion *vregion = + container_of(work, struct mlxsw_sp_acl_tcam_vregion, + rehash.dw.work); + int credits = MLXSW_SP_ACL_TCAM_VREGION_REHASH_CREDITS; + + mlxsw_sp_acl_tcam_vregion_rehash(vregion->mlxsw_sp, vregion, &credits); + if (credits < 0) + /* Rehash gone out of credits so it was interrupted. + * Schedule the work as soon as possible to continue. + */ + mlxsw_core_schedule_dw(&vregion->rehash.dw, 0); + else + mlxsw_sp_acl_tcam_vregion_rehash_work_schedule(vregion); +} + +static void +mlxsw_sp_acl_tcam_rehash_ctx_vchunk_changed(struct mlxsw_sp_acl_tcam_vchunk *vchunk) +{ + struct mlxsw_sp_acl_tcam_vregion *vregion = vchunk->vregion; + + /* If a rule was added or deleted from vchunk which is currently + * under rehash migration, we have to reset the ventry pointers + * to make sure all rules are properly migrated. + */ + if (vregion->rehash.ctx.current_vchunk == vchunk) { + vregion->rehash.ctx.start_ventry = NULL; + vregion->rehash.ctx.stop_ventry = NULL; + } +} + +static void +mlxsw_sp_acl_tcam_rehash_ctx_vregion_changed(struct mlxsw_sp_acl_tcam_vregion *vregion) +{ + /* If a chunk was added or deleted from vregion we have to reset + * the current chunk pointer to make sure all chunks + * are properly migrated. + */ + vregion->rehash.ctx.current_vchunk = NULL; +} + +static struct mlxsw_sp_acl_tcam_vregion * +mlxsw_sp_acl_tcam_vregion_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vgroup *vgroup, + unsigned int priority, + struct mlxsw_afk_element_usage *elusage) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + struct mlxsw_afk *afk = mlxsw_sp_acl_afk(mlxsw_sp->acl); + struct mlxsw_sp_acl_tcam *tcam = vgroup->group.tcam; + struct mlxsw_sp_acl_tcam_vregion *vregion; + int err; + + vregion = kzalloc(sizeof(*vregion), GFP_KERNEL); + if (!vregion) + return ERR_PTR(-ENOMEM); + INIT_LIST_HEAD(&vregion->vchunk_list); + mutex_init(&vregion->lock); + vregion->tcam = tcam; + vregion->mlxsw_sp = mlxsw_sp; + vregion->vgroup = vgroup; + vregion->ref_count = 1; + + vregion->key_info = mlxsw_afk_key_info_get(afk, elusage); + if (IS_ERR(vregion->key_info)) { + err = PTR_ERR(vregion->key_info); + goto err_key_info_get; + } + + vregion->region = mlxsw_sp_acl_tcam_region_create(mlxsw_sp, tcam, + vregion, NULL); + if (IS_ERR(vregion->region)) { + err = PTR_ERR(vregion->region); + goto err_region_create; + } + + err = mlxsw_sp_acl_tcam_vgroup_vregion_attach(mlxsw_sp, vgroup, vregion, + priority); + if (err) + goto err_vgroup_vregion_attach; + + if (vgroup->vregion_rehash_enabled && ops->region_rehash_hints_get) { + /* Create the delayed work for vregion periodic rehash */ + INIT_DELAYED_WORK(&vregion->rehash.dw, + mlxsw_sp_acl_tcam_vregion_rehash_work); + mlxsw_sp_acl_tcam_vregion_rehash_work_schedule(vregion); + mutex_lock(&tcam->lock); + list_add_tail(&vregion->tlist, &tcam->vregion_list); + mutex_unlock(&tcam->lock); + } + + return vregion; + +err_vgroup_vregion_attach: + mlxsw_sp_acl_tcam_region_destroy(mlxsw_sp, vregion->region); +err_region_create: + mlxsw_afk_key_info_put(vregion->key_info); +err_key_info_get: + kfree(vregion); + return ERR_PTR(err); +} + +static void +mlxsw_sp_acl_tcam_vregion_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vregion *vregion) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + struct mlxsw_sp_acl_tcam_vgroup *vgroup = vregion->vgroup; + struct mlxsw_sp_acl_tcam *tcam = vregion->tcam; + + if (vgroup->vregion_rehash_enabled && ops->region_rehash_hints_get) { + mutex_lock(&tcam->lock); + list_del(&vregion->tlist); + mutex_unlock(&tcam->lock); + cancel_delayed_work_sync(&vregion->rehash.dw); + } + mlxsw_sp_acl_tcam_vgroup_vregion_detach(mlxsw_sp, vregion); + if (vregion->region2) + mlxsw_sp_acl_tcam_region_destroy(mlxsw_sp, vregion->region2); + mlxsw_sp_acl_tcam_region_destroy(mlxsw_sp, vregion->region); + mlxsw_afk_key_info_put(vregion->key_info); + mutex_destroy(&vregion->lock); + kfree(vregion); +} + +u32 mlxsw_sp_acl_tcam_vregion_rehash_intrvl_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + u32 vregion_rehash_intrvl; + + if (WARN_ON(!ops->region_rehash_hints_get)) + return 0; + vregion_rehash_intrvl = tcam->vregion_rehash_intrvl; + return vregion_rehash_intrvl; +} + +int mlxsw_sp_acl_tcam_vregion_rehash_intrvl_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam, + u32 val) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + struct mlxsw_sp_acl_tcam_vregion *vregion; + + if (val < MLXSW_SP_ACL_TCAM_VREGION_REHASH_INTRVL_MIN && val) + return -EINVAL; + if (WARN_ON(!ops->region_rehash_hints_get)) + return -EOPNOTSUPP; + tcam->vregion_rehash_intrvl = val; + mutex_lock(&tcam->lock); + list_for_each_entry(vregion, &tcam->vregion_list, tlist) { + if (val) + mlxsw_core_schedule_dw(&vregion->rehash.dw, 0); + else + cancel_delayed_work_sync(&vregion->rehash.dw); + } + mutex_unlock(&tcam->lock); + return 0; +} + +static struct mlxsw_sp_acl_tcam_vregion * +mlxsw_sp_acl_tcam_vregion_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vgroup *vgroup, + unsigned int priority, + struct mlxsw_afk_element_usage *elusage) +{ + struct mlxsw_afk_element_usage vregion_elusage; + struct mlxsw_sp_acl_tcam_vregion *vregion; + bool need_split; + + vregion = mlxsw_sp_acl_tcam_vgroup_vregion_find(vgroup, priority, + elusage, &need_split); + if (vregion) { + if (need_split) { + /* According to priority, new vchunk should belong to + * an existing vregion. However, this vchunk needs + * elements that vregion does not contain. We need + * to split the existing vregion into two and create + * a new vregion for the new vchunk in between. + * This is not supported now. + */ + return ERR_PTR(-EOPNOTSUPP); + } + vregion->ref_count++; + return vregion; + } + + mlxsw_sp_acl_tcam_vgroup_use_patterns(vgroup, elusage, + &vregion_elusage); + + return mlxsw_sp_acl_tcam_vregion_create(mlxsw_sp, vgroup, priority, + &vregion_elusage); +} + +static void +mlxsw_sp_acl_tcam_vregion_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vregion *vregion) +{ + if (--vregion->ref_count) + return; + mlxsw_sp_acl_tcam_vregion_destroy(mlxsw_sp, vregion); +} + +static struct mlxsw_sp_acl_tcam_chunk * +mlxsw_sp_acl_tcam_chunk_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vchunk *vchunk, + struct mlxsw_sp_acl_tcam_region *region) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + struct mlxsw_sp_acl_tcam_chunk *chunk; + + chunk = kzalloc(sizeof(*chunk) + ops->chunk_priv_size, GFP_KERNEL); + if (!chunk) + return ERR_PTR(-ENOMEM); + chunk->vchunk = vchunk; + chunk->region = region; + + ops->chunk_init(region->priv, chunk->priv, vchunk->priority); + return chunk; +} + +static void +mlxsw_sp_acl_tcam_chunk_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_chunk *chunk) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + + ops->chunk_fini(chunk->priv); + kfree(chunk); +} + +static struct mlxsw_sp_acl_tcam_vchunk * +mlxsw_sp_acl_tcam_vchunk_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vgroup *vgroup, + unsigned int priority, + struct mlxsw_afk_element_usage *elusage) +{ + struct mlxsw_sp_acl_tcam_vchunk *vchunk, *vchunk2; + struct mlxsw_sp_acl_tcam_vregion *vregion; + struct list_head *pos; + int err; + + if (priority == MLXSW_SP_ACL_TCAM_CATCHALL_PRIO) + return ERR_PTR(-EINVAL); + + vchunk = kzalloc(sizeof(*vchunk), GFP_KERNEL); + if (!vchunk) + return ERR_PTR(-ENOMEM); + INIT_LIST_HEAD(&vchunk->ventry_list); + vchunk->priority = priority; + vchunk->vgroup = vgroup; + vchunk->ref_count = 1; + + vregion = mlxsw_sp_acl_tcam_vregion_get(mlxsw_sp, vgroup, + priority, elusage); + if (IS_ERR(vregion)) { + err = PTR_ERR(vregion); + goto err_vregion_get; + } + + vchunk->vregion = vregion; + + err = rhashtable_insert_fast(&vgroup->vchunk_ht, &vchunk->ht_node, + mlxsw_sp_acl_tcam_vchunk_ht_params); + if (err) + goto err_rhashtable_insert; + + mutex_lock(&vregion->lock); + vchunk->chunk = mlxsw_sp_acl_tcam_chunk_create(mlxsw_sp, vchunk, + vchunk->vregion->region); + if (IS_ERR(vchunk->chunk)) { + mutex_unlock(&vregion->lock); + err = PTR_ERR(vchunk->chunk); + goto err_chunk_create; + } + + mlxsw_sp_acl_tcam_rehash_ctx_vregion_changed(vregion); + + /* Position the vchunk inside the list according to priority */ + list_for_each(pos, &vregion->vchunk_list) { + vchunk2 = list_entry(pos, typeof(*vchunk2), list); + if (vchunk2->priority > priority) + break; + } + list_add_tail(&vchunk->list, pos); + mutex_unlock(&vregion->lock); + mlxsw_sp_acl_tcam_vgroup_prio_update(vgroup); + + return vchunk; + +err_chunk_create: + rhashtable_remove_fast(&vgroup->vchunk_ht, &vchunk->ht_node, + mlxsw_sp_acl_tcam_vchunk_ht_params); +err_rhashtable_insert: + mlxsw_sp_acl_tcam_vregion_put(mlxsw_sp, vregion); +err_vregion_get: + kfree(vchunk); + return ERR_PTR(err); +} + +static void +mlxsw_sp_acl_tcam_vchunk_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vchunk *vchunk) +{ + struct mlxsw_sp_acl_tcam_vregion *vregion = vchunk->vregion; + struct mlxsw_sp_acl_tcam_vgroup *vgroup = vchunk->vgroup; + + mutex_lock(&vregion->lock); + mlxsw_sp_acl_tcam_rehash_ctx_vregion_changed(vregion); + list_del(&vchunk->list); + if (vchunk->chunk2) + mlxsw_sp_acl_tcam_chunk_destroy(mlxsw_sp, vchunk->chunk2); + mlxsw_sp_acl_tcam_chunk_destroy(mlxsw_sp, vchunk->chunk); + mutex_unlock(&vregion->lock); + rhashtable_remove_fast(&vgroup->vchunk_ht, &vchunk->ht_node, + mlxsw_sp_acl_tcam_vchunk_ht_params); + mlxsw_sp_acl_tcam_vregion_put(mlxsw_sp, vchunk->vregion); + kfree(vchunk); + mlxsw_sp_acl_tcam_vgroup_prio_update(vgroup); +} + +static struct mlxsw_sp_acl_tcam_vchunk * +mlxsw_sp_acl_tcam_vchunk_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vgroup *vgroup, + unsigned int priority, + struct mlxsw_afk_element_usage *elusage) +{ + struct mlxsw_sp_acl_tcam_vchunk *vchunk; + + vchunk = rhashtable_lookup_fast(&vgroup->vchunk_ht, &priority, + mlxsw_sp_acl_tcam_vchunk_ht_params); + if (vchunk) { + if (WARN_ON(!mlxsw_afk_key_info_subset(vchunk->vregion->key_info, + elusage))) + return ERR_PTR(-EINVAL); + vchunk->ref_count++; + return vchunk; + } + return mlxsw_sp_acl_tcam_vchunk_create(mlxsw_sp, vgroup, + priority, elusage); +} + +static void +mlxsw_sp_acl_tcam_vchunk_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vchunk *vchunk) +{ + if (--vchunk->ref_count) + return; + mlxsw_sp_acl_tcam_vchunk_destroy(mlxsw_sp, vchunk); +} + +static struct mlxsw_sp_acl_tcam_entry * +mlxsw_sp_acl_tcam_entry_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_ventry *ventry, + struct mlxsw_sp_acl_tcam_chunk *chunk) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + struct mlxsw_sp_acl_tcam_entry *entry; + int err; + + entry = kzalloc(sizeof(*entry) + ops->entry_priv_size, GFP_KERNEL); + if (!entry) + return ERR_PTR(-ENOMEM); + entry->ventry = ventry; + entry->chunk = chunk; + + err = ops->entry_add(mlxsw_sp, chunk->region->priv, chunk->priv, + entry->priv, ventry->rulei); + if (err) + goto err_entry_add; + + return entry; + +err_entry_add: + kfree(entry); + return ERR_PTR(err); +} + +static void mlxsw_sp_acl_tcam_entry_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_entry *entry) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + + ops->entry_del(mlxsw_sp, entry->chunk->region->priv, + entry->chunk->priv, entry->priv); + kfree(entry); +} + +static int +mlxsw_sp_acl_tcam_entry_action_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_region *region, + struct mlxsw_sp_acl_tcam_entry *entry, + struct mlxsw_sp_acl_rule_info *rulei) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + + return ops->entry_action_replace(mlxsw_sp, region->priv, + entry->priv, rulei); +} + +static int +mlxsw_sp_acl_tcam_entry_activity_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_entry *entry, + bool *activity) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + + return ops->entry_activity_get(mlxsw_sp, entry->chunk->region->priv, + entry->priv, activity); +} + +static int mlxsw_sp_acl_tcam_ventry_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vgroup *vgroup, + struct mlxsw_sp_acl_tcam_ventry *ventry, + struct mlxsw_sp_acl_rule_info *rulei) +{ + struct mlxsw_sp_acl_tcam_vregion *vregion; + struct mlxsw_sp_acl_tcam_vchunk *vchunk; + int err; + + vchunk = mlxsw_sp_acl_tcam_vchunk_get(mlxsw_sp, vgroup, rulei->priority, + &rulei->values.elusage); + if (IS_ERR(vchunk)) + return PTR_ERR(vchunk); + + ventry->vchunk = vchunk; + ventry->rulei = rulei; + vregion = vchunk->vregion; + + mutex_lock(&vregion->lock); + ventry->entry = mlxsw_sp_acl_tcam_entry_create(mlxsw_sp, ventry, + vchunk->chunk); + if (IS_ERR(ventry->entry)) { + mutex_unlock(&vregion->lock); + err = PTR_ERR(ventry->entry); + goto err_entry_create; + } + + list_add_tail(&ventry->list, &vchunk->ventry_list); + mlxsw_sp_acl_tcam_rehash_ctx_vchunk_changed(vchunk); + mutex_unlock(&vregion->lock); + + return 0; + +err_entry_create: + mlxsw_sp_acl_tcam_vchunk_put(mlxsw_sp, vchunk); + return err; +} + +static void mlxsw_sp_acl_tcam_ventry_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_ventry *ventry) +{ + struct mlxsw_sp_acl_tcam_vchunk *vchunk = ventry->vchunk; + struct mlxsw_sp_acl_tcam_vregion *vregion = vchunk->vregion; + + mutex_lock(&vregion->lock); + mlxsw_sp_acl_tcam_rehash_ctx_vchunk_changed(vchunk); + list_del(&ventry->list); + mlxsw_sp_acl_tcam_entry_destroy(mlxsw_sp, ventry->entry); + mutex_unlock(&vregion->lock); + mlxsw_sp_acl_tcam_vchunk_put(mlxsw_sp, vchunk); +} + +static int +mlxsw_sp_acl_tcam_ventry_action_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_ventry *ventry, + struct mlxsw_sp_acl_rule_info *rulei) +{ + struct mlxsw_sp_acl_tcam_vchunk *vchunk = ventry->vchunk; + + return mlxsw_sp_acl_tcam_entry_action_replace(mlxsw_sp, + vchunk->vregion->region, + ventry->entry, rulei); +} + +static int +mlxsw_sp_acl_tcam_ventry_activity_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_ventry *ventry, + bool *activity) +{ + return mlxsw_sp_acl_tcam_entry_activity_get(mlxsw_sp, + ventry->entry, activity); +} + +static int +mlxsw_sp_acl_tcam_ventry_migrate(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_ventry *ventry, + struct mlxsw_sp_acl_tcam_chunk *chunk, + int *credits) +{ + struct mlxsw_sp_acl_tcam_entry *new_entry; + + /* First check if the entry is not already where we want it to be. */ + if (ventry->entry->chunk == chunk) + return 0; + + if (--(*credits) < 0) + return 0; + + new_entry = mlxsw_sp_acl_tcam_entry_create(mlxsw_sp, ventry, chunk); + if (IS_ERR(new_entry)) + return PTR_ERR(new_entry); + mlxsw_sp_acl_tcam_entry_destroy(mlxsw_sp, ventry->entry); + ventry->entry = new_entry; + return 0; +} + +static int +mlxsw_sp_acl_tcam_vchunk_migrate_start(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vchunk *vchunk, + struct mlxsw_sp_acl_tcam_region *region, + struct mlxsw_sp_acl_tcam_rehash_ctx *ctx) +{ + struct mlxsw_sp_acl_tcam_chunk *new_chunk; + + new_chunk = mlxsw_sp_acl_tcam_chunk_create(mlxsw_sp, vchunk, region); + if (IS_ERR(new_chunk)) + return PTR_ERR(new_chunk); + vchunk->chunk2 = vchunk->chunk; + vchunk->chunk = new_chunk; + ctx->current_vchunk = vchunk; + ctx->start_ventry = NULL; + ctx->stop_ventry = NULL; + return 0; +} + +static void +mlxsw_sp_acl_tcam_vchunk_migrate_end(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vchunk *vchunk, + struct mlxsw_sp_acl_tcam_rehash_ctx *ctx) +{ + mlxsw_sp_acl_tcam_chunk_destroy(mlxsw_sp, vchunk->chunk2); + vchunk->chunk2 = NULL; + ctx->current_vchunk = NULL; +} + +static int +mlxsw_sp_acl_tcam_vchunk_migrate_one(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vchunk *vchunk, + struct mlxsw_sp_acl_tcam_region *region, + struct mlxsw_sp_acl_tcam_rehash_ctx *ctx, + int *credits) +{ + struct mlxsw_sp_acl_tcam_ventry *ventry; + int err; + + if (vchunk->chunk->region != region) { + err = mlxsw_sp_acl_tcam_vchunk_migrate_start(mlxsw_sp, vchunk, + region, ctx); + if (err) + return err; + } else if (!vchunk->chunk2) { + /* The chunk is already as it should be, nothing to do. */ + return 0; + } + + /* If the migration got interrupted, we have the ventry to start from + * stored in context. + */ + if (ctx->start_ventry) + ventry = ctx->start_ventry; + else + ventry = list_first_entry(&vchunk->ventry_list, + typeof(*ventry), list); + + list_for_each_entry_from(ventry, &vchunk->ventry_list, list) { + /* During rollback, once we reach the ventry that failed + * to migrate, we are done. + */ + if (ventry == ctx->stop_ventry) + break; + + err = mlxsw_sp_acl_tcam_ventry_migrate(mlxsw_sp, ventry, + vchunk->chunk, credits); + if (err) { + if (ctx->this_is_rollback) { + /* Save the ventry which we ended with and try + * to continue later on. + */ + ctx->start_ventry = ventry; + return err; + } + /* Swap the chunk and chunk2 pointers so the follow-up + * rollback call will see the original chunk pointer + * in vchunk->chunk. + */ + swap(vchunk->chunk, vchunk->chunk2); + /* The rollback has to be done from beginning of the + * chunk, that is why we have to null the start_ventry. + * However, we know where to stop the rollback, + * at the current ventry. + */ + ctx->start_ventry = NULL; + ctx->stop_ventry = ventry; + return err; + } else if (*credits < 0) { + /* We are out of credits, the rest of the ventries + * will be migrated later. Save the ventry + * which we ended with. + */ + ctx->start_ventry = ventry; + return 0; + } + } + + mlxsw_sp_acl_tcam_vchunk_migrate_end(mlxsw_sp, vchunk, ctx); + return 0; +} + +static int +mlxsw_sp_acl_tcam_vchunk_migrate_all(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vregion *vregion, + struct mlxsw_sp_acl_tcam_rehash_ctx *ctx, + int *credits) +{ + struct mlxsw_sp_acl_tcam_vchunk *vchunk; + int err; + + /* If the migration got interrupted, we have the vchunk + * we are working on stored in context. + */ + if (ctx->current_vchunk) + vchunk = ctx->current_vchunk; + else + vchunk = list_first_entry(&vregion->vchunk_list, + typeof(*vchunk), list); + + list_for_each_entry_from(vchunk, &vregion->vchunk_list, list) { + err = mlxsw_sp_acl_tcam_vchunk_migrate_one(mlxsw_sp, vchunk, + vregion->region, + ctx, credits); + if (err || *credits < 0) + return err; + } + return 0; +} + +static int +mlxsw_sp_acl_tcam_vregion_migrate(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vregion *vregion, + struct mlxsw_sp_acl_tcam_rehash_ctx *ctx, + int *credits) +{ + int err, err2; + + trace_mlxsw_sp_acl_tcam_vregion_migrate(mlxsw_sp, vregion); + mutex_lock(&vregion->lock); + err = mlxsw_sp_acl_tcam_vchunk_migrate_all(mlxsw_sp, vregion, + ctx, credits); + if (err) { + /* In case migration was not successful, we need to swap + * so the original region pointer is assigned again + * to vregion->region. + */ + swap(vregion->region, vregion->region2); + ctx->current_vchunk = NULL; + ctx->this_is_rollback = true; + err2 = mlxsw_sp_acl_tcam_vchunk_migrate_all(mlxsw_sp, vregion, + ctx, credits); + if (err2) { + trace_mlxsw_sp_acl_tcam_vregion_rehash_rollback_failed(mlxsw_sp, + vregion); + dev_err(mlxsw_sp->bus_info->dev, "Failed to rollback during vregion migration fail\n"); + /* Let the rollback to be continued later on. */ + } + } + mutex_unlock(&vregion->lock); + trace_mlxsw_sp_acl_tcam_vregion_migrate_end(mlxsw_sp, vregion); + return err; +} + +static bool +mlxsw_sp_acl_tcam_vregion_rehash_in_progress(const struct mlxsw_sp_acl_tcam_rehash_ctx *ctx) +{ + return ctx->hints_priv; +} + +static int +mlxsw_sp_acl_tcam_vregion_rehash_start(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vregion *vregion, + struct mlxsw_sp_acl_tcam_rehash_ctx *ctx) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + unsigned int priority = mlxsw_sp_acl_tcam_vregion_prio(vregion); + struct mlxsw_sp_acl_tcam_region *new_region; + void *hints_priv; + int err; + + trace_mlxsw_sp_acl_tcam_vregion_rehash(mlxsw_sp, vregion); + + hints_priv = ops->region_rehash_hints_get(vregion->region->priv); + if (IS_ERR(hints_priv)) + return PTR_ERR(hints_priv); + + new_region = mlxsw_sp_acl_tcam_region_create(mlxsw_sp, vregion->tcam, + vregion, hints_priv); + if (IS_ERR(new_region)) { + err = PTR_ERR(new_region); + goto err_region_create; + } + + /* vregion->region contains the pointer to the new region + * we are going to migrate to. + */ + vregion->region2 = vregion->region; + vregion->region = new_region; + err = mlxsw_sp_acl_tcam_group_region_attach(mlxsw_sp, + vregion->region2->group, + new_region, priority, + vregion->region2); + if (err) + goto err_group_region_attach; + + ctx->hints_priv = hints_priv; + ctx->this_is_rollback = false; + + return 0; + +err_group_region_attach: + vregion->region = vregion->region2; + vregion->region2 = NULL; + mlxsw_sp_acl_tcam_region_destroy(mlxsw_sp, new_region); +err_region_create: + ops->region_rehash_hints_put(hints_priv); + return err; +} + +static void +mlxsw_sp_acl_tcam_vregion_rehash_end(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vregion *vregion, + struct mlxsw_sp_acl_tcam_rehash_ctx *ctx) +{ + struct mlxsw_sp_acl_tcam_region *unused_region = vregion->region2; + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + + vregion->region2 = NULL; + mlxsw_sp_acl_tcam_group_region_detach(mlxsw_sp, unused_region); + mlxsw_sp_acl_tcam_region_destroy(mlxsw_sp, unused_region); + ops->region_rehash_hints_put(ctx->hints_priv); + ctx->hints_priv = NULL; +} + +static void +mlxsw_sp_acl_tcam_vregion_rehash(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam_vregion *vregion, + int *credits) +{ + struct mlxsw_sp_acl_tcam_rehash_ctx *ctx = &vregion->rehash.ctx; + int err; + + /* Check if the previous rehash work was interrupted + * which means we have to continue it now. + * If not, start a new rehash. + */ + if (!mlxsw_sp_acl_tcam_vregion_rehash_in_progress(ctx)) { + err = mlxsw_sp_acl_tcam_vregion_rehash_start(mlxsw_sp, + vregion, ctx); + if (err) { + if (err != -EAGAIN) + dev_err(mlxsw_sp->bus_info->dev, "Failed get rehash hints\n"); + return; + } + } + + err = mlxsw_sp_acl_tcam_vregion_migrate(mlxsw_sp, vregion, + ctx, credits); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to migrate vregion\n"); + } + + if (*credits >= 0) + mlxsw_sp_acl_tcam_vregion_rehash_end(mlxsw_sp, vregion, ctx); +} + +static const enum mlxsw_afk_element mlxsw_sp_acl_tcam_pattern_ipv4[] = { + MLXSW_AFK_ELEMENT_SRC_SYS_PORT, + MLXSW_AFK_ELEMENT_DMAC_32_47, + MLXSW_AFK_ELEMENT_DMAC_0_31, + MLXSW_AFK_ELEMENT_SMAC_32_47, + MLXSW_AFK_ELEMENT_SMAC_0_31, + MLXSW_AFK_ELEMENT_ETHERTYPE, + MLXSW_AFK_ELEMENT_IP_PROTO, + MLXSW_AFK_ELEMENT_SRC_IP_0_31, + MLXSW_AFK_ELEMENT_DST_IP_0_31, + MLXSW_AFK_ELEMENT_DST_L4_PORT, + MLXSW_AFK_ELEMENT_SRC_L4_PORT, + MLXSW_AFK_ELEMENT_VID, + MLXSW_AFK_ELEMENT_PCP, + MLXSW_AFK_ELEMENT_TCP_FLAGS, + MLXSW_AFK_ELEMENT_IP_TTL_, + MLXSW_AFK_ELEMENT_IP_ECN, + MLXSW_AFK_ELEMENT_IP_DSCP, +}; + +static const enum mlxsw_afk_element mlxsw_sp_acl_tcam_pattern_ipv6[] = { + MLXSW_AFK_ELEMENT_ETHERTYPE, + MLXSW_AFK_ELEMENT_IP_PROTO, + MLXSW_AFK_ELEMENT_SRC_IP_96_127, + MLXSW_AFK_ELEMENT_SRC_IP_64_95, + MLXSW_AFK_ELEMENT_SRC_IP_32_63, + MLXSW_AFK_ELEMENT_SRC_IP_0_31, + MLXSW_AFK_ELEMENT_DST_IP_96_127, + MLXSW_AFK_ELEMENT_DST_IP_64_95, + MLXSW_AFK_ELEMENT_DST_IP_32_63, + MLXSW_AFK_ELEMENT_DST_IP_0_31, + MLXSW_AFK_ELEMENT_DST_L4_PORT, + MLXSW_AFK_ELEMENT_SRC_L4_PORT, +}; + +static const struct mlxsw_sp_acl_tcam_pattern mlxsw_sp_acl_tcam_patterns[] = { + { + .elements = mlxsw_sp_acl_tcam_pattern_ipv4, + .elements_count = ARRAY_SIZE(mlxsw_sp_acl_tcam_pattern_ipv4), + }, + { + .elements = mlxsw_sp_acl_tcam_pattern_ipv6, + .elements_count = ARRAY_SIZE(mlxsw_sp_acl_tcam_pattern_ipv6), + }, +}; + +#define MLXSW_SP_ACL_TCAM_PATTERNS_COUNT \ + ARRAY_SIZE(mlxsw_sp_acl_tcam_patterns) + +struct mlxsw_sp_acl_tcam_flower_ruleset { + struct mlxsw_sp_acl_tcam_vgroup vgroup; +}; + +struct mlxsw_sp_acl_tcam_flower_rule { + struct mlxsw_sp_acl_tcam_ventry ventry; +}; + +static int +mlxsw_sp_acl_tcam_flower_ruleset_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam, + void *ruleset_priv, + struct mlxsw_afk_element_usage *tmplt_elusage, + unsigned int *p_min_prio, + unsigned int *p_max_prio) +{ + struct mlxsw_sp_acl_tcam_flower_ruleset *ruleset = ruleset_priv; + + return mlxsw_sp_acl_tcam_vgroup_add(mlxsw_sp, tcam, &ruleset->vgroup, + mlxsw_sp_acl_tcam_patterns, + MLXSW_SP_ACL_TCAM_PATTERNS_COUNT, + tmplt_elusage, true, + p_min_prio, p_max_prio); +} + +static void +mlxsw_sp_acl_tcam_flower_ruleset_del(struct mlxsw_sp *mlxsw_sp, + void *ruleset_priv) +{ + struct mlxsw_sp_acl_tcam_flower_ruleset *ruleset = ruleset_priv; + + mlxsw_sp_acl_tcam_vgroup_del(&ruleset->vgroup); +} + +static int +mlxsw_sp_acl_tcam_flower_ruleset_bind(struct mlxsw_sp *mlxsw_sp, + void *ruleset_priv, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + struct mlxsw_sp_acl_tcam_flower_ruleset *ruleset = ruleset_priv; + + return mlxsw_sp_acl_tcam_group_bind(mlxsw_sp, &ruleset->vgroup.group, + mlxsw_sp_port, ingress); +} + +static void +mlxsw_sp_acl_tcam_flower_ruleset_unbind(struct mlxsw_sp *mlxsw_sp, + void *ruleset_priv, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + struct mlxsw_sp_acl_tcam_flower_ruleset *ruleset = ruleset_priv; + + mlxsw_sp_acl_tcam_group_unbind(mlxsw_sp, &ruleset->vgroup.group, + mlxsw_sp_port, ingress); +} + +static u16 +mlxsw_sp_acl_tcam_flower_ruleset_group_id(void *ruleset_priv) +{ + struct mlxsw_sp_acl_tcam_flower_ruleset *ruleset = ruleset_priv; + + return mlxsw_sp_acl_tcam_group_id(&ruleset->vgroup.group); +} + +static int +mlxsw_sp_acl_tcam_flower_rule_add(struct mlxsw_sp *mlxsw_sp, + void *ruleset_priv, void *rule_priv, + struct mlxsw_sp_acl_rule_info *rulei) +{ + struct mlxsw_sp_acl_tcam_flower_ruleset *ruleset = ruleset_priv; + struct mlxsw_sp_acl_tcam_flower_rule *rule = rule_priv; + + return mlxsw_sp_acl_tcam_ventry_add(mlxsw_sp, &ruleset->vgroup, + &rule->ventry, rulei); +} + +static void +mlxsw_sp_acl_tcam_flower_rule_del(struct mlxsw_sp *mlxsw_sp, void *rule_priv) +{ + struct mlxsw_sp_acl_tcam_flower_rule *rule = rule_priv; + + mlxsw_sp_acl_tcam_ventry_del(mlxsw_sp, &rule->ventry); +} + +static int +mlxsw_sp_acl_tcam_flower_rule_action_replace(struct mlxsw_sp *mlxsw_sp, + void *rule_priv, + struct mlxsw_sp_acl_rule_info *rulei) +{ + return -EOPNOTSUPP; +} + +static int +mlxsw_sp_acl_tcam_flower_rule_activity_get(struct mlxsw_sp *mlxsw_sp, + void *rule_priv, bool *activity) +{ + struct mlxsw_sp_acl_tcam_flower_rule *rule = rule_priv; + + return mlxsw_sp_acl_tcam_ventry_activity_get(mlxsw_sp, &rule->ventry, + activity); +} + +static const struct mlxsw_sp_acl_profile_ops mlxsw_sp_acl_tcam_flower_ops = { + .ruleset_priv_size = sizeof(struct mlxsw_sp_acl_tcam_flower_ruleset), + .ruleset_add = mlxsw_sp_acl_tcam_flower_ruleset_add, + .ruleset_del = mlxsw_sp_acl_tcam_flower_ruleset_del, + .ruleset_bind = mlxsw_sp_acl_tcam_flower_ruleset_bind, + .ruleset_unbind = mlxsw_sp_acl_tcam_flower_ruleset_unbind, + .ruleset_group_id = mlxsw_sp_acl_tcam_flower_ruleset_group_id, + .rule_priv_size = sizeof(struct mlxsw_sp_acl_tcam_flower_rule), + .rule_add = mlxsw_sp_acl_tcam_flower_rule_add, + .rule_del = mlxsw_sp_acl_tcam_flower_rule_del, + .rule_action_replace = mlxsw_sp_acl_tcam_flower_rule_action_replace, + .rule_activity_get = mlxsw_sp_acl_tcam_flower_rule_activity_get, +}; + +struct mlxsw_sp_acl_tcam_mr_ruleset { + struct mlxsw_sp_acl_tcam_vchunk *vchunk; + struct mlxsw_sp_acl_tcam_vgroup vgroup; +}; + +struct mlxsw_sp_acl_tcam_mr_rule { + struct mlxsw_sp_acl_tcam_ventry ventry; +}; + +static int +mlxsw_sp_acl_tcam_mr_ruleset_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam, + void *ruleset_priv, + struct mlxsw_afk_element_usage *tmplt_elusage, + unsigned int *p_min_prio, + unsigned int *p_max_prio) +{ + struct mlxsw_sp_acl_tcam_mr_ruleset *ruleset = ruleset_priv; + int err; + + err = mlxsw_sp_acl_tcam_vgroup_add(mlxsw_sp, tcam, &ruleset->vgroup, + mlxsw_sp_acl_tcam_patterns, + MLXSW_SP_ACL_TCAM_PATTERNS_COUNT, + tmplt_elusage, false, + p_min_prio, p_max_prio); + if (err) + return err; + + /* For most of the TCAM clients it would make sense to take a tcam chunk + * only when the first rule is written. This is not the case for + * multicast router as it is required to bind the multicast router to a + * specific ACL Group ID which must exist in HW before multicast router + * is initialized. + */ + ruleset->vchunk = mlxsw_sp_acl_tcam_vchunk_get(mlxsw_sp, + &ruleset->vgroup, 1, + tmplt_elusage); + if (IS_ERR(ruleset->vchunk)) { + err = PTR_ERR(ruleset->vchunk); + goto err_chunk_get; + } + + return 0; + +err_chunk_get: + mlxsw_sp_acl_tcam_vgroup_del(&ruleset->vgroup); + return err; +} + +static void +mlxsw_sp_acl_tcam_mr_ruleset_del(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv) +{ + struct mlxsw_sp_acl_tcam_mr_ruleset *ruleset = ruleset_priv; + + mlxsw_sp_acl_tcam_vchunk_put(mlxsw_sp, ruleset->vchunk); + mlxsw_sp_acl_tcam_vgroup_del(&ruleset->vgroup); +} + +static int +mlxsw_sp_acl_tcam_mr_ruleset_bind(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + /* Binding is done when initializing multicast router */ + return 0; +} + +static void +mlxsw_sp_acl_tcam_mr_ruleset_unbind(struct mlxsw_sp *mlxsw_sp, + void *ruleset_priv, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ +} + +static u16 +mlxsw_sp_acl_tcam_mr_ruleset_group_id(void *ruleset_priv) +{ + struct mlxsw_sp_acl_tcam_mr_ruleset *ruleset = ruleset_priv; + + return mlxsw_sp_acl_tcam_group_id(&ruleset->vgroup.group); +} + +static int +mlxsw_sp_acl_tcam_mr_rule_add(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv, + void *rule_priv, + struct mlxsw_sp_acl_rule_info *rulei) +{ + struct mlxsw_sp_acl_tcam_mr_ruleset *ruleset = ruleset_priv; + struct mlxsw_sp_acl_tcam_mr_rule *rule = rule_priv; + + return mlxsw_sp_acl_tcam_ventry_add(mlxsw_sp, &ruleset->vgroup, + &rule->ventry, rulei); +} + +static void +mlxsw_sp_acl_tcam_mr_rule_del(struct mlxsw_sp *mlxsw_sp, void *rule_priv) +{ + struct mlxsw_sp_acl_tcam_mr_rule *rule = rule_priv; + + mlxsw_sp_acl_tcam_ventry_del(mlxsw_sp, &rule->ventry); +} + +static int +mlxsw_sp_acl_tcam_mr_rule_action_replace(struct mlxsw_sp *mlxsw_sp, + void *rule_priv, + struct mlxsw_sp_acl_rule_info *rulei) +{ + struct mlxsw_sp_acl_tcam_mr_rule *rule = rule_priv; + + return mlxsw_sp_acl_tcam_ventry_action_replace(mlxsw_sp, &rule->ventry, + rulei); +} + +static int +mlxsw_sp_acl_tcam_mr_rule_activity_get(struct mlxsw_sp *mlxsw_sp, + void *rule_priv, bool *activity) +{ + *activity = false; + + return 0; +} + +static const struct mlxsw_sp_acl_profile_ops mlxsw_sp_acl_tcam_mr_ops = { + .ruleset_priv_size = sizeof(struct mlxsw_sp_acl_tcam_mr_ruleset), + .ruleset_add = mlxsw_sp_acl_tcam_mr_ruleset_add, + .ruleset_del = mlxsw_sp_acl_tcam_mr_ruleset_del, + .ruleset_bind = mlxsw_sp_acl_tcam_mr_ruleset_bind, + .ruleset_unbind = mlxsw_sp_acl_tcam_mr_ruleset_unbind, + .ruleset_group_id = mlxsw_sp_acl_tcam_mr_ruleset_group_id, + .rule_priv_size = sizeof(struct mlxsw_sp_acl_tcam_mr_rule), + .rule_add = mlxsw_sp_acl_tcam_mr_rule_add, + .rule_del = mlxsw_sp_acl_tcam_mr_rule_del, + .rule_action_replace = mlxsw_sp_acl_tcam_mr_rule_action_replace, + .rule_activity_get = mlxsw_sp_acl_tcam_mr_rule_activity_get, +}; + +static const struct mlxsw_sp_acl_profile_ops * +mlxsw_sp_acl_tcam_profile_ops_arr[] = { + [MLXSW_SP_ACL_PROFILE_FLOWER] = &mlxsw_sp_acl_tcam_flower_ops, + [MLXSW_SP_ACL_PROFILE_MR] = &mlxsw_sp_acl_tcam_mr_ops, +}; + +const struct mlxsw_sp_acl_profile_ops * +mlxsw_sp_acl_tcam_profile_ops(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_acl_profile profile) +{ + const struct mlxsw_sp_acl_profile_ops *ops; + + if (WARN_ON(profile >= ARRAY_SIZE(mlxsw_sp_acl_tcam_profile_ops_arr))) + return NULL; + ops = mlxsw_sp_acl_tcam_profile_ops_arr[profile]; + if (WARN_ON(!ops)) + return NULL; + return ops; +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.h new file mode 100644 index 000000000..edbbc89e7 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.h @@ -0,0 +1,312 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_SPECTRUM_ACL_TCAM_H +#define _MLXSW_SPECTRUM_ACL_TCAM_H + +#include <linux/list.h> +#include <linux/parman.h> + +#include "reg.h" +#include "spectrum.h" +#include "core_acl_flex_keys.h" + +struct mlxsw_sp_acl_tcam { + unsigned long *used_regions; /* bit array */ + unsigned int max_regions; + unsigned long *used_groups; /* bit array */ + unsigned int max_groups; + unsigned int max_group_size; + struct mutex lock; /* guards vregion list */ + struct list_head vregion_list; + u32 vregion_rehash_intrvl; /* ms */ + unsigned long priv[]; + /* priv has to be always the last item */ +}; + +size_t mlxsw_sp_acl_tcam_priv_size(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_acl_tcam_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam); +void mlxsw_sp_acl_tcam_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam); +u32 mlxsw_sp_acl_tcam_vregion_rehash_intrvl_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam); +int mlxsw_sp_acl_tcam_vregion_rehash_intrvl_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam, + u32 val); +int mlxsw_sp_acl_tcam_priority_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + u32 *priority, bool fillup_priority); + +struct mlxsw_sp_acl_profile_ops { + size_t ruleset_priv_size; + int (*ruleset_add)(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam, void *ruleset_priv, + struct mlxsw_afk_element_usage *tmplt_elusage, + unsigned int *p_min_prio, unsigned int *p_max_prio); + void (*ruleset_del)(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv); + int (*ruleset_bind)(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress); + void (*ruleset_unbind)(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress); + u16 (*ruleset_group_id)(void *ruleset_priv); + size_t rule_priv_size; + int (*rule_add)(struct mlxsw_sp *mlxsw_sp, + void *ruleset_priv, void *rule_priv, + struct mlxsw_sp_acl_rule_info *rulei); + void (*rule_del)(struct mlxsw_sp *mlxsw_sp, void *rule_priv); + int (*rule_action_replace)(struct mlxsw_sp *mlxsw_sp, void *rule_priv, + struct mlxsw_sp_acl_rule_info *rulei); + int (*rule_activity_get)(struct mlxsw_sp *mlxsw_sp, void *rule_priv, + bool *activity); +}; + +const struct mlxsw_sp_acl_profile_ops * +mlxsw_sp_acl_tcam_profile_ops(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_acl_profile profile); + +#define MLXSW_SP_ACL_TCAM_REGION_BASE_COUNT 16 +#define MLXSW_SP_ACL_TCAM_REGION_RESIZE_STEP 16 + +#define MLXSW_SP_ACL_TCAM_CATCHALL_PRIO (~0U) + +#define MLXSW_SP_ACL_TCAM_MASK_LEN \ + (MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN * BITS_PER_BYTE) + +struct mlxsw_sp_acl_tcam_group; +struct mlxsw_sp_acl_tcam_vregion; + +struct mlxsw_sp_acl_tcam_region { + struct mlxsw_sp_acl_tcam_vregion *vregion; + struct mlxsw_sp_acl_tcam_group *group; + struct list_head list; /* Member of a TCAM group */ + enum mlxsw_reg_ptar_key_type key_type; + u16 id; /* ACL ID and region ID - they are same */ + char tcam_region_info[MLXSW_REG_PXXX_TCAM_REGION_INFO_LEN]; + struct mlxsw_afk_key_info *key_info; + struct mlxsw_sp *mlxsw_sp; + unsigned long priv[]; + /* priv has to be always the last item */ +}; + +struct mlxsw_sp_acl_ctcam_region { + struct parman *parman; + const struct mlxsw_sp_acl_ctcam_region_ops *ops; + struct mlxsw_sp_acl_tcam_region *region; +}; + +struct mlxsw_sp_acl_ctcam_chunk { + struct parman_prio parman_prio; +}; + +struct mlxsw_sp_acl_ctcam_entry { + struct parman_item parman_item; +}; + +struct mlxsw_sp_acl_ctcam_region_ops { + int (*entry_insert)(struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_entry *centry, + const char *mask); + void (*entry_remove)(struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_entry *centry); +}; + +int +mlxsw_sp_acl_ctcam_region_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_tcam_region *region, + const struct mlxsw_sp_acl_ctcam_region_ops *ops); +void mlxsw_sp_acl_ctcam_region_fini(struct mlxsw_sp_acl_ctcam_region *cregion); +void mlxsw_sp_acl_ctcam_chunk_init(struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_chunk *cchunk, + unsigned int priority); +void mlxsw_sp_acl_ctcam_chunk_fini(struct mlxsw_sp_acl_ctcam_chunk *cchunk); +int mlxsw_sp_acl_ctcam_entry_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_chunk *cchunk, + struct mlxsw_sp_acl_ctcam_entry *centry, + struct mlxsw_sp_acl_rule_info *rulei, + bool fillup_priority); +void mlxsw_sp_acl_ctcam_entry_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_chunk *cchunk, + struct mlxsw_sp_acl_ctcam_entry *centry); +int mlxsw_sp_acl_ctcam_entry_action_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ctcam_region *cregion, + struct mlxsw_sp_acl_ctcam_entry *centry, + struct mlxsw_sp_acl_rule_info *rulei); +static inline unsigned int +mlxsw_sp_acl_ctcam_entry_offset(struct mlxsw_sp_acl_ctcam_entry *centry) +{ + return centry->parman_item.index; +} + +enum mlxsw_sp_acl_atcam_region_type { + MLXSW_SP_ACL_ATCAM_REGION_TYPE_2KB, + MLXSW_SP_ACL_ATCAM_REGION_TYPE_4KB, + MLXSW_SP_ACL_ATCAM_REGION_TYPE_8KB, + MLXSW_SP_ACL_ATCAM_REGION_TYPE_12KB, + __MLXSW_SP_ACL_ATCAM_REGION_TYPE_MAX, +}; + +#define MLXSW_SP_ACL_ATCAM_REGION_TYPE_MAX \ + (__MLXSW_SP_ACL_ATCAM_REGION_TYPE_MAX - 1) + +struct mlxsw_sp_acl_atcam { + struct mlxsw_sp_acl_erp_core *erp_core; +}; + +struct mlxsw_sp_acl_atcam_region { + struct rhashtable entries_ht; /* A-TCAM only */ + struct list_head entries_list; /* A-TCAM only */ + struct mlxsw_sp_acl_ctcam_region cregion; + const struct mlxsw_sp_acl_atcam_region_ops *ops; + struct mlxsw_sp_acl_tcam_region *region; + struct mlxsw_sp_acl_atcam *atcam; + enum mlxsw_sp_acl_atcam_region_type type; + struct mlxsw_sp_acl_erp_table *erp_table; + void *priv; +}; + +struct mlxsw_sp_acl_atcam_entry_ht_key { + char full_enc_key[MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN]; /* Encoded + * key. + */ + u8 erp_id; +}; + +struct mlxsw_sp_acl_atcam_chunk { + struct mlxsw_sp_acl_ctcam_chunk cchunk; +}; + +struct mlxsw_sp_acl_atcam_entry { + struct rhash_head ht_node; + struct list_head list; /* Member in entries_list */ + struct mlxsw_sp_acl_atcam_entry_ht_key ht_key; + char enc_key[MLXSW_REG_PTCEX_FLEX_KEY_BLOCKS_LEN]; /* Encoded key, + * minus delta bits. + */ + struct { + u16 start; + u8 mask; + u8 value; + } delta_info; + struct mlxsw_sp_acl_ctcam_entry centry; + struct mlxsw_sp_acl_atcam_lkey_id *lkey_id; + struct mlxsw_sp_acl_erp_mask *erp_mask; +}; + +static inline struct mlxsw_sp_acl_atcam_region * +mlxsw_sp_acl_tcam_cregion_aregion(struct mlxsw_sp_acl_ctcam_region *cregion) +{ + return container_of(cregion, struct mlxsw_sp_acl_atcam_region, cregion); +} + +static inline struct mlxsw_sp_acl_atcam_entry * +mlxsw_sp_acl_tcam_centry_aentry(struct mlxsw_sp_acl_ctcam_entry *centry) +{ + return container_of(centry, struct mlxsw_sp_acl_atcam_entry, centry); +} + +int mlxsw_sp_acl_atcam_region_associate(struct mlxsw_sp *mlxsw_sp, + u16 region_id); +int +mlxsw_sp_acl_atcam_region_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam *atcam, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_tcam_region *region, + void *hints_priv, + const struct mlxsw_sp_acl_ctcam_region_ops *ops); +void mlxsw_sp_acl_atcam_region_fini(struct mlxsw_sp_acl_atcam_region *aregion); +void mlxsw_sp_acl_atcam_chunk_init(struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_chunk *achunk, + unsigned int priority); +void mlxsw_sp_acl_atcam_chunk_fini(struct mlxsw_sp_acl_atcam_chunk *achunk); +int mlxsw_sp_acl_atcam_entry_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_chunk *achunk, + struct mlxsw_sp_acl_atcam_entry *aentry, + struct mlxsw_sp_acl_rule_info *rulei); +void mlxsw_sp_acl_atcam_entry_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_chunk *achunk, + struct mlxsw_sp_acl_atcam_entry *aentry); +int mlxsw_sp_acl_atcam_entry_action_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry, + struct mlxsw_sp_acl_rule_info *rulei); +int mlxsw_sp_acl_atcam_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam *atcam); +void mlxsw_sp_acl_atcam_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam *atcam); +void * +mlxsw_sp_acl_atcam_rehash_hints_get(struct mlxsw_sp_acl_atcam_region *aregion); +void mlxsw_sp_acl_atcam_rehash_hints_put(void *hints_priv); + +struct mlxsw_sp_acl_erp_delta; + +u16 mlxsw_sp_acl_erp_delta_start(const struct mlxsw_sp_acl_erp_delta *delta); +u8 mlxsw_sp_acl_erp_delta_mask(const struct mlxsw_sp_acl_erp_delta *delta); +u8 mlxsw_sp_acl_erp_delta_value(const struct mlxsw_sp_acl_erp_delta *delta, + const char *enc_key); +void mlxsw_sp_acl_erp_delta_clear(const struct mlxsw_sp_acl_erp_delta *delta, + const char *enc_key); + +struct mlxsw_sp_acl_erp_mask; + +bool +mlxsw_sp_acl_erp_mask_is_ctcam(const struct mlxsw_sp_acl_erp_mask *erp_mask); +u8 mlxsw_sp_acl_erp_mask_erp_id(const struct mlxsw_sp_acl_erp_mask *erp_mask); +const struct mlxsw_sp_acl_erp_delta * +mlxsw_sp_acl_erp_delta(const struct mlxsw_sp_acl_erp_mask *erp_mask); +struct mlxsw_sp_acl_erp_mask * +mlxsw_sp_acl_erp_mask_get(struct mlxsw_sp_acl_atcam_region *aregion, + const char *mask, bool ctcam); +void mlxsw_sp_acl_erp_mask_put(struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_erp_mask *erp_mask); +int mlxsw_sp_acl_erp_bf_insert(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_erp_mask *erp_mask, + struct mlxsw_sp_acl_atcam_entry *aentry); +void mlxsw_sp_acl_erp_bf_remove(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_erp_mask *erp_mask, + struct mlxsw_sp_acl_atcam_entry *aentry); +void * +mlxsw_sp_acl_erp_rehash_hints_get(struct mlxsw_sp_acl_atcam_region *aregion); +void mlxsw_sp_acl_erp_rehash_hints_put(void *hints_priv); +int mlxsw_sp_acl_erp_region_init(struct mlxsw_sp_acl_atcam_region *aregion, + void *hints_priv); +void mlxsw_sp_acl_erp_region_fini(struct mlxsw_sp_acl_atcam_region *aregion); +int mlxsw_sp_acl_erps_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam *atcam); +void mlxsw_sp_acl_erps_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_atcam *atcam); + +struct mlxsw_sp_acl_bf; + +struct mlxsw_sp_acl_bf_ops { + unsigned int (*index_get)(struct mlxsw_sp_acl_bf *bf, + struct mlxsw_sp_acl_atcam_region *aregion, + struct mlxsw_sp_acl_atcam_entry *aentry); +}; + +int +mlxsw_sp_acl_bf_entry_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_bf *bf, + struct mlxsw_sp_acl_atcam_region *aregion, + unsigned int erp_bank, + struct mlxsw_sp_acl_atcam_entry *aentry); +void +mlxsw_sp_acl_bf_entry_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_bf *bf, + struct mlxsw_sp_acl_atcam_region *aregion, + unsigned int erp_bank, + struct mlxsw_sp_acl_atcam_entry *aentry); +struct mlxsw_sp_acl_bf * +mlxsw_sp_acl_bf_init(struct mlxsw_sp *mlxsw_sp, unsigned int num_erp_banks); +void mlxsw_sp_acl_bf_fini(struct mlxsw_sp_acl_bf *bf); + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_buffers.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_buffers.c new file mode 100644 index 000000000..c9f1c79f3 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_buffers.c @@ -0,0 +1,1786 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/types.h> +#include <linux/dcbnl.h> +#include <linux/if_ether.h> +#include <linux/list.h> +#include <linux/netlink.h> + +#include "spectrum.h" +#include "core.h" +#include "port.h" +#include "reg.h" + +struct mlxsw_sp_sb_pr { + enum mlxsw_reg_sbpr_mode mode; + u32 size; + u8 freeze_mode:1, + freeze_size:1; +}; + +struct mlxsw_cp_sb_occ { + u32 cur; + u32 max; +}; + +struct mlxsw_sp_sb_cm { + u32 min_buff; + u32 max_buff; + u16 pool_index; + struct mlxsw_cp_sb_occ occ; + u8 freeze_pool:1, + freeze_thresh:1; +}; + +#define MLXSW_SP_SB_INFI -1U +#define MLXSW_SP_SB_REST -2U + +struct mlxsw_sp_sb_pm { + u32 min_buff; + u32 max_buff; + struct mlxsw_cp_sb_occ occ; +}; + +struct mlxsw_sp_sb_mm { + u32 min_buff; + u32 max_buff; + u16 pool_index; +}; + +struct mlxsw_sp_sb_pool_des { + enum mlxsw_reg_sbxx_dir dir; + u8 pool; +}; + +#define MLXSW_SP_SB_POOL_ING 0 +#define MLXSW_SP_SB_POOL_EGR 4 +#define MLXSW_SP_SB_POOL_EGR_MC 8 +#define MLXSW_SP_SB_POOL_ING_CPU 9 +#define MLXSW_SP_SB_POOL_EGR_CPU 10 + +static const struct mlxsw_sp_sb_pool_des mlxsw_sp1_sb_pool_dess[] = { + {MLXSW_REG_SBXX_DIR_INGRESS, 0}, + {MLXSW_REG_SBXX_DIR_INGRESS, 1}, + {MLXSW_REG_SBXX_DIR_INGRESS, 2}, + {MLXSW_REG_SBXX_DIR_INGRESS, 3}, + {MLXSW_REG_SBXX_DIR_EGRESS, 0}, + {MLXSW_REG_SBXX_DIR_EGRESS, 1}, + {MLXSW_REG_SBXX_DIR_EGRESS, 2}, + {MLXSW_REG_SBXX_DIR_EGRESS, 3}, + {MLXSW_REG_SBXX_DIR_EGRESS, 15}, + {MLXSW_REG_SBXX_DIR_INGRESS, 4}, + {MLXSW_REG_SBXX_DIR_EGRESS, 4}, +}; + +static const struct mlxsw_sp_sb_pool_des mlxsw_sp2_sb_pool_dess[] = { + {MLXSW_REG_SBXX_DIR_INGRESS, 0}, + {MLXSW_REG_SBXX_DIR_INGRESS, 1}, + {MLXSW_REG_SBXX_DIR_INGRESS, 2}, + {MLXSW_REG_SBXX_DIR_INGRESS, 3}, + {MLXSW_REG_SBXX_DIR_EGRESS, 0}, + {MLXSW_REG_SBXX_DIR_EGRESS, 1}, + {MLXSW_REG_SBXX_DIR_EGRESS, 2}, + {MLXSW_REG_SBXX_DIR_EGRESS, 3}, + {MLXSW_REG_SBXX_DIR_EGRESS, 15}, + {MLXSW_REG_SBXX_DIR_INGRESS, 4}, + {MLXSW_REG_SBXX_DIR_EGRESS, 4}, +}; + +#define MLXSW_SP_SB_ING_TC_COUNT 8 +#define MLXSW_SP_SB_EG_TC_COUNT 16 + +struct mlxsw_sp_sb_port { + struct mlxsw_sp_sb_cm ing_cms[MLXSW_SP_SB_ING_TC_COUNT]; + struct mlxsw_sp_sb_cm eg_cms[MLXSW_SP_SB_EG_TC_COUNT]; + struct mlxsw_sp_sb_pm *pms; +}; + +struct mlxsw_sp_sb { + struct mlxsw_sp_sb_pr *prs; + struct mlxsw_sp_sb_port *ports; + u32 cell_size; + u32 max_headroom_cells; + u64 sb_size; +}; + +struct mlxsw_sp_sb_vals { + unsigned int pool_count; + const struct mlxsw_sp_sb_pool_des *pool_dess; + const struct mlxsw_sp_sb_pm *pms; + const struct mlxsw_sp_sb_pm *pms_cpu; + const struct mlxsw_sp_sb_pr *prs; + const struct mlxsw_sp_sb_mm *mms; + const struct mlxsw_sp_sb_cm *cms_ingress; + const struct mlxsw_sp_sb_cm *cms_egress; + const struct mlxsw_sp_sb_cm *cms_cpu; + unsigned int mms_count; + unsigned int cms_ingress_count; + unsigned int cms_egress_count; + unsigned int cms_cpu_count; +}; + +struct mlxsw_sp_sb_ops { + u32 (*int_buf_size_get)(int mtu, u32 speed); +}; + +u32 mlxsw_sp_cells_bytes(const struct mlxsw_sp *mlxsw_sp, u32 cells) +{ + return mlxsw_sp->sb->cell_size * cells; +} + +u32 mlxsw_sp_bytes_cells(const struct mlxsw_sp *mlxsw_sp, u32 bytes) +{ + return DIV_ROUND_UP(bytes, mlxsw_sp->sb->cell_size); +} + +static u32 mlxsw_sp_port_headroom_8x_adjust(const struct mlxsw_sp_port *mlxsw_sp_port, + u32 size_cells) +{ + /* Ports with eight lanes use two headroom buffers between which the + * configured headroom size is split. Therefore, multiply the calculated + * headroom size by two. + */ + return mlxsw_sp_port->mapping.width == 8 ? 2 * size_cells : size_cells; +} + +static struct mlxsw_sp_sb_pr *mlxsw_sp_sb_pr_get(struct mlxsw_sp *mlxsw_sp, + u16 pool_index) +{ + return &mlxsw_sp->sb->prs[pool_index]; +} + +static bool mlxsw_sp_sb_cm_exists(u8 pg_buff, enum mlxsw_reg_sbxx_dir dir) +{ + if (dir == MLXSW_REG_SBXX_DIR_INGRESS) + return pg_buff < MLXSW_SP_SB_ING_TC_COUNT; + else + return pg_buff < MLXSW_SP_SB_EG_TC_COUNT; +} + +static struct mlxsw_sp_sb_cm *mlxsw_sp_sb_cm_get(struct mlxsw_sp *mlxsw_sp, + u16 local_port, u8 pg_buff, + enum mlxsw_reg_sbxx_dir dir) +{ + struct mlxsw_sp_sb_port *sb_port = &mlxsw_sp->sb->ports[local_port]; + + WARN_ON(!mlxsw_sp_sb_cm_exists(pg_buff, dir)); + if (dir == MLXSW_REG_SBXX_DIR_INGRESS) + return &sb_port->ing_cms[pg_buff]; + else + return &sb_port->eg_cms[pg_buff]; +} + +static struct mlxsw_sp_sb_pm *mlxsw_sp_sb_pm_get(struct mlxsw_sp *mlxsw_sp, + u16 local_port, u16 pool_index) +{ + return &mlxsw_sp->sb->ports[local_port].pms[pool_index]; +} + +static int mlxsw_sp_sb_pr_write(struct mlxsw_sp *mlxsw_sp, u16 pool_index, + enum mlxsw_reg_sbpr_mode mode, + u32 size, bool infi_size) +{ + const struct mlxsw_sp_sb_pool_des *des = + &mlxsw_sp->sb_vals->pool_dess[pool_index]; + char sbpr_pl[MLXSW_REG_SBPR_LEN]; + struct mlxsw_sp_sb_pr *pr; + int err; + + mlxsw_reg_sbpr_pack(sbpr_pl, des->pool, des->dir, mode, + size, infi_size); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sbpr), sbpr_pl); + if (err) + return err; + + if (infi_size) + size = mlxsw_sp_bytes_cells(mlxsw_sp, mlxsw_sp->sb->sb_size); + pr = mlxsw_sp_sb_pr_get(mlxsw_sp, pool_index); + pr->mode = mode; + pr->size = size; + return 0; +} + +static int mlxsw_sp_sb_pr_desc_write(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_reg_sbxx_dir dir, + enum mlxsw_reg_sbpr_mode mode, + u32 size, bool infi_size) +{ + char sbpr_pl[MLXSW_REG_SBPR_LEN]; + + /* The FW default descriptor buffer configuration uses only pool 14 for + * descriptors. + */ + mlxsw_reg_sbpr_pack(sbpr_pl, 14, dir, mode, size, infi_size); + mlxsw_reg_sbpr_desc_set(sbpr_pl, true); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sbpr), sbpr_pl); +} + +static int mlxsw_sp_sb_cm_write(struct mlxsw_sp *mlxsw_sp, u16 local_port, + u8 pg_buff, u32 min_buff, u32 max_buff, + bool infi_max, u16 pool_index) +{ + const struct mlxsw_sp_sb_pool_des *des = + &mlxsw_sp->sb_vals->pool_dess[pool_index]; + char sbcm_pl[MLXSW_REG_SBCM_LEN]; + struct mlxsw_sp_sb_cm *cm; + int err; + + mlxsw_reg_sbcm_pack(sbcm_pl, local_port, pg_buff, des->dir, + min_buff, max_buff, infi_max, des->pool); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sbcm), sbcm_pl); + if (err) + return err; + + if (mlxsw_sp_sb_cm_exists(pg_buff, des->dir)) { + if (infi_max) + max_buff = mlxsw_sp_bytes_cells(mlxsw_sp, + mlxsw_sp->sb->sb_size); + + cm = mlxsw_sp_sb_cm_get(mlxsw_sp, local_port, pg_buff, + des->dir); + cm->min_buff = min_buff; + cm->max_buff = max_buff; + cm->pool_index = pool_index; + } + return 0; +} + +static int mlxsw_sp_sb_pm_write(struct mlxsw_sp *mlxsw_sp, u16 local_port, + u16 pool_index, u32 min_buff, u32 max_buff) +{ + const struct mlxsw_sp_sb_pool_des *des = + &mlxsw_sp->sb_vals->pool_dess[pool_index]; + char sbpm_pl[MLXSW_REG_SBPM_LEN]; + struct mlxsw_sp_sb_pm *pm; + int err; + + mlxsw_reg_sbpm_pack(sbpm_pl, local_port, des->pool, des->dir, false, + min_buff, max_buff); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sbpm), sbpm_pl); + if (err) + return err; + + pm = mlxsw_sp_sb_pm_get(mlxsw_sp, local_port, pool_index); + pm->min_buff = min_buff; + pm->max_buff = max_buff; + return 0; +} + +static int mlxsw_sp_sb_pm_occ_clear(struct mlxsw_sp *mlxsw_sp, u16 local_port, + u16 pool_index, struct list_head *bulk_list) +{ + const struct mlxsw_sp_sb_pool_des *des = + &mlxsw_sp->sb_vals->pool_dess[pool_index]; + char sbpm_pl[MLXSW_REG_SBPM_LEN]; + + if (local_port == MLXSW_PORT_CPU_PORT && + des->dir == MLXSW_REG_SBXX_DIR_INGRESS) + return 0; + + mlxsw_reg_sbpm_pack(sbpm_pl, local_port, des->pool, des->dir, + true, 0, 0); + return mlxsw_reg_trans_query(mlxsw_sp->core, MLXSW_REG(sbpm), sbpm_pl, + bulk_list, NULL, 0); +} + +static void mlxsw_sp_sb_pm_occ_query_cb(struct mlxsw_core *mlxsw_core, + char *sbpm_pl, size_t sbpm_pl_len, + unsigned long cb_priv) +{ + struct mlxsw_sp_sb_pm *pm = (struct mlxsw_sp_sb_pm *) cb_priv; + + mlxsw_reg_sbpm_unpack(sbpm_pl, &pm->occ.cur, &pm->occ.max); +} + +static int mlxsw_sp_sb_pm_occ_query(struct mlxsw_sp *mlxsw_sp, u16 local_port, + u16 pool_index, struct list_head *bulk_list) +{ + const struct mlxsw_sp_sb_pool_des *des = + &mlxsw_sp->sb_vals->pool_dess[pool_index]; + char sbpm_pl[MLXSW_REG_SBPM_LEN]; + struct mlxsw_sp_sb_pm *pm; + + if (local_port == MLXSW_PORT_CPU_PORT && + des->dir == MLXSW_REG_SBXX_DIR_INGRESS) + return 0; + + pm = mlxsw_sp_sb_pm_get(mlxsw_sp, local_port, pool_index); + mlxsw_reg_sbpm_pack(sbpm_pl, local_port, des->pool, des->dir, + false, 0, 0); + return mlxsw_reg_trans_query(mlxsw_sp->core, MLXSW_REG(sbpm), sbpm_pl, + bulk_list, + mlxsw_sp_sb_pm_occ_query_cb, + (unsigned long) pm); +} + +void mlxsw_sp_hdroom_prios_reset_buf_idx(struct mlxsw_sp_hdroom *hdroom) +{ + int prio; + + for (prio = 0; prio < IEEE_8021QAZ_MAX_TCS; prio++) { + switch (hdroom->mode) { + case MLXSW_SP_HDROOM_MODE_DCB: + hdroom->prios.prio[prio].buf_idx = hdroom->prios.prio[prio].ets_buf_idx; + break; + case MLXSW_SP_HDROOM_MODE_TC: + hdroom->prios.prio[prio].buf_idx = hdroom->prios.prio[prio].set_buf_idx; + break; + } + } +} + +void mlxsw_sp_hdroom_bufs_reset_lossiness(struct mlxsw_sp_hdroom *hdroom) +{ + int prio; + int i; + + for (i = 0; i < DCBX_MAX_BUFFERS; i++) + hdroom->bufs.buf[i].lossy = true; + + for (prio = 0; prio < IEEE_8021Q_MAX_PRIORITIES; prio++) { + if (!hdroom->prios.prio[prio].lossy) + hdroom->bufs.buf[hdroom->prios.prio[prio].buf_idx].lossy = false; + } +} + +static u16 mlxsw_sp_hdroom_buf_threshold_get(const struct mlxsw_sp *mlxsw_sp, int mtu) +{ + return 2 * mlxsw_sp_bytes_cells(mlxsw_sp, mtu); +} + +static void mlxsw_sp_hdroom_buf_pack(char *pbmc_pl, int index, u16 size, u16 thres, bool lossy) +{ + if (lossy) + mlxsw_reg_pbmc_lossy_buffer_pack(pbmc_pl, index, size); + else + mlxsw_reg_pbmc_lossless_buffer_pack(pbmc_pl, index, size, + thres); +} + +static u16 mlxsw_sp_hdroom_buf_delay_get(const struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_hdroom *hdroom) +{ + u16 delay_cells; + + delay_cells = mlxsw_sp_bytes_cells(mlxsw_sp, hdroom->delay_bytes); + + /* In the worst case scenario the delay will be made up of packets that + * are all of size CELL_SIZE + 1, which means each packet will require + * almost twice its true size when buffered in the switch. We therefore + * multiply this value by the "cell factor", which is close to 2. + * + * Another MTU is added in case the transmitting host already started + * transmitting a maximum length frame when the PFC packet was received. + */ + return 2 * delay_cells + mlxsw_sp_bytes_cells(mlxsw_sp, hdroom->mtu); +} + +static u32 mlxsw_sp_hdroom_int_buf_size_get(struct mlxsw_sp *mlxsw_sp, int mtu, u32 speed) +{ + u32 buffsize = mlxsw_sp->sb_ops->int_buf_size_get(mtu, speed); + + return mlxsw_sp_bytes_cells(mlxsw_sp, buffsize) + 1; +} + +static bool mlxsw_sp_hdroom_buf_is_used(const struct mlxsw_sp_hdroom *hdroom, int buf) +{ + int prio; + + for (prio = 0; prio < IEEE_8021QAZ_MAX_TCS; prio++) { + if (hdroom->prios.prio[prio].buf_idx == buf) + return true; + } + return false; +} + +void mlxsw_sp_hdroom_bufs_reset_sizes(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_hdroom *hdroom) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 reserve_cells; + int i; + + /* Internal buffer. */ + reserve_cells = mlxsw_sp_hdroom_int_buf_size_get(mlxsw_sp, mlxsw_sp_port->max_mtu, + mlxsw_sp_port->max_speed); + reserve_cells = mlxsw_sp_port_headroom_8x_adjust(mlxsw_sp_port, reserve_cells); + hdroom->int_buf.reserve_cells = reserve_cells; + + if (hdroom->int_buf.enable) + hdroom->int_buf.size_cells = reserve_cells; + else + hdroom->int_buf.size_cells = 0; + + /* PG buffers. */ + for (i = 0; i < DCBX_MAX_BUFFERS; i++) { + struct mlxsw_sp_hdroom_buf *buf = &hdroom->bufs.buf[i]; + u16 thres_cells; + u16 delay_cells; + + if (!mlxsw_sp_hdroom_buf_is_used(hdroom, i)) { + thres_cells = 0; + delay_cells = 0; + } else if (buf->lossy) { + thres_cells = mlxsw_sp_hdroom_buf_threshold_get(mlxsw_sp, hdroom->mtu); + delay_cells = 0; + } else { + thres_cells = mlxsw_sp_hdroom_buf_threshold_get(mlxsw_sp, hdroom->mtu); + delay_cells = mlxsw_sp_hdroom_buf_delay_get(mlxsw_sp, hdroom); + } + + thres_cells = mlxsw_sp_port_headroom_8x_adjust(mlxsw_sp_port, thres_cells); + delay_cells = mlxsw_sp_port_headroom_8x_adjust(mlxsw_sp_port, delay_cells); + + buf->thres_cells = thres_cells; + if (hdroom->mode == MLXSW_SP_HDROOM_MODE_DCB) { + buf->size_cells = thres_cells + delay_cells; + } else { + /* Do not allow going below the minimum size, even if + * the user requested it. + */ + buf->size_cells = max(buf->set_size_cells, buf->thres_cells); + } + } +} + +#define MLXSW_SP_PB_UNUSED 8 + +static int mlxsw_sp_hdroom_configure_buffers(struct mlxsw_sp_port *mlxsw_sp_port, + const struct mlxsw_sp_hdroom *hdroom, bool force) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char pbmc_pl[MLXSW_REG_PBMC_LEN]; + bool dirty; + int err; + int i; + + dirty = memcmp(&mlxsw_sp_port->hdroom->bufs, &hdroom->bufs, sizeof(hdroom->bufs)); + if (!dirty && !force) + return 0; + + mlxsw_reg_pbmc_pack(pbmc_pl, mlxsw_sp_port->local_port, 0xffff, 0xffff / 2); + for (i = 0; i < MLXSW_SP_PB_COUNT; i++) { + const struct mlxsw_sp_hdroom_buf *buf = &hdroom->bufs.buf[i]; + + if (i == MLXSW_SP_PB_UNUSED) + continue; + + mlxsw_sp_hdroom_buf_pack(pbmc_pl, i, buf->size_cells, buf->thres_cells, buf->lossy); + } + + mlxsw_reg_pbmc_lossy_buffer_pack(pbmc_pl, MLXSW_REG_PBMC_PORT_SHARED_BUF_IDX, 0); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pbmc), pbmc_pl); + if (err) + return err; + + mlxsw_sp_port->hdroom->bufs = hdroom->bufs; + return 0; +} + +static int mlxsw_sp_hdroom_configure_priomap(struct mlxsw_sp_port *mlxsw_sp_port, + const struct mlxsw_sp_hdroom *hdroom, bool force) +{ + char pptb_pl[MLXSW_REG_PPTB_LEN]; + bool dirty; + int prio; + int err; + + dirty = memcmp(&mlxsw_sp_port->hdroom->prios, &hdroom->prios, sizeof(hdroom->prios)); + if (!dirty && !force) + return 0; + + mlxsw_reg_pptb_pack(pptb_pl, mlxsw_sp_port->local_port); + for (prio = 0; prio < IEEE_8021QAZ_MAX_TCS; prio++) + mlxsw_reg_pptb_prio_to_buff_pack(pptb_pl, prio, hdroom->prios.prio[prio].buf_idx); + + err = mlxsw_reg_write(mlxsw_sp_port->mlxsw_sp->core, MLXSW_REG(pptb), pptb_pl); + if (err) + return err; + + mlxsw_sp_port->hdroom->prios = hdroom->prios; + return 0; +} + +static int mlxsw_sp_hdroom_configure_int_buf(struct mlxsw_sp_port *mlxsw_sp_port, + const struct mlxsw_sp_hdroom *hdroom, bool force) +{ + char sbib_pl[MLXSW_REG_SBIB_LEN]; + bool dirty; + int err; + + dirty = memcmp(&mlxsw_sp_port->hdroom->int_buf, &hdroom->int_buf, sizeof(hdroom->int_buf)); + if (!dirty && !force) + return 0; + + mlxsw_reg_sbib_pack(sbib_pl, mlxsw_sp_port->local_port, hdroom->int_buf.size_cells); + err = mlxsw_reg_write(mlxsw_sp_port->mlxsw_sp->core, MLXSW_REG(sbib), sbib_pl); + if (err) + return err; + + mlxsw_sp_port->hdroom->int_buf = hdroom->int_buf; + return 0; +} + +static bool mlxsw_sp_hdroom_bufs_fit(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_hdroom *hdroom) +{ + u32 taken_headroom_cells = 0; + int i; + + for (i = 0; i < MLXSW_SP_PB_COUNT; i++) + taken_headroom_cells += hdroom->bufs.buf[i].size_cells; + + taken_headroom_cells += hdroom->int_buf.reserve_cells; + return taken_headroom_cells <= mlxsw_sp->sb->max_headroom_cells; +} + +static int __mlxsw_sp_hdroom_configure(struct mlxsw_sp_port *mlxsw_sp_port, + const struct mlxsw_sp_hdroom *hdroom, bool force) +{ + struct mlxsw_sp_hdroom orig_hdroom; + struct mlxsw_sp_hdroom tmp_hdroom; + int err; + int i; + + /* Port buffers need to be configured in three steps. First, all buffers + * with non-zero size are configured. Then, prio-to-buffer map is + * updated, allowing traffic to flow to the now non-zero buffers. + * Finally, zero-sized buffers are configured, because now no traffic + * should be directed to them anymore. This way, in a non-congested + * system, no packet drops are introduced by the reconfiguration. + */ + + orig_hdroom = *mlxsw_sp_port->hdroom; + tmp_hdroom = orig_hdroom; + for (i = 0; i < MLXSW_SP_PB_COUNT; i++) { + if (hdroom->bufs.buf[i].size_cells) + tmp_hdroom.bufs.buf[i] = hdroom->bufs.buf[i]; + } + + if (!mlxsw_sp_hdroom_bufs_fit(mlxsw_sp_port->mlxsw_sp, &tmp_hdroom) || + !mlxsw_sp_hdroom_bufs_fit(mlxsw_sp_port->mlxsw_sp, hdroom)) + return -ENOBUFS; + + err = mlxsw_sp_hdroom_configure_buffers(mlxsw_sp_port, &tmp_hdroom, force); + if (err) + return err; + + err = mlxsw_sp_hdroom_configure_priomap(mlxsw_sp_port, hdroom, force); + if (err) + goto err_configure_priomap; + + err = mlxsw_sp_hdroom_configure_buffers(mlxsw_sp_port, hdroom, false); + if (err) + goto err_configure_buffers; + + err = mlxsw_sp_hdroom_configure_int_buf(mlxsw_sp_port, hdroom, false); + if (err) + goto err_configure_int_buf; + + *mlxsw_sp_port->hdroom = *hdroom; + return 0; + +err_configure_int_buf: + mlxsw_sp_hdroom_configure_buffers(mlxsw_sp_port, &tmp_hdroom, false); +err_configure_buffers: + mlxsw_sp_hdroom_configure_priomap(mlxsw_sp_port, &tmp_hdroom, false); +err_configure_priomap: + mlxsw_sp_hdroom_configure_buffers(mlxsw_sp_port, &orig_hdroom, false); + return err; +} + +int mlxsw_sp_hdroom_configure(struct mlxsw_sp_port *mlxsw_sp_port, + const struct mlxsw_sp_hdroom *hdroom) +{ + return __mlxsw_sp_hdroom_configure(mlxsw_sp_port, hdroom, false); +} + +static int mlxsw_sp_port_headroom_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_hdroom hdroom = {}; + u32 size9; + int prio; + + hdroom.mtu = mlxsw_sp_port->dev->mtu; + hdroom.mode = MLXSW_SP_HDROOM_MODE_DCB; + for (prio = 0; prio < IEEE_8021QAZ_MAX_TCS; prio++) + hdroom.prios.prio[prio].lossy = true; + + mlxsw_sp_hdroom_bufs_reset_lossiness(&hdroom); + mlxsw_sp_hdroom_bufs_reset_sizes(mlxsw_sp_port, &hdroom); + + /* Buffer 9 is used for control traffic. */ + size9 = mlxsw_sp_port_headroom_8x_adjust(mlxsw_sp_port, mlxsw_sp_port->max_mtu); + hdroom.bufs.buf[9].size_cells = mlxsw_sp_bytes_cells(mlxsw_sp, size9); + + return __mlxsw_sp_hdroom_configure(mlxsw_sp_port, &hdroom, true); +} + +static int mlxsw_sp_sb_port_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_sb_port *sb_port) +{ + struct mlxsw_sp_sb_pm *pms; + + pms = kcalloc(mlxsw_sp->sb_vals->pool_count, sizeof(*pms), + GFP_KERNEL); + if (!pms) + return -ENOMEM; + sb_port->pms = pms; + return 0; +} + +static void mlxsw_sp_sb_port_fini(struct mlxsw_sp_sb_port *sb_port) +{ + kfree(sb_port->pms); +} + +static int mlxsw_sp_sb_ports_init(struct mlxsw_sp *mlxsw_sp) +{ + unsigned int max_ports = mlxsw_core_max_ports(mlxsw_sp->core); + struct mlxsw_sp_sb_pr *prs; + int i; + int err; + + mlxsw_sp->sb->ports = kcalloc(max_ports, + sizeof(struct mlxsw_sp_sb_port), + GFP_KERNEL); + if (!mlxsw_sp->sb->ports) + return -ENOMEM; + + prs = kcalloc(mlxsw_sp->sb_vals->pool_count, sizeof(*prs), + GFP_KERNEL); + if (!prs) { + err = -ENOMEM; + goto err_alloc_prs; + } + mlxsw_sp->sb->prs = prs; + + for (i = 0; i < max_ports; i++) { + err = mlxsw_sp_sb_port_init(mlxsw_sp, &mlxsw_sp->sb->ports[i]); + if (err) + goto err_sb_port_init; + } + + return 0; + +err_sb_port_init: + for (i--; i >= 0; i--) + mlxsw_sp_sb_port_fini(&mlxsw_sp->sb->ports[i]); + kfree(mlxsw_sp->sb->prs); +err_alloc_prs: + kfree(mlxsw_sp->sb->ports); + return err; +} + +static void mlxsw_sp_sb_ports_fini(struct mlxsw_sp *mlxsw_sp) +{ + int max_ports = mlxsw_core_max_ports(mlxsw_sp->core); + int i; + + for (i = max_ports - 1; i >= 0; i--) + mlxsw_sp_sb_port_fini(&mlxsw_sp->sb->ports[i]); + kfree(mlxsw_sp->sb->prs); + kfree(mlxsw_sp->sb->ports); +} + +#define MLXSW_SP_SB_PR(_mode, _size) \ + { \ + .mode = _mode, \ + .size = _size, \ + } + +#define MLXSW_SP_SB_PR_EXT(_mode, _size, _freeze_mode, _freeze_size) \ + { \ + .mode = _mode, \ + .size = _size, \ + .freeze_mode = _freeze_mode, \ + .freeze_size = _freeze_size, \ + } + +#define MLXSW_SP1_SB_PR_CPU_SIZE (256 * 1000) + +/* Order according to mlxsw_sp1_sb_pool_dess */ +static const struct mlxsw_sp_sb_pr mlxsw_sp1_sb_prs[] = { + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_DYNAMIC, MLXSW_SP_SB_REST), + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_DYNAMIC, 0), + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_DYNAMIC, 0), + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_DYNAMIC, 0), + MLXSW_SP_SB_PR_EXT(MLXSW_REG_SBPR_MODE_DYNAMIC, MLXSW_SP_SB_REST, + true, false), + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_DYNAMIC, 0), + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_DYNAMIC, 0), + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_DYNAMIC, 0), + MLXSW_SP_SB_PR_EXT(MLXSW_REG_SBPR_MODE_STATIC, MLXSW_SP_SB_INFI, + true, true), + MLXSW_SP_SB_PR_EXT(MLXSW_REG_SBPR_MODE_DYNAMIC, + MLXSW_SP1_SB_PR_CPU_SIZE, true, false), + MLXSW_SP_SB_PR_EXT(MLXSW_REG_SBPR_MODE_DYNAMIC, + MLXSW_SP1_SB_PR_CPU_SIZE, true, false), +}; + +#define MLXSW_SP2_SB_PR_CPU_SIZE (256 * 1000) + +/* Order according to mlxsw_sp2_sb_pool_dess */ +static const struct mlxsw_sp_sb_pr mlxsw_sp2_sb_prs[] = { + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_DYNAMIC, MLXSW_SP_SB_REST), + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_STATIC, 0), + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_STATIC, 0), + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_STATIC, 0), + MLXSW_SP_SB_PR_EXT(MLXSW_REG_SBPR_MODE_DYNAMIC, MLXSW_SP_SB_REST, + true, false), + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_STATIC, 0), + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_STATIC, 0), + MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_STATIC, 0), + MLXSW_SP_SB_PR_EXT(MLXSW_REG_SBPR_MODE_STATIC, MLXSW_SP_SB_INFI, + true, true), + MLXSW_SP_SB_PR_EXT(MLXSW_REG_SBPR_MODE_DYNAMIC, + MLXSW_SP2_SB_PR_CPU_SIZE, true, false), + MLXSW_SP_SB_PR_EXT(MLXSW_REG_SBPR_MODE_DYNAMIC, + MLXSW_SP2_SB_PR_CPU_SIZE, true, false), +}; + +static int mlxsw_sp_sb_prs_init(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_sb_pr *prs, + const struct mlxsw_sp_sb_pool_des *pool_dess, + size_t prs_len) +{ + /* Round down, unlike mlxsw_sp_bytes_cells(). */ + u32 sb_cells = div_u64(mlxsw_sp->sb->sb_size, mlxsw_sp->sb->cell_size); + u32 rest_cells[2] = {sb_cells, sb_cells}; + int i; + int err; + + /* Calculate how much space to give to the "REST" pools in either + * direction. + */ + for (i = 0; i < prs_len; i++) { + enum mlxsw_reg_sbxx_dir dir = pool_dess[i].dir; + u32 size = prs[i].size; + u32 size_cells; + + if (size == MLXSW_SP_SB_INFI || size == MLXSW_SP_SB_REST) + continue; + + size_cells = mlxsw_sp_bytes_cells(mlxsw_sp, size); + if (WARN_ON_ONCE(size_cells > rest_cells[dir])) + continue; + + rest_cells[dir] -= size_cells; + } + + for (i = 0; i < prs_len; i++) { + u32 size = prs[i].size; + u32 size_cells; + + if (size == MLXSW_SP_SB_INFI) { + err = mlxsw_sp_sb_pr_write(mlxsw_sp, i, prs[i].mode, + 0, true); + } else if (size == MLXSW_SP_SB_REST) { + size_cells = rest_cells[pool_dess[i].dir]; + err = mlxsw_sp_sb_pr_write(mlxsw_sp, i, prs[i].mode, + size_cells, false); + } else { + size_cells = mlxsw_sp_bytes_cells(mlxsw_sp, size); + err = mlxsw_sp_sb_pr_write(mlxsw_sp, i, prs[i].mode, + size_cells, false); + } + if (err) + return err; + } + + err = mlxsw_sp_sb_pr_desc_write(mlxsw_sp, MLXSW_REG_SBXX_DIR_INGRESS, + MLXSW_REG_SBPR_MODE_DYNAMIC, 0, true); + if (err) + return err; + + err = mlxsw_sp_sb_pr_desc_write(mlxsw_sp, MLXSW_REG_SBXX_DIR_EGRESS, + MLXSW_REG_SBPR_MODE_DYNAMIC, 0, true); + if (err) + return err; + + return 0; +} + +#define MLXSW_SP_SB_CM(_min_buff, _max_buff, _pool) \ + { \ + .min_buff = _min_buff, \ + .max_buff = _max_buff, \ + .pool_index = _pool, \ + } + +#define MLXSW_SP_SB_CM_ING(_min_buff, _max_buff) \ + { \ + .min_buff = _min_buff, \ + .max_buff = _max_buff, \ + .pool_index = MLXSW_SP_SB_POOL_ING, \ + } + +#define MLXSW_SP_SB_CM_EGR(_min_buff, _max_buff) \ + { \ + .min_buff = _min_buff, \ + .max_buff = _max_buff, \ + .pool_index = MLXSW_SP_SB_POOL_EGR, \ + } + +#define MLXSW_SP_SB_CM_EGR_MC(_min_buff, _max_buff) \ + { \ + .min_buff = _min_buff, \ + .max_buff = _max_buff, \ + .pool_index = MLXSW_SP_SB_POOL_EGR_MC, \ + .freeze_pool = true, \ + .freeze_thresh = true, \ + } + +static const struct mlxsw_sp_sb_cm mlxsw_sp1_sb_cms_ingress[] = { + MLXSW_SP_SB_CM_ING(10000, 8), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, 0), /* dummy, this PG does not exist */ + MLXSW_SP_SB_CM(10000, 8, MLXSW_SP_SB_POOL_ING_CPU), +}; + +static const struct mlxsw_sp_sb_cm mlxsw_sp2_sb_cms_ingress[] = { + MLXSW_SP_SB_CM_ING(0, 7), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_CM_ING(0, 0), /* dummy, this PG does not exist */ + MLXSW_SP_SB_CM(10000, 8, MLXSW_SP_SB_POOL_ING_CPU), +}; + +static const struct mlxsw_sp_sb_cm mlxsw_sp1_sb_cms_egress[] = { + MLXSW_SP_SB_CM_EGR(1500, 9), + MLXSW_SP_SB_CM_EGR(1500, 9), + MLXSW_SP_SB_CM_EGR(1500, 9), + MLXSW_SP_SB_CM_EGR(1500, 9), + MLXSW_SP_SB_CM_EGR(1500, 9), + MLXSW_SP_SB_CM_EGR(1500, 9), + MLXSW_SP_SB_CM_EGR(1500, 9), + MLXSW_SP_SB_CM_EGR(1500, 9), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR(1, 0xff), +}; + +static const struct mlxsw_sp_sb_cm mlxsw_sp2_sb_cms_egress[] = { + MLXSW_SP_SB_CM_EGR(0, 7), + MLXSW_SP_SB_CM_EGR(0, 7), + MLXSW_SP_SB_CM_EGR(0, 7), + MLXSW_SP_SB_CM_EGR(0, 7), + MLXSW_SP_SB_CM_EGR(0, 7), + MLXSW_SP_SB_CM_EGR(0, 7), + MLXSW_SP_SB_CM_EGR(0, 7), + MLXSW_SP_SB_CM_EGR(0, 7), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR_MC(0, MLXSW_SP_SB_INFI), + MLXSW_SP_SB_CM_EGR(1, 0xff), +}; + +#define MLXSW_SP_CPU_PORT_SB_CM MLXSW_SP_SB_CM(0, 0, MLXSW_SP_SB_POOL_EGR_CPU) + +static const struct mlxsw_sp_sb_cm mlxsw_sp_cpu_port_sb_cms[] = { + MLXSW_SP_SB_CM(1000, 8, MLXSW_SP_SB_POOL_EGR_CPU), + MLXSW_SP_SB_CM(1000, 8, MLXSW_SP_SB_POOL_EGR_CPU), + MLXSW_SP_SB_CM(1000, 8, MLXSW_SP_SB_POOL_EGR_CPU), + MLXSW_SP_SB_CM(1000, 8, MLXSW_SP_SB_POOL_EGR_CPU), + MLXSW_SP_SB_CM(1000, 8, MLXSW_SP_SB_POOL_EGR_CPU), + MLXSW_SP_SB_CM(1000, 8, MLXSW_SP_SB_POOL_EGR_CPU), + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_SB_CM(1000, 8, MLXSW_SP_SB_POOL_EGR_CPU), + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, + MLXSW_SP_CPU_PORT_SB_CM, +}; + +static bool +mlxsw_sp_sb_pool_is_static(struct mlxsw_sp *mlxsw_sp, u16 pool_index) +{ + struct mlxsw_sp_sb_pr *pr = mlxsw_sp_sb_pr_get(mlxsw_sp, pool_index); + + return pr->mode == MLXSW_REG_SBPR_MODE_STATIC; +} + +static int __mlxsw_sp_sb_cms_init(struct mlxsw_sp *mlxsw_sp, u16 local_port, + enum mlxsw_reg_sbxx_dir dir, + const struct mlxsw_sp_sb_cm *cms, + size_t cms_len) +{ + const struct mlxsw_sp_sb_vals *sb_vals = mlxsw_sp->sb_vals; + int i; + int err; + + for (i = 0; i < cms_len; i++) { + const struct mlxsw_sp_sb_cm *cm; + u32 min_buff; + u32 max_buff; + + if (i == 8 && dir == MLXSW_REG_SBXX_DIR_INGRESS) + continue; /* PG number 8 does not exist, skip it */ + cm = &cms[i]; + if (WARN_ON(sb_vals->pool_dess[cm->pool_index].dir != dir)) + continue; + + min_buff = mlxsw_sp_bytes_cells(mlxsw_sp, cm->min_buff); + max_buff = cm->max_buff; + if (max_buff == MLXSW_SP_SB_INFI) { + err = mlxsw_sp_sb_cm_write(mlxsw_sp, local_port, i, + min_buff, 0, + true, cm->pool_index); + } else { + if (mlxsw_sp_sb_pool_is_static(mlxsw_sp, + cm->pool_index)) + max_buff = mlxsw_sp_bytes_cells(mlxsw_sp, + max_buff); + err = mlxsw_sp_sb_cm_write(mlxsw_sp, local_port, i, + min_buff, max_buff, + false, cm->pool_index); + } + if (err) + return err; + } + return 0; +} + +static int mlxsw_sp_port_sb_cms_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + int err; + + err = __mlxsw_sp_sb_cms_init(mlxsw_sp, + mlxsw_sp_port->local_port, + MLXSW_REG_SBXX_DIR_INGRESS, + mlxsw_sp->sb_vals->cms_ingress, + mlxsw_sp->sb_vals->cms_ingress_count); + if (err) + return err; + return __mlxsw_sp_sb_cms_init(mlxsw_sp_port->mlxsw_sp, + mlxsw_sp_port->local_port, + MLXSW_REG_SBXX_DIR_EGRESS, + mlxsw_sp->sb_vals->cms_egress, + mlxsw_sp->sb_vals->cms_egress_count); +} + +static int mlxsw_sp_cpu_port_sb_cms_init(struct mlxsw_sp *mlxsw_sp) +{ + return __mlxsw_sp_sb_cms_init(mlxsw_sp, 0, MLXSW_REG_SBXX_DIR_EGRESS, + mlxsw_sp->sb_vals->cms_cpu, + mlxsw_sp->sb_vals->cms_cpu_count); +} + +#define MLXSW_SP_SB_PM(_min_buff, _max_buff) \ + { \ + .min_buff = _min_buff, \ + .max_buff = _max_buff, \ + } + +/* Order according to mlxsw_sp1_sb_pool_dess */ +static const struct mlxsw_sp_sb_pm mlxsw_sp1_sb_pms[] = { + MLXSW_SP_SB_PM(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MAX), + MLXSW_SP_SB_PM(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_PM(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_PM(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_PM(0, 7), + MLXSW_SP_SB_PM(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_PM(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_PM(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), + MLXSW_SP_SB_PM(10000, 90000), + MLXSW_SP_SB_PM(0, 8), /* 50% occupancy */ + MLXSW_SP_SB_PM(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), +}; + +/* Order according to mlxsw_sp2_sb_pool_dess */ +static const struct mlxsw_sp_sb_pm mlxsw_sp2_sb_pms[] = { + MLXSW_SP_SB_PM(0, 7), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 7), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(10000, 90000), + MLXSW_SP_SB_PM(0, 8), /* 50% occupancy */ + MLXSW_SP_SB_PM(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN), +}; + +/* Order according to mlxsw_sp*_sb_pool_dess */ +static const struct mlxsw_sp_sb_pm mlxsw_sp_cpu_port_sb_pms[] = { + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, 90000), + MLXSW_SP_SB_PM(0, 0), + MLXSW_SP_SB_PM(0, MLXSW_REG_SBXX_DYN_MAX_BUFF_MAX), +}; + +static int mlxsw_sp_sb_pms_init(struct mlxsw_sp *mlxsw_sp, u16 local_port, + const struct mlxsw_sp_sb_pm *pms, + bool skip_ingress) +{ + int i, err; + + for (i = 0; i < mlxsw_sp->sb_vals->pool_count; i++) { + const struct mlxsw_sp_sb_pm *pm = &pms[i]; + const struct mlxsw_sp_sb_pool_des *des; + u32 max_buff; + u32 min_buff; + + des = &mlxsw_sp->sb_vals->pool_dess[i]; + if (skip_ingress && des->dir == MLXSW_REG_SBXX_DIR_INGRESS) + continue; + + min_buff = mlxsw_sp_bytes_cells(mlxsw_sp, pm->min_buff); + max_buff = pm->max_buff; + if (mlxsw_sp_sb_pool_is_static(mlxsw_sp, i)) + max_buff = mlxsw_sp_bytes_cells(mlxsw_sp, max_buff); + err = mlxsw_sp_sb_pm_write(mlxsw_sp, local_port, i, min_buff, + max_buff); + if (err) + return err; + } + return 0; +} + +static int mlxsw_sp_port_sb_pms_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + + return mlxsw_sp_sb_pms_init(mlxsw_sp, mlxsw_sp_port->local_port, + mlxsw_sp->sb_vals->pms, false); +} + +static int mlxsw_sp_cpu_port_sb_pms_init(struct mlxsw_sp *mlxsw_sp) +{ + return mlxsw_sp_sb_pms_init(mlxsw_sp, 0, mlxsw_sp->sb_vals->pms_cpu, + true); +} + +#define MLXSW_SP_SB_MM(_min_buff, _max_buff) \ + { \ + .min_buff = _min_buff, \ + .max_buff = _max_buff, \ + .pool_index = MLXSW_SP_SB_POOL_EGR, \ + } + +static const struct mlxsw_sp_sb_mm mlxsw_sp_sb_mms[] = { + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), + MLXSW_SP_SB_MM(0, 6), +}; + +static int mlxsw_sp_sb_mms_init(struct mlxsw_sp *mlxsw_sp) +{ + char sbmm_pl[MLXSW_REG_SBMM_LEN]; + int i; + int err; + + for (i = 0; i < mlxsw_sp->sb_vals->mms_count; i++) { + const struct mlxsw_sp_sb_pool_des *des; + const struct mlxsw_sp_sb_mm *mc; + u32 min_buff; + + mc = &mlxsw_sp->sb_vals->mms[i]; + des = &mlxsw_sp->sb_vals->pool_dess[mc->pool_index]; + /* All pools used by sb_mm's are initialized using dynamic + * thresholds, therefore 'max_buff' isn't specified in cells. + */ + min_buff = mlxsw_sp_bytes_cells(mlxsw_sp, mc->min_buff); + mlxsw_reg_sbmm_pack(sbmm_pl, i, min_buff, mc->max_buff, + des->pool); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sbmm), sbmm_pl); + if (err) + return err; + } + return 0; +} + +static void mlxsw_sp_pool_count(struct mlxsw_sp *mlxsw_sp, + u16 *p_ingress_len, u16 *p_egress_len) +{ + int i; + + for (i = 0; i < mlxsw_sp->sb_vals->pool_count; ++i) { + if (mlxsw_sp->sb_vals->pool_dess[i].dir == + MLXSW_REG_SBXX_DIR_INGRESS) + (*p_ingress_len)++; + else + (*p_egress_len)++; + } + + WARN(*p_egress_len == 0, "No egress pools\n"); +} + +const struct mlxsw_sp_sb_vals mlxsw_sp1_sb_vals = { + .pool_count = ARRAY_SIZE(mlxsw_sp1_sb_pool_dess), + .pool_dess = mlxsw_sp1_sb_pool_dess, + .pms = mlxsw_sp1_sb_pms, + .pms_cpu = mlxsw_sp_cpu_port_sb_pms, + .prs = mlxsw_sp1_sb_prs, + .mms = mlxsw_sp_sb_mms, + .cms_ingress = mlxsw_sp1_sb_cms_ingress, + .cms_egress = mlxsw_sp1_sb_cms_egress, + .cms_cpu = mlxsw_sp_cpu_port_sb_cms, + .mms_count = ARRAY_SIZE(mlxsw_sp_sb_mms), + .cms_ingress_count = ARRAY_SIZE(mlxsw_sp1_sb_cms_ingress), + .cms_egress_count = ARRAY_SIZE(mlxsw_sp1_sb_cms_egress), + .cms_cpu_count = ARRAY_SIZE(mlxsw_sp_cpu_port_sb_cms), +}; + +const struct mlxsw_sp_sb_vals mlxsw_sp2_sb_vals = { + .pool_count = ARRAY_SIZE(mlxsw_sp2_sb_pool_dess), + .pool_dess = mlxsw_sp2_sb_pool_dess, + .pms = mlxsw_sp2_sb_pms, + .pms_cpu = mlxsw_sp_cpu_port_sb_pms, + .prs = mlxsw_sp2_sb_prs, + .mms = mlxsw_sp_sb_mms, + .cms_ingress = mlxsw_sp2_sb_cms_ingress, + .cms_egress = mlxsw_sp2_sb_cms_egress, + .cms_cpu = mlxsw_sp_cpu_port_sb_cms, + .mms_count = ARRAY_SIZE(mlxsw_sp_sb_mms), + .cms_ingress_count = ARRAY_SIZE(mlxsw_sp2_sb_cms_ingress), + .cms_egress_count = ARRAY_SIZE(mlxsw_sp2_sb_cms_egress), + .cms_cpu_count = ARRAY_SIZE(mlxsw_sp_cpu_port_sb_cms), +}; + +static u32 mlxsw_sp1_pb_int_buf_size_get(int mtu, u32 speed) +{ + return mtu * 5 / 2; +} + +static u32 __mlxsw_sp_pb_int_buf_size_get(int mtu, u32 speed, u32 buffer_factor) +{ + return 3 * mtu + buffer_factor * speed / 1000; +} + +#define MLXSW_SP2_SPAN_EG_MIRROR_BUFFER_FACTOR 38 + +static u32 mlxsw_sp2_pb_int_buf_size_get(int mtu, u32 speed) +{ + int factor = MLXSW_SP2_SPAN_EG_MIRROR_BUFFER_FACTOR; + + return __mlxsw_sp_pb_int_buf_size_get(mtu, speed, factor); +} + +#define MLXSW_SP3_SPAN_EG_MIRROR_BUFFER_FACTOR 50 + +static u32 mlxsw_sp3_pb_int_buf_size_get(int mtu, u32 speed) +{ + int factor = MLXSW_SP3_SPAN_EG_MIRROR_BUFFER_FACTOR; + + return __mlxsw_sp_pb_int_buf_size_get(mtu, speed, factor); +} + +const struct mlxsw_sp_sb_ops mlxsw_sp1_sb_ops = { + .int_buf_size_get = mlxsw_sp1_pb_int_buf_size_get, +}; + +const struct mlxsw_sp_sb_ops mlxsw_sp2_sb_ops = { + .int_buf_size_get = mlxsw_sp2_pb_int_buf_size_get, +}; + +const struct mlxsw_sp_sb_ops mlxsw_sp3_sb_ops = { + .int_buf_size_get = mlxsw_sp3_pb_int_buf_size_get, +}; + +int mlxsw_sp_buffers_init(struct mlxsw_sp *mlxsw_sp) +{ + u32 max_headroom_size; + u16 ing_pool_count = 0; + u16 eg_pool_count = 0; + int err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, CELL_SIZE)) + return -EIO; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, GUARANTEED_SHARED_BUFFER)) + return -EIO; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, MAX_HEADROOM_SIZE)) + return -EIO; + + mlxsw_sp->sb = kzalloc(sizeof(*mlxsw_sp->sb), GFP_KERNEL); + if (!mlxsw_sp->sb) + return -ENOMEM; + mlxsw_sp->sb->cell_size = MLXSW_CORE_RES_GET(mlxsw_sp->core, CELL_SIZE); + mlxsw_sp->sb->sb_size = MLXSW_CORE_RES_GET(mlxsw_sp->core, + GUARANTEED_SHARED_BUFFER); + max_headroom_size = MLXSW_CORE_RES_GET(mlxsw_sp->core, + MAX_HEADROOM_SIZE); + /* Round down, because this limit must not be overstepped. */ + mlxsw_sp->sb->max_headroom_cells = max_headroom_size / + mlxsw_sp->sb->cell_size; + + err = mlxsw_sp_sb_ports_init(mlxsw_sp); + if (err) + goto err_sb_ports_init; + err = mlxsw_sp_sb_prs_init(mlxsw_sp, mlxsw_sp->sb_vals->prs, + mlxsw_sp->sb_vals->pool_dess, + mlxsw_sp->sb_vals->pool_count); + if (err) + goto err_sb_prs_init; + err = mlxsw_sp_cpu_port_sb_cms_init(mlxsw_sp); + if (err) + goto err_sb_cpu_port_sb_cms_init; + err = mlxsw_sp_cpu_port_sb_pms_init(mlxsw_sp); + if (err) + goto err_sb_cpu_port_pms_init; + err = mlxsw_sp_sb_mms_init(mlxsw_sp); + if (err) + goto err_sb_mms_init; + mlxsw_sp_pool_count(mlxsw_sp, &ing_pool_count, &eg_pool_count); + err = devl_sb_register(priv_to_devlink(mlxsw_sp->core), 0, + mlxsw_sp->sb->sb_size, + ing_pool_count, + eg_pool_count, + MLXSW_SP_SB_ING_TC_COUNT, + MLXSW_SP_SB_EG_TC_COUNT); + if (err) + goto err_devlink_sb_register; + + return 0; + +err_devlink_sb_register: +err_sb_mms_init: +err_sb_cpu_port_pms_init: +err_sb_cpu_port_sb_cms_init: +err_sb_prs_init: + mlxsw_sp_sb_ports_fini(mlxsw_sp); +err_sb_ports_init: + kfree(mlxsw_sp->sb); + return err; +} + +void mlxsw_sp_buffers_fini(struct mlxsw_sp *mlxsw_sp) +{ + devl_sb_unregister(priv_to_devlink(mlxsw_sp->core), 0); + mlxsw_sp_sb_ports_fini(mlxsw_sp); + kfree(mlxsw_sp->sb); +} + +int mlxsw_sp_port_buffers_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + int err; + + mlxsw_sp_port->hdroom = kzalloc(sizeof(*mlxsw_sp_port->hdroom), GFP_KERNEL); + if (!mlxsw_sp_port->hdroom) + return -ENOMEM; + mlxsw_sp_port->hdroom->mtu = mlxsw_sp_port->dev->mtu; + + err = mlxsw_sp_port_headroom_init(mlxsw_sp_port); + if (err) + goto err_headroom_init; + err = mlxsw_sp_port_sb_cms_init(mlxsw_sp_port); + if (err) + goto err_port_sb_cms_init; + err = mlxsw_sp_port_sb_pms_init(mlxsw_sp_port); + if (err) + goto err_port_sb_pms_init; + return 0; + +err_port_sb_pms_init: +err_port_sb_cms_init: +err_headroom_init: + kfree(mlxsw_sp_port->hdroom); + return err; +} + +void mlxsw_sp_port_buffers_fini(struct mlxsw_sp_port *mlxsw_sp_port) +{ + kfree(mlxsw_sp_port->hdroom); +} + +int mlxsw_sp_sb_pool_get(struct mlxsw_core *mlxsw_core, + unsigned int sb_index, u16 pool_index, + struct devlink_sb_pool_info *pool_info) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + enum mlxsw_reg_sbxx_dir dir; + struct mlxsw_sp_sb_pr *pr; + + dir = mlxsw_sp->sb_vals->pool_dess[pool_index].dir; + pr = mlxsw_sp_sb_pr_get(mlxsw_sp, pool_index); + pool_info->pool_type = (enum devlink_sb_pool_type) dir; + pool_info->size = mlxsw_sp_cells_bytes(mlxsw_sp, pr->size); + pool_info->threshold_type = (enum devlink_sb_threshold_type) pr->mode; + pool_info->cell_size = mlxsw_sp->sb->cell_size; + return 0; +} + +int mlxsw_sp_sb_pool_set(struct mlxsw_core *mlxsw_core, + unsigned int sb_index, u16 pool_index, u32 size, + enum devlink_sb_threshold_type threshold_type, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + u32 pool_size = mlxsw_sp_bytes_cells(mlxsw_sp, size); + const struct mlxsw_sp_sb_pr *pr; + enum mlxsw_reg_sbpr_mode mode; + + mode = (enum mlxsw_reg_sbpr_mode) threshold_type; + pr = &mlxsw_sp->sb_vals->prs[pool_index]; + + if (size > MLXSW_CORE_RES_GET(mlxsw_sp->core, + GUARANTEED_SHARED_BUFFER)) { + NL_SET_ERR_MSG_MOD(extack, "Exceeded shared buffer size"); + return -EINVAL; + } + + if (pr->freeze_mode && pr->mode != mode) { + NL_SET_ERR_MSG_MOD(extack, "Changing this pool's threshold type is forbidden"); + return -EINVAL; + } + + if (pr->freeze_size && pr->size != size) { + NL_SET_ERR_MSG_MOD(extack, "Changing this pool's size is forbidden"); + return -EINVAL; + } + + return mlxsw_sp_sb_pr_write(mlxsw_sp, pool_index, mode, + pool_size, false); +} + +#define MLXSW_SP_SB_THRESHOLD_TO_ALPHA_OFFSET (-2) /* 3->1, 16->14 */ + +static u32 mlxsw_sp_sb_threshold_out(struct mlxsw_sp *mlxsw_sp, u16 pool_index, + u32 max_buff) +{ + struct mlxsw_sp_sb_pr *pr = mlxsw_sp_sb_pr_get(mlxsw_sp, pool_index); + + if (pr->mode == MLXSW_REG_SBPR_MODE_DYNAMIC) + return max_buff - MLXSW_SP_SB_THRESHOLD_TO_ALPHA_OFFSET; + return mlxsw_sp_cells_bytes(mlxsw_sp, max_buff); +} + +static int mlxsw_sp_sb_threshold_in(struct mlxsw_sp *mlxsw_sp, u16 pool_index, + u32 threshold, u32 *p_max_buff, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_sb_pr *pr = mlxsw_sp_sb_pr_get(mlxsw_sp, pool_index); + + if (pr->mode == MLXSW_REG_SBPR_MODE_DYNAMIC) { + int val; + + val = threshold + MLXSW_SP_SB_THRESHOLD_TO_ALPHA_OFFSET; + if (val < MLXSW_REG_SBXX_DYN_MAX_BUFF_MIN || + val > MLXSW_REG_SBXX_DYN_MAX_BUFF_MAX) { + NL_SET_ERR_MSG_MOD(extack, "Invalid dynamic threshold value"); + return -EINVAL; + } + *p_max_buff = val; + } else { + *p_max_buff = mlxsw_sp_bytes_cells(mlxsw_sp, threshold); + } + return 0; +} + +int mlxsw_sp_sb_port_pool_get(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 pool_index, + u32 *p_threshold) +{ + struct mlxsw_sp_port *mlxsw_sp_port = + mlxsw_core_port_driver_priv(mlxsw_core_port); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 local_port = mlxsw_sp_port->local_port; + struct mlxsw_sp_sb_pm *pm = mlxsw_sp_sb_pm_get(mlxsw_sp, local_port, + pool_index); + + *p_threshold = mlxsw_sp_sb_threshold_out(mlxsw_sp, pool_index, + pm->max_buff); + return 0; +} + +int mlxsw_sp_sb_port_pool_set(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 pool_index, + u32 threshold, struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port *mlxsw_sp_port = + mlxsw_core_port_driver_priv(mlxsw_core_port); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 local_port = mlxsw_sp_port->local_port; + u32 max_buff; + int err; + + if (local_port == MLXSW_PORT_CPU_PORT) { + NL_SET_ERR_MSG_MOD(extack, "Changing CPU port's threshold is forbidden"); + return -EINVAL; + } + + err = mlxsw_sp_sb_threshold_in(mlxsw_sp, pool_index, + threshold, &max_buff, extack); + if (err) + return err; + + return mlxsw_sp_sb_pm_write(mlxsw_sp, local_port, pool_index, + 0, max_buff); +} + +int mlxsw_sp_sb_tc_pool_bind_get(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 tc_index, + enum devlink_sb_pool_type pool_type, + u16 *p_pool_index, u32 *p_threshold) +{ + struct mlxsw_sp_port *mlxsw_sp_port = + mlxsw_core_port_driver_priv(mlxsw_core_port); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 local_port = mlxsw_sp_port->local_port; + u8 pg_buff = tc_index; + enum mlxsw_reg_sbxx_dir dir = (enum mlxsw_reg_sbxx_dir) pool_type; + struct mlxsw_sp_sb_cm *cm = mlxsw_sp_sb_cm_get(mlxsw_sp, local_port, + pg_buff, dir); + + *p_threshold = mlxsw_sp_sb_threshold_out(mlxsw_sp, cm->pool_index, + cm->max_buff); + *p_pool_index = cm->pool_index; + return 0; +} + +int mlxsw_sp_sb_tc_pool_bind_set(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 tc_index, + enum devlink_sb_pool_type pool_type, + u16 pool_index, u32 threshold, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port *mlxsw_sp_port = + mlxsw_core_port_driver_priv(mlxsw_core_port); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 local_port = mlxsw_sp_port->local_port; + const struct mlxsw_sp_sb_cm *cm; + u8 pg_buff = tc_index; + enum mlxsw_reg_sbxx_dir dir = (enum mlxsw_reg_sbxx_dir) pool_type; + u32 max_buff; + int err; + + if (local_port == MLXSW_PORT_CPU_PORT) { + NL_SET_ERR_MSG_MOD(extack, "Changing CPU port's binding is forbidden"); + return -EINVAL; + } + + if (dir != mlxsw_sp->sb_vals->pool_dess[pool_index].dir) { + NL_SET_ERR_MSG_MOD(extack, "Binding egress TC to ingress pool and vice versa is forbidden"); + return -EINVAL; + } + + if (dir == MLXSW_REG_SBXX_DIR_INGRESS) + cm = &mlxsw_sp->sb_vals->cms_ingress[tc_index]; + else + cm = &mlxsw_sp->sb_vals->cms_egress[tc_index]; + + if (cm->freeze_pool && cm->pool_index != pool_index) { + NL_SET_ERR_MSG_MOD(extack, "Binding this TC to a different pool is forbidden"); + return -EINVAL; + } + + if (cm->freeze_thresh && cm->max_buff != threshold) { + NL_SET_ERR_MSG_MOD(extack, "Changing this TC's threshold is forbidden"); + return -EINVAL; + } + + err = mlxsw_sp_sb_threshold_in(mlxsw_sp, pool_index, + threshold, &max_buff, extack); + if (err) + return err; + + return mlxsw_sp_sb_cm_write(mlxsw_sp, local_port, pg_buff, + 0, max_buff, false, pool_index); +} + +#define MASKED_COUNT_MAX \ + (MLXSW_REG_SBSR_REC_MAX_COUNT / \ + (MLXSW_SP_SB_ING_TC_COUNT + MLXSW_SP_SB_EG_TC_COUNT)) + +struct mlxsw_sp_sb_sr_occ_query_cb_ctx { + u8 masked_count; + u16 local_port_1; +}; + +static void mlxsw_sp_sb_sr_occ_query_cb(struct mlxsw_core *mlxsw_core, + char *sbsr_pl, size_t sbsr_pl_len, + unsigned long cb_priv) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + struct mlxsw_sp_sb_sr_occ_query_cb_ctx cb_ctx; + u8 masked_count; + u16 local_port; + int rec_index = 0; + struct mlxsw_sp_sb_cm *cm; + int i; + + memcpy(&cb_ctx, &cb_priv, sizeof(cb_ctx)); + + masked_count = 0; + for (local_port = cb_ctx.local_port_1; + local_port < mlxsw_core_max_ports(mlxsw_core); local_port++) { + if (!mlxsw_sp->ports[local_port]) + continue; + if (local_port == MLXSW_PORT_CPU_PORT) { + /* Ingress quotas are not supported for the CPU port */ + masked_count++; + continue; + } + for (i = 0; i < MLXSW_SP_SB_ING_TC_COUNT; i++) { + cm = mlxsw_sp_sb_cm_get(mlxsw_sp, local_port, i, + MLXSW_REG_SBXX_DIR_INGRESS); + mlxsw_reg_sbsr_rec_unpack(sbsr_pl, rec_index++, + &cm->occ.cur, &cm->occ.max); + } + if (++masked_count == cb_ctx.masked_count) + break; + } + masked_count = 0; + for (local_port = cb_ctx.local_port_1; + local_port < mlxsw_core_max_ports(mlxsw_core); local_port++) { + if (!mlxsw_sp->ports[local_port]) + continue; + for (i = 0; i < MLXSW_SP_SB_EG_TC_COUNT; i++) { + cm = mlxsw_sp_sb_cm_get(mlxsw_sp, local_port, i, + MLXSW_REG_SBXX_DIR_EGRESS); + mlxsw_reg_sbsr_rec_unpack(sbsr_pl, rec_index++, + &cm->occ.cur, &cm->occ.max); + } + if (++masked_count == cb_ctx.masked_count) + break; + } +} + +int mlxsw_sp_sb_occ_snapshot(struct mlxsw_core *mlxsw_core, + unsigned int sb_index) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + u16 local_port, local_port_1, last_local_port; + struct mlxsw_sp_sb_sr_occ_query_cb_ctx cb_ctx; + u8 masked_count, current_page = 0; + unsigned long cb_priv = 0; + LIST_HEAD(bulk_list); + char *sbsr_pl; + int i; + int err; + int err2; + + sbsr_pl = kmalloc(MLXSW_REG_SBSR_LEN, GFP_KERNEL); + if (!sbsr_pl) + return -ENOMEM; + + local_port = MLXSW_PORT_CPU_PORT; +next_batch: + local_port_1 = local_port; + masked_count = 0; + mlxsw_reg_sbsr_pack(sbsr_pl, false); + mlxsw_reg_sbsr_port_page_set(sbsr_pl, current_page); + last_local_port = current_page * MLXSW_REG_SBSR_NUM_PORTS_IN_PAGE + + MLXSW_REG_SBSR_NUM_PORTS_IN_PAGE - 1; + + for (i = 0; i < MLXSW_SP_SB_ING_TC_COUNT; i++) + mlxsw_reg_sbsr_pg_buff_mask_set(sbsr_pl, i, 1); + for (i = 0; i < MLXSW_SP_SB_EG_TC_COUNT; i++) + mlxsw_reg_sbsr_tclass_mask_set(sbsr_pl, i, 1); + for (; local_port < mlxsw_core_max_ports(mlxsw_core); local_port++) { + if (!mlxsw_sp->ports[local_port]) + continue; + if (local_port > last_local_port) { + current_page++; + goto do_query; + } + if (local_port != MLXSW_PORT_CPU_PORT) { + /* Ingress quotas are not supported for the CPU port */ + mlxsw_reg_sbsr_ingress_port_mask_set(sbsr_pl, + local_port, 1); + } + mlxsw_reg_sbsr_egress_port_mask_set(sbsr_pl, local_port, 1); + for (i = 0; i < mlxsw_sp->sb_vals->pool_count; i++) { + err = mlxsw_sp_sb_pm_occ_query(mlxsw_sp, local_port, i, + &bulk_list); + if (err) + goto out; + } + if (++masked_count == MASKED_COUNT_MAX) + goto do_query; + } + +do_query: + cb_ctx.masked_count = masked_count; + cb_ctx.local_port_1 = local_port_1; + memcpy(&cb_priv, &cb_ctx, sizeof(cb_ctx)); + err = mlxsw_reg_trans_query(mlxsw_core, MLXSW_REG(sbsr), sbsr_pl, + &bulk_list, mlxsw_sp_sb_sr_occ_query_cb, + cb_priv); + if (err) + goto out; + if (local_port < mlxsw_core_max_ports(mlxsw_core)) { + local_port++; + goto next_batch; + } + +out: + err2 = mlxsw_reg_trans_bulk_wait(&bulk_list); + if (!err) + err = err2; + kfree(sbsr_pl); + return err; +} + +int mlxsw_sp_sb_occ_max_clear(struct mlxsw_core *mlxsw_core, + unsigned int sb_index) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + u16 local_port, last_local_port; + LIST_HEAD(bulk_list); + unsigned int masked_count; + u8 current_page = 0; + char *sbsr_pl; + int i; + int err; + int err2; + + sbsr_pl = kmalloc(MLXSW_REG_SBSR_LEN, GFP_KERNEL); + if (!sbsr_pl) + return -ENOMEM; + + local_port = MLXSW_PORT_CPU_PORT; +next_batch: + masked_count = 0; + mlxsw_reg_sbsr_pack(sbsr_pl, true); + mlxsw_reg_sbsr_port_page_set(sbsr_pl, current_page); + last_local_port = current_page * MLXSW_REG_SBSR_NUM_PORTS_IN_PAGE + + MLXSW_REG_SBSR_NUM_PORTS_IN_PAGE - 1; + + for (i = 0; i < MLXSW_SP_SB_ING_TC_COUNT; i++) + mlxsw_reg_sbsr_pg_buff_mask_set(sbsr_pl, i, 1); + for (i = 0; i < MLXSW_SP_SB_EG_TC_COUNT; i++) + mlxsw_reg_sbsr_tclass_mask_set(sbsr_pl, i, 1); + for (; local_port < mlxsw_core_max_ports(mlxsw_core); local_port++) { + if (!mlxsw_sp->ports[local_port]) + continue; + if (local_port > last_local_port) { + current_page++; + goto do_query; + } + if (local_port != MLXSW_PORT_CPU_PORT) { + /* Ingress quotas are not supported for the CPU port */ + mlxsw_reg_sbsr_ingress_port_mask_set(sbsr_pl, + local_port, 1); + } + mlxsw_reg_sbsr_egress_port_mask_set(sbsr_pl, local_port, 1); + for (i = 0; i < mlxsw_sp->sb_vals->pool_count; i++) { + err = mlxsw_sp_sb_pm_occ_clear(mlxsw_sp, local_port, i, + &bulk_list); + if (err) + goto out; + } + if (++masked_count == MASKED_COUNT_MAX) + goto do_query; + } + +do_query: + err = mlxsw_reg_trans_query(mlxsw_core, MLXSW_REG(sbsr), sbsr_pl, + &bulk_list, NULL, 0); + if (err) + goto out; + if (local_port < mlxsw_core_max_ports(mlxsw_core)) { + local_port++; + goto next_batch; + } + +out: + err2 = mlxsw_reg_trans_bulk_wait(&bulk_list); + if (!err) + err = err2; + kfree(sbsr_pl); + return err; +} + +int mlxsw_sp_sb_occ_port_pool_get(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 pool_index, + u32 *p_cur, u32 *p_max) +{ + struct mlxsw_sp_port *mlxsw_sp_port = + mlxsw_core_port_driver_priv(mlxsw_core_port); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 local_port = mlxsw_sp_port->local_port; + struct mlxsw_sp_sb_pm *pm = mlxsw_sp_sb_pm_get(mlxsw_sp, local_port, + pool_index); + + *p_cur = mlxsw_sp_cells_bytes(mlxsw_sp, pm->occ.cur); + *p_max = mlxsw_sp_cells_bytes(mlxsw_sp, pm->occ.max); + return 0; +} + +int mlxsw_sp_sb_occ_tc_port_bind_get(struct mlxsw_core_port *mlxsw_core_port, + unsigned int sb_index, u16 tc_index, + enum devlink_sb_pool_type pool_type, + u32 *p_cur, u32 *p_max) +{ + struct mlxsw_sp_port *mlxsw_sp_port = + mlxsw_core_port_driver_priv(mlxsw_core_port); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 local_port = mlxsw_sp_port->local_port; + u8 pg_buff = tc_index; + enum mlxsw_reg_sbxx_dir dir = (enum mlxsw_reg_sbxx_dir) pool_type; + struct mlxsw_sp_sb_cm *cm = mlxsw_sp_sb_cm_get(mlxsw_sp, local_port, + pg_buff, dir); + + *p_cur = mlxsw_sp_cells_bytes(mlxsw_sp, cm->occ.cur); + *p_max = mlxsw_sp_cells_bytes(mlxsw_sp, cm->occ.max); + return 0; +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_cnt.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_cnt.c new file mode 100644 index 000000000..ee59c7915 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_cnt.c @@ -0,0 +1,306 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/bitops.h> +#include <linux/spinlock.h> + +#include "spectrum_cnt.h" + +struct mlxsw_sp_counter_sub_pool { + u64 size; + unsigned int base_index; + enum mlxsw_res_id entry_size_res_id; + const char *resource_name; /* devlink resource name */ + u64 resource_id; /* devlink resource id */ + unsigned int entry_size; + unsigned int bank_count; + atomic_t active_entries_count; +}; + +struct mlxsw_sp_counter_pool { + u64 pool_size; + unsigned long *usage; /* Usage bitmap */ + spinlock_t counter_pool_lock; /* Protects counter pool allocations */ + atomic_t active_entries_count; + unsigned int sub_pools_count; + struct mlxsw_sp_counter_sub_pool sub_pools[]; +}; + +static const struct mlxsw_sp_counter_sub_pool mlxsw_sp_counter_sub_pools[] = { + [MLXSW_SP_COUNTER_SUB_POOL_FLOW] = { + .entry_size_res_id = MLXSW_RES_ID_COUNTER_SIZE_PACKETS_BYTES, + .resource_name = MLXSW_SP_RESOURCE_NAME_COUNTERS_FLOW, + .resource_id = MLXSW_SP_RESOURCE_COUNTERS_FLOW, + .bank_count = 6, + }, + [MLXSW_SP_COUNTER_SUB_POOL_RIF] = { + .entry_size_res_id = MLXSW_RES_ID_COUNTER_SIZE_ROUTER_BASIC, + .resource_name = MLXSW_SP_RESOURCE_NAME_COUNTERS_RIF, + .resource_id = MLXSW_SP_RESOURCE_COUNTERS_RIF, + .bank_count = 2, + } +}; + +static u64 mlxsw_sp_counter_sub_pool_occ_get(void *priv) +{ + const struct mlxsw_sp_counter_sub_pool *sub_pool = priv; + + return atomic_read(&sub_pool->active_entries_count); +} + +static int mlxsw_sp_counter_sub_pools_init(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_counter_pool *pool = mlxsw_sp->counter_pool; + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + struct mlxsw_sp_counter_sub_pool *sub_pool; + unsigned int base_index = 0; + enum mlxsw_res_id res_id; + int err; + int i; + + for (i = 0; i < pool->sub_pools_count; i++) { + sub_pool = &pool->sub_pools[i]; + res_id = sub_pool->entry_size_res_id; + + if (!mlxsw_core_res_valid(mlxsw_sp->core, res_id)) + return -EIO; + sub_pool->entry_size = mlxsw_core_res_get(mlxsw_sp->core, + res_id); + err = devl_resource_size_get(devlink, + sub_pool->resource_id, + &sub_pool->size); + if (err) + goto err_resource_size_get; + + devl_resource_occ_get_register(devlink, + sub_pool->resource_id, + mlxsw_sp_counter_sub_pool_occ_get, + sub_pool); + + sub_pool->base_index = base_index; + base_index += sub_pool->size; + atomic_set(&sub_pool->active_entries_count, 0); + } + return 0; + +err_resource_size_get: + for (i--; i >= 0; i--) { + sub_pool = &pool->sub_pools[i]; + + devl_resource_occ_get_unregister(devlink, + sub_pool->resource_id); + } + return err; +} + +static void mlxsw_sp_counter_sub_pools_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_counter_pool *pool = mlxsw_sp->counter_pool; + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + struct mlxsw_sp_counter_sub_pool *sub_pool; + int i; + + for (i = 0; i < pool->sub_pools_count; i++) { + sub_pool = &pool->sub_pools[i]; + + WARN_ON(atomic_read(&sub_pool->active_entries_count)); + devl_resource_occ_get_unregister(devlink, + sub_pool->resource_id); + } +} + +static u64 mlxsw_sp_counter_pool_occ_get(void *priv) +{ + const struct mlxsw_sp_counter_pool *pool = priv; + + return atomic_read(&pool->active_entries_count); +} + +int mlxsw_sp_counter_pool_init(struct mlxsw_sp *mlxsw_sp) +{ + unsigned int sub_pools_count = ARRAY_SIZE(mlxsw_sp_counter_sub_pools); + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + struct mlxsw_sp_counter_pool *pool; + int err; + + pool = kzalloc(struct_size(pool, sub_pools, sub_pools_count), + GFP_KERNEL); + if (!pool) + return -ENOMEM; + mlxsw_sp->counter_pool = pool; + pool->sub_pools_count = sub_pools_count; + memcpy(pool->sub_pools, mlxsw_sp_counter_sub_pools, + flex_array_size(pool, sub_pools, pool->sub_pools_count)); + spin_lock_init(&pool->counter_pool_lock); + atomic_set(&pool->active_entries_count, 0); + + err = devl_resource_size_get(devlink, MLXSW_SP_RESOURCE_COUNTERS, + &pool->pool_size); + if (err) + goto err_pool_resource_size_get; + devl_resource_occ_get_register(devlink, MLXSW_SP_RESOURCE_COUNTERS, + mlxsw_sp_counter_pool_occ_get, pool); + + pool->usage = bitmap_zalloc(pool->pool_size, GFP_KERNEL); + if (!pool->usage) { + err = -ENOMEM; + goto err_usage_alloc; + } + + err = mlxsw_sp_counter_sub_pools_init(mlxsw_sp); + if (err) + goto err_sub_pools_init; + + return 0; + +err_sub_pools_init: + bitmap_free(pool->usage); +err_usage_alloc: + devl_resource_occ_get_unregister(devlink, + MLXSW_SP_RESOURCE_COUNTERS); +err_pool_resource_size_get: + kfree(pool); + return err; +} + +void mlxsw_sp_counter_pool_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_counter_pool *pool = mlxsw_sp->counter_pool; + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + + mlxsw_sp_counter_sub_pools_fini(mlxsw_sp); + WARN_ON(find_first_bit(pool->usage, pool->pool_size) != + pool->pool_size); + WARN_ON(atomic_read(&pool->active_entries_count)); + bitmap_free(pool->usage); + devl_resource_occ_get_unregister(devlink, + MLXSW_SP_RESOURCE_COUNTERS); + kfree(pool); +} + +int mlxsw_sp_counter_alloc(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_counter_sub_pool_id sub_pool_id, + unsigned int *p_counter_index) +{ + struct mlxsw_sp_counter_pool *pool = mlxsw_sp->counter_pool; + struct mlxsw_sp_counter_sub_pool *sub_pool; + unsigned int entry_index; + unsigned int stop_index; + int i, err; + + sub_pool = &pool->sub_pools[sub_pool_id]; + stop_index = sub_pool->base_index + sub_pool->size; + entry_index = sub_pool->base_index; + + spin_lock(&pool->counter_pool_lock); + entry_index = find_next_zero_bit(pool->usage, stop_index, entry_index); + if (entry_index == stop_index) { + err = -ENOBUFS; + goto err_alloc; + } + /* The sub-pools can contain non-integer number of entries + * so we must check for overflow + */ + if (entry_index + sub_pool->entry_size > stop_index) { + err = -ENOBUFS; + goto err_alloc; + } + for (i = 0; i < sub_pool->entry_size; i++) + __set_bit(entry_index + i, pool->usage); + spin_unlock(&pool->counter_pool_lock); + + *p_counter_index = entry_index; + atomic_add(sub_pool->entry_size, &sub_pool->active_entries_count); + atomic_add(sub_pool->entry_size, &pool->active_entries_count); + return 0; + +err_alloc: + spin_unlock(&pool->counter_pool_lock); + return err; +} + +void mlxsw_sp_counter_free(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_counter_sub_pool_id sub_pool_id, + unsigned int counter_index) +{ + struct mlxsw_sp_counter_pool *pool = mlxsw_sp->counter_pool; + struct mlxsw_sp_counter_sub_pool *sub_pool; + int i; + + if (WARN_ON(counter_index >= pool->pool_size)) + return; + sub_pool = &pool->sub_pools[sub_pool_id]; + spin_lock(&pool->counter_pool_lock); + for (i = 0; i < sub_pool->entry_size; i++) + __clear_bit(counter_index + i, pool->usage); + spin_unlock(&pool->counter_pool_lock); + atomic_sub(sub_pool->entry_size, &sub_pool->active_entries_count); + atomic_sub(sub_pool->entry_size, &pool->active_entries_count); +} + +int mlxsw_sp_counter_resources_register(struct mlxsw_core *mlxsw_core) +{ + static struct devlink_resource_size_params size_params; + struct devlink *devlink = priv_to_devlink(mlxsw_core); + const struct mlxsw_sp_counter_sub_pool *sub_pool; + unsigned int total_bank_config; + u64 sub_pool_size; + u64 base_index; + u64 pool_size; + u64 bank_size; + int err; + int i; + + if (!MLXSW_CORE_RES_VALID(mlxsw_core, COUNTER_POOL_SIZE) || + !MLXSW_CORE_RES_VALID(mlxsw_core, COUNTER_BANK_SIZE)) + return -EIO; + + pool_size = MLXSW_CORE_RES_GET(mlxsw_core, COUNTER_POOL_SIZE); + bank_size = MLXSW_CORE_RES_GET(mlxsw_core, COUNTER_BANK_SIZE); + + devlink_resource_size_params_init(&size_params, pool_size, + pool_size, bank_size, + DEVLINK_RESOURCE_UNIT_ENTRY); + err = devl_resource_register(devlink, + MLXSW_SP_RESOURCE_NAME_COUNTERS, + pool_size, + MLXSW_SP_RESOURCE_COUNTERS, + DEVLINK_RESOURCE_ID_PARENT_TOP, + &size_params); + if (err) + return err; + + /* Allocation is based on bank count which should be + * specified for each sub pool statically. + */ + total_bank_config = 0; + base_index = 0; + for (i = 0; i < ARRAY_SIZE(mlxsw_sp_counter_sub_pools); i++) { + sub_pool = &mlxsw_sp_counter_sub_pools[i]; + sub_pool_size = sub_pool->bank_count * bank_size; + /* The last bank can't be fully used */ + if (base_index + sub_pool_size > pool_size) + sub_pool_size = pool_size - base_index; + base_index += sub_pool_size; + + devlink_resource_size_params_init(&size_params, sub_pool_size, + sub_pool_size, bank_size, + DEVLINK_RESOURCE_UNIT_ENTRY); + err = devl_resource_register(devlink, + sub_pool->resource_name, + sub_pool_size, + sub_pool->resource_id, + MLXSW_SP_RESOURCE_COUNTERS, + &size_params); + if (err) + return err; + total_bank_config += sub_pool->bank_count; + } + + /* Check config is valid, no bank over subscription */ + if (WARN_ON(total_bank_config > div64_u64(pool_size, bank_size) + 1)) + return -EINVAL; + + return 0; +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_cnt.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_cnt.h new file mode 100644 index 000000000..15c8d4de8 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_cnt.h @@ -0,0 +1,25 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_SPECTRUM_CNT_H +#define _MLXSW_SPECTRUM_CNT_H + +#include "core.h" +#include "spectrum.h" + +enum mlxsw_sp_counter_sub_pool_id { + MLXSW_SP_COUNTER_SUB_POOL_RIF, + MLXSW_SP_COUNTER_SUB_POOL_FLOW, +}; + +int mlxsw_sp_counter_alloc(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_counter_sub_pool_id sub_pool_id, + unsigned int *p_counter_index); +void mlxsw_sp_counter_free(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_counter_sub_pool_id sub_pool_id, + unsigned int counter_index); +int mlxsw_sp_counter_pool_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_counter_pool_fini(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_counter_resources_register(struct mlxsw_core *mlxsw_core); + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_dcb.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_dcb.c new file mode 100644 index 000000000..aff6d4f35 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_dcb.c @@ -0,0 +1,740 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2016-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/netdevice.h> +#include <linux/string.h> +#include <linux/bitops.h> +#include <net/dcbnl.h> + +#include "spectrum.h" +#include "reg.h" + +static u8 mlxsw_sp_dcbnl_getdcbx(struct net_device __always_unused *dev) +{ + return DCB_CAP_DCBX_HOST | DCB_CAP_DCBX_VER_IEEE; +} + +static u8 mlxsw_sp_dcbnl_setdcbx(struct net_device __always_unused *dev, + u8 mode) +{ + return (mode != (DCB_CAP_DCBX_HOST | DCB_CAP_DCBX_VER_IEEE)) ? 1 : 0; +} + +static int mlxsw_sp_dcbnl_ieee_getets(struct net_device *dev, + struct ieee_ets *ets) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + + memcpy(ets, mlxsw_sp_port->dcb.ets, sizeof(*ets)); + + return 0; +} + +static int mlxsw_sp_port_ets_validate(struct mlxsw_sp_port *mlxsw_sp_port, + struct ieee_ets *ets) +{ + struct net_device *dev = mlxsw_sp_port->dev; + bool has_ets_tc = false; + int i, tx_bw_sum = 0; + + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + switch (ets->tc_tsa[i]) { + case IEEE_8021QAZ_TSA_STRICT: + break; + case IEEE_8021QAZ_TSA_ETS: + has_ets_tc = true; + tx_bw_sum += ets->tc_tx_bw[i]; + break; + default: + netdev_err(dev, "Only strict priority and ETS are supported\n"); + return -EINVAL; + } + + if (ets->prio_tc[i] >= IEEE_8021QAZ_MAX_TCS) { + netdev_err(dev, "Invalid TC\n"); + return -EINVAL; + } + } + + if (has_ets_tc && tx_bw_sum != 100) { + netdev_err(dev, "Total ETS bandwidth should equal 100\n"); + return -EINVAL; + } + + return 0; +} + +static int mlxsw_sp_port_headroom_ets_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct ieee_ets *ets) +{ + struct net_device *dev = mlxsw_sp_port->dev; + struct mlxsw_sp_hdroom hdroom; + int prio; + int err; + + hdroom = *mlxsw_sp_port->hdroom; + for (prio = 0; prio < IEEE_8021QAZ_MAX_TCS; prio++) + hdroom.prios.prio[prio].ets_buf_idx = ets->prio_tc[prio]; + mlxsw_sp_hdroom_prios_reset_buf_idx(&hdroom); + mlxsw_sp_hdroom_bufs_reset_lossiness(&hdroom); + mlxsw_sp_hdroom_bufs_reset_sizes(mlxsw_sp_port, &hdroom); + + err = mlxsw_sp_hdroom_configure(mlxsw_sp_port, &hdroom); + if (err) { + netdev_err(dev, "Failed to configure port's headroom\n"); + return err; + } + + return 0; +} + +static int __mlxsw_sp_dcbnl_ieee_setets(struct mlxsw_sp_port *mlxsw_sp_port, + struct ieee_ets *ets) +{ + struct ieee_ets *my_ets = mlxsw_sp_port->dcb.ets; + struct net_device *dev = mlxsw_sp_port->dev; + int i, err; + + /* Egress configuration. */ + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + bool dwrr = ets->tc_tsa[i] == IEEE_8021QAZ_TSA_ETS; + u8 weight = ets->tc_tx_bw[i]; + + err = mlxsw_sp_port_ets_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, i, + 0, dwrr, weight); + if (err) { + netdev_err(dev, "Failed to link subgroup ETS element %d to group\n", + i); + goto err_port_ets_set; + } + } + + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + err = mlxsw_sp_port_prio_tc_set(mlxsw_sp_port, i, + ets->prio_tc[i]); + if (err) { + netdev_err(dev, "Failed to map prio %d to TC %d\n", i, + ets->prio_tc[i]); + goto err_port_prio_tc_set; + } + } + + /* Ingress configuration. */ + err = mlxsw_sp_port_headroom_ets_set(mlxsw_sp_port, ets); + if (err) + goto err_port_headroom_set; + + return 0; + +err_port_headroom_set: + i = IEEE_8021QAZ_MAX_TCS; +err_port_prio_tc_set: + for (i--; i >= 0; i--) + mlxsw_sp_port_prio_tc_set(mlxsw_sp_port, i, my_ets->prio_tc[i]); + i = IEEE_8021QAZ_MAX_TCS; +err_port_ets_set: + for (i--; i >= 0; i--) { + bool dwrr = my_ets->tc_tsa[i] == IEEE_8021QAZ_TSA_ETS; + u8 weight = my_ets->tc_tx_bw[i]; + + err = mlxsw_sp_port_ets_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, i, + 0, dwrr, weight); + } + return err; +} + +static int mlxsw_sp_dcbnl_ieee_setets(struct net_device *dev, + struct ieee_ets *ets) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + int err; + + err = mlxsw_sp_port_ets_validate(mlxsw_sp_port, ets); + if (err) + return err; + + err = __mlxsw_sp_dcbnl_ieee_setets(mlxsw_sp_port, ets); + if (err) + return err; + + memcpy(mlxsw_sp_port->dcb.ets, ets, sizeof(*ets)); + mlxsw_sp_port->dcb.ets->ets_cap = IEEE_8021QAZ_MAX_TCS; + + return 0; +} + +static int mlxsw_sp_dcbnl_app_validate(struct net_device *dev, + struct dcb_app *app) +{ + if (app->priority >= IEEE_8021QAZ_MAX_TCS) { + netdev_err(dev, "APP entry with priority value %u is invalid\n", + app->priority); + return -EINVAL; + } + + switch (app->selector) { + case IEEE_8021QAZ_APP_SEL_DSCP: + if (app->protocol >= 64) { + netdev_err(dev, "DSCP APP entry with protocol value %u is invalid\n", + app->protocol); + return -EINVAL; + } + break; + + case IEEE_8021QAZ_APP_SEL_ETHERTYPE: + if (app->protocol) { + netdev_err(dev, "EtherType APP entries with protocol value != 0 not supported\n"); + return -EINVAL; + } + break; + + default: + netdev_err(dev, "APP entries with selector %u not supported\n", + app->selector); + return -EINVAL; + } + + return 0; +} + +static u8 +mlxsw_sp_port_dcb_app_default_prio(struct mlxsw_sp_port *mlxsw_sp_port) +{ + u8 prio_mask; + + prio_mask = dcb_ieee_getapp_default_prio_mask(mlxsw_sp_port->dev); + if (prio_mask) + /* Take the highest configured priority. */ + return fls(prio_mask) - 1; + + return 0; +} + +static void +mlxsw_sp_port_dcb_app_dscp_prio_map(struct mlxsw_sp_port *mlxsw_sp_port, + u8 default_prio, + struct dcb_ieee_app_dscp_map *map) +{ + int i; + + dcb_ieee_getapp_dscp_prio_mask_map(mlxsw_sp_port->dev, map); + for (i = 0; i < ARRAY_SIZE(map->map); ++i) { + if (map->map[i]) + map->map[i] = fls(map->map[i]) - 1; + else + map->map[i] = default_prio; + } +} + +static bool +mlxsw_sp_port_dcb_app_prio_dscp_map(struct mlxsw_sp_port *mlxsw_sp_port, + struct dcb_ieee_app_prio_map *map) +{ + bool have_dscp = false; + int i; + + dcb_ieee_getapp_prio_dscp_mask_map(mlxsw_sp_port->dev, map); + for (i = 0; i < ARRAY_SIZE(map->map); ++i) { + if (map->map[i]) { + map->map[i] = fls64(map->map[i]) - 1; + have_dscp = true; + } + } + + return have_dscp; +} + +static int +mlxsw_sp_port_dcb_app_update_qpts(struct mlxsw_sp_port *mlxsw_sp_port, + enum mlxsw_reg_qpts_trust_state ts) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char qpts_pl[MLXSW_REG_QPTS_LEN]; + + mlxsw_reg_qpts_pack(qpts_pl, mlxsw_sp_port->local_port, ts); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qpts), qpts_pl); +} + +static int +mlxsw_sp_port_dcb_app_update_qrwe(struct mlxsw_sp_port *mlxsw_sp_port, + bool rewrite_dscp) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char qrwe_pl[MLXSW_REG_QRWE_LEN]; + + mlxsw_reg_qrwe_pack(qrwe_pl, mlxsw_sp_port->local_port, + false, rewrite_dscp); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qrwe), qrwe_pl); +} + +static int +mlxsw_sp_port_dcb_toggle_trust(struct mlxsw_sp_port *mlxsw_sp_port, + enum mlxsw_reg_qpts_trust_state ts) +{ + bool rewrite_dscp = ts == MLXSW_REG_QPTS_TRUST_STATE_DSCP; + int err; + + if (mlxsw_sp_port->dcb.trust_state == ts) + return 0; + + err = mlxsw_sp_port_dcb_app_update_qpts(mlxsw_sp_port, ts); + if (err) + return err; + + err = mlxsw_sp_port_dcb_app_update_qrwe(mlxsw_sp_port, rewrite_dscp); + if (err) + goto err_update_qrwe; + + mlxsw_sp_port->dcb.trust_state = ts; + return 0; + +err_update_qrwe: + mlxsw_sp_port_dcb_app_update_qpts(mlxsw_sp_port, + mlxsw_sp_port->dcb.trust_state); + return err; +} + +static int +mlxsw_sp_port_dcb_app_update_qpdp(struct mlxsw_sp_port *mlxsw_sp_port, + u8 default_prio) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char qpdp_pl[MLXSW_REG_QPDP_LEN]; + + mlxsw_reg_qpdp_pack(qpdp_pl, mlxsw_sp_port->local_port, default_prio); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qpdp), qpdp_pl); +} + +static int +mlxsw_sp_port_dcb_app_update_qpdpm(struct mlxsw_sp_port *mlxsw_sp_port, + struct dcb_ieee_app_dscp_map *map) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char qpdpm_pl[MLXSW_REG_QPDPM_LEN]; + short int i; + + mlxsw_reg_qpdpm_pack(qpdpm_pl, mlxsw_sp_port->local_port); + for (i = 0; i < ARRAY_SIZE(map->map); ++i) + mlxsw_reg_qpdpm_dscp_pack(qpdpm_pl, i, map->map[i]); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qpdpm), qpdpm_pl); +} + +static int +mlxsw_sp_port_dcb_app_update_qpdsm(struct mlxsw_sp_port *mlxsw_sp_port, + struct dcb_ieee_app_prio_map *map) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char qpdsm_pl[MLXSW_REG_QPDSM_LEN]; + short int i; + + mlxsw_reg_qpdsm_pack(qpdsm_pl, mlxsw_sp_port->local_port); + for (i = 0; i < ARRAY_SIZE(map->map); ++i) + mlxsw_reg_qpdsm_prio_pack(qpdsm_pl, i, map->map[i]); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qpdsm), qpdsm_pl); +} + +static int mlxsw_sp_port_dcb_app_update(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct dcb_ieee_app_prio_map prio_map; + struct dcb_ieee_app_dscp_map dscp_map; + u8 default_prio; + bool have_dscp; + int err; + + default_prio = mlxsw_sp_port_dcb_app_default_prio(mlxsw_sp_port); + err = mlxsw_sp_port_dcb_app_update_qpdp(mlxsw_sp_port, default_prio); + if (err) { + netdev_err(mlxsw_sp_port->dev, "Couldn't configure port default priority\n"); + return err; + } + + have_dscp = mlxsw_sp_port_dcb_app_prio_dscp_map(mlxsw_sp_port, + &prio_map); + + mlxsw_sp_port_dcb_app_dscp_prio_map(mlxsw_sp_port, default_prio, + &dscp_map); + err = mlxsw_sp_port_dcb_app_update_qpdpm(mlxsw_sp_port, + &dscp_map); + if (err) { + netdev_err(mlxsw_sp_port->dev, "Couldn't configure priority map\n"); + return err; + } + + err = mlxsw_sp_port_dcb_app_update_qpdsm(mlxsw_sp_port, + &prio_map); + if (err) { + netdev_err(mlxsw_sp_port->dev, "Couldn't configure DSCP rewrite map\n"); + return err; + } + + if (!have_dscp) { + err = mlxsw_sp_port_dcb_toggle_trust(mlxsw_sp_port, + MLXSW_REG_QPTS_TRUST_STATE_PCP); + if (err) + netdev_err(mlxsw_sp_port->dev, "Couldn't switch to trust L2\n"); + return err; + } + + err = mlxsw_sp_port_dcb_toggle_trust(mlxsw_sp_port, + MLXSW_REG_QPTS_TRUST_STATE_DSCP); + if (err) { + /* A failure to set trust DSCP means that the QPDPM and QPDSM + * maps installed above are not in effect. And since we are here + * attempting to set trust DSCP, we couldn't have attempted to + * switch trust to PCP. Thus no cleanup is necessary. + */ + netdev_err(mlxsw_sp_port->dev, "Couldn't switch to trust L3\n"); + return err; + } + + return 0; +} + +static int mlxsw_sp_dcbnl_ieee_setapp(struct net_device *dev, + struct dcb_app *app) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + int err; + + err = mlxsw_sp_dcbnl_app_validate(dev, app); + if (err) + return err; + + err = dcb_ieee_setapp(dev, app); + if (err) + return err; + + err = mlxsw_sp_port_dcb_app_update(mlxsw_sp_port); + if (err) + goto err_update; + + return 0; + +err_update: + dcb_ieee_delapp(dev, app); + return err; +} + +static int mlxsw_sp_dcbnl_ieee_delapp(struct net_device *dev, + struct dcb_app *app) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + int err; + + err = dcb_ieee_delapp(dev, app); + if (err) + return err; + + err = mlxsw_sp_port_dcb_app_update(mlxsw_sp_port); + if (err) + netdev_err(dev, "Failed to update DCB APP configuration\n"); + return 0; +} + +static int mlxsw_sp_dcbnl_ieee_getmaxrate(struct net_device *dev, + struct ieee_maxrate *maxrate) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + + memcpy(maxrate, mlxsw_sp_port->dcb.maxrate, sizeof(*maxrate)); + + return 0; +} + +static int mlxsw_sp_dcbnl_ieee_setmaxrate(struct net_device *dev, + struct ieee_maxrate *maxrate) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct ieee_maxrate *my_maxrate = mlxsw_sp_port->dcb.maxrate; + int err, i; + + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + err = mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, + i, 0, + maxrate->tc_maxrate[i], 0); + if (err) { + netdev_err(dev, "Failed to set maxrate for TC %d\n", i); + goto err_port_ets_maxrate_set; + } + } + + memcpy(mlxsw_sp_port->dcb.maxrate, maxrate, sizeof(*maxrate)); + + return 0; + +err_port_ets_maxrate_set: + for (i--; i >= 0; i--) + mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, + i, 0, + my_maxrate->tc_maxrate[i], 0); + return err; +} + +static int mlxsw_sp_port_pfc_cnt_get(struct mlxsw_sp_port *mlxsw_sp_port, + u8 prio) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct ieee_pfc *my_pfc = mlxsw_sp_port->dcb.pfc; + char ppcnt_pl[MLXSW_REG_PPCNT_LEN]; + int err; + + mlxsw_reg_ppcnt_pack(ppcnt_pl, mlxsw_sp_port->local_port, + MLXSW_REG_PPCNT_PRIO_CNT, prio); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ppcnt), ppcnt_pl); + if (err) + return err; + + my_pfc->requests[prio] = mlxsw_reg_ppcnt_tx_pause_get(ppcnt_pl); + my_pfc->indications[prio] = mlxsw_reg_ppcnt_rx_pause_get(ppcnt_pl); + + return 0; +} + +static int mlxsw_sp_dcbnl_ieee_getpfc(struct net_device *dev, + struct ieee_pfc *pfc) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + int err, i; + + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + err = mlxsw_sp_port_pfc_cnt_get(mlxsw_sp_port, i); + if (err) { + netdev_err(dev, "Failed to get PFC count for priority %d\n", + i); + return err; + } + } + + memcpy(pfc, mlxsw_sp_port->dcb.pfc, sizeof(*pfc)); + + return 0; +} + +static int mlxsw_sp_port_pfc_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct ieee_pfc *pfc) +{ + char pfcc_pl[MLXSW_REG_PFCC_LEN]; + + mlxsw_reg_pfcc_pack(pfcc_pl, mlxsw_sp_port->local_port); + mlxsw_reg_pfcc_pprx_set(pfcc_pl, mlxsw_sp_port->link.rx_pause); + mlxsw_reg_pfcc_pptx_set(pfcc_pl, mlxsw_sp_port->link.tx_pause); + mlxsw_reg_pfcc_prio_pack(pfcc_pl, pfc->pfc_en); + + return mlxsw_reg_write(mlxsw_sp_port->mlxsw_sp->core, MLXSW_REG(pfcc), + pfcc_pl); +} + +static int mlxsw_sp_dcbnl_ieee_setpfc(struct net_device *dev, + struct ieee_pfc *pfc) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + bool pause_en = mlxsw_sp_port_is_pause_en(mlxsw_sp_port); + struct mlxsw_sp_hdroom orig_hdroom; + struct mlxsw_sp_hdroom hdroom; + int prio; + int err; + + if (pause_en && pfc->pfc_en) { + netdev_err(dev, "PAUSE frames already enabled on port\n"); + return -EINVAL; + } + + orig_hdroom = *mlxsw_sp_port->hdroom; + + hdroom = orig_hdroom; + if (pfc->pfc_en) + hdroom.delay_bytes = DIV_ROUND_UP(pfc->delay, BITS_PER_BYTE); + else + hdroom.delay_bytes = 0; + + for (prio = 0; prio < IEEE_8021QAZ_MAX_TCS; prio++) + hdroom.prios.prio[prio].lossy = !(pfc->pfc_en & BIT(prio)); + + mlxsw_sp_hdroom_bufs_reset_lossiness(&hdroom); + mlxsw_sp_hdroom_bufs_reset_sizes(mlxsw_sp_port, &hdroom); + + err = mlxsw_sp_hdroom_configure(mlxsw_sp_port, &hdroom); + if (err) { + netdev_err(dev, "Failed to configure port's headroom for PFC\n"); + return err; + } + + err = mlxsw_sp_port_pfc_set(mlxsw_sp_port, pfc); + if (err) { + netdev_err(dev, "Failed to configure PFC\n"); + goto err_port_pfc_set; + } + + memcpy(mlxsw_sp_port->dcb.pfc, pfc, sizeof(*pfc)); + mlxsw_sp_port->dcb.pfc->pfc_cap = IEEE_8021QAZ_MAX_TCS; + + return 0; + +err_port_pfc_set: + mlxsw_sp_hdroom_configure(mlxsw_sp_port, &orig_hdroom); + return err; +} + +static int mlxsw_sp_dcbnl_getbuffer(struct net_device *dev, struct dcbnl_buffer *buf) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp_hdroom *hdroom = mlxsw_sp_port->hdroom; + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + int prio; + int i; + + buf->total_size = 0; + + BUILD_BUG_ON(DCBX_MAX_BUFFERS > MLXSW_SP_PB_COUNT); + for (i = 0; i < MLXSW_SP_PB_COUNT; i++) { + u32 bytes = mlxsw_sp_cells_bytes(mlxsw_sp, hdroom->bufs.buf[i].size_cells); + + if (i < DCBX_MAX_BUFFERS) + buf->buffer_size[i] = bytes; + buf->total_size += bytes; + } + + buf->total_size += mlxsw_sp_cells_bytes(mlxsw_sp, hdroom->int_buf.size_cells); + + for (prio = 0; prio < IEEE_8021Q_MAX_PRIORITIES; prio++) + buf->prio2buffer[prio] = hdroom->prios.prio[prio].buf_idx; + + return 0; +} + +static int mlxsw_sp_dcbnl_setbuffer(struct net_device *dev, struct dcbnl_buffer *buf) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_hdroom hdroom; + int prio; + int i; + + hdroom = *mlxsw_sp_port->hdroom; + + if (hdroom.mode != MLXSW_SP_HDROOM_MODE_TC) { + netdev_err(dev, "The use of dcbnl_setbuffer is only allowed if egress is configured using TC\n"); + return -EINVAL; + } + + for (prio = 0; prio < IEEE_8021Q_MAX_PRIORITIES; prio++) + hdroom.prios.prio[prio].set_buf_idx = buf->prio2buffer[prio]; + + BUILD_BUG_ON(DCBX_MAX_BUFFERS > MLXSW_SP_PB_COUNT); + for (i = 0; i < DCBX_MAX_BUFFERS; i++) + hdroom.bufs.buf[i].set_size_cells = mlxsw_sp_bytes_cells(mlxsw_sp, + buf->buffer_size[i]); + + mlxsw_sp_hdroom_prios_reset_buf_idx(&hdroom); + mlxsw_sp_hdroom_bufs_reset_lossiness(&hdroom); + mlxsw_sp_hdroom_bufs_reset_sizes(mlxsw_sp_port, &hdroom); + return mlxsw_sp_hdroom_configure(mlxsw_sp_port, &hdroom); +} + +static const struct dcbnl_rtnl_ops mlxsw_sp_dcbnl_ops = { + .ieee_getets = mlxsw_sp_dcbnl_ieee_getets, + .ieee_setets = mlxsw_sp_dcbnl_ieee_setets, + .ieee_getmaxrate = mlxsw_sp_dcbnl_ieee_getmaxrate, + .ieee_setmaxrate = mlxsw_sp_dcbnl_ieee_setmaxrate, + .ieee_getpfc = mlxsw_sp_dcbnl_ieee_getpfc, + .ieee_setpfc = mlxsw_sp_dcbnl_ieee_setpfc, + .ieee_setapp = mlxsw_sp_dcbnl_ieee_setapp, + .ieee_delapp = mlxsw_sp_dcbnl_ieee_delapp, + + .getdcbx = mlxsw_sp_dcbnl_getdcbx, + .setdcbx = mlxsw_sp_dcbnl_setdcbx, + + .dcbnl_getbuffer = mlxsw_sp_dcbnl_getbuffer, + .dcbnl_setbuffer = mlxsw_sp_dcbnl_setbuffer, +}; + +static int mlxsw_sp_port_ets_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + mlxsw_sp_port->dcb.ets = kzalloc(sizeof(*mlxsw_sp_port->dcb.ets), + GFP_KERNEL); + if (!mlxsw_sp_port->dcb.ets) + return -ENOMEM; + + mlxsw_sp_port->dcb.ets->ets_cap = IEEE_8021QAZ_MAX_TCS; + + return 0; +} + +static void mlxsw_sp_port_ets_fini(struct mlxsw_sp_port *mlxsw_sp_port) +{ + kfree(mlxsw_sp_port->dcb.ets); +} + +static int mlxsw_sp_port_maxrate_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + int i; + + mlxsw_sp_port->dcb.maxrate = kmalloc(sizeof(*mlxsw_sp_port->dcb.maxrate), + GFP_KERNEL); + if (!mlxsw_sp_port->dcb.maxrate) + return -ENOMEM; + + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) + mlxsw_sp_port->dcb.maxrate->tc_maxrate[i] = MLXSW_REG_QEEC_MAS_DIS; + + return 0; +} + +static void mlxsw_sp_port_maxrate_fini(struct mlxsw_sp_port *mlxsw_sp_port) +{ + kfree(mlxsw_sp_port->dcb.maxrate); +} + +static int mlxsw_sp_port_pfc_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + mlxsw_sp_port->dcb.pfc = kzalloc(sizeof(*mlxsw_sp_port->dcb.pfc), + GFP_KERNEL); + if (!mlxsw_sp_port->dcb.pfc) + return -ENOMEM; + + mlxsw_sp_port->dcb.pfc->pfc_cap = IEEE_8021QAZ_MAX_TCS; + + return 0; +} + +static void mlxsw_sp_port_pfc_fini(struct mlxsw_sp_port *mlxsw_sp_port) +{ + kfree(mlxsw_sp_port->dcb.pfc); +} + +int mlxsw_sp_port_dcb_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + int err; + + err = mlxsw_sp_port_ets_init(mlxsw_sp_port); + if (err) + return err; + err = mlxsw_sp_port_maxrate_init(mlxsw_sp_port); + if (err) + goto err_port_maxrate_init; + err = mlxsw_sp_port_pfc_init(mlxsw_sp_port); + if (err) + goto err_port_pfc_init; + + mlxsw_sp_port->dcb.trust_state = MLXSW_REG_QPTS_TRUST_STATE_PCP; + mlxsw_sp_port->dev->dcbnl_ops = &mlxsw_sp_dcbnl_ops; + + return 0; + +err_port_pfc_init: + mlxsw_sp_port_maxrate_fini(mlxsw_sp_port); +err_port_maxrate_init: + mlxsw_sp_port_ets_fini(mlxsw_sp_port); + return err; +} + +void mlxsw_sp_port_dcb_fini(struct mlxsw_sp_port *mlxsw_sp_port) +{ + mlxsw_sp_port_pfc_fini(mlxsw_sp_port); + mlxsw_sp_port_maxrate_fini(mlxsw_sp_port); + mlxsw_sp_port_ets_fini(mlxsw_sp_port); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.c new file mode 100644 index 000000000..5416093c0 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.c @@ -0,0 +1,1308 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/mutex.h> +#include <net/devlink.h> + +#include "spectrum.h" +#include "spectrum_dpipe.h" +#include "spectrum_router.h" + +enum mlxsw_sp_field_metadata_id { + MLXSW_SP_DPIPE_FIELD_METADATA_ERIF_PORT, + MLXSW_SP_DPIPE_FIELD_METADATA_L3_FORWARD, + MLXSW_SP_DPIPE_FIELD_METADATA_L3_DROP, + MLXSW_SP_DPIPE_FIELD_METADATA_ADJ_INDEX, + MLXSW_SP_DPIPE_FIELD_METADATA_ADJ_SIZE, + MLXSW_SP_DPIPE_FIELD_METADATA_ADJ_HASH_INDEX, +}; + +static struct devlink_dpipe_field mlxsw_sp_dpipe_fields_metadata[] = { + { + .name = "erif_port", + .id = MLXSW_SP_DPIPE_FIELD_METADATA_ERIF_PORT, + .bitwidth = 32, + .mapping_type = DEVLINK_DPIPE_FIELD_MAPPING_TYPE_IFINDEX, + }, + { + .name = "l3_forward", + .id = MLXSW_SP_DPIPE_FIELD_METADATA_L3_FORWARD, + .bitwidth = 1, + }, + { + .name = "l3_drop", + .id = MLXSW_SP_DPIPE_FIELD_METADATA_L3_DROP, + .bitwidth = 1, + }, + { + .name = "adj_index", + .id = MLXSW_SP_DPIPE_FIELD_METADATA_ADJ_INDEX, + .bitwidth = 32, + }, + { + .name = "adj_size", + .id = MLXSW_SP_DPIPE_FIELD_METADATA_ADJ_SIZE, + .bitwidth = 32, + }, + { + .name = "adj_hash_index", + .id = MLXSW_SP_DPIPE_FIELD_METADATA_ADJ_HASH_INDEX, + .bitwidth = 32, + }, +}; + +enum mlxsw_sp_dpipe_header_id { + MLXSW_SP_DPIPE_HEADER_METADATA, +}; + +static struct devlink_dpipe_header mlxsw_sp_dpipe_header_metadata = { + .name = "mlxsw_meta", + .id = MLXSW_SP_DPIPE_HEADER_METADATA, + .fields = mlxsw_sp_dpipe_fields_metadata, + .fields_count = ARRAY_SIZE(mlxsw_sp_dpipe_fields_metadata), +}; + +static struct devlink_dpipe_header *mlxsw_dpipe_headers[] = { + &mlxsw_sp_dpipe_header_metadata, + &devlink_dpipe_header_ethernet, + &devlink_dpipe_header_ipv4, + &devlink_dpipe_header_ipv6, +}; + +static struct devlink_dpipe_headers mlxsw_sp_dpipe_headers = { + .headers = mlxsw_dpipe_headers, + .headers_count = ARRAY_SIZE(mlxsw_dpipe_headers), +}; + +static int mlxsw_sp_dpipe_table_erif_actions_dump(void *priv, + struct sk_buff *skb) +{ + struct devlink_dpipe_action action = {0}; + int err; + + action.type = DEVLINK_DPIPE_ACTION_TYPE_FIELD_MODIFY; + action.header = &mlxsw_sp_dpipe_header_metadata; + action.field_id = MLXSW_SP_DPIPE_FIELD_METADATA_L3_FORWARD; + + err = devlink_dpipe_action_put(skb, &action); + if (err) + return err; + + action.type = DEVLINK_DPIPE_ACTION_TYPE_FIELD_MODIFY; + action.header = &mlxsw_sp_dpipe_header_metadata; + action.field_id = MLXSW_SP_DPIPE_FIELD_METADATA_L3_DROP; + + return devlink_dpipe_action_put(skb, &action); +} + +static int mlxsw_sp_dpipe_table_erif_matches_dump(void *priv, + struct sk_buff *skb) +{ + struct devlink_dpipe_match match = {0}; + + match.type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + match.header = &mlxsw_sp_dpipe_header_metadata; + match.field_id = MLXSW_SP_DPIPE_FIELD_METADATA_ERIF_PORT; + + return devlink_dpipe_match_put(skb, &match); +} + +static void +mlxsw_sp_erif_match_action_prepare(struct devlink_dpipe_match *match, + struct devlink_dpipe_action *action) +{ + action->type = DEVLINK_DPIPE_ACTION_TYPE_FIELD_MODIFY; + action->header = &mlxsw_sp_dpipe_header_metadata; + action->field_id = MLXSW_SP_DPIPE_FIELD_METADATA_L3_FORWARD; + + match->type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + match->header = &mlxsw_sp_dpipe_header_metadata; + match->field_id = MLXSW_SP_DPIPE_FIELD_METADATA_ERIF_PORT; +} + +static int mlxsw_sp_erif_entry_prepare(struct devlink_dpipe_entry *entry, + struct devlink_dpipe_value *match_value, + struct devlink_dpipe_match *match, + struct devlink_dpipe_value *action_value, + struct devlink_dpipe_action *action) +{ + entry->match_values = match_value; + entry->match_values_count = 1; + + entry->action_values = action_value; + entry->action_values_count = 1; + + match_value->match = match; + match_value->value_size = sizeof(u32); + match_value->value = kmalloc(match_value->value_size, GFP_KERNEL); + if (!match_value->value) + return -ENOMEM; + + action_value->action = action; + action_value->value_size = sizeof(u32); + action_value->value = kmalloc(action_value->value_size, GFP_KERNEL); + if (!action_value->value) + goto err_action_alloc; + return 0; + +err_action_alloc: + kfree(match_value->value); + return -ENOMEM; +} + +static int mlxsw_sp_erif_entry_get(struct mlxsw_sp *mlxsw_sp, + struct devlink_dpipe_entry *entry, + struct mlxsw_sp_rif *rif, + bool counters_enabled) +{ + u32 *action_value; + u32 *rif_value; + u64 cnt; + int err; + + /* Set Match RIF index */ + rif_value = entry->match_values->value; + *rif_value = mlxsw_sp_rif_index(rif); + entry->match_values->mapping_value = mlxsw_sp_rif_dev_ifindex(rif); + entry->match_values->mapping_valid = true; + + /* Set Action Forwarding */ + action_value = entry->action_values->value; + *action_value = 1; + + entry->counter_valid = false; + entry->counter = 0; + entry->index = mlxsw_sp_rif_index(rif); + + if (!counters_enabled) + return 0; + + err = mlxsw_sp_rif_counter_value_get(mlxsw_sp, rif, + MLXSW_SP_RIF_COUNTER_EGRESS, + &cnt); + if (!err) { + entry->counter = cnt; + entry->counter_valid = true; + } + return 0; +} + +static int +mlxsw_sp_dpipe_table_erif_entries_dump(void *priv, bool counters_enabled, + struct devlink_dpipe_dump_ctx *dump_ctx) +{ + struct devlink_dpipe_value match_value, action_value; + struct devlink_dpipe_action action = {0}; + struct devlink_dpipe_match match = {0}; + struct devlink_dpipe_entry entry = {0}; + struct mlxsw_sp *mlxsw_sp = priv; + unsigned int rif_count; + int i, j; + int err; + + memset(&match_value, 0, sizeof(match_value)); + memset(&action_value, 0, sizeof(action_value)); + + mlxsw_sp_erif_match_action_prepare(&match, &action); + err = mlxsw_sp_erif_entry_prepare(&entry, &match_value, &match, + &action_value, &action); + if (err) + return err; + + rif_count = MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); + mutex_lock(&mlxsw_sp->router->lock); + i = 0; +start_again: + err = devlink_dpipe_entry_ctx_prepare(dump_ctx); + if (err) + goto err_ctx_prepare; + j = 0; + for (; i < rif_count; i++) { + struct mlxsw_sp_rif *rif = mlxsw_sp_rif_by_index(mlxsw_sp, i); + + if (!rif || !mlxsw_sp_rif_dev(rif)) + continue; + err = mlxsw_sp_erif_entry_get(mlxsw_sp, &entry, rif, + counters_enabled); + if (err) + goto err_entry_get; + err = devlink_dpipe_entry_ctx_append(dump_ctx, &entry); + if (err) { + if (err == -EMSGSIZE) { + if (!j) + goto err_entry_append; + break; + } + goto err_entry_append; + } + j++; + } + + devlink_dpipe_entry_ctx_close(dump_ctx); + if (i != rif_count) + goto start_again; + mutex_unlock(&mlxsw_sp->router->lock); + + devlink_dpipe_entry_clear(&entry); + return 0; +err_entry_append: +err_entry_get: +err_ctx_prepare: + mutex_unlock(&mlxsw_sp->router->lock); + devlink_dpipe_entry_clear(&entry); + return err; +} + +static int mlxsw_sp_dpipe_table_erif_counters_update(void *priv, bool enable) +{ + struct mlxsw_sp *mlxsw_sp = priv; + int i; + + mutex_lock(&mlxsw_sp->router->lock); + for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); i++) { + struct mlxsw_sp_rif *rif = mlxsw_sp_rif_by_index(mlxsw_sp, i); + + if (!rif) + continue; + if (enable) + mlxsw_sp_rif_counter_alloc(rif, + MLXSW_SP_RIF_COUNTER_EGRESS); + else + mlxsw_sp_rif_counter_free(rif, + MLXSW_SP_RIF_COUNTER_EGRESS); + } + mutex_unlock(&mlxsw_sp->router->lock); + return 0; +} + +static u64 mlxsw_sp_dpipe_table_erif_size_get(void *priv) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + return MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); +} + +static struct devlink_dpipe_table_ops mlxsw_sp_erif_ops = { + .matches_dump = mlxsw_sp_dpipe_table_erif_matches_dump, + .actions_dump = mlxsw_sp_dpipe_table_erif_actions_dump, + .entries_dump = mlxsw_sp_dpipe_table_erif_entries_dump, + .counters_set_update = mlxsw_sp_dpipe_table_erif_counters_update, + .size_get = mlxsw_sp_dpipe_table_erif_size_get, +}; + +static int mlxsw_sp_dpipe_erif_table_init(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + + return devl_dpipe_table_register(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_ERIF, + &mlxsw_sp_erif_ops, + mlxsw_sp, false); +} + +static void mlxsw_sp_dpipe_erif_table_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + + devl_dpipe_table_unregister(devlink, MLXSW_SP_DPIPE_TABLE_NAME_ERIF); +} + +static int mlxsw_sp_dpipe_table_host_matches_dump(struct sk_buff *skb, int type) +{ + struct devlink_dpipe_match match = {0}; + int err; + + match.type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + match.header = &mlxsw_sp_dpipe_header_metadata; + match.field_id = MLXSW_SP_DPIPE_FIELD_METADATA_ERIF_PORT; + + err = devlink_dpipe_match_put(skb, &match); + if (err) + return err; + + switch (type) { + case AF_INET: + match.type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + match.header = &devlink_dpipe_header_ipv4; + match.field_id = DEVLINK_DPIPE_FIELD_IPV4_DST_IP; + break; + case AF_INET6: + match.type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + match.header = &devlink_dpipe_header_ipv6; + match.field_id = DEVLINK_DPIPE_FIELD_IPV6_DST_IP; + break; + default: + WARN_ON(1); + return -EINVAL; + } + + return devlink_dpipe_match_put(skb, &match); +} + +static int +mlxsw_sp_dpipe_table_host4_matches_dump(void *priv, struct sk_buff *skb) +{ + return mlxsw_sp_dpipe_table_host_matches_dump(skb, AF_INET); +} + +static int +mlxsw_sp_dpipe_table_host_actions_dump(void *priv, struct sk_buff *skb) +{ + struct devlink_dpipe_action action = {0}; + + action.type = DEVLINK_DPIPE_ACTION_TYPE_FIELD_MODIFY; + action.header = &devlink_dpipe_header_ethernet; + action.field_id = DEVLINK_DPIPE_FIELD_ETHERNET_DST_MAC; + + return devlink_dpipe_action_put(skb, &action); +} + +enum mlxsw_sp_dpipe_table_host_match { + MLXSW_SP_DPIPE_TABLE_HOST_MATCH_RIF, + MLXSW_SP_DPIPE_TABLE_HOST_MATCH_DIP, + MLXSW_SP_DPIPE_TABLE_HOST_MATCH_COUNT, +}; + +static void +mlxsw_sp_dpipe_table_host_match_action_prepare(struct devlink_dpipe_match *matches, + struct devlink_dpipe_action *action, + int type) +{ + struct devlink_dpipe_match *match; + + match = &matches[MLXSW_SP_DPIPE_TABLE_HOST_MATCH_RIF]; + match->type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + match->header = &mlxsw_sp_dpipe_header_metadata; + match->field_id = MLXSW_SP_DPIPE_FIELD_METADATA_ERIF_PORT; + + match = &matches[MLXSW_SP_DPIPE_TABLE_HOST_MATCH_DIP]; + match->type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + switch (type) { + case AF_INET: + match->header = &devlink_dpipe_header_ipv4; + match->field_id = DEVLINK_DPIPE_FIELD_IPV4_DST_IP; + break; + case AF_INET6: + match->header = &devlink_dpipe_header_ipv6; + match->field_id = DEVLINK_DPIPE_FIELD_IPV6_DST_IP; + break; + default: + WARN_ON(1); + return; + } + + action->type = DEVLINK_DPIPE_ACTION_TYPE_FIELD_MODIFY; + action->header = &devlink_dpipe_header_ethernet; + action->field_id = DEVLINK_DPIPE_FIELD_ETHERNET_DST_MAC; +} + +static int +mlxsw_sp_dpipe_table_host_entry_prepare(struct devlink_dpipe_entry *entry, + struct devlink_dpipe_value *match_values, + struct devlink_dpipe_match *matches, + struct devlink_dpipe_value *action_value, + struct devlink_dpipe_action *action, + int type) +{ + struct devlink_dpipe_value *match_value; + struct devlink_dpipe_match *match; + + entry->match_values = match_values; + entry->match_values_count = MLXSW_SP_DPIPE_TABLE_HOST_MATCH_COUNT; + + entry->action_values = action_value; + entry->action_values_count = 1; + + match = &matches[MLXSW_SP_DPIPE_TABLE_HOST_MATCH_RIF]; + match_value = &match_values[MLXSW_SP_DPIPE_TABLE_HOST_MATCH_RIF]; + + match_value->match = match; + match_value->value_size = sizeof(u32); + match_value->value = kmalloc(match_value->value_size, GFP_KERNEL); + if (!match_value->value) + return -ENOMEM; + + match = &matches[MLXSW_SP_DPIPE_TABLE_HOST_MATCH_DIP]; + match_value = &match_values[MLXSW_SP_DPIPE_TABLE_HOST_MATCH_DIP]; + + match_value->match = match; + switch (type) { + case AF_INET: + match_value->value_size = sizeof(u32); + break; + case AF_INET6: + match_value->value_size = sizeof(struct in6_addr); + break; + default: + WARN_ON(1); + return -EINVAL; + } + + match_value->value = kmalloc(match_value->value_size, GFP_KERNEL); + if (!match_value->value) + return -ENOMEM; + + action_value->action = action; + action_value->value_size = sizeof(u64); + action_value->value = kmalloc(action_value->value_size, GFP_KERNEL); + if (!action_value->value) + return -ENOMEM; + + return 0; +} + +static void +__mlxsw_sp_dpipe_table_host_entry_fill(struct devlink_dpipe_entry *entry, + struct mlxsw_sp_rif *rif, + unsigned char *ha, void *dip) +{ + struct devlink_dpipe_value *value; + u32 *rif_value; + u8 *ha_value; + + /* Set Match RIF index */ + value = &entry->match_values[MLXSW_SP_DPIPE_TABLE_HOST_MATCH_RIF]; + + rif_value = value->value; + *rif_value = mlxsw_sp_rif_index(rif); + value->mapping_value = mlxsw_sp_rif_dev_ifindex(rif); + value->mapping_valid = true; + + /* Set Match DIP */ + value = &entry->match_values[MLXSW_SP_DPIPE_TABLE_HOST_MATCH_DIP]; + memcpy(value->value, dip, value->value_size); + + /* Set Action DMAC */ + value = entry->action_values; + ha_value = value->value; + ether_addr_copy(ha_value, ha); +} + +static void +mlxsw_sp_dpipe_table_host4_entry_fill(struct devlink_dpipe_entry *entry, + struct mlxsw_sp_neigh_entry *neigh_entry, + struct mlxsw_sp_rif *rif) +{ + unsigned char *ha; + u32 dip; + + ha = mlxsw_sp_neigh_entry_ha(neigh_entry); + dip = mlxsw_sp_neigh4_entry_dip(neigh_entry); + __mlxsw_sp_dpipe_table_host_entry_fill(entry, rif, ha, &dip); +} + +static void +mlxsw_sp_dpipe_table_host6_entry_fill(struct devlink_dpipe_entry *entry, + struct mlxsw_sp_neigh_entry *neigh_entry, + struct mlxsw_sp_rif *rif) +{ + struct in6_addr *dip; + unsigned char *ha; + + ha = mlxsw_sp_neigh_entry_ha(neigh_entry); + dip = mlxsw_sp_neigh6_entry_dip(neigh_entry); + + __mlxsw_sp_dpipe_table_host_entry_fill(entry, rif, ha, dip); +} + +static void +mlxsw_sp_dpipe_table_host_entry_fill(struct mlxsw_sp *mlxsw_sp, + struct devlink_dpipe_entry *entry, + struct mlxsw_sp_neigh_entry *neigh_entry, + struct mlxsw_sp_rif *rif, + int type) +{ + int err; + + switch (type) { + case AF_INET: + mlxsw_sp_dpipe_table_host4_entry_fill(entry, neigh_entry, rif); + break; + case AF_INET6: + mlxsw_sp_dpipe_table_host6_entry_fill(entry, neigh_entry, rif); + break; + default: + WARN_ON(1); + return; + } + + err = mlxsw_sp_neigh_counter_get(mlxsw_sp, neigh_entry, + &entry->counter); + if (!err) + entry->counter_valid = true; +} + +static int +mlxsw_sp_dpipe_table_host_entries_get(struct mlxsw_sp *mlxsw_sp, + struct devlink_dpipe_entry *entry, + bool counters_enabled, + struct devlink_dpipe_dump_ctx *dump_ctx, + int type) +{ + int rif_neigh_count = 0; + int rif_neigh_skip = 0; + int neigh_count = 0; + int rif_count; + int i, j; + int err; + + mutex_lock(&mlxsw_sp->router->lock); + i = 0; + rif_count = MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); +start_again: + err = devlink_dpipe_entry_ctx_prepare(dump_ctx); + if (err) + goto err_ctx_prepare; + j = 0; + rif_neigh_skip = rif_neigh_count; + for (; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); i++) { + struct mlxsw_sp_rif *rif = mlxsw_sp_rif_by_index(mlxsw_sp, i); + struct mlxsw_sp_neigh_entry *neigh_entry; + + if (!rif) + continue; + + rif_neigh_count = 0; + mlxsw_sp_rif_neigh_for_each(neigh_entry, rif) { + int neigh_type = mlxsw_sp_neigh_entry_type(neigh_entry); + + if (neigh_type != type) + continue; + + if (neigh_type == AF_INET6 && + mlxsw_sp_neigh_ipv6_ignore(neigh_entry)) + continue; + + if (rif_neigh_count < rif_neigh_skip) + goto skip; + + mlxsw_sp_dpipe_table_host_entry_fill(mlxsw_sp, entry, + neigh_entry, rif, + type); + entry->index = neigh_count; + err = devlink_dpipe_entry_ctx_append(dump_ctx, entry); + if (err) { + if (err == -EMSGSIZE) { + if (!j) + goto err_entry_append; + else + goto out; + } + goto err_entry_append; + } + neigh_count++; + j++; +skip: + rif_neigh_count++; + } + rif_neigh_skip = 0; + } +out: + devlink_dpipe_entry_ctx_close(dump_ctx); + if (i != rif_count) + goto start_again; + + mutex_unlock(&mlxsw_sp->router->lock); + return 0; + +err_ctx_prepare: +err_entry_append: + mutex_unlock(&mlxsw_sp->router->lock); + return err; +} + +static int +mlxsw_sp_dpipe_table_host_entries_dump(struct mlxsw_sp *mlxsw_sp, + bool counters_enabled, + struct devlink_dpipe_dump_ctx *dump_ctx, + int type) +{ + struct devlink_dpipe_value match_values[MLXSW_SP_DPIPE_TABLE_HOST_MATCH_COUNT]; + struct devlink_dpipe_match matches[MLXSW_SP_DPIPE_TABLE_HOST_MATCH_COUNT]; + struct devlink_dpipe_value action_value; + struct devlink_dpipe_action action = {0}; + struct devlink_dpipe_entry entry = {0}; + int err; + + memset(matches, 0, MLXSW_SP_DPIPE_TABLE_HOST_MATCH_COUNT * + sizeof(matches[0])); + memset(match_values, 0, MLXSW_SP_DPIPE_TABLE_HOST_MATCH_COUNT * + sizeof(match_values[0])); + memset(&action_value, 0, sizeof(action_value)); + + mlxsw_sp_dpipe_table_host_match_action_prepare(matches, &action, type); + err = mlxsw_sp_dpipe_table_host_entry_prepare(&entry, match_values, + matches, &action_value, + &action, type); + if (err) + goto out; + + err = mlxsw_sp_dpipe_table_host_entries_get(mlxsw_sp, &entry, + counters_enabled, dump_ctx, + type); +out: + devlink_dpipe_entry_clear(&entry); + return err; +} + +static int +mlxsw_sp_dpipe_table_host4_entries_dump(void *priv, bool counters_enabled, + struct devlink_dpipe_dump_ctx *dump_ctx) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + return mlxsw_sp_dpipe_table_host_entries_dump(mlxsw_sp, + counters_enabled, + dump_ctx, AF_INET); +} + +static void +mlxsw_sp_dpipe_table_host_counters_update(struct mlxsw_sp *mlxsw_sp, + bool enable, int type) +{ + int i; + + mutex_lock(&mlxsw_sp->router->lock); + for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); i++) { + struct mlxsw_sp_rif *rif = mlxsw_sp_rif_by_index(mlxsw_sp, i); + struct mlxsw_sp_neigh_entry *neigh_entry; + + if (!rif) + continue; + mlxsw_sp_rif_neigh_for_each(neigh_entry, rif) { + int neigh_type = mlxsw_sp_neigh_entry_type(neigh_entry); + + if (neigh_type != type) + continue; + + if (neigh_type == AF_INET6 && + mlxsw_sp_neigh_ipv6_ignore(neigh_entry)) + continue; + + mlxsw_sp_neigh_entry_counter_update(mlxsw_sp, + neigh_entry, + enable); + } + } + mutex_unlock(&mlxsw_sp->router->lock); +} + +static int mlxsw_sp_dpipe_table_host4_counters_update(void *priv, bool enable) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + mlxsw_sp_dpipe_table_host_counters_update(mlxsw_sp, enable, AF_INET); + return 0; +} + +static u64 +mlxsw_sp_dpipe_table_host_size_get(struct mlxsw_sp *mlxsw_sp, int type) +{ + u64 size = 0; + int i; + + mutex_lock(&mlxsw_sp->router->lock); + for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); i++) { + struct mlxsw_sp_rif *rif = mlxsw_sp_rif_by_index(mlxsw_sp, i); + struct mlxsw_sp_neigh_entry *neigh_entry; + + if (!rif) + continue; + mlxsw_sp_rif_neigh_for_each(neigh_entry, rif) { + int neigh_type = mlxsw_sp_neigh_entry_type(neigh_entry); + + if (neigh_type != type) + continue; + + if (neigh_type == AF_INET6 && + mlxsw_sp_neigh_ipv6_ignore(neigh_entry)) + continue; + + size++; + } + } + mutex_unlock(&mlxsw_sp->router->lock); + + return size; +} + +static u64 mlxsw_sp_dpipe_table_host4_size_get(void *priv) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + return mlxsw_sp_dpipe_table_host_size_get(mlxsw_sp, AF_INET); +} + +static struct devlink_dpipe_table_ops mlxsw_sp_host4_ops = { + .matches_dump = mlxsw_sp_dpipe_table_host4_matches_dump, + .actions_dump = mlxsw_sp_dpipe_table_host_actions_dump, + .entries_dump = mlxsw_sp_dpipe_table_host4_entries_dump, + .counters_set_update = mlxsw_sp_dpipe_table_host4_counters_update, + .size_get = mlxsw_sp_dpipe_table_host4_size_get, +}; + +#define MLXSW_SP_DPIPE_TABLE_RESOURCE_UNIT_HOST4 1 + +static int mlxsw_sp_dpipe_host4_table_init(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + int err; + + err = devl_dpipe_table_register(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST4, + &mlxsw_sp_host4_ops, + mlxsw_sp, false); + if (err) + return err; + + err = devl_dpipe_table_resource_set(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST4, + MLXSW_SP_RESOURCE_KVD_HASH_SINGLE, + MLXSW_SP_DPIPE_TABLE_RESOURCE_UNIT_HOST4); + if (err) + goto err_resource_set; + + return 0; + +err_resource_set: + devl_dpipe_table_unregister(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST4); + return err; +} + +static void mlxsw_sp_dpipe_host4_table_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + + devl_dpipe_table_unregister(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST4); +} + +static int +mlxsw_sp_dpipe_table_host6_matches_dump(void *priv, struct sk_buff *skb) +{ + return mlxsw_sp_dpipe_table_host_matches_dump(skb, AF_INET6); +} + +static int +mlxsw_sp_dpipe_table_host6_entries_dump(void *priv, bool counters_enabled, + struct devlink_dpipe_dump_ctx *dump_ctx) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + return mlxsw_sp_dpipe_table_host_entries_dump(mlxsw_sp, + counters_enabled, + dump_ctx, AF_INET6); +} + +static int mlxsw_sp_dpipe_table_host6_counters_update(void *priv, bool enable) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + mlxsw_sp_dpipe_table_host_counters_update(mlxsw_sp, enable, AF_INET6); + return 0; +} + +static u64 mlxsw_sp_dpipe_table_host6_size_get(void *priv) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + return mlxsw_sp_dpipe_table_host_size_get(mlxsw_sp, AF_INET6); +} + +static struct devlink_dpipe_table_ops mlxsw_sp_host6_ops = { + .matches_dump = mlxsw_sp_dpipe_table_host6_matches_dump, + .actions_dump = mlxsw_sp_dpipe_table_host_actions_dump, + .entries_dump = mlxsw_sp_dpipe_table_host6_entries_dump, + .counters_set_update = mlxsw_sp_dpipe_table_host6_counters_update, + .size_get = mlxsw_sp_dpipe_table_host6_size_get, +}; + +#define MLXSW_SP_DPIPE_TABLE_RESOURCE_UNIT_HOST6 2 + +static int mlxsw_sp_dpipe_host6_table_init(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + int err; + + err = devl_dpipe_table_register(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST6, + &mlxsw_sp_host6_ops, + mlxsw_sp, false); + if (err) + return err; + + err = devl_dpipe_table_resource_set(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST6, + MLXSW_SP_RESOURCE_KVD_HASH_DOUBLE, + MLXSW_SP_DPIPE_TABLE_RESOURCE_UNIT_HOST6); + if (err) + goto err_resource_set; + + return 0; + +err_resource_set: + devl_dpipe_table_unregister(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST6); + return err; +} + +static void mlxsw_sp_dpipe_host6_table_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + + devl_dpipe_table_unregister(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST6); +} + +static int mlxsw_sp_dpipe_table_adj_matches_dump(void *priv, + struct sk_buff *skb) +{ + struct devlink_dpipe_match match = {0}; + int err; + + match.type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + match.header = &mlxsw_sp_dpipe_header_metadata; + match.field_id = MLXSW_SP_DPIPE_FIELD_METADATA_ADJ_INDEX; + + err = devlink_dpipe_match_put(skb, &match); + if (err) + return err; + + match.type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + match.header = &mlxsw_sp_dpipe_header_metadata; + match.field_id = MLXSW_SP_DPIPE_FIELD_METADATA_ADJ_SIZE; + + err = devlink_dpipe_match_put(skb, &match); + if (err) + return err; + + match.type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + match.header = &mlxsw_sp_dpipe_header_metadata; + match.field_id = MLXSW_SP_DPIPE_FIELD_METADATA_ADJ_HASH_INDEX; + + return devlink_dpipe_match_put(skb, &match); +} + +static int mlxsw_sp_dpipe_table_adj_actions_dump(void *priv, + struct sk_buff *skb) +{ + struct devlink_dpipe_action action = {0}; + int err; + + action.type = DEVLINK_DPIPE_ACTION_TYPE_FIELD_MODIFY; + action.header = &devlink_dpipe_header_ethernet; + action.field_id = DEVLINK_DPIPE_FIELD_ETHERNET_DST_MAC; + + err = devlink_dpipe_action_put(skb, &action); + if (err) + return err; + + action.type = DEVLINK_DPIPE_ACTION_TYPE_FIELD_MODIFY; + action.header = &mlxsw_sp_dpipe_header_metadata; + action.field_id = MLXSW_SP_DPIPE_FIELD_METADATA_ERIF_PORT; + + return devlink_dpipe_action_put(skb, &action); +} + +static u64 mlxsw_sp_dpipe_table_adj_size(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_nexthop *nh; + u64 size = 0; + + mlxsw_sp_nexthop_for_each(nh, mlxsw_sp->router) + if (mlxsw_sp_nexthop_is_forward(nh) && + !mlxsw_sp_nexthop_group_has_ipip(nh)) + size++; + return size; +} + +enum mlxsw_sp_dpipe_table_adj_match { + MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_INDEX, + MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_SIZE, + MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_HASH_INDEX, + MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_COUNT, +}; + +enum mlxsw_sp_dpipe_table_adj_action { + MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_DST_MAC, + MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_ERIF_PORT, + MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_COUNT, +}; + +static void +mlxsw_sp_dpipe_table_adj_match_action_prepare(struct devlink_dpipe_match *matches, + struct devlink_dpipe_action *actions) +{ + struct devlink_dpipe_action *action; + struct devlink_dpipe_match *match; + + match = &matches[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_INDEX]; + match->type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + match->header = &mlxsw_sp_dpipe_header_metadata; + match->field_id = MLXSW_SP_DPIPE_FIELD_METADATA_ADJ_INDEX; + + match = &matches[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_SIZE]; + match->type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + match->header = &mlxsw_sp_dpipe_header_metadata; + match->field_id = MLXSW_SP_DPIPE_FIELD_METADATA_ADJ_SIZE; + + match = &matches[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_HASH_INDEX]; + match->type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT; + match->header = &mlxsw_sp_dpipe_header_metadata; + match->field_id = MLXSW_SP_DPIPE_FIELD_METADATA_ADJ_HASH_INDEX; + + action = &actions[MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_DST_MAC]; + action->type = DEVLINK_DPIPE_ACTION_TYPE_FIELD_MODIFY; + action->header = &devlink_dpipe_header_ethernet; + action->field_id = DEVLINK_DPIPE_FIELD_ETHERNET_DST_MAC; + + action = &actions[MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_ERIF_PORT]; + action->type = DEVLINK_DPIPE_ACTION_TYPE_FIELD_MODIFY; + action->header = &mlxsw_sp_dpipe_header_metadata; + action->field_id = MLXSW_SP_DPIPE_FIELD_METADATA_ERIF_PORT; +} + +static int +mlxsw_sp_dpipe_table_adj_entry_prepare(struct devlink_dpipe_entry *entry, + struct devlink_dpipe_value *match_values, + struct devlink_dpipe_match *matches, + struct devlink_dpipe_value *action_values, + struct devlink_dpipe_action *actions) +{ struct devlink_dpipe_value *action_value; + struct devlink_dpipe_value *match_value; + struct devlink_dpipe_action *action; + struct devlink_dpipe_match *match; + + entry->match_values = match_values; + entry->match_values_count = MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_COUNT; + + entry->action_values = action_values; + entry->action_values_count = MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_COUNT; + + match = &matches[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_INDEX]; + match_value = &match_values[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_INDEX]; + + match_value->match = match; + match_value->value_size = sizeof(u32); + match_value->value = kmalloc(match_value->value_size, GFP_KERNEL); + if (!match_value->value) + return -ENOMEM; + + match = &matches[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_SIZE]; + match_value = &match_values[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_SIZE]; + + match_value->match = match; + match_value->value_size = sizeof(u32); + match_value->value = kmalloc(match_value->value_size, GFP_KERNEL); + if (!match_value->value) + return -ENOMEM; + + match = &matches[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_HASH_INDEX]; + match_value = &match_values[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_HASH_INDEX]; + + match_value->match = match; + match_value->value_size = sizeof(u32); + match_value->value = kmalloc(match_value->value_size, GFP_KERNEL); + if (!match_value->value) + return -ENOMEM; + + action = &actions[MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_DST_MAC]; + action_value = &action_values[MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_DST_MAC]; + + action_value->action = action; + action_value->value_size = sizeof(u64); + action_value->value = kmalloc(action_value->value_size, GFP_KERNEL); + if (!action_value->value) + return -ENOMEM; + + action = &actions[MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_ERIF_PORT]; + action_value = &action_values[MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_ERIF_PORT]; + + action_value->action = action; + action_value->value_size = sizeof(u32); + action_value->value = kmalloc(action_value->value_size, GFP_KERNEL); + if (!action_value->value) + return -ENOMEM; + + return 0; +} + +static void +__mlxsw_sp_dpipe_table_adj_entry_fill(struct devlink_dpipe_entry *entry, + u32 adj_index, u32 adj_size, + u32 adj_hash_index, unsigned char *ha, + struct mlxsw_sp_rif *rif) +{ + struct devlink_dpipe_value *value; + u32 *p_rif_value; + u32 *p_index; + + value = &entry->match_values[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_INDEX]; + p_index = value->value; + *p_index = adj_index; + + value = &entry->match_values[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_SIZE]; + p_index = value->value; + *p_index = adj_size; + + value = &entry->match_values[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_HASH_INDEX]; + p_index = value->value; + *p_index = adj_hash_index; + + value = &entry->action_values[MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_DST_MAC]; + ether_addr_copy(value->value, ha); + + value = &entry->action_values[MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_ERIF_PORT]; + p_rif_value = value->value; + *p_rif_value = mlxsw_sp_rif_index(rif); + value->mapping_value = mlxsw_sp_rif_dev_ifindex(rif); + value->mapping_valid = true; +} + +static void mlxsw_sp_dpipe_table_adj_entry_fill(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh, + struct devlink_dpipe_entry *entry) +{ + struct mlxsw_sp_rif *rif = mlxsw_sp_nexthop_rif(nh); + unsigned char *ha = mlxsw_sp_nexthop_ha(nh); + u32 adj_hash_index = 0; + u32 adj_index = 0; + u32 adj_size = 0; + int err; + + mlxsw_sp_nexthop_indexes(nh, &adj_index, &adj_size, &adj_hash_index); + __mlxsw_sp_dpipe_table_adj_entry_fill(entry, adj_index, adj_size, + adj_hash_index, ha, rif); + err = mlxsw_sp_nexthop_counter_get(mlxsw_sp, nh, &entry->counter); + if (!err) + entry->counter_valid = true; +} + +static int +mlxsw_sp_dpipe_table_adj_entries_get(struct mlxsw_sp *mlxsw_sp, + struct devlink_dpipe_entry *entry, + bool counters_enabled, + struct devlink_dpipe_dump_ctx *dump_ctx) +{ + struct mlxsw_sp_nexthop *nh; + int entry_index = 0; + int nh_count_max; + int nh_count = 0; + int nh_skip; + int j; + int err; + + mutex_lock(&mlxsw_sp->router->lock); + nh_count_max = mlxsw_sp_dpipe_table_adj_size(mlxsw_sp); +start_again: + err = devlink_dpipe_entry_ctx_prepare(dump_ctx); + if (err) + goto err_ctx_prepare; + j = 0; + nh_skip = nh_count; + nh_count = 0; + mlxsw_sp_nexthop_for_each(nh, mlxsw_sp->router) { + if (!mlxsw_sp_nexthop_is_forward(nh) || + mlxsw_sp_nexthop_group_has_ipip(nh)) + continue; + + if (nh_count < nh_skip) + goto skip; + + mlxsw_sp_dpipe_table_adj_entry_fill(mlxsw_sp, nh, entry); + entry->index = entry_index; + err = devlink_dpipe_entry_ctx_append(dump_ctx, entry); + if (err) { + if (err == -EMSGSIZE) { + if (!j) + goto err_entry_append; + break; + } + goto err_entry_append; + } + entry_index++; + j++; +skip: + nh_count++; + } + + devlink_dpipe_entry_ctx_close(dump_ctx); + if (nh_count != nh_count_max) + goto start_again; + mutex_unlock(&mlxsw_sp->router->lock); + + return 0; + +err_ctx_prepare: +err_entry_append: + mutex_unlock(&mlxsw_sp->router->lock); + return err; +} + +static int +mlxsw_sp_dpipe_table_adj_entries_dump(void *priv, bool counters_enabled, + struct devlink_dpipe_dump_ctx *dump_ctx) +{ + struct devlink_dpipe_value action_values[MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_COUNT]; + struct devlink_dpipe_value match_values[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_COUNT]; + struct devlink_dpipe_action actions[MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_COUNT]; + struct devlink_dpipe_match matches[MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_COUNT]; + struct devlink_dpipe_entry entry = {0}; + struct mlxsw_sp *mlxsw_sp = priv; + int err; + + memset(matches, 0, MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_COUNT * + sizeof(matches[0])); + memset(match_values, 0, MLXSW_SP_DPIPE_TABLE_ADJ_MATCH_COUNT * + sizeof(match_values[0])); + memset(actions, 0, MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_COUNT * + sizeof(actions[0])); + memset(action_values, 0, MLXSW_SP_DPIPE_TABLE_ADJ_ACTION_COUNT * + sizeof(action_values[0])); + + mlxsw_sp_dpipe_table_adj_match_action_prepare(matches, actions); + err = mlxsw_sp_dpipe_table_adj_entry_prepare(&entry, + match_values, matches, + action_values, actions); + if (err) + goto out; + + err = mlxsw_sp_dpipe_table_adj_entries_get(mlxsw_sp, &entry, + counters_enabled, dump_ctx); +out: + devlink_dpipe_entry_clear(&entry); + return err; +} + +static int mlxsw_sp_dpipe_table_adj_counters_update(void *priv, bool enable) +{ + char ratr_pl[MLXSW_REG_RATR_LEN]; + struct mlxsw_sp *mlxsw_sp = priv; + struct mlxsw_sp_nexthop *nh; + u32 adj_hash_index = 0; + u32 adj_index = 0; + u32 adj_size = 0; + + mlxsw_sp_nexthop_for_each(nh, mlxsw_sp->router) { + if (!mlxsw_sp_nexthop_is_forward(nh) || + mlxsw_sp_nexthop_group_has_ipip(nh)) + continue; + + mlxsw_sp_nexthop_indexes(nh, &adj_index, &adj_size, + &adj_hash_index); + if (enable) + mlxsw_sp_nexthop_counter_alloc(mlxsw_sp, nh); + else + mlxsw_sp_nexthop_counter_free(mlxsw_sp, nh); + mlxsw_sp_nexthop_eth_update(mlxsw_sp, + adj_index + adj_hash_index, nh, + true, ratr_pl); + } + return 0; +} + +static u64 +mlxsw_sp_dpipe_table_adj_size_get(void *priv) +{ + struct mlxsw_sp *mlxsw_sp = priv; + u64 size; + + mutex_lock(&mlxsw_sp->router->lock); + size = mlxsw_sp_dpipe_table_adj_size(mlxsw_sp); + mutex_unlock(&mlxsw_sp->router->lock); + + return size; +} + +static struct devlink_dpipe_table_ops mlxsw_sp_dpipe_table_adj_ops = { + .matches_dump = mlxsw_sp_dpipe_table_adj_matches_dump, + .actions_dump = mlxsw_sp_dpipe_table_adj_actions_dump, + .entries_dump = mlxsw_sp_dpipe_table_adj_entries_dump, + .counters_set_update = mlxsw_sp_dpipe_table_adj_counters_update, + .size_get = mlxsw_sp_dpipe_table_adj_size_get, +}; + +#define MLXSW_SP_DPIPE_TABLE_RESOURCE_UNIT_ADJ 1 + +static int mlxsw_sp_dpipe_adj_table_init(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + int err; + + err = devl_dpipe_table_register(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_ADJ, + &mlxsw_sp_dpipe_table_adj_ops, + mlxsw_sp, false); + if (err) + return err; + + err = devl_dpipe_table_resource_set(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_ADJ, + MLXSW_SP_RESOURCE_KVD_LINEAR, + MLXSW_SP_DPIPE_TABLE_RESOURCE_UNIT_ADJ); + if (err) + goto err_resource_set; + + return 0; + +err_resource_set: + devl_dpipe_table_unregister(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_ADJ); + return err; +} + +static void mlxsw_sp_dpipe_adj_table_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + + devl_dpipe_table_unregister(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_ADJ); +} + +int mlxsw_sp_dpipe_init(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + int err; + + devl_dpipe_headers_register(devlink, &mlxsw_sp_dpipe_headers); + + err = mlxsw_sp_dpipe_erif_table_init(mlxsw_sp); + if (err) + goto err_erif_table_init; + + err = mlxsw_sp_dpipe_host4_table_init(mlxsw_sp); + if (err) + goto err_host4_table_init; + + err = mlxsw_sp_dpipe_host6_table_init(mlxsw_sp); + if (err) + goto err_host6_table_init; + + err = mlxsw_sp_dpipe_adj_table_init(mlxsw_sp); + if (err) + goto err_adj_table_init; + + return 0; +err_adj_table_init: + mlxsw_sp_dpipe_host6_table_fini(mlxsw_sp); +err_host6_table_init: + mlxsw_sp_dpipe_host4_table_fini(mlxsw_sp); +err_host4_table_init: + mlxsw_sp_dpipe_erif_table_fini(mlxsw_sp); +err_erif_table_init: + devl_dpipe_headers_unregister(priv_to_devlink(mlxsw_sp->core)); + return err; +} + +void mlxsw_sp_dpipe_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + + mlxsw_sp_dpipe_adj_table_fini(mlxsw_sp); + mlxsw_sp_dpipe_host6_table_fini(mlxsw_sp); + mlxsw_sp_dpipe_host4_table_fini(mlxsw_sp); + mlxsw_sp_dpipe_erif_table_fini(mlxsw_sp); + devl_dpipe_headers_unregister(devlink); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.h new file mode 100644 index 000000000..246dbb3c0 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.h @@ -0,0 +1,15 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_PIPELINE_H_ +#define _MLXSW_PIPELINE_H_ + +int mlxsw_sp_dpipe_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_dpipe_fini(struct mlxsw_sp *mlxsw_sp); + +#define MLXSW_SP_DPIPE_TABLE_NAME_ERIF "mlxsw_erif" +#define MLXSW_SP_DPIPE_TABLE_NAME_HOST4 "mlxsw_host4" +#define MLXSW_SP_DPIPE_TABLE_NAME_HOST6 "mlxsw_host6" +#define MLXSW_SP_DPIPE_TABLE_NAME_ADJ "mlxsw_adj" + +#endif /* _MLXSW_PIPELINE_H_*/ diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c new file mode 100644 index 000000000..dcd79d7e2 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c @@ -0,0 +1,2023 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2020 Mellanox Technologies. All rights reserved */ + +#include "reg.h" +#include "core.h" +#include "spectrum.h" +#include "core_env.h" + +static const char mlxsw_sp_driver_version[] = "1.0"; + +static void mlxsw_sp_port_get_drvinfo(struct net_device *dev, + struct ethtool_drvinfo *drvinfo) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + + strscpy(drvinfo->driver, mlxsw_sp->bus_info->device_kind, + sizeof(drvinfo->driver)); + strscpy(drvinfo->version, mlxsw_sp_driver_version, + sizeof(drvinfo->version)); + snprintf(drvinfo->fw_version, sizeof(drvinfo->fw_version), + "%d.%d.%d", + mlxsw_sp->bus_info->fw_rev.major, + mlxsw_sp->bus_info->fw_rev.minor, + mlxsw_sp->bus_info->fw_rev.subminor); + strscpy(drvinfo->bus_info, mlxsw_sp->bus_info->device_name, + sizeof(drvinfo->bus_info)); +} + +struct mlxsw_sp_ethtool_link_ext_state_opcode_mapping { + u32 status_opcode; + enum ethtool_link_ext_state link_ext_state; + u8 link_ext_substate; +}; + +static const struct mlxsw_sp_ethtool_link_ext_state_opcode_mapping +mlxsw_sp_link_ext_state_opcode_map[] = { + {2, ETHTOOL_LINK_EXT_STATE_AUTONEG, + ETHTOOL_LINK_EXT_SUBSTATE_AN_NO_PARTNER_DETECTED}, + {3, ETHTOOL_LINK_EXT_STATE_AUTONEG, + ETHTOOL_LINK_EXT_SUBSTATE_AN_ACK_NOT_RECEIVED}, + {4, ETHTOOL_LINK_EXT_STATE_AUTONEG, + ETHTOOL_LINK_EXT_SUBSTATE_AN_NEXT_PAGE_EXCHANGE_FAILED}, + {36, ETHTOOL_LINK_EXT_STATE_AUTONEG, + ETHTOOL_LINK_EXT_SUBSTATE_AN_NO_PARTNER_DETECTED_FORCE_MODE}, + {38, ETHTOOL_LINK_EXT_STATE_AUTONEG, + ETHTOOL_LINK_EXT_SUBSTATE_AN_FEC_MISMATCH_DURING_OVERRIDE}, + {39, ETHTOOL_LINK_EXT_STATE_AUTONEG, + ETHTOOL_LINK_EXT_SUBSTATE_AN_NO_HCD}, + + {5, ETHTOOL_LINK_EXT_STATE_LINK_TRAINING_FAILURE, + ETHTOOL_LINK_EXT_SUBSTATE_LT_KR_FRAME_LOCK_NOT_ACQUIRED}, + {6, ETHTOOL_LINK_EXT_STATE_LINK_TRAINING_FAILURE, + ETHTOOL_LINK_EXT_SUBSTATE_LT_KR_LINK_INHIBIT_TIMEOUT}, + {7, ETHTOOL_LINK_EXT_STATE_LINK_TRAINING_FAILURE, + ETHTOOL_LINK_EXT_SUBSTATE_LT_KR_LINK_PARTNER_DID_NOT_SET_RECEIVER_READY}, + {8, ETHTOOL_LINK_EXT_STATE_LINK_TRAINING_FAILURE, 0}, + {14, ETHTOOL_LINK_EXT_STATE_LINK_TRAINING_FAILURE, + ETHTOOL_LINK_EXT_SUBSTATE_LT_REMOTE_FAULT}, + + {9, ETHTOOL_LINK_EXT_STATE_LINK_LOGICAL_MISMATCH, + ETHTOOL_LINK_EXT_SUBSTATE_LLM_PCS_DID_NOT_ACQUIRE_BLOCK_LOCK}, + {10, ETHTOOL_LINK_EXT_STATE_LINK_LOGICAL_MISMATCH, + ETHTOOL_LINK_EXT_SUBSTATE_LLM_PCS_DID_NOT_ACQUIRE_AM_LOCK}, + {11, ETHTOOL_LINK_EXT_STATE_LINK_LOGICAL_MISMATCH, + ETHTOOL_LINK_EXT_SUBSTATE_LLM_PCS_DID_NOT_GET_ALIGN_STATUS}, + {12, ETHTOOL_LINK_EXT_STATE_LINK_LOGICAL_MISMATCH, + ETHTOOL_LINK_EXT_SUBSTATE_LLM_FC_FEC_IS_NOT_LOCKED}, + {13, ETHTOOL_LINK_EXT_STATE_LINK_LOGICAL_MISMATCH, + ETHTOOL_LINK_EXT_SUBSTATE_LLM_RS_FEC_IS_NOT_LOCKED}, + + {15, ETHTOOL_LINK_EXT_STATE_BAD_SIGNAL_INTEGRITY, 0}, + {17, ETHTOOL_LINK_EXT_STATE_BAD_SIGNAL_INTEGRITY, + ETHTOOL_LINK_EXT_SUBSTATE_BSI_LARGE_NUMBER_OF_PHYSICAL_ERRORS}, + {42, ETHTOOL_LINK_EXT_STATE_BAD_SIGNAL_INTEGRITY, + ETHTOOL_LINK_EXT_SUBSTATE_BSI_UNSUPPORTED_RATE}, + + {1024, ETHTOOL_LINK_EXT_STATE_NO_CABLE, 0}, + + {16, ETHTOOL_LINK_EXT_STATE_CABLE_ISSUE, + ETHTOOL_LINK_EXT_SUBSTATE_CI_UNSUPPORTED_CABLE}, + {20, ETHTOOL_LINK_EXT_STATE_CABLE_ISSUE, + ETHTOOL_LINK_EXT_SUBSTATE_CI_UNSUPPORTED_CABLE}, + {29, ETHTOOL_LINK_EXT_STATE_CABLE_ISSUE, + ETHTOOL_LINK_EXT_SUBSTATE_CI_UNSUPPORTED_CABLE}, + {1025, ETHTOOL_LINK_EXT_STATE_CABLE_ISSUE, + ETHTOOL_LINK_EXT_SUBSTATE_CI_UNSUPPORTED_CABLE}, + {1029, ETHTOOL_LINK_EXT_STATE_CABLE_ISSUE, + ETHTOOL_LINK_EXT_SUBSTATE_CI_UNSUPPORTED_CABLE}, + {1031, ETHTOOL_LINK_EXT_STATE_CABLE_ISSUE, 0}, + + {1027, ETHTOOL_LINK_EXT_STATE_EEPROM_ISSUE, 0}, + + {23, ETHTOOL_LINK_EXT_STATE_CALIBRATION_FAILURE, 0}, + + {1032, ETHTOOL_LINK_EXT_STATE_POWER_BUDGET_EXCEEDED, 0}, + + {1030, ETHTOOL_LINK_EXT_STATE_OVERHEAT, 0}, + + {1042, ETHTOOL_LINK_EXT_STATE_MODULE, + ETHTOOL_LINK_EXT_SUBSTATE_MODULE_CMIS_NOT_READY}, +}; + +static void +mlxsw_sp_port_set_link_ext_state(struct mlxsw_sp_ethtool_link_ext_state_opcode_mapping + link_ext_state_mapping, + struct ethtool_link_ext_state_info *link_ext_state_info) +{ + switch (link_ext_state_mapping.link_ext_state) { + case ETHTOOL_LINK_EXT_STATE_AUTONEG: + link_ext_state_info->autoneg = + link_ext_state_mapping.link_ext_substate; + break; + case ETHTOOL_LINK_EXT_STATE_LINK_TRAINING_FAILURE: + link_ext_state_info->link_training = + link_ext_state_mapping.link_ext_substate; + break; + case ETHTOOL_LINK_EXT_STATE_LINK_LOGICAL_MISMATCH: + link_ext_state_info->link_logical_mismatch = + link_ext_state_mapping.link_ext_substate; + break; + case ETHTOOL_LINK_EXT_STATE_BAD_SIGNAL_INTEGRITY: + link_ext_state_info->bad_signal_integrity = + link_ext_state_mapping.link_ext_substate; + break; + case ETHTOOL_LINK_EXT_STATE_CABLE_ISSUE: + link_ext_state_info->cable_issue = + link_ext_state_mapping.link_ext_substate; + break; + case ETHTOOL_LINK_EXT_STATE_MODULE: + link_ext_state_info->module = + link_ext_state_mapping.link_ext_substate; + break; + default: + break; + } + + link_ext_state_info->link_ext_state = link_ext_state_mapping.link_ext_state; +} + +static int +mlxsw_sp_port_get_link_ext_state(struct net_device *dev, + struct ethtool_link_ext_state_info *link_ext_state_info) +{ + struct mlxsw_sp_ethtool_link_ext_state_opcode_mapping link_ext_state_mapping; + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + char pddr_pl[MLXSW_REG_PDDR_LEN]; + int opcode, err, i; + u32 status_opcode; + + if (netif_carrier_ok(dev)) + return -ENODATA; + + mlxsw_reg_pddr_pack(pddr_pl, mlxsw_sp_port->local_port, + MLXSW_REG_PDDR_PAGE_SELECT_TROUBLESHOOTING_INFO); + + opcode = MLXSW_REG_PDDR_TRBLSH_GROUP_OPCODE_MONITOR; + mlxsw_reg_pddr_trblsh_group_opcode_set(pddr_pl, opcode); + + err = mlxsw_reg_query(mlxsw_sp_port->mlxsw_sp->core, MLXSW_REG(pddr), + pddr_pl); + if (err) + return err; + + status_opcode = mlxsw_reg_pddr_trblsh_status_opcode_get(pddr_pl); + if (!status_opcode) + return -ENODATA; + + for (i = 0; i < ARRAY_SIZE(mlxsw_sp_link_ext_state_opcode_map); i++) { + link_ext_state_mapping = mlxsw_sp_link_ext_state_opcode_map[i]; + if (link_ext_state_mapping.status_opcode == status_opcode) { + mlxsw_sp_port_set_link_ext_state(link_ext_state_mapping, + link_ext_state_info); + return 0; + } + } + + return -ENODATA; +} + +static void mlxsw_sp_port_get_pauseparam(struct net_device *dev, + struct ethtool_pauseparam *pause) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + + pause->rx_pause = mlxsw_sp_port->link.rx_pause; + pause->tx_pause = mlxsw_sp_port->link.tx_pause; +} + +static int mlxsw_sp_port_pause_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct ethtool_pauseparam *pause) +{ + char pfcc_pl[MLXSW_REG_PFCC_LEN]; + + mlxsw_reg_pfcc_pack(pfcc_pl, mlxsw_sp_port->local_port); + mlxsw_reg_pfcc_pprx_set(pfcc_pl, pause->rx_pause); + mlxsw_reg_pfcc_pptx_set(pfcc_pl, pause->tx_pause); + + return mlxsw_reg_write(mlxsw_sp_port->mlxsw_sp->core, MLXSW_REG(pfcc), + pfcc_pl); +} + +/* Maximum delay buffer needed in case of PAUSE frames. Similar to PFC delay, but is + * measured in bytes. Assumes 100m cable and does not take into account MTU. + */ +#define MLXSW_SP_PAUSE_DELAY_BYTES 19476 + +static int mlxsw_sp_port_set_pauseparam(struct net_device *dev, + struct ethtool_pauseparam *pause) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + bool pause_en = pause->tx_pause || pause->rx_pause; + struct mlxsw_sp_hdroom orig_hdroom; + struct mlxsw_sp_hdroom hdroom; + int prio; + int err; + + if (mlxsw_sp_port->dcb.pfc && mlxsw_sp_port->dcb.pfc->pfc_en) { + netdev_err(dev, "PFC already enabled on port\n"); + return -EINVAL; + } + + if (pause->autoneg) { + netdev_err(dev, "PAUSE frames autonegotiation isn't supported\n"); + return -EINVAL; + } + + orig_hdroom = *mlxsw_sp_port->hdroom; + + hdroom = orig_hdroom; + if (pause_en) + hdroom.delay_bytes = MLXSW_SP_PAUSE_DELAY_BYTES; + else + hdroom.delay_bytes = 0; + + for (prio = 0; prio < IEEE_8021QAZ_MAX_TCS; prio++) + hdroom.prios.prio[prio].lossy = !pause_en; + + mlxsw_sp_hdroom_bufs_reset_lossiness(&hdroom); + mlxsw_sp_hdroom_bufs_reset_sizes(mlxsw_sp_port, &hdroom); + + err = mlxsw_sp_hdroom_configure(mlxsw_sp_port, &hdroom); + if (err) { + netdev_err(dev, "Failed to configure port's headroom\n"); + return err; + } + + err = mlxsw_sp_port_pause_set(mlxsw_sp_port, pause); + if (err) { + netdev_err(dev, "Failed to set PAUSE parameters\n"); + goto err_port_pause_configure; + } + + mlxsw_sp_port->link.rx_pause = pause->rx_pause; + mlxsw_sp_port->link.tx_pause = pause->tx_pause; + + return 0; + +err_port_pause_configure: + mlxsw_sp_hdroom_configure(mlxsw_sp_port, &orig_hdroom); + return err; +} + +struct mlxsw_sp_port_hw_stats { + char str[ETH_GSTRING_LEN]; + u64 (*getter)(const char *payload); + bool cells_bytes; +}; + +static struct mlxsw_sp_port_hw_stats mlxsw_sp_port_hw_stats[] = { + { + .str = "a_frames_transmitted_ok", + .getter = mlxsw_reg_ppcnt_a_frames_transmitted_ok_get, + }, + { + .str = "a_frames_received_ok", + .getter = mlxsw_reg_ppcnt_a_frames_received_ok_get, + }, + { + .str = "a_frame_check_sequence_errors", + .getter = mlxsw_reg_ppcnt_a_frame_check_sequence_errors_get, + }, + { + .str = "a_alignment_errors", + .getter = mlxsw_reg_ppcnt_a_alignment_errors_get, + }, + { + .str = "a_octets_transmitted_ok", + .getter = mlxsw_reg_ppcnt_a_octets_transmitted_ok_get, + }, + { + .str = "a_octets_received_ok", + .getter = mlxsw_reg_ppcnt_a_octets_received_ok_get, + }, + { + .str = "a_multicast_frames_xmitted_ok", + .getter = mlxsw_reg_ppcnt_a_multicast_frames_xmitted_ok_get, + }, + { + .str = "a_broadcast_frames_xmitted_ok", + .getter = mlxsw_reg_ppcnt_a_broadcast_frames_xmitted_ok_get, + }, + { + .str = "a_multicast_frames_received_ok", + .getter = mlxsw_reg_ppcnt_a_multicast_frames_received_ok_get, + }, + { + .str = "a_broadcast_frames_received_ok", + .getter = mlxsw_reg_ppcnt_a_broadcast_frames_received_ok_get, + }, + { + .str = "a_in_range_length_errors", + .getter = mlxsw_reg_ppcnt_a_in_range_length_errors_get, + }, + { + .str = "a_out_of_range_length_field", + .getter = mlxsw_reg_ppcnt_a_out_of_range_length_field_get, + }, + { + .str = "a_frame_too_long_errors", + .getter = mlxsw_reg_ppcnt_a_frame_too_long_errors_get, + }, + { + .str = "a_symbol_error_during_carrier", + .getter = mlxsw_reg_ppcnt_a_symbol_error_during_carrier_get, + }, + { + .str = "a_mac_control_frames_transmitted", + .getter = mlxsw_reg_ppcnt_a_mac_control_frames_transmitted_get, + }, + { + .str = "a_mac_control_frames_received", + .getter = mlxsw_reg_ppcnt_a_mac_control_frames_received_get, + }, + { + .str = "a_unsupported_opcodes_received", + .getter = mlxsw_reg_ppcnt_a_unsupported_opcodes_received_get, + }, + { + .str = "a_pause_mac_ctrl_frames_received", + .getter = mlxsw_reg_ppcnt_a_pause_mac_ctrl_frames_received_get, + }, + { + .str = "a_pause_mac_ctrl_frames_xmitted", + .getter = mlxsw_reg_ppcnt_a_pause_mac_ctrl_frames_transmitted_get, + }, +}; + +#define MLXSW_SP_PORT_HW_STATS_LEN ARRAY_SIZE(mlxsw_sp_port_hw_stats) + +static struct mlxsw_sp_port_hw_stats mlxsw_sp_port_hw_rfc_2863_stats[] = { + { + .str = "if_in_discards", + .getter = mlxsw_reg_ppcnt_if_in_discards_get, + }, + { + .str = "if_out_discards", + .getter = mlxsw_reg_ppcnt_if_out_discards_get, + }, + { + .str = "if_out_errors", + .getter = mlxsw_reg_ppcnt_if_out_errors_get, + }, +}; + +#define MLXSW_SP_PORT_HW_RFC_2863_STATS_LEN \ + ARRAY_SIZE(mlxsw_sp_port_hw_rfc_2863_stats) + +static struct mlxsw_sp_port_hw_stats mlxsw_sp_port_hw_rfc_2819_stats[] = { + { + .str = "ether_stats_undersize_pkts", + .getter = mlxsw_reg_ppcnt_ether_stats_undersize_pkts_get, + }, + { + .str = "ether_stats_oversize_pkts", + .getter = mlxsw_reg_ppcnt_ether_stats_oversize_pkts_get, + }, + { + .str = "ether_stats_fragments", + .getter = mlxsw_reg_ppcnt_ether_stats_fragments_get, + }, + { + .str = "ether_pkts64octets", + .getter = mlxsw_reg_ppcnt_ether_stats_pkts64octets_get, + }, + { + .str = "ether_pkts65to127octets", + .getter = mlxsw_reg_ppcnt_ether_stats_pkts65to127octets_get, + }, + { + .str = "ether_pkts128to255octets", + .getter = mlxsw_reg_ppcnt_ether_stats_pkts128to255octets_get, + }, + { + .str = "ether_pkts256to511octets", + .getter = mlxsw_reg_ppcnt_ether_stats_pkts256to511octets_get, + }, + { + .str = "ether_pkts512to1023octets", + .getter = mlxsw_reg_ppcnt_ether_stats_pkts512to1023octets_get, + }, + { + .str = "ether_pkts1024to1518octets", + .getter = mlxsw_reg_ppcnt_ether_stats_pkts1024to1518octets_get, + }, + { + .str = "ether_pkts1519to2047octets", + .getter = mlxsw_reg_ppcnt_ether_stats_pkts1519to2047octets_get, + }, + { + .str = "ether_pkts2048to4095octets", + .getter = mlxsw_reg_ppcnt_ether_stats_pkts2048to4095octets_get, + }, + { + .str = "ether_pkts4096to8191octets", + .getter = mlxsw_reg_ppcnt_ether_stats_pkts4096to8191octets_get, + }, + { + .str = "ether_pkts8192to10239octets", + .getter = mlxsw_reg_ppcnt_ether_stats_pkts8192to10239octets_get, + }, +}; + +#define MLXSW_SP_PORT_HW_RFC_2819_STATS_LEN \ + ARRAY_SIZE(mlxsw_sp_port_hw_rfc_2819_stats) + +static struct mlxsw_sp_port_hw_stats mlxsw_sp_port_hw_rfc_3635_stats[] = { + { + .str = "dot3stats_fcs_errors", + .getter = mlxsw_reg_ppcnt_dot3stats_fcs_errors_get, + }, + { + .str = "dot3stats_symbol_errors", + .getter = mlxsw_reg_ppcnt_dot3stats_symbol_errors_get, + }, + { + .str = "dot3control_in_unknown_opcodes", + .getter = mlxsw_reg_ppcnt_dot3control_in_unknown_opcodes_get, + }, + { + .str = "dot3in_pause_frames", + .getter = mlxsw_reg_ppcnt_dot3in_pause_frames_get, + }, +}; + +#define MLXSW_SP_PORT_HW_RFC_3635_STATS_LEN \ + ARRAY_SIZE(mlxsw_sp_port_hw_rfc_3635_stats) + +static struct mlxsw_sp_port_hw_stats mlxsw_sp_port_hw_ext_stats[] = { + { + .str = "ecn_marked", + .getter = mlxsw_reg_ppcnt_ecn_marked_get, + }, +}; + +#define MLXSW_SP_PORT_HW_EXT_STATS_LEN ARRAY_SIZE(mlxsw_sp_port_hw_ext_stats) + +static struct mlxsw_sp_port_hw_stats mlxsw_sp_port_hw_discard_stats[] = { + { + .str = "discard_ingress_general", + .getter = mlxsw_reg_ppcnt_ingress_general_get, + }, + { + .str = "discard_ingress_policy_engine", + .getter = mlxsw_reg_ppcnt_ingress_policy_engine_get, + }, + { + .str = "discard_ingress_vlan_membership", + .getter = mlxsw_reg_ppcnt_ingress_vlan_membership_get, + }, + { + .str = "discard_ingress_tag_frame_type", + .getter = mlxsw_reg_ppcnt_ingress_tag_frame_type_get, + }, + { + .str = "discard_egress_vlan_membership", + .getter = mlxsw_reg_ppcnt_egress_vlan_membership_get, + }, + { + .str = "discard_loopback_filter", + .getter = mlxsw_reg_ppcnt_loopback_filter_get, + }, + { + .str = "discard_egress_general", + .getter = mlxsw_reg_ppcnt_egress_general_get, + }, + { + .str = "discard_egress_hoq", + .getter = mlxsw_reg_ppcnt_egress_hoq_get, + }, + { + .str = "discard_egress_policy_engine", + .getter = mlxsw_reg_ppcnt_egress_policy_engine_get, + }, + { + .str = "discard_ingress_tx_link_down", + .getter = mlxsw_reg_ppcnt_ingress_tx_link_down_get, + }, + { + .str = "discard_egress_stp_filter", + .getter = mlxsw_reg_ppcnt_egress_stp_filter_get, + }, + { + .str = "discard_egress_sll", + .getter = mlxsw_reg_ppcnt_egress_sll_get, + }, +}; + +#define MLXSW_SP_PORT_HW_DISCARD_STATS_LEN \ + ARRAY_SIZE(mlxsw_sp_port_hw_discard_stats) + +static struct mlxsw_sp_port_hw_stats mlxsw_sp_port_hw_prio_stats[] = { + { + .str = "rx_octets_prio", + .getter = mlxsw_reg_ppcnt_rx_octets_get, + }, + { + .str = "rx_frames_prio", + .getter = mlxsw_reg_ppcnt_rx_frames_get, + }, + { + .str = "tx_octets_prio", + .getter = mlxsw_reg_ppcnt_tx_octets_get, + }, + { + .str = "tx_frames_prio", + .getter = mlxsw_reg_ppcnt_tx_frames_get, + }, + { + .str = "rx_pause_prio", + .getter = mlxsw_reg_ppcnt_rx_pause_get, + }, + { + .str = "rx_pause_duration_prio", + .getter = mlxsw_reg_ppcnt_rx_pause_duration_get, + }, + { + .str = "tx_pause_prio", + .getter = mlxsw_reg_ppcnt_tx_pause_get, + }, + { + .str = "tx_pause_duration_prio", + .getter = mlxsw_reg_ppcnt_tx_pause_duration_get, + }, +}; + +#define MLXSW_SP_PORT_HW_PRIO_STATS_LEN ARRAY_SIZE(mlxsw_sp_port_hw_prio_stats) + +static struct mlxsw_sp_port_hw_stats mlxsw_sp_port_hw_tc_stats[] = { + { + .str = "tc_transmit_queue_tc", + .getter = mlxsw_reg_ppcnt_tc_transmit_queue_get, + .cells_bytes = true, + }, + { + .str = "tc_no_buffer_discard_uc_tc", + .getter = mlxsw_reg_ppcnt_tc_no_buffer_discard_uc_get, + }, +}; + +#define MLXSW_SP_PORT_HW_TC_STATS_LEN ARRAY_SIZE(mlxsw_sp_port_hw_tc_stats) + +struct mlxsw_sp_port_stats { + char str[ETH_GSTRING_LEN]; + u64 (*getter)(struct mlxsw_sp_port *mlxsw_sp_port); +}; + +static u64 +mlxsw_sp_port_get_transceiver_overheat_stats(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_core *mlxsw_core = mlxsw_sp_port->mlxsw_sp->core; + u8 slot_index = mlxsw_sp_port->mapping.slot_index; + u8 module = mlxsw_sp_port->mapping.module; + u64 stats; + int err; + + err = mlxsw_env_module_overheat_counter_get(mlxsw_core, slot_index, + module, &stats); + if (err) + return mlxsw_sp_port->module_overheat_initial_val; + + return stats - mlxsw_sp_port->module_overheat_initial_val; +} + +static struct mlxsw_sp_port_stats mlxsw_sp_port_transceiver_stats[] = { + { + .str = "transceiver_overheat", + .getter = mlxsw_sp_port_get_transceiver_overheat_stats, + }, +}; + +#define MLXSW_SP_PORT_HW_TRANSCEIVER_STATS_LEN ARRAY_SIZE(mlxsw_sp_port_transceiver_stats) + +#define MLXSW_SP_PORT_ETHTOOL_STATS_LEN (MLXSW_SP_PORT_HW_STATS_LEN + \ + MLXSW_SP_PORT_HW_RFC_2863_STATS_LEN + \ + MLXSW_SP_PORT_HW_RFC_2819_STATS_LEN + \ + MLXSW_SP_PORT_HW_RFC_3635_STATS_LEN + \ + MLXSW_SP_PORT_HW_EXT_STATS_LEN + \ + MLXSW_SP_PORT_HW_DISCARD_STATS_LEN + \ + (MLXSW_SP_PORT_HW_PRIO_STATS_LEN * \ + IEEE_8021QAZ_MAX_TCS) + \ + (MLXSW_SP_PORT_HW_TC_STATS_LEN * \ + TC_MAX_QUEUE) + \ + MLXSW_SP_PORT_HW_TRANSCEIVER_STATS_LEN) + +static void mlxsw_sp_port_get_prio_strings(u8 **p, int prio) +{ + int i; + + for (i = 0; i < MLXSW_SP_PORT_HW_PRIO_STATS_LEN; i++) { + snprintf(*p, ETH_GSTRING_LEN, "%.29s_%.1d", + mlxsw_sp_port_hw_prio_stats[i].str, prio); + *p += ETH_GSTRING_LEN; + } +} + +static void mlxsw_sp_port_get_tc_strings(u8 **p, int tc) +{ + int i; + + for (i = 0; i < MLXSW_SP_PORT_HW_TC_STATS_LEN; i++) { + snprintf(*p, ETH_GSTRING_LEN, "%.29s_%.1d", + mlxsw_sp_port_hw_tc_stats[i].str, tc); + *p += ETH_GSTRING_LEN; + } +} + +static void mlxsw_sp_port_get_strings(struct net_device *dev, + u32 stringset, u8 *data) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + u8 *p = data; + int i; + + switch (stringset) { + case ETH_SS_STATS: + for (i = 0; i < MLXSW_SP_PORT_HW_STATS_LEN; i++) { + memcpy(p, mlxsw_sp_port_hw_stats[i].str, + ETH_GSTRING_LEN); + p += ETH_GSTRING_LEN; + } + + for (i = 0; i < MLXSW_SP_PORT_HW_RFC_2863_STATS_LEN; i++) { + memcpy(p, mlxsw_sp_port_hw_rfc_2863_stats[i].str, + ETH_GSTRING_LEN); + p += ETH_GSTRING_LEN; + } + + for (i = 0; i < MLXSW_SP_PORT_HW_RFC_2819_STATS_LEN; i++) { + memcpy(p, mlxsw_sp_port_hw_rfc_2819_stats[i].str, + ETH_GSTRING_LEN); + p += ETH_GSTRING_LEN; + } + + for (i = 0; i < MLXSW_SP_PORT_HW_RFC_3635_STATS_LEN; i++) { + memcpy(p, mlxsw_sp_port_hw_rfc_3635_stats[i].str, + ETH_GSTRING_LEN); + p += ETH_GSTRING_LEN; + } + + for (i = 0; i < MLXSW_SP_PORT_HW_EXT_STATS_LEN; i++) { + memcpy(p, mlxsw_sp_port_hw_ext_stats[i].str, + ETH_GSTRING_LEN); + p += ETH_GSTRING_LEN; + } + + for (i = 0; i < MLXSW_SP_PORT_HW_DISCARD_STATS_LEN; i++) { + memcpy(p, mlxsw_sp_port_hw_discard_stats[i].str, + ETH_GSTRING_LEN); + p += ETH_GSTRING_LEN; + } + + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) + mlxsw_sp_port_get_prio_strings(&p, i); + + for (i = 0; i < TC_MAX_QUEUE; i++) + mlxsw_sp_port_get_tc_strings(&p, i); + + mlxsw_sp_port->mlxsw_sp->ptp_ops->get_stats_strings(&p); + + for (i = 0; i < MLXSW_SP_PORT_HW_TRANSCEIVER_STATS_LEN; i++) { + memcpy(p, mlxsw_sp_port_transceiver_stats[i].str, + ETH_GSTRING_LEN); + p += ETH_GSTRING_LEN; + } + break; + } +} + +static int mlxsw_sp_port_set_phys_id(struct net_device *dev, + enum ethtool_phys_id_state state) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char mlcr_pl[MLXSW_REG_MLCR_LEN]; + bool active; + + switch (state) { + case ETHTOOL_ID_ACTIVE: + active = true; + break; + case ETHTOOL_ID_INACTIVE: + active = false; + break; + default: + return -EOPNOTSUPP; + } + + mlxsw_reg_mlcr_pack(mlcr_pl, mlxsw_sp_port->local_port, active); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mlcr), mlcr_pl); +} + +static int +mlxsw_sp_get_hw_stats_by_group(struct mlxsw_sp_port_hw_stats **p_hw_stats, + int *p_len, enum mlxsw_reg_ppcnt_grp grp) +{ + switch (grp) { + case MLXSW_REG_PPCNT_IEEE_8023_CNT: + *p_hw_stats = mlxsw_sp_port_hw_stats; + *p_len = MLXSW_SP_PORT_HW_STATS_LEN; + break; + case MLXSW_REG_PPCNT_RFC_2863_CNT: + *p_hw_stats = mlxsw_sp_port_hw_rfc_2863_stats; + *p_len = MLXSW_SP_PORT_HW_RFC_2863_STATS_LEN; + break; + case MLXSW_REG_PPCNT_RFC_2819_CNT: + *p_hw_stats = mlxsw_sp_port_hw_rfc_2819_stats; + *p_len = MLXSW_SP_PORT_HW_RFC_2819_STATS_LEN; + break; + case MLXSW_REG_PPCNT_RFC_3635_CNT: + *p_hw_stats = mlxsw_sp_port_hw_rfc_3635_stats; + *p_len = MLXSW_SP_PORT_HW_RFC_3635_STATS_LEN; + break; + case MLXSW_REG_PPCNT_EXT_CNT: + *p_hw_stats = mlxsw_sp_port_hw_ext_stats; + *p_len = MLXSW_SP_PORT_HW_EXT_STATS_LEN; + break; + case MLXSW_REG_PPCNT_DISCARD_CNT: + *p_hw_stats = mlxsw_sp_port_hw_discard_stats; + *p_len = MLXSW_SP_PORT_HW_DISCARD_STATS_LEN; + break; + case MLXSW_REG_PPCNT_PRIO_CNT: + *p_hw_stats = mlxsw_sp_port_hw_prio_stats; + *p_len = MLXSW_SP_PORT_HW_PRIO_STATS_LEN; + break; + case MLXSW_REG_PPCNT_TC_CNT: + *p_hw_stats = mlxsw_sp_port_hw_tc_stats; + *p_len = MLXSW_SP_PORT_HW_TC_STATS_LEN; + break; + default: + WARN_ON(1); + return -EOPNOTSUPP; + } + return 0; +} + +static void __mlxsw_sp_port_get_stats(struct net_device *dev, + enum mlxsw_reg_ppcnt_grp grp, int prio, + u64 *data, int data_index) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_port_hw_stats *hw_stats; + char ppcnt_pl[MLXSW_REG_PPCNT_LEN]; + int i, len; + int err; + + err = mlxsw_sp_get_hw_stats_by_group(&hw_stats, &len, grp); + if (err) + return; + mlxsw_sp_port_get_stats_raw(dev, grp, prio, ppcnt_pl); + for (i = 0; i < len; i++) { + data[data_index + i] = hw_stats[i].getter(ppcnt_pl); + if (!hw_stats[i].cells_bytes) + continue; + data[data_index + i] = mlxsw_sp_cells_bytes(mlxsw_sp, + data[data_index + i]); + } +} + +static void __mlxsw_sp_port_get_env_stats(struct net_device *dev, u64 *data, int data_index, + struct mlxsw_sp_port_stats *port_stats, + int len) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + int i; + + for (i = 0; i < len; i++) + data[data_index + i] = port_stats[i].getter(mlxsw_sp_port); +} + +static void mlxsw_sp_port_get_stats(struct net_device *dev, + struct ethtool_stats *stats, u64 *data) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + int i, data_index = 0; + + /* IEEE 802.3 Counters */ + __mlxsw_sp_port_get_stats(dev, MLXSW_REG_PPCNT_IEEE_8023_CNT, 0, + data, data_index); + data_index = MLXSW_SP_PORT_HW_STATS_LEN; + + /* RFC 2863 Counters */ + __mlxsw_sp_port_get_stats(dev, MLXSW_REG_PPCNT_RFC_2863_CNT, 0, + data, data_index); + data_index += MLXSW_SP_PORT_HW_RFC_2863_STATS_LEN; + + /* RFC 2819 Counters */ + __mlxsw_sp_port_get_stats(dev, MLXSW_REG_PPCNT_RFC_2819_CNT, 0, + data, data_index); + data_index += MLXSW_SP_PORT_HW_RFC_2819_STATS_LEN; + + /* RFC 3635 Counters */ + __mlxsw_sp_port_get_stats(dev, MLXSW_REG_PPCNT_RFC_3635_CNT, 0, + data, data_index); + data_index += MLXSW_SP_PORT_HW_RFC_3635_STATS_LEN; + + /* Extended Counters */ + __mlxsw_sp_port_get_stats(dev, MLXSW_REG_PPCNT_EXT_CNT, 0, + data, data_index); + data_index += MLXSW_SP_PORT_HW_EXT_STATS_LEN; + + /* Discard Counters */ + __mlxsw_sp_port_get_stats(dev, MLXSW_REG_PPCNT_DISCARD_CNT, 0, + data, data_index); + data_index += MLXSW_SP_PORT_HW_DISCARD_STATS_LEN; + + /* Per-Priority Counters */ + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + __mlxsw_sp_port_get_stats(dev, MLXSW_REG_PPCNT_PRIO_CNT, i, + data, data_index); + data_index += MLXSW_SP_PORT_HW_PRIO_STATS_LEN; + } + + /* Per-TC Counters */ + for (i = 0; i < TC_MAX_QUEUE; i++) { + __mlxsw_sp_port_get_stats(dev, MLXSW_REG_PPCNT_TC_CNT, i, + data, data_index); + data_index += MLXSW_SP_PORT_HW_TC_STATS_LEN; + } + + /* PTP counters */ + mlxsw_sp_port->mlxsw_sp->ptp_ops->get_stats(mlxsw_sp_port, + data, data_index); + data_index += mlxsw_sp_port->mlxsw_sp->ptp_ops->get_stats_count(); + + /* Transceiver counters */ + __mlxsw_sp_port_get_env_stats(dev, data, data_index, mlxsw_sp_port_transceiver_stats, + MLXSW_SP_PORT_HW_TRANSCEIVER_STATS_LEN); + data_index += MLXSW_SP_PORT_HW_TRANSCEIVER_STATS_LEN; +} + +static int mlxsw_sp_port_get_sset_count(struct net_device *dev, int sset) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + + switch (sset) { + case ETH_SS_STATS: + return MLXSW_SP_PORT_ETHTOOL_STATS_LEN + + mlxsw_sp_port->mlxsw_sp->ptp_ops->get_stats_count(); + default: + return -EOPNOTSUPP; + } +} + +static void +mlxsw_sp_port_get_link_supported(struct mlxsw_sp *mlxsw_sp, u32 eth_proto_cap, + struct ethtool_link_ksettings *cmd) +{ + const struct mlxsw_sp_port_type_speed_ops *ops; + + ops = mlxsw_sp->port_type_speed_ops; + + ethtool_link_ksettings_add_link_mode(cmd, supported, Asym_Pause); + ethtool_link_ksettings_add_link_mode(cmd, supported, Autoneg); + ethtool_link_ksettings_add_link_mode(cmd, supported, Pause); + + ops->from_ptys_supported_port(mlxsw_sp, eth_proto_cap, cmd); + ops->from_ptys_link(mlxsw_sp, eth_proto_cap, + cmd->link_modes.supported); +} + +static void +mlxsw_sp_port_get_link_advertise(struct mlxsw_sp *mlxsw_sp, + u32 eth_proto_admin, bool autoneg, + struct ethtool_link_ksettings *cmd) +{ + const struct mlxsw_sp_port_type_speed_ops *ops; + + ops = mlxsw_sp->port_type_speed_ops; + + if (!autoneg) + return; + + ethtool_link_ksettings_add_link_mode(cmd, advertising, Autoneg); + ops->from_ptys_link(mlxsw_sp, eth_proto_admin, + cmd->link_modes.advertising); +} + +static u8 +mlxsw_sp_port_connector_port(enum mlxsw_reg_ptys_connector_type connector_type) +{ + switch (connector_type) { + case MLXSW_REG_PTYS_CONNECTOR_TYPE_UNKNOWN_OR_NO_CONNECTOR: + return PORT_OTHER; + case MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_NONE: + return PORT_NONE; + case MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_TP: + return PORT_TP; + case MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_AUI: + return PORT_AUI; + case MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_BNC: + return PORT_BNC; + case MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_MII: + return PORT_MII; + case MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_FIBRE: + return PORT_FIBRE; + case MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_DA: + return PORT_DA; + case MLXSW_REG_PTYS_CONNECTOR_TYPE_PORT_OTHER: + return PORT_OTHER; + default: + WARN_ON_ONCE(1); + return PORT_OTHER; + } +} + +static int mlxsw_sp_port_ptys_query(struct mlxsw_sp_port *mlxsw_sp_port, + u32 *p_eth_proto_cap, u32 *p_eth_proto_admin, + u32 *p_eth_proto_oper, u8 *p_connector_type) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + const struct mlxsw_sp_port_type_speed_ops *ops; + char ptys_pl[MLXSW_REG_PTYS_LEN]; + int err; + + ops = mlxsw_sp->port_type_speed_ops; + + ops->reg_ptys_eth_pack(mlxsw_sp, ptys_pl, mlxsw_sp_port->local_port, 0, false); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ptys), ptys_pl); + if (err) + return err; + + ops->reg_ptys_eth_unpack(mlxsw_sp, ptys_pl, p_eth_proto_cap, p_eth_proto_admin, + p_eth_proto_oper); + if (p_connector_type) + *p_connector_type = mlxsw_reg_ptys_connector_type_get(ptys_pl); + return 0; +} + +static int mlxsw_sp_port_get_link_ksettings(struct net_device *dev, + struct ethtool_link_ksettings *cmd) +{ + u32 eth_proto_cap, eth_proto_admin, eth_proto_oper; + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + const struct mlxsw_sp_port_type_speed_ops *ops; + u8 connector_type; + bool autoneg; + int err; + + err = mlxsw_sp_port_ptys_query(mlxsw_sp_port, ð_proto_cap, ð_proto_admin, + ð_proto_oper, &connector_type); + if (err) + return err; + + ops = mlxsw_sp->port_type_speed_ops; + autoneg = mlxsw_sp_port->link.autoneg; + + mlxsw_sp_port_get_link_supported(mlxsw_sp, eth_proto_cap, cmd); + + mlxsw_sp_port_get_link_advertise(mlxsw_sp, eth_proto_admin, autoneg, cmd); + + cmd->base.autoneg = autoneg ? AUTONEG_ENABLE : AUTONEG_DISABLE; + cmd->base.port = mlxsw_sp_port_connector_port(connector_type); + ops->from_ptys_link_mode(mlxsw_sp, netif_carrier_ok(dev), + eth_proto_oper, cmd); + + return 0; +} + +static int +mlxsw_sp_port_set_link_ksettings(struct net_device *dev, + const struct ethtool_link_ksettings *cmd) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + const struct mlxsw_sp_port_type_speed_ops *ops; + char ptys_pl[MLXSW_REG_PTYS_LEN]; + u32 eth_proto_cap, eth_proto_new; + bool autoneg; + int err; + + ops = mlxsw_sp->port_type_speed_ops; + + ops->reg_ptys_eth_pack(mlxsw_sp, ptys_pl, mlxsw_sp_port->local_port, + 0, false); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ptys), ptys_pl); + if (err) + return err; + ops->reg_ptys_eth_unpack(mlxsw_sp, ptys_pl, ð_proto_cap, NULL, NULL); + + autoneg = cmd->base.autoneg == AUTONEG_ENABLE; + eth_proto_new = autoneg ? + ops->to_ptys_advert_link(mlxsw_sp, cmd) : + ops->to_ptys_speed_lanes(mlxsw_sp, mlxsw_sp_port->mapping.width, + cmd); + + eth_proto_new = eth_proto_new & eth_proto_cap; + if (!eth_proto_new) { + netdev_err(dev, "No supported speed or lanes requested\n"); + return -EINVAL; + } + + ops->reg_ptys_eth_pack(mlxsw_sp, ptys_pl, mlxsw_sp_port->local_port, + eth_proto_new, autoneg); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptys), ptys_pl); + if (err) + return err; + + mlxsw_sp_port->link.autoneg = autoneg; + + if (!netif_running(dev)) + return 0; + + mlxsw_sp_port_admin_status_set(mlxsw_sp_port, false); + mlxsw_sp_port_admin_status_set(mlxsw_sp_port, true); + + return 0; +} + +static int mlxsw_sp_get_module_info(struct net_device *netdev, + struct ethtool_modinfo *modinfo) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(netdev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + + return mlxsw_env_get_module_info(netdev, mlxsw_sp->core, + mlxsw_sp_port->mapping.slot_index, + mlxsw_sp_port->mapping.module, + modinfo); +} + +static int mlxsw_sp_get_module_eeprom(struct net_device *netdev, + struct ethtool_eeprom *ee, u8 *data) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(netdev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u8 slot_index = mlxsw_sp_port->mapping.slot_index; + u8 module = mlxsw_sp_port->mapping.module; + + return mlxsw_env_get_module_eeprom(netdev, mlxsw_sp->core, slot_index, + module, ee, data); +} + +static int +mlxsw_sp_get_module_eeprom_by_page(struct net_device *dev, + const struct ethtool_module_eeprom *page, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u8 slot_index = mlxsw_sp_port->mapping.slot_index; + u8 module = mlxsw_sp_port->mapping.module; + + return mlxsw_env_get_module_eeprom_by_page(mlxsw_sp->core, slot_index, + module, page, extack); +} + +static int +mlxsw_sp_get_ts_info(struct net_device *netdev, struct ethtool_ts_info *info) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(netdev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + + return mlxsw_sp->ptp_ops->get_ts_info(mlxsw_sp, info); +} + +static void +mlxsw_sp_get_eth_phy_stats(struct net_device *dev, + struct ethtool_eth_phy_stats *phy_stats) +{ + char ppcnt_pl[MLXSW_REG_PPCNT_LEN]; + + if (mlxsw_sp_port_get_stats_raw(dev, MLXSW_REG_PPCNT_IEEE_8023_CNT, + 0, ppcnt_pl)) + return; + + phy_stats->SymbolErrorDuringCarrier = + mlxsw_reg_ppcnt_a_symbol_error_during_carrier_get(ppcnt_pl); +} + +static void +mlxsw_sp_get_eth_mac_stats(struct net_device *dev, + struct ethtool_eth_mac_stats *mac_stats) +{ + char ppcnt_pl[MLXSW_REG_PPCNT_LEN]; + + if (mlxsw_sp_port_get_stats_raw(dev, MLXSW_REG_PPCNT_IEEE_8023_CNT, + 0, ppcnt_pl)) + return; + + mac_stats->FramesTransmittedOK = + mlxsw_reg_ppcnt_a_frames_transmitted_ok_get(ppcnt_pl); + mac_stats->FramesReceivedOK = + mlxsw_reg_ppcnt_a_frames_received_ok_get(ppcnt_pl); + mac_stats->FrameCheckSequenceErrors = + mlxsw_reg_ppcnt_a_frame_check_sequence_errors_get(ppcnt_pl); + mac_stats->AlignmentErrors = + mlxsw_reg_ppcnt_a_alignment_errors_get(ppcnt_pl); + mac_stats->OctetsTransmittedOK = + mlxsw_reg_ppcnt_a_octets_transmitted_ok_get(ppcnt_pl); + mac_stats->OctetsReceivedOK = + mlxsw_reg_ppcnt_a_octets_received_ok_get(ppcnt_pl); + mac_stats->MulticastFramesXmittedOK = + mlxsw_reg_ppcnt_a_multicast_frames_xmitted_ok_get(ppcnt_pl); + mac_stats->BroadcastFramesXmittedOK = + mlxsw_reg_ppcnt_a_broadcast_frames_xmitted_ok_get(ppcnt_pl); + mac_stats->MulticastFramesReceivedOK = + mlxsw_reg_ppcnt_a_multicast_frames_received_ok_get(ppcnt_pl); + mac_stats->BroadcastFramesReceivedOK = + mlxsw_reg_ppcnt_a_broadcast_frames_received_ok_get(ppcnt_pl); + mac_stats->InRangeLengthErrors = + mlxsw_reg_ppcnt_a_in_range_length_errors_get(ppcnt_pl); + mac_stats->OutOfRangeLengthField = + mlxsw_reg_ppcnt_a_out_of_range_length_field_get(ppcnt_pl); + mac_stats->FrameTooLongErrors = + mlxsw_reg_ppcnt_a_frame_too_long_errors_get(ppcnt_pl); +} + +static void +mlxsw_sp_get_eth_ctrl_stats(struct net_device *dev, + struct ethtool_eth_ctrl_stats *ctrl_stats) +{ + char ppcnt_pl[MLXSW_REG_PPCNT_LEN]; + + if (mlxsw_sp_port_get_stats_raw(dev, MLXSW_REG_PPCNT_IEEE_8023_CNT, + 0, ppcnt_pl)) + return; + + ctrl_stats->MACControlFramesTransmitted = + mlxsw_reg_ppcnt_a_mac_control_frames_transmitted_get(ppcnt_pl); + ctrl_stats->MACControlFramesReceived = + mlxsw_reg_ppcnt_a_mac_control_frames_received_get(ppcnt_pl); + ctrl_stats->UnsupportedOpcodesReceived = + mlxsw_reg_ppcnt_a_unsupported_opcodes_received_get(ppcnt_pl); +} + +static const struct ethtool_rmon_hist_range mlxsw_rmon_ranges[] = { + { 0, 64 }, + { 65, 127 }, + { 128, 255 }, + { 256, 511 }, + { 512, 1023 }, + { 1024, 1518 }, + { 1519, 2047 }, + { 2048, 4095 }, + { 4096, 8191 }, + { 8192, 10239 }, + {} +}; + +static void +mlxsw_sp_get_rmon_stats(struct net_device *dev, + struct ethtool_rmon_stats *rmon, + const struct ethtool_rmon_hist_range **ranges) +{ + char ppcnt_pl[MLXSW_REG_PPCNT_LEN]; + + if (mlxsw_sp_port_get_stats_raw(dev, MLXSW_REG_PPCNT_RFC_2819_CNT, + 0, ppcnt_pl)) + return; + + rmon->undersize_pkts = + mlxsw_reg_ppcnt_ether_stats_undersize_pkts_get(ppcnt_pl); + rmon->oversize_pkts = + mlxsw_reg_ppcnt_ether_stats_oversize_pkts_get(ppcnt_pl); + rmon->fragments = + mlxsw_reg_ppcnt_ether_stats_fragments_get(ppcnt_pl); + + rmon->hist[0] = mlxsw_reg_ppcnt_ether_stats_pkts64octets_get(ppcnt_pl); + rmon->hist[1] = + mlxsw_reg_ppcnt_ether_stats_pkts65to127octets_get(ppcnt_pl); + rmon->hist[2] = + mlxsw_reg_ppcnt_ether_stats_pkts128to255octets_get(ppcnt_pl); + rmon->hist[3] = + mlxsw_reg_ppcnt_ether_stats_pkts256to511octets_get(ppcnt_pl); + rmon->hist[4] = + mlxsw_reg_ppcnt_ether_stats_pkts512to1023octets_get(ppcnt_pl); + rmon->hist[5] = + mlxsw_reg_ppcnt_ether_stats_pkts1024to1518octets_get(ppcnt_pl); + rmon->hist[6] = + mlxsw_reg_ppcnt_ether_stats_pkts1519to2047octets_get(ppcnt_pl); + rmon->hist[7] = + mlxsw_reg_ppcnt_ether_stats_pkts2048to4095octets_get(ppcnt_pl); + rmon->hist[8] = + mlxsw_reg_ppcnt_ether_stats_pkts4096to8191octets_get(ppcnt_pl); + rmon->hist[9] = + mlxsw_reg_ppcnt_ether_stats_pkts8192to10239octets_get(ppcnt_pl); + + *ranges = mlxsw_rmon_ranges; +} + +static int mlxsw_sp_reset(struct net_device *dev, u32 *flags) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u8 slot_index = mlxsw_sp_port->mapping.slot_index; + u8 module = mlxsw_sp_port->mapping.module; + + return mlxsw_env_reset_module(dev, mlxsw_sp->core, slot_index, + module, flags); +} + +static int +mlxsw_sp_get_module_power_mode(struct net_device *dev, + struct ethtool_module_power_mode_params *params, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u8 slot_index = mlxsw_sp_port->mapping.slot_index; + u8 module = mlxsw_sp_port->mapping.module; + + return mlxsw_env_get_module_power_mode(mlxsw_sp->core, slot_index, + module, params, extack); +} + +static int +mlxsw_sp_set_module_power_mode(struct net_device *dev, + const struct ethtool_module_power_mode_params *params, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u8 slot_index = mlxsw_sp_port->mapping.slot_index; + u8 module = mlxsw_sp_port->mapping.module; + + return mlxsw_env_set_module_power_mode(mlxsw_sp->core, slot_index, + module, params->policy, extack); +} + +const struct ethtool_ops mlxsw_sp_port_ethtool_ops = { + .cap_link_lanes_supported = true, + .get_drvinfo = mlxsw_sp_port_get_drvinfo, + .get_link = ethtool_op_get_link, + .get_link_ext_state = mlxsw_sp_port_get_link_ext_state, + .get_pauseparam = mlxsw_sp_port_get_pauseparam, + .set_pauseparam = mlxsw_sp_port_set_pauseparam, + .get_strings = mlxsw_sp_port_get_strings, + .set_phys_id = mlxsw_sp_port_set_phys_id, + .get_ethtool_stats = mlxsw_sp_port_get_stats, + .get_sset_count = mlxsw_sp_port_get_sset_count, + .get_link_ksettings = mlxsw_sp_port_get_link_ksettings, + .set_link_ksettings = mlxsw_sp_port_set_link_ksettings, + .get_module_info = mlxsw_sp_get_module_info, + .get_module_eeprom = mlxsw_sp_get_module_eeprom, + .get_module_eeprom_by_page = mlxsw_sp_get_module_eeprom_by_page, + .get_ts_info = mlxsw_sp_get_ts_info, + .get_eth_phy_stats = mlxsw_sp_get_eth_phy_stats, + .get_eth_mac_stats = mlxsw_sp_get_eth_mac_stats, + .get_eth_ctrl_stats = mlxsw_sp_get_eth_ctrl_stats, + .get_rmon_stats = mlxsw_sp_get_rmon_stats, + .reset = mlxsw_sp_reset, + .get_module_power_mode = mlxsw_sp_get_module_power_mode, + .set_module_power_mode = mlxsw_sp_set_module_power_mode, +}; + +struct mlxsw_sp1_port_link_mode { + enum ethtool_link_mode_bit_indices mask_ethtool; + u32 mask; + u32 speed; +}; + +static const struct mlxsw_sp1_port_link_mode mlxsw_sp1_port_link_mode[] = { + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_100BASE_T, + .mask_ethtool = ETHTOOL_LINK_MODE_100baseT_Full_BIT, + .speed = SPEED_100, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_SGMII | + MLXSW_REG_PTYS_ETH_SPEED_1000BASE_KX, + .mask_ethtool = ETHTOOL_LINK_MODE_1000baseKX_Full_BIT, + .speed = SPEED_1000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_1000BASE_T, + .mask_ethtool = ETHTOOL_LINK_MODE_1000baseT_Full_BIT, + .speed = SPEED_1000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_10GBASE_CX4 | + MLXSW_REG_PTYS_ETH_SPEED_10GBASE_KX4, + .mask_ethtool = ETHTOOL_LINK_MODE_10000baseKX4_Full_BIT, + .speed = SPEED_10000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_10GBASE_KR | + MLXSW_REG_PTYS_ETH_SPEED_10GBASE_CR | + MLXSW_REG_PTYS_ETH_SPEED_10GBASE_SR | + MLXSW_REG_PTYS_ETH_SPEED_10GBASE_ER_LR, + .mask_ethtool = ETHTOOL_LINK_MODE_10000baseKR_Full_BIT, + .speed = SPEED_10000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_40GBASE_CR4, + .mask_ethtool = ETHTOOL_LINK_MODE_40000baseCR4_Full_BIT, + .speed = SPEED_40000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_40GBASE_KR4, + .mask_ethtool = ETHTOOL_LINK_MODE_40000baseKR4_Full_BIT, + .speed = SPEED_40000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_40GBASE_SR4, + .mask_ethtool = ETHTOOL_LINK_MODE_40000baseSR4_Full_BIT, + .speed = SPEED_40000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_40GBASE_LR4_ER4, + .mask_ethtool = ETHTOOL_LINK_MODE_40000baseLR4_Full_BIT, + .speed = SPEED_40000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_25GBASE_CR, + .mask_ethtool = ETHTOOL_LINK_MODE_25000baseCR_Full_BIT, + .speed = SPEED_25000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_25GBASE_KR, + .mask_ethtool = ETHTOOL_LINK_MODE_25000baseKR_Full_BIT, + .speed = SPEED_25000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_25GBASE_SR, + .mask_ethtool = ETHTOOL_LINK_MODE_25000baseSR_Full_BIT, + .speed = SPEED_25000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_50GBASE_CR2, + .mask_ethtool = ETHTOOL_LINK_MODE_50000baseCR2_Full_BIT, + .speed = SPEED_50000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_50GBASE_KR2, + .mask_ethtool = ETHTOOL_LINK_MODE_50000baseKR2_Full_BIT, + .speed = SPEED_50000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_50GBASE_SR2, + .mask_ethtool = ETHTOOL_LINK_MODE_50000baseSR2_Full_BIT, + .speed = SPEED_50000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_100GBASE_CR4, + .mask_ethtool = ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT, + .speed = SPEED_100000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_100GBASE_SR4, + .mask_ethtool = ETHTOOL_LINK_MODE_100000baseSR4_Full_BIT, + .speed = SPEED_100000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_100GBASE_KR4, + .mask_ethtool = ETHTOOL_LINK_MODE_100000baseKR4_Full_BIT, + .speed = SPEED_100000, + }, + { + .mask = MLXSW_REG_PTYS_ETH_SPEED_100GBASE_LR4_ER4, + .mask_ethtool = ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT, + .speed = SPEED_100000, + }, +}; + +#define MLXSW_SP1_PORT_LINK_MODE_LEN ARRAY_SIZE(mlxsw_sp1_port_link_mode) + +static void +mlxsw_sp1_from_ptys_supported_port(struct mlxsw_sp *mlxsw_sp, + u32 ptys_eth_proto, + struct ethtool_link_ksettings *cmd) +{ + if (ptys_eth_proto & (MLXSW_REG_PTYS_ETH_SPEED_10GBASE_CR | + MLXSW_REG_PTYS_ETH_SPEED_10GBASE_SR | + MLXSW_REG_PTYS_ETH_SPEED_40GBASE_CR4 | + MLXSW_REG_PTYS_ETH_SPEED_40GBASE_SR4 | + MLXSW_REG_PTYS_ETH_SPEED_100GBASE_SR4 | + MLXSW_REG_PTYS_ETH_SPEED_SGMII)) + ethtool_link_ksettings_add_link_mode(cmd, supported, FIBRE); + + if (ptys_eth_proto & (MLXSW_REG_PTYS_ETH_SPEED_10GBASE_KR | + MLXSW_REG_PTYS_ETH_SPEED_10GBASE_KX4 | + MLXSW_REG_PTYS_ETH_SPEED_40GBASE_KR4 | + MLXSW_REG_PTYS_ETH_SPEED_100GBASE_KR4 | + MLXSW_REG_PTYS_ETH_SPEED_1000BASE_KX)) + ethtool_link_ksettings_add_link_mode(cmd, supported, Backplane); +} + +static void +mlxsw_sp1_from_ptys_link(struct mlxsw_sp *mlxsw_sp, u32 ptys_eth_proto, + unsigned long *mode) +{ + int i; + + for (i = 0; i < MLXSW_SP1_PORT_LINK_MODE_LEN; i++) { + if (ptys_eth_proto & mlxsw_sp1_port_link_mode[i].mask) + __set_bit(mlxsw_sp1_port_link_mode[i].mask_ethtool, + mode); + } +} + +static u32 +mlxsw_sp1_from_ptys_speed(struct mlxsw_sp *mlxsw_sp, u32 ptys_eth_proto) +{ + int i; + + for (i = 0; i < MLXSW_SP1_PORT_LINK_MODE_LEN; i++) { + if (ptys_eth_proto & mlxsw_sp1_port_link_mode[i].mask) + return mlxsw_sp1_port_link_mode[i].speed; + } + + return SPEED_UNKNOWN; +} + +static void +mlxsw_sp1_from_ptys_link_mode(struct mlxsw_sp *mlxsw_sp, bool carrier_ok, + u32 ptys_eth_proto, + struct ethtool_link_ksettings *cmd) +{ + struct mlxsw_sp1_port_link_mode link; + int i; + + cmd->base.speed = SPEED_UNKNOWN; + cmd->base.duplex = DUPLEX_UNKNOWN; + cmd->lanes = 0; + + if (!carrier_ok) + return; + + for (i = 0; i < MLXSW_SP1_PORT_LINK_MODE_LEN; i++) { + if (ptys_eth_proto & mlxsw_sp1_port_link_mode[i].mask) { + link = mlxsw_sp1_port_link_mode[i]; + ethtool_params_from_link_mode(cmd, + link.mask_ethtool); + } + } +} + +static int mlxsw_sp1_ptys_max_speed(struct mlxsw_sp_port *mlxsw_sp_port, u32 *p_max_speed) +{ + u32 eth_proto_cap; + u32 max_speed = 0; + int err; + int i; + + err = mlxsw_sp_port_ptys_query(mlxsw_sp_port, ð_proto_cap, NULL, NULL, NULL); + if (err) + return err; + + for (i = 0; i < MLXSW_SP1_PORT_LINK_MODE_LEN; i++) { + if ((eth_proto_cap & mlxsw_sp1_port_link_mode[i].mask) && + mlxsw_sp1_port_link_mode[i].speed > max_speed) + max_speed = mlxsw_sp1_port_link_mode[i].speed; + } + + *p_max_speed = max_speed; + return 0; +} + +static u32 +mlxsw_sp1_to_ptys_advert_link(struct mlxsw_sp *mlxsw_sp, + const struct ethtool_link_ksettings *cmd) +{ + u32 ptys_proto = 0; + int i; + + for (i = 0; i < MLXSW_SP1_PORT_LINK_MODE_LEN; i++) { + if (test_bit(mlxsw_sp1_port_link_mode[i].mask_ethtool, + cmd->link_modes.advertising)) + ptys_proto |= mlxsw_sp1_port_link_mode[i].mask; + } + return ptys_proto; +} + +static u32 mlxsw_sp1_to_ptys_speed_lanes(struct mlxsw_sp *mlxsw_sp, u8 width, + const struct ethtool_link_ksettings *cmd) +{ + u32 ptys_proto = 0; + int i; + + if (cmd->lanes > width) + return ptys_proto; + + for (i = 0; i < MLXSW_SP1_PORT_LINK_MODE_LEN; i++) { + if (cmd->base.speed == mlxsw_sp1_port_link_mode[i].speed) + ptys_proto |= mlxsw_sp1_port_link_mode[i].mask; + } + return ptys_proto; +} + +static void +mlxsw_sp1_reg_ptys_eth_pack(struct mlxsw_sp *mlxsw_sp, char *payload, + u16 local_port, u32 proto_admin, bool autoneg) +{ + mlxsw_reg_ptys_eth_pack(payload, local_port, proto_admin, autoneg); +} + +static void +mlxsw_sp1_reg_ptys_eth_unpack(struct mlxsw_sp *mlxsw_sp, char *payload, + u32 *p_eth_proto_cap, u32 *p_eth_proto_admin, + u32 *p_eth_proto_oper) +{ + mlxsw_reg_ptys_eth_unpack(payload, p_eth_proto_cap, p_eth_proto_admin, + p_eth_proto_oper); +} + +static u32 mlxsw_sp1_ptys_proto_cap_masked_get(u32 eth_proto_cap) +{ + u32 ptys_proto_cap_masked = 0; + int i; + + for (i = 0; i < MLXSW_SP1_PORT_LINK_MODE_LEN; i++) { + if (mlxsw_sp1_port_link_mode[i].mask & eth_proto_cap) + ptys_proto_cap_masked |= + mlxsw_sp1_port_link_mode[i].mask; + } + + return ptys_proto_cap_masked; +} + +const struct mlxsw_sp_port_type_speed_ops mlxsw_sp1_port_type_speed_ops = { + .from_ptys_supported_port = mlxsw_sp1_from_ptys_supported_port, + .from_ptys_link = mlxsw_sp1_from_ptys_link, + .from_ptys_speed = mlxsw_sp1_from_ptys_speed, + .from_ptys_link_mode = mlxsw_sp1_from_ptys_link_mode, + .ptys_max_speed = mlxsw_sp1_ptys_max_speed, + .to_ptys_advert_link = mlxsw_sp1_to_ptys_advert_link, + .to_ptys_speed_lanes = mlxsw_sp1_to_ptys_speed_lanes, + .reg_ptys_eth_pack = mlxsw_sp1_reg_ptys_eth_pack, + .reg_ptys_eth_unpack = mlxsw_sp1_reg_ptys_eth_unpack, + .ptys_proto_cap_masked_get = mlxsw_sp1_ptys_proto_cap_masked_get, +}; + +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_sgmii_100m[] = { + ETHTOOL_LINK_MODE_100baseT_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_SGMII_100M_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_sgmii_100m) + +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_1000base_x_sgmii[] = { + ETHTOOL_LINK_MODE_1000baseT_Full_BIT, + ETHTOOL_LINK_MODE_1000baseKX_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_1000BASE_X_SGMII_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_1000base_x_sgmii) + +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_5gbase_r[] = { + ETHTOOL_LINK_MODE_5000baseT_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_5GBASE_R_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_5gbase_r) + +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_xfi_xaui_1_10g[] = { + ETHTOOL_LINK_MODE_10000baseT_Full_BIT, + ETHTOOL_LINK_MODE_10000baseKR_Full_BIT, + ETHTOOL_LINK_MODE_10000baseR_FEC_BIT, + ETHTOOL_LINK_MODE_10000baseCR_Full_BIT, + ETHTOOL_LINK_MODE_10000baseSR_Full_BIT, + ETHTOOL_LINK_MODE_10000baseLR_Full_BIT, + ETHTOOL_LINK_MODE_10000baseER_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_XFI_XAUI_1_10G_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_xfi_xaui_1_10g) + +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_xlaui_4_xlppi_4_40g[] = { + ETHTOOL_LINK_MODE_40000baseKR4_Full_BIT, + ETHTOOL_LINK_MODE_40000baseCR4_Full_BIT, + ETHTOOL_LINK_MODE_40000baseSR4_Full_BIT, + ETHTOOL_LINK_MODE_40000baseLR4_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_XLAUI_4_XLPPI_4_40G_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_xlaui_4_xlppi_4_40g) + +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_25gaui_1_25gbase_cr_kr[] = { + ETHTOOL_LINK_MODE_25000baseCR_Full_BIT, + ETHTOOL_LINK_MODE_25000baseKR_Full_BIT, + ETHTOOL_LINK_MODE_25000baseSR_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_25GAUI_1_25GBASE_CR_KR_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_25gaui_1_25gbase_cr_kr) + +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_50gaui_2_laui_2_50gbase_cr2_kr2[] = { + ETHTOOL_LINK_MODE_50000baseCR2_Full_BIT, + ETHTOOL_LINK_MODE_50000baseKR2_Full_BIT, + ETHTOOL_LINK_MODE_50000baseSR2_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_50GAUI_2_LAUI_2_50GBASE_CR2_KR2_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_50gaui_2_laui_2_50gbase_cr2_kr2) + +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_50gaui_1_laui_1_50gbase_cr_kr[] = { + ETHTOOL_LINK_MODE_50000baseKR_Full_BIT, + ETHTOOL_LINK_MODE_50000baseSR_Full_BIT, + ETHTOOL_LINK_MODE_50000baseCR_Full_BIT, + ETHTOOL_LINK_MODE_50000baseLR_ER_FR_Full_BIT, + ETHTOOL_LINK_MODE_50000baseDR_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_50GAUI_1_LAUI_1_50GBASE_CR_KR_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_50gaui_1_laui_1_50gbase_cr_kr) + +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_caui_4_100gbase_cr4_kr4[] = { + ETHTOOL_LINK_MODE_100000baseKR4_Full_BIT, + ETHTOOL_LINK_MODE_100000baseSR4_Full_BIT, + ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT, + ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_CAUI_4_100GBASE_CR4_KR4_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_caui_4_100gbase_cr4_kr4) + +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_100gaui_2_100gbase_cr2_kr2[] = { + ETHTOOL_LINK_MODE_100000baseKR2_Full_BIT, + ETHTOOL_LINK_MODE_100000baseSR2_Full_BIT, + ETHTOOL_LINK_MODE_100000baseCR2_Full_BIT, + ETHTOOL_LINK_MODE_100000baseLR2_ER2_FR2_Full_BIT, + ETHTOOL_LINK_MODE_100000baseDR2_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_100GAUI_2_100GBASE_CR2_KR2_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_100gaui_2_100gbase_cr2_kr2) + +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_200gaui_4_200gbase_cr4_kr4[] = { + ETHTOOL_LINK_MODE_200000baseKR4_Full_BIT, + ETHTOOL_LINK_MODE_200000baseSR4_Full_BIT, + ETHTOOL_LINK_MODE_200000baseLR4_ER4_FR4_Full_BIT, + ETHTOOL_LINK_MODE_200000baseDR4_Full_BIT, + ETHTOOL_LINK_MODE_200000baseCR4_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_200GAUI_4_200GBASE_CR4_KR4_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_200gaui_4_200gbase_cr4_kr4) + +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_400gaui_8[] = { + ETHTOOL_LINK_MODE_400000baseKR8_Full_BIT, + ETHTOOL_LINK_MODE_400000baseSR8_Full_BIT, + ETHTOOL_LINK_MODE_400000baseLR8_ER8_FR8_Full_BIT, + ETHTOOL_LINK_MODE_400000baseDR8_Full_BIT, + ETHTOOL_LINK_MODE_400000baseCR8_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_400GAUI_8_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_400gaui_8) + +#define MLXSW_SP_PORT_MASK_WIDTH_1X BIT(0) +#define MLXSW_SP_PORT_MASK_WIDTH_2X BIT(1) +#define MLXSW_SP_PORT_MASK_WIDTH_4X BIT(2) +#define MLXSW_SP_PORT_MASK_WIDTH_8X BIT(3) + +static u8 mlxsw_sp_port_mask_width_get(u8 width) +{ + switch (width) { + case 1: + return MLXSW_SP_PORT_MASK_WIDTH_1X; + case 2: + return MLXSW_SP_PORT_MASK_WIDTH_2X; + case 4: + return MLXSW_SP_PORT_MASK_WIDTH_4X; + case 8: + return MLXSW_SP_PORT_MASK_WIDTH_8X; + default: + WARN_ON_ONCE(1); + return 0; + } +} + +struct mlxsw_sp2_port_link_mode { + const enum ethtool_link_mode_bit_indices *mask_ethtool; + int m_ethtool_len; + u32 mask; + u32 speed; + u32 width; + u8 mask_sup_width; +}; + +static const struct mlxsw_sp2_port_link_mode mlxsw_sp2_port_link_mode[] = { + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_SGMII_100M, + .mask_ethtool = mlxsw_sp2_mask_ethtool_sgmii_100m, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_SGMII_100M_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_1X | + MLXSW_SP_PORT_MASK_WIDTH_2X | + MLXSW_SP_PORT_MASK_WIDTH_4X | + MLXSW_SP_PORT_MASK_WIDTH_8X, + .speed = SPEED_100, + .width = 1, + }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_1000BASE_X_SGMII, + .mask_ethtool = mlxsw_sp2_mask_ethtool_1000base_x_sgmii, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_1000BASE_X_SGMII_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_1X | + MLXSW_SP_PORT_MASK_WIDTH_2X | + MLXSW_SP_PORT_MASK_WIDTH_4X | + MLXSW_SP_PORT_MASK_WIDTH_8X, + .speed = SPEED_1000, + .width = 1, + }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_5GBASE_R, + .mask_ethtool = mlxsw_sp2_mask_ethtool_5gbase_r, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_5GBASE_R_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_1X | + MLXSW_SP_PORT_MASK_WIDTH_2X | + MLXSW_SP_PORT_MASK_WIDTH_4X | + MLXSW_SP_PORT_MASK_WIDTH_8X, + .speed = SPEED_5000, + .width = 1, + }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_XFI_XAUI_1_10G, + .mask_ethtool = mlxsw_sp2_mask_ethtool_xfi_xaui_1_10g, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_XFI_XAUI_1_10G_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_1X | + MLXSW_SP_PORT_MASK_WIDTH_2X | + MLXSW_SP_PORT_MASK_WIDTH_4X | + MLXSW_SP_PORT_MASK_WIDTH_8X, + .speed = SPEED_10000, + .width = 1, + }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_XLAUI_4_XLPPI_4_40G, + .mask_ethtool = mlxsw_sp2_mask_ethtool_xlaui_4_xlppi_4_40g, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_XLAUI_4_XLPPI_4_40G_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_4X | + MLXSW_SP_PORT_MASK_WIDTH_8X, + .speed = SPEED_40000, + .width = 4, + }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_25GAUI_1_25GBASE_CR_KR, + .mask_ethtool = mlxsw_sp2_mask_ethtool_25gaui_1_25gbase_cr_kr, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_25GAUI_1_25GBASE_CR_KR_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_1X | + MLXSW_SP_PORT_MASK_WIDTH_2X | + MLXSW_SP_PORT_MASK_WIDTH_4X | + MLXSW_SP_PORT_MASK_WIDTH_8X, + .speed = SPEED_25000, + .width = 1, + }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_50GAUI_2_LAUI_2_50GBASE_CR2_KR2, + .mask_ethtool = mlxsw_sp2_mask_ethtool_50gaui_2_laui_2_50gbase_cr2_kr2, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_50GAUI_2_LAUI_2_50GBASE_CR2_KR2_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_2X | + MLXSW_SP_PORT_MASK_WIDTH_4X | + MLXSW_SP_PORT_MASK_WIDTH_8X, + .speed = SPEED_50000, + .width = 2, + }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_50GAUI_1_LAUI_1_50GBASE_CR_KR, + .mask_ethtool = mlxsw_sp2_mask_ethtool_50gaui_1_laui_1_50gbase_cr_kr, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_50GAUI_1_LAUI_1_50GBASE_CR_KR_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_1X, + .speed = SPEED_50000, + .width = 1, + }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_CAUI_4_100GBASE_CR4_KR4, + .mask_ethtool = mlxsw_sp2_mask_ethtool_caui_4_100gbase_cr4_kr4, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_CAUI_4_100GBASE_CR4_KR4_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_4X | + MLXSW_SP_PORT_MASK_WIDTH_8X, + .speed = SPEED_100000, + .width = 4, + }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_100GAUI_2_100GBASE_CR2_KR2, + .mask_ethtool = mlxsw_sp2_mask_ethtool_100gaui_2_100gbase_cr2_kr2, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_100GAUI_2_100GBASE_CR2_KR2_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_2X, + .speed = SPEED_100000, + .width = 2, + }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_200GAUI_4_200GBASE_CR4_KR4, + .mask_ethtool = mlxsw_sp2_mask_ethtool_200gaui_4_200gbase_cr4_kr4, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_200GAUI_4_200GBASE_CR4_KR4_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_4X | + MLXSW_SP_PORT_MASK_WIDTH_8X, + .speed = SPEED_200000, + .width = 4, + }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_400GAUI_8, + .mask_ethtool = mlxsw_sp2_mask_ethtool_400gaui_8, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_400GAUI_8_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_8X, + .speed = SPEED_400000, + .width = 8, + }, +}; + +#define MLXSW_SP2_PORT_LINK_MODE_LEN ARRAY_SIZE(mlxsw_sp2_port_link_mode) + +static void +mlxsw_sp2_from_ptys_supported_port(struct mlxsw_sp *mlxsw_sp, + u32 ptys_eth_proto, + struct ethtool_link_ksettings *cmd) +{ + ethtool_link_ksettings_add_link_mode(cmd, supported, FIBRE); + ethtool_link_ksettings_add_link_mode(cmd, supported, Backplane); +} + +static void +mlxsw_sp2_set_bit_ethtool(const struct mlxsw_sp2_port_link_mode *link_mode, + unsigned long *mode) +{ + int i; + + for (i = 0; i < link_mode->m_ethtool_len; i++) + __set_bit(link_mode->mask_ethtool[i], mode); +} + +static void +mlxsw_sp2_from_ptys_link(struct mlxsw_sp *mlxsw_sp, u32 ptys_eth_proto, + unsigned long *mode) +{ + int i; + + for (i = 0; i < MLXSW_SP2_PORT_LINK_MODE_LEN; i++) { + if (ptys_eth_proto & mlxsw_sp2_port_link_mode[i].mask) + mlxsw_sp2_set_bit_ethtool(&mlxsw_sp2_port_link_mode[i], + mode); + } +} + +static u32 +mlxsw_sp2_from_ptys_speed(struct mlxsw_sp *mlxsw_sp, u32 ptys_eth_proto) +{ + int i; + + for (i = 0; i < MLXSW_SP2_PORT_LINK_MODE_LEN; i++) { + if (ptys_eth_proto & mlxsw_sp2_port_link_mode[i].mask) + return mlxsw_sp2_port_link_mode[i].speed; + } + + return SPEED_UNKNOWN; +} + +static void +mlxsw_sp2_from_ptys_link_mode(struct mlxsw_sp *mlxsw_sp, bool carrier_ok, + u32 ptys_eth_proto, + struct ethtool_link_ksettings *cmd) +{ + struct mlxsw_sp2_port_link_mode link; + int i; + + cmd->base.speed = SPEED_UNKNOWN; + cmd->base.duplex = DUPLEX_UNKNOWN; + cmd->lanes = 0; + + if (!carrier_ok) + return; + + for (i = 0; i < MLXSW_SP2_PORT_LINK_MODE_LEN; i++) { + if (ptys_eth_proto & mlxsw_sp2_port_link_mode[i].mask) { + link = mlxsw_sp2_port_link_mode[i]; + ethtool_params_from_link_mode(cmd, + link.mask_ethtool[1]); + } + } +} + +static int mlxsw_sp2_ptys_max_speed(struct mlxsw_sp_port *mlxsw_sp_port, u32 *p_max_speed) +{ + u32 eth_proto_cap; + u32 max_speed = 0; + int err; + int i; + + err = mlxsw_sp_port_ptys_query(mlxsw_sp_port, ð_proto_cap, NULL, NULL, NULL); + if (err) + return err; + + for (i = 0; i < MLXSW_SP2_PORT_LINK_MODE_LEN; i++) { + if ((eth_proto_cap & mlxsw_sp2_port_link_mode[i].mask) && + mlxsw_sp2_port_link_mode[i].speed > max_speed) + max_speed = mlxsw_sp2_port_link_mode[i].speed; + } + + *p_max_speed = max_speed; + return 0; +} + +static bool +mlxsw_sp2_test_bit_ethtool(const struct mlxsw_sp2_port_link_mode *link_mode, + const unsigned long *mode) +{ + int cnt = 0; + int i; + + for (i = 0; i < link_mode->m_ethtool_len; i++) { + if (test_bit(link_mode->mask_ethtool[i], mode)) + cnt++; + } + + return cnt == link_mode->m_ethtool_len; +} + +static u32 +mlxsw_sp2_to_ptys_advert_link(struct mlxsw_sp *mlxsw_sp, + const struct ethtool_link_ksettings *cmd) +{ + u32 ptys_proto = 0; + int i; + + for (i = 0; i < MLXSW_SP2_PORT_LINK_MODE_LEN; i++) { + if (mlxsw_sp2_test_bit_ethtool(&mlxsw_sp2_port_link_mode[i], + cmd->link_modes.advertising)) + ptys_proto |= mlxsw_sp2_port_link_mode[i].mask; + } + return ptys_proto; +} + +static u32 mlxsw_sp2_to_ptys_speed_lanes(struct mlxsw_sp *mlxsw_sp, u8 width, + const struct ethtool_link_ksettings *cmd) +{ + u8 mask_width = mlxsw_sp_port_mask_width_get(width); + struct mlxsw_sp2_port_link_mode link_mode; + u32 ptys_proto = 0; + int i; + + if (cmd->lanes > width) + return ptys_proto; + + for (i = 0; i < MLXSW_SP2_PORT_LINK_MODE_LEN; i++) { + if (cmd->base.speed == mlxsw_sp2_port_link_mode[i].speed) { + link_mode = mlxsw_sp2_port_link_mode[i]; + + if (!cmd->lanes) { + /* If number of lanes was not set by user space, + * choose the link mode that supports the width + * of the port. + */ + if (mask_width & link_mode.mask_sup_width) + ptys_proto |= link_mode.mask; + } else if (cmd->lanes == link_mode.width) { + /* Else if the number of lanes was set, choose + * the link mode that its actual width equals to + * it. + */ + ptys_proto |= link_mode.mask; + } + } + } + return ptys_proto; +} + +static void +mlxsw_sp2_reg_ptys_eth_pack(struct mlxsw_sp *mlxsw_sp, char *payload, + u16 local_port, u32 proto_admin, + bool autoneg) +{ + mlxsw_reg_ptys_ext_eth_pack(payload, local_port, proto_admin, autoneg); +} + +static void +mlxsw_sp2_reg_ptys_eth_unpack(struct mlxsw_sp *mlxsw_sp, char *payload, + u32 *p_eth_proto_cap, u32 *p_eth_proto_admin, + u32 *p_eth_proto_oper) +{ + mlxsw_reg_ptys_ext_eth_unpack(payload, p_eth_proto_cap, + p_eth_proto_admin, p_eth_proto_oper); +} + +static u32 mlxsw_sp2_ptys_proto_cap_masked_get(u32 eth_proto_cap) +{ + u32 ptys_proto_cap_masked = 0; + int i; + + for (i = 0; i < MLXSW_SP2_PORT_LINK_MODE_LEN; i++) { + if (mlxsw_sp2_port_link_mode[i].mask & eth_proto_cap) + ptys_proto_cap_masked |= + mlxsw_sp2_port_link_mode[i].mask; + } + + return ptys_proto_cap_masked; +} + +const struct mlxsw_sp_port_type_speed_ops mlxsw_sp2_port_type_speed_ops = { + .from_ptys_supported_port = mlxsw_sp2_from_ptys_supported_port, + .from_ptys_link = mlxsw_sp2_from_ptys_link, + .from_ptys_speed = mlxsw_sp2_from_ptys_speed, + .from_ptys_link_mode = mlxsw_sp2_from_ptys_link_mode, + .ptys_max_speed = mlxsw_sp2_ptys_max_speed, + .to_ptys_advert_link = mlxsw_sp2_to_ptys_advert_link, + .to_ptys_speed_lanes = mlxsw_sp2_to_ptys_speed_lanes, + .reg_ptys_eth_pack = mlxsw_sp2_reg_ptys_eth_pack, + .reg_ptys_eth_unpack = mlxsw_sp2_reg_ptys_eth_unpack, + .ptys_proto_cap_masked_get = mlxsw_sp2_ptys_proto_cap_masked_get, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_fid.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_fid.c new file mode 100644 index 000000000..b6ee2d658 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_fid.c @@ -0,0 +1,1876 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/bitops.h> +#include <linux/if_vlan.h> +#include <linux/if_bridge.h> +#include <linux/netdevice.h> +#include <linux/rhashtable.h> +#include <linux/rtnetlink.h> +#include <linux/refcount.h> + +#include "spectrum.h" +#include "reg.h" + +struct mlxsw_sp_fid_family; + +struct mlxsw_sp_fid_core { + struct rhashtable fid_ht; + struct rhashtable vni_ht; + struct mlxsw_sp_fid_family *fid_family_arr[MLXSW_SP_FID_TYPE_MAX]; + unsigned int *port_fid_mappings; +}; + +struct mlxsw_sp_fid_port_vid { + struct list_head list; + u16 local_port; + u16 vid; +}; + +struct mlxsw_sp_fid { + struct list_head list; + struct mlxsw_sp_rif *rif; + refcount_t ref_count; + u16 fid_index; + u16 fid_offset; + struct mlxsw_sp_fid_family *fid_family; + struct rhash_head ht_node; + + struct rhash_head vni_ht_node; + enum mlxsw_sp_nve_type nve_type; + __be32 vni; + u32 nve_flood_index; + int nve_ifindex; + u8 vni_valid:1, + nve_flood_index_valid:1; + struct list_head port_vid_list; /* Ordered by local port. */ +}; + +struct mlxsw_sp_fid_8021q { + struct mlxsw_sp_fid common; + u16 vid; +}; + +struct mlxsw_sp_fid_8021d { + struct mlxsw_sp_fid common; + int br_ifindex; +}; + +static const struct rhashtable_params mlxsw_sp_fid_ht_params = { + .key_len = sizeof_field(struct mlxsw_sp_fid, fid_index), + .key_offset = offsetof(struct mlxsw_sp_fid, fid_index), + .head_offset = offsetof(struct mlxsw_sp_fid, ht_node), +}; + +static const struct rhashtable_params mlxsw_sp_fid_vni_ht_params = { + .key_len = sizeof_field(struct mlxsw_sp_fid, vni), + .key_offset = offsetof(struct mlxsw_sp_fid, vni), + .head_offset = offsetof(struct mlxsw_sp_fid, vni_ht_node), +}; + +struct mlxsw_sp_flood_table { + enum mlxsw_sp_flood_type packet_type; + enum mlxsw_flood_table_type table_type; + int table_index; +}; + +struct mlxsw_sp_fid_ops { + void (*setup)(struct mlxsw_sp_fid *fid, const void *arg); + int (*configure)(struct mlxsw_sp_fid *fid); + void (*deconfigure)(struct mlxsw_sp_fid *fid); + int (*index_alloc)(struct mlxsw_sp_fid *fid, const void *arg, + u16 *p_fid_index); + bool (*compare)(const struct mlxsw_sp_fid *fid, + const void *arg); + int (*port_vid_map)(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_port *port, u16 vid); + void (*port_vid_unmap)(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_port *port, u16 vid); + int (*vni_set)(struct mlxsw_sp_fid *fid); + void (*vni_clear)(struct mlxsw_sp_fid *fid); + int (*nve_flood_index_set)(struct mlxsw_sp_fid *fid); + void (*nve_flood_index_clear)(struct mlxsw_sp_fid *fid); + void (*fdb_clear_offload)(const struct mlxsw_sp_fid *fid, + const struct net_device *nve_dev); + int (*vid_to_fid_rif_update)(const struct mlxsw_sp_fid *fid, + const struct mlxsw_sp_rif *rif); +}; + +struct mlxsw_sp_fid_family { + enum mlxsw_sp_fid_type type; + size_t fid_size; + u16 start_index; + u16 end_index; + struct list_head fids_list; + unsigned long *fids_bitmap; + const struct mlxsw_sp_flood_table *flood_tables; + int nr_flood_tables; + enum mlxsw_sp_rif_type rif_type; + const struct mlxsw_sp_fid_ops *ops; + struct mlxsw_sp *mlxsw_sp; + bool flood_rsp; + enum mlxsw_reg_bridge_type bridge_type; + u16 pgt_base; + bool smpe_index_valid; +}; + +static const int mlxsw_sp_sfgc_uc_packet_types[MLXSW_REG_SFGC_TYPE_MAX] = { + [MLXSW_REG_SFGC_TYPE_UNKNOWN_UNICAST] = 1, +}; + +static const int mlxsw_sp_sfgc_bc_packet_types[MLXSW_REG_SFGC_TYPE_MAX] = { + [MLXSW_REG_SFGC_TYPE_BROADCAST] = 1, + [MLXSW_REG_SFGC_TYPE_UNREGISTERED_MULTICAST_NON_IP] = 1, + [MLXSW_REG_SFGC_TYPE_IPV4_LINK_LOCAL] = 1, + [MLXSW_REG_SFGC_TYPE_IPV6_ALL_HOST] = 1, + [MLXSW_REG_SFGC_TYPE_UNREGISTERED_MULTICAST_IPV6] = 1, +}; + +static const int mlxsw_sp_sfgc_mc_packet_types[MLXSW_REG_SFGC_TYPE_MAX] = { + [MLXSW_REG_SFGC_TYPE_UNREGISTERED_MULTICAST_IPV4] = 1, +}; + +static const int *mlxsw_sp_packet_type_sfgc_types[] = { + [MLXSW_SP_FLOOD_TYPE_UC] = mlxsw_sp_sfgc_uc_packet_types, + [MLXSW_SP_FLOOD_TYPE_BC] = mlxsw_sp_sfgc_bc_packet_types, + [MLXSW_SP_FLOOD_TYPE_MC] = mlxsw_sp_sfgc_mc_packet_types, +}; + +bool mlxsw_sp_fid_is_dummy(struct mlxsw_sp *mlxsw_sp, u16 fid_index) +{ + enum mlxsw_sp_fid_type fid_type = MLXSW_SP_FID_TYPE_DUMMY; + struct mlxsw_sp_fid_family *fid_family; + + fid_family = mlxsw_sp->fid_core->fid_family_arr[fid_type]; + + return fid_family->start_index == fid_index; +} + +struct mlxsw_sp_fid *mlxsw_sp_fid_lookup_by_index(struct mlxsw_sp *mlxsw_sp, + u16 fid_index) +{ + struct mlxsw_sp_fid *fid; + + fid = rhashtable_lookup_fast(&mlxsw_sp->fid_core->fid_ht, &fid_index, + mlxsw_sp_fid_ht_params); + if (fid) + refcount_inc(&fid->ref_count); + + return fid; +} + +int mlxsw_sp_fid_nve_ifindex(const struct mlxsw_sp_fid *fid, int *nve_ifindex) +{ + if (!fid->vni_valid) + return -EINVAL; + + *nve_ifindex = fid->nve_ifindex; + + return 0; +} + +int mlxsw_sp_fid_nve_type(const struct mlxsw_sp_fid *fid, + enum mlxsw_sp_nve_type *p_type) +{ + if (!fid->vni_valid) + return -EINVAL; + + *p_type = fid->nve_type; + + return 0; +} + +struct mlxsw_sp_fid *mlxsw_sp_fid_lookup_by_vni(struct mlxsw_sp *mlxsw_sp, + __be32 vni) +{ + struct mlxsw_sp_fid *fid; + + fid = rhashtable_lookup_fast(&mlxsw_sp->fid_core->vni_ht, &vni, + mlxsw_sp_fid_vni_ht_params); + if (fid) + refcount_inc(&fid->ref_count); + + return fid; +} + +int mlxsw_sp_fid_vni(const struct mlxsw_sp_fid *fid, __be32 *vni) +{ + if (!fid->vni_valid) + return -EINVAL; + + *vni = fid->vni; + + return 0; +} + +int mlxsw_sp_fid_nve_flood_index_set(struct mlxsw_sp_fid *fid, + u32 nve_flood_index) +{ + struct mlxsw_sp_fid_family *fid_family = fid->fid_family; + const struct mlxsw_sp_fid_ops *ops = fid_family->ops; + int err; + + if (WARN_ON(fid->nve_flood_index_valid)) + return -EINVAL; + + fid->nve_flood_index = nve_flood_index; + fid->nve_flood_index_valid = true; + err = ops->nve_flood_index_set(fid); + if (err) + goto err_nve_flood_index_set; + + return 0; + +err_nve_flood_index_set: + fid->nve_flood_index_valid = false; + return err; +} + +void mlxsw_sp_fid_nve_flood_index_clear(struct mlxsw_sp_fid *fid) +{ + struct mlxsw_sp_fid_family *fid_family = fid->fid_family; + const struct mlxsw_sp_fid_ops *ops = fid_family->ops; + + if (WARN_ON(!fid->nve_flood_index_valid)) + return; + + fid->nve_flood_index_valid = false; + ops->nve_flood_index_clear(fid); +} + +bool mlxsw_sp_fid_nve_flood_index_is_set(const struct mlxsw_sp_fid *fid) +{ + return fid->nve_flood_index_valid; +} + +int mlxsw_sp_fid_vni_set(struct mlxsw_sp_fid *fid, enum mlxsw_sp_nve_type type, + __be32 vni, int nve_ifindex) +{ + struct mlxsw_sp_fid_family *fid_family = fid->fid_family; + const struct mlxsw_sp_fid_ops *ops = fid_family->ops; + struct mlxsw_sp *mlxsw_sp = fid_family->mlxsw_sp; + int err; + + if (WARN_ON(fid->vni_valid)) + return -EINVAL; + + fid->nve_type = type; + fid->nve_ifindex = nve_ifindex; + fid->vni = vni; + err = rhashtable_lookup_insert_fast(&mlxsw_sp->fid_core->vni_ht, + &fid->vni_ht_node, + mlxsw_sp_fid_vni_ht_params); + if (err) + return err; + + fid->vni_valid = true; + err = ops->vni_set(fid); + if (err) + goto err_vni_set; + + return 0; + +err_vni_set: + fid->vni_valid = false; + rhashtable_remove_fast(&mlxsw_sp->fid_core->vni_ht, &fid->vni_ht_node, + mlxsw_sp_fid_vni_ht_params); + return err; +} + +void mlxsw_sp_fid_vni_clear(struct mlxsw_sp_fid *fid) +{ + struct mlxsw_sp_fid_family *fid_family = fid->fid_family; + const struct mlxsw_sp_fid_ops *ops = fid_family->ops; + struct mlxsw_sp *mlxsw_sp = fid_family->mlxsw_sp; + + if (WARN_ON(!fid->vni_valid)) + return; + + fid->vni_valid = false; + ops->vni_clear(fid); + rhashtable_remove_fast(&mlxsw_sp->fid_core->vni_ht, &fid->vni_ht_node, + mlxsw_sp_fid_vni_ht_params); +} + +bool mlxsw_sp_fid_vni_is_set(const struct mlxsw_sp_fid *fid) +{ + return fid->vni_valid; +} + +void mlxsw_sp_fid_fdb_clear_offload(const struct mlxsw_sp_fid *fid, + const struct net_device *nve_dev) +{ + struct mlxsw_sp_fid_family *fid_family = fid->fid_family; + const struct mlxsw_sp_fid_ops *ops = fid_family->ops; + + if (ops->fdb_clear_offload) + ops->fdb_clear_offload(fid, nve_dev); +} + +static const struct mlxsw_sp_flood_table * +mlxsw_sp_fid_flood_table_lookup(const struct mlxsw_sp_fid *fid, + enum mlxsw_sp_flood_type packet_type) +{ + struct mlxsw_sp_fid_family *fid_family = fid->fid_family; + int i; + + for (i = 0; i < fid_family->nr_flood_tables; i++) { + if (fid_family->flood_tables[i].packet_type != packet_type) + continue; + return &fid_family->flood_tables[i]; + } + + return NULL; +} + +static u16 +mlxsw_sp_fid_family_num_fids(const struct mlxsw_sp_fid_family *fid_family) +{ + return fid_family->end_index - fid_family->start_index + 1; +} + +static u16 +mlxsw_sp_fid_flood_table_mid(const struct mlxsw_sp_fid_family *fid_family, + const struct mlxsw_sp_flood_table *flood_table, + u16 fid_offset) +{ + u16 num_fids; + + num_fids = mlxsw_sp_fid_family_num_fids(fid_family); + return fid_family->pgt_base + num_fids * flood_table->table_index + + fid_offset; +} + +int mlxsw_sp_fid_flood_set(struct mlxsw_sp_fid *fid, + enum mlxsw_sp_flood_type packet_type, u16 local_port, + bool member) +{ + struct mlxsw_sp_fid_family *fid_family = fid->fid_family; + const struct mlxsw_sp_flood_table *flood_table; + u16 mid_index; + + if (WARN_ON(!fid_family->flood_tables)) + return -EINVAL; + + flood_table = mlxsw_sp_fid_flood_table_lookup(fid, packet_type); + if (!flood_table) + return -ESRCH; + + mid_index = mlxsw_sp_fid_flood_table_mid(fid_family, flood_table, + fid->fid_offset); + return mlxsw_sp_pgt_entry_port_set(fid_family->mlxsw_sp, mid_index, + fid->fid_index, local_port, member); +} + +int mlxsw_sp_fid_port_vid_map(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_port *mlxsw_sp_port, u16 vid) +{ + if (WARN_ON(!fid->fid_family->ops->port_vid_map)) + return -EINVAL; + return fid->fid_family->ops->port_vid_map(fid, mlxsw_sp_port, vid); +} + +void mlxsw_sp_fid_port_vid_unmap(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_port *mlxsw_sp_port, u16 vid) +{ + fid->fid_family->ops->port_vid_unmap(fid, mlxsw_sp_port, vid); +} + +u16 mlxsw_sp_fid_index(const struct mlxsw_sp_fid *fid) +{ + return fid->fid_index; +} + +enum mlxsw_sp_fid_type mlxsw_sp_fid_type(const struct mlxsw_sp_fid *fid) +{ + return fid->fid_family->type; +} + +struct mlxsw_sp_rif *mlxsw_sp_fid_rif(const struct mlxsw_sp_fid *fid) +{ + return fid->rif; +} + +enum mlxsw_sp_rif_type +mlxsw_sp_fid_type_rif_type(const struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_fid_type type) +{ + struct mlxsw_sp_fid_core *fid_core = mlxsw_sp->fid_core; + + return fid_core->fid_family_arr[type]->rif_type; +} + +static struct mlxsw_sp_fid_8021q * +mlxsw_sp_fid_8021q_fid(const struct mlxsw_sp_fid *fid) +{ + return container_of(fid, struct mlxsw_sp_fid_8021q, common); +} + +u16 mlxsw_sp_fid_8021q_vid(const struct mlxsw_sp_fid *fid) +{ + return mlxsw_sp_fid_8021q_fid(fid)->vid; +} + +static void mlxsw_sp_fid_8021q_setup(struct mlxsw_sp_fid *fid, const void *arg) +{ + u16 vid = *(u16 *) arg; + + mlxsw_sp_fid_8021q_fid(fid)->vid = vid; + fid->fid_offset = fid->fid_index - fid->fid_family->start_index; +} + +static enum mlxsw_reg_sfmr_op mlxsw_sp_sfmr_op(bool valid) +{ + return valid ? MLXSW_REG_SFMR_OP_CREATE_FID : + MLXSW_REG_SFMR_OP_DESTROY_FID; +} + +static int mlxsw_sp_fid_op(const struct mlxsw_sp_fid *fid, bool valid) +{ + struct mlxsw_sp *mlxsw_sp = fid->fid_family->mlxsw_sp; + char sfmr_pl[MLXSW_REG_SFMR_LEN]; + u16 smpe; + + smpe = fid->fid_family->smpe_index_valid ? fid->fid_index : 0; + + mlxsw_reg_sfmr_pack(sfmr_pl, mlxsw_sp_sfmr_op(valid), fid->fid_index, + fid->fid_offset, fid->fid_family->flood_rsp, + fid->fid_family->bridge_type, + fid->fid_family->smpe_index_valid, smpe); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sfmr), sfmr_pl); +} + +static int mlxsw_sp_fid_edit_op(const struct mlxsw_sp_fid *fid, + const struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp *mlxsw_sp = fid->fid_family->mlxsw_sp; + char sfmr_pl[MLXSW_REG_SFMR_LEN]; + u16 smpe; + + smpe = fid->fid_family->smpe_index_valid ? fid->fid_index : 0; + + mlxsw_reg_sfmr_pack(sfmr_pl, MLXSW_REG_SFMR_OP_CREATE_FID, + fid->fid_index, fid->fid_offset, + fid->fid_family->flood_rsp, + fid->fid_family->bridge_type, + fid->fid_family->smpe_index_valid, smpe); + mlxsw_reg_sfmr_vv_set(sfmr_pl, fid->vni_valid); + mlxsw_reg_sfmr_vni_set(sfmr_pl, be32_to_cpu(fid->vni)); + mlxsw_reg_sfmr_vtfp_set(sfmr_pl, fid->nve_flood_index_valid); + mlxsw_reg_sfmr_nve_tunnel_flood_ptr_set(sfmr_pl, fid->nve_flood_index); + + if (rif) { + mlxsw_reg_sfmr_irif_v_set(sfmr_pl, true); + mlxsw_reg_sfmr_irif_set(sfmr_pl, mlxsw_sp_rif_index(rif)); + } + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sfmr), sfmr_pl); +} + +static int mlxsw_sp_fid_vni_to_fid_map(const struct mlxsw_sp_fid *fid, + const struct mlxsw_sp_rif *rif, + bool valid) +{ + struct mlxsw_sp *mlxsw_sp = fid->fid_family->mlxsw_sp; + char svfa_pl[MLXSW_REG_SVFA_LEN]; + bool irif_valid; + u16 irif_index; + + irif_valid = !!rif; + irif_index = rif ? mlxsw_sp_rif_index(rif) : 0; + + mlxsw_reg_svfa_vni_pack(svfa_pl, valid, fid->fid_index, + be32_to_cpu(fid->vni), irif_valid, irif_index); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(svfa), svfa_pl); +} + +static int mlxsw_sp_fid_to_fid_rif_update(const struct mlxsw_sp_fid *fid, + const struct mlxsw_sp_rif *rif) +{ + return mlxsw_sp_fid_edit_op(fid, rif); +} + +static int mlxsw_sp_fid_vni_to_fid_rif_update(const struct mlxsw_sp_fid *fid, + const struct mlxsw_sp_rif *rif) +{ + if (!fid->vni_valid) + return 0; + + return mlxsw_sp_fid_vni_to_fid_map(fid, rif, fid->vni_valid); +} + +static int +mlxsw_sp_fid_vid_to_fid_map(const struct mlxsw_sp_fid *fid, u16 vid, bool valid, + const struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp *mlxsw_sp = fid->fid_family->mlxsw_sp; + char svfa_pl[MLXSW_REG_SVFA_LEN]; + bool irif_valid; + u16 irif_index; + + irif_valid = !!rif; + irif_index = rif ? mlxsw_sp_rif_index(rif) : 0; + + mlxsw_reg_svfa_vid_pack(svfa_pl, valid, fid->fid_index, vid, irif_valid, + irif_index); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(svfa), svfa_pl); +} + +static int +mlxsw_sp_fid_8021q_vid_to_fid_rif_update(const struct mlxsw_sp_fid *fid, + const struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp_fid_8021q *fid_8021q = mlxsw_sp_fid_8021q_fid(fid); + + /* Update the global VID => FID mapping we created when the FID was + * configured. + */ + return mlxsw_sp_fid_vid_to_fid_map(fid, fid_8021q->vid, true, rif); +} + +static int +mlxsw_sp_fid_port_vid_to_fid_rif_update_one(const struct mlxsw_sp_fid *fid, + struct mlxsw_sp_fid_port_vid *pv, + bool irif_valid, u16 irif_index) +{ + struct mlxsw_sp *mlxsw_sp = fid->fid_family->mlxsw_sp; + char svfa_pl[MLXSW_REG_SVFA_LEN]; + + mlxsw_reg_svfa_port_vid_pack(svfa_pl, pv->local_port, true, + fid->fid_index, pv->vid, irif_valid, + irif_index); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(svfa), svfa_pl); +} + +static int mlxsw_sp_fid_vid_to_fid_rif_set(const struct mlxsw_sp_fid *fid, + const struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp *mlxsw_sp = fid->fid_family->mlxsw_sp; + struct mlxsw_sp_fid_port_vid *pv; + u16 irif_index; + int err; + + err = fid->fid_family->ops->vid_to_fid_rif_update(fid, rif); + if (err) + return err; + + irif_index = mlxsw_sp_rif_index(rif); + + list_for_each_entry(pv, &fid->port_vid_list, list) { + /* If port is not in virtual mode, then it does not have any + * {Port, VID}->FID mappings that need to be updated with the + * ingress RIF. + */ + if (!mlxsw_sp->fid_core->port_fid_mappings[pv->local_port]) + continue; + + err = mlxsw_sp_fid_port_vid_to_fid_rif_update_one(fid, pv, + true, + irif_index); + if (err) + goto err_port_vid_to_fid_rif_update_one; + } + + return 0; + +err_port_vid_to_fid_rif_update_one: + list_for_each_entry_continue_reverse(pv, &fid->port_vid_list, list) { + if (!mlxsw_sp->fid_core->port_fid_mappings[pv->local_port]) + continue; + + mlxsw_sp_fid_port_vid_to_fid_rif_update_one(fid, pv, false, 0); + } + + fid->fid_family->ops->vid_to_fid_rif_update(fid, NULL); + return err; +} + +static void mlxsw_sp_fid_vid_to_fid_rif_unset(const struct mlxsw_sp_fid *fid) +{ + struct mlxsw_sp *mlxsw_sp = fid->fid_family->mlxsw_sp; + struct mlxsw_sp_fid_port_vid *pv; + + list_for_each_entry(pv, &fid->port_vid_list, list) { + /* If port is not in virtual mode, then it does not have any + * {Port, VID}->FID mappings that need to be updated. + */ + if (!mlxsw_sp->fid_core->port_fid_mappings[pv->local_port]) + continue; + + mlxsw_sp_fid_port_vid_to_fid_rif_update_one(fid, pv, false, 0); + } + + fid->fid_family->ops->vid_to_fid_rif_update(fid, NULL); +} + +static int mlxsw_sp_fid_reiv_handle(struct mlxsw_sp_fid *fid, u16 rif_index, + bool valid, u8 port_page) +{ + u16 local_port_end = (port_page + 1) * MLXSW_REG_REIV_REC_MAX_COUNT - 1; + u16 local_port_start = port_page * MLXSW_REG_REIV_REC_MAX_COUNT; + struct mlxsw_sp *mlxsw_sp = fid->fid_family->mlxsw_sp; + struct mlxsw_sp_fid_port_vid *port_vid; + u8 rec_num, entries_num = 0; + char *reiv_pl; + int err; + + reiv_pl = kmalloc(MLXSW_REG_REIV_LEN, GFP_KERNEL); + if (!reiv_pl) + return -ENOMEM; + + mlxsw_reg_reiv_pack(reiv_pl, port_page, rif_index); + + list_for_each_entry(port_vid, &fid->port_vid_list, list) { + /* port_vid_list is sorted by local_port. */ + if (port_vid->local_port < local_port_start) + continue; + + if (port_vid->local_port > local_port_end) + break; + + rec_num = port_vid->local_port % MLXSW_REG_REIV_REC_MAX_COUNT; + mlxsw_reg_reiv_rec_update_set(reiv_pl, rec_num, true); + mlxsw_reg_reiv_rec_evid_set(reiv_pl, rec_num, + valid ? port_vid->vid : 0); + entries_num++; + } + + if (!entries_num) { + kfree(reiv_pl); + return 0; + } + + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(reiv), reiv_pl); + if (err) + goto err_reg_write; + + kfree(reiv_pl); + return 0; + +err_reg_write: + kfree(reiv_pl); + return err; +} + +static int mlxsw_sp_fid_erif_eport_to_vid_map(struct mlxsw_sp_fid *fid, + u16 rif_index, bool valid) +{ + struct mlxsw_sp *mlxsw_sp = fid->fid_family->mlxsw_sp; + u8 num_port_pages; + int err, i; + + num_port_pages = mlxsw_core_max_ports(mlxsw_sp->core) / + MLXSW_REG_REIV_REC_MAX_COUNT + 1; + + for (i = 0; i < num_port_pages; i++) { + err = mlxsw_sp_fid_reiv_handle(fid, rif_index, valid, i); + if (err) + goto err_reiv_handle; + } + + return 0; + +err_reiv_handle: + for (; i >= 0; i--) + mlxsw_sp_fid_reiv_handle(fid, rif_index, !valid, i); + return err; +} + +int mlxsw_sp_fid_rif_set(struct mlxsw_sp_fid *fid, struct mlxsw_sp_rif *rif) +{ + u16 rif_index = mlxsw_sp_rif_index(rif); + int err; + + err = mlxsw_sp_fid_to_fid_rif_update(fid, rif); + if (err) + return err; + + err = mlxsw_sp_fid_vni_to_fid_rif_update(fid, rif); + if (err) + goto err_vni_to_fid_rif_update; + + err = mlxsw_sp_fid_vid_to_fid_rif_set(fid, rif); + if (err) + goto err_vid_to_fid_rif_set; + + err = mlxsw_sp_fid_erif_eport_to_vid_map(fid, rif_index, true); + if (err) + goto err_erif_eport_to_vid_map; + + fid->rif = rif; + return 0; + +err_erif_eport_to_vid_map: + mlxsw_sp_fid_vid_to_fid_rif_unset(fid); +err_vid_to_fid_rif_set: + mlxsw_sp_fid_vni_to_fid_rif_update(fid, NULL); +err_vni_to_fid_rif_update: + mlxsw_sp_fid_to_fid_rif_update(fid, NULL); + return err; +} + +void mlxsw_sp_fid_rif_unset(struct mlxsw_sp_fid *fid) +{ + u16 rif_index; + + if (!fid->rif) + return; + + rif_index = mlxsw_sp_rif_index(fid->rif); + fid->rif = NULL; + + mlxsw_sp_fid_erif_eport_to_vid_map(fid, rif_index, false); + mlxsw_sp_fid_vid_to_fid_rif_unset(fid); + mlxsw_sp_fid_vni_to_fid_rif_update(fid, NULL); + mlxsw_sp_fid_to_fid_rif_update(fid, NULL); +} + +static int mlxsw_sp_fid_vni_op(const struct mlxsw_sp_fid *fid) +{ + int err; + + err = mlxsw_sp_fid_vni_to_fid_map(fid, fid->rif, fid->vni_valid); + if (err) + return err; + + err = mlxsw_sp_fid_edit_op(fid, fid->rif); + if (err) + goto err_fid_edit_op; + + return 0; + +err_fid_edit_op: + mlxsw_sp_fid_vni_to_fid_map(fid, fid->rif, !fid->vni_valid); + return err; +} + +static int __mlxsw_sp_fid_port_vid_map(const struct mlxsw_sp_fid *fid, + u16 local_port, u16 vid, bool valid) +{ + struct mlxsw_sp *mlxsw_sp = fid->fid_family->mlxsw_sp; + char svfa_pl[MLXSW_REG_SVFA_LEN]; + bool irif_valid = false; + u16 irif_index = 0; + + if (fid->rif) { + irif_valid = true; + irif_index = mlxsw_sp_rif_index(fid->rif); + } + + mlxsw_reg_svfa_port_vid_pack(svfa_pl, local_port, valid, fid->fid_index, + vid, irif_valid, irif_index); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(svfa), svfa_pl); +} + +static struct mlxsw_sp_fid_8021d * +mlxsw_sp_fid_8021d_fid(const struct mlxsw_sp_fid *fid) +{ + return container_of(fid, struct mlxsw_sp_fid_8021d, common); +} + +static void mlxsw_sp_fid_8021d_setup(struct mlxsw_sp_fid *fid, const void *arg) +{ + int br_ifindex = *(int *) arg; + + mlxsw_sp_fid_8021d_fid(fid)->br_ifindex = br_ifindex; + fid->fid_offset = fid->fid_index - fid->fid_family->start_index; +} + +static int mlxsw_sp_fid_8021d_configure(struct mlxsw_sp_fid *fid) +{ + return mlxsw_sp_fid_op(fid, true); +} + +static void mlxsw_sp_fid_8021d_deconfigure(struct mlxsw_sp_fid *fid) +{ + if (fid->vni_valid) + mlxsw_sp_nve_fid_disable(fid->fid_family->mlxsw_sp, fid); + mlxsw_sp_fid_op(fid, false); +} + +static int mlxsw_sp_fid_8021d_index_alloc(struct mlxsw_sp_fid *fid, + const void *arg, u16 *p_fid_index) +{ + struct mlxsw_sp_fid_family *fid_family = fid->fid_family; + u16 nr_fids, fid_index; + + nr_fids = fid_family->end_index - fid_family->start_index + 1; + fid_index = find_first_zero_bit(fid_family->fids_bitmap, nr_fids); + if (fid_index == nr_fids) + return -ENOBUFS; + *p_fid_index = fid_family->start_index + fid_index; + + return 0; +} + +static bool +mlxsw_sp_fid_8021d_compare(const struct mlxsw_sp_fid *fid, const void *arg) +{ + int br_ifindex = *(int *) arg; + + return mlxsw_sp_fid_8021d_fid(fid)->br_ifindex == br_ifindex; +} + +static int mlxsw_sp_port_vp_mode_trans(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + int err; + + list_for_each_entry(mlxsw_sp_port_vlan, &mlxsw_sp_port->vlans_list, + list) { + struct mlxsw_sp_fid *fid = mlxsw_sp_port_vlan->fid; + u16 vid = mlxsw_sp_port_vlan->vid; + + if (!fid) + continue; + + err = __mlxsw_sp_fid_port_vid_map(fid, + mlxsw_sp_port->local_port, + vid, true); + if (err) + goto err_fid_port_vid_map; + } + + err = mlxsw_sp_port_vp_mode_set(mlxsw_sp_port, true); + if (err) + goto err_port_vp_mode_set; + + return 0; + +err_port_vp_mode_set: +err_fid_port_vid_map: + list_for_each_entry_continue_reverse(mlxsw_sp_port_vlan, + &mlxsw_sp_port->vlans_list, list) { + struct mlxsw_sp_fid *fid = mlxsw_sp_port_vlan->fid; + u16 vid = mlxsw_sp_port_vlan->vid; + + if (!fid) + continue; + + __mlxsw_sp_fid_port_vid_map(fid, mlxsw_sp_port->local_port, vid, + false); + } + return err; +} + +static void mlxsw_sp_port_vlan_mode_trans(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + + mlxsw_sp_port_vp_mode_set(mlxsw_sp_port, false); + + list_for_each_entry_reverse(mlxsw_sp_port_vlan, + &mlxsw_sp_port->vlans_list, list) { + struct mlxsw_sp_fid *fid = mlxsw_sp_port_vlan->fid; + u16 vid = mlxsw_sp_port_vlan->vid; + + if (!fid) + continue; + + __mlxsw_sp_fid_port_vid_map(fid, mlxsw_sp_port->local_port, vid, + false); + } +} + +static int +mlxsw_sp_fid_port_vid_list_add(struct mlxsw_sp_fid *fid, u16 local_port, + u16 vid) +{ + struct mlxsw_sp_fid_port_vid *port_vid, *tmp_port_vid; + + port_vid = kzalloc(sizeof(*port_vid), GFP_KERNEL); + if (!port_vid) + return -ENOMEM; + + port_vid->local_port = local_port; + port_vid->vid = vid; + + list_for_each_entry(tmp_port_vid, &fid->port_vid_list, list) { + if (tmp_port_vid->local_port > local_port) + break; + } + + list_add_tail(&port_vid->list, &tmp_port_vid->list); + return 0; +} + +static void +mlxsw_sp_fid_port_vid_list_del(struct mlxsw_sp_fid *fid, u16 local_port, + u16 vid) +{ + struct mlxsw_sp_fid_port_vid *port_vid, *tmp; + + list_for_each_entry_safe(port_vid, tmp, &fid->port_vid_list, list) { + if (port_vid->local_port != local_port || port_vid->vid != vid) + continue; + + list_del(&port_vid->list); + kfree(port_vid); + return; + } +} + +static int +mlxsw_sp_fid_mpe_table_map(const struct mlxsw_sp_fid *fid, u16 local_port, + u16 vid, bool valid) +{ + struct mlxsw_sp *mlxsw_sp = fid->fid_family->mlxsw_sp; + char smpe_pl[MLXSW_REG_SMPE_LEN]; + + mlxsw_reg_smpe_pack(smpe_pl, local_port, fid->fid_index, + valid ? vid : 0); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(smpe), smpe_pl); +} + +static int +mlxsw_sp_fid_erif_eport_to_vid_map_one(const struct mlxsw_sp_fid *fid, + u16 local_port, u16 vid, bool valid) +{ + u8 port_page = local_port / MLXSW_REG_REIV_REC_MAX_COUNT; + u8 rec_num = local_port % MLXSW_REG_REIV_REC_MAX_COUNT; + struct mlxsw_sp *mlxsw_sp = fid->fid_family->mlxsw_sp; + u16 rif_index = mlxsw_sp_rif_index(fid->rif); + char *reiv_pl; + int err; + + reiv_pl = kmalloc(MLXSW_REG_REIV_LEN, GFP_KERNEL); + if (!reiv_pl) + return -ENOMEM; + + mlxsw_reg_reiv_pack(reiv_pl, port_page, rif_index); + mlxsw_reg_reiv_rec_update_set(reiv_pl, rec_num, true); + mlxsw_reg_reiv_rec_evid_set(reiv_pl, rec_num, valid ? vid : 0); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(reiv), reiv_pl); + kfree(reiv_pl); + return err; +} + +static int mlxsw_sp_fid_evid_map(const struct mlxsw_sp_fid *fid, u16 local_port, + u16 vid, bool valid) +{ + int err; + + err = mlxsw_sp_fid_mpe_table_map(fid, local_port, vid, valid); + if (err) + return err; + + if (!fid->rif) + return 0; + + err = mlxsw_sp_fid_erif_eport_to_vid_map_one(fid, local_port, vid, + valid); + if (err) + goto err_erif_eport_to_vid_map_one; + + return 0; + +err_erif_eport_to_vid_map_one: + mlxsw_sp_fid_mpe_table_map(fid, local_port, vid, !valid); + return err; +} + +static int mlxsw_sp_fid_8021d_port_vid_map(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_port *mlxsw_sp_port, + u16 vid) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 local_port = mlxsw_sp_port->local_port; + int err; + + err = __mlxsw_sp_fid_port_vid_map(fid, mlxsw_sp_port->local_port, vid, + true); + if (err) + return err; + + err = mlxsw_sp_fid_evid_map(fid, local_port, vid, true); + if (err) + goto err_fid_evid_map; + + err = mlxsw_sp_fid_port_vid_list_add(fid, mlxsw_sp_port->local_port, + vid); + if (err) + goto err_port_vid_list_add; + + if (mlxsw_sp->fid_core->port_fid_mappings[local_port]++ == 0) { + err = mlxsw_sp_port_vp_mode_trans(mlxsw_sp_port); + if (err) + goto err_port_vp_mode_trans; + } + + return 0; + +err_port_vp_mode_trans: + mlxsw_sp->fid_core->port_fid_mappings[local_port]--; + mlxsw_sp_fid_port_vid_list_del(fid, mlxsw_sp_port->local_port, vid); +err_port_vid_list_add: + mlxsw_sp_fid_evid_map(fid, local_port, vid, false); +err_fid_evid_map: + __mlxsw_sp_fid_port_vid_map(fid, mlxsw_sp_port->local_port, vid, false); + return err; +} + +static void +mlxsw_sp_fid_8021d_port_vid_unmap(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_port *mlxsw_sp_port, u16 vid) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 local_port = mlxsw_sp_port->local_port; + + if (mlxsw_sp->fid_core->port_fid_mappings[local_port] == 1) + mlxsw_sp_port_vlan_mode_trans(mlxsw_sp_port); + mlxsw_sp->fid_core->port_fid_mappings[local_port]--; + mlxsw_sp_fid_port_vid_list_del(fid, mlxsw_sp_port->local_port, vid); + mlxsw_sp_fid_evid_map(fid, local_port, vid, false); + __mlxsw_sp_fid_port_vid_map(fid, mlxsw_sp_port->local_port, vid, false); +} + +static int mlxsw_sp_fid_8021d_vni_set(struct mlxsw_sp_fid *fid) +{ + return mlxsw_sp_fid_vni_op(fid); +} + +static void mlxsw_sp_fid_8021d_vni_clear(struct mlxsw_sp_fid *fid) +{ + mlxsw_sp_fid_vni_op(fid); +} + +static int mlxsw_sp_fid_8021d_nve_flood_index_set(struct mlxsw_sp_fid *fid) +{ + return mlxsw_sp_fid_edit_op(fid, fid->rif); +} + +static void mlxsw_sp_fid_8021d_nve_flood_index_clear(struct mlxsw_sp_fid *fid) +{ + mlxsw_sp_fid_edit_op(fid, fid->rif); +} + +static void +mlxsw_sp_fid_8021d_fdb_clear_offload(const struct mlxsw_sp_fid *fid, + const struct net_device *nve_dev) +{ + br_fdb_clear_offload(nve_dev, 0); +} + +static int +mlxsw_sp_fid_8021d_vid_to_fid_rif_update(const struct mlxsw_sp_fid *fid, + const struct mlxsw_sp_rif *rif) +{ + return 0; +} + +static const struct mlxsw_sp_fid_ops mlxsw_sp_fid_8021d_ops = { + .setup = mlxsw_sp_fid_8021d_setup, + .configure = mlxsw_sp_fid_8021d_configure, + .deconfigure = mlxsw_sp_fid_8021d_deconfigure, + .index_alloc = mlxsw_sp_fid_8021d_index_alloc, + .compare = mlxsw_sp_fid_8021d_compare, + .port_vid_map = mlxsw_sp_fid_8021d_port_vid_map, + .port_vid_unmap = mlxsw_sp_fid_8021d_port_vid_unmap, + .vni_set = mlxsw_sp_fid_8021d_vni_set, + .vni_clear = mlxsw_sp_fid_8021d_vni_clear, + .nve_flood_index_set = mlxsw_sp_fid_8021d_nve_flood_index_set, + .nve_flood_index_clear = mlxsw_sp_fid_8021d_nve_flood_index_clear, + .fdb_clear_offload = mlxsw_sp_fid_8021d_fdb_clear_offload, + .vid_to_fid_rif_update = mlxsw_sp_fid_8021d_vid_to_fid_rif_update, +}; + +#define MLXSW_SP_FID_8021Q_MAX (VLAN_N_VID - 2) +#define MLXSW_SP_FID_RFID_MAX (11 * 1024) +#define MLXSW_SP_FID_8021Q_PGT_BASE 0 +#define MLXSW_SP_FID_8021D_PGT_BASE (3 * MLXSW_SP_FID_8021Q_MAX) + +static const struct mlxsw_sp_flood_table mlxsw_sp_fid_8021d_flood_tables[] = { + { + .packet_type = MLXSW_SP_FLOOD_TYPE_UC, + .table_type = MLXSW_REG_SFGC_TABLE_TYPE_FID_OFFSET, + .table_index = 0, + }, + { + .packet_type = MLXSW_SP_FLOOD_TYPE_MC, + .table_type = MLXSW_REG_SFGC_TABLE_TYPE_FID_OFFSET, + .table_index = 1, + }, + { + .packet_type = MLXSW_SP_FLOOD_TYPE_BC, + .table_type = MLXSW_REG_SFGC_TABLE_TYPE_FID_OFFSET, + .table_index = 2, + }, +}; + +static bool +mlxsw_sp_fid_8021q_compare(const struct mlxsw_sp_fid *fid, const void *arg) +{ + u16 vid = *(u16 *) arg; + + return mlxsw_sp_fid_8021q_fid(fid)->vid == vid; +} + +static void +mlxsw_sp_fid_8021q_fdb_clear_offload(const struct mlxsw_sp_fid *fid, + const struct net_device *nve_dev) +{ + br_fdb_clear_offload(nve_dev, mlxsw_sp_fid_8021q_vid(fid)); +} + +static void mlxsw_sp_fid_rfid_setup(struct mlxsw_sp_fid *fid, const void *arg) +{ + fid->fid_offset = 0; +} + +static int mlxsw_sp_fid_rfid_configure(struct mlxsw_sp_fid *fid) +{ + return mlxsw_sp_fid_op(fid, true); +} + +static void mlxsw_sp_fid_rfid_deconfigure(struct mlxsw_sp_fid *fid) +{ + mlxsw_sp_fid_op(fid, false); +} + +static int mlxsw_sp_fid_rfid_index_alloc(struct mlxsw_sp_fid *fid, + const void *arg, u16 *p_fid_index) +{ + u16 rif_index = *(u16 *) arg; + + *p_fid_index = fid->fid_family->start_index + rif_index; + + return 0; +} + +static bool mlxsw_sp_fid_rfid_compare(const struct mlxsw_sp_fid *fid, + const void *arg) +{ + u16 rif_index = *(u16 *) arg; + + return fid->fid_index == rif_index + fid->fid_family->start_index; +} + +static int mlxsw_sp_fid_rfid_port_vid_map(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_port *mlxsw_sp_port, + u16 vid) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 local_port = mlxsw_sp_port->local_port; + int err; + + err = mlxsw_sp_fid_port_vid_list_add(fid, mlxsw_sp_port->local_port, + vid); + if (err) + return err; + + /* Using legacy bridge model, we only need to transition the port to + * virtual mode since {Port, VID} => FID is done by the firmware upon + * RIF creation. Using unified bridge model, we need to map + * {Port, VID} => FID and map egress VID. + */ + err = __mlxsw_sp_fid_port_vid_map(fid, mlxsw_sp_port->local_port, vid, + true); + if (err) + goto err_port_vid_map; + + if (fid->rif) { + err = mlxsw_sp_fid_erif_eport_to_vid_map_one(fid, local_port, + vid, true); + if (err) + goto err_erif_eport_to_vid_map_one; + } + + if (mlxsw_sp->fid_core->port_fid_mappings[local_port]++ == 0) { + err = mlxsw_sp_port_vp_mode_trans(mlxsw_sp_port); + if (err) + goto err_port_vp_mode_trans; + } + + return 0; + +err_port_vp_mode_trans: + mlxsw_sp->fid_core->port_fid_mappings[local_port]--; + if (fid->rif) + mlxsw_sp_fid_erif_eport_to_vid_map_one(fid, local_port, vid, + false); +err_erif_eport_to_vid_map_one: + __mlxsw_sp_fid_port_vid_map(fid, mlxsw_sp_port->local_port, vid, false); +err_port_vid_map: + mlxsw_sp_fid_port_vid_list_del(fid, mlxsw_sp_port->local_port, vid); + return err; +} + +static void +mlxsw_sp_fid_rfid_port_vid_unmap(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_port *mlxsw_sp_port, u16 vid) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 local_port = mlxsw_sp_port->local_port; + + if (mlxsw_sp->fid_core->port_fid_mappings[local_port] == 1) + mlxsw_sp_port_vlan_mode_trans(mlxsw_sp_port); + mlxsw_sp->fid_core->port_fid_mappings[local_port]--; + + if (fid->rif) + mlxsw_sp_fid_erif_eport_to_vid_map_one(fid, local_port, vid, + false); + __mlxsw_sp_fid_port_vid_map(fid, mlxsw_sp_port->local_port, vid, false); + mlxsw_sp_fid_port_vid_list_del(fid, mlxsw_sp_port->local_port, vid); +} + +static int mlxsw_sp_fid_rfid_vni_set(struct mlxsw_sp_fid *fid) +{ + return -EOPNOTSUPP; +} + +static void mlxsw_sp_fid_rfid_vni_clear(struct mlxsw_sp_fid *fid) +{ + WARN_ON_ONCE(1); +} + +static int mlxsw_sp_fid_rfid_nve_flood_index_set(struct mlxsw_sp_fid *fid) +{ + return -EOPNOTSUPP; +} + +static void mlxsw_sp_fid_rfid_nve_flood_index_clear(struct mlxsw_sp_fid *fid) +{ + WARN_ON_ONCE(1); +} + +static int +mlxsw_sp_fid_rfid_vid_to_fid_rif_update(const struct mlxsw_sp_fid *fid, + const struct mlxsw_sp_rif *rif) +{ + return 0; +} + +static const struct mlxsw_sp_fid_ops mlxsw_sp_fid_rfid_ops = { + .setup = mlxsw_sp_fid_rfid_setup, + .configure = mlxsw_sp_fid_rfid_configure, + .deconfigure = mlxsw_sp_fid_rfid_deconfigure, + .index_alloc = mlxsw_sp_fid_rfid_index_alloc, + .compare = mlxsw_sp_fid_rfid_compare, + .port_vid_map = mlxsw_sp_fid_rfid_port_vid_map, + .port_vid_unmap = mlxsw_sp_fid_rfid_port_vid_unmap, + .vni_set = mlxsw_sp_fid_rfid_vni_set, + .vni_clear = mlxsw_sp_fid_rfid_vni_clear, + .nve_flood_index_set = mlxsw_sp_fid_rfid_nve_flood_index_set, + .nve_flood_index_clear = mlxsw_sp_fid_rfid_nve_flood_index_clear, + .vid_to_fid_rif_update = mlxsw_sp_fid_rfid_vid_to_fid_rif_update, +}; + +static void mlxsw_sp_fid_dummy_setup(struct mlxsw_sp_fid *fid, const void *arg) +{ + fid->fid_offset = 0; +} + +static int mlxsw_sp_fid_dummy_configure(struct mlxsw_sp_fid *fid) +{ + return mlxsw_sp_fid_op(fid, true); +} + +static void mlxsw_sp_fid_dummy_deconfigure(struct mlxsw_sp_fid *fid) +{ + mlxsw_sp_fid_op(fid, false); +} + +static int mlxsw_sp_fid_dummy_index_alloc(struct mlxsw_sp_fid *fid, + const void *arg, u16 *p_fid_index) +{ + *p_fid_index = fid->fid_family->start_index; + + return 0; +} + +static bool mlxsw_sp_fid_dummy_compare(const struct mlxsw_sp_fid *fid, + const void *arg) +{ + return true; +} + +static int mlxsw_sp_fid_dummy_vni_set(struct mlxsw_sp_fid *fid) +{ + return -EOPNOTSUPP; +} + +static void mlxsw_sp_fid_dummy_vni_clear(struct mlxsw_sp_fid *fid) +{ + WARN_ON_ONCE(1); +} + +static int mlxsw_sp_fid_dummy_nve_flood_index_set(struct mlxsw_sp_fid *fid) +{ + return -EOPNOTSUPP; +} + +static void mlxsw_sp_fid_dummy_nve_flood_index_clear(struct mlxsw_sp_fid *fid) +{ + WARN_ON_ONCE(1); +} + +static const struct mlxsw_sp_fid_ops mlxsw_sp_fid_dummy_ops = { + .setup = mlxsw_sp_fid_dummy_setup, + .configure = mlxsw_sp_fid_dummy_configure, + .deconfigure = mlxsw_sp_fid_dummy_deconfigure, + .index_alloc = mlxsw_sp_fid_dummy_index_alloc, + .compare = mlxsw_sp_fid_dummy_compare, + .vni_set = mlxsw_sp_fid_dummy_vni_set, + .vni_clear = mlxsw_sp_fid_dummy_vni_clear, + .nve_flood_index_set = mlxsw_sp_fid_dummy_nve_flood_index_set, + .nve_flood_index_clear = mlxsw_sp_fid_dummy_nve_flood_index_clear, +}; + +static int mlxsw_sp_fid_8021q_configure(struct mlxsw_sp_fid *fid) +{ + struct mlxsw_sp_fid_8021q *fid_8021q = mlxsw_sp_fid_8021q_fid(fid); + int err; + + err = mlxsw_sp_fid_op(fid, true); + if (err) + return err; + + err = mlxsw_sp_fid_vid_to_fid_map(fid, fid_8021q->vid, true, fid->rif); + if (err) + goto err_vid_to_fid_map; + + return 0; + +err_vid_to_fid_map: + mlxsw_sp_fid_op(fid, false); + return err; +} + +static void mlxsw_sp_fid_8021q_deconfigure(struct mlxsw_sp_fid *fid) +{ + struct mlxsw_sp_fid_8021q *fid_8021q = mlxsw_sp_fid_8021q_fid(fid); + + if (fid->vni_valid) + mlxsw_sp_nve_fid_disable(fid->fid_family->mlxsw_sp, fid); + + mlxsw_sp_fid_vid_to_fid_map(fid, fid_8021q->vid, false, NULL); + mlxsw_sp_fid_op(fid, false); +} + +static int mlxsw_sp_fid_8021q_port_vid_map(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_port *mlxsw_sp_port, + u16 vid) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 local_port = mlxsw_sp_port->local_port; + int err; + + /* In case there are no {Port, VID} => FID mappings on the port, + * we can use the global VID => FID mapping we created when the + * FID was configured, otherwise, configure new mapping. + */ + if (mlxsw_sp->fid_core->port_fid_mappings[local_port]) { + err = __mlxsw_sp_fid_port_vid_map(fid, local_port, vid, true); + if (err) + return err; + } + + err = mlxsw_sp_fid_evid_map(fid, local_port, vid, true); + if (err) + goto err_fid_evid_map; + + err = mlxsw_sp_fid_port_vid_list_add(fid, mlxsw_sp_port->local_port, + vid); + if (err) + goto err_port_vid_list_add; + + return 0; + +err_port_vid_list_add: + mlxsw_sp_fid_evid_map(fid, local_port, vid, false); +err_fid_evid_map: + if (mlxsw_sp->fid_core->port_fid_mappings[local_port]) + __mlxsw_sp_fid_port_vid_map(fid, local_port, vid, false); + return err; +} + +static void +mlxsw_sp_fid_8021q_port_vid_unmap(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_port *mlxsw_sp_port, u16 vid) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u16 local_port = mlxsw_sp_port->local_port; + + mlxsw_sp_fid_port_vid_list_del(fid, mlxsw_sp_port->local_port, vid); + mlxsw_sp_fid_evid_map(fid, local_port, vid, false); + if (mlxsw_sp->fid_core->port_fid_mappings[local_port]) + __mlxsw_sp_fid_port_vid_map(fid, local_port, vid, false); +} + +static const struct mlxsw_sp_fid_ops mlxsw_sp_fid_8021q_ops = { + .setup = mlxsw_sp_fid_8021q_setup, + .configure = mlxsw_sp_fid_8021q_configure, + .deconfigure = mlxsw_sp_fid_8021q_deconfigure, + .index_alloc = mlxsw_sp_fid_8021d_index_alloc, + .compare = mlxsw_sp_fid_8021q_compare, + .port_vid_map = mlxsw_sp_fid_8021q_port_vid_map, + .port_vid_unmap = mlxsw_sp_fid_8021q_port_vid_unmap, + .vni_set = mlxsw_sp_fid_8021d_vni_set, + .vni_clear = mlxsw_sp_fid_8021d_vni_clear, + .nve_flood_index_set = mlxsw_sp_fid_8021d_nve_flood_index_set, + .nve_flood_index_clear = mlxsw_sp_fid_8021d_nve_flood_index_clear, + .fdb_clear_offload = mlxsw_sp_fid_8021q_fdb_clear_offload, + .vid_to_fid_rif_update = mlxsw_sp_fid_8021q_vid_to_fid_rif_update, +}; + +/* There are 4K-2 802.1Q FIDs */ +#define MLXSW_SP_FID_8021Q_START 1 /* FID 0 is reserved. */ +#define MLXSW_SP_FID_8021Q_END (MLXSW_SP_FID_8021Q_START + \ + MLXSW_SP_FID_8021Q_MAX - 1) + +/* There are 1K 802.1D FIDs */ +#define MLXSW_SP_FID_8021D_START (MLXSW_SP_FID_8021Q_END + 1) +#define MLXSW_SP_FID_8021D_END (MLXSW_SP_FID_8021D_START + \ + MLXSW_SP_FID_8021D_MAX - 1) + +/* There is one dummy FID */ +#define MLXSW_SP_FID_DUMMY (MLXSW_SP_FID_8021D_END + 1) + +/* There are 11K rFIDs */ +#define MLXSW_SP_RFID_START (MLXSW_SP_FID_DUMMY + 1) +#define MLXSW_SP_RFID_END (MLXSW_SP_RFID_START + \ + MLXSW_SP_FID_RFID_MAX - 1) + +static const struct mlxsw_sp_fid_family mlxsw_sp1_fid_8021q_family = { + .type = MLXSW_SP_FID_TYPE_8021Q, + .fid_size = sizeof(struct mlxsw_sp_fid_8021q), + .start_index = MLXSW_SP_FID_8021Q_START, + .end_index = MLXSW_SP_FID_8021Q_END, + .flood_tables = mlxsw_sp_fid_8021d_flood_tables, + .nr_flood_tables = ARRAY_SIZE(mlxsw_sp_fid_8021d_flood_tables), + .rif_type = MLXSW_SP_RIF_TYPE_VLAN, + .ops = &mlxsw_sp_fid_8021q_ops, + .flood_rsp = false, + .bridge_type = MLXSW_REG_BRIDGE_TYPE_0, + .pgt_base = MLXSW_SP_FID_8021Q_PGT_BASE, + .smpe_index_valid = false, +}; + +static const struct mlxsw_sp_fid_family mlxsw_sp1_fid_8021d_family = { + .type = MLXSW_SP_FID_TYPE_8021D, + .fid_size = sizeof(struct mlxsw_sp_fid_8021d), + .start_index = MLXSW_SP_FID_8021D_START, + .end_index = MLXSW_SP_FID_8021D_END, + .flood_tables = mlxsw_sp_fid_8021d_flood_tables, + .nr_flood_tables = ARRAY_SIZE(mlxsw_sp_fid_8021d_flood_tables), + .rif_type = MLXSW_SP_RIF_TYPE_FID, + .ops = &mlxsw_sp_fid_8021d_ops, + .bridge_type = MLXSW_REG_BRIDGE_TYPE_1, + .pgt_base = MLXSW_SP_FID_8021D_PGT_BASE, + .smpe_index_valid = false, +}; + +static const struct mlxsw_sp_fid_family mlxsw_sp1_fid_dummy_family = { + .type = MLXSW_SP_FID_TYPE_DUMMY, + .fid_size = sizeof(struct mlxsw_sp_fid), + .start_index = MLXSW_SP_FID_DUMMY, + .end_index = MLXSW_SP_FID_DUMMY, + .ops = &mlxsw_sp_fid_dummy_ops, + .smpe_index_valid = false, +}; + +static const struct mlxsw_sp_fid_family mlxsw_sp_fid_rfid_family = { + .type = MLXSW_SP_FID_TYPE_RFID, + .fid_size = sizeof(struct mlxsw_sp_fid), + .start_index = MLXSW_SP_RFID_START, + .end_index = MLXSW_SP_RFID_END, + .rif_type = MLXSW_SP_RIF_TYPE_SUBPORT, + .ops = &mlxsw_sp_fid_rfid_ops, + .flood_rsp = true, + .smpe_index_valid = false, +}; + +const struct mlxsw_sp_fid_family *mlxsw_sp1_fid_family_arr[] = { + [MLXSW_SP_FID_TYPE_8021Q] = &mlxsw_sp1_fid_8021q_family, + [MLXSW_SP_FID_TYPE_8021D] = &mlxsw_sp1_fid_8021d_family, + [MLXSW_SP_FID_TYPE_DUMMY] = &mlxsw_sp1_fid_dummy_family, + [MLXSW_SP_FID_TYPE_RFID] = &mlxsw_sp_fid_rfid_family, +}; + +static const struct mlxsw_sp_fid_family mlxsw_sp2_fid_8021q_family = { + .type = MLXSW_SP_FID_TYPE_8021Q, + .fid_size = sizeof(struct mlxsw_sp_fid_8021q), + .start_index = MLXSW_SP_FID_8021Q_START, + .end_index = MLXSW_SP_FID_8021Q_END, + .flood_tables = mlxsw_sp_fid_8021d_flood_tables, + .nr_flood_tables = ARRAY_SIZE(mlxsw_sp_fid_8021d_flood_tables), + .rif_type = MLXSW_SP_RIF_TYPE_VLAN, + .ops = &mlxsw_sp_fid_8021q_ops, + .flood_rsp = false, + .bridge_type = MLXSW_REG_BRIDGE_TYPE_0, + .pgt_base = MLXSW_SP_FID_8021Q_PGT_BASE, + .smpe_index_valid = true, +}; + +static const struct mlxsw_sp_fid_family mlxsw_sp2_fid_8021d_family = { + .type = MLXSW_SP_FID_TYPE_8021D, + .fid_size = sizeof(struct mlxsw_sp_fid_8021d), + .start_index = MLXSW_SP_FID_8021D_START, + .end_index = MLXSW_SP_FID_8021D_END, + .flood_tables = mlxsw_sp_fid_8021d_flood_tables, + .nr_flood_tables = ARRAY_SIZE(mlxsw_sp_fid_8021d_flood_tables), + .rif_type = MLXSW_SP_RIF_TYPE_FID, + .ops = &mlxsw_sp_fid_8021d_ops, + .bridge_type = MLXSW_REG_BRIDGE_TYPE_1, + .pgt_base = MLXSW_SP_FID_8021D_PGT_BASE, + .smpe_index_valid = true, +}; + +static const struct mlxsw_sp_fid_family mlxsw_sp2_fid_dummy_family = { + .type = MLXSW_SP_FID_TYPE_DUMMY, + .fid_size = sizeof(struct mlxsw_sp_fid), + .start_index = MLXSW_SP_FID_DUMMY, + .end_index = MLXSW_SP_FID_DUMMY, + .ops = &mlxsw_sp_fid_dummy_ops, + .smpe_index_valid = false, +}; + +const struct mlxsw_sp_fid_family *mlxsw_sp2_fid_family_arr[] = { + [MLXSW_SP_FID_TYPE_8021Q] = &mlxsw_sp2_fid_8021q_family, + [MLXSW_SP_FID_TYPE_8021D] = &mlxsw_sp2_fid_8021d_family, + [MLXSW_SP_FID_TYPE_DUMMY] = &mlxsw_sp2_fid_dummy_family, + [MLXSW_SP_FID_TYPE_RFID] = &mlxsw_sp_fid_rfid_family, +}; + +static struct mlxsw_sp_fid *mlxsw_sp_fid_lookup(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_fid_type type, + const void *arg) +{ + struct mlxsw_sp_fid_family *fid_family; + struct mlxsw_sp_fid *fid; + + fid_family = mlxsw_sp->fid_core->fid_family_arr[type]; + list_for_each_entry(fid, &fid_family->fids_list, list) { + if (!fid->fid_family->ops->compare(fid, arg)) + continue; + refcount_inc(&fid->ref_count); + return fid; + } + + return NULL; +} + +static struct mlxsw_sp_fid *mlxsw_sp_fid_get(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_fid_type type, + const void *arg) +{ + struct mlxsw_sp_fid_family *fid_family; + struct mlxsw_sp_fid *fid; + u16 fid_index; + int err; + + fid = mlxsw_sp_fid_lookup(mlxsw_sp, type, arg); + if (fid) + return fid; + + fid_family = mlxsw_sp->fid_core->fid_family_arr[type]; + fid = kzalloc(fid_family->fid_size, GFP_KERNEL); + if (!fid) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&fid->port_vid_list); + fid->fid_family = fid_family; + + err = fid->fid_family->ops->index_alloc(fid, arg, &fid_index); + if (err) + goto err_index_alloc; + fid->fid_index = fid_index; + __set_bit(fid_index - fid_family->start_index, fid_family->fids_bitmap); + + fid->fid_family->ops->setup(fid, arg); + + err = fid->fid_family->ops->configure(fid); + if (err) + goto err_configure; + + err = rhashtable_insert_fast(&mlxsw_sp->fid_core->fid_ht, &fid->ht_node, + mlxsw_sp_fid_ht_params); + if (err) + goto err_rhashtable_insert; + + list_add(&fid->list, &fid_family->fids_list); + refcount_set(&fid->ref_count, 1); + return fid; + +err_rhashtable_insert: + fid->fid_family->ops->deconfigure(fid); +err_configure: + __clear_bit(fid_index - fid_family->start_index, + fid_family->fids_bitmap); +err_index_alloc: + kfree(fid); + return ERR_PTR(err); +} + +void mlxsw_sp_fid_put(struct mlxsw_sp_fid *fid) +{ + struct mlxsw_sp_fid_family *fid_family = fid->fid_family; + struct mlxsw_sp *mlxsw_sp = fid_family->mlxsw_sp; + + if (!refcount_dec_and_test(&fid->ref_count)) + return; + + list_del(&fid->list); + rhashtable_remove_fast(&mlxsw_sp->fid_core->fid_ht, + &fid->ht_node, mlxsw_sp_fid_ht_params); + fid->fid_family->ops->deconfigure(fid); + __clear_bit(fid->fid_index - fid_family->start_index, + fid_family->fids_bitmap); + WARN_ON_ONCE(!list_empty(&fid->port_vid_list)); + kfree(fid); +} + +struct mlxsw_sp_fid *mlxsw_sp_fid_8021q_get(struct mlxsw_sp *mlxsw_sp, u16 vid) +{ + return mlxsw_sp_fid_get(mlxsw_sp, MLXSW_SP_FID_TYPE_8021Q, &vid); +} + +struct mlxsw_sp_fid *mlxsw_sp_fid_8021d_get(struct mlxsw_sp *mlxsw_sp, + int br_ifindex) +{ + return mlxsw_sp_fid_get(mlxsw_sp, MLXSW_SP_FID_TYPE_8021D, &br_ifindex); +} + +struct mlxsw_sp_fid *mlxsw_sp_fid_8021q_lookup(struct mlxsw_sp *mlxsw_sp, + u16 vid) +{ + return mlxsw_sp_fid_lookup(mlxsw_sp, MLXSW_SP_FID_TYPE_8021Q, &vid); +} + +struct mlxsw_sp_fid *mlxsw_sp_fid_8021d_lookup(struct mlxsw_sp *mlxsw_sp, + int br_ifindex) +{ + return mlxsw_sp_fid_lookup(mlxsw_sp, MLXSW_SP_FID_TYPE_8021D, + &br_ifindex); +} + +struct mlxsw_sp_fid *mlxsw_sp_fid_rfid_get(struct mlxsw_sp *mlxsw_sp, + u16 rif_index) +{ + return mlxsw_sp_fid_get(mlxsw_sp, MLXSW_SP_FID_TYPE_RFID, &rif_index); +} + +struct mlxsw_sp_fid *mlxsw_sp_fid_dummy_get(struct mlxsw_sp *mlxsw_sp) +{ + return mlxsw_sp_fid_get(mlxsw_sp, MLXSW_SP_FID_TYPE_DUMMY, NULL); +} + +static int +mlxsw_sp_fid_flood_table_init(struct mlxsw_sp_fid_family *fid_family, + const struct mlxsw_sp_flood_table *flood_table) +{ + enum mlxsw_sp_flood_type packet_type = flood_table->packet_type; + struct mlxsw_sp *mlxsw_sp = fid_family->mlxsw_sp; + const int *sfgc_packet_types; + u16 num_fids, mid_base; + int err, i; + + mid_base = mlxsw_sp_fid_flood_table_mid(fid_family, flood_table, 0); + num_fids = mlxsw_sp_fid_family_num_fids(fid_family); + err = mlxsw_sp_pgt_mid_alloc_range(mlxsw_sp, mid_base, num_fids); + if (err) + return err; + + sfgc_packet_types = mlxsw_sp_packet_type_sfgc_types[packet_type]; + for (i = 0; i < MLXSW_REG_SFGC_TYPE_MAX; i++) { + char sfgc_pl[MLXSW_REG_SFGC_LEN]; + + if (!sfgc_packet_types[i]) + continue; + + mlxsw_reg_sfgc_pack(sfgc_pl, i, fid_family->bridge_type, + flood_table->table_type, 0, mid_base); + + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sfgc), sfgc_pl); + if (err) + goto err_reg_write; + } + + return 0; + +err_reg_write: + mlxsw_sp_pgt_mid_free_range(mlxsw_sp, mid_base, num_fids); + return err; +} + +static void +mlxsw_sp_fid_flood_table_fini(struct mlxsw_sp_fid_family *fid_family, + const struct mlxsw_sp_flood_table *flood_table) +{ + struct mlxsw_sp *mlxsw_sp = fid_family->mlxsw_sp; + u16 num_fids, mid_base; + + mid_base = mlxsw_sp_fid_flood_table_mid(fid_family, flood_table, 0); + num_fids = mlxsw_sp_fid_family_num_fids(fid_family); + mlxsw_sp_pgt_mid_free_range(mlxsw_sp, mid_base, num_fids); +} + +static int +mlxsw_sp_fid_flood_tables_init(struct mlxsw_sp_fid_family *fid_family) +{ + int i; + + for (i = 0; i < fid_family->nr_flood_tables; i++) { + const struct mlxsw_sp_flood_table *flood_table; + int err; + + flood_table = &fid_family->flood_tables[i]; + err = mlxsw_sp_fid_flood_table_init(fid_family, flood_table); + if (err) + return err; + } + + return 0; +} + +static void +mlxsw_sp_fid_flood_tables_fini(struct mlxsw_sp_fid_family *fid_family) +{ + int i; + + for (i = 0; i < fid_family->nr_flood_tables; i++) { + const struct mlxsw_sp_flood_table *flood_table; + + flood_table = &fid_family->flood_tables[i]; + mlxsw_sp_fid_flood_table_fini(fid_family, flood_table); + } +} + +static int mlxsw_sp_fid_family_register(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_fid_family *tmpl) +{ + u16 nr_fids = tmpl->end_index - tmpl->start_index + 1; + struct mlxsw_sp_fid_family *fid_family; + int err; + + fid_family = kmemdup(tmpl, sizeof(*fid_family), GFP_KERNEL); + if (!fid_family) + return -ENOMEM; + + fid_family->mlxsw_sp = mlxsw_sp; + INIT_LIST_HEAD(&fid_family->fids_list); + fid_family->fids_bitmap = bitmap_zalloc(nr_fids, GFP_KERNEL); + if (!fid_family->fids_bitmap) { + err = -ENOMEM; + goto err_alloc_fids_bitmap; + } + + if (fid_family->flood_tables) { + err = mlxsw_sp_fid_flood_tables_init(fid_family); + if (err) + goto err_fid_flood_tables_init; + } + + mlxsw_sp->fid_core->fid_family_arr[tmpl->type] = fid_family; + + return 0; + +err_fid_flood_tables_init: + bitmap_free(fid_family->fids_bitmap); +err_alloc_fids_bitmap: + kfree(fid_family); + return err; +} + +static void +mlxsw_sp_fid_family_unregister(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fid_family *fid_family) +{ + mlxsw_sp->fid_core->fid_family_arr[fid_family->type] = NULL; + + if (fid_family->flood_tables) + mlxsw_sp_fid_flood_tables_fini(fid_family); + + bitmap_free(fid_family->fids_bitmap); + WARN_ON_ONCE(!list_empty(&fid_family->fids_list)); + kfree(fid_family); +} + +int mlxsw_sp_port_fids_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + + /* Track number of FIDs configured on the port with mapping type + * PORT_VID_TO_FID, so that we know when to transition the port + * back to non-virtual (VLAN) mode. + */ + mlxsw_sp->fid_core->port_fid_mappings[mlxsw_sp_port->local_port] = 0; + + return mlxsw_sp_port_vp_mode_set(mlxsw_sp_port, false); +} + +void mlxsw_sp_port_fids_fini(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + + mlxsw_sp->fid_core->port_fid_mappings[mlxsw_sp_port->local_port] = 0; +} + +int mlxsw_sp_fids_init(struct mlxsw_sp *mlxsw_sp) +{ + unsigned int max_ports = mlxsw_core_max_ports(mlxsw_sp->core); + struct mlxsw_sp_fid_core *fid_core; + int err, i; + + fid_core = kzalloc(sizeof(*mlxsw_sp->fid_core), GFP_KERNEL); + if (!fid_core) + return -ENOMEM; + mlxsw_sp->fid_core = fid_core; + + err = rhashtable_init(&fid_core->fid_ht, &mlxsw_sp_fid_ht_params); + if (err) + goto err_rhashtable_fid_init; + + err = rhashtable_init(&fid_core->vni_ht, &mlxsw_sp_fid_vni_ht_params); + if (err) + goto err_rhashtable_vni_init; + + fid_core->port_fid_mappings = kcalloc(max_ports, sizeof(unsigned int), + GFP_KERNEL); + if (!fid_core->port_fid_mappings) { + err = -ENOMEM; + goto err_alloc_port_fid_mappings; + } + + for (i = 0; i < MLXSW_SP_FID_TYPE_MAX; i++) { + err = mlxsw_sp_fid_family_register(mlxsw_sp, + mlxsw_sp->fid_family_arr[i]); + + if (err) + goto err_fid_ops_register; + } + + return 0; + +err_fid_ops_register: + for (i--; i >= 0; i--) { + struct mlxsw_sp_fid_family *fid_family; + + fid_family = fid_core->fid_family_arr[i]; + mlxsw_sp_fid_family_unregister(mlxsw_sp, fid_family); + } + kfree(fid_core->port_fid_mappings); +err_alloc_port_fid_mappings: + rhashtable_destroy(&fid_core->vni_ht); +err_rhashtable_vni_init: + rhashtable_destroy(&fid_core->fid_ht); +err_rhashtable_fid_init: + kfree(fid_core); + return err; +} + +void mlxsw_sp_fids_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_fid_core *fid_core = mlxsw_sp->fid_core; + int i; + + for (i = 0; i < MLXSW_SP_FID_TYPE_MAX; i++) + mlxsw_sp_fid_family_unregister(mlxsw_sp, + fid_core->fid_family_arr[i]); + kfree(fid_core->port_fid_mappings); + rhashtable_destroy(&fid_core->vni_ht); + rhashtable_destroy(&fid_core->fid_ht); + kfree(fid_core); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_flow.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_flow.c new file mode 100644 index 000000000..9e50c823a --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_flow.c @@ -0,0 +1,295 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2020 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/slab.h> +#include <linux/errno.h> +#include <linux/list.h> +#include <net/net_namespace.h> + +#include "spectrum.h" + +struct mlxsw_sp_flow_block * +mlxsw_sp_flow_block_create(struct mlxsw_sp *mlxsw_sp, struct net *net) +{ + struct mlxsw_sp_flow_block *block; + + block = kzalloc(sizeof(*block), GFP_KERNEL); + if (!block) + return NULL; + INIT_LIST_HEAD(&block->binding_list); + INIT_LIST_HEAD(&block->mall.list); + block->mlxsw_sp = mlxsw_sp; + block->net = net; + return block; +} + +void mlxsw_sp_flow_block_destroy(struct mlxsw_sp_flow_block *block) +{ + WARN_ON(!list_empty(&block->binding_list)); + kfree(block); +} + +static struct mlxsw_sp_flow_block_binding * +mlxsw_sp_flow_block_lookup(struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, bool ingress) +{ + struct mlxsw_sp_flow_block_binding *binding; + + list_for_each_entry(binding, &block->binding_list, list) + if (binding->mlxsw_sp_port == mlxsw_sp_port && + binding->ingress == ingress) + return binding; + return NULL; +} + +static bool +mlxsw_sp_flow_block_ruleset_bound(const struct mlxsw_sp_flow_block *block) +{ + return block->ruleset_zero; +} + +static int mlxsw_sp_flow_block_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_flow_block_binding *binding; + int err; + + if (WARN_ON(mlxsw_sp_flow_block_lookup(block, mlxsw_sp_port, ingress))) + return -EEXIST; + + if (ingress && block->ingress_blocker_rule_count) { + NL_SET_ERR_MSG_MOD(extack, "Block cannot be bound to ingress because it contains unsupported rules"); + return -EOPNOTSUPP; + } + + if (!ingress && block->egress_blocker_rule_count) { + NL_SET_ERR_MSG_MOD(extack, "Block cannot be bound to egress because it contains unsupported rules"); + return -EOPNOTSUPP; + } + + err = mlxsw_sp_mall_port_bind(block, mlxsw_sp_port, extack); + if (err) + return err; + + binding = kzalloc(sizeof(*binding), GFP_KERNEL); + if (!binding) { + err = -ENOMEM; + goto err_binding_alloc; + } + binding->mlxsw_sp_port = mlxsw_sp_port; + binding->ingress = ingress; + + if (mlxsw_sp_flow_block_ruleset_bound(block)) { + err = mlxsw_sp_acl_ruleset_bind(mlxsw_sp, block, binding); + if (err) + goto err_ruleset_bind; + } + + if (ingress) + block->ingress_binding_count++; + else + block->egress_binding_count++; + list_add(&binding->list, &block->binding_list); + return 0; + +err_ruleset_bind: + kfree(binding); +err_binding_alloc: + mlxsw_sp_mall_port_unbind(block, mlxsw_sp_port); + + return err; +} + +static int mlxsw_sp_flow_block_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + struct mlxsw_sp_flow_block_binding *binding; + + binding = mlxsw_sp_flow_block_lookup(block, mlxsw_sp_port, ingress); + if (!binding) + return -ENOENT; + + list_del(&binding->list); + + if (ingress) + block->ingress_binding_count--; + else + block->egress_binding_count--; + + if (mlxsw_sp_flow_block_ruleset_bound(block)) + mlxsw_sp_acl_ruleset_unbind(mlxsw_sp, block, binding); + + kfree(binding); + + mlxsw_sp_mall_port_unbind(block, mlxsw_sp_port); + + return 0; +} + +static int mlxsw_sp_flow_block_mall_cb(struct mlxsw_sp_flow_block *flow_block, + struct tc_cls_matchall_offload *f) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_flow_block_mlxsw_sp(flow_block); + + switch (f->command) { + case TC_CLSMATCHALL_REPLACE: + return mlxsw_sp_mall_replace(mlxsw_sp, flow_block, f); + case TC_CLSMATCHALL_DESTROY: + mlxsw_sp_mall_destroy(flow_block, f); + return 0; + default: + return -EOPNOTSUPP; + } +} + +static int mlxsw_sp_flow_block_flower_cb(struct mlxsw_sp_flow_block *flow_block, + struct flow_cls_offload *f) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_flow_block_mlxsw_sp(flow_block); + + switch (f->command) { + case FLOW_CLS_REPLACE: + return mlxsw_sp_flower_replace(mlxsw_sp, flow_block, f); + case FLOW_CLS_DESTROY: + mlxsw_sp_flower_destroy(mlxsw_sp, flow_block, f); + return 0; + case FLOW_CLS_STATS: + return mlxsw_sp_flower_stats(mlxsw_sp, flow_block, f); + case FLOW_CLS_TMPLT_CREATE: + return mlxsw_sp_flower_tmplt_create(mlxsw_sp, flow_block, f); + case FLOW_CLS_TMPLT_DESTROY: + mlxsw_sp_flower_tmplt_destroy(mlxsw_sp, flow_block, f); + return 0; + default: + return -EOPNOTSUPP; + } +} + +static int mlxsw_sp_flow_block_cb(enum tc_setup_type type, + void *type_data, void *cb_priv) +{ + struct mlxsw_sp_flow_block *flow_block = cb_priv; + + if (mlxsw_sp_flow_block_disabled(flow_block)) + return -EOPNOTSUPP; + + switch (type) { + case TC_SETUP_CLSMATCHALL: + return mlxsw_sp_flow_block_mall_cb(flow_block, type_data); + case TC_SETUP_CLSFLOWER: + return mlxsw_sp_flow_block_flower_cb(flow_block, type_data); + default: + return -EOPNOTSUPP; + } +} + +static void mlxsw_sp_tc_block_release(void *cb_priv) +{ + struct mlxsw_sp_flow_block *flow_block = cb_priv; + + mlxsw_sp_flow_block_destroy(flow_block); +} + +static LIST_HEAD(mlxsw_sp_block_cb_list); + +static int mlxsw_sp_setup_tc_block_bind(struct mlxsw_sp_port *mlxsw_sp_port, + struct flow_block_offload *f, + bool ingress) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_flow_block *flow_block; + struct flow_block_cb *block_cb; + bool register_block = false; + int err; + + block_cb = flow_block_cb_lookup(f->block, mlxsw_sp_flow_block_cb, + mlxsw_sp); + if (!block_cb) { + flow_block = mlxsw_sp_flow_block_create(mlxsw_sp, f->net); + if (!flow_block) + return -ENOMEM; + block_cb = flow_block_cb_alloc(mlxsw_sp_flow_block_cb, + mlxsw_sp, flow_block, + mlxsw_sp_tc_block_release); + if (IS_ERR(block_cb)) { + mlxsw_sp_flow_block_destroy(flow_block); + return PTR_ERR(block_cb); + } + register_block = true; + } else { + flow_block = flow_block_cb_priv(block_cb); + } + flow_block_cb_incref(block_cb); + err = mlxsw_sp_flow_block_bind(mlxsw_sp, flow_block, + mlxsw_sp_port, ingress, f->extack); + if (err) + goto err_block_bind; + + if (ingress) + mlxsw_sp_port->ing_flow_block = flow_block; + else + mlxsw_sp_port->eg_flow_block = flow_block; + + if (register_block) { + flow_block_cb_add(block_cb, f); + list_add_tail(&block_cb->driver_list, &mlxsw_sp_block_cb_list); + } + + return 0; + +err_block_bind: + if (!flow_block_cb_decref(block_cb)) + flow_block_cb_free(block_cb); + return err; +} + +static void mlxsw_sp_setup_tc_block_unbind(struct mlxsw_sp_port *mlxsw_sp_port, + struct flow_block_offload *f, + bool ingress) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_flow_block *flow_block; + struct flow_block_cb *block_cb; + int err; + + block_cb = flow_block_cb_lookup(f->block, mlxsw_sp_flow_block_cb, + mlxsw_sp); + if (!block_cb) + return; + + if (ingress) + mlxsw_sp_port->ing_flow_block = NULL; + else + mlxsw_sp_port->eg_flow_block = NULL; + + flow_block = flow_block_cb_priv(block_cb); + err = mlxsw_sp_flow_block_unbind(mlxsw_sp, flow_block, + mlxsw_sp_port, ingress); + if (!err && !flow_block_cb_decref(block_cb)) { + flow_block_cb_remove(block_cb, f); + list_del(&block_cb->driver_list); + } +} + +int mlxsw_sp_setup_tc_block_clsact(struct mlxsw_sp_port *mlxsw_sp_port, + struct flow_block_offload *f, + bool ingress) +{ + f->driver_block_list = &mlxsw_sp_block_cb_list; + + switch (f->command) { + case FLOW_BLOCK_BIND: + return mlxsw_sp_setup_tc_block_bind(mlxsw_sp_port, f, ingress); + case FLOW_BLOCK_UNBIND: + mlxsw_sp_setup_tc_block_unbind(mlxsw_sp_port, f, ingress); + return 0; + default: + return -EOPNOTSUPP; + } +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c new file mode 100644 index 000000000..e91fb205e --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c @@ -0,0 +1,796 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/netdevice.h> +#include <linux/log2.h> +#include <net/net_namespace.h> +#include <net/flow_dissector.h> +#include <net/pkt_cls.h> +#include <net/tc_act/tc_gact.h> +#include <net/tc_act/tc_mirred.h> +#include <net/tc_act/tc_vlan.h> + +#include "spectrum.h" +#include "core_acl_flex_keys.h" + +static int mlxsw_sp_policer_validate(const struct flow_action *action, + const struct flow_action_entry *act, + struct netlink_ext_ack *extack) +{ + if (act->police.exceed.act_id != FLOW_ACTION_DROP) { + NL_SET_ERR_MSG_MOD(extack, + "Offload not supported when exceed action is not drop"); + return -EOPNOTSUPP; + } + + if (act->police.notexceed.act_id != FLOW_ACTION_PIPE && + act->police.notexceed.act_id != FLOW_ACTION_ACCEPT) { + NL_SET_ERR_MSG_MOD(extack, + "Offload not supported when conform action is not pipe or ok"); + return -EOPNOTSUPP; + } + + if (act->police.notexceed.act_id == FLOW_ACTION_ACCEPT && + !flow_action_is_last_entry(action, act)) { + NL_SET_ERR_MSG_MOD(extack, + "Offload not supported when conform action is ok, but action is not last"); + return -EOPNOTSUPP; + } + + if (act->police.peakrate_bytes_ps || + act->police.avrate || act->police.overhead) { + NL_SET_ERR_MSG_MOD(extack, + "Offload not supported when peakrate/avrate/overhead is configured"); + return -EOPNOTSUPP; + } + + if (act->police.rate_pkt_ps) { + NL_SET_ERR_MSG_MOD(extack, + "QoS offload not support packets per second"); + return -EOPNOTSUPP; + } + + return 0; +} + +static int mlxsw_sp_flower_parse_actions(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_acl_rule_info *rulei, + struct flow_action *flow_action, + struct netlink_ext_ack *extack) +{ + const struct flow_action_entry *act; + int mirror_act_count = 0; + int police_act_count = 0; + int sample_act_count = 0; + int err, i; + + if (!flow_action_has_entries(flow_action)) + return 0; + if (!flow_action_mixed_hw_stats_check(flow_action, extack)) + return -EOPNOTSUPP; + + act = flow_action_first_entry_get(flow_action); + if (act->hw_stats & FLOW_ACTION_HW_STATS_DISABLED) { + /* Nothing to do */ + } else if (act->hw_stats & FLOW_ACTION_HW_STATS_IMMEDIATE) { + /* Count action is inserted first */ + err = mlxsw_sp_acl_rulei_act_count(mlxsw_sp, rulei, extack); + if (err) + return err; + } else { + NL_SET_ERR_MSG_MOD(extack, "Unsupported action HW stats type"); + return -EOPNOTSUPP; + } + + flow_action_for_each(i, act, flow_action) { + switch (act->id) { + case FLOW_ACTION_ACCEPT: + err = mlxsw_sp_acl_rulei_act_terminate(rulei); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append terminate action"); + return err; + } + break; + case FLOW_ACTION_DROP: { + bool ingress; + + if (mlxsw_sp_flow_block_is_mixed_bound(block)) { + NL_SET_ERR_MSG_MOD(extack, "Drop action is not supported when block is bound to ingress and egress"); + return -EOPNOTSUPP; + } + ingress = mlxsw_sp_flow_block_is_ingress_bound(block); + err = mlxsw_sp_acl_rulei_act_drop(rulei, ingress, + act->cookie, extack); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append drop action"); + return err; + } + + /* Forbid block with this rulei to be bound + * to ingress/egress in future. Ingress rule is + * a blocker for egress and vice versa. + */ + if (ingress) + rulei->egress_bind_blocker = 1; + else + rulei->ingress_bind_blocker = 1; + } + break; + case FLOW_ACTION_TRAP: + err = mlxsw_sp_acl_rulei_act_trap(rulei); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append trap action"); + return err; + } + break; + case FLOW_ACTION_GOTO: { + u32 chain_index = act->chain_index; + struct mlxsw_sp_acl_ruleset *ruleset; + u16 group_id; + + ruleset = mlxsw_sp_acl_ruleset_lookup(mlxsw_sp, block, + chain_index, + MLXSW_SP_ACL_PROFILE_FLOWER); + if (IS_ERR(ruleset)) + return PTR_ERR(ruleset); + + group_id = mlxsw_sp_acl_ruleset_group_id(ruleset); + err = mlxsw_sp_acl_rulei_act_jump(rulei, group_id); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Cannot append jump action"); + return err; + } + } + break; + case FLOW_ACTION_REDIRECT: { + struct net_device *out_dev; + struct mlxsw_sp_fid *fid; + u16 fid_index; + + if (mlxsw_sp_flow_block_is_egress_bound(block)) { + NL_SET_ERR_MSG_MOD(extack, "Redirect action is not supported on egress"); + return -EOPNOTSUPP; + } + + /* Forbid block with this rulei to be bound + * to egress in future. + */ + rulei->egress_bind_blocker = 1; + + fid = mlxsw_sp_acl_dummy_fid(mlxsw_sp); + fid_index = mlxsw_sp_fid_index(fid); + err = mlxsw_sp_acl_rulei_act_fid_set(mlxsw_sp, rulei, + fid_index, extack); + if (err) + return err; + + out_dev = act->dev; + err = mlxsw_sp_acl_rulei_act_fwd(mlxsw_sp, rulei, + out_dev, extack); + if (err) + return err; + } + break; + case FLOW_ACTION_MIRRED: { + struct net_device *out_dev = act->dev; + + if (mirror_act_count++) { + NL_SET_ERR_MSG_MOD(extack, "Multiple mirror actions per rule are not supported"); + return -EOPNOTSUPP; + } + + err = mlxsw_sp_acl_rulei_act_mirror(mlxsw_sp, rulei, + block, out_dev, + extack); + if (err) + return err; + } + break; + case FLOW_ACTION_VLAN_MANGLE: { + u16 proto = be16_to_cpu(act->vlan.proto); + u8 prio = act->vlan.prio; + u16 vid = act->vlan.vid; + + err = mlxsw_sp_acl_rulei_act_vlan(mlxsw_sp, rulei, + act->id, vid, + proto, prio, extack); + if (err) + return err; + break; + } + case FLOW_ACTION_PRIORITY: + err = mlxsw_sp_acl_rulei_act_priority(mlxsw_sp, rulei, + act->priority, + extack); + if (err) + return err; + break; + case FLOW_ACTION_MANGLE: { + enum flow_action_mangle_base htype = act->mangle.htype; + __be32 be_mask = (__force __be32) act->mangle.mask; + __be32 be_val = (__force __be32) act->mangle.val; + u32 offset = act->mangle.offset; + u32 mask = be32_to_cpu(be_mask); + u32 val = be32_to_cpu(be_val); + + err = mlxsw_sp_acl_rulei_act_mangle(mlxsw_sp, rulei, + htype, offset, + mask, val, extack); + if (err) + return err; + break; + } + case FLOW_ACTION_POLICE: { + u32 burst; + + if (police_act_count++) { + NL_SET_ERR_MSG_MOD(extack, "Multiple police actions per rule are not supported"); + return -EOPNOTSUPP; + } + + err = mlxsw_sp_policer_validate(flow_action, act, extack); + if (err) + return err; + + /* The kernel might adjust the requested burst size so + * that it is not exactly a power of two. Re-adjust it + * here since the hardware only supports burst sizes + * that are a power of two. + */ + burst = roundup_pow_of_two(act->police.burst); + err = mlxsw_sp_acl_rulei_act_police(mlxsw_sp, rulei, + act->hw_index, + act->police.rate_bytes_ps, + burst, extack); + if (err) + return err; + break; + } + case FLOW_ACTION_SAMPLE: { + if (sample_act_count++) { + NL_SET_ERR_MSG_MOD(extack, "Multiple sample actions per rule are not supported"); + return -EOPNOTSUPP; + } + + err = mlxsw_sp_acl_rulei_act_sample(mlxsw_sp, rulei, + block, + act->sample.psample_group, + act->sample.rate, + act->sample.trunc_size, + act->sample.truncate, + extack); + if (err) + return err; + break; + } + default: + NL_SET_ERR_MSG_MOD(extack, "Unsupported action"); + dev_err(mlxsw_sp->bus_info->dev, "Unsupported action\n"); + return -EOPNOTSUPP; + } + } + + if (rulei->ipv6_valid) { + NL_SET_ERR_MSG_MOD(extack, "Unsupported mangle field"); + return -EOPNOTSUPP; + } + + return 0; +} + +static int mlxsw_sp_flower_parse_meta(struct mlxsw_sp_acl_rule_info *rulei, + struct flow_cls_offload *f, + struct mlxsw_sp_flow_block *block) +{ + struct flow_rule *rule = flow_cls_offload_flow_rule(f); + struct mlxsw_sp_port *mlxsw_sp_port; + struct net_device *ingress_dev; + struct flow_match_meta match; + + if (!flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_META)) + return 0; + + flow_rule_match_meta(rule, &match); + if (match.mask->ingress_ifindex != 0xFFFFFFFF) { + NL_SET_ERR_MSG_MOD(f->common.extack, "Unsupported ingress ifindex mask"); + return -EINVAL; + } + + ingress_dev = __dev_get_by_index(block->net, + match.key->ingress_ifindex); + if (!ingress_dev) { + NL_SET_ERR_MSG_MOD(f->common.extack, "Can't find specified ingress port to match on"); + return -EINVAL; + } + + if (!mlxsw_sp_port_dev_check(ingress_dev)) { + NL_SET_ERR_MSG_MOD(f->common.extack, "Can't match on non-mlxsw ingress port"); + return -EINVAL; + } + + mlxsw_sp_port = netdev_priv(ingress_dev); + if (mlxsw_sp_port->mlxsw_sp != block->mlxsw_sp) { + NL_SET_ERR_MSG_MOD(f->common.extack, "Can't match on a port from different device"); + return -EINVAL; + } + + mlxsw_sp_acl_rulei_keymask_u32(rulei, + MLXSW_AFK_ELEMENT_SRC_SYS_PORT, + mlxsw_sp_port->local_port, + 0xFFFFFFFF); + return 0; +} + +static void mlxsw_sp_flower_parse_ipv4(struct mlxsw_sp_acl_rule_info *rulei, + struct flow_cls_offload *f) +{ + struct flow_match_ipv4_addrs match; + + flow_rule_match_ipv4_addrs(f->rule, &match); + + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_SRC_IP_0_31, + (char *) &match.key->src, + (char *) &match.mask->src, 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_DST_IP_0_31, + (char *) &match.key->dst, + (char *) &match.mask->dst, 4); +} + +static void mlxsw_sp_flower_parse_ipv6(struct mlxsw_sp_acl_rule_info *rulei, + struct flow_cls_offload *f) +{ + struct flow_match_ipv6_addrs match; + + flow_rule_match_ipv6_addrs(f->rule, &match); + + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_SRC_IP_96_127, + &match.key->src.s6_addr[0x0], + &match.mask->src.s6_addr[0x0], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_SRC_IP_64_95, + &match.key->src.s6_addr[0x4], + &match.mask->src.s6_addr[0x4], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_SRC_IP_32_63, + &match.key->src.s6_addr[0x8], + &match.mask->src.s6_addr[0x8], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_SRC_IP_0_31, + &match.key->src.s6_addr[0xC], + &match.mask->src.s6_addr[0xC], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_DST_IP_96_127, + &match.key->dst.s6_addr[0x0], + &match.mask->dst.s6_addr[0x0], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_DST_IP_64_95, + &match.key->dst.s6_addr[0x4], + &match.mask->dst.s6_addr[0x4], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_DST_IP_32_63, + &match.key->dst.s6_addr[0x8], + &match.mask->dst.s6_addr[0x8], 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, MLXSW_AFK_ELEMENT_DST_IP_0_31, + &match.key->dst.s6_addr[0xC], + &match.mask->dst.s6_addr[0xC], 4); +} + +static int mlxsw_sp_flower_parse_ports(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct flow_cls_offload *f, + u8 ip_proto) +{ + const struct flow_rule *rule = flow_cls_offload_flow_rule(f); + struct flow_match_ports match; + + if (!flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_PORTS)) + return 0; + + if (ip_proto != IPPROTO_TCP && ip_proto != IPPROTO_UDP) { + NL_SET_ERR_MSG_MOD(f->common.extack, "Only UDP and TCP keys are supported"); + dev_err(mlxsw_sp->bus_info->dev, "Only UDP and TCP keys are supported\n"); + return -EINVAL; + } + + flow_rule_match_ports(rule, &match); + mlxsw_sp_acl_rulei_keymask_u32(rulei, MLXSW_AFK_ELEMENT_DST_L4_PORT, + ntohs(match.key->dst), + ntohs(match.mask->dst)); + mlxsw_sp_acl_rulei_keymask_u32(rulei, MLXSW_AFK_ELEMENT_SRC_L4_PORT, + ntohs(match.key->src), + ntohs(match.mask->src)); + return 0; +} + +static int mlxsw_sp_flower_parse_tcp(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct flow_cls_offload *f, + u8 ip_proto) +{ + const struct flow_rule *rule = flow_cls_offload_flow_rule(f); + struct flow_match_tcp match; + + if (!flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_TCP)) + return 0; + + if (ip_proto != IPPROTO_TCP) { + NL_SET_ERR_MSG_MOD(f->common.extack, "TCP keys supported only for TCP"); + dev_err(mlxsw_sp->bus_info->dev, "TCP keys supported only for TCP\n"); + return -EINVAL; + } + + flow_rule_match_tcp(rule, &match); + + if (match.mask->flags & htons(0x0E00)) { + NL_SET_ERR_MSG_MOD(f->common.extack, "TCP flags match not supported on reserved bits"); + dev_err(mlxsw_sp->bus_info->dev, "TCP flags match not supported on reserved bits\n"); + return -EINVAL; + } + + mlxsw_sp_acl_rulei_keymask_u32(rulei, MLXSW_AFK_ELEMENT_TCP_FLAGS, + ntohs(match.key->flags), + ntohs(match.mask->flags)); + return 0; +} + +static int mlxsw_sp_flower_parse_ip(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct flow_cls_offload *f, + u16 n_proto) +{ + const struct flow_rule *rule = flow_cls_offload_flow_rule(f); + struct flow_match_ip match; + + if (!flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_IP)) + return 0; + + if (n_proto != ETH_P_IP && n_proto != ETH_P_IPV6) { + NL_SET_ERR_MSG_MOD(f->common.extack, "IP keys supported only for IPv4/6"); + dev_err(mlxsw_sp->bus_info->dev, "IP keys supported only for IPv4/6\n"); + return -EINVAL; + } + + flow_rule_match_ip(rule, &match); + + mlxsw_sp_acl_rulei_keymask_u32(rulei, MLXSW_AFK_ELEMENT_IP_TTL_, + match.key->ttl, match.mask->ttl); + + mlxsw_sp_acl_rulei_keymask_u32(rulei, MLXSW_AFK_ELEMENT_IP_ECN, + match.key->tos & 0x3, + match.mask->tos & 0x3); + + mlxsw_sp_acl_rulei_keymask_u32(rulei, MLXSW_AFK_ELEMENT_IP_DSCP, + match.key->tos >> 2, + match.mask->tos >> 2); + + return 0; +} + +static int mlxsw_sp_flower_parse(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_acl_rule_info *rulei, + struct flow_cls_offload *f) +{ + struct flow_rule *rule = flow_cls_offload_flow_rule(f); + struct flow_dissector *dissector = rule->match.dissector; + u16 n_proto_mask = 0; + u16 n_proto_key = 0; + u16 addr_type = 0; + u8 ip_proto = 0; + int err; + + if (dissector->used_keys & + ~(BIT(FLOW_DISSECTOR_KEY_META) | + BIT(FLOW_DISSECTOR_KEY_CONTROL) | + BIT(FLOW_DISSECTOR_KEY_BASIC) | + BIT(FLOW_DISSECTOR_KEY_ETH_ADDRS) | + BIT(FLOW_DISSECTOR_KEY_IPV4_ADDRS) | + BIT(FLOW_DISSECTOR_KEY_IPV6_ADDRS) | + BIT(FLOW_DISSECTOR_KEY_PORTS) | + BIT(FLOW_DISSECTOR_KEY_TCP) | + BIT(FLOW_DISSECTOR_KEY_IP) | + BIT(FLOW_DISSECTOR_KEY_VLAN))) { + dev_err(mlxsw_sp->bus_info->dev, "Unsupported key\n"); + NL_SET_ERR_MSG_MOD(f->common.extack, "Unsupported key"); + return -EOPNOTSUPP; + } + + mlxsw_sp_acl_rulei_priority(rulei, f->common.prio); + + err = mlxsw_sp_flower_parse_meta(rulei, f, block); + if (err) + return err; + + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_CONTROL)) { + struct flow_match_control match; + + flow_rule_match_control(rule, &match); + addr_type = match.key->addr_type; + } + + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_BASIC)) { + struct flow_match_basic match; + + flow_rule_match_basic(rule, &match); + n_proto_key = ntohs(match.key->n_proto); + n_proto_mask = ntohs(match.mask->n_proto); + + if (n_proto_key == ETH_P_ALL) { + n_proto_key = 0; + n_proto_mask = 0; + } + mlxsw_sp_acl_rulei_keymask_u32(rulei, + MLXSW_AFK_ELEMENT_ETHERTYPE, + n_proto_key, n_proto_mask); + + ip_proto = match.key->ip_proto; + mlxsw_sp_acl_rulei_keymask_u32(rulei, + MLXSW_AFK_ELEMENT_IP_PROTO, + match.key->ip_proto, + match.mask->ip_proto); + } + + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ETH_ADDRS)) { + struct flow_match_eth_addrs match; + + flow_rule_match_eth_addrs(rule, &match); + mlxsw_sp_acl_rulei_keymask_buf(rulei, + MLXSW_AFK_ELEMENT_DMAC_32_47, + match.key->dst, + match.mask->dst, 2); + mlxsw_sp_acl_rulei_keymask_buf(rulei, + MLXSW_AFK_ELEMENT_DMAC_0_31, + match.key->dst + 2, + match.mask->dst + 2, 4); + mlxsw_sp_acl_rulei_keymask_buf(rulei, + MLXSW_AFK_ELEMENT_SMAC_32_47, + match.key->src, + match.mask->src, 2); + mlxsw_sp_acl_rulei_keymask_buf(rulei, + MLXSW_AFK_ELEMENT_SMAC_0_31, + match.key->src + 2, + match.mask->src + 2, 4); + } + + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_VLAN)) { + struct flow_match_vlan match; + + flow_rule_match_vlan(rule, &match); + if (mlxsw_sp_flow_block_is_egress_bound(block) && + match.mask->vlan_id) { + NL_SET_ERR_MSG_MOD(f->common.extack, "vlan_id key is not supported on egress"); + return -EOPNOTSUPP; + } + + /* Forbid block with this rulei to be bound + * to egress in future. + */ + rulei->egress_bind_blocker = 1; + + if (match.mask->vlan_id != 0) + mlxsw_sp_acl_rulei_keymask_u32(rulei, + MLXSW_AFK_ELEMENT_VID, + match.key->vlan_id, + match.mask->vlan_id); + if (match.mask->vlan_priority != 0) + mlxsw_sp_acl_rulei_keymask_u32(rulei, + MLXSW_AFK_ELEMENT_PCP, + match.key->vlan_priority, + match.mask->vlan_priority); + } + + if (addr_type == FLOW_DISSECTOR_KEY_IPV4_ADDRS) + mlxsw_sp_flower_parse_ipv4(rulei, f); + + if (addr_type == FLOW_DISSECTOR_KEY_IPV6_ADDRS) + mlxsw_sp_flower_parse_ipv6(rulei, f); + + err = mlxsw_sp_flower_parse_ports(mlxsw_sp, rulei, f, ip_proto); + if (err) + return err; + err = mlxsw_sp_flower_parse_tcp(mlxsw_sp, rulei, f, ip_proto); + if (err) + return err; + + err = mlxsw_sp_flower_parse_ip(mlxsw_sp, rulei, f, n_proto_key & n_proto_mask); + if (err) + return err; + + return mlxsw_sp_flower_parse_actions(mlxsw_sp, block, rulei, + &f->rule->action, + f->common.extack); +} + +static int mlxsw_sp_flower_mall_prio_check(struct mlxsw_sp_flow_block *block, + struct flow_cls_offload *f) +{ + bool ingress = mlxsw_sp_flow_block_is_ingress_bound(block); + unsigned int mall_min_prio; + unsigned int mall_max_prio; + int err; + + err = mlxsw_sp_mall_prio_get(block, f->common.chain_index, + &mall_min_prio, &mall_max_prio); + if (err) { + if (err == -ENOENT) + /* No matchall filters installed on this chain. */ + return 0; + NL_SET_ERR_MSG(f->common.extack, "Failed to get matchall priorities"); + return err; + } + if (ingress && f->common.prio <= mall_min_prio) { + NL_SET_ERR_MSG(f->common.extack, "Failed to add in front of existing matchall rules"); + return -EOPNOTSUPP; + } + if (!ingress && f->common.prio >= mall_max_prio) { + NL_SET_ERR_MSG(f->common.extack, "Failed to add behind of existing matchall rules"); + return -EOPNOTSUPP; + } + return 0; +} + +int mlxsw_sp_flower_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct flow_cls_offload *f) +{ + struct mlxsw_sp_acl_rule_info *rulei; + struct mlxsw_sp_acl_ruleset *ruleset; + struct mlxsw_sp_acl_rule *rule; + int err; + + err = mlxsw_sp_flower_mall_prio_check(block, f); + if (err) + return err; + + ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, block, + f->common.chain_index, + MLXSW_SP_ACL_PROFILE_FLOWER, NULL); + if (IS_ERR(ruleset)) + return PTR_ERR(ruleset); + + rule = mlxsw_sp_acl_rule_create(mlxsw_sp, ruleset, f->cookie, NULL, + f->common.extack); + if (IS_ERR(rule)) { + err = PTR_ERR(rule); + goto err_rule_create; + } + + rulei = mlxsw_sp_acl_rule_rulei(rule); + err = mlxsw_sp_flower_parse(mlxsw_sp, block, rulei, f); + if (err) + goto err_flower_parse; + + err = mlxsw_sp_acl_rulei_commit(rulei); + if (err) + goto err_rulei_commit; + + err = mlxsw_sp_acl_rule_add(mlxsw_sp, rule); + if (err) + goto err_rule_add; + + mlxsw_sp_acl_ruleset_put(mlxsw_sp, ruleset); + return 0; + +err_rule_add: +err_rulei_commit: +err_flower_parse: + mlxsw_sp_acl_rule_destroy(mlxsw_sp, rule); +err_rule_create: + mlxsw_sp_acl_ruleset_put(mlxsw_sp, ruleset); + return err; +} + +void mlxsw_sp_flower_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct flow_cls_offload *f) +{ + struct mlxsw_sp_acl_ruleset *ruleset; + struct mlxsw_sp_acl_rule *rule; + + ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, block, + f->common.chain_index, + MLXSW_SP_ACL_PROFILE_FLOWER, NULL); + if (IS_ERR(ruleset)) + return; + + rule = mlxsw_sp_acl_rule_lookup(mlxsw_sp, ruleset, f->cookie); + if (rule) { + mlxsw_sp_acl_rule_del(mlxsw_sp, rule); + mlxsw_sp_acl_rule_destroy(mlxsw_sp, rule); + } + + mlxsw_sp_acl_ruleset_put(mlxsw_sp, ruleset); +} + +int mlxsw_sp_flower_stats(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct flow_cls_offload *f) +{ + enum flow_action_hw_stats used_hw_stats = FLOW_ACTION_HW_STATS_DISABLED; + struct mlxsw_sp_acl_ruleset *ruleset; + struct mlxsw_sp_acl_rule *rule; + u64 packets; + u64 lastuse; + u64 bytes; + u64 drops; + int err; + + ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, block, + f->common.chain_index, + MLXSW_SP_ACL_PROFILE_FLOWER, NULL); + if (WARN_ON(IS_ERR(ruleset))) + return -EINVAL; + + rule = mlxsw_sp_acl_rule_lookup(mlxsw_sp, ruleset, f->cookie); + if (!rule) + return -EINVAL; + + err = mlxsw_sp_acl_rule_get_stats(mlxsw_sp, rule, &packets, &bytes, + &drops, &lastuse, &used_hw_stats); + if (err) + goto err_rule_get_stats; + + flow_stats_update(&f->stats, bytes, packets, drops, lastuse, + used_hw_stats); + + mlxsw_sp_acl_ruleset_put(mlxsw_sp, ruleset); + return 0; + +err_rule_get_stats: + mlxsw_sp_acl_ruleset_put(mlxsw_sp, ruleset); + return err; +} + +int mlxsw_sp_flower_tmplt_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct flow_cls_offload *f) +{ + struct mlxsw_sp_acl_ruleset *ruleset; + struct mlxsw_sp_acl_rule_info rulei; + int err; + + memset(&rulei, 0, sizeof(rulei)); + err = mlxsw_sp_flower_parse(mlxsw_sp, block, &rulei, f); + if (err) + return err; + ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, block, + f->common.chain_index, + MLXSW_SP_ACL_PROFILE_FLOWER, + &rulei.values.elusage); + + /* keep the reference to the ruleset */ + return PTR_ERR_OR_ZERO(ruleset); +} + +void mlxsw_sp_flower_tmplt_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct flow_cls_offload *f) +{ + struct mlxsw_sp_acl_ruleset *ruleset; + + ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, block, + f->common.chain_index, + MLXSW_SP_ACL_PROFILE_FLOWER, NULL); + if (IS_ERR(ruleset)) + return; + /* put the reference to the ruleset kept in create */ + mlxsw_sp_acl_ruleset_put(mlxsw_sp, ruleset); + mlxsw_sp_acl_ruleset_put(mlxsw_sp, ruleset); +} + +int mlxsw_sp_flower_prio_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + u32 chain_index, unsigned int *p_min_prio, + unsigned int *p_max_prio) +{ + struct mlxsw_sp_acl_ruleset *ruleset; + + ruleset = mlxsw_sp_acl_ruleset_lookup(mlxsw_sp, block, + chain_index, + MLXSW_SP_ACL_PROFILE_FLOWER); + if (IS_ERR(ruleset)) + /* In case there are no flower rules, the caller + * receives -ENOENT to indicate there is no need + * to check the priorities. + */ + return PTR_ERR(ruleset); + mlxsw_sp_acl_ruleset_prio_get(ruleset, p_min_prio, p_max_prio); + return 0; +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c new file mode 100644 index 000000000..a2ee695a3 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c @@ -0,0 +1,676 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <net/ip_tunnels.h> +#include <net/ip6_tunnel.h> +#include <net/inet_ecn.h> + +#include "spectrum_ipip.h" +#include "reg.h" + +struct ip_tunnel_parm +mlxsw_sp_ipip_netdev_parms4(const struct net_device *ol_dev) +{ + struct ip_tunnel *tun = netdev_priv(ol_dev); + + return tun->parms; +} + +struct __ip6_tnl_parm +mlxsw_sp_ipip_netdev_parms6(const struct net_device *ol_dev) +{ + struct ip6_tnl *tun = netdev_priv(ol_dev); + + return tun->parms; +} + +static bool mlxsw_sp_ipip_parms4_has_ikey(const struct ip_tunnel_parm *parms) +{ + return !!(parms->i_flags & TUNNEL_KEY); +} + +static bool mlxsw_sp_ipip_parms6_has_ikey(const struct __ip6_tnl_parm *parms) +{ + return !!(parms->i_flags & TUNNEL_KEY); +} + +static bool mlxsw_sp_ipip_parms4_has_okey(const struct ip_tunnel_parm *parms) +{ + return !!(parms->o_flags & TUNNEL_KEY); +} + +static bool mlxsw_sp_ipip_parms6_has_okey(const struct __ip6_tnl_parm *parms) +{ + return !!(parms->o_flags & TUNNEL_KEY); +} + +static u32 mlxsw_sp_ipip_parms4_ikey(const struct ip_tunnel_parm *parms) +{ + return mlxsw_sp_ipip_parms4_has_ikey(parms) ? + be32_to_cpu(parms->i_key) : 0; +} + +static u32 mlxsw_sp_ipip_parms6_ikey(const struct __ip6_tnl_parm *parms) +{ + return mlxsw_sp_ipip_parms6_has_ikey(parms) ? + be32_to_cpu(parms->i_key) : 0; +} + +static u32 mlxsw_sp_ipip_parms4_okey(const struct ip_tunnel_parm *parms) +{ + return mlxsw_sp_ipip_parms4_has_okey(parms) ? + be32_to_cpu(parms->o_key) : 0; +} + +static u32 mlxsw_sp_ipip_parms6_okey(const struct __ip6_tnl_parm *parms) +{ + return mlxsw_sp_ipip_parms6_has_okey(parms) ? + be32_to_cpu(parms->o_key) : 0; +} + +static union mlxsw_sp_l3addr +mlxsw_sp_ipip_parms4_saddr(const struct ip_tunnel_parm *parms) +{ + return (union mlxsw_sp_l3addr) { .addr4 = parms->iph.saddr }; +} + +static union mlxsw_sp_l3addr +mlxsw_sp_ipip_parms6_saddr(const struct __ip6_tnl_parm *parms) +{ + return (union mlxsw_sp_l3addr) { .addr6 = parms->laddr }; +} + +static union mlxsw_sp_l3addr +mlxsw_sp_ipip_parms4_daddr(const struct ip_tunnel_parm *parms) +{ + return (union mlxsw_sp_l3addr) { .addr4 = parms->iph.daddr }; +} + +static union mlxsw_sp_l3addr +mlxsw_sp_ipip_parms6_daddr(const struct __ip6_tnl_parm *parms) +{ + return (union mlxsw_sp_l3addr) { .addr6 = parms->raddr }; +} + +union mlxsw_sp_l3addr +mlxsw_sp_ipip_netdev_saddr(enum mlxsw_sp_l3proto proto, + const struct net_device *ol_dev) +{ + struct ip_tunnel_parm parms4; + struct __ip6_tnl_parm parms6; + + switch (proto) { + case MLXSW_SP_L3_PROTO_IPV4: + parms4 = mlxsw_sp_ipip_netdev_parms4(ol_dev); + return mlxsw_sp_ipip_parms4_saddr(&parms4); + case MLXSW_SP_L3_PROTO_IPV6: + parms6 = mlxsw_sp_ipip_netdev_parms6(ol_dev); + return mlxsw_sp_ipip_parms6_saddr(&parms6); + } + + WARN_ON(1); + return (union mlxsw_sp_l3addr) {0}; +} + +static __be32 mlxsw_sp_ipip_netdev_daddr4(const struct net_device *ol_dev) +{ + + struct ip_tunnel_parm parms4 = mlxsw_sp_ipip_netdev_parms4(ol_dev); + + return mlxsw_sp_ipip_parms4_daddr(&parms4).addr4; +} + +static union mlxsw_sp_l3addr +mlxsw_sp_ipip_netdev_daddr(enum mlxsw_sp_l3proto proto, + const struct net_device *ol_dev) +{ + struct ip_tunnel_parm parms4; + struct __ip6_tnl_parm parms6; + + switch (proto) { + case MLXSW_SP_L3_PROTO_IPV4: + parms4 = mlxsw_sp_ipip_netdev_parms4(ol_dev); + return mlxsw_sp_ipip_parms4_daddr(&parms4); + case MLXSW_SP_L3_PROTO_IPV6: + parms6 = mlxsw_sp_ipip_netdev_parms6(ol_dev); + return mlxsw_sp_ipip_parms6_daddr(&parms6); + } + + WARN_ON(1); + return (union mlxsw_sp_l3addr) {0}; +} + +bool mlxsw_sp_l3addr_is_zero(union mlxsw_sp_l3addr addr) +{ + union mlxsw_sp_l3addr naddr = {0}; + + return !memcmp(&addr, &naddr, sizeof(naddr)); +} + +static struct mlxsw_sp_ipip_parms +mlxsw_sp_ipip_netdev_parms_init_gre4(const struct net_device *ol_dev) +{ + struct ip_tunnel_parm parms = mlxsw_sp_ipip_netdev_parms4(ol_dev); + + return (struct mlxsw_sp_ipip_parms) { + .proto = MLXSW_SP_L3_PROTO_IPV4, + .saddr = mlxsw_sp_ipip_parms4_saddr(&parms), + .daddr = mlxsw_sp_ipip_parms4_daddr(&parms), + .link = parms.link, + .ikey = mlxsw_sp_ipip_parms4_ikey(&parms), + .okey = mlxsw_sp_ipip_parms4_okey(&parms), + }; +} + +static int +mlxsw_sp_ipip_nexthop_update_gre4(struct mlxsw_sp *mlxsw_sp, u32 adj_index, + struct mlxsw_sp_ipip_entry *ipip_entry, + bool force, char *ratr_pl) +{ + u16 rif_index = mlxsw_sp_ipip_lb_rif_index(ipip_entry->ol_lb); + __be32 daddr4 = mlxsw_sp_ipip_netdev_daddr4(ipip_entry->ol_dev); + enum mlxsw_reg_ratr_op op; + + op = force ? MLXSW_REG_RATR_OP_WRITE_WRITE_ENTRY : + MLXSW_REG_RATR_OP_WRITE_WRITE_ENTRY_ON_ACTIVITY; + mlxsw_reg_ratr_pack(ratr_pl, op, true, MLXSW_REG_RATR_TYPE_IPIP, + adj_index, rif_index); + mlxsw_reg_ratr_ipip4_entry_pack(ratr_pl, be32_to_cpu(daddr4)); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ratr), ratr_pl); +} + +static int +mlxsw_sp_ipip_decap_config_gre4(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + u32 tunnel_index) +{ + u16 rif_index = mlxsw_sp_ipip_lb_rif_index(ipip_entry->ol_lb); + u16 ul_rif_id = mlxsw_sp_ipip_lb_ul_rif_id(ipip_entry->ol_lb); + char rtdp_pl[MLXSW_REG_RTDP_LEN]; + struct ip_tunnel_parm parms; + unsigned int type_check; + bool has_ikey; + u32 daddr4; + u32 ikey; + + parms = mlxsw_sp_ipip_netdev_parms4(ipip_entry->ol_dev); + has_ikey = mlxsw_sp_ipip_parms4_has_ikey(&parms); + ikey = mlxsw_sp_ipip_parms4_ikey(&parms); + + mlxsw_reg_rtdp_pack(rtdp_pl, MLXSW_REG_RTDP_TYPE_IPIP, tunnel_index); + mlxsw_reg_rtdp_egress_router_interface_set(rtdp_pl, ul_rif_id); + + type_check = has_ikey ? + MLXSW_REG_RTDP_IPIP_TYPE_CHECK_ALLOW_GRE_KEY : + MLXSW_REG_RTDP_IPIP_TYPE_CHECK_ALLOW_GRE; + + /* Linux demuxes tunnels based on packet SIP (which must match tunnel + * remote IP). Thus configure decap so that it filters out packets that + * are not IPv4 or have the wrong SIP. IPIP_DECAP_ERROR trap is + * generated for packets that fail this criterion. Linux then handles + * such packets in slow path and generates ICMP destination unreachable. + */ + daddr4 = be32_to_cpu(mlxsw_sp_ipip_netdev_daddr4(ipip_entry->ol_dev)); + mlxsw_reg_rtdp_ipip4_pack(rtdp_pl, rif_index, + MLXSW_REG_RTDP_IPIP_SIP_CHECK_FILTER_IPV4, + type_check, has_ikey, daddr4, ikey); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rtdp), rtdp_pl); +} + +static bool mlxsw_sp_ipip_tunnel_complete(enum mlxsw_sp_l3proto proto, + const struct net_device *ol_dev) +{ + union mlxsw_sp_l3addr saddr = mlxsw_sp_ipip_netdev_saddr(proto, ol_dev); + union mlxsw_sp_l3addr daddr = mlxsw_sp_ipip_netdev_daddr(proto, ol_dev); + + /* Tunnels with unset local or remote address are valid in Linux and + * used for lightweight tunnels (LWT) and Non-Broadcast Multi-Access + * (NBMA) tunnels. In principle these can be offloaded, but the driver + * currently doesn't support this. So punt. + */ + return !mlxsw_sp_l3addr_is_zero(saddr) && + !mlxsw_sp_l3addr_is_zero(daddr); +} + +static bool mlxsw_sp_ipip_can_offload_gre4(const struct mlxsw_sp *mlxsw_sp, + const struct net_device *ol_dev) +{ + struct ip_tunnel *tunnel = netdev_priv(ol_dev); + __be16 okflags = TUNNEL_KEY; /* We can't offload any other features. */ + bool inherit_ttl = tunnel->parms.iph.ttl == 0; + bool inherit_tos = tunnel->parms.iph.tos & 0x1; + + return (tunnel->parms.i_flags & ~okflags) == 0 && + (tunnel->parms.o_flags & ~okflags) == 0 && + inherit_ttl && inherit_tos && + mlxsw_sp_ipip_tunnel_complete(MLXSW_SP_L3_PROTO_IPV4, ol_dev); +} + +static struct mlxsw_sp_rif_ipip_lb_config +mlxsw_sp_ipip_ol_loopback_config_gre4(struct mlxsw_sp *mlxsw_sp, + const struct net_device *ol_dev) +{ + struct ip_tunnel_parm parms = mlxsw_sp_ipip_netdev_parms4(ol_dev); + enum mlxsw_reg_ritr_loopback_ipip_type lb_ipipt; + + lb_ipipt = mlxsw_sp_ipip_parms4_has_okey(&parms) ? + MLXSW_REG_RITR_LOOPBACK_IPIP_TYPE_IP_IN_GRE_KEY_IN_IP : + MLXSW_REG_RITR_LOOPBACK_IPIP_TYPE_IP_IN_GRE_IN_IP; + return (struct mlxsw_sp_rif_ipip_lb_config){ + .lb_ipipt = lb_ipipt, + .okey = mlxsw_sp_ipip_parms4_okey(&parms), + .ul_protocol = MLXSW_SP_L3_PROTO_IPV4, + .saddr = mlxsw_sp_ipip_netdev_saddr(MLXSW_SP_L3_PROTO_IPV4, + ol_dev), + }; +} + +static int +mlxsw_sp_ipip_ol_netdev_change_gre(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + const struct mlxsw_sp_ipip_parms *new_parms, + struct netlink_ext_ack *extack) +{ + const struct mlxsw_sp_ipip_parms *old_parms = &ipip_entry->parms; + bool update_tunnel = false; + bool update_decap = false; + bool update_nhs = false; + int err = 0; + + if (!mlxsw_sp_l3addr_eq(&new_parms->saddr, &old_parms->saddr)) { + u16 ul_tb_id = mlxsw_sp_ipip_dev_ul_tb_id(ipip_entry->ol_dev); + + /* Since the local address has changed, if there is another + * tunnel with a matching saddr, both need to be demoted. + */ + if (mlxsw_sp_ipip_demote_tunnel_by_saddr(mlxsw_sp, + new_parms->proto, + new_parms->saddr, + ul_tb_id, + ipip_entry)) { + mlxsw_sp_ipip_entry_demote_tunnel(mlxsw_sp, ipip_entry); + return 0; + } + + update_tunnel = true; + } else if (old_parms->okey != new_parms->okey || + old_parms->link != new_parms->link) { + update_tunnel = true; + } else if (!mlxsw_sp_l3addr_eq(&new_parms->daddr, &old_parms->daddr)) { + update_nhs = true; + } else if (old_parms->ikey != new_parms->ikey) { + update_decap = true; + } + + if (update_tunnel) + err = __mlxsw_sp_ipip_entry_update_tunnel(mlxsw_sp, ipip_entry, + true, true, true, + extack); + else if (update_nhs) + err = __mlxsw_sp_ipip_entry_update_tunnel(mlxsw_sp, ipip_entry, + false, false, true, + extack); + else if (update_decap) + err = __mlxsw_sp_ipip_entry_update_tunnel(mlxsw_sp, ipip_entry, + false, false, false, + extack); + if (err) + return err; + + ipip_entry->parms = *new_parms; + return 0; +} + +static int +mlxsw_sp_ipip_ol_netdev_change_gre4(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_ipip_parms new_parms; + + new_parms = mlxsw_sp_ipip_netdev_parms_init_gre4(ipip_entry->ol_dev); + return mlxsw_sp_ipip_ol_netdev_change_gre(mlxsw_sp, ipip_entry, + &new_parms, extack); +} + +static int +mlxsw_sp_ipip_rem_addr_set_gre4(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + return 0; +} + +static void +mlxsw_sp_ipip_rem_addr_unset_gre4(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_ipip_entry *ipip_entry) +{ +} + +static const struct mlxsw_sp_ipip_ops mlxsw_sp_ipip_gre4_ops = { + .dev_type = ARPHRD_IPGRE, + .ul_proto = MLXSW_SP_L3_PROTO_IPV4, + .inc_parsing_depth = false, + .parms_init = mlxsw_sp_ipip_netdev_parms_init_gre4, + .nexthop_update = mlxsw_sp_ipip_nexthop_update_gre4, + .decap_config = mlxsw_sp_ipip_decap_config_gre4, + .can_offload = mlxsw_sp_ipip_can_offload_gre4, + .ol_loopback_config = mlxsw_sp_ipip_ol_loopback_config_gre4, + .ol_netdev_change = mlxsw_sp_ipip_ol_netdev_change_gre4, + .rem_ip_addr_set = mlxsw_sp_ipip_rem_addr_set_gre4, + .rem_ip_addr_unset = mlxsw_sp_ipip_rem_addr_unset_gre4, +}; + +static struct mlxsw_sp_ipip_parms +mlxsw_sp1_ipip_netdev_parms_init_gre6(const struct net_device *ol_dev) +{ + struct mlxsw_sp_ipip_parms parms = {0}; + + WARN_ON_ONCE(1); + return parms; +} + +static int +mlxsw_sp1_ipip_nexthop_update_gre6(struct mlxsw_sp *mlxsw_sp, u32 adj_index, + struct mlxsw_sp_ipip_entry *ipip_entry, + bool force, char *ratr_pl) +{ + WARN_ON_ONCE(1); + return -EINVAL; +} + +static int +mlxsw_sp1_ipip_decap_config_gre6(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + u32 tunnel_index) +{ + WARN_ON_ONCE(1); + return -EINVAL; +} + +static bool mlxsw_sp1_ipip_can_offload_gre6(const struct mlxsw_sp *mlxsw_sp, + const struct net_device *ol_dev) +{ + return false; +} + +static struct mlxsw_sp_rif_ipip_lb_config +mlxsw_sp1_ipip_ol_loopback_config_gre6(struct mlxsw_sp *mlxsw_sp, + const struct net_device *ol_dev) +{ + struct mlxsw_sp_rif_ipip_lb_config config = {0}; + + WARN_ON_ONCE(1); + return config; +} + +static int +mlxsw_sp1_ipip_ol_netdev_change_gre6(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + struct netlink_ext_ack *extack) +{ + WARN_ON_ONCE(1); + return -EINVAL; +} + +static int +mlxsw_sp1_ipip_rem_addr_set_gre6(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + WARN_ON_ONCE(1); + return -EINVAL; +} + +static void +mlxsw_sp1_ipip_rem_addr_unset_gre6(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_ipip_entry *ipip_entry) +{ + WARN_ON_ONCE(1); +} + +static const struct mlxsw_sp_ipip_ops mlxsw_sp1_ipip_gre6_ops = { + .dev_type = ARPHRD_IP6GRE, + .ul_proto = MLXSW_SP_L3_PROTO_IPV6, + .inc_parsing_depth = true, + .parms_init = mlxsw_sp1_ipip_netdev_parms_init_gre6, + .nexthop_update = mlxsw_sp1_ipip_nexthop_update_gre6, + .decap_config = mlxsw_sp1_ipip_decap_config_gre6, + .can_offload = mlxsw_sp1_ipip_can_offload_gre6, + .ol_loopback_config = mlxsw_sp1_ipip_ol_loopback_config_gre6, + .ol_netdev_change = mlxsw_sp1_ipip_ol_netdev_change_gre6, + .rem_ip_addr_set = mlxsw_sp1_ipip_rem_addr_set_gre6, + .rem_ip_addr_unset = mlxsw_sp1_ipip_rem_addr_unset_gre6, +}; + +const struct mlxsw_sp_ipip_ops *mlxsw_sp1_ipip_ops_arr[] = { + [MLXSW_SP_IPIP_TYPE_GRE4] = &mlxsw_sp_ipip_gre4_ops, + [MLXSW_SP_IPIP_TYPE_GRE6] = &mlxsw_sp1_ipip_gre6_ops, +}; + +static struct mlxsw_sp_ipip_parms +mlxsw_sp2_ipip_netdev_parms_init_gre6(const struct net_device *ol_dev) +{ + struct __ip6_tnl_parm parms = mlxsw_sp_ipip_netdev_parms6(ol_dev); + + return (struct mlxsw_sp_ipip_parms) { + .proto = MLXSW_SP_L3_PROTO_IPV6, + .saddr = mlxsw_sp_ipip_parms6_saddr(&parms), + .daddr = mlxsw_sp_ipip_parms6_daddr(&parms), + .link = parms.link, + .ikey = mlxsw_sp_ipip_parms6_ikey(&parms), + .okey = mlxsw_sp_ipip_parms6_okey(&parms), + }; +} + +static int +mlxsw_sp2_ipip_nexthop_update_gre6(struct mlxsw_sp *mlxsw_sp, u32 adj_index, + struct mlxsw_sp_ipip_entry *ipip_entry, + bool force, char *ratr_pl) +{ + u16 rif_index = mlxsw_sp_ipip_lb_rif_index(ipip_entry->ol_lb); + enum mlxsw_reg_ratr_op op; + + op = force ? MLXSW_REG_RATR_OP_WRITE_WRITE_ENTRY : + MLXSW_REG_RATR_OP_WRITE_WRITE_ENTRY_ON_ACTIVITY; + mlxsw_reg_ratr_pack(ratr_pl, op, true, MLXSW_REG_RATR_TYPE_IPIP, + adj_index, rif_index); + mlxsw_reg_ratr_ipip6_entry_pack(ratr_pl, + ipip_entry->dip_kvdl_index); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ratr), ratr_pl); +} + +static int +mlxsw_sp2_ipip_decap_config_gre6(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + u32 tunnel_index) +{ + u16 rif_index = mlxsw_sp_ipip_lb_rif_index(ipip_entry->ol_lb); + u16 ul_rif_id = mlxsw_sp_ipip_lb_ul_rif_id(ipip_entry->ol_lb); + char rtdp_pl[MLXSW_REG_RTDP_LEN]; + struct __ip6_tnl_parm parms; + unsigned int type_check; + bool has_ikey; + u32 ikey; + + parms = mlxsw_sp_ipip_netdev_parms6(ipip_entry->ol_dev); + has_ikey = mlxsw_sp_ipip_parms6_has_ikey(&parms); + ikey = mlxsw_sp_ipip_parms6_ikey(&parms); + + mlxsw_reg_rtdp_pack(rtdp_pl, MLXSW_REG_RTDP_TYPE_IPIP, tunnel_index); + mlxsw_reg_rtdp_egress_router_interface_set(rtdp_pl, ul_rif_id); + + type_check = has_ikey ? + MLXSW_REG_RTDP_IPIP_TYPE_CHECK_ALLOW_GRE_KEY : + MLXSW_REG_RTDP_IPIP_TYPE_CHECK_ALLOW_GRE; + + /* Linux demuxes tunnels based on packet SIP (which must match tunnel + * remote IP). Thus configure decap so that it filters out packets that + * are not IPv6 or have the wrong SIP. IPIP_DECAP_ERROR trap is + * generated for packets that fail this criterion. Linux then handles + * such packets in slow path and generates ICMP destination unreachable. + */ + mlxsw_reg_rtdp_ipip6_pack(rtdp_pl, rif_index, + MLXSW_REG_RTDP_IPIP_SIP_CHECK_FILTER_IPV6, + type_check, has_ikey, + ipip_entry->dip_kvdl_index, ikey); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rtdp), rtdp_pl); +} + +static bool mlxsw_sp2_ipip_can_offload_gre6(const struct mlxsw_sp *mlxsw_sp, + const struct net_device *ol_dev) +{ + struct __ip6_tnl_parm tparm = mlxsw_sp_ipip_netdev_parms6(ol_dev); + bool inherit_tos = tparm.flags & IP6_TNL_F_USE_ORIG_TCLASS; + bool inherit_ttl = tparm.hop_limit == 0; + __be16 okflags = TUNNEL_KEY; /* We can't offload any other features. */ + + return (tparm.i_flags & ~okflags) == 0 && + (tparm.o_flags & ~okflags) == 0 && + inherit_ttl && inherit_tos && + mlxsw_sp_ipip_tunnel_complete(MLXSW_SP_L3_PROTO_IPV6, ol_dev); +} + +static struct mlxsw_sp_rif_ipip_lb_config +mlxsw_sp2_ipip_ol_loopback_config_gre6(struct mlxsw_sp *mlxsw_sp, + const struct net_device *ol_dev) +{ + struct __ip6_tnl_parm parms = mlxsw_sp_ipip_netdev_parms6(ol_dev); + enum mlxsw_reg_ritr_loopback_ipip_type lb_ipipt; + + lb_ipipt = mlxsw_sp_ipip_parms6_has_okey(&parms) ? + MLXSW_REG_RITR_LOOPBACK_IPIP_TYPE_IP_IN_GRE_KEY_IN_IP : + MLXSW_REG_RITR_LOOPBACK_IPIP_TYPE_IP_IN_GRE_IN_IP; + return (struct mlxsw_sp_rif_ipip_lb_config){ + .lb_ipipt = lb_ipipt, + .okey = mlxsw_sp_ipip_parms6_okey(&parms), + .ul_protocol = MLXSW_SP_L3_PROTO_IPV6, + .saddr = mlxsw_sp_ipip_netdev_saddr(MLXSW_SP_L3_PROTO_IPV6, + ol_dev), + }; +} + +static int +mlxsw_sp2_ipip_ol_netdev_change_gre6(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_ipip_parms new_parms; + + new_parms = mlxsw_sp2_ipip_netdev_parms_init_gre6(ipip_entry->ol_dev); + return mlxsw_sp_ipip_ol_netdev_change_gre(mlxsw_sp, ipip_entry, + &new_parms, extack); +} + +static int +mlxsw_sp2_ipip_rem_addr_set_gre6(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + return mlxsw_sp_ipv6_addr_kvdl_index_get(mlxsw_sp, + &ipip_entry->parms.daddr.addr6, + &ipip_entry->dip_kvdl_index); +} + +static void +mlxsw_sp2_ipip_rem_addr_unset_gre6(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_ipip_entry *ipip_entry) +{ + mlxsw_sp_ipv6_addr_put(mlxsw_sp, &ipip_entry->parms.daddr.addr6); +} + +static const struct mlxsw_sp_ipip_ops mlxsw_sp2_ipip_gre6_ops = { + .dev_type = ARPHRD_IP6GRE, + .ul_proto = MLXSW_SP_L3_PROTO_IPV6, + .inc_parsing_depth = true, + .parms_init = mlxsw_sp2_ipip_netdev_parms_init_gre6, + .nexthop_update = mlxsw_sp2_ipip_nexthop_update_gre6, + .decap_config = mlxsw_sp2_ipip_decap_config_gre6, + .can_offload = mlxsw_sp2_ipip_can_offload_gre6, + .ol_loopback_config = mlxsw_sp2_ipip_ol_loopback_config_gre6, + .ol_netdev_change = mlxsw_sp2_ipip_ol_netdev_change_gre6, + .rem_ip_addr_set = mlxsw_sp2_ipip_rem_addr_set_gre6, + .rem_ip_addr_unset = mlxsw_sp2_ipip_rem_addr_unset_gre6, +}; + +const struct mlxsw_sp_ipip_ops *mlxsw_sp2_ipip_ops_arr[] = { + [MLXSW_SP_IPIP_TYPE_GRE4] = &mlxsw_sp_ipip_gre4_ops, + [MLXSW_SP_IPIP_TYPE_GRE6] = &mlxsw_sp2_ipip_gre6_ops, +}; + +static int mlxsw_sp_ipip_ecn_encap_init_one(struct mlxsw_sp *mlxsw_sp, + u8 inner_ecn, u8 outer_ecn) +{ + char tieem_pl[MLXSW_REG_TIEEM_LEN]; + + mlxsw_reg_tieem_pack(tieem_pl, inner_ecn, outer_ecn); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tieem), tieem_pl); +} + +int mlxsw_sp_ipip_ecn_encap_init(struct mlxsw_sp *mlxsw_sp) +{ + int i; + + /* Iterate over inner ECN values */ + for (i = INET_ECN_NOT_ECT; i <= INET_ECN_CE; i++) { + u8 outer_ecn = INET_ECN_encapsulate(0, i); + int err; + + err = mlxsw_sp_ipip_ecn_encap_init_one(mlxsw_sp, i, outer_ecn); + if (err) + return err; + } + + return 0; +} + +static int mlxsw_sp_ipip_ecn_decap_init_one(struct mlxsw_sp *mlxsw_sp, + u8 inner_ecn, u8 outer_ecn) +{ + char tidem_pl[MLXSW_REG_TIDEM_LEN]; + u8 new_inner_ecn; + bool trap_en; + + new_inner_ecn = mlxsw_sp_tunnel_ecn_decap(outer_ecn, inner_ecn, + &trap_en); + mlxsw_reg_tidem_pack(tidem_pl, outer_ecn, inner_ecn, new_inner_ecn, + trap_en, trap_en ? MLXSW_TRAP_ID_DECAP_ECN0 : 0); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tidem), tidem_pl); +} + +int mlxsw_sp_ipip_ecn_decap_init(struct mlxsw_sp *mlxsw_sp) +{ + int i, j, err; + + /* Iterate over inner ECN values */ + for (i = INET_ECN_NOT_ECT; i <= INET_ECN_CE; i++) { + /* Iterate over outer ECN values */ + for (j = INET_ECN_NOT_ECT; j <= INET_ECN_CE; j++) { + err = mlxsw_sp_ipip_ecn_decap_init_one(mlxsw_sp, i, j); + if (err) + return err; + } + } + + return 0; +} + +struct net_device * +mlxsw_sp_ipip_netdev_ul_dev_get(const struct net_device *ol_dev) +{ + struct net *net = dev_net(ol_dev); + struct ip_tunnel *tun4; + struct ip6_tnl *tun6; + + switch (ol_dev->type) { + case ARPHRD_IPGRE: + tun4 = netdev_priv(ol_dev); + return dev_get_by_index_rcu(net, tun4->parms.link); + case ARPHRD_IP6GRE: + tun6 = netdev_priv(ol_dev); + return dev_get_by_index_rcu(net, tun6->parms.link); + default: + return NULL; + } +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h new file mode 100644 index 000000000..8cc259dcc --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h @@ -0,0 +1,84 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_IPIP_H_ +#define _MLXSW_IPIP_H_ + +#include "spectrum_router.h" +#include <net/ip_fib.h> +#include <linux/if_tunnel.h> +#include <net/ip6_tunnel.h> + +struct ip_tunnel_parm +mlxsw_sp_ipip_netdev_parms4(const struct net_device *ol_dev); +struct __ip6_tnl_parm +mlxsw_sp_ipip_netdev_parms6(const struct net_device *ol_dev); + +union mlxsw_sp_l3addr +mlxsw_sp_ipip_netdev_saddr(enum mlxsw_sp_l3proto proto, + const struct net_device *ol_dev); + +bool mlxsw_sp_l3addr_is_zero(union mlxsw_sp_l3addr addr); + +enum mlxsw_sp_ipip_type { + MLXSW_SP_IPIP_TYPE_GRE4, + MLXSW_SP_IPIP_TYPE_GRE6, + MLXSW_SP_IPIP_TYPE_MAX, +}; + +struct mlxsw_sp_ipip_parms { + enum mlxsw_sp_l3proto proto; + union mlxsw_sp_l3addr saddr; + union mlxsw_sp_l3addr daddr; + int link; + u32 ikey; + u32 okey; +}; + +struct mlxsw_sp_ipip_entry { + enum mlxsw_sp_ipip_type ipipt; + struct net_device *ol_dev; /* Overlay. */ + struct mlxsw_sp_rif_ipip_lb *ol_lb; + struct mlxsw_sp_fib_entry *decap_fib_entry; + struct list_head ipip_list_node; + struct mlxsw_sp_ipip_parms parms; + u32 dip_kvdl_index; +}; + +struct mlxsw_sp_ipip_ops { + int dev_type; + enum mlxsw_sp_l3proto ul_proto; /* Underlay. */ + bool inc_parsing_depth; + + struct mlxsw_sp_ipip_parms + (*parms_init)(const struct net_device *ol_dev); + + int (*nexthop_update)(struct mlxsw_sp *mlxsw_sp, u32 adj_index, + struct mlxsw_sp_ipip_entry *ipip_entry, + bool force, char *ratr_pl); + + bool (*can_offload)(const struct mlxsw_sp *mlxsw_sp, + const struct net_device *ol_dev); + + /* Return a configuration for creating an overlay loopback RIF. */ + struct mlxsw_sp_rif_ipip_lb_config + (*ol_loopback_config)(struct mlxsw_sp *mlxsw_sp, + const struct net_device *ol_dev); + + int (*decap_config)(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + u32 tunnel_index); + + int (*ol_netdev_change)(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + struct netlink_ext_ack *extack); + int (*rem_ip_addr_set)(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry); + void (*rem_ip_addr_unset)(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_ipip_entry *ipip_entry); +}; + +extern const struct mlxsw_sp_ipip_ops *mlxsw_sp1_ipip_ops_arr[]; +extern const struct mlxsw_sp_ipip_ops *mlxsw_sp2_ipip_ops_arr[]; + +#endif /* _MLXSW_IPIP_H_*/ diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_kvdl.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_kvdl.c new file mode 100644 index 000000000..20d72f1c0 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_kvdl.c @@ -0,0 +1,87 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2016-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/mutex.h> +#include <linux/slab.h> + +#include "spectrum.h" + +struct mlxsw_sp_kvdl { + const struct mlxsw_sp_kvdl_ops *kvdl_ops; + struct mutex kvdl_lock; /* Protects kvdl allocations */ + unsigned long priv[]; + /* priv has to be always the last item */ +}; + +int mlxsw_sp_kvdl_init(struct mlxsw_sp *mlxsw_sp) +{ + const struct mlxsw_sp_kvdl_ops *kvdl_ops = mlxsw_sp->kvdl_ops; + struct mlxsw_sp_kvdl *kvdl; + int err; + + kvdl = kzalloc(sizeof(*mlxsw_sp->kvdl) + kvdl_ops->priv_size, + GFP_KERNEL); + if (!kvdl) + return -ENOMEM; + mutex_init(&kvdl->kvdl_lock); + kvdl->kvdl_ops = kvdl_ops; + mlxsw_sp->kvdl = kvdl; + + err = kvdl_ops->init(mlxsw_sp, kvdl->priv); + if (err) + goto err_init; + return 0; + +err_init: + mutex_destroy(&kvdl->kvdl_lock); + kfree(kvdl); + return err; +} + +void mlxsw_sp_kvdl_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_kvdl *kvdl = mlxsw_sp->kvdl; + + kvdl->kvdl_ops->fini(mlxsw_sp, kvdl->priv); + mutex_destroy(&kvdl->kvdl_lock); + kfree(kvdl); +} + +int mlxsw_sp_kvdl_alloc(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, u32 *p_entry_index) +{ + struct mlxsw_sp_kvdl *kvdl = mlxsw_sp->kvdl; + int err; + + mutex_lock(&kvdl->kvdl_lock); + err = kvdl->kvdl_ops->alloc(mlxsw_sp, kvdl->priv, type, + entry_count, p_entry_index); + mutex_unlock(&kvdl->kvdl_lock); + + return err; +} + +void mlxsw_sp_kvdl_free(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, int entry_index) +{ + struct mlxsw_sp_kvdl *kvdl = mlxsw_sp->kvdl; + + mutex_lock(&kvdl->kvdl_lock); + kvdl->kvdl_ops->free(mlxsw_sp, kvdl->priv, type, + entry_count, entry_index); + mutex_unlock(&kvdl->kvdl_lock); +} + +int mlxsw_sp_kvdl_alloc_count_query(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_kvdl_entry_type type, + unsigned int entry_count, + unsigned int *p_alloc_count) +{ + struct mlxsw_sp_kvdl *kvdl = mlxsw_sp->kvdl; + + return kvdl->kvdl_ops->alloc_size_query(mlxsw_sp, kvdl->priv, type, + entry_count, p_alloc_count); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_matchall.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_matchall.c new file mode 100644 index 000000000..07b371cd9 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_matchall.c @@ -0,0 +1,478 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2020 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/netdevice.h> +#include <net/flow_offload.h> + +#include "spectrum.h" +#include "spectrum_span.h" +#include "reg.h" + +static struct mlxsw_sp_mall_entry * +mlxsw_sp_mall_entry_find(struct mlxsw_sp_flow_block *block, unsigned long cookie) +{ + struct mlxsw_sp_mall_entry *mall_entry; + + list_for_each_entry(mall_entry, &block->mall.list, list) + if (mall_entry->cookie == cookie) + return mall_entry; + + return NULL; +} + +static int +mlxsw_sp_mall_port_mirror_add(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_mall_entry *mall_entry, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_span_agent_parms agent_parms = {}; + struct mlxsw_sp_span_trigger_parms parms; + enum mlxsw_sp_span_trigger trigger; + int err; + + if (!mall_entry->mirror.to_dev) { + NL_SET_ERR_MSG(extack, "Could not find requested device"); + return -EINVAL; + } + + agent_parms.to_dev = mall_entry->mirror.to_dev; + err = mlxsw_sp_span_agent_get(mlxsw_sp, &mall_entry->mirror.span_id, + &agent_parms); + if (err) { + NL_SET_ERR_MSG(extack, "Failed to get SPAN agent"); + return err; + } + + err = mlxsw_sp_span_analyzed_port_get(mlxsw_sp_port, + mall_entry->ingress); + if (err) { + NL_SET_ERR_MSG(extack, "Failed to get analyzed port"); + goto err_analyzed_port_get; + } + + trigger = mall_entry->ingress ? MLXSW_SP_SPAN_TRIGGER_INGRESS : + MLXSW_SP_SPAN_TRIGGER_EGRESS; + parms.span_id = mall_entry->mirror.span_id; + parms.probability_rate = 1; + err = mlxsw_sp_span_agent_bind(mlxsw_sp, trigger, mlxsw_sp_port, + &parms); + if (err) { + NL_SET_ERR_MSG(extack, "Failed to bind SPAN agent"); + goto err_agent_bind; + } + + return 0; + +err_agent_bind: + mlxsw_sp_span_analyzed_port_put(mlxsw_sp_port, mall_entry->ingress); +err_analyzed_port_get: + mlxsw_sp_span_agent_put(mlxsw_sp, mall_entry->mirror.span_id); + return err; +} + +static void +mlxsw_sp_mall_port_mirror_del(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_mall_entry *mall_entry) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_span_trigger_parms parms; + enum mlxsw_sp_span_trigger trigger; + + trigger = mall_entry->ingress ? MLXSW_SP_SPAN_TRIGGER_INGRESS : + MLXSW_SP_SPAN_TRIGGER_EGRESS; + parms.span_id = mall_entry->mirror.span_id; + mlxsw_sp_span_agent_unbind(mlxsw_sp, trigger, mlxsw_sp_port, &parms); + mlxsw_sp_span_analyzed_port_put(mlxsw_sp_port, mall_entry->ingress); + mlxsw_sp_span_agent_put(mlxsw_sp, mall_entry->mirror.span_id); +} + +static int mlxsw_sp_mall_port_sample_set(struct mlxsw_sp_port *mlxsw_sp_port, + bool enable, u32 rate) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char mpsc_pl[MLXSW_REG_MPSC_LEN]; + + mlxsw_reg_mpsc_pack(mpsc_pl, mlxsw_sp_port->local_port, enable, rate); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mpsc), mpsc_pl); +} + +static int +mlxsw_sp_mall_port_sample_add(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_mall_entry *mall_entry, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_sample_trigger trigger; + int err; + + if (mall_entry->ingress) + trigger.type = MLXSW_SP_SAMPLE_TRIGGER_TYPE_INGRESS; + else + trigger.type = MLXSW_SP_SAMPLE_TRIGGER_TYPE_EGRESS; + trigger.local_port = mlxsw_sp_port->local_port; + err = mlxsw_sp_sample_trigger_params_set(mlxsw_sp, &trigger, + &mall_entry->sample.params, + extack); + if (err) + return err; + + err = mlxsw_sp->mall_ops->sample_add(mlxsw_sp, mlxsw_sp_port, + mall_entry, extack); + if (err) + goto err_port_sample_set; + return 0; + +err_port_sample_set: + mlxsw_sp_sample_trigger_params_unset(mlxsw_sp, &trigger); + return err; +} + +static void +mlxsw_sp_mall_port_sample_del(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_mall_entry *mall_entry) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_sample_trigger trigger; + + if (mall_entry->ingress) + trigger.type = MLXSW_SP_SAMPLE_TRIGGER_TYPE_INGRESS; + else + trigger.type = MLXSW_SP_SAMPLE_TRIGGER_TYPE_EGRESS; + trigger.local_port = mlxsw_sp_port->local_port; + + mlxsw_sp->mall_ops->sample_del(mlxsw_sp, mlxsw_sp_port, mall_entry); + mlxsw_sp_sample_trigger_params_unset(mlxsw_sp, &trigger); +} + +static int +mlxsw_sp_mall_port_rule_add(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_mall_entry *mall_entry, + struct netlink_ext_ack *extack) +{ + switch (mall_entry->type) { + case MLXSW_SP_MALL_ACTION_TYPE_MIRROR: + return mlxsw_sp_mall_port_mirror_add(mlxsw_sp_port, mall_entry, + extack); + case MLXSW_SP_MALL_ACTION_TYPE_SAMPLE: + return mlxsw_sp_mall_port_sample_add(mlxsw_sp_port, mall_entry, + extack); + default: + WARN_ON(1); + return -EINVAL; + } +} + +static void +mlxsw_sp_mall_port_rule_del(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_mall_entry *mall_entry) +{ + switch (mall_entry->type) { + case MLXSW_SP_MALL_ACTION_TYPE_MIRROR: + mlxsw_sp_mall_port_mirror_del(mlxsw_sp_port, mall_entry); + break; + case MLXSW_SP_MALL_ACTION_TYPE_SAMPLE: + mlxsw_sp_mall_port_sample_del(mlxsw_sp_port, mall_entry); + break; + default: + WARN_ON(1); + } +} + +static void mlxsw_sp_mall_prio_update(struct mlxsw_sp_flow_block *block) +{ + struct mlxsw_sp_mall_entry *mall_entry; + + if (list_empty(&block->mall.list)) + return; + block->mall.min_prio = UINT_MAX; + block->mall.max_prio = 0; + list_for_each_entry(mall_entry, &block->mall.list, list) { + if (mall_entry->priority < block->mall.min_prio) + block->mall.min_prio = mall_entry->priority; + if (mall_entry->priority > block->mall.max_prio) + block->mall.max_prio = mall_entry->priority; + } +} + +int mlxsw_sp_mall_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_flow_block *block, + struct tc_cls_matchall_offload *f) +{ + struct mlxsw_sp_flow_block_binding *binding; + struct mlxsw_sp_mall_entry *mall_entry; + __be16 protocol = f->common.protocol; + struct flow_action_entry *act; + unsigned int flower_min_prio; + unsigned int flower_max_prio; + bool flower_prio_valid; + int err; + + if (!flow_offload_has_one_action(&f->rule->action)) { + NL_SET_ERR_MSG(f->common.extack, "Only singular actions are supported"); + return -EOPNOTSUPP; + } + + if (f->common.chain_index) { + NL_SET_ERR_MSG(f->common.extack, "Only chain 0 is supported"); + return -EOPNOTSUPP; + } + + if (mlxsw_sp_flow_block_is_mixed_bound(block)) { + NL_SET_ERR_MSG(f->common.extack, "Only not mixed bound blocks are supported"); + return -EOPNOTSUPP; + } + + err = mlxsw_sp_flower_prio_get(mlxsw_sp, block, f->common.chain_index, + &flower_min_prio, &flower_max_prio); + if (err) { + if (err != -ENOENT) { + NL_SET_ERR_MSG(f->common.extack, "Failed to get flower priorities"); + return err; + } + flower_prio_valid = false; + /* No flower filters are installed in specified chain. */ + } else { + flower_prio_valid = true; + } + + if (protocol != htons(ETH_P_ALL)) { + NL_SET_ERR_MSG(f->common.extack, "matchall rules only supported with 'all' protocol"); + return -EOPNOTSUPP; + } + + mall_entry = kzalloc(sizeof(*mall_entry), GFP_KERNEL); + if (!mall_entry) + return -ENOMEM; + mall_entry->cookie = f->cookie; + mall_entry->priority = f->common.prio; + mall_entry->ingress = mlxsw_sp_flow_block_is_ingress_bound(block); + + if (flower_prio_valid && mall_entry->ingress && + mall_entry->priority >= flower_min_prio) { + NL_SET_ERR_MSG(f->common.extack, "Failed to add behind existing flower rules"); + err = -EOPNOTSUPP; + goto errout; + } + if (flower_prio_valid && !mall_entry->ingress && + mall_entry->priority <= flower_max_prio) { + NL_SET_ERR_MSG(f->common.extack, "Failed to add in front of existing flower rules"); + err = -EOPNOTSUPP; + goto errout; + } + + act = &f->rule->action.entries[0]; + + switch (act->id) { + case FLOW_ACTION_MIRRED: + mall_entry->type = MLXSW_SP_MALL_ACTION_TYPE_MIRROR; + mall_entry->mirror.to_dev = act->dev; + break; + case FLOW_ACTION_SAMPLE: + mall_entry->type = MLXSW_SP_MALL_ACTION_TYPE_SAMPLE; + mall_entry->sample.params.psample_group = act->sample.psample_group; + mall_entry->sample.params.truncate = act->sample.truncate; + mall_entry->sample.params.trunc_size = act->sample.trunc_size; + mall_entry->sample.params.rate = act->sample.rate; + break; + default: + err = -EOPNOTSUPP; + goto errout; + } + + list_for_each_entry(binding, &block->binding_list, list) { + err = mlxsw_sp_mall_port_rule_add(binding->mlxsw_sp_port, + mall_entry, f->common.extack); + if (err) + goto rollback; + } + + block->rule_count++; + if (mall_entry->ingress) + block->egress_blocker_rule_count++; + else + block->ingress_blocker_rule_count++; + list_add_tail(&mall_entry->list, &block->mall.list); + mlxsw_sp_mall_prio_update(block); + return 0; + +rollback: + list_for_each_entry_continue_reverse(binding, &block->binding_list, + list) + mlxsw_sp_mall_port_rule_del(binding->mlxsw_sp_port, mall_entry); +errout: + kfree(mall_entry); + return err; +} + +void mlxsw_sp_mall_destroy(struct mlxsw_sp_flow_block *block, + struct tc_cls_matchall_offload *f) +{ + struct mlxsw_sp_flow_block_binding *binding; + struct mlxsw_sp_mall_entry *mall_entry; + + mall_entry = mlxsw_sp_mall_entry_find(block, f->cookie); + if (!mall_entry) { + NL_SET_ERR_MSG(f->common.extack, "Entry not found"); + return; + } + + list_del(&mall_entry->list); + if (mall_entry->ingress) + block->egress_blocker_rule_count--; + else + block->ingress_blocker_rule_count--; + block->rule_count--; + list_for_each_entry(binding, &block->binding_list, list) + mlxsw_sp_mall_port_rule_del(binding->mlxsw_sp_port, mall_entry); + kfree_rcu(mall_entry, rcu); /* sample RX packets may be in-flight */ + mlxsw_sp_mall_prio_update(block); +} + +int mlxsw_sp_mall_port_bind(struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_mall_entry *mall_entry; + int err; + + list_for_each_entry(mall_entry, &block->mall.list, list) { + err = mlxsw_sp_mall_port_rule_add(mlxsw_sp_port, mall_entry, + extack); + if (err) + goto rollback; + } + return 0; + +rollback: + list_for_each_entry_continue_reverse(mall_entry, &block->mall.list, + list) + mlxsw_sp_mall_port_rule_del(mlxsw_sp_port, mall_entry); + return err; +} + +void mlxsw_sp_mall_port_unbind(struct mlxsw_sp_flow_block *block, + struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp_mall_entry *mall_entry; + + list_for_each_entry(mall_entry, &block->mall.list, list) + mlxsw_sp_mall_port_rule_del(mlxsw_sp_port, mall_entry); +} + +int mlxsw_sp_mall_prio_get(struct mlxsw_sp_flow_block *block, u32 chain_index, + unsigned int *p_min_prio, unsigned int *p_max_prio) +{ + if (chain_index || list_empty(&block->mall.list)) + /* In case there are no matchall rules, the caller + * receives -ENOENT to indicate there is no need + * to check the priorities. + */ + return -ENOENT; + *p_min_prio = block->mall.min_prio; + *p_max_prio = block->mall.max_prio; + return 0; +} + +static int mlxsw_sp1_mall_sample_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_mall_entry *mall_entry, + struct netlink_ext_ack *extack) +{ + u32 rate = mall_entry->sample.params.rate; + + if (!mall_entry->ingress) { + NL_SET_ERR_MSG(extack, "Sampling is not supported on egress"); + return -EOPNOTSUPP; + } + + if (rate > MLXSW_REG_MPSC_RATE_MAX) { + NL_SET_ERR_MSG(extack, "Unsupported sampling rate"); + return -EOPNOTSUPP; + } + + return mlxsw_sp_mall_port_sample_set(mlxsw_sp_port, true, rate); +} + +static void mlxsw_sp1_mall_sample_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_mall_entry *mall_entry) +{ + mlxsw_sp_mall_port_sample_set(mlxsw_sp_port, false, 1); +} + +const struct mlxsw_sp_mall_ops mlxsw_sp1_mall_ops = { + .sample_add = mlxsw_sp1_mall_sample_add, + .sample_del = mlxsw_sp1_mall_sample_del, +}; + +static int mlxsw_sp2_mall_sample_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_mall_entry *mall_entry, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_span_trigger_parms trigger_parms = {}; + struct mlxsw_sp_span_agent_parms agent_parms = { + .to_dev = NULL, /* Mirror to CPU. */ + .session_id = MLXSW_SP_SPAN_SESSION_ID_SAMPLING, + }; + u32 rate = mall_entry->sample.params.rate; + enum mlxsw_sp_span_trigger span_trigger; + int err; + + err = mlxsw_sp_span_agent_get(mlxsw_sp, &mall_entry->sample.span_id, + &agent_parms); + if (err) { + NL_SET_ERR_MSG(extack, "Failed to get SPAN agent"); + return err; + } + + err = mlxsw_sp_span_analyzed_port_get(mlxsw_sp_port, + mall_entry->ingress); + if (err) { + NL_SET_ERR_MSG(extack, "Failed to get analyzed port"); + goto err_analyzed_port_get; + } + + span_trigger = mall_entry->ingress ? MLXSW_SP_SPAN_TRIGGER_INGRESS : + MLXSW_SP_SPAN_TRIGGER_EGRESS; + trigger_parms.span_id = mall_entry->sample.span_id; + trigger_parms.probability_rate = rate; + err = mlxsw_sp_span_agent_bind(mlxsw_sp, span_trigger, mlxsw_sp_port, + &trigger_parms); + if (err) { + NL_SET_ERR_MSG(extack, "Failed to bind SPAN agent"); + goto err_agent_bind; + } + + return 0; + +err_agent_bind: + mlxsw_sp_span_analyzed_port_put(mlxsw_sp_port, mall_entry->ingress); +err_analyzed_port_get: + mlxsw_sp_span_agent_put(mlxsw_sp, mall_entry->sample.span_id); + return err; +} + +static void mlxsw_sp2_mall_sample_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_mall_entry *mall_entry) +{ + struct mlxsw_sp_span_trigger_parms trigger_parms = {}; + enum mlxsw_sp_span_trigger span_trigger; + + span_trigger = mall_entry->ingress ? MLXSW_SP_SPAN_TRIGGER_INGRESS : + MLXSW_SP_SPAN_TRIGGER_EGRESS; + trigger_parms.span_id = mall_entry->sample.span_id; + mlxsw_sp_span_agent_unbind(mlxsw_sp, span_trigger, mlxsw_sp_port, + &trigger_parms); + mlxsw_sp_span_analyzed_port_put(mlxsw_sp_port, mall_entry->ingress); + mlxsw_sp_span_agent_put(mlxsw_sp, mall_entry->sample.span_id); +} + +const struct mlxsw_sp_mall_ops mlxsw_sp2_mall_ops = { + .sample_add = mlxsw_sp2_mall_sample_add, + .sample_del = mlxsw_sp2_mall_sample_del, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.c new file mode 100644 index 000000000..1f6bc0c7e --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.c @@ -0,0 +1,1075 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/mutex.h> +#include <linux/rhashtable.h> +#include <net/ipv6.h> + +#include "spectrum_mr.h" +#include "spectrum_router.h" + +struct mlxsw_sp_mr { + const struct mlxsw_sp_mr_ops *mr_ops; + void *catchall_route_priv; + struct delayed_work stats_update_dw; + struct list_head table_list; + struct mutex table_list_lock; /* Protects table_list */ +#define MLXSW_SP_MR_ROUTES_COUNTER_UPDATE_INTERVAL 5000 /* ms */ + unsigned long priv[]; + /* priv has to be always the last item */ +}; + +struct mlxsw_sp_mr_vif; +struct mlxsw_sp_mr_vif_ops { + bool (*is_regular)(const struct mlxsw_sp_mr_vif *vif); +}; + +struct mlxsw_sp_mr_vif { + struct net_device *dev; + const struct mlxsw_sp_rif *rif; + unsigned long vif_flags; + + /* A list of route_vif_entry structs that point to routes that the VIF + * instance is used as one of the egress VIFs + */ + struct list_head route_evif_list; + + /* A list of route_vif_entry structs that point to routes that the VIF + * instance is used as an ingress VIF + */ + struct list_head route_ivif_list; + + /* Protocol specific operations for a VIF */ + const struct mlxsw_sp_mr_vif_ops *ops; +}; + +struct mlxsw_sp_mr_route_vif_entry { + struct list_head vif_node; + struct list_head route_node; + struct mlxsw_sp_mr_vif *mr_vif; + struct mlxsw_sp_mr_route *mr_route; +}; + +struct mlxsw_sp_mr_table; +struct mlxsw_sp_mr_table_ops { + bool (*is_route_valid)(const struct mlxsw_sp_mr_table *mr_table, + const struct mr_mfc *mfc); + void (*key_create)(struct mlxsw_sp_mr_table *mr_table, + struct mlxsw_sp_mr_route_key *key, + struct mr_mfc *mfc); + bool (*is_route_starg)(const struct mlxsw_sp_mr_table *mr_table, + const struct mlxsw_sp_mr_route *mr_route); +}; + +struct mlxsw_sp_mr_table { + struct list_head node; + enum mlxsw_sp_l3proto proto; + struct mlxsw_sp *mlxsw_sp; + u32 vr_id; + struct mlxsw_sp_mr_vif vifs[MAXVIFS]; + struct list_head route_list; + struct mutex route_list_lock; /* Protects route_list */ + struct rhashtable route_ht; + const struct mlxsw_sp_mr_table_ops *ops; + char catchall_route_priv[]; + /* catchall_route_priv has to be always the last item */ +}; + +struct mlxsw_sp_mr_route { + struct list_head node; + struct rhash_head ht_node; + struct mlxsw_sp_mr_route_key key; + enum mlxsw_sp_mr_route_action route_action; + u16 min_mtu; + struct mr_mfc *mfc; + void *route_priv; + const struct mlxsw_sp_mr_table *mr_table; + /* A list of route_vif_entry structs that point to the egress VIFs */ + struct list_head evif_list; + /* A route_vif_entry struct that point to the ingress VIF */ + struct mlxsw_sp_mr_route_vif_entry ivif; +}; + +static const struct rhashtable_params mlxsw_sp_mr_route_ht_params = { + .key_len = sizeof(struct mlxsw_sp_mr_route_key), + .key_offset = offsetof(struct mlxsw_sp_mr_route, key), + .head_offset = offsetof(struct mlxsw_sp_mr_route, ht_node), + .automatic_shrinking = true, +}; + +static bool mlxsw_sp_mr_vif_valid(const struct mlxsw_sp_mr_vif *vif) +{ + return vif->ops->is_regular(vif) && vif->dev && vif->rif; +} + +static bool mlxsw_sp_mr_vif_exists(const struct mlxsw_sp_mr_vif *vif) +{ + return vif->dev; +} + +static bool +mlxsw_sp_mr_route_ivif_in_evifs(const struct mlxsw_sp_mr_route *mr_route) +{ + vifi_t ivif = mr_route->mfc->mfc_parent; + + return mr_route->mfc->mfc_un.res.ttls[ivif] != 255; +} + +static int +mlxsw_sp_mr_route_valid_evifs_num(const struct mlxsw_sp_mr_route *mr_route) +{ + struct mlxsw_sp_mr_route_vif_entry *rve; + int valid_evifs; + + valid_evifs = 0; + list_for_each_entry(rve, &mr_route->evif_list, route_node) + if (mlxsw_sp_mr_vif_valid(rve->mr_vif)) + valid_evifs++; + return valid_evifs; +} + +static enum mlxsw_sp_mr_route_action +mlxsw_sp_mr_route_action(const struct mlxsw_sp_mr_route *mr_route) +{ + struct mlxsw_sp_mr_route_vif_entry *rve; + + /* If the ingress port is not regular and resolved, trap the route */ + if (!mlxsw_sp_mr_vif_valid(mr_route->ivif.mr_vif)) + return MLXSW_SP_MR_ROUTE_ACTION_TRAP; + + /* The kernel does not match a (*,G) route that the ingress interface is + * not one of the egress interfaces, so trap these kind of routes. + */ + if (mr_route->mr_table->ops->is_route_starg(mr_route->mr_table, + mr_route) && + !mlxsw_sp_mr_route_ivif_in_evifs(mr_route)) + return MLXSW_SP_MR_ROUTE_ACTION_TRAP; + + /* If the route has no valid eVIFs, trap it. */ + if (!mlxsw_sp_mr_route_valid_evifs_num(mr_route)) + return MLXSW_SP_MR_ROUTE_ACTION_TRAP; + + /* If one of the eVIFs has no RIF, trap-and-forward the route as there + * is some more routing to do in software too. + */ + list_for_each_entry(rve, &mr_route->evif_list, route_node) + if (mlxsw_sp_mr_vif_exists(rve->mr_vif) && !rve->mr_vif->rif) + return MLXSW_SP_MR_ROUTE_ACTION_TRAP_AND_FORWARD; + + return MLXSW_SP_MR_ROUTE_ACTION_FORWARD; +} + +static enum mlxsw_sp_mr_route_prio +mlxsw_sp_mr_route_prio(const struct mlxsw_sp_mr_route *mr_route) +{ + return mr_route->mr_table->ops->is_route_starg(mr_route->mr_table, + mr_route) ? + MLXSW_SP_MR_ROUTE_PRIO_STARG : MLXSW_SP_MR_ROUTE_PRIO_SG; +} + +static int mlxsw_sp_mr_route_evif_link(struct mlxsw_sp_mr_route *mr_route, + struct mlxsw_sp_mr_vif *mr_vif) +{ + struct mlxsw_sp_mr_route_vif_entry *rve; + + rve = kzalloc(sizeof(*rve), GFP_KERNEL); + if (!rve) + return -ENOMEM; + rve->mr_route = mr_route; + rve->mr_vif = mr_vif; + list_add_tail(&rve->route_node, &mr_route->evif_list); + list_add_tail(&rve->vif_node, &mr_vif->route_evif_list); + return 0; +} + +static void +mlxsw_sp_mr_route_evif_unlink(struct mlxsw_sp_mr_route_vif_entry *rve) +{ + list_del(&rve->route_node); + list_del(&rve->vif_node); + kfree(rve); +} + +static void mlxsw_sp_mr_route_ivif_link(struct mlxsw_sp_mr_route *mr_route, + struct mlxsw_sp_mr_vif *mr_vif) +{ + mr_route->ivif.mr_route = mr_route; + mr_route->ivif.mr_vif = mr_vif; + list_add_tail(&mr_route->ivif.vif_node, &mr_vif->route_ivif_list); +} + +static void mlxsw_sp_mr_route_ivif_unlink(struct mlxsw_sp_mr_route *mr_route) +{ + list_del(&mr_route->ivif.vif_node); +} + +static int +mlxsw_sp_mr_route_info_create(struct mlxsw_sp_mr_table *mr_table, + struct mlxsw_sp_mr_route *mr_route, + struct mlxsw_sp_mr_route_info *route_info) +{ + struct mlxsw_sp_mr_route_vif_entry *rve; + u16 *erif_indices; + u16 irif_index; + u16 erif = 0; + + erif_indices = kmalloc_array(MAXVIFS, sizeof(*erif_indices), + GFP_KERNEL); + if (!erif_indices) + return -ENOMEM; + + list_for_each_entry(rve, &mr_route->evif_list, route_node) { + if (mlxsw_sp_mr_vif_valid(rve->mr_vif)) { + u16 rifi = mlxsw_sp_rif_index(rve->mr_vif->rif); + + erif_indices[erif++] = rifi; + } + } + + if (mlxsw_sp_mr_vif_valid(mr_route->ivif.mr_vif)) + irif_index = mlxsw_sp_rif_index(mr_route->ivif.mr_vif->rif); + else + irif_index = 0; + + route_info->irif_index = irif_index; + route_info->erif_indices = erif_indices; + route_info->min_mtu = mr_route->min_mtu; + route_info->route_action = mr_route->route_action; + route_info->erif_num = erif; + return 0; +} + +static void +mlxsw_sp_mr_route_info_destroy(struct mlxsw_sp_mr_route_info *route_info) +{ + kfree(route_info->erif_indices); +} + +static int mlxsw_sp_mr_route_write(struct mlxsw_sp_mr_table *mr_table, + struct mlxsw_sp_mr_route *mr_route, + bool replace) +{ + struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp; + struct mlxsw_sp_mr_route_info route_info; + struct mlxsw_sp_mr *mr = mlxsw_sp->mr; + int err; + + err = mlxsw_sp_mr_route_info_create(mr_table, mr_route, &route_info); + if (err) + return err; + + if (!replace) { + struct mlxsw_sp_mr_route_params route_params; + + mr_route->route_priv = kzalloc(mr->mr_ops->route_priv_size, + GFP_KERNEL); + if (!mr_route->route_priv) { + err = -ENOMEM; + goto out; + } + + route_params.key = mr_route->key; + route_params.value = route_info; + route_params.prio = mlxsw_sp_mr_route_prio(mr_route); + err = mr->mr_ops->route_create(mlxsw_sp, mr->priv, + mr_route->route_priv, + &route_params); + if (err) + kfree(mr_route->route_priv); + } else { + err = mr->mr_ops->route_update(mlxsw_sp, mr_route->route_priv, + &route_info); + } +out: + mlxsw_sp_mr_route_info_destroy(&route_info); + return err; +} + +static void mlxsw_sp_mr_route_erase(struct mlxsw_sp_mr_table *mr_table, + struct mlxsw_sp_mr_route *mr_route) +{ + struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp; + struct mlxsw_sp_mr *mr = mlxsw_sp->mr; + + mr->mr_ops->route_destroy(mlxsw_sp, mr->priv, mr_route->route_priv); + kfree(mr_route->route_priv); +} + +static struct mlxsw_sp_mr_route * +mlxsw_sp_mr_route_create(struct mlxsw_sp_mr_table *mr_table, + struct mr_mfc *mfc) +{ + struct mlxsw_sp_mr_route_vif_entry *rve, *tmp; + struct mlxsw_sp_mr_route *mr_route; + int err = 0; + int i; + + /* Allocate and init a new route and fill it with parameters */ + mr_route = kzalloc(sizeof(*mr_route), GFP_KERNEL); + if (!mr_route) + return ERR_PTR(-ENOMEM); + INIT_LIST_HEAD(&mr_route->evif_list); + + /* Find min_mtu and link iVIF and eVIFs */ + mr_route->min_mtu = ETH_MAX_MTU; + mr_cache_hold(mfc); + mr_route->mfc = mfc; + mr_table->ops->key_create(mr_table, &mr_route->key, mr_route->mfc); + + mr_route->mr_table = mr_table; + for (i = 0; i < MAXVIFS; i++) { + if (mfc->mfc_un.res.ttls[i] != 255) { + err = mlxsw_sp_mr_route_evif_link(mr_route, + &mr_table->vifs[i]); + if (err) + goto err; + if (mr_table->vifs[i].dev && + mr_table->vifs[i].dev->mtu < mr_route->min_mtu) + mr_route->min_mtu = mr_table->vifs[i].dev->mtu; + } + } + mlxsw_sp_mr_route_ivif_link(mr_route, + &mr_table->vifs[mfc->mfc_parent]); + + mr_route->route_action = mlxsw_sp_mr_route_action(mr_route); + return mr_route; +err: + mr_cache_put(mfc); + list_for_each_entry_safe(rve, tmp, &mr_route->evif_list, route_node) + mlxsw_sp_mr_route_evif_unlink(rve); + kfree(mr_route); + return ERR_PTR(err); +} + +static void mlxsw_sp_mr_route_destroy(struct mlxsw_sp_mr_table *mr_table, + struct mlxsw_sp_mr_route *mr_route) +{ + struct mlxsw_sp_mr_route_vif_entry *rve, *tmp; + + mlxsw_sp_mr_route_ivif_unlink(mr_route); + mr_cache_put(mr_route->mfc); + list_for_each_entry_safe(rve, tmp, &mr_route->evif_list, route_node) + mlxsw_sp_mr_route_evif_unlink(rve); + kfree(mr_route); +} + +static void mlxsw_sp_mr_mfc_offload_set(struct mlxsw_sp_mr_route *mr_route, + bool offload) +{ + if (offload) + mr_route->mfc->mfc_flags |= MFC_OFFLOAD; + else + mr_route->mfc->mfc_flags &= ~MFC_OFFLOAD; +} + +static void mlxsw_sp_mr_mfc_offload_update(struct mlxsw_sp_mr_route *mr_route) +{ + bool offload; + + offload = mr_route->route_action != MLXSW_SP_MR_ROUTE_ACTION_TRAP; + mlxsw_sp_mr_mfc_offload_set(mr_route, offload); +} + +static void __mlxsw_sp_mr_route_del(struct mlxsw_sp_mr_table *mr_table, + struct mlxsw_sp_mr_route *mr_route) +{ + WARN_ON_ONCE(!mutex_is_locked(&mr_table->route_list_lock)); + + mlxsw_sp_mr_mfc_offload_set(mr_route, false); + rhashtable_remove_fast(&mr_table->route_ht, &mr_route->ht_node, + mlxsw_sp_mr_route_ht_params); + list_del(&mr_route->node); + mlxsw_sp_mr_route_erase(mr_table, mr_route); + mlxsw_sp_mr_route_destroy(mr_table, mr_route); +} + +int mlxsw_sp_mr_route_add(struct mlxsw_sp_mr_table *mr_table, + struct mr_mfc *mfc, bool replace) +{ + struct mlxsw_sp_mr_route *mr_orig_route = NULL; + struct mlxsw_sp_mr_route *mr_route; + int err; + + if (!mr_table->ops->is_route_valid(mr_table, mfc)) + return -EINVAL; + + /* Create a new route */ + mr_route = mlxsw_sp_mr_route_create(mr_table, mfc); + if (IS_ERR(mr_route)) + return PTR_ERR(mr_route); + + /* Find any route with a matching key */ + mr_orig_route = rhashtable_lookup_fast(&mr_table->route_ht, + &mr_route->key, + mlxsw_sp_mr_route_ht_params); + if (replace) { + /* On replace case, make the route point to the new route_priv. + */ + if (WARN_ON(!mr_orig_route)) { + err = -ENOENT; + goto err_no_orig_route; + } + mr_route->route_priv = mr_orig_route->route_priv; + } else if (mr_orig_route) { + /* On non replace case, if another route with the same key was + * found, abort, as duplicate routes are used for proxy routes. + */ + dev_warn(mr_table->mlxsw_sp->bus_info->dev, + "Offloading proxy routes is not supported.\n"); + err = -EINVAL; + goto err_duplicate_route; + } + + /* Write the route to the hardware */ + err = mlxsw_sp_mr_route_write(mr_table, mr_route, replace); + if (err) + goto err_mr_route_write; + + /* Put it in the table data-structures */ + mutex_lock(&mr_table->route_list_lock); + list_add_tail(&mr_route->node, &mr_table->route_list); + mutex_unlock(&mr_table->route_list_lock); + err = rhashtable_insert_fast(&mr_table->route_ht, + &mr_route->ht_node, + mlxsw_sp_mr_route_ht_params); + if (err) + goto err_rhashtable_insert; + + /* Destroy the original route */ + if (replace) { + rhashtable_remove_fast(&mr_table->route_ht, + &mr_orig_route->ht_node, + mlxsw_sp_mr_route_ht_params); + list_del(&mr_orig_route->node); + mlxsw_sp_mr_route_destroy(mr_table, mr_orig_route); + } + + mlxsw_sp_mr_mfc_offload_update(mr_route); + return 0; + +err_rhashtable_insert: + mutex_lock(&mr_table->route_list_lock); + list_del(&mr_route->node); + mutex_unlock(&mr_table->route_list_lock); + mlxsw_sp_mr_route_erase(mr_table, mr_route); +err_mr_route_write: +err_no_orig_route: +err_duplicate_route: + mlxsw_sp_mr_route_destroy(mr_table, mr_route); + return err; +} + +void mlxsw_sp_mr_route_del(struct mlxsw_sp_mr_table *mr_table, + struct mr_mfc *mfc) +{ + struct mlxsw_sp_mr_route *mr_route; + struct mlxsw_sp_mr_route_key key; + + mr_table->ops->key_create(mr_table, &key, mfc); + mr_route = rhashtable_lookup_fast(&mr_table->route_ht, &key, + mlxsw_sp_mr_route_ht_params); + if (mr_route) { + mutex_lock(&mr_table->route_list_lock); + __mlxsw_sp_mr_route_del(mr_table, mr_route); + mutex_unlock(&mr_table->route_list_lock); + } +} + +/* Should be called after the VIF struct is updated */ +static int +mlxsw_sp_mr_route_ivif_resolve(struct mlxsw_sp_mr_table *mr_table, + struct mlxsw_sp_mr_route_vif_entry *rve) +{ + struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp; + enum mlxsw_sp_mr_route_action route_action; + struct mlxsw_sp_mr *mr = mlxsw_sp->mr; + u16 irif_index; + int err; + + route_action = mlxsw_sp_mr_route_action(rve->mr_route); + if (route_action == MLXSW_SP_MR_ROUTE_ACTION_TRAP) + return 0; + + /* rve->mr_vif->rif is guaranteed to be valid at this stage */ + irif_index = mlxsw_sp_rif_index(rve->mr_vif->rif); + err = mr->mr_ops->route_irif_update(mlxsw_sp, rve->mr_route->route_priv, + irif_index); + if (err) + return err; + + err = mr->mr_ops->route_action_update(mlxsw_sp, + rve->mr_route->route_priv, + route_action); + if (err) + /* No need to rollback here because the iRIF change only takes + * place after the action has been updated. + */ + return err; + + rve->mr_route->route_action = route_action; + mlxsw_sp_mr_mfc_offload_update(rve->mr_route); + return 0; +} + +static void +mlxsw_sp_mr_route_ivif_unresolve(struct mlxsw_sp_mr_table *mr_table, + struct mlxsw_sp_mr_route_vif_entry *rve) +{ + struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp; + struct mlxsw_sp_mr *mr = mlxsw_sp->mr; + + mr->mr_ops->route_action_update(mlxsw_sp, rve->mr_route->route_priv, + MLXSW_SP_MR_ROUTE_ACTION_TRAP); + rve->mr_route->route_action = MLXSW_SP_MR_ROUTE_ACTION_TRAP; + mlxsw_sp_mr_mfc_offload_update(rve->mr_route); +} + +/* Should be called after the RIF struct is updated */ +static int +mlxsw_sp_mr_route_evif_resolve(struct mlxsw_sp_mr_table *mr_table, + struct mlxsw_sp_mr_route_vif_entry *rve) +{ + struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp; + enum mlxsw_sp_mr_route_action route_action; + struct mlxsw_sp_mr *mr = mlxsw_sp->mr; + u16 erif_index = 0; + int err; + + /* Add the eRIF */ + if (mlxsw_sp_mr_vif_valid(rve->mr_vif)) { + erif_index = mlxsw_sp_rif_index(rve->mr_vif->rif); + err = mr->mr_ops->route_erif_add(mlxsw_sp, + rve->mr_route->route_priv, + erif_index); + if (err) + return err; + } + + /* Update the route action, as the new eVIF can be a tunnel or a pimreg + * device which will require updating the action. + */ + route_action = mlxsw_sp_mr_route_action(rve->mr_route); + if (route_action != rve->mr_route->route_action) { + err = mr->mr_ops->route_action_update(mlxsw_sp, + rve->mr_route->route_priv, + route_action); + if (err) + goto err_route_action_update; + } + + /* Update the minimum MTU */ + if (rve->mr_vif->dev->mtu < rve->mr_route->min_mtu) { + rve->mr_route->min_mtu = rve->mr_vif->dev->mtu; + err = mr->mr_ops->route_min_mtu_update(mlxsw_sp, + rve->mr_route->route_priv, + rve->mr_route->min_mtu); + if (err) + goto err_route_min_mtu_update; + } + + rve->mr_route->route_action = route_action; + mlxsw_sp_mr_mfc_offload_update(rve->mr_route); + return 0; + +err_route_min_mtu_update: + if (route_action != rve->mr_route->route_action) + mr->mr_ops->route_action_update(mlxsw_sp, + rve->mr_route->route_priv, + rve->mr_route->route_action); +err_route_action_update: + if (mlxsw_sp_mr_vif_valid(rve->mr_vif)) + mr->mr_ops->route_erif_del(mlxsw_sp, rve->mr_route->route_priv, + erif_index); + return err; +} + +/* Should be called before the RIF struct is updated */ +static void +mlxsw_sp_mr_route_evif_unresolve(struct mlxsw_sp_mr_table *mr_table, + struct mlxsw_sp_mr_route_vif_entry *rve) +{ + struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp; + enum mlxsw_sp_mr_route_action route_action; + struct mlxsw_sp_mr *mr = mlxsw_sp->mr; + u16 rifi; + + /* If the unresolved RIF was not valid, no need to delete it */ + if (!mlxsw_sp_mr_vif_valid(rve->mr_vif)) + return; + + /* Update the route action: if there is only one valid eVIF in the + * route, set the action to trap as the VIF deletion will lead to zero + * valid eVIFs. On any other case, use the mlxsw_sp_mr_route_action to + * determine the route action. + */ + if (mlxsw_sp_mr_route_valid_evifs_num(rve->mr_route) == 1) + route_action = MLXSW_SP_MR_ROUTE_ACTION_TRAP; + else + route_action = mlxsw_sp_mr_route_action(rve->mr_route); + if (route_action != rve->mr_route->route_action) + mr->mr_ops->route_action_update(mlxsw_sp, + rve->mr_route->route_priv, + route_action); + + /* Delete the erif from the route */ + rifi = mlxsw_sp_rif_index(rve->mr_vif->rif); + mr->mr_ops->route_erif_del(mlxsw_sp, rve->mr_route->route_priv, rifi); + rve->mr_route->route_action = route_action; + mlxsw_sp_mr_mfc_offload_update(rve->mr_route); +} + +static int mlxsw_sp_mr_vif_resolve(struct mlxsw_sp_mr_table *mr_table, + struct net_device *dev, + struct mlxsw_sp_mr_vif *mr_vif, + unsigned long vif_flags, + const struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp_mr_route_vif_entry *irve, *erve; + int err; + + /* Update the VIF */ + mr_vif->dev = dev; + mr_vif->rif = rif; + mr_vif->vif_flags = vif_flags; + + /* Update all routes where this VIF is used as an unresolved iRIF */ + list_for_each_entry(irve, &mr_vif->route_ivif_list, vif_node) { + err = mlxsw_sp_mr_route_ivif_resolve(mr_table, irve); + if (err) + goto err_irif_unresolve; + } + + /* Update all routes where this VIF is used as an unresolved eRIF */ + list_for_each_entry(erve, &mr_vif->route_evif_list, vif_node) { + err = mlxsw_sp_mr_route_evif_resolve(mr_table, erve); + if (err) + goto err_erif_unresolve; + } + return 0; + +err_erif_unresolve: + list_for_each_entry_continue_reverse(erve, &mr_vif->route_evif_list, + vif_node) + mlxsw_sp_mr_route_evif_unresolve(mr_table, erve); +err_irif_unresolve: + list_for_each_entry_continue_reverse(irve, &mr_vif->route_ivif_list, + vif_node) + mlxsw_sp_mr_route_ivif_unresolve(mr_table, irve); + mr_vif->rif = NULL; + return err; +} + +static void mlxsw_sp_mr_vif_unresolve(struct mlxsw_sp_mr_table *mr_table, + struct net_device *dev, + struct mlxsw_sp_mr_vif *mr_vif) +{ + struct mlxsw_sp_mr_route_vif_entry *rve; + + /* Update all routes where this VIF is used as an unresolved eRIF */ + list_for_each_entry(rve, &mr_vif->route_evif_list, vif_node) + mlxsw_sp_mr_route_evif_unresolve(mr_table, rve); + + /* Update all routes where this VIF is used as an unresolved iRIF */ + list_for_each_entry(rve, &mr_vif->route_ivif_list, vif_node) + mlxsw_sp_mr_route_ivif_unresolve(mr_table, rve); + + /* Update the VIF */ + mr_vif->dev = dev; + mr_vif->rif = NULL; +} + +int mlxsw_sp_mr_vif_add(struct mlxsw_sp_mr_table *mr_table, + struct net_device *dev, vifi_t vif_index, + unsigned long vif_flags, const struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp_mr_vif *mr_vif = &mr_table->vifs[vif_index]; + + if (WARN_ON(vif_index >= MAXVIFS)) + return -EINVAL; + if (mr_vif->dev) + return -EEXIST; + return mlxsw_sp_mr_vif_resolve(mr_table, dev, mr_vif, vif_flags, rif); +} + +void mlxsw_sp_mr_vif_del(struct mlxsw_sp_mr_table *mr_table, vifi_t vif_index) +{ + struct mlxsw_sp_mr_vif *mr_vif = &mr_table->vifs[vif_index]; + + if (WARN_ON(vif_index >= MAXVIFS)) + return; + if (WARN_ON(!mr_vif->dev)) + return; + mlxsw_sp_mr_vif_unresolve(mr_table, NULL, mr_vif); +} + +static struct mlxsw_sp_mr_vif * +mlxsw_sp_mr_dev_vif_lookup(struct mlxsw_sp_mr_table *mr_table, + const struct net_device *dev) +{ + vifi_t vif_index; + + for (vif_index = 0; vif_index < MAXVIFS; vif_index++) + if (mr_table->vifs[vif_index].dev == dev) + return &mr_table->vifs[vif_index]; + return NULL; +} + +int mlxsw_sp_mr_rif_add(struct mlxsw_sp_mr_table *mr_table, + const struct mlxsw_sp_rif *rif) +{ + const struct net_device *rif_dev = mlxsw_sp_rif_dev(rif); + struct mlxsw_sp_mr_vif *mr_vif; + + if (!rif_dev) + return 0; + + mr_vif = mlxsw_sp_mr_dev_vif_lookup(mr_table, rif_dev); + if (!mr_vif) + return 0; + return mlxsw_sp_mr_vif_resolve(mr_table, mr_vif->dev, mr_vif, + mr_vif->vif_flags, rif); +} + +void mlxsw_sp_mr_rif_del(struct mlxsw_sp_mr_table *mr_table, + const struct mlxsw_sp_rif *rif) +{ + const struct net_device *rif_dev = mlxsw_sp_rif_dev(rif); + struct mlxsw_sp_mr_vif *mr_vif; + + if (!rif_dev) + return; + + mr_vif = mlxsw_sp_mr_dev_vif_lookup(mr_table, rif_dev); + if (!mr_vif) + return; + mlxsw_sp_mr_vif_unresolve(mr_table, mr_vif->dev, mr_vif); +} + +void mlxsw_sp_mr_rif_mtu_update(struct mlxsw_sp_mr_table *mr_table, + const struct mlxsw_sp_rif *rif, int mtu) +{ + const struct net_device *rif_dev = mlxsw_sp_rif_dev(rif); + struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp; + struct mlxsw_sp_mr_route_vif_entry *rve; + struct mlxsw_sp_mr *mr = mlxsw_sp->mr; + struct mlxsw_sp_mr_vif *mr_vif; + + if (!rif_dev) + return; + + /* Search for a VIF that use that RIF */ + mr_vif = mlxsw_sp_mr_dev_vif_lookup(mr_table, rif_dev); + if (!mr_vif) + return; + + /* Update all the routes that uses that VIF as eVIF */ + list_for_each_entry(rve, &mr_vif->route_evif_list, vif_node) { + if (mtu < rve->mr_route->min_mtu) { + rve->mr_route->min_mtu = mtu; + mr->mr_ops->route_min_mtu_update(mlxsw_sp, + rve->mr_route->route_priv, + mtu); + } + } +} + +/* Protocol specific functions */ +static bool +mlxsw_sp_mr_route4_validate(const struct mlxsw_sp_mr_table *mr_table, + const struct mr_mfc *c) +{ + struct mfc_cache *mfc = (struct mfc_cache *) c; + + /* If the route is a (*,*) route, abort, as these kind of routes are + * used for proxy routes. + */ + if (mfc->mfc_origin == htonl(INADDR_ANY) && + mfc->mfc_mcastgrp == htonl(INADDR_ANY)) { + dev_warn(mr_table->mlxsw_sp->bus_info->dev, + "Offloading proxy routes is not supported.\n"); + return false; + } + return true; +} + +static void mlxsw_sp_mr_route4_key(struct mlxsw_sp_mr_table *mr_table, + struct mlxsw_sp_mr_route_key *key, + struct mr_mfc *c) +{ + const struct mfc_cache *mfc = (struct mfc_cache *) c; + bool starg; + + starg = (mfc->mfc_origin == htonl(INADDR_ANY)); + + memset(key, 0, sizeof(*key)); + key->vrid = mr_table->vr_id; + key->proto = MLXSW_SP_L3_PROTO_IPV4; + key->group.addr4 = mfc->mfc_mcastgrp; + key->group_mask.addr4 = htonl(0xffffffff); + key->source.addr4 = mfc->mfc_origin; + key->source_mask.addr4 = htonl(starg ? 0 : 0xffffffff); +} + +static bool mlxsw_sp_mr_route4_starg(const struct mlxsw_sp_mr_table *mr_table, + const struct mlxsw_sp_mr_route *mr_route) +{ + return mr_route->key.source_mask.addr4 == htonl(INADDR_ANY); +} + +static bool mlxsw_sp_mr_vif4_is_regular(const struct mlxsw_sp_mr_vif *vif) +{ + return !(vif->vif_flags & (VIFF_TUNNEL | VIFF_REGISTER)); +} + +static bool +mlxsw_sp_mr_route6_validate(const struct mlxsw_sp_mr_table *mr_table, + const struct mr_mfc *c) +{ + struct mfc6_cache *mfc = (struct mfc6_cache *) c; + + /* If the route is a (*,*) route, abort, as these kind of routes are + * used for proxy routes. + */ + if (ipv6_addr_any(&mfc->mf6c_origin) && + ipv6_addr_any(&mfc->mf6c_mcastgrp)) { + dev_warn(mr_table->mlxsw_sp->bus_info->dev, + "Offloading proxy routes is not supported.\n"); + return false; + } + return true; +} + +static void mlxsw_sp_mr_route6_key(struct mlxsw_sp_mr_table *mr_table, + struct mlxsw_sp_mr_route_key *key, + struct mr_mfc *c) +{ + const struct mfc6_cache *mfc = (struct mfc6_cache *) c; + + memset(key, 0, sizeof(*key)); + key->vrid = mr_table->vr_id; + key->proto = MLXSW_SP_L3_PROTO_IPV6; + key->group.addr6 = mfc->mf6c_mcastgrp; + memset(&key->group_mask.addr6, 0xff, sizeof(key->group_mask.addr6)); + key->source.addr6 = mfc->mf6c_origin; + if (!ipv6_addr_any(&mfc->mf6c_origin)) + memset(&key->source_mask.addr6, 0xff, + sizeof(key->source_mask.addr6)); +} + +static bool mlxsw_sp_mr_route6_starg(const struct mlxsw_sp_mr_table *mr_table, + const struct mlxsw_sp_mr_route *mr_route) +{ + return ipv6_addr_any(&mr_route->key.source_mask.addr6); +} + +static bool mlxsw_sp_mr_vif6_is_regular(const struct mlxsw_sp_mr_vif *vif) +{ + return !(vif->vif_flags & MIFF_REGISTER); +} + +static struct +mlxsw_sp_mr_vif_ops mlxsw_sp_mr_vif_ops_arr[] = { + { + .is_regular = mlxsw_sp_mr_vif4_is_regular, + }, + { + .is_regular = mlxsw_sp_mr_vif6_is_regular, + }, +}; + +static struct +mlxsw_sp_mr_table_ops mlxsw_sp_mr_table_ops_arr[] = { + { + .is_route_valid = mlxsw_sp_mr_route4_validate, + .key_create = mlxsw_sp_mr_route4_key, + .is_route_starg = mlxsw_sp_mr_route4_starg, + }, + { + .is_route_valid = mlxsw_sp_mr_route6_validate, + .key_create = mlxsw_sp_mr_route6_key, + .is_route_starg = mlxsw_sp_mr_route6_starg, + }, + +}; + +struct mlxsw_sp_mr_table *mlxsw_sp_mr_table_create(struct mlxsw_sp *mlxsw_sp, + u32 vr_id, + enum mlxsw_sp_l3proto proto) +{ + struct mlxsw_sp_mr_route_params catchall_route_params = { + .prio = MLXSW_SP_MR_ROUTE_PRIO_CATCHALL, + .key = { + .vrid = vr_id, + .proto = proto, + }, + .value = { + .route_action = MLXSW_SP_MR_ROUTE_ACTION_TRAP, + } + }; + struct mlxsw_sp_mr *mr = mlxsw_sp->mr; + struct mlxsw_sp_mr_table *mr_table; + int err; + int i; + + mr_table = kzalloc(sizeof(*mr_table) + mr->mr_ops->route_priv_size, + GFP_KERNEL); + if (!mr_table) + return ERR_PTR(-ENOMEM); + + mr_table->vr_id = vr_id; + mr_table->mlxsw_sp = mlxsw_sp; + mr_table->proto = proto; + mr_table->ops = &mlxsw_sp_mr_table_ops_arr[proto]; + INIT_LIST_HEAD(&mr_table->route_list); + mutex_init(&mr_table->route_list_lock); + + err = rhashtable_init(&mr_table->route_ht, + &mlxsw_sp_mr_route_ht_params); + if (err) + goto err_route_rhashtable_init; + + for (i = 0; i < MAXVIFS; i++) { + INIT_LIST_HEAD(&mr_table->vifs[i].route_evif_list); + INIT_LIST_HEAD(&mr_table->vifs[i].route_ivif_list); + mr_table->vifs[i].ops = &mlxsw_sp_mr_vif_ops_arr[proto]; + } + + err = mr->mr_ops->route_create(mlxsw_sp, mr->priv, + mr_table->catchall_route_priv, + &catchall_route_params); + if (err) + goto err_ops_route_create; + mutex_lock(&mr->table_list_lock); + list_add_tail(&mr_table->node, &mr->table_list); + mutex_unlock(&mr->table_list_lock); + return mr_table; + +err_ops_route_create: + rhashtable_destroy(&mr_table->route_ht); +err_route_rhashtable_init: + mutex_destroy(&mr_table->route_list_lock); + kfree(mr_table); + return ERR_PTR(err); +} + +void mlxsw_sp_mr_table_destroy(struct mlxsw_sp_mr_table *mr_table) +{ + struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp; + struct mlxsw_sp_mr *mr = mlxsw_sp->mr; + + WARN_ON(!mlxsw_sp_mr_table_empty(mr_table)); + mutex_lock(&mr->table_list_lock); + list_del(&mr_table->node); + mutex_unlock(&mr->table_list_lock); + mr->mr_ops->route_destroy(mlxsw_sp, mr->priv, + &mr_table->catchall_route_priv); + rhashtable_destroy(&mr_table->route_ht); + mutex_destroy(&mr_table->route_list_lock); + kfree(mr_table); +} + +void mlxsw_sp_mr_table_flush(struct mlxsw_sp_mr_table *mr_table) +{ + struct mlxsw_sp_mr_route *mr_route, *tmp; + int i; + + mutex_lock(&mr_table->route_list_lock); + list_for_each_entry_safe(mr_route, tmp, &mr_table->route_list, node) + __mlxsw_sp_mr_route_del(mr_table, mr_route); + mutex_unlock(&mr_table->route_list_lock); + + for (i = 0; i < MAXVIFS; i++) { + mr_table->vifs[i].dev = NULL; + mr_table->vifs[i].rif = NULL; + } +} + +bool mlxsw_sp_mr_table_empty(const struct mlxsw_sp_mr_table *mr_table) +{ + int i; + + for (i = 0; i < MAXVIFS; i++) + if (mr_table->vifs[i].dev) + return false; + return list_empty(&mr_table->route_list); +} + +static void mlxsw_sp_mr_route_stats_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mr_route *mr_route) +{ + struct mlxsw_sp_mr *mr = mlxsw_sp->mr; + u64 packets, bytes; + + if (mr_route->route_action == MLXSW_SP_MR_ROUTE_ACTION_TRAP) + return; + + mr->mr_ops->route_stats(mlxsw_sp, mr_route->route_priv, &packets, + &bytes); + + if (mr_route->mfc->mfc_un.res.pkt != packets) + mr_route->mfc->mfc_un.res.lastuse = jiffies; + mr_route->mfc->mfc_un.res.pkt = packets; + mr_route->mfc->mfc_un.res.bytes = bytes; +} + +static void mlxsw_sp_mr_stats_update(struct work_struct *work) +{ + struct mlxsw_sp_mr *mr = container_of(work, struct mlxsw_sp_mr, + stats_update_dw.work); + struct mlxsw_sp_mr_table *mr_table; + struct mlxsw_sp_mr_route *mr_route; + unsigned long interval; + + mutex_lock(&mr->table_list_lock); + list_for_each_entry(mr_table, &mr->table_list, node) { + mutex_lock(&mr_table->route_list_lock); + list_for_each_entry(mr_route, &mr_table->route_list, node) + mlxsw_sp_mr_route_stats_update(mr_table->mlxsw_sp, + mr_route); + mutex_unlock(&mr_table->route_list_lock); + } + mutex_unlock(&mr->table_list_lock); + + interval = msecs_to_jiffies(MLXSW_SP_MR_ROUTES_COUNTER_UPDATE_INTERVAL); + mlxsw_core_schedule_dw(&mr->stats_update_dw, interval); +} + +int mlxsw_sp_mr_init(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_mr_ops *mr_ops) +{ + struct mlxsw_sp_mr *mr; + unsigned long interval; + int err; + + mr = kzalloc(sizeof(*mr) + mr_ops->priv_size, GFP_KERNEL); + if (!mr) + return -ENOMEM; + mr->mr_ops = mr_ops; + mlxsw_sp->mr = mr; + INIT_LIST_HEAD(&mr->table_list); + mutex_init(&mr->table_list_lock); + + err = mr_ops->init(mlxsw_sp, mr->priv); + if (err) + goto err; + + /* Create the delayed work for counter updates */ + INIT_DELAYED_WORK(&mr->stats_update_dw, mlxsw_sp_mr_stats_update); + interval = msecs_to_jiffies(MLXSW_SP_MR_ROUTES_COUNTER_UPDATE_INTERVAL); + mlxsw_core_schedule_dw(&mr->stats_update_dw, interval); + return 0; +err: + mutex_destroy(&mr->table_list_lock); + kfree(mr); + return err; +} + +void mlxsw_sp_mr_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_mr *mr = mlxsw_sp->mr; + + cancel_delayed_work_sync(&mr->stats_update_dw); + mr->mr_ops->fini(mlxsw_sp, mr->priv); + mutex_destroy(&mr->table_list_lock); + kfree(mr); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.h new file mode 100644 index 000000000..3cde3671f --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.h @@ -0,0 +1,95 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_SPECTRUM_MCROUTER_H +#define _MLXSW_SPECTRUM_MCROUTER_H + +#include <linux/mroute.h> +#include <linux/mroute6.h> +#include "spectrum_router.h" +#include "spectrum.h" + +enum mlxsw_sp_mr_route_action { + MLXSW_SP_MR_ROUTE_ACTION_FORWARD, + MLXSW_SP_MR_ROUTE_ACTION_TRAP, + MLXSW_SP_MR_ROUTE_ACTION_TRAP_AND_FORWARD, +}; + +struct mlxsw_sp_mr_route_key { + int vrid; + enum mlxsw_sp_l3proto proto; + union mlxsw_sp_l3addr group; + union mlxsw_sp_l3addr group_mask; + union mlxsw_sp_l3addr source; + union mlxsw_sp_l3addr source_mask; +}; + +struct mlxsw_sp_mr_route_info { + enum mlxsw_sp_mr_route_action route_action; + u16 irif_index; + u16 *erif_indices; + size_t erif_num; + u16 min_mtu; +}; + +struct mlxsw_sp_mr_route_params { + struct mlxsw_sp_mr_route_key key; + struct mlxsw_sp_mr_route_info value; + enum mlxsw_sp_mr_route_prio prio; +}; + +struct mlxsw_sp_mr_ops { + int priv_size; + int route_priv_size; + int (*init)(struct mlxsw_sp *mlxsw_sp, void *priv); + int (*route_create)(struct mlxsw_sp *mlxsw_sp, void *priv, + void *route_priv, + struct mlxsw_sp_mr_route_params *route_params); + int (*route_update)(struct mlxsw_sp *mlxsw_sp, void *route_priv, + struct mlxsw_sp_mr_route_info *route_info); + int (*route_stats)(struct mlxsw_sp *mlxsw_sp, void *route_priv, + u64 *packets, u64 *bytes); + int (*route_action_update)(struct mlxsw_sp *mlxsw_sp, void *route_priv, + enum mlxsw_sp_mr_route_action route_action); + int (*route_min_mtu_update)(struct mlxsw_sp *mlxsw_sp, void *route_priv, + u16 min_mtu); + int (*route_irif_update)(struct mlxsw_sp *mlxsw_sp, void *route_priv, + u16 irif_index); + int (*route_erif_add)(struct mlxsw_sp *mlxsw_sp, void *route_priv, + u16 erif_index); + int (*route_erif_del)(struct mlxsw_sp *mlxsw_sp, void *route_priv, + u16 erif_index); + void (*route_destroy)(struct mlxsw_sp *mlxsw_sp, void *priv, + void *route_priv); + void (*fini)(struct mlxsw_sp *mlxsw_sp, void *priv); +}; + +struct mlxsw_sp_mr; +struct mlxsw_sp_mr_table; + +int mlxsw_sp_mr_init(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_mr_ops *mr_ops); +void mlxsw_sp_mr_fini(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_mr_route_add(struct mlxsw_sp_mr_table *mr_table, + struct mr_mfc *mfc, bool replace); +void mlxsw_sp_mr_route_del(struct mlxsw_sp_mr_table *mr_table, + struct mr_mfc *mfc); +int mlxsw_sp_mr_vif_add(struct mlxsw_sp_mr_table *mr_table, + struct net_device *dev, vifi_t vif_index, + unsigned long vif_flags, + const struct mlxsw_sp_rif *rif); +void mlxsw_sp_mr_vif_del(struct mlxsw_sp_mr_table *mr_table, vifi_t vif_index); +int mlxsw_sp_mr_rif_add(struct mlxsw_sp_mr_table *mr_table, + const struct mlxsw_sp_rif *rif); +void mlxsw_sp_mr_rif_del(struct mlxsw_sp_mr_table *mr_table, + const struct mlxsw_sp_rif *rif); +void mlxsw_sp_mr_rif_mtu_update(struct mlxsw_sp_mr_table *mr_table, + const struct mlxsw_sp_rif *rif, int mtu); +struct mlxsw_sp_mr_table *mlxsw_sp_mr_table_create(struct mlxsw_sp *mlxsw_sp, + u32 tb_id, + enum mlxsw_sp_l3proto proto); +void mlxsw_sp_mr_table_destroy(struct mlxsw_sp_mr_table *mr_table); +void mlxsw_sp_mr_table_flush(struct mlxsw_sp_mr_table *mr_table); +bool mlxsw_sp_mr_table_empty(const struct mlxsw_sp_mr_table *mr_table); + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c new file mode 100644 index 000000000..221aa6a47 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c @@ -0,0 +1,615 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/list.h> +#include <linux/netdevice.h> + +#include "spectrum_mr_tcam.h" +#include "reg.h" +#include "spectrum.h" +#include "core_acl_flex_actions.h" +#include "spectrum_mr.h" + +struct mlxsw_sp_mr_tcam { + void *priv; +}; + +/* This struct maps to one RIGR2 register entry */ +struct mlxsw_sp_mr_erif_sublist { + struct list_head list; + u32 rigr2_kvdl_index; + int num_erifs; + u16 erif_indices[MLXSW_REG_RIGR2_MAX_ERIFS]; + bool synced; +}; + +struct mlxsw_sp_mr_tcam_erif_list { + struct list_head erif_sublists; + u32 kvdl_index; +}; + +static bool +mlxsw_sp_mr_erif_sublist_full(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mr_erif_sublist *erif_sublist) +{ + int erif_list_entries = MLXSW_CORE_RES_GET(mlxsw_sp->core, + MC_ERIF_LIST_ENTRIES); + + return erif_sublist->num_erifs == erif_list_entries; +} + +static void +mlxsw_sp_mr_erif_list_init(struct mlxsw_sp_mr_tcam_erif_list *erif_list) +{ + INIT_LIST_HEAD(&erif_list->erif_sublists); +} + +static struct mlxsw_sp_mr_erif_sublist * +mlxsw_sp_mr_erif_sublist_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mr_tcam_erif_list *erif_list) +{ + struct mlxsw_sp_mr_erif_sublist *erif_sublist; + int err; + + erif_sublist = kzalloc(sizeof(*erif_sublist), GFP_KERNEL); + if (!erif_sublist) + return ERR_PTR(-ENOMEM); + err = mlxsw_sp_kvdl_alloc(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_MCRIGR, + 1, &erif_sublist->rigr2_kvdl_index); + if (err) { + kfree(erif_sublist); + return ERR_PTR(err); + } + + list_add_tail(&erif_sublist->list, &erif_list->erif_sublists); + return erif_sublist; +} + +static void +mlxsw_sp_mr_erif_sublist_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mr_erif_sublist *erif_sublist) +{ + list_del(&erif_sublist->list); + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_MCRIGR, + 1, erif_sublist->rigr2_kvdl_index); + kfree(erif_sublist); +} + +static int +mlxsw_sp_mr_erif_list_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mr_tcam_erif_list *erif_list, + u16 erif_index) +{ + struct mlxsw_sp_mr_erif_sublist *sublist; + + /* If either there is no erif_entry or the last one is full, allocate a + * new one. + */ + if (list_empty(&erif_list->erif_sublists)) { + sublist = mlxsw_sp_mr_erif_sublist_create(mlxsw_sp, erif_list); + if (IS_ERR(sublist)) + return PTR_ERR(sublist); + erif_list->kvdl_index = sublist->rigr2_kvdl_index; + } else { + sublist = list_last_entry(&erif_list->erif_sublists, + struct mlxsw_sp_mr_erif_sublist, + list); + sublist->synced = false; + if (mlxsw_sp_mr_erif_sublist_full(mlxsw_sp, sublist)) { + sublist = mlxsw_sp_mr_erif_sublist_create(mlxsw_sp, + erif_list); + if (IS_ERR(sublist)) + return PTR_ERR(sublist); + } + } + + /* Add the eRIF to the last entry's last index */ + sublist->erif_indices[sublist->num_erifs++] = erif_index; + return 0; +} + +static void +mlxsw_sp_mr_erif_list_flush(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mr_tcam_erif_list *erif_list) +{ + struct mlxsw_sp_mr_erif_sublist *erif_sublist, *tmp; + + list_for_each_entry_safe(erif_sublist, tmp, &erif_list->erif_sublists, + list) + mlxsw_sp_mr_erif_sublist_destroy(mlxsw_sp, erif_sublist); +} + +static int +mlxsw_sp_mr_erif_list_commit(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mr_tcam_erif_list *erif_list) +{ + struct mlxsw_sp_mr_erif_sublist *curr_sublist; + char rigr2_pl[MLXSW_REG_RIGR2_LEN]; + int err; + int i; + + list_for_each_entry(curr_sublist, &erif_list->erif_sublists, list) { + if (curr_sublist->synced) + continue; + + /* If the sublist is not the last one, pack the next index */ + if (list_is_last(&curr_sublist->list, + &erif_list->erif_sublists)) { + mlxsw_reg_rigr2_pack(rigr2_pl, + curr_sublist->rigr2_kvdl_index, + false, 0); + } else { + struct mlxsw_sp_mr_erif_sublist *next_sublist; + + next_sublist = list_next_entry(curr_sublist, list); + mlxsw_reg_rigr2_pack(rigr2_pl, + curr_sublist->rigr2_kvdl_index, + true, + next_sublist->rigr2_kvdl_index); + } + + /* Pack all the erifs */ + for (i = 0; i < curr_sublist->num_erifs; i++) { + u16 erif_index = curr_sublist->erif_indices[i]; + + mlxsw_reg_rigr2_erif_entry_pack(rigr2_pl, i, true, + erif_index); + } + + /* Write the entry */ + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rigr2), + rigr2_pl); + if (err) + /* No need of a rollback here because this + * hardware entry should not be pointed yet. + */ + return err; + curr_sublist->synced = true; + } + return 0; +} + +static void mlxsw_sp_mr_erif_list_move(struct mlxsw_sp_mr_tcam_erif_list *to, + struct mlxsw_sp_mr_tcam_erif_list *from) +{ + list_splice(&from->erif_sublists, &to->erif_sublists); + to->kvdl_index = from->kvdl_index; +} + +struct mlxsw_sp_mr_tcam_route { + struct mlxsw_sp_mr_tcam_erif_list erif_list; + struct mlxsw_afa_block *afa_block; + u32 counter_index; + enum mlxsw_sp_mr_route_action action; + struct mlxsw_sp_mr_route_key key; + u16 irif_index; + u16 min_mtu; + void *priv; +}; + +static struct mlxsw_afa_block * +mlxsw_sp_mr_tcam_afa_block_create(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_mr_route_action route_action, + u16 irif_index, u32 counter_index, + u16 min_mtu, + struct mlxsw_sp_mr_tcam_erif_list *erif_list) +{ + struct mlxsw_afa_block *afa_block; + int err; + + afa_block = mlxsw_afa_block_create(mlxsw_sp->afa); + if (IS_ERR(afa_block)) + return afa_block; + + err = mlxsw_afa_block_append_allocated_counter(afa_block, + counter_index); + if (err) + goto err; + + switch (route_action) { + case MLXSW_SP_MR_ROUTE_ACTION_TRAP: + err = mlxsw_afa_block_append_trap(afa_block, + MLXSW_TRAP_ID_ACL1); + if (err) + goto err; + break; + case MLXSW_SP_MR_ROUTE_ACTION_TRAP_AND_FORWARD: + case MLXSW_SP_MR_ROUTE_ACTION_FORWARD: + /* If we are about to append a multicast router action, commit + * the erif_list. + */ + err = mlxsw_sp_mr_erif_list_commit(mlxsw_sp, erif_list); + if (err) + goto err; + + err = mlxsw_afa_block_append_mcrouter(afa_block, irif_index, + min_mtu, false, + erif_list->kvdl_index); + if (err) + goto err; + + if (route_action == MLXSW_SP_MR_ROUTE_ACTION_TRAP_AND_FORWARD) { + err = mlxsw_afa_block_append_trap_and_forward(afa_block, + MLXSW_TRAP_ID_ACL2); + if (err) + goto err; + } + break; + default: + err = -EINVAL; + goto err; + } + + err = mlxsw_afa_block_commit(afa_block); + if (err) + goto err; + return afa_block; +err: + mlxsw_afa_block_destroy(afa_block); + return ERR_PTR(err); +} + +static void +mlxsw_sp_mr_tcam_afa_block_destroy(struct mlxsw_afa_block *afa_block) +{ + mlxsw_afa_block_destroy(afa_block); +} + +static int +mlxsw_sp_mr_tcam_erif_populate(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mr_tcam_erif_list *erif_list, + struct mlxsw_sp_mr_route_info *route_info) +{ + int err; + int i; + + for (i = 0; i < route_info->erif_num; i++) { + u16 erif_index = route_info->erif_indices[i]; + + err = mlxsw_sp_mr_erif_list_add(mlxsw_sp, erif_list, + erif_index); + if (err) + return err; + } + return 0; +} + +static int +mlxsw_sp_mr_tcam_route_create(struct mlxsw_sp *mlxsw_sp, void *priv, + void *route_priv, + struct mlxsw_sp_mr_route_params *route_params) +{ + const struct mlxsw_sp_mr_tcam_ops *ops = mlxsw_sp->mr_tcam_ops; + struct mlxsw_sp_mr_tcam_route *route = route_priv; + struct mlxsw_sp_mr_tcam *mr_tcam = priv; + int err; + + route->key = route_params->key; + route->irif_index = route_params->value.irif_index; + route->min_mtu = route_params->value.min_mtu; + route->action = route_params->value.route_action; + + /* Create the egress RIFs list */ + mlxsw_sp_mr_erif_list_init(&route->erif_list); + err = mlxsw_sp_mr_tcam_erif_populate(mlxsw_sp, &route->erif_list, + &route_params->value); + if (err) + goto err_erif_populate; + + /* Create the flow counter */ + err = mlxsw_sp_flow_counter_alloc(mlxsw_sp, &route->counter_index); + if (err) + goto err_counter_alloc; + + /* Create the flexible action block */ + route->afa_block = mlxsw_sp_mr_tcam_afa_block_create(mlxsw_sp, + route->action, + route->irif_index, + route->counter_index, + route->min_mtu, + &route->erif_list); + if (IS_ERR(route->afa_block)) { + err = PTR_ERR(route->afa_block); + goto err_afa_block_create; + } + + route->priv = kzalloc(ops->route_priv_size, GFP_KERNEL); + if (!route->priv) { + err = -ENOMEM; + goto err_route_priv_alloc; + } + + /* Write the route to the TCAM */ + err = ops->route_create(mlxsw_sp, mr_tcam->priv, route->priv, + &route->key, route->afa_block, + route_params->prio); + if (err) + goto err_route_create; + return 0; + +err_route_create: + kfree(route->priv); +err_route_priv_alloc: + mlxsw_sp_mr_tcam_afa_block_destroy(route->afa_block); +err_afa_block_create: + mlxsw_sp_flow_counter_free(mlxsw_sp, route->counter_index); +err_erif_populate: +err_counter_alloc: + mlxsw_sp_mr_erif_list_flush(mlxsw_sp, &route->erif_list); + return err; +} + +static void mlxsw_sp_mr_tcam_route_destroy(struct mlxsw_sp *mlxsw_sp, + void *priv, void *route_priv) +{ + const struct mlxsw_sp_mr_tcam_ops *ops = mlxsw_sp->mr_tcam_ops; + struct mlxsw_sp_mr_tcam_route *route = route_priv; + struct mlxsw_sp_mr_tcam *mr_tcam = priv; + + ops->route_destroy(mlxsw_sp, mr_tcam->priv, route->priv, &route->key); + kfree(route->priv); + mlxsw_sp_mr_tcam_afa_block_destroy(route->afa_block); + mlxsw_sp_flow_counter_free(mlxsw_sp, route->counter_index); + mlxsw_sp_mr_erif_list_flush(mlxsw_sp, &route->erif_list); +} + +static int mlxsw_sp_mr_tcam_route_stats(struct mlxsw_sp *mlxsw_sp, + void *route_priv, u64 *packets, + u64 *bytes) +{ + struct mlxsw_sp_mr_tcam_route *route = route_priv; + + return mlxsw_sp_flow_counter_get(mlxsw_sp, route->counter_index, + packets, bytes); +} + +static int +mlxsw_sp_mr_tcam_route_action_update(struct mlxsw_sp *mlxsw_sp, + void *route_priv, + enum mlxsw_sp_mr_route_action route_action) +{ + const struct mlxsw_sp_mr_tcam_ops *ops = mlxsw_sp->mr_tcam_ops; + struct mlxsw_sp_mr_tcam_route *route = route_priv; + struct mlxsw_afa_block *afa_block; + int err; + + /* Create a new flexible action block */ + afa_block = mlxsw_sp_mr_tcam_afa_block_create(mlxsw_sp, route_action, + route->irif_index, + route->counter_index, + route->min_mtu, + &route->erif_list); + if (IS_ERR(afa_block)) + return PTR_ERR(afa_block); + + /* Update the TCAM route entry */ + err = ops->route_update(mlxsw_sp, route->priv, &route->key, afa_block); + if (err) + goto err; + + /* Delete the old one */ + mlxsw_sp_mr_tcam_afa_block_destroy(route->afa_block); + route->afa_block = afa_block; + route->action = route_action; + return 0; +err: + mlxsw_sp_mr_tcam_afa_block_destroy(afa_block); + return err; +} + +static int mlxsw_sp_mr_tcam_route_min_mtu_update(struct mlxsw_sp *mlxsw_sp, + void *route_priv, u16 min_mtu) +{ + const struct mlxsw_sp_mr_tcam_ops *ops = mlxsw_sp->mr_tcam_ops; + struct mlxsw_sp_mr_tcam_route *route = route_priv; + struct mlxsw_afa_block *afa_block; + int err; + + /* Create a new flexible action block */ + afa_block = mlxsw_sp_mr_tcam_afa_block_create(mlxsw_sp, + route->action, + route->irif_index, + route->counter_index, + min_mtu, + &route->erif_list); + if (IS_ERR(afa_block)) + return PTR_ERR(afa_block); + + /* Update the TCAM route entry */ + err = ops->route_update(mlxsw_sp, route->priv, &route->key, afa_block); + if (err) + goto err; + + /* Delete the old one */ + mlxsw_sp_mr_tcam_afa_block_destroy(route->afa_block); + route->afa_block = afa_block; + route->min_mtu = min_mtu; + return 0; +err: + mlxsw_sp_mr_tcam_afa_block_destroy(afa_block); + return err; +} + +static int mlxsw_sp_mr_tcam_route_irif_update(struct mlxsw_sp *mlxsw_sp, + void *route_priv, u16 irif_index) +{ + struct mlxsw_sp_mr_tcam_route *route = route_priv; + + if (route->action != MLXSW_SP_MR_ROUTE_ACTION_TRAP) + return -EINVAL; + route->irif_index = irif_index; + return 0; +} + +static int mlxsw_sp_mr_tcam_route_erif_add(struct mlxsw_sp *mlxsw_sp, + void *route_priv, u16 erif_index) +{ + struct mlxsw_sp_mr_tcam_route *route = route_priv; + int err; + + err = mlxsw_sp_mr_erif_list_add(mlxsw_sp, &route->erif_list, + erif_index); + if (err) + return err; + + /* Commit the action only if the route action is not TRAP */ + if (route->action != MLXSW_SP_MR_ROUTE_ACTION_TRAP) + return mlxsw_sp_mr_erif_list_commit(mlxsw_sp, + &route->erif_list); + return 0; +} + +static int mlxsw_sp_mr_tcam_route_erif_del(struct mlxsw_sp *mlxsw_sp, + void *route_priv, u16 erif_index) +{ + const struct mlxsw_sp_mr_tcam_ops *ops = mlxsw_sp->mr_tcam_ops; + struct mlxsw_sp_mr_tcam_route *route = route_priv; + struct mlxsw_sp_mr_erif_sublist *erif_sublist; + struct mlxsw_sp_mr_tcam_erif_list erif_list; + struct mlxsw_afa_block *afa_block; + int err; + int i; + + /* Create a copy of the original erif_list without the deleted entry */ + mlxsw_sp_mr_erif_list_init(&erif_list); + list_for_each_entry(erif_sublist, &route->erif_list.erif_sublists, list) { + for (i = 0; i < erif_sublist->num_erifs; i++) { + u16 curr_erif = erif_sublist->erif_indices[i]; + + if (curr_erif == erif_index) + continue; + err = mlxsw_sp_mr_erif_list_add(mlxsw_sp, &erif_list, + curr_erif); + if (err) + goto err_erif_list_add; + } + } + + /* Create the flexible action block pointing to the new erif_list */ + afa_block = mlxsw_sp_mr_tcam_afa_block_create(mlxsw_sp, route->action, + route->irif_index, + route->counter_index, + route->min_mtu, + &erif_list); + if (IS_ERR(afa_block)) { + err = PTR_ERR(afa_block); + goto err_afa_block_create; + } + + /* Update the TCAM route entry */ + err = ops->route_update(mlxsw_sp, route->priv, &route->key, afa_block); + if (err) + goto err_route_write; + + mlxsw_sp_mr_tcam_afa_block_destroy(route->afa_block); + mlxsw_sp_mr_erif_list_flush(mlxsw_sp, &route->erif_list); + route->afa_block = afa_block; + mlxsw_sp_mr_erif_list_move(&route->erif_list, &erif_list); + return 0; + +err_route_write: + mlxsw_sp_mr_tcam_afa_block_destroy(afa_block); +err_afa_block_create: +err_erif_list_add: + mlxsw_sp_mr_erif_list_flush(mlxsw_sp, &erif_list); + return err; +} + +static int +mlxsw_sp_mr_tcam_route_update(struct mlxsw_sp *mlxsw_sp, void *route_priv, + struct mlxsw_sp_mr_route_info *route_info) +{ + const struct mlxsw_sp_mr_tcam_ops *ops = mlxsw_sp->mr_tcam_ops; + struct mlxsw_sp_mr_tcam_route *route = route_priv; + struct mlxsw_sp_mr_tcam_erif_list erif_list; + struct mlxsw_afa_block *afa_block; + int err; + + /* Create a new erif_list */ + mlxsw_sp_mr_erif_list_init(&erif_list); + err = mlxsw_sp_mr_tcam_erif_populate(mlxsw_sp, &erif_list, route_info); + if (err) + goto err_erif_populate; + + /* Create the flexible action block pointing to the new erif_list */ + afa_block = mlxsw_sp_mr_tcam_afa_block_create(mlxsw_sp, + route_info->route_action, + route_info->irif_index, + route->counter_index, + route_info->min_mtu, + &erif_list); + if (IS_ERR(afa_block)) { + err = PTR_ERR(afa_block); + goto err_afa_block_create; + } + + /* Update the TCAM route entry */ + err = ops->route_update(mlxsw_sp, route->priv, &route->key, afa_block); + if (err) + goto err_route_write; + + mlxsw_sp_mr_tcam_afa_block_destroy(route->afa_block); + mlxsw_sp_mr_erif_list_flush(mlxsw_sp, &route->erif_list); + route->afa_block = afa_block; + mlxsw_sp_mr_erif_list_move(&route->erif_list, &erif_list); + route->action = route_info->route_action; + route->irif_index = route_info->irif_index; + route->min_mtu = route_info->min_mtu; + return 0; + +err_route_write: + mlxsw_sp_mr_tcam_afa_block_destroy(afa_block); +err_afa_block_create: +err_erif_populate: + mlxsw_sp_mr_erif_list_flush(mlxsw_sp, &erif_list); + return err; +} + +static int mlxsw_sp_mr_tcam_init(struct mlxsw_sp *mlxsw_sp, void *priv) +{ + const struct mlxsw_sp_mr_tcam_ops *ops = mlxsw_sp->mr_tcam_ops; + struct mlxsw_sp_mr_tcam *mr_tcam = priv; + int err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, MC_ERIF_LIST_ENTRIES)) + return -EIO; + + mr_tcam->priv = kzalloc(ops->priv_size, GFP_KERNEL); + if (!mr_tcam->priv) + return -ENOMEM; + + err = ops->init(mlxsw_sp, mr_tcam->priv); + if (err) + goto err_init; + return 0; + +err_init: + kfree(mr_tcam->priv); + return err; +} + +static void mlxsw_sp_mr_tcam_fini(struct mlxsw_sp *mlxsw_sp, void *priv) +{ + const struct mlxsw_sp_mr_tcam_ops *ops = mlxsw_sp->mr_tcam_ops; + struct mlxsw_sp_mr_tcam *mr_tcam = priv; + + ops->fini(mr_tcam->priv); + kfree(mr_tcam->priv); +} + +const struct mlxsw_sp_mr_ops mlxsw_sp_mr_tcam_ops = { + .priv_size = sizeof(struct mlxsw_sp_mr_tcam), + .route_priv_size = sizeof(struct mlxsw_sp_mr_tcam_route), + .init = mlxsw_sp_mr_tcam_init, + .route_create = mlxsw_sp_mr_tcam_route_create, + .route_update = mlxsw_sp_mr_tcam_route_update, + .route_stats = mlxsw_sp_mr_tcam_route_stats, + .route_action_update = mlxsw_sp_mr_tcam_route_action_update, + .route_min_mtu_update = mlxsw_sp_mr_tcam_route_min_mtu_update, + .route_irif_update = mlxsw_sp_mr_tcam_route_irif_update, + .route_erif_add = mlxsw_sp_mr_tcam_route_erif_add, + .route_erif_del = mlxsw_sp_mr_tcam_route_erif_del, + .route_destroy = mlxsw_sp_mr_tcam_route_destroy, + .fini = mlxsw_sp_mr_tcam_fini, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.h new file mode 100644 index 000000000..3c84151b4 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.h @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_SPECTRUM_MCROUTER_TCAM_H +#define _MLXSW_SPECTRUM_MCROUTER_TCAM_H + +#include "spectrum.h" +#include "spectrum_mr.h" + +extern const struct mlxsw_sp_mr_ops mlxsw_sp_mr_tcam_ops; + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_nve.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_nve.c new file mode 100644 index 000000000..d2b57a045 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_nve.c @@ -0,0 +1,1173 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#include <linux/err.h> +#include <linux/gfp.h> +#include <linux/kernel.h> +#include <linux/list.h> +#include <linux/netlink.h> +#include <linux/rtnetlink.h> +#include <linux/slab.h> +#include <net/inet_ecn.h> +#include <net/ipv6.h> + +#include "reg.h" +#include "spectrum.h" +#include "spectrum_nve.h" + +const struct mlxsw_sp_nve_ops *mlxsw_sp1_nve_ops_arr[] = { + [MLXSW_SP_NVE_TYPE_VXLAN] = &mlxsw_sp1_nve_vxlan_ops, +}; + +const struct mlxsw_sp_nve_ops *mlxsw_sp2_nve_ops_arr[] = { + [MLXSW_SP_NVE_TYPE_VXLAN] = &mlxsw_sp2_nve_vxlan_ops, +}; + +struct mlxsw_sp_nve_mc_entry; +struct mlxsw_sp_nve_mc_record; +struct mlxsw_sp_nve_mc_list; + +struct mlxsw_sp_nve_mc_record_ops { + enum mlxsw_reg_tnumt_record_type type; + int (*entry_add)(struct mlxsw_sp_nve_mc_record *mc_record, + struct mlxsw_sp_nve_mc_entry *mc_entry, + const union mlxsw_sp_l3addr *addr); + void (*entry_del)(const struct mlxsw_sp_nve_mc_record *mc_record, + const struct mlxsw_sp_nve_mc_entry *mc_entry); + void (*entry_set)(const struct mlxsw_sp_nve_mc_record *mc_record, + const struct mlxsw_sp_nve_mc_entry *mc_entry, + char *tnumt_pl, unsigned int entry_index); + bool (*entry_compare)(const struct mlxsw_sp_nve_mc_record *mc_record, + const struct mlxsw_sp_nve_mc_entry *mc_entry, + const union mlxsw_sp_l3addr *addr); +}; + +struct mlxsw_sp_nve_mc_list_key { + u16 fid_index; +}; + +struct mlxsw_sp_nve_mc_ipv6_entry { + struct in6_addr addr6; + u32 addr6_kvdl_index; +}; + +struct mlxsw_sp_nve_mc_entry { + union { + __be32 addr4; + struct mlxsw_sp_nve_mc_ipv6_entry ipv6_entry; + }; + u8 valid:1; +}; + +struct mlxsw_sp_nve_mc_record { + struct list_head list; + enum mlxsw_sp_l3proto proto; + unsigned int num_entries; + struct mlxsw_sp *mlxsw_sp; + struct mlxsw_sp_nve_mc_list *mc_list; + const struct mlxsw_sp_nve_mc_record_ops *ops; + u32 kvdl_index; + struct mlxsw_sp_nve_mc_entry entries[]; +}; + +struct mlxsw_sp_nve_mc_list { + struct list_head records_list; + struct rhash_head ht_node; + struct mlxsw_sp_nve_mc_list_key key; +}; + +static const struct rhashtable_params mlxsw_sp_nve_mc_list_ht_params = { + .key_len = sizeof(struct mlxsw_sp_nve_mc_list_key), + .key_offset = offsetof(struct mlxsw_sp_nve_mc_list, key), + .head_offset = offsetof(struct mlxsw_sp_nve_mc_list, ht_node), +}; + +static int +mlxsw_sp_nve_mc_record_ipv4_entry_add(struct mlxsw_sp_nve_mc_record *mc_record, + struct mlxsw_sp_nve_mc_entry *mc_entry, + const union mlxsw_sp_l3addr *addr) +{ + mc_entry->addr4 = addr->addr4; + + return 0; +} + +static void +mlxsw_sp_nve_mc_record_ipv4_entry_del(const struct mlxsw_sp_nve_mc_record *mc_record, + const struct mlxsw_sp_nve_mc_entry *mc_entry) +{ +} + +static void +mlxsw_sp_nve_mc_record_ipv4_entry_set(const struct mlxsw_sp_nve_mc_record *mc_record, + const struct mlxsw_sp_nve_mc_entry *mc_entry, + char *tnumt_pl, unsigned int entry_index) +{ + u32 udip = be32_to_cpu(mc_entry->addr4); + + mlxsw_reg_tnumt_udip_set(tnumt_pl, entry_index, udip); +} + +static bool +mlxsw_sp_nve_mc_record_ipv4_entry_compare(const struct mlxsw_sp_nve_mc_record *mc_record, + const struct mlxsw_sp_nve_mc_entry *mc_entry, + const union mlxsw_sp_l3addr *addr) +{ + return mc_entry->addr4 == addr->addr4; +} + +static const struct mlxsw_sp_nve_mc_record_ops +mlxsw_sp_nve_mc_record_ipv4_ops = { + .type = MLXSW_REG_TNUMT_RECORD_TYPE_IPV4, + .entry_add = &mlxsw_sp_nve_mc_record_ipv4_entry_add, + .entry_del = &mlxsw_sp_nve_mc_record_ipv4_entry_del, + .entry_set = &mlxsw_sp_nve_mc_record_ipv4_entry_set, + .entry_compare = &mlxsw_sp_nve_mc_record_ipv4_entry_compare, +}; + +static int +mlxsw_sp_nve_mc_record_ipv6_entry_add(struct mlxsw_sp_nve_mc_record *mc_record, + struct mlxsw_sp_nve_mc_entry *mc_entry, + const union mlxsw_sp_l3addr *addr) +{ + u32 kvdl_index; + int err; + + err = mlxsw_sp_ipv6_addr_kvdl_index_get(mc_record->mlxsw_sp, + &addr->addr6, &kvdl_index); + if (err) + return err; + + mc_entry->ipv6_entry.addr6 = addr->addr6; + mc_entry->ipv6_entry.addr6_kvdl_index = kvdl_index; + return 0; +} + +static void +mlxsw_sp_nve_mc_record_ipv6_entry_del(const struct mlxsw_sp_nve_mc_record *mc_record, + const struct mlxsw_sp_nve_mc_entry *mc_entry) +{ + mlxsw_sp_ipv6_addr_put(mc_record->mlxsw_sp, + &mc_entry->ipv6_entry.addr6); +} + +static void +mlxsw_sp_nve_mc_record_ipv6_entry_set(const struct mlxsw_sp_nve_mc_record *mc_record, + const struct mlxsw_sp_nve_mc_entry *mc_entry, + char *tnumt_pl, unsigned int entry_index) +{ + u32 udip_ptr = mc_entry->ipv6_entry.addr6_kvdl_index; + + mlxsw_reg_tnumt_udip_ptr_set(tnumt_pl, entry_index, udip_ptr); +} + +static bool +mlxsw_sp_nve_mc_record_ipv6_entry_compare(const struct mlxsw_sp_nve_mc_record *mc_record, + const struct mlxsw_sp_nve_mc_entry *mc_entry, + const union mlxsw_sp_l3addr *addr) +{ + return ipv6_addr_equal(&mc_entry->ipv6_entry.addr6, &addr->addr6); +} + +static const struct mlxsw_sp_nve_mc_record_ops +mlxsw_sp_nve_mc_record_ipv6_ops = { + .type = MLXSW_REG_TNUMT_RECORD_TYPE_IPV6, + .entry_add = &mlxsw_sp_nve_mc_record_ipv6_entry_add, + .entry_del = &mlxsw_sp_nve_mc_record_ipv6_entry_del, + .entry_set = &mlxsw_sp_nve_mc_record_ipv6_entry_set, + .entry_compare = &mlxsw_sp_nve_mc_record_ipv6_entry_compare, +}; + +static const struct mlxsw_sp_nve_mc_record_ops * +mlxsw_sp_nve_mc_record_ops_arr[] = { + [MLXSW_SP_L3_PROTO_IPV4] = &mlxsw_sp_nve_mc_record_ipv4_ops, + [MLXSW_SP_L3_PROTO_IPV6] = &mlxsw_sp_nve_mc_record_ipv6_ops, +}; + +int mlxsw_sp_nve_learned_ip_resolve(struct mlxsw_sp *mlxsw_sp, u32 uip, + enum mlxsw_sp_l3proto proto, + union mlxsw_sp_l3addr *addr) +{ + switch (proto) { + case MLXSW_SP_L3_PROTO_IPV4: + addr->addr4 = cpu_to_be32(uip); + return 0; + default: + WARN_ON(1); + return -EINVAL; + } +} + +static struct mlxsw_sp_nve_mc_list * +mlxsw_sp_nve_mc_list_find(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_nve_mc_list_key *key) +{ + struct mlxsw_sp_nve *nve = mlxsw_sp->nve; + + return rhashtable_lookup_fast(&nve->mc_list_ht, key, + mlxsw_sp_nve_mc_list_ht_params); +} + +static struct mlxsw_sp_nve_mc_list * +mlxsw_sp_nve_mc_list_create(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_nve_mc_list_key *key) +{ + struct mlxsw_sp_nve *nve = mlxsw_sp->nve; + struct mlxsw_sp_nve_mc_list *mc_list; + int err; + + mc_list = kmalloc(sizeof(*mc_list), GFP_KERNEL); + if (!mc_list) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&mc_list->records_list); + mc_list->key = *key; + + err = rhashtable_insert_fast(&nve->mc_list_ht, &mc_list->ht_node, + mlxsw_sp_nve_mc_list_ht_params); + if (err) + goto err_rhashtable_insert; + + return mc_list; + +err_rhashtable_insert: + kfree(mc_list); + return ERR_PTR(err); +} + +static void mlxsw_sp_nve_mc_list_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nve_mc_list *mc_list) +{ + struct mlxsw_sp_nve *nve = mlxsw_sp->nve; + + rhashtable_remove_fast(&nve->mc_list_ht, &mc_list->ht_node, + mlxsw_sp_nve_mc_list_ht_params); + WARN_ON(!list_empty(&mc_list->records_list)); + kfree(mc_list); +} + +static struct mlxsw_sp_nve_mc_list * +mlxsw_sp_nve_mc_list_get(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_nve_mc_list_key *key) +{ + struct mlxsw_sp_nve_mc_list *mc_list; + + mc_list = mlxsw_sp_nve_mc_list_find(mlxsw_sp, key); + if (mc_list) + return mc_list; + + return mlxsw_sp_nve_mc_list_create(mlxsw_sp, key); +} + +static void +mlxsw_sp_nve_mc_list_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nve_mc_list *mc_list) +{ + if (!list_empty(&mc_list->records_list)) + return; + mlxsw_sp_nve_mc_list_destroy(mlxsw_sp, mc_list); +} + +static struct mlxsw_sp_nve_mc_record * +mlxsw_sp_nve_mc_record_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nve_mc_list *mc_list, + enum mlxsw_sp_l3proto proto) +{ + unsigned int num_max_entries = mlxsw_sp->nve->num_max_mc_entries[proto]; + struct mlxsw_sp_nve_mc_record *mc_record; + int err; + + mc_record = kzalloc(struct_size(mc_record, entries, num_max_entries), + GFP_KERNEL); + if (!mc_record) + return ERR_PTR(-ENOMEM); + + err = mlxsw_sp_kvdl_alloc(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_TNUMT, 1, + &mc_record->kvdl_index); + if (err) + goto err_kvdl_alloc; + + mc_record->ops = mlxsw_sp_nve_mc_record_ops_arr[proto]; + mc_record->mlxsw_sp = mlxsw_sp; + mc_record->mc_list = mc_list; + mc_record->proto = proto; + list_add_tail(&mc_record->list, &mc_list->records_list); + + return mc_record; + +err_kvdl_alloc: + kfree(mc_record); + return ERR_PTR(err); +} + +static void +mlxsw_sp_nve_mc_record_destroy(struct mlxsw_sp_nve_mc_record *mc_record) +{ + struct mlxsw_sp *mlxsw_sp = mc_record->mlxsw_sp; + + list_del(&mc_record->list); + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_TNUMT, 1, + mc_record->kvdl_index); + WARN_ON(mc_record->num_entries); + kfree(mc_record); +} + +static struct mlxsw_sp_nve_mc_record * +mlxsw_sp_nve_mc_record_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nve_mc_list *mc_list, + enum mlxsw_sp_l3proto proto) +{ + struct mlxsw_sp_nve_mc_record *mc_record; + + list_for_each_entry_reverse(mc_record, &mc_list->records_list, list) { + unsigned int num_entries = mc_record->num_entries; + struct mlxsw_sp_nve *nve = mlxsw_sp->nve; + + if (mc_record->proto == proto && + num_entries < nve->num_max_mc_entries[proto]) + return mc_record; + } + + return mlxsw_sp_nve_mc_record_create(mlxsw_sp, mc_list, proto); +} + +static void +mlxsw_sp_nve_mc_record_put(struct mlxsw_sp_nve_mc_record *mc_record) +{ + if (mc_record->num_entries != 0) + return; + + mlxsw_sp_nve_mc_record_destroy(mc_record); +} + +static struct mlxsw_sp_nve_mc_entry * +mlxsw_sp_nve_mc_free_entry_find(struct mlxsw_sp_nve_mc_record *mc_record) +{ + struct mlxsw_sp_nve *nve = mc_record->mlxsw_sp->nve; + unsigned int num_max_entries; + int i; + + num_max_entries = nve->num_max_mc_entries[mc_record->proto]; + for (i = 0; i < num_max_entries; i++) { + if (mc_record->entries[i].valid) + continue; + return &mc_record->entries[i]; + } + + return NULL; +} + +static int +mlxsw_sp_nve_mc_record_refresh(struct mlxsw_sp_nve_mc_record *mc_record) +{ + enum mlxsw_reg_tnumt_record_type type = mc_record->ops->type; + struct mlxsw_sp_nve_mc_list *mc_list = mc_record->mc_list; + struct mlxsw_sp *mlxsw_sp = mc_record->mlxsw_sp; + char tnumt_pl[MLXSW_REG_TNUMT_LEN]; + unsigned int num_max_entries; + unsigned int num_entries = 0; + u32 next_kvdl_index = 0; + bool next_valid = false; + int i; + + if (!list_is_last(&mc_record->list, &mc_list->records_list)) { + struct mlxsw_sp_nve_mc_record *next_record; + + next_record = list_next_entry(mc_record, list); + next_kvdl_index = next_record->kvdl_index; + next_valid = true; + } + + mlxsw_reg_tnumt_pack(tnumt_pl, type, MLXSW_REG_TUNNEL_PORT_NVE, + mc_record->kvdl_index, next_valid, + next_kvdl_index, mc_record->num_entries); + + num_max_entries = mlxsw_sp->nve->num_max_mc_entries[mc_record->proto]; + for (i = 0; i < num_max_entries; i++) { + struct mlxsw_sp_nve_mc_entry *mc_entry; + + mc_entry = &mc_record->entries[i]; + if (!mc_entry->valid) + continue; + mc_record->ops->entry_set(mc_record, mc_entry, tnumt_pl, + num_entries++); + } + + WARN_ON(num_entries != mc_record->num_entries); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tnumt), tnumt_pl); +} + +static bool +mlxsw_sp_nve_mc_record_is_first(struct mlxsw_sp_nve_mc_record *mc_record) +{ + struct mlxsw_sp_nve_mc_list *mc_list = mc_record->mc_list; + struct mlxsw_sp_nve_mc_record *first_record; + + first_record = list_first_entry(&mc_list->records_list, + struct mlxsw_sp_nve_mc_record, list); + + return mc_record == first_record; +} + +static struct mlxsw_sp_nve_mc_entry * +mlxsw_sp_nve_mc_entry_find(struct mlxsw_sp_nve_mc_record *mc_record, + union mlxsw_sp_l3addr *addr) +{ + struct mlxsw_sp_nve *nve = mc_record->mlxsw_sp->nve; + unsigned int num_max_entries; + int i; + + num_max_entries = nve->num_max_mc_entries[mc_record->proto]; + for (i = 0; i < num_max_entries; i++) { + struct mlxsw_sp_nve_mc_entry *mc_entry; + + mc_entry = &mc_record->entries[i]; + if (!mc_entry->valid) + continue; + if (mc_record->ops->entry_compare(mc_record, mc_entry, addr)) + return mc_entry; + } + + return NULL; +} + +static int +mlxsw_sp_nve_mc_record_ip_add(struct mlxsw_sp_nve_mc_record *mc_record, + union mlxsw_sp_l3addr *addr) +{ + struct mlxsw_sp_nve_mc_entry *mc_entry = NULL; + int err; + + mc_entry = mlxsw_sp_nve_mc_free_entry_find(mc_record); + if (WARN_ON(!mc_entry)) + return -EINVAL; + + err = mc_record->ops->entry_add(mc_record, mc_entry, addr); + if (err) + return err; + mc_record->num_entries++; + mc_entry->valid = true; + + err = mlxsw_sp_nve_mc_record_refresh(mc_record); + if (err) + goto err_record_refresh; + + /* If this is a new record and not the first one, then we need to + * update the next pointer of the previous entry + */ + if (mc_record->num_entries != 1 || + mlxsw_sp_nve_mc_record_is_first(mc_record)) + return 0; + + err = mlxsw_sp_nve_mc_record_refresh(list_prev_entry(mc_record, list)); + if (err) + goto err_prev_record_refresh; + + return 0; + +err_prev_record_refresh: +err_record_refresh: + mc_entry->valid = false; + mc_record->num_entries--; + mc_record->ops->entry_del(mc_record, mc_entry); + return err; +} + +static void +mlxsw_sp_nve_mc_record_entry_del(struct mlxsw_sp_nve_mc_record *mc_record, + struct mlxsw_sp_nve_mc_entry *mc_entry) +{ + struct mlxsw_sp_nve_mc_list *mc_list = mc_record->mc_list; + + mc_entry->valid = false; + mc_record->num_entries--; + + /* When the record continues to exist we only need to invalidate + * the requested entry + */ + if (mc_record->num_entries != 0) { + mlxsw_sp_nve_mc_record_refresh(mc_record); + mc_record->ops->entry_del(mc_record, mc_entry); + return; + } + + /* If the record needs to be deleted, but it is not the first, + * then we need to make sure that the previous record no longer + * points to it. Remove deleted record from the list to reflect + * that and then re-add it at the end, so that it could be + * properly removed by the record destruction code + */ + if (!mlxsw_sp_nve_mc_record_is_first(mc_record)) { + struct mlxsw_sp_nve_mc_record *prev_record; + + prev_record = list_prev_entry(mc_record, list); + list_del(&mc_record->list); + mlxsw_sp_nve_mc_record_refresh(prev_record); + list_add_tail(&mc_record->list, &mc_list->records_list); + mc_record->ops->entry_del(mc_record, mc_entry); + return; + } + + /* If the first record needs to be deleted, but the list is not + * singular, then the second record needs to be written in the + * first record's address, as this address is stored as a property + * of the FID + */ + if (mlxsw_sp_nve_mc_record_is_first(mc_record) && + !list_is_singular(&mc_list->records_list)) { + struct mlxsw_sp_nve_mc_record *next_record; + + next_record = list_next_entry(mc_record, list); + swap(mc_record->kvdl_index, next_record->kvdl_index); + mlxsw_sp_nve_mc_record_refresh(next_record); + mc_record->ops->entry_del(mc_record, mc_entry); + return; + } + + /* This is the last case where the last remaining record needs to + * be deleted. Simply delete the entry + */ + mc_record->ops->entry_del(mc_record, mc_entry); +} + +static struct mlxsw_sp_nve_mc_record * +mlxsw_sp_nve_mc_record_find(struct mlxsw_sp_nve_mc_list *mc_list, + enum mlxsw_sp_l3proto proto, + union mlxsw_sp_l3addr *addr, + struct mlxsw_sp_nve_mc_entry **mc_entry) +{ + struct mlxsw_sp_nve_mc_record *mc_record; + + list_for_each_entry(mc_record, &mc_list->records_list, list) { + if (mc_record->proto != proto) + continue; + + *mc_entry = mlxsw_sp_nve_mc_entry_find(mc_record, addr); + if (*mc_entry) + return mc_record; + } + + return NULL; +} + +static int mlxsw_sp_nve_mc_list_ip_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nve_mc_list *mc_list, + enum mlxsw_sp_l3proto proto, + union mlxsw_sp_l3addr *addr) +{ + struct mlxsw_sp_nve_mc_record *mc_record; + int err; + + mc_record = mlxsw_sp_nve_mc_record_get(mlxsw_sp, mc_list, proto); + if (IS_ERR(mc_record)) + return PTR_ERR(mc_record); + + err = mlxsw_sp_nve_mc_record_ip_add(mc_record, addr); + if (err) + goto err_ip_add; + + return 0; + +err_ip_add: + mlxsw_sp_nve_mc_record_put(mc_record); + return err; +} + +static void mlxsw_sp_nve_mc_list_ip_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nve_mc_list *mc_list, + enum mlxsw_sp_l3proto proto, + union mlxsw_sp_l3addr *addr) +{ + struct mlxsw_sp_nve_mc_record *mc_record; + struct mlxsw_sp_nve_mc_entry *mc_entry; + + mc_record = mlxsw_sp_nve_mc_record_find(mc_list, proto, addr, + &mc_entry); + if (!mc_record) + return; + + mlxsw_sp_nve_mc_record_entry_del(mc_record, mc_entry); + mlxsw_sp_nve_mc_record_put(mc_record); +} + +static int +mlxsw_sp_nve_fid_flood_index_set(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_nve_mc_list *mc_list) +{ + struct mlxsw_sp_nve_mc_record *mc_record; + + /* The address of the first record in the list is a property of + * the FID and we never change it. It only needs to be set when + * a new list is created + */ + if (mlxsw_sp_fid_nve_flood_index_is_set(fid)) + return 0; + + mc_record = list_first_entry(&mc_list->records_list, + struct mlxsw_sp_nve_mc_record, list); + + return mlxsw_sp_fid_nve_flood_index_set(fid, mc_record->kvdl_index); +} + +static void +mlxsw_sp_nve_fid_flood_index_clear(struct mlxsw_sp_fid *fid, + struct mlxsw_sp_nve_mc_list *mc_list) +{ + struct mlxsw_sp_nve_mc_record *mc_record; + + /* The address of the first record needs to be invalidated only when + * the last record is about to be removed + */ + if (!list_is_singular(&mc_list->records_list)) + return; + + mc_record = list_first_entry(&mc_list->records_list, + struct mlxsw_sp_nve_mc_record, list); + if (mc_record->num_entries != 1) + return; + + return mlxsw_sp_fid_nve_flood_index_clear(fid); +} + +int mlxsw_sp_nve_flood_ip_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fid *fid, + enum mlxsw_sp_l3proto proto, + union mlxsw_sp_l3addr *addr) +{ + struct mlxsw_sp_nve_mc_list_key key = { 0 }; + struct mlxsw_sp_nve_mc_list *mc_list; + int err; + + key.fid_index = mlxsw_sp_fid_index(fid); + mc_list = mlxsw_sp_nve_mc_list_get(mlxsw_sp, &key); + if (IS_ERR(mc_list)) + return PTR_ERR(mc_list); + + err = mlxsw_sp_nve_mc_list_ip_add(mlxsw_sp, mc_list, proto, addr); + if (err) + goto err_add_ip; + + err = mlxsw_sp_nve_fid_flood_index_set(fid, mc_list); + if (err) + goto err_fid_flood_index_set; + + return 0; + +err_fid_flood_index_set: + mlxsw_sp_nve_mc_list_ip_del(mlxsw_sp, mc_list, proto, addr); +err_add_ip: + mlxsw_sp_nve_mc_list_put(mlxsw_sp, mc_list); + return err; +} + +void mlxsw_sp_nve_flood_ip_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fid *fid, + enum mlxsw_sp_l3proto proto, + union mlxsw_sp_l3addr *addr) +{ + struct mlxsw_sp_nve_mc_list_key key = { 0 }; + struct mlxsw_sp_nve_mc_list *mc_list; + + key.fid_index = mlxsw_sp_fid_index(fid); + mc_list = mlxsw_sp_nve_mc_list_find(mlxsw_sp, &key); + if (!mc_list) + return; + + mlxsw_sp_nve_fid_flood_index_clear(fid, mc_list); + mlxsw_sp_nve_mc_list_ip_del(mlxsw_sp, mc_list, proto, addr); + mlxsw_sp_nve_mc_list_put(mlxsw_sp, mc_list); +} + +static void +mlxsw_sp_nve_mc_record_delete(struct mlxsw_sp_nve_mc_record *mc_record) +{ + struct mlxsw_sp_nve *nve = mc_record->mlxsw_sp->nve; + unsigned int num_max_entries; + int i; + + num_max_entries = nve->num_max_mc_entries[mc_record->proto]; + for (i = 0; i < num_max_entries; i++) { + struct mlxsw_sp_nve_mc_entry *mc_entry = &mc_record->entries[i]; + + if (!mc_entry->valid) + continue; + mlxsw_sp_nve_mc_record_entry_del(mc_record, mc_entry); + } + + WARN_ON(mc_record->num_entries); + mlxsw_sp_nve_mc_record_put(mc_record); +} + +static void mlxsw_sp_nve_flood_ip_flush(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fid *fid) +{ + struct mlxsw_sp_nve_mc_record *mc_record, *tmp; + struct mlxsw_sp_nve_mc_list_key key = { 0 }; + struct mlxsw_sp_nve_mc_list *mc_list; + + if (!mlxsw_sp_fid_nve_flood_index_is_set(fid)) + return; + + mlxsw_sp_fid_nve_flood_index_clear(fid); + + key.fid_index = mlxsw_sp_fid_index(fid); + mc_list = mlxsw_sp_nve_mc_list_find(mlxsw_sp, &key); + if (WARN_ON(!mc_list)) + return; + + list_for_each_entry_safe(mc_record, tmp, &mc_list->records_list, list) + mlxsw_sp_nve_mc_record_delete(mc_record); + + WARN_ON(!list_empty(&mc_list->records_list)); + mlxsw_sp_nve_mc_list_put(mlxsw_sp, mc_list); +} + +static int mlxsw_sp_nve_tunnel_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nve_config *config) +{ + struct mlxsw_sp_nve *nve = mlxsw_sp->nve; + const struct mlxsw_sp_nve_ops *ops; + int err; + + if (nve->num_nve_tunnels++ != 0) + return 0; + + nve->config = *config; + + err = mlxsw_sp_kvdl_alloc(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, 1, + &nve->tunnel_index); + if (err) + goto err_kvdl_alloc; + + ops = nve->nve_ops_arr[config->type]; + err = ops->init(nve, config); + if (err) + goto err_ops_init; + + return 0; + +err_ops_init: + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, 1, + nve->tunnel_index); +err_kvdl_alloc: + memset(&nve->config, 0, sizeof(nve->config)); + nve->num_nve_tunnels--; + return err; +} + +static void mlxsw_sp_nve_tunnel_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_nve *nve = mlxsw_sp->nve; + const struct mlxsw_sp_nve_ops *ops; + + ops = nve->nve_ops_arr[nve->config.type]; + + if (mlxsw_sp->nve->num_nve_tunnels == 1) { + ops->fini(nve); + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, 1, + nve->tunnel_index); + memset(&nve->config, 0, sizeof(nve->config)); + } + nve->num_nve_tunnels--; +} + +static void mlxsw_sp_nve_fdb_flush_by_fid(struct mlxsw_sp *mlxsw_sp, + u16 fid_index) +{ + char sfdf_pl[MLXSW_REG_SFDF_LEN]; + + mlxsw_reg_sfdf_pack(sfdf_pl, MLXSW_REG_SFDF_FLUSH_PER_NVE_AND_FID); + mlxsw_reg_sfdf_fid_set(sfdf_pl, fid_index); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sfdf), sfdf_pl); +} + +static void mlxsw_sp_nve_fdb_clear_offload(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_fid *fid, + const struct net_device *nve_dev, + __be32 vni) +{ + const struct mlxsw_sp_nve_ops *ops; + enum mlxsw_sp_nve_type type; + + if (WARN_ON(mlxsw_sp_fid_nve_type(fid, &type))) + return; + + ops = mlxsw_sp->nve->nve_ops_arr[type]; + ops->fdb_clear_offload(nve_dev, vni); +} + +struct mlxsw_sp_nve_ipv6_ht_key { + u8 mac[ETH_ALEN]; + u16 fid_index; +}; + +struct mlxsw_sp_nve_ipv6_ht_node { + struct rhash_head ht_node; + struct list_head list; + struct mlxsw_sp_nve_ipv6_ht_key key; + struct in6_addr addr6; +}; + +static const struct rhashtable_params mlxsw_sp_nve_ipv6_ht_params = { + .key_len = sizeof(struct mlxsw_sp_nve_ipv6_ht_key), + .key_offset = offsetof(struct mlxsw_sp_nve_ipv6_ht_node, key), + .head_offset = offsetof(struct mlxsw_sp_nve_ipv6_ht_node, ht_node), +}; + +int mlxsw_sp_nve_ipv6_addr_kvdl_set(struct mlxsw_sp *mlxsw_sp, + const struct in6_addr *addr6, + u32 *p_kvdl_index) +{ + return mlxsw_sp_ipv6_addr_kvdl_index_get(mlxsw_sp, addr6, p_kvdl_index); +} + +void mlxsw_sp_nve_ipv6_addr_kvdl_unset(struct mlxsw_sp *mlxsw_sp, + const struct in6_addr *addr6) +{ + mlxsw_sp_ipv6_addr_put(mlxsw_sp, addr6); +} + +static struct mlxsw_sp_nve_ipv6_ht_node * +mlxsw_sp_nve_ipv6_ht_node_lookup(struct mlxsw_sp *mlxsw_sp, const char *mac, + u16 fid_index) +{ + struct mlxsw_sp_nve_ipv6_ht_key key = {}; + + ether_addr_copy(key.mac, mac); + key.fid_index = fid_index; + return rhashtable_lookup_fast(&mlxsw_sp->nve->ipv6_ht, &key, + mlxsw_sp_nve_ipv6_ht_params); +} + +static int mlxsw_sp_nve_ipv6_ht_insert(struct mlxsw_sp *mlxsw_sp, + const char *mac, u16 fid_index, + const struct in6_addr *addr6) +{ + struct mlxsw_sp_nve_ipv6_ht_node *ipv6_ht_node; + struct mlxsw_sp_nve *nve = mlxsw_sp->nve; + int err; + + ipv6_ht_node = kzalloc(sizeof(*ipv6_ht_node), GFP_KERNEL); + if (!ipv6_ht_node) + return -ENOMEM; + + ether_addr_copy(ipv6_ht_node->key.mac, mac); + ipv6_ht_node->key.fid_index = fid_index; + ipv6_ht_node->addr6 = *addr6; + + err = rhashtable_insert_fast(&nve->ipv6_ht, &ipv6_ht_node->ht_node, + mlxsw_sp_nve_ipv6_ht_params); + if (err) + goto err_rhashtable_insert; + + list_add(&ipv6_ht_node->list, &nve->ipv6_addr_list); + + return 0; + +err_rhashtable_insert: + kfree(ipv6_ht_node); + return err; +} + +static void +mlxsw_sp_nve_ipv6_ht_remove(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nve_ipv6_ht_node *ipv6_ht_node) +{ + struct mlxsw_sp_nve *nve = mlxsw_sp->nve; + + list_del(&ipv6_ht_node->list); + rhashtable_remove_fast(&nve->ipv6_ht, &ipv6_ht_node->ht_node, + mlxsw_sp_nve_ipv6_ht_params); + kfree(ipv6_ht_node); +} + +int +mlxsw_sp_nve_ipv6_addr_map_replace(struct mlxsw_sp *mlxsw_sp, const char *mac, + u16 fid_index, + const struct in6_addr *new_addr6) +{ + struct mlxsw_sp_nve_ipv6_ht_node *ipv6_ht_node; + + ASSERT_RTNL(); + + ipv6_ht_node = mlxsw_sp_nve_ipv6_ht_node_lookup(mlxsw_sp, mac, + fid_index); + if (!ipv6_ht_node) + return mlxsw_sp_nve_ipv6_ht_insert(mlxsw_sp, mac, fid_index, + new_addr6); + + mlxsw_sp_ipv6_addr_put(mlxsw_sp, &ipv6_ht_node->addr6); + ipv6_ht_node->addr6 = *new_addr6; + return 0; +} + +void mlxsw_sp_nve_ipv6_addr_map_del(struct mlxsw_sp *mlxsw_sp, const char *mac, + u16 fid_index) +{ + struct mlxsw_sp_nve_ipv6_ht_node *ipv6_ht_node; + + ASSERT_RTNL(); + + ipv6_ht_node = mlxsw_sp_nve_ipv6_ht_node_lookup(mlxsw_sp, mac, + fid_index); + if (WARN_ON(!ipv6_ht_node)) + return; + + mlxsw_sp_nve_ipv6_ht_remove(mlxsw_sp, ipv6_ht_node); +} + +static void mlxsw_sp_nve_ipv6_addr_flush_by_fid(struct mlxsw_sp *mlxsw_sp, + u16 fid_index) +{ + struct mlxsw_sp_nve_ipv6_ht_node *ipv6_ht_node, *tmp; + struct mlxsw_sp_nve *nve = mlxsw_sp->nve; + + list_for_each_entry_safe(ipv6_ht_node, tmp, &nve->ipv6_addr_list, + list) { + if (ipv6_ht_node->key.fid_index != fid_index) + continue; + + mlxsw_sp_ipv6_addr_put(mlxsw_sp, &ipv6_ht_node->addr6); + mlxsw_sp_nve_ipv6_ht_remove(mlxsw_sp, ipv6_ht_node); + } +} + +int mlxsw_sp_nve_fid_enable(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fid *fid, + struct mlxsw_sp_nve_params *params, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_nve *nve = mlxsw_sp->nve; + const struct mlxsw_sp_nve_ops *ops; + struct mlxsw_sp_nve_config config; + int err; + + ops = nve->nve_ops_arr[params->type]; + + if (!ops->can_offload(nve, params, extack)) + return -EINVAL; + + memset(&config, 0, sizeof(config)); + ops->nve_config(nve, params, &config); + if (nve->num_nve_tunnels && + memcmp(&config, &nve->config, sizeof(config))) { + NL_SET_ERR_MSG_MOD(extack, "Conflicting NVE tunnels configuration"); + return -EINVAL; + } + + err = mlxsw_sp_nve_tunnel_init(mlxsw_sp, &config); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to initialize NVE tunnel"); + return err; + } + + err = mlxsw_sp_fid_vni_set(fid, params->type, params->vni, + params->dev->ifindex); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to set VNI on FID"); + goto err_fid_vni_set; + } + + err = ops->fdb_replay(params->dev, params->vni, extack); + if (err) + goto err_fdb_replay; + + return 0; + +err_fdb_replay: + mlxsw_sp_fid_vni_clear(fid); +err_fid_vni_set: + mlxsw_sp_nve_tunnel_fini(mlxsw_sp); + return err; +} + +void mlxsw_sp_nve_fid_disable(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fid *fid) +{ + u16 fid_index = mlxsw_sp_fid_index(fid); + struct net_device *nve_dev; + int nve_ifindex; + __be32 vni; + + mlxsw_sp_nve_flood_ip_flush(mlxsw_sp, fid); + mlxsw_sp_nve_fdb_flush_by_fid(mlxsw_sp, fid_index); + mlxsw_sp_nve_ipv6_addr_flush_by_fid(mlxsw_sp, fid_index); + + if (WARN_ON(mlxsw_sp_fid_nve_ifindex(fid, &nve_ifindex) || + mlxsw_sp_fid_vni(fid, &vni))) + goto out; + + nve_dev = dev_get_by_index(mlxsw_sp_net(mlxsw_sp), nve_ifindex); + if (!nve_dev) + goto out; + + mlxsw_sp_nve_fdb_clear_offload(mlxsw_sp, fid, nve_dev, vni); + mlxsw_sp_fid_fdb_clear_offload(fid, nve_dev); + + dev_put(nve_dev); + +out: + mlxsw_sp_fid_vni_clear(fid); + mlxsw_sp_nve_tunnel_fini(mlxsw_sp); +} + +int mlxsw_sp_port_nve_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char tnqdr_pl[MLXSW_REG_TNQDR_LEN]; + + mlxsw_reg_tnqdr_pack(tnqdr_pl, mlxsw_sp_port->local_port); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tnqdr), tnqdr_pl); +} + +void mlxsw_sp_port_nve_fini(struct mlxsw_sp_port *mlxsw_sp_port) +{ +} + +static int mlxsw_sp_nve_qos_init(struct mlxsw_sp *mlxsw_sp) +{ + char tnqcr_pl[MLXSW_REG_TNQCR_LEN]; + + mlxsw_reg_tnqcr_pack(tnqcr_pl); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tnqcr), tnqcr_pl); +} + +static int mlxsw_sp_nve_ecn_encap_init(struct mlxsw_sp *mlxsw_sp) +{ + int i; + + /* Iterate over inner ECN values */ + for (i = INET_ECN_NOT_ECT; i <= INET_ECN_CE; i++) { + u8 outer_ecn = INET_ECN_encapsulate(0, i); + char tneem_pl[MLXSW_REG_TNEEM_LEN]; + int err; + + mlxsw_reg_tneem_pack(tneem_pl, i, outer_ecn); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tneem), + tneem_pl); + if (err) + return err; + } + + return 0; +} + +static int __mlxsw_sp_nve_ecn_decap_init(struct mlxsw_sp *mlxsw_sp, + u8 inner_ecn, u8 outer_ecn) +{ + char tndem_pl[MLXSW_REG_TNDEM_LEN]; + u8 new_inner_ecn; + bool trap_en; + + new_inner_ecn = mlxsw_sp_tunnel_ecn_decap(outer_ecn, inner_ecn, + &trap_en); + mlxsw_reg_tndem_pack(tndem_pl, outer_ecn, inner_ecn, new_inner_ecn, + trap_en, trap_en ? MLXSW_TRAP_ID_DECAP_ECN0 : 0); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tndem), tndem_pl); +} + +static int mlxsw_sp_nve_ecn_decap_init(struct mlxsw_sp *mlxsw_sp) +{ + int i; + + /* Iterate over inner ECN values */ + for (i = INET_ECN_NOT_ECT; i <= INET_ECN_CE; i++) { + int j; + + /* Iterate over outer ECN values */ + for (j = INET_ECN_NOT_ECT; j <= INET_ECN_CE; j++) { + int err; + + err = __mlxsw_sp_nve_ecn_decap_init(mlxsw_sp, i, j); + if (err) + return err; + } + } + + return 0; +} + +static int mlxsw_sp_nve_ecn_init(struct mlxsw_sp *mlxsw_sp) +{ + int err; + + err = mlxsw_sp_nve_ecn_encap_init(mlxsw_sp); + if (err) + return err; + + return mlxsw_sp_nve_ecn_decap_init(mlxsw_sp); +} + +static int mlxsw_sp_nve_resources_query(struct mlxsw_sp *mlxsw_sp) +{ + unsigned int max; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, MAX_NVE_MC_ENTRIES_IPV4) || + !MLXSW_CORE_RES_VALID(mlxsw_sp->core, MAX_NVE_MC_ENTRIES_IPV6)) + return -EIO; + max = MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_NVE_MC_ENTRIES_IPV4); + mlxsw_sp->nve->num_max_mc_entries[MLXSW_SP_L3_PROTO_IPV4] = max; + max = MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_NVE_MC_ENTRIES_IPV6); + mlxsw_sp->nve->num_max_mc_entries[MLXSW_SP_L3_PROTO_IPV6] = max; + + return 0; +} + +int mlxsw_sp_nve_init(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_nve *nve; + int err; + + nve = kzalloc(sizeof(*mlxsw_sp->nve), GFP_KERNEL); + if (!nve) + return -ENOMEM; + mlxsw_sp->nve = nve; + nve->mlxsw_sp = mlxsw_sp; + nve->nve_ops_arr = mlxsw_sp->nve_ops_arr; + + err = rhashtable_init(&nve->mc_list_ht, + &mlxsw_sp_nve_mc_list_ht_params); + if (err) + goto err_mc_rhashtable_init; + + err = rhashtable_init(&nve->ipv6_ht, &mlxsw_sp_nve_ipv6_ht_params); + if (err) + goto err_ipv6_rhashtable_init; + + INIT_LIST_HEAD(&nve->ipv6_addr_list); + + err = mlxsw_sp_nve_qos_init(mlxsw_sp); + if (err) + goto err_nve_qos_init; + + err = mlxsw_sp_nve_ecn_init(mlxsw_sp); + if (err) + goto err_nve_ecn_init; + + err = mlxsw_sp_nve_resources_query(mlxsw_sp); + if (err) + goto err_nve_resources_query; + + return 0; + +err_nve_resources_query: +err_nve_ecn_init: +err_nve_qos_init: + rhashtable_destroy(&nve->ipv6_ht); +err_ipv6_rhashtable_init: + rhashtable_destroy(&nve->mc_list_ht); +err_mc_rhashtable_init: + mlxsw_sp->nve = NULL; + kfree(nve); + return err; +} + +void mlxsw_sp_nve_fini(struct mlxsw_sp *mlxsw_sp) +{ + WARN_ON(mlxsw_sp->nve->num_nve_tunnels); + WARN_ON(!list_empty(&mlxsw_sp->nve->ipv6_addr_list)); + rhashtable_destroy(&mlxsw_sp->nve->ipv6_ht); + rhashtable_destroy(&mlxsw_sp->nve->mc_list_ht); + kfree(mlxsw_sp->nve); + mlxsw_sp->nve = NULL; +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_nve.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_nve.h new file mode 100644 index 000000000..0d21de1d0 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_nve.h @@ -0,0 +1,55 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_SPECTRUM_NVE_H +#define _MLXSW_SPECTRUM_NVE_H + +#include <linux/netlink.h> +#include <linux/rhashtable.h> + +#include "spectrum.h" + +struct mlxsw_sp_nve_config { + enum mlxsw_sp_nve_type type; + u8 ttl; + u8 learning_en:1; + __be16 udp_dport; + __be32 flowlabel; + u32 ul_tb_id; + enum mlxsw_sp_l3proto ul_proto; + union mlxsw_sp_l3addr ul_sip; +}; + +struct mlxsw_sp_nve { + struct mlxsw_sp_nve_config config; + struct rhashtable mc_list_ht; + struct rhashtable ipv6_ht; + struct list_head ipv6_addr_list; /* Saves hash table nodes. */ + struct mlxsw_sp *mlxsw_sp; + const struct mlxsw_sp_nve_ops **nve_ops_arr; + unsigned int num_nve_tunnels; /* Protected by RTNL */ + unsigned int num_max_mc_entries[MLXSW_SP_L3_PROTO_MAX]; + u32 tunnel_index; + u16 ul_rif_index; /* Reserved for Spectrum */ +}; + +struct mlxsw_sp_nve_ops { + enum mlxsw_sp_nve_type type; + bool (*can_offload)(const struct mlxsw_sp_nve *nve, + const struct mlxsw_sp_nve_params *params, + struct netlink_ext_ack *extack); + void (*nve_config)(const struct mlxsw_sp_nve *nve, + const struct mlxsw_sp_nve_params *params, + struct mlxsw_sp_nve_config *config); + int (*init)(struct mlxsw_sp_nve *nve, + const struct mlxsw_sp_nve_config *config); + void (*fini)(struct mlxsw_sp_nve *nve); + int (*fdb_replay)(const struct net_device *nve_dev, __be32 vni, + struct netlink_ext_ack *extack); + void (*fdb_clear_offload)(const struct net_device *nve_dev, __be32 vni); +}; + +extern const struct mlxsw_sp_nve_ops mlxsw_sp1_nve_vxlan_ops; +extern const struct mlxsw_sp_nve_ops mlxsw_sp2_nve_vxlan_ops; + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_nve_vxlan.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_nve_vxlan.c new file mode 100644 index 000000000..cdd8818b4 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_nve_vxlan.c @@ -0,0 +1,473 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#include <linux/netdevice.h> +#include <linux/netlink.h> +#include <linux/random.h> +#include <net/vxlan.h> + +#include "reg.h" +#include "spectrum.h" +#include "spectrum_nve.h" + +#define MLXSW_SP_NVE_VXLAN_IPV4_SUPPORTED_FLAGS (VXLAN_F_UDP_ZERO_CSUM_TX | \ + VXLAN_F_LEARN) +#define MLXSW_SP_NVE_VXLAN_IPV6_SUPPORTED_FLAGS (VXLAN_F_IPV6 | \ + VXLAN_F_UDP_ZERO_CSUM6_TX | \ + VXLAN_F_UDP_ZERO_CSUM6_RX) + +static bool mlxsw_sp_nve_vxlan_ipv4_flags_check(const struct vxlan_config *cfg, + struct netlink_ext_ack *extack) +{ + if (!(cfg->flags & VXLAN_F_UDP_ZERO_CSUM_TX)) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: Zero UDP checksum must be allowed for TX"); + return false; + } + + if (cfg->flags & ~MLXSW_SP_NVE_VXLAN_IPV4_SUPPORTED_FLAGS) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: Unsupported flag"); + return false; + } + + return true; +} + +static bool mlxsw_sp_nve_vxlan_ipv6_flags_check(const struct vxlan_config *cfg, + struct netlink_ext_ack *extack) +{ + if (!(cfg->flags & VXLAN_F_UDP_ZERO_CSUM6_TX)) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: Zero UDP checksum must be allowed for TX"); + return false; + } + + if (!(cfg->flags & VXLAN_F_UDP_ZERO_CSUM6_RX)) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: Zero UDP checksum must be allowed for RX"); + return false; + } + + if (cfg->flags & ~MLXSW_SP_NVE_VXLAN_IPV6_SUPPORTED_FLAGS) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: Unsupported flag"); + return false; + } + + return true; +} + +static bool mlxsw_sp_nve_vxlan_can_offload(const struct mlxsw_sp_nve *nve, + const struct mlxsw_sp_nve_params *params, + struct netlink_ext_ack *extack) +{ + struct vxlan_dev *vxlan = netdev_priv(params->dev); + struct vxlan_config *cfg = &vxlan->cfg; + + if (vxlan_addr_multicast(&cfg->remote_ip)) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: Multicast destination IP is not supported"); + return false; + } + + if (vxlan_addr_any(&cfg->saddr)) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: Source address must be specified"); + return false; + } + + if (cfg->remote_ifindex) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: Local interface is not supported"); + return false; + } + + if (cfg->port_min || cfg->port_max) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: Only default UDP source port range is supported"); + return false; + } + + if (cfg->tos != 1) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: TOS must be configured to inherit"); + return false; + } + + if (cfg->flags & VXLAN_F_TTL_INHERIT) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: TTL must not be configured to inherit"); + return false; + } + + switch (cfg->saddr.sa.sa_family) { + case AF_INET: + if (!mlxsw_sp_nve_vxlan_ipv4_flags_check(cfg, extack)) + return false; + break; + case AF_INET6: + if (!mlxsw_sp_nve_vxlan_ipv6_flags_check(cfg, extack)) + return false; + break; + } + + if (cfg->ttl == 0) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: TTL must not be configured to 0"); + return false; + } + + if (cfg->label != 0) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: Flow label must be configured to 0"); + return false; + } + + return true; +} + +static bool mlxsw_sp1_nve_vxlan_can_offload(const struct mlxsw_sp_nve *nve, + const struct mlxsw_sp_nve_params *params, + struct netlink_ext_ack *extack) +{ + if (params->ethertype == ETH_P_8021AD) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: 802.1ad bridge is not supported with VxLAN"); + return false; + } + + return mlxsw_sp_nve_vxlan_can_offload(nve, params, extack); +} + +static void +mlxsw_sp_nve_vxlan_ul_proto_sip_config(const struct vxlan_config *cfg, + struct mlxsw_sp_nve_config *config) +{ + switch (cfg->saddr.sa.sa_family) { + case AF_INET: + config->ul_proto = MLXSW_SP_L3_PROTO_IPV4; + config->ul_sip.addr4 = cfg->saddr.sin.sin_addr.s_addr; + break; + case AF_INET6: + config->ul_proto = MLXSW_SP_L3_PROTO_IPV6; + config->ul_sip.addr6 = cfg->saddr.sin6.sin6_addr; + break; + } +} + +static void mlxsw_sp_nve_vxlan_config(const struct mlxsw_sp_nve *nve, + const struct mlxsw_sp_nve_params *params, + struct mlxsw_sp_nve_config *config) +{ + struct vxlan_dev *vxlan = netdev_priv(params->dev); + struct vxlan_config *cfg = &vxlan->cfg; + + config->type = MLXSW_SP_NVE_TYPE_VXLAN; + config->ttl = cfg->ttl; + config->flowlabel = cfg->label; + config->learning_en = cfg->flags & VXLAN_F_LEARN ? 1 : 0; + config->ul_tb_id = RT_TABLE_MAIN; + mlxsw_sp_nve_vxlan_ul_proto_sip_config(cfg, config); + config->udp_dport = cfg->dst_port; +} + +static void +mlxsw_sp_nve_vxlan_config_prepare(char *tngcr_pl, + const struct mlxsw_sp_nve_config *config) +{ + struct in6_addr addr6; + u8 udp_sport; + + mlxsw_reg_tngcr_pack(tngcr_pl, MLXSW_REG_TNGCR_TYPE_VXLAN, true, + config->ttl); + /* VxLAN driver's default UDP source port range is 32768 (0x8000) + * to 60999 (0xee47). Set the upper 8 bits of the UDP source port + * to a random number between 0x80 and 0xee + */ + get_random_bytes(&udp_sport, sizeof(udp_sport)); + udp_sport = (udp_sport % (0xee - 0x80 + 1)) + 0x80; + mlxsw_reg_tngcr_nve_udp_sport_prefix_set(tngcr_pl, udp_sport); + + switch (config->ul_proto) { + case MLXSW_SP_L3_PROTO_IPV4: + mlxsw_reg_tngcr_usipv4_set(tngcr_pl, + be32_to_cpu(config->ul_sip.addr4)); + break; + case MLXSW_SP_L3_PROTO_IPV6: + addr6 = config->ul_sip.addr6; + mlxsw_reg_tngcr_usipv6_memcpy_to(tngcr_pl, + (const char *)&addr6); + break; + } +} + +static int +mlxsw_sp1_nve_vxlan_config_set(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_nve_config *config) +{ + char tngcr_pl[MLXSW_REG_TNGCR_LEN]; + u16 ul_vr_id; + int err; + + err = mlxsw_sp_router_tb_id_vr_id(mlxsw_sp, config->ul_tb_id, + &ul_vr_id); + if (err) + return err; + + mlxsw_sp_nve_vxlan_config_prepare(tngcr_pl, config); + mlxsw_reg_tngcr_learn_enable_set(tngcr_pl, config->learning_en); + mlxsw_reg_tngcr_underlay_virtual_router_set(tngcr_pl, ul_vr_id); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tngcr), tngcr_pl); +} + +static void mlxsw_sp1_nve_vxlan_config_clear(struct mlxsw_sp *mlxsw_sp) +{ + char tngcr_pl[MLXSW_REG_TNGCR_LEN]; + + mlxsw_reg_tngcr_pack(tngcr_pl, MLXSW_REG_TNGCR_TYPE_VXLAN, false, 0); + + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tngcr), tngcr_pl); +} + +static int mlxsw_sp1_nve_vxlan_rtdp_set(struct mlxsw_sp *mlxsw_sp, + unsigned int tunnel_index) +{ + char rtdp_pl[MLXSW_REG_RTDP_LEN]; + + mlxsw_reg_rtdp_pack(rtdp_pl, MLXSW_REG_RTDP_TYPE_NVE, tunnel_index); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rtdp), rtdp_pl); +} + +static int mlxsw_sp1_nve_vxlan_init(struct mlxsw_sp_nve *nve, + const struct mlxsw_sp_nve_config *config) +{ + struct mlxsw_sp *mlxsw_sp = nve->mlxsw_sp; + int err; + + err = mlxsw_sp_parsing_vxlan_udp_dport_set(mlxsw_sp, config->udp_dport); + if (err) + return err; + + err = mlxsw_sp_parsing_depth_inc(mlxsw_sp); + if (err) + goto err_parsing_depth_inc; + + err = mlxsw_sp1_nve_vxlan_config_set(mlxsw_sp, config); + if (err) + goto err_config_set; + + err = mlxsw_sp1_nve_vxlan_rtdp_set(mlxsw_sp, nve->tunnel_index); + if (err) + goto err_rtdp_set; + + err = mlxsw_sp_router_nve_promote_decap(mlxsw_sp, config->ul_tb_id, + config->ul_proto, + &config->ul_sip, + nve->tunnel_index); + if (err) + goto err_promote_decap; + + return 0; + +err_promote_decap: +err_rtdp_set: + mlxsw_sp1_nve_vxlan_config_clear(mlxsw_sp); +err_config_set: + mlxsw_sp_parsing_depth_dec(mlxsw_sp); +err_parsing_depth_inc: + mlxsw_sp_parsing_vxlan_udp_dport_set(mlxsw_sp, 0); + return err; +} + +static void mlxsw_sp1_nve_vxlan_fini(struct mlxsw_sp_nve *nve) +{ + struct mlxsw_sp_nve_config *config = &nve->config; + struct mlxsw_sp *mlxsw_sp = nve->mlxsw_sp; + + mlxsw_sp_router_nve_demote_decap(mlxsw_sp, config->ul_tb_id, + config->ul_proto, &config->ul_sip); + mlxsw_sp1_nve_vxlan_config_clear(mlxsw_sp); + mlxsw_sp_parsing_depth_dec(mlxsw_sp); + mlxsw_sp_parsing_vxlan_udp_dport_set(mlxsw_sp, 0); +} + +static int +mlxsw_sp_nve_vxlan_fdb_replay(const struct net_device *nve_dev, __be32 vni, + struct netlink_ext_ack *extack) +{ + if (WARN_ON(!netif_is_vxlan(nve_dev))) + return -EINVAL; + return vxlan_fdb_replay(nve_dev, vni, &mlxsw_sp_switchdev_notifier, + extack); +} + +static void +mlxsw_sp_nve_vxlan_clear_offload(const struct net_device *nve_dev, __be32 vni) +{ + if (WARN_ON(!netif_is_vxlan(nve_dev))) + return; + vxlan_fdb_clear_offload(nve_dev, vni); +} + +const struct mlxsw_sp_nve_ops mlxsw_sp1_nve_vxlan_ops = { + .type = MLXSW_SP_NVE_TYPE_VXLAN, + .can_offload = mlxsw_sp1_nve_vxlan_can_offload, + .nve_config = mlxsw_sp_nve_vxlan_config, + .init = mlxsw_sp1_nve_vxlan_init, + .fini = mlxsw_sp1_nve_vxlan_fini, + .fdb_replay = mlxsw_sp_nve_vxlan_fdb_replay, + .fdb_clear_offload = mlxsw_sp_nve_vxlan_clear_offload, +}; + +static int mlxsw_sp2_nve_vxlan_learning_set(struct mlxsw_sp *mlxsw_sp, + bool learning_en) +{ + char tnpc_pl[MLXSW_REG_TNPC_LEN]; + + mlxsw_reg_tnpc_pack(tnpc_pl, MLXSW_REG_TUNNEL_PORT_NVE, + learning_en); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tnpc), tnpc_pl); +} + +static int +mlxsw_sp2_nve_decap_ethertype_set(struct mlxsw_sp *mlxsw_sp) +{ + char spvid_pl[MLXSW_REG_SPVID_LEN] = {}; + + mlxsw_reg_spvid_tport_set(spvid_pl, true); + mlxsw_reg_spvid_local_port_set(spvid_pl, + MLXSW_REG_TUNNEL_PORT_NVE); + mlxsw_reg_spvid_egr_et_set_set(spvid_pl, true); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(spvid), spvid_pl); +} + +static int +mlxsw_sp2_nve_vxlan_config_set(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_nve_config *config) +{ + char tngcr_pl[MLXSW_REG_TNGCR_LEN]; + char spvtr_pl[MLXSW_REG_SPVTR_LEN]; + u16 ul_rif_index; + int err; + + err = mlxsw_sp_router_ul_rif_get(mlxsw_sp, config->ul_tb_id, + &ul_rif_index); + if (err) + return err; + mlxsw_sp->nve->ul_rif_index = ul_rif_index; + + err = mlxsw_sp2_nve_vxlan_learning_set(mlxsw_sp, config->learning_en); + if (err) + goto err_vxlan_learning_set; + + mlxsw_sp_nve_vxlan_config_prepare(tngcr_pl, config); + mlxsw_reg_tngcr_underlay_rif_set(tngcr_pl, ul_rif_index); + + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tngcr), tngcr_pl); + if (err) + goto err_tngcr_write; + + mlxsw_reg_spvtr_pack(spvtr_pl, true, MLXSW_REG_TUNNEL_PORT_NVE, + MLXSW_REG_SPVTR_IPVID_MODE_ALWAYS_PUSH_VLAN); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(spvtr), spvtr_pl); + if (err) + goto err_spvtr_write; + + err = mlxsw_sp2_nve_decap_ethertype_set(mlxsw_sp); + if (err) + goto err_decap_ethertype_set; + + return 0; + +err_decap_ethertype_set: + mlxsw_reg_spvtr_pack(spvtr_pl, true, MLXSW_REG_TUNNEL_PORT_NVE, + MLXSW_REG_SPVTR_IPVID_MODE_IEEE_COMPLIANT_PVID); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(spvtr), spvtr_pl); +err_spvtr_write: + mlxsw_reg_tngcr_pack(tngcr_pl, MLXSW_REG_TNGCR_TYPE_VXLAN, false, 0); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tngcr), tngcr_pl); +err_tngcr_write: + mlxsw_sp2_nve_vxlan_learning_set(mlxsw_sp, false); +err_vxlan_learning_set: + mlxsw_sp_router_ul_rif_put(mlxsw_sp, ul_rif_index); + return err; +} + +static void mlxsw_sp2_nve_vxlan_config_clear(struct mlxsw_sp *mlxsw_sp) +{ + char spvtr_pl[MLXSW_REG_SPVTR_LEN]; + char tngcr_pl[MLXSW_REG_TNGCR_LEN]; + + mlxsw_reg_spvtr_pack(spvtr_pl, true, MLXSW_REG_TUNNEL_PORT_NVE, + MLXSW_REG_SPVTR_IPVID_MODE_IEEE_COMPLIANT_PVID); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(spvtr), spvtr_pl); + mlxsw_reg_tngcr_pack(tngcr_pl, MLXSW_REG_TNGCR_TYPE_VXLAN, false, 0); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tngcr), tngcr_pl); + mlxsw_sp2_nve_vxlan_learning_set(mlxsw_sp, false); + mlxsw_sp_router_ul_rif_put(mlxsw_sp, mlxsw_sp->nve->ul_rif_index); +} + +static int mlxsw_sp2_nve_vxlan_rtdp_set(struct mlxsw_sp *mlxsw_sp, + unsigned int tunnel_index, + u16 ul_rif_index) +{ + char rtdp_pl[MLXSW_REG_RTDP_LEN]; + + mlxsw_reg_rtdp_pack(rtdp_pl, MLXSW_REG_RTDP_TYPE_NVE, tunnel_index); + mlxsw_reg_rtdp_egress_router_interface_set(rtdp_pl, ul_rif_index); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rtdp), rtdp_pl); +} + +static int mlxsw_sp2_nve_vxlan_init(struct mlxsw_sp_nve *nve, + const struct mlxsw_sp_nve_config *config) +{ + struct mlxsw_sp *mlxsw_sp = nve->mlxsw_sp; + int err; + + err = mlxsw_sp_parsing_vxlan_udp_dport_set(mlxsw_sp, config->udp_dport); + if (err) + return err; + + err = mlxsw_sp_parsing_depth_inc(mlxsw_sp); + if (err) + goto err_parsing_depth_inc; + + err = mlxsw_sp2_nve_vxlan_config_set(mlxsw_sp, config); + if (err) + goto err_config_set; + + err = mlxsw_sp2_nve_vxlan_rtdp_set(mlxsw_sp, nve->tunnel_index, + nve->ul_rif_index); + if (err) + goto err_rtdp_set; + + err = mlxsw_sp_router_nve_promote_decap(mlxsw_sp, config->ul_tb_id, + config->ul_proto, + &config->ul_sip, + nve->tunnel_index); + if (err) + goto err_promote_decap; + + return 0; + +err_promote_decap: +err_rtdp_set: + mlxsw_sp2_nve_vxlan_config_clear(mlxsw_sp); +err_config_set: + mlxsw_sp_parsing_depth_dec(mlxsw_sp); +err_parsing_depth_inc: + mlxsw_sp_parsing_vxlan_udp_dport_set(mlxsw_sp, 0); + return err; +} + +static void mlxsw_sp2_nve_vxlan_fini(struct mlxsw_sp_nve *nve) +{ + struct mlxsw_sp_nve_config *config = &nve->config; + struct mlxsw_sp *mlxsw_sp = nve->mlxsw_sp; + + mlxsw_sp_router_nve_demote_decap(mlxsw_sp, config->ul_tb_id, + config->ul_proto, &config->ul_sip); + mlxsw_sp2_nve_vxlan_config_clear(mlxsw_sp); + mlxsw_sp_parsing_depth_dec(mlxsw_sp); + mlxsw_sp_parsing_vxlan_udp_dport_set(mlxsw_sp, 0); +} + +const struct mlxsw_sp_nve_ops mlxsw_sp2_nve_vxlan_ops = { + .type = MLXSW_SP_NVE_TYPE_VXLAN, + .can_offload = mlxsw_sp_nve_vxlan_can_offload, + .nve_config = mlxsw_sp_nve_vxlan_config, + .init = mlxsw_sp2_nve_vxlan_init, + .fini = mlxsw_sp2_nve_vxlan_fini, + .fdb_replay = mlxsw_sp_nve_vxlan_fdb_replay, + .fdb_clear_offload = mlxsw_sp_nve_vxlan_clear_offload, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_pgt.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_pgt.c new file mode 100644 index 000000000..7dd3dba0f --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_pgt.c @@ -0,0 +1,346 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2022, NVIDIA CORPORATION & AFFILIATES. All rights reserved. */ + +#include <linux/refcount.h> +#include <linux/idr.h> + +#include "spectrum.h" +#include "reg.h" + +struct mlxsw_sp_pgt { + struct idr pgt_idr; + u16 end_index; /* Exclusive. */ + struct mutex lock; /* Protects PGT. */ + bool smpe_index_valid; +}; + +struct mlxsw_sp_pgt_entry { + struct list_head ports_list; + u16 index; + u16 smpe_index; +}; + +struct mlxsw_sp_pgt_entry_port { + struct list_head list; /* Member of 'ports_list'. */ + u16 local_port; +}; + +int mlxsw_sp_pgt_mid_alloc(struct mlxsw_sp *mlxsw_sp, u16 *p_mid) +{ + int index, err = 0; + + mutex_lock(&mlxsw_sp->pgt->lock); + index = idr_alloc(&mlxsw_sp->pgt->pgt_idr, NULL, 0, + mlxsw_sp->pgt->end_index, GFP_KERNEL); + + if (index < 0) { + err = index; + goto err_idr_alloc; + } + + *p_mid = index; + mutex_unlock(&mlxsw_sp->pgt->lock); + return 0; + +err_idr_alloc: + mutex_unlock(&mlxsw_sp->pgt->lock); + return err; +} + +void mlxsw_sp_pgt_mid_free(struct mlxsw_sp *mlxsw_sp, u16 mid_base) +{ + mutex_lock(&mlxsw_sp->pgt->lock); + WARN_ON(idr_remove(&mlxsw_sp->pgt->pgt_idr, mid_base)); + mutex_unlock(&mlxsw_sp->pgt->lock); +} + +int +mlxsw_sp_pgt_mid_alloc_range(struct mlxsw_sp *mlxsw_sp, u16 mid_base, u16 count) +{ + unsigned int idr_cursor; + int i, err; + + mutex_lock(&mlxsw_sp->pgt->lock); + + /* This function is supposed to be called several times as part of + * driver init, in specific order. Verify that the mid_index is the + * first free index in the idr, to be able to free the indexes in case + * of error. + */ + idr_cursor = idr_get_cursor(&mlxsw_sp->pgt->pgt_idr); + if (WARN_ON(idr_cursor != mid_base)) { + err = -EINVAL; + goto err_idr_cursor; + } + + for (i = 0; i < count; i++) { + err = idr_alloc_cyclic(&mlxsw_sp->pgt->pgt_idr, NULL, + mid_base, mid_base + count, GFP_KERNEL); + if (err < 0) + goto err_idr_alloc_cyclic; + } + + mutex_unlock(&mlxsw_sp->pgt->lock); + return 0; + +err_idr_alloc_cyclic: + for (i--; i >= 0; i--) + idr_remove(&mlxsw_sp->pgt->pgt_idr, mid_base + i); +err_idr_cursor: + mutex_unlock(&mlxsw_sp->pgt->lock); + return err; +} + +void +mlxsw_sp_pgt_mid_free_range(struct mlxsw_sp *mlxsw_sp, u16 mid_base, u16 count) +{ + struct idr *pgt_idr = &mlxsw_sp->pgt->pgt_idr; + int i; + + mutex_lock(&mlxsw_sp->pgt->lock); + + for (i = 0; i < count; i++) + WARN_ON_ONCE(idr_remove(pgt_idr, mid_base + i)); + + mutex_unlock(&mlxsw_sp->pgt->lock); +} + +static struct mlxsw_sp_pgt_entry_port * +mlxsw_sp_pgt_entry_port_lookup(struct mlxsw_sp_pgt_entry *pgt_entry, + u16 local_port) +{ + struct mlxsw_sp_pgt_entry_port *pgt_entry_port; + + list_for_each_entry(pgt_entry_port, &pgt_entry->ports_list, list) { + if (pgt_entry_port->local_port == local_port) + return pgt_entry_port; + } + + return NULL; +} + +static struct mlxsw_sp_pgt_entry * +mlxsw_sp_pgt_entry_create(struct mlxsw_sp_pgt *pgt, u16 mid, u16 smpe) +{ + struct mlxsw_sp_pgt_entry *pgt_entry; + void *ret; + int err; + + pgt_entry = kzalloc(sizeof(*pgt_entry), GFP_KERNEL); + if (!pgt_entry) + return ERR_PTR(-ENOMEM); + + ret = idr_replace(&pgt->pgt_idr, pgt_entry, mid); + if (IS_ERR(ret)) { + err = PTR_ERR(ret); + goto err_idr_replace; + } + + INIT_LIST_HEAD(&pgt_entry->ports_list); + pgt_entry->index = mid; + pgt_entry->smpe_index = smpe; + return pgt_entry; + +err_idr_replace: + kfree(pgt_entry); + return ERR_PTR(err); +} + +static void mlxsw_sp_pgt_entry_destroy(struct mlxsw_sp_pgt *pgt, + struct mlxsw_sp_pgt_entry *pgt_entry) +{ + WARN_ON(!list_empty(&pgt_entry->ports_list)); + + pgt_entry = idr_replace(&pgt->pgt_idr, NULL, pgt_entry->index); + if (WARN_ON(IS_ERR(pgt_entry))) + return; + + kfree(pgt_entry); +} + +static struct mlxsw_sp_pgt_entry * +mlxsw_sp_pgt_entry_get(struct mlxsw_sp_pgt *pgt, u16 mid, u16 smpe) +{ + struct mlxsw_sp_pgt_entry *pgt_entry; + + pgt_entry = idr_find(&pgt->pgt_idr, mid); + if (pgt_entry) + return pgt_entry; + + return mlxsw_sp_pgt_entry_create(pgt, mid, smpe); +} + +static void mlxsw_sp_pgt_entry_put(struct mlxsw_sp_pgt *pgt, u16 mid) +{ + struct mlxsw_sp_pgt_entry *pgt_entry; + + pgt_entry = idr_find(&pgt->pgt_idr, mid); + if (WARN_ON(!pgt_entry)) + return; + + if (list_empty(&pgt_entry->ports_list)) + mlxsw_sp_pgt_entry_destroy(pgt, pgt_entry); +} + +static void mlxsw_sp_pgt_smid2_port_set(char *smid2_pl, u16 local_port, + bool member) +{ + mlxsw_reg_smid2_port_set(smid2_pl, local_port, member); + mlxsw_reg_smid2_port_mask_set(smid2_pl, local_port, 1); +} + +static int +mlxsw_sp_pgt_entry_port_write(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_pgt_entry *pgt_entry, + u16 local_port, bool member) +{ + char *smid2_pl; + int err; + + smid2_pl = kmalloc(MLXSW_REG_SMID2_LEN, GFP_KERNEL); + if (!smid2_pl) + return -ENOMEM; + + mlxsw_reg_smid2_pack(smid2_pl, pgt_entry->index, 0, 0, + mlxsw_sp->pgt->smpe_index_valid, + pgt_entry->smpe_index); + + mlxsw_sp_pgt_smid2_port_set(smid2_pl, local_port, member); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(smid2), smid2_pl); + + kfree(smid2_pl); + + return err; +} + +static struct mlxsw_sp_pgt_entry_port * +mlxsw_sp_pgt_entry_port_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_pgt_entry *pgt_entry, + u16 local_port) +{ + struct mlxsw_sp_pgt_entry_port *pgt_entry_port; + int err; + + pgt_entry_port = kzalloc(sizeof(*pgt_entry_port), GFP_KERNEL); + if (!pgt_entry_port) + return ERR_PTR(-ENOMEM); + + err = mlxsw_sp_pgt_entry_port_write(mlxsw_sp, pgt_entry, local_port, + true); + if (err) + goto err_pgt_entry_port_write; + + pgt_entry_port->local_port = local_port; + list_add(&pgt_entry_port->list, &pgt_entry->ports_list); + + return pgt_entry_port; + +err_pgt_entry_port_write: + kfree(pgt_entry_port); + return ERR_PTR(err); +} + +static void +mlxsw_sp_pgt_entry_port_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_pgt_entry *pgt_entry, + struct mlxsw_sp_pgt_entry_port *pgt_entry_port) + +{ + list_del(&pgt_entry_port->list); + mlxsw_sp_pgt_entry_port_write(mlxsw_sp, pgt_entry, + pgt_entry_port->local_port, false); + kfree(pgt_entry_port); +} + +static int mlxsw_sp_pgt_entry_port_add(struct mlxsw_sp *mlxsw_sp, u16 mid, + u16 smpe, u16 local_port) +{ + struct mlxsw_sp_pgt_entry_port *pgt_entry_port; + struct mlxsw_sp_pgt_entry *pgt_entry; + int err; + + mutex_lock(&mlxsw_sp->pgt->lock); + + pgt_entry = mlxsw_sp_pgt_entry_get(mlxsw_sp->pgt, mid, smpe); + if (IS_ERR(pgt_entry)) { + err = PTR_ERR(pgt_entry); + goto err_pgt_entry_get; + } + + pgt_entry_port = mlxsw_sp_pgt_entry_port_create(mlxsw_sp, pgt_entry, + local_port); + if (IS_ERR(pgt_entry_port)) { + err = PTR_ERR(pgt_entry_port); + goto err_pgt_entry_port_get; + } + + mutex_unlock(&mlxsw_sp->pgt->lock); + return 0; + +err_pgt_entry_port_get: + mlxsw_sp_pgt_entry_put(mlxsw_sp->pgt, mid); +err_pgt_entry_get: + mutex_unlock(&mlxsw_sp->pgt->lock); + return err; +} + +static void mlxsw_sp_pgt_entry_port_del(struct mlxsw_sp *mlxsw_sp, + u16 mid, u16 smpe, u16 local_port) +{ + struct mlxsw_sp_pgt_entry_port *pgt_entry_port; + struct mlxsw_sp_pgt_entry *pgt_entry; + + mutex_lock(&mlxsw_sp->pgt->lock); + + pgt_entry = idr_find(&mlxsw_sp->pgt->pgt_idr, mid); + if (!pgt_entry) + goto out; + + pgt_entry_port = mlxsw_sp_pgt_entry_port_lookup(pgt_entry, local_port); + if (!pgt_entry_port) + goto out; + + mlxsw_sp_pgt_entry_port_destroy(mlxsw_sp, pgt_entry, pgt_entry_port); + mlxsw_sp_pgt_entry_put(mlxsw_sp->pgt, mid); + +out: + mutex_unlock(&mlxsw_sp->pgt->lock); +} + +int mlxsw_sp_pgt_entry_port_set(struct mlxsw_sp *mlxsw_sp, u16 mid, + u16 smpe, u16 local_port, bool member) +{ + if (member) + return mlxsw_sp_pgt_entry_port_add(mlxsw_sp, mid, smpe, + local_port); + + mlxsw_sp_pgt_entry_port_del(mlxsw_sp, mid, smpe, local_port); + return 0; +} + +int mlxsw_sp_pgt_init(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_pgt *pgt; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, PGT_SIZE)) + return -EIO; + + pgt = kzalloc(sizeof(*mlxsw_sp->pgt), GFP_KERNEL); + if (!pgt) + return -ENOMEM; + + idr_init(&pgt->pgt_idr); + pgt->end_index = MLXSW_CORE_RES_GET(mlxsw_sp->core, PGT_SIZE); + mutex_init(&pgt->lock); + pgt->smpe_index_valid = mlxsw_sp->pgt_smpe_index_valid; + mlxsw_sp->pgt = pgt; + return 0; +} + +void mlxsw_sp_pgt_fini(struct mlxsw_sp *mlxsw_sp) +{ + mutex_destroy(&mlxsw_sp->pgt->lock); + WARN_ON(!idr_is_empty(&mlxsw_sp->pgt->pgt_idr)); + idr_destroy(&mlxsw_sp->pgt->pgt_idr); + kfree(mlxsw_sp->pgt); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_policer.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_policer.c new file mode 100644 index 000000000..22ebb207c --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_policer.c @@ -0,0 +1,468 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2020 Mellanox Technologies. All rights reserved */ + +#include <linux/idr.h> +#include <linux/log2.h> +#include <linux/mutex.h> +#include <linux/netlink.h> +#include <net/devlink.h> + +#include "spectrum.h" + +struct mlxsw_sp_policer_family { + enum mlxsw_sp_policer_type type; + enum mlxsw_reg_qpcr_g qpcr_type; + struct mlxsw_sp *mlxsw_sp; + u16 start_index; /* Inclusive */ + u16 end_index; /* Exclusive */ + struct idr policer_idr; + struct mutex lock; /* Protects policer_idr */ + atomic_t policers_count; + const struct mlxsw_sp_policer_family_ops *ops; +}; + +struct mlxsw_sp_policer { + struct mlxsw_sp_policer_params params; + u16 index; +}; + +struct mlxsw_sp_policer_family_ops { + int (*init)(struct mlxsw_sp_policer_family *family); + void (*fini)(struct mlxsw_sp_policer_family *family); + int (*policer_index_alloc)(struct mlxsw_sp_policer_family *family, + struct mlxsw_sp_policer *policer); + struct mlxsw_sp_policer * (*policer_index_free)(struct mlxsw_sp_policer_family *family, + u16 policer_index); + int (*policer_init)(struct mlxsw_sp_policer_family *family, + const struct mlxsw_sp_policer *policer); + int (*policer_params_check)(const struct mlxsw_sp_policer_family *family, + const struct mlxsw_sp_policer_params *params, + struct netlink_ext_ack *extack); +}; + +struct mlxsw_sp_policer_core { + struct mlxsw_sp_policer_family *family_arr[MLXSW_SP_POLICER_TYPE_MAX + 1]; + const struct mlxsw_sp_policer_core_ops *ops; + u8 lowest_bs_bits; + u8 highest_bs_bits; +}; + +struct mlxsw_sp_policer_core_ops { + int (*init)(struct mlxsw_sp_policer_core *policer_core); +}; + +static u64 mlxsw_sp_policer_rate_bytes_ps_kbps(u64 rate_bytes_ps) +{ + return div_u64(rate_bytes_ps, 1000) * BITS_PER_BYTE; +} + +static u8 mlxsw_sp_policer_burst_bytes_hw_units(u64 burst_bytes) +{ + /* Provided burst size is in bytes. The ASIC burst size value is + * (2 ^ bs) * 512 bits. Convert the provided size to 512-bit units. + */ + u64 bs512 = div_u64(burst_bytes, 64); + + if (!bs512) + return 0; + + return fls64(bs512) - 1; +} + +static u64 mlxsw_sp_policer_single_rate_occ_get(void *priv) +{ + struct mlxsw_sp_policer_family *family = priv; + + return atomic_read(&family->policers_count); +} + +static int +mlxsw_sp_policer_single_rate_family_init(struct mlxsw_sp_policer_family *family) +{ + struct mlxsw_core *core = family->mlxsw_sp->core; + struct devlink *devlink; + + /* CPU policers are allocated from the first N policers in the global + * range, so skip them. + */ + if (!MLXSW_CORE_RES_VALID(core, MAX_GLOBAL_POLICERS) || + !MLXSW_CORE_RES_VALID(core, MAX_CPU_POLICERS)) + return -EIO; + + family->start_index = MLXSW_CORE_RES_GET(core, MAX_CPU_POLICERS); + family->end_index = MLXSW_CORE_RES_GET(core, MAX_GLOBAL_POLICERS); + + atomic_set(&family->policers_count, 0); + devlink = priv_to_devlink(core); + devl_resource_occ_get_register(devlink, + MLXSW_SP_RESOURCE_SINGLE_RATE_POLICERS, + mlxsw_sp_policer_single_rate_occ_get, + family); + + return 0; +} + +static void +mlxsw_sp_policer_single_rate_family_fini(struct mlxsw_sp_policer_family *family) +{ + struct devlink *devlink = priv_to_devlink(family->mlxsw_sp->core); + + devl_resource_occ_get_unregister(devlink, + MLXSW_SP_RESOURCE_SINGLE_RATE_POLICERS); + WARN_ON(atomic_read(&family->policers_count) != 0); +} + +static int +mlxsw_sp_policer_single_rate_index_alloc(struct mlxsw_sp_policer_family *family, + struct mlxsw_sp_policer *policer) +{ + int id; + + mutex_lock(&family->lock); + id = idr_alloc(&family->policer_idr, policer, family->start_index, + family->end_index, GFP_KERNEL); + mutex_unlock(&family->lock); + + if (id < 0) + return id; + + atomic_inc(&family->policers_count); + policer->index = id; + + return 0; +} + +static struct mlxsw_sp_policer * +mlxsw_sp_policer_single_rate_index_free(struct mlxsw_sp_policer_family *family, + u16 policer_index) +{ + struct mlxsw_sp_policer *policer; + + atomic_dec(&family->policers_count); + + mutex_lock(&family->lock); + policer = idr_remove(&family->policer_idr, policer_index); + mutex_unlock(&family->lock); + + WARN_ON(!policer); + + return policer; +} + +static int +mlxsw_sp_policer_single_rate_init(struct mlxsw_sp_policer_family *family, + const struct mlxsw_sp_policer *policer) +{ + u64 rate_kbps = mlxsw_sp_policer_rate_bytes_ps_kbps(policer->params.rate); + u8 bs = mlxsw_sp_policer_burst_bytes_hw_units(policer->params.burst); + struct mlxsw_sp *mlxsw_sp = family->mlxsw_sp; + char qpcr_pl[MLXSW_REG_QPCR_LEN]; + + mlxsw_reg_qpcr_pack(qpcr_pl, policer->index, MLXSW_REG_QPCR_IR_UNITS_K, + true, rate_kbps, bs); + mlxsw_reg_qpcr_clear_counter_set(qpcr_pl, true); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qpcr), qpcr_pl); +} + +static int +mlxsw_sp_policer_single_rate_params_check(const struct mlxsw_sp_policer_family *family, + const struct mlxsw_sp_policer_params *params, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_policer_core *policer_core = family->mlxsw_sp->policer_core; + u64 rate_bps = params->rate * BITS_PER_BYTE; + u8 bs; + + if (!params->bytes) { + NL_SET_ERR_MSG_MOD(extack, "Only bandwidth policing is currently supported by single rate policers"); + return -EINVAL; + } + + if (!is_power_of_2(params->burst)) { + NL_SET_ERR_MSG_MOD(extack, "Policer burst size is not power of two"); + return -EINVAL; + } + + bs = mlxsw_sp_policer_burst_bytes_hw_units(params->burst); + + if (bs < policer_core->lowest_bs_bits) { + NL_SET_ERR_MSG_MOD(extack, "Policer burst size lower than limit"); + return -EINVAL; + } + + if (bs > policer_core->highest_bs_bits) { + NL_SET_ERR_MSG_MOD(extack, "Policer burst size higher than limit"); + return -EINVAL; + } + + if (rate_bps < MLXSW_REG_QPCR_LOWEST_CIR_BITS) { + NL_SET_ERR_MSG_MOD(extack, "Policer rate lower than limit"); + return -EINVAL; + } + + if (rate_bps > MLXSW_REG_QPCR_HIGHEST_CIR_BITS) { + NL_SET_ERR_MSG_MOD(extack, "Policer rate higher than limit"); + return -EINVAL; + } + + return 0; +} + +static const struct mlxsw_sp_policer_family_ops mlxsw_sp_policer_single_rate_ops = { + .init = mlxsw_sp_policer_single_rate_family_init, + .fini = mlxsw_sp_policer_single_rate_family_fini, + .policer_index_alloc = mlxsw_sp_policer_single_rate_index_alloc, + .policer_index_free = mlxsw_sp_policer_single_rate_index_free, + .policer_init = mlxsw_sp_policer_single_rate_init, + .policer_params_check = mlxsw_sp_policer_single_rate_params_check, +}; + +static const struct mlxsw_sp_policer_family mlxsw_sp_policer_single_rate_family = { + .type = MLXSW_SP_POLICER_TYPE_SINGLE_RATE, + .qpcr_type = MLXSW_REG_QPCR_G_GLOBAL, + .ops = &mlxsw_sp_policer_single_rate_ops, +}; + +static const struct mlxsw_sp_policer_family *mlxsw_sp_policer_family_arr[] = { + [MLXSW_SP_POLICER_TYPE_SINGLE_RATE] = &mlxsw_sp_policer_single_rate_family, +}; + +int mlxsw_sp_policer_add(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_policer_type type, + const struct mlxsw_sp_policer_params *params, + struct netlink_ext_ack *extack, u16 *p_policer_index) +{ + struct mlxsw_sp_policer_family *family; + struct mlxsw_sp_policer *policer; + int err; + + family = mlxsw_sp->policer_core->family_arr[type]; + + err = family->ops->policer_params_check(family, params, extack); + if (err) + return err; + + policer = kmalloc(sizeof(*policer), GFP_KERNEL); + if (!policer) + return -ENOMEM; + policer->params = *params; + + err = family->ops->policer_index_alloc(family, policer); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to allocate policer index"); + goto err_policer_index_alloc; + } + + err = family->ops->policer_init(family, policer); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to initialize policer"); + goto err_policer_init; + } + + *p_policer_index = policer->index; + + return 0; + +err_policer_init: + family->ops->policer_index_free(family, policer->index); +err_policer_index_alloc: + kfree(policer); + return err; +} + +void mlxsw_sp_policer_del(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_policer_type type, u16 policer_index) +{ + struct mlxsw_sp_policer_family *family; + struct mlxsw_sp_policer *policer; + + family = mlxsw_sp->policer_core->family_arr[type]; + policer = family->ops->policer_index_free(family, policer_index); + kfree(policer); +} + +int mlxsw_sp_policer_drops_counter_get(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_policer_type type, + u16 policer_index, u64 *p_drops) +{ + struct mlxsw_sp_policer_family *family; + char qpcr_pl[MLXSW_REG_QPCR_LEN]; + int err; + + family = mlxsw_sp->policer_core->family_arr[type]; + + MLXSW_REG_ZERO(qpcr, qpcr_pl); + mlxsw_reg_qpcr_pid_set(qpcr_pl, policer_index); + mlxsw_reg_qpcr_g_set(qpcr_pl, family->qpcr_type); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(qpcr), qpcr_pl); + if (err) + return err; + + *p_drops = mlxsw_reg_qpcr_violate_count_get(qpcr_pl); + + return 0; +} + +static int +mlxsw_sp_policer_family_register(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_policer_family *tmpl) +{ + struct mlxsw_sp_policer_family *family; + int err; + + family = kmemdup(tmpl, sizeof(*family), GFP_KERNEL); + if (!family) + return -ENOMEM; + + family->mlxsw_sp = mlxsw_sp; + idr_init(&family->policer_idr); + mutex_init(&family->lock); + + err = family->ops->init(family); + if (err) + goto err_family_init; + + if (WARN_ON(family->start_index >= family->end_index)) { + err = -EINVAL; + goto err_index_check; + } + + mlxsw_sp->policer_core->family_arr[tmpl->type] = family; + + return 0; + +err_index_check: + family->ops->fini(family); +err_family_init: + mutex_destroy(&family->lock); + idr_destroy(&family->policer_idr); + kfree(family); + return err; +} + +static void +mlxsw_sp_policer_family_unregister(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_policer_family *family) +{ + family->ops->fini(family); + mutex_destroy(&family->lock); + WARN_ON(!idr_is_empty(&family->policer_idr)); + idr_destroy(&family->policer_idr); + kfree(family); +} + +int mlxsw_sp_policers_init(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_policer_core *policer_core; + int i, err; + + policer_core = kzalloc(sizeof(*policer_core), GFP_KERNEL); + if (!policer_core) + return -ENOMEM; + mlxsw_sp->policer_core = policer_core; + policer_core->ops = mlxsw_sp->policer_core_ops; + + err = policer_core->ops->init(policer_core); + if (err) + goto err_init; + + for (i = 0; i < MLXSW_SP_POLICER_TYPE_MAX + 1; i++) { + err = mlxsw_sp_policer_family_register(mlxsw_sp, mlxsw_sp_policer_family_arr[i]); + if (err) + goto err_family_register; + } + + return 0; + +err_family_register: + for (i--; i >= 0; i--) { + struct mlxsw_sp_policer_family *family; + + family = mlxsw_sp->policer_core->family_arr[i]; + mlxsw_sp_policer_family_unregister(mlxsw_sp, family); + } +err_init: + kfree(mlxsw_sp->policer_core); + return err; +} + +void mlxsw_sp_policers_fini(struct mlxsw_sp *mlxsw_sp) +{ + int i; + + for (i = MLXSW_SP_POLICER_TYPE_MAX; i >= 0; i--) { + struct mlxsw_sp_policer_family *family; + + family = mlxsw_sp->policer_core->family_arr[i]; + mlxsw_sp_policer_family_unregister(mlxsw_sp, family); + } + + kfree(mlxsw_sp->policer_core); +} + +int mlxsw_sp_policer_resources_register(struct mlxsw_core *mlxsw_core) +{ + u64 global_policers, cpu_policers, single_rate_policers; + struct devlink *devlink = priv_to_devlink(mlxsw_core); + struct devlink_resource_size_params size_params; + int err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_core, MAX_GLOBAL_POLICERS) || + !MLXSW_CORE_RES_VALID(mlxsw_core, MAX_CPU_POLICERS)) + return -EIO; + + global_policers = MLXSW_CORE_RES_GET(mlxsw_core, MAX_GLOBAL_POLICERS); + cpu_policers = MLXSW_CORE_RES_GET(mlxsw_core, MAX_CPU_POLICERS); + single_rate_policers = global_policers - cpu_policers; + + devlink_resource_size_params_init(&size_params, global_policers, + global_policers, 1, + DEVLINK_RESOURCE_UNIT_ENTRY); + err = devl_resource_register(devlink, "global_policers", + global_policers, + MLXSW_SP_RESOURCE_GLOBAL_POLICERS, + DEVLINK_RESOURCE_ID_PARENT_TOP, + &size_params); + if (err) + return err; + + devlink_resource_size_params_init(&size_params, single_rate_policers, + single_rate_policers, 1, + DEVLINK_RESOURCE_UNIT_ENTRY); + err = devl_resource_register(devlink, "single_rate_policers", + single_rate_policers, + MLXSW_SP_RESOURCE_SINGLE_RATE_POLICERS, + MLXSW_SP_RESOURCE_GLOBAL_POLICERS, + &size_params); + if (err) + return err; + + return 0; +} + +static int +mlxsw_sp1_policer_core_init(struct mlxsw_sp_policer_core *policer_core) +{ + policer_core->lowest_bs_bits = MLXSW_REG_QPCR_LOWEST_CBS_BITS_SP1; + policer_core->highest_bs_bits = MLXSW_REG_QPCR_HIGHEST_CBS_BITS_SP1; + + return 0; +} + +const struct mlxsw_sp_policer_core_ops mlxsw_sp1_policer_core_ops = { + .init = mlxsw_sp1_policer_core_init, +}; + +static int +mlxsw_sp2_policer_core_init(struct mlxsw_sp_policer_core *policer_core) +{ + policer_core->lowest_bs_bits = MLXSW_REG_QPCR_LOWEST_CBS_BITS_SP2; + policer_core->highest_bs_bits = MLXSW_REG_QPCR_HIGHEST_CBS_BITS_SP2; + + return 0; +} + +const struct mlxsw_sp_policer_core_ops mlxsw_sp2_policer_core_ops = { + .init = mlxsw_sp2_policer_core_init, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.c new file mode 100644 index 000000000..7b01b9c20 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.c @@ -0,0 +1,1726 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2019 Mellanox Technologies. All rights reserved */ + +#include <linux/ptp_clock_kernel.h> +#include <linux/clocksource.h> +#include <linux/timecounter.h> +#include <linux/spinlock.h> +#include <linux/device.h> +#include <linux/rhashtable.h> +#include <linux/ptp_classify.h> +#include <linux/if_ether.h> +#include <linux/if_vlan.h> +#include <linux/net_tstamp.h> +#include <linux/refcount.h> + +#include "spectrum.h" +#include "spectrum_ptp.h" +#include "core.h" + +#define MLXSW_SP1_PTP_CLOCK_CYCLES_SHIFT 29 +#define MLXSW_SP1_PTP_CLOCK_FREQ_KHZ 156257 /* 6.4nSec */ +#define MLXSW_SP1_PTP_CLOCK_MASK 64 + +#define MLXSW_SP1_PTP_HT_GC_INTERVAL 500 /* ms */ + +/* How long, approximately, should the unmatched entries stay in the hash table + * before they are collected. Should be evenly divisible by the GC interval. + */ +#define MLXSW_SP1_PTP_HT_GC_TIMEOUT 1000 /* ms */ + +struct mlxsw_sp_ptp_state { + struct mlxsw_sp *mlxsw_sp; +}; + +struct mlxsw_sp1_ptp_state { + struct mlxsw_sp_ptp_state common; + struct rhltable unmatched_ht; + spinlock_t unmatched_lock; /* protects the HT */ + struct delayed_work ht_gc_dw; + u32 gc_cycle; +}; + +struct mlxsw_sp2_ptp_state { + struct mlxsw_sp_ptp_state common; + refcount_t ptp_port_enabled_ref; /* Number of ports with time stamping + * enabled. + */ + struct hwtstamp_config config; + struct mutex lock; /* Protects 'config' and HW configuration. */ +}; + +struct mlxsw_sp1_ptp_key { + u16 local_port; + u8 message_type; + u16 sequence_id; + u8 domain_number; + bool ingress; +}; + +struct mlxsw_sp1_ptp_unmatched { + struct mlxsw_sp1_ptp_key key; + struct rhlist_head ht_node; + struct rcu_head rcu; + struct sk_buff *skb; + u64 timestamp; + u32 gc_cycle; +}; + +static const struct rhashtable_params mlxsw_sp1_ptp_unmatched_ht_params = { + .key_len = sizeof_field(struct mlxsw_sp1_ptp_unmatched, key), + .key_offset = offsetof(struct mlxsw_sp1_ptp_unmatched, key), + .head_offset = offsetof(struct mlxsw_sp1_ptp_unmatched, ht_node), +}; + +struct mlxsw_sp_ptp_clock { + struct mlxsw_core *core; + struct ptp_clock *ptp; + struct ptp_clock_info ptp_info; +}; + +struct mlxsw_sp1_ptp_clock { + struct mlxsw_sp_ptp_clock common; + spinlock_t lock; /* protect this structure */ + struct cyclecounter cycles; + struct timecounter tc; + u32 nominal_c_mult; + unsigned long overflow_period; + struct delayed_work overflow_work; +}; + +static struct mlxsw_sp1_ptp_state * +mlxsw_sp1_ptp_state(struct mlxsw_sp *mlxsw_sp) +{ + return container_of(mlxsw_sp->ptp_state, struct mlxsw_sp1_ptp_state, + common); +} + +static struct mlxsw_sp2_ptp_state * +mlxsw_sp2_ptp_state(struct mlxsw_sp *mlxsw_sp) +{ + return container_of(mlxsw_sp->ptp_state, struct mlxsw_sp2_ptp_state, + common); +} + +static struct mlxsw_sp1_ptp_clock * +mlxsw_sp1_ptp_clock(struct ptp_clock_info *ptp) +{ + return container_of(ptp, struct mlxsw_sp1_ptp_clock, common.ptp_info); +} + +static u64 __mlxsw_sp1_ptp_read_frc(struct mlxsw_sp1_ptp_clock *clock, + struct ptp_system_timestamp *sts) +{ + struct mlxsw_core *mlxsw_core = clock->common.core; + u32 frc_h1, frc_h2, frc_l; + + frc_h1 = mlxsw_core_read_frc_h(mlxsw_core); + ptp_read_system_prets(sts); + frc_l = mlxsw_core_read_frc_l(mlxsw_core); + ptp_read_system_postts(sts); + frc_h2 = mlxsw_core_read_frc_h(mlxsw_core); + + if (frc_h1 != frc_h2) { + /* wrap around */ + ptp_read_system_prets(sts); + frc_l = mlxsw_core_read_frc_l(mlxsw_core); + ptp_read_system_postts(sts); + } + + return (u64) frc_l | (u64) frc_h2 << 32; +} + +static u64 mlxsw_sp1_ptp_read_frc(const struct cyclecounter *cc) +{ + struct mlxsw_sp1_ptp_clock *clock = + container_of(cc, struct mlxsw_sp1_ptp_clock, cycles); + + return __mlxsw_sp1_ptp_read_frc(clock, NULL) & cc->mask; +} + +static int +mlxsw_sp_ptp_phc_adjfreq(struct mlxsw_sp_ptp_clock *clock, int freq_adj) +{ + struct mlxsw_core *mlxsw_core = clock->core; + char mtutc_pl[MLXSW_REG_MTUTC_LEN]; + + mlxsw_reg_mtutc_pack(mtutc_pl, MLXSW_REG_MTUTC_OPERATION_ADJUST_FREQ, + freq_adj, 0, 0, 0); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mtutc), mtutc_pl); +} + +static u64 mlxsw_sp1_ptp_ns2cycles(const struct timecounter *tc, u64 nsec) +{ + u64 cycles = (u64) nsec; + + cycles <<= tc->cc->shift; + cycles = div_u64(cycles, tc->cc->mult); + + return cycles; +} + +static int +mlxsw_sp1_ptp_phc_settime(struct mlxsw_sp1_ptp_clock *clock, u64 nsec) +{ + struct mlxsw_core *mlxsw_core = clock->common.core; + u64 next_sec, next_sec_in_nsec, cycles; + char mtutc_pl[MLXSW_REG_MTUTC_LEN]; + char mtpps_pl[MLXSW_REG_MTPPS_LEN]; + int err; + + next_sec = div_u64(nsec, NSEC_PER_SEC) + 1; + next_sec_in_nsec = next_sec * NSEC_PER_SEC; + + spin_lock_bh(&clock->lock); + cycles = mlxsw_sp1_ptp_ns2cycles(&clock->tc, next_sec_in_nsec); + spin_unlock_bh(&clock->lock); + + mlxsw_reg_mtpps_vpin_pack(mtpps_pl, cycles); + err = mlxsw_reg_write(mlxsw_core, MLXSW_REG(mtpps), mtpps_pl); + if (err) + return err; + + mlxsw_reg_mtutc_pack(mtutc_pl, + MLXSW_REG_MTUTC_OPERATION_SET_TIME_AT_NEXT_SEC, + 0, next_sec, 0, 0); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mtutc), mtutc_pl); +} + +static int mlxsw_sp1_ptp_adjfine(struct ptp_clock_info *ptp, long scaled_ppm) +{ + struct mlxsw_sp1_ptp_clock *clock = mlxsw_sp1_ptp_clock(ptp); + int neg_adj = 0; + u32 diff; + u64 adj; + s32 ppb; + + ppb = scaled_ppm_to_ppb(scaled_ppm); + + if (ppb < 0) { + neg_adj = 1; + ppb = -ppb; + } + + adj = clock->nominal_c_mult; + adj *= ppb; + diff = div_u64(adj, NSEC_PER_SEC); + + spin_lock_bh(&clock->lock); + timecounter_read(&clock->tc); + clock->cycles.mult = neg_adj ? clock->nominal_c_mult - diff : + clock->nominal_c_mult + diff; + spin_unlock_bh(&clock->lock); + + return mlxsw_sp_ptp_phc_adjfreq(&clock->common, neg_adj ? -ppb : ppb); +} + +static int mlxsw_sp1_ptp_adjtime(struct ptp_clock_info *ptp, s64 delta) +{ + struct mlxsw_sp1_ptp_clock *clock = mlxsw_sp1_ptp_clock(ptp); + u64 nsec; + + spin_lock_bh(&clock->lock); + timecounter_adjtime(&clock->tc, delta); + nsec = timecounter_read(&clock->tc); + spin_unlock_bh(&clock->lock); + + return mlxsw_sp1_ptp_phc_settime(clock, nsec); +} + +static int mlxsw_sp1_ptp_gettimex(struct ptp_clock_info *ptp, + struct timespec64 *ts, + struct ptp_system_timestamp *sts) +{ + struct mlxsw_sp1_ptp_clock *clock = mlxsw_sp1_ptp_clock(ptp); + u64 cycles, nsec; + + spin_lock_bh(&clock->lock); + cycles = __mlxsw_sp1_ptp_read_frc(clock, sts); + nsec = timecounter_cyc2time(&clock->tc, cycles); + spin_unlock_bh(&clock->lock); + + *ts = ns_to_timespec64(nsec); + + return 0; +} + +static int mlxsw_sp1_ptp_settime(struct ptp_clock_info *ptp, + const struct timespec64 *ts) +{ + struct mlxsw_sp1_ptp_clock *clock = mlxsw_sp1_ptp_clock(ptp); + u64 nsec = timespec64_to_ns(ts); + + spin_lock_bh(&clock->lock); + timecounter_init(&clock->tc, &clock->cycles, nsec); + nsec = timecounter_read(&clock->tc); + spin_unlock_bh(&clock->lock); + + return mlxsw_sp1_ptp_phc_settime(clock, nsec); +} + +static const struct ptp_clock_info mlxsw_sp1_ptp_clock_info = { + .owner = THIS_MODULE, + .name = "mlxsw_sp_clock", + .max_adj = 100000000, + .adjfine = mlxsw_sp1_ptp_adjfine, + .adjtime = mlxsw_sp1_ptp_adjtime, + .gettimex64 = mlxsw_sp1_ptp_gettimex, + .settime64 = mlxsw_sp1_ptp_settime, +}; + +static void mlxsw_sp1_ptp_clock_overflow(struct work_struct *work) +{ + struct delayed_work *dwork = to_delayed_work(work); + struct mlxsw_sp1_ptp_clock *clock; + + clock = container_of(dwork, struct mlxsw_sp1_ptp_clock, overflow_work); + + spin_lock_bh(&clock->lock); + timecounter_read(&clock->tc); + spin_unlock_bh(&clock->lock); + mlxsw_core_schedule_dw(&clock->overflow_work, clock->overflow_period); +} + +struct mlxsw_sp_ptp_clock * +mlxsw_sp1_ptp_clock_init(struct mlxsw_sp *mlxsw_sp, struct device *dev) +{ + u64 overflow_cycles, nsec, frac = 0; + struct mlxsw_sp1_ptp_clock *clock; + int err; + + clock = kzalloc(sizeof(*clock), GFP_KERNEL); + if (!clock) + return ERR_PTR(-ENOMEM); + + spin_lock_init(&clock->lock); + clock->cycles.read = mlxsw_sp1_ptp_read_frc; + clock->cycles.shift = MLXSW_SP1_PTP_CLOCK_CYCLES_SHIFT; + clock->cycles.mult = clocksource_khz2mult(MLXSW_SP1_PTP_CLOCK_FREQ_KHZ, + clock->cycles.shift); + clock->nominal_c_mult = clock->cycles.mult; + clock->cycles.mask = CLOCKSOURCE_MASK(MLXSW_SP1_PTP_CLOCK_MASK); + clock->common.core = mlxsw_sp->core; + + timecounter_init(&clock->tc, &clock->cycles, 0); + + /* Calculate period in seconds to call the overflow watchdog - to make + * sure counter is checked at least twice every wrap around. + * The period is calculated as the minimum between max HW cycles count + * (The clock source mask) and max amount of cycles that can be + * multiplied by clock multiplier where the result doesn't exceed + * 64bits. + */ + overflow_cycles = div64_u64(~0ULL >> 1, clock->cycles.mult); + overflow_cycles = min(overflow_cycles, div_u64(clock->cycles.mask, 3)); + + nsec = cyclecounter_cyc2ns(&clock->cycles, overflow_cycles, 0, &frac); + clock->overflow_period = nsecs_to_jiffies(nsec); + + INIT_DELAYED_WORK(&clock->overflow_work, mlxsw_sp1_ptp_clock_overflow); + mlxsw_core_schedule_dw(&clock->overflow_work, 0); + + clock->common.ptp_info = mlxsw_sp1_ptp_clock_info; + clock->common.ptp = ptp_clock_register(&clock->common.ptp_info, dev); + if (IS_ERR(clock->common.ptp)) { + err = PTR_ERR(clock->common.ptp); + dev_err(dev, "ptp_clock_register failed %d\n", err); + goto err_ptp_clock_register; + } + + return &clock->common; + +err_ptp_clock_register: + cancel_delayed_work_sync(&clock->overflow_work); + kfree(clock); + return ERR_PTR(err); +} + +void mlxsw_sp1_ptp_clock_fini(struct mlxsw_sp_ptp_clock *clock_common) +{ + struct mlxsw_sp1_ptp_clock *clock = + container_of(clock_common, struct mlxsw_sp1_ptp_clock, common); + + ptp_clock_unregister(clock_common->ptp); + cancel_delayed_work_sync(&clock->overflow_work); + kfree(clock); +} + +static u64 mlxsw_sp2_ptp_read_utc(struct mlxsw_sp_ptp_clock *clock, + struct ptp_system_timestamp *sts) +{ + struct mlxsw_core *mlxsw_core = clock->core; + u32 utc_sec1, utc_sec2, utc_nsec; + + utc_sec1 = mlxsw_core_read_utc_sec(mlxsw_core); + ptp_read_system_prets(sts); + utc_nsec = mlxsw_core_read_utc_nsec(mlxsw_core); + ptp_read_system_postts(sts); + utc_sec2 = mlxsw_core_read_utc_sec(mlxsw_core); + + if (utc_sec1 != utc_sec2) { + /* Wrap around. */ + ptp_read_system_prets(sts); + utc_nsec = mlxsw_core_read_utc_nsec(mlxsw_core); + ptp_read_system_postts(sts); + } + + return (u64)utc_sec2 * NSEC_PER_SEC + utc_nsec; +} + +static int +mlxsw_sp2_ptp_phc_settime(struct mlxsw_sp_ptp_clock *clock, u64 nsec) +{ + struct mlxsw_core *mlxsw_core = clock->core; + char mtutc_pl[MLXSW_REG_MTUTC_LEN]; + u32 sec, nsec_rem; + + sec = div_u64_rem(nsec, NSEC_PER_SEC, &nsec_rem); + mlxsw_reg_mtutc_pack(mtutc_pl, + MLXSW_REG_MTUTC_OPERATION_SET_TIME_IMMEDIATE, + 0, sec, nsec_rem, 0); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mtutc), mtutc_pl); +} + +static int mlxsw_sp2_ptp_adjfine(struct ptp_clock_info *ptp, long scaled_ppm) +{ + struct mlxsw_sp_ptp_clock *clock = + container_of(ptp, struct mlxsw_sp_ptp_clock, ptp_info); + s32 ppb = scaled_ppm_to_ppb(scaled_ppm); + + /* In Spectrum-2 and newer ASICs, the frequency adjustment in MTUTC is + * reversed, positive values mean to decrease the frequency. Adjust the + * sign of PPB to this behavior. + */ + return mlxsw_sp_ptp_phc_adjfreq(clock, -ppb); +} + +static int mlxsw_sp2_ptp_adjtime(struct ptp_clock_info *ptp, s64 delta) +{ + struct mlxsw_sp_ptp_clock *clock = + container_of(ptp, struct mlxsw_sp_ptp_clock, ptp_info); + struct mlxsw_core *mlxsw_core = clock->core; + char mtutc_pl[MLXSW_REG_MTUTC_LEN]; + + /* HW time adjustment range is s16. If out of range, set time instead. */ + if (delta < S16_MIN || delta > S16_MAX) { + u64 nsec; + + nsec = mlxsw_sp2_ptp_read_utc(clock, NULL); + nsec += delta; + + return mlxsw_sp2_ptp_phc_settime(clock, nsec); + } + + mlxsw_reg_mtutc_pack(mtutc_pl, + MLXSW_REG_MTUTC_OPERATION_ADJUST_TIME, + 0, 0, 0, delta); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(mtutc), mtutc_pl); +} + +static int mlxsw_sp2_ptp_gettimex(struct ptp_clock_info *ptp, + struct timespec64 *ts, + struct ptp_system_timestamp *sts) +{ + struct mlxsw_sp_ptp_clock *clock = + container_of(ptp, struct mlxsw_sp_ptp_clock, ptp_info); + u64 nsec; + + nsec = mlxsw_sp2_ptp_read_utc(clock, sts); + *ts = ns_to_timespec64(nsec); + + return 0; +} + +static int mlxsw_sp2_ptp_settime(struct ptp_clock_info *ptp, + const struct timespec64 *ts) +{ + struct mlxsw_sp_ptp_clock *clock = + container_of(ptp, struct mlxsw_sp_ptp_clock, ptp_info); + u64 nsec = timespec64_to_ns(ts); + + return mlxsw_sp2_ptp_phc_settime(clock, nsec); +} + +static const struct ptp_clock_info mlxsw_sp2_ptp_clock_info = { + .owner = THIS_MODULE, + .name = "mlxsw_sp_clock", + .max_adj = MLXSW_REG_MTUTC_MAX_FREQ_ADJ, + .adjfine = mlxsw_sp2_ptp_adjfine, + .adjtime = mlxsw_sp2_ptp_adjtime, + .gettimex64 = mlxsw_sp2_ptp_gettimex, + .settime64 = mlxsw_sp2_ptp_settime, +}; + +struct mlxsw_sp_ptp_clock * +mlxsw_sp2_ptp_clock_init(struct mlxsw_sp *mlxsw_sp, struct device *dev) +{ + struct mlxsw_sp_ptp_clock *clock; + int err; + + clock = kzalloc(sizeof(*clock), GFP_KERNEL); + if (!clock) + return ERR_PTR(-ENOMEM); + + clock->core = mlxsw_sp->core; + + clock->ptp_info = mlxsw_sp2_ptp_clock_info; + + err = mlxsw_sp2_ptp_phc_settime(clock, 0); + if (err) { + dev_err(dev, "setting UTC time failed %d\n", err); + goto err_ptp_phc_settime; + } + + clock->ptp = ptp_clock_register(&clock->ptp_info, dev); + if (IS_ERR(clock->ptp)) { + err = PTR_ERR(clock->ptp); + dev_err(dev, "ptp_clock_register failed %d\n", err); + goto err_ptp_clock_register; + } + + return clock; + +err_ptp_clock_register: +err_ptp_phc_settime: + kfree(clock); + return ERR_PTR(err); +} + +void mlxsw_sp2_ptp_clock_fini(struct mlxsw_sp_ptp_clock *clock) +{ + ptp_clock_unregister(clock->ptp); + kfree(clock); +} + +static int mlxsw_sp_ptp_parse(struct sk_buff *skb, + u8 *p_domain_number, + u8 *p_message_type, + u16 *p_sequence_id) +{ + unsigned int ptp_class; + struct ptp_header *hdr; + + ptp_class = ptp_classify_raw(skb); + + switch (ptp_class & PTP_CLASS_VMASK) { + case PTP_CLASS_V1: + case PTP_CLASS_V2: + break; + default: + return -ERANGE; + } + + hdr = ptp_parse_header(skb, ptp_class); + if (!hdr) + return -EINVAL; + + *p_message_type = ptp_get_msgtype(hdr, ptp_class); + *p_domain_number = hdr->domain_number; + *p_sequence_id = be16_to_cpu(hdr->sequence_id); + + return 0; +} + +/* Returns NULL on successful insertion, a pointer on conflict, or an ERR_PTR on + * error. + */ +static int +mlxsw_sp1_ptp_unmatched_save(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp1_ptp_key key, + struct sk_buff *skb, + u64 timestamp) +{ + int cycles = MLXSW_SP1_PTP_HT_GC_TIMEOUT / MLXSW_SP1_PTP_HT_GC_INTERVAL; + struct mlxsw_sp1_ptp_state *ptp_state = mlxsw_sp1_ptp_state(mlxsw_sp); + struct mlxsw_sp1_ptp_unmatched *unmatched; + int err; + + unmatched = kzalloc(sizeof(*unmatched), GFP_ATOMIC); + if (!unmatched) + return -ENOMEM; + + unmatched->key = key; + unmatched->skb = skb; + unmatched->timestamp = timestamp; + unmatched->gc_cycle = ptp_state->gc_cycle + cycles; + + err = rhltable_insert(&ptp_state->unmatched_ht, &unmatched->ht_node, + mlxsw_sp1_ptp_unmatched_ht_params); + if (err) + kfree(unmatched); + + return err; +} + +static struct mlxsw_sp1_ptp_unmatched * +mlxsw_sp1_ptp_unmatched_lookup(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp1_ptp_key key, int *p_length) +{ + struct mlxsw_sp1_ptp_state *ptp_state = mlxsw_sp1_ptp_state(mlxsw_sp); + struct mlxsw_sp1_ptp_unmatched *unmatched, *last = NULL; + struct rhlist_head *tmp, *list; + int length = 0; + + list = rhltable_lookup(&ptp_state->unmatched_ht, &key, + mlxsw_sp1_ptp_unmatched_ht_params); + rhl_for_each_entry_rcu(unmatched, tmp, list, ht_node) { + last = unmatched; + length++; + } + + *p_length = length; + return last; +} + +static int +mlxsw_sp1_ptp_unmatched_remove(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp1_ptp_unmatched *unmatched) +{ + struct mlxsw_sp1_ptp_state *ptp_state = mlxsw_sp1_ptp_state(mlxsw_sp); + + return rhltable_remove(&ptp_state->unmatched_ht, + &unmatched->ht_node, + mlxsw_sp1_ptp_unmatched_ht_params); +} + +/* This function is called in the following scenarios: + * + * 1) When a packet is matched with its timestamp. + * 2) In several situation when it is necessary to immediately pass on + * an SKB without a timestamp. + * 3) From GC indirectly through mlxsw_sp1_ptp_unmatched_finish(). + * This case is similar to 2) above. + */ +static void mlxsw_sp1_ptp_packet_finish(struct mlxsw_sp *mlxsw_sp, + struct sk_buff *skb, u16 local_port, + bool ingress, + struct skb_shared_hwtstamps *hwtstamps) +{ + struct mlxsw_sp_port *mlxsw_sp_port; + + /* Between capturing the packet and finishing it, there is a window of + * opportunity for the originating port to go away (e.g. due to a + * split). Also make sure the SKB device reference is still valid. + */ + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + if (!(mlxsw_sp_port && (!skb->dev || skb->dev == mlxsw_sp_port->dev))) { + dev_kfree_skb_any(skb); + return; + } + + if (ingress) { + if (hwtstamps) + *skb_hwtstamps(skb) = *hwtstamps; + mlxsw_sp_rx_listener_no_mark_func(skb, local_port, mlxsw_sp); + } else { + /* skb_tstamp_tx() allows hwtstamps to be NULL. */ + skb_tstamp_tx(skb, hwtstamps); + dev_kfree_skb_any(skb); + } +} + +static void mlxsw_sp1_packet_timestamp(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp1_ptp_key key, + struct sk_buff *skb, + u64 timestamp) +{ + struct mlxsw_sp_ptp_clock *clock_common = mlxsw_sp->clock; + struct mlxsw_sp1_ptp_clock *clock = + container_of(clock_common, struct mlxsw_sp1_ptp_clock, common); + + struct skb_shared_hwtstamps hwtstamps; + u64 nsec; + + spin_lock_bh(&clock->lock); + nsec = timecounter_cyc2time(&clock->tc, timestamp); + spin_unlock_bh(&clock->lock); + + hwtstamps.hwtstamp = ns_to_ktime(nsec); + mlxsw_sp1_ptp_packet_finish(mlxsw_sp, skb, + key.local_port, key.ingress, &hwtstamps); +} + +static void +mlxsw_sp1_ptp_unmatched_finish(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp1_ptp_unmatched *unmatched) +{ + if (unmatched->skb && unmatched->timestamp) + mlxsw_sp1_packet_timestamp(mlxsw_sp, unmatched->key, + unmatched->skb, + unmatched->timestamp); + else if (unmatched->skb) + mlxsw_sp1_ptp_packet_finish(mlxsw_sp, unmatched->skb, + unmatched->key.local_port, + unmatched->key.ingress, NULL); + kfree_rcu(unmatched, rcu); +} + +static void mlxsw_sp1_ptp_unmatched_free_fn(void *ptr, void *arg) +{ + struct mlxsw_sp1_ptp_unmatched *unmatched = ptr; + + /* This is invoked at a point where the ports are gone already. Nothing + * to do with whatever is left in the HT but to free it. + */ + if (unmatched->skb) + dev_kfree_skb_any(unmatched->skb); + kfree_rcu(unmatched, rcu); +} + +static void mlxsw_sp1_ptp_got_piece(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp1_ptp_key key, + struct sk_buff *skb, u64 timestamp) +{ + struct mlxsw_sp1_ptp_state *ptp_state = mlxsw_sp1_ptp_state(mlxsw_sp); + struct mlxsw_sp1_ptp_unmatched *unmatched; + int length; + int err; + + rcu_read_lock(); + + spin_lock(&ptp_state->unmatched_lock); + + unmatched = mlxsw_sp1_ptp_unmatched_lookup(mlxsw_sp, key, &length); + if (skb && unmatched && unmatched->timestamp) { + unmatched->skb = skb; + } else if (timestamp && unmatched && unmatched->skb) { + unmatched->timestamp = timestamp; + } else { + /* Either there is no entry to match, or one that is there is + * incompatible. + */ + if (length < 100) + err = mlxsw_sp1_ptp_unmatched_save(mlxsw_sp, key, + skb, timestamp); + else + err = -E2BIG; + if (err && skb) + mlxsw_sp1_ptp_packet_finish(mlxsw_sp, skb, + key.local_port, + key.ingress, NULL); + unmatched = NULL; + } + + if (unmatched) { + err = mlxsw_sp1_ptp_unmatched_remove(mlxsw_sp, unmatched); + WARN_ON_ONCE(err); + } + + spin_unlock(&ptp_state->unmatched_lock); + + if (unmatched) + mlxsw_sp1_ptp_unmatched_finish(mlxsw_sp, unmatched); + + rcu_read_unlock(); +} + +static void mlxsw_sp1_ptp_got_packet(struct mlxsw_sp *mlxsw_sp, + struct sk_buff *skb, u16 local_port, + bool ingress) +{ + struct mlxsw_sp_port *mlxsw_sp_port; + struct mlxsw_sp1_ptp_key key; + u8 types; + int err; + + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + if (!mlxsw_sp_port) + goto immediate; + + types = ingress ? mlxsw_sp_port->ptp.ing_types : + mlxsw_sp_port->ptp.egr_types; + if (!types) + goto immediate; + + memset(&key, 0, sizeof(key)); + key.local_port = local_port; + key.ingress = ingress; + + err = mlxsw_sp_ptp_parse(skb, &key.domain_number, &key.message_type, + &key.sequence_id); + if (err) + goto immediate; + + /* For packets whose timestamping was not enabled on this port, don't + * bother trying to match the timestamp. + */ + if (!((1 << key.message_type) & types)) + goto immediate; + + mlxsw_sp1_ptp_got_piece(mlxsw_sp, key, skb, 0); + return; + +immediate: + mlxsw_sp1_ptp_packet_finish(mlxsw_sp, skb, local_port, ingress, NULL); +} + +void mlxsw_sp1_ptp_got_timestamp(struct mlxsw_sp *mlxsw_sp, bool ingress, + u16 local_port, u8 message_type, + u8 domain_number, u16 sequence_id, + u64 timestamp) +{ + struct mlxsw_sp_port *mlxsw_sp_port; + struct mlxsw_sp1_ptp_key key; + u8 types; + + if (WARN_ON_ONCE(!mlxsw_sp_local_port_is_valid(mlxsw_sp, local_port))) + return; + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + if (!mlxsw_sp_port) + return; + + types = ingress ? mlxsw_sp_port->ptp.ing_types : + mlxsw_sp_port->ptp.egr_types; + + /* For message types whose timestamping was not enabled on this port, + * don't bother with the timestamp. + */ + if (!((1 << message_type) & types)) + return; + + memset(&key, 0, sizeof(key)); + key.local_port = local_port; + key.domain_number = domain_number; + key.message_type = message_type; + key.sequence_id = sequence_id; + key.ingress = ingress; + + mlxsw_sp1_ptp_got_piece(mlxsw_sp, key, NULL, timestamp); +} + +void mlxsw_sp1_ptp_receive(struct mlxsw_sp *mlxsw_sp, struct sk_buff *skb, + u16 local_port) +{ + skb_reset_mac_header(skb); + mlxsw_sp1_ptp_got_packet(mlxsw_sp, skb, local_port, true); +} + +void mlxsw_sp1_ptp_transmitted(struct mlxsw_sp *mlxsw_sp, + struct sk_buff *skb, u16 local_port) +{ + mlxsw_sp1_ptp_got_packet(mlxsw_sp, skb, local_port, false); +} + +static void +mlxsw_sp1_ptp_ht_gc_collect(struct mlxsw_sp1_ptp_state *ptp_state, + struct mlxsw_sp1_ptp_unmatched *unmatched) +{ + struct mlxsw_sp *mlxsw_sp = ptp_state->common.mlxsw_sp; + struct mlxsw_sp_ptp_port_dir_stats *stats; + struct mlxsw_sp_port *mlxsw_sp_port; + int err; + + /* If an unmatched entry has an SKB, it has to be handed over to the + * networking stack. This is usually done from a trap handler, which is + * invoked in a softirq context. Here we are going to do it in process + * context. If that were to be interrupted by a softirq, it could cause + * a deadlock when an attempt is made to take an already-taken lock + * somewhere along the sending path. Disable softirqs to prevent this. + */ + local_bh_disable(); + + spin_lock(&ptp_state->unmatched_lock); + err = rhltable_remove(&ptp_state->unmatched_ht, &unmatched->ht_node, + mlxsw_sp1_ptp_unmatched_ht_params); + spin_unlock(&ptp_state->unmatched_lock); + + if (err) + /* The packet was matched with timestamp during the walk. */ + goto out; + + mlxsw_sp_port = mlxsw_sp->ports[unmatched->key.local_port]; + if (mlxsw_sp_port) { + stats = unmatched->key.ingress ? + &mlxsw_sp_port->ptp.stats.rx_gcd : + &mlxsw_sp_port->ptp.stats.tx_gcd; + if (unmatched->skb) + stats->packets++; + else + stats->timestamps++; + } + + /* mlxsw_sp1_ptp_unmatched_finish() invokes netif_receive_skb(). While + * the comment at that function states that it can only be called in + * soft IRQ context, this pattern of local_bh_disable() + + * netif_receive_skb(), in process context, is seen elsewhere in the + * kernel, notably in pktgen. + */ + mlxsw_sp1_ptp_unmatched_finish(mlxsw_sp, unmatched); + +out: + local_bh_enable(); +} + +static void mlxsw_sp1_ptp_ht_gc(struct work_struct *work) +{ + struct delayed_work *dwork = to_delayed_work(work); + struct mlxsw_sp1_ptp_unmatched *unmatched; + struct mlxsw_sp1_ptp_state *ptp_state; + struct rhashtable_iter iter; + u32 gc_cycle; + void *obj; + + ptp_state = container_of(dwork, struct mlxsw_sp1_ptp_state, ht_gc_dw); + gc_cycle = ptp_state->gc_cycle++; + + rhltable_walk_enter(&ptp_state->unmatched_ht, &iter); + rhashtable_walk_start(&iter); + while ((obj = rhashtable_walk_next(&iter))) { + if (IS_ERR(obj)) + continue; + + unmatched = obj; + if (unmatched->gc_cycle <= gc_cycle) + mlxsw_sp1_ptp_ht_gc_collect(ptp_state, unmatched); + } + rhashtable_walk_stop(&iter); + rhashtable_walk_exit(&iter); + + mlxsw_core_schedule_dw(&ptp_state->ht_gc_dw, + MLXSW_SP1_PTP_HT_GC_INTERVAL); +} + +static int mlxsw_sp_ptp_mtptpt_set(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_reg_mtptpt_trap_id trap_id, + u16 message_type) +{ + char mtptpt_pl[MLXSW_REG_MTPTPT_LEN]; + + mlxsw_reg_mtptpt_pack(mtptpt_pl, trap_id, message_type); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mtptpt), mtptpt_pl); +} + +static int mlxsw_sp1_ptp_set_fifo_clr_on_trap(struct mlxsw_sp *mlxsw_sp, + bool clr) +{ + char mogcr_pl[MLXSW_REG_MOGCR_LEN] = {0}; + int err; + + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(mogcr), mogcr_pl); + if (err) + return err; + + mlxsw_reg_mogcr_ptp_iftc_set(mogcr_pl, clr); + mlxsw_reg_mogcr_ptp_eftc_set(mogcr_pl, clr); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mogcr), mogcr_pl); +} + +static int mlxsw_sp1_ptp_mtpppc_set(struct mlxsw_sp *mlxsw_sp, + u16 ing_types, u16 egr_types) +{ + char mtpppc_pl[MLXSW_REG_MTPPPC_LEN]; + + mlxsw_reg_mtpppc_pack(mtpppc_pl, ing_types, egr_types); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mtpppc), mtpppc_pl); +} + +struct mlxsw_sp1_ptp_shaper_params { + u32 ethtool_speed; + enum mlxsw_reg_qpsc_port_speed port_speed; + u8 shaper_time_exp; + u8 shaper_time_mantissa; + u8 shaper_inc; + u8 shaper_bs; + u8 port_to_shaper_credits; + int ing_timestamp_inc; + int egr_timestamp_inc; +}; + +static const struct mlxsw_sp1_ptp_shaper_params +mlxsw_sp1_ptp_shaper_params[] = { + { + .ethtool_speed = SPEED_100, + .port_speed = MLXSW_REG_QPSC_PORT_SPEED_100M, + .shaper_time_exp = 4, + .shaper_time_mantissa = 12, + .shaper_inc = 9, + .shaper_bs = 1, + .port_to_shaper_credits = 1, + .ing_timestamp_inc = -313, + .egr_timestamp_inc = 313, + }, + { + .ethtool_speed = SPEED_1000, + .port_speed = MLXSW_REG_QPSC_PORT_SPEED_1G, + .shaper_time_exp = 0, + .shaper_time_mantissa = 12, + .shaper_inc = 6, + .shaper_bs = 0, + .port_to_shaper_credits = 1, + .ing_timestamp_inc = -35, + .egr_timestamp_inc = 35, + }, + { + .ethtool_speed = SPEED_10000, + .port_speed = MLXSW_REG_QPSC_PORT_SPEED_10G, + .shaper_time_exp = 0, + .shaper_time_mantissa = 2, + .shaper_inc = 14, + .shaper_bs = 1, + .port_to_shaper_credits = 1, + .ing_timestamp_inc = -11, + .egr_timestamp_inc = 11, + }, + { + .ethtool_speed = SPEED_25000, + .port_speed = MLXSW_REG_QPSC_PORT_SPEED_25G, + .shaper_time_exp = 0, + .shaper_time_mantissa = 0, + .shaper_inc = 11, + .shaper_bs = 1, + .port_to_shaper_credits = 1, + .ing_timestamp_inc = -14, + .egr_timestamp_inc = 14, + }, +}; + +#define MLXSW_SP1_PTP_SHAPER_PARAMS_LEN ARRAY_SIZE(mlxsw_sp1_ptp_shaper_params) + +static int mlxsw_sp1_ptp_shaper_params_set(struct mlxsw_sp *mlxsw_sp) +{ + const struct mlxsw_sp1_ptp_shaper_params *params; + char qpsc_pl[MLXSW_REG_QPSC_LEN]; + int i, err; + + for (i = 0; i < MLXSW_SP1_PTP_SHAPER_PARAMS_LEN; i++) { + params = &mlxsw_sp1_ptp_shaper_params[i]; + mlxsw_reg_qpsc_pack(qpsc_pl, params->port_speed, + params->shaper_time_exp, + params->shaper_time_mantissa, + params->shaper_inc, params->shaper_bs, + params->port_to_shaper_credits, + params->ing_timestamp_inc, + params->egr_timestamp_inc); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qpsc), qpsc_pl); + if (err) + return err; + } + + return 0; +} + +static int mlxsw_sp_ptp_traps_set(struct mlxsw_sp *mlxsw_sp) +{ + u16 event_message_type; + int err; + + /* Deliver these message types as PTP0. */ + event_message_type = BIT(PTP_MSGTYPE_SYNC) | + BIT(PTP_MSGTYPE_DELAY_REQ) | + BIT(PTP_MSGTYPE_PDELAY_REQ) | + BIT(PTP_MSGTYPE_PDELAY_RESP); + + err = mlxsw_sp_ptp_mtptpt_set(mlxsw_sp, MLXSW_REG_MTPTPT_TRAP_ID_PTP0, + event_message_type); + if (err) + return err; + + /* Everything else is PTP1. */ + err = mlxsw_sp_ptp_mtptpt_set(mlxsw_sp, MLXSW_REG_MTPTPT_TRAP_ID_PTP1, + ~event_message_type); + if (err) + goto err_mtptpt1_set; + + return 0; + +err_mtptpt1_set: + mlxsw_sp_ptp_mtptpt_set(mlxsw_sp, MLXSW_REG_MTPTPT_TRAP_ID_PTP0, 0); + return err; +} + +static void mlxsw_sp_ptp_traps_unset(struct mlxsw_sp *mlxsw_sp) +{ + mlxsw_sp_ptp_mtptpt_set(mlxsw_sp, MLXSW_REG_MTPTPT_TRAP_ID_PTP1, 0); + mlxsw_sp_ptp_mtptpt_set(mlxsw_sp, MLXSW_REG_MTPTPT_TRAP_ID_PTP0, 0); +} + +struct mlxsw_sp_ptp_state *mlxsw_sp1_ptp_init(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp1_ptp_state *ptp_state; + int err; + + err = mlxsw_sp1_ptp_shaper_params_set(mlxsw_sp); + if (err) + return ERR_PTR(err); + + ptp_state = kzalloc(sizeof(*ptp_state), GFP_KERNEL); + if (!ptp_state) + return ERR_PTR(-ENOMEM); + ptp_state->common.mlxsw_sp = mlxsw_sp; + + spin_lock_init(&ptp_state->unmatched_lock); + + err = rhltable_init(&ptp_state->unmatched_ht, + &mlxsw_sp1_ptp_unmatched_ht_params); + if (err) + goto err_hashtable_init; + + err = mlxsw_sp_ptp_traps_set(mlxsw_sp); + if (err) + goto err_ptp_traps_set; + + err = mlxsw_sp1_ptp_set_fifo_clr_on_trap(mlxsw_sp, true); + if (err) + goto err_fifo_clr; + + INIT_DELAYED_WORK(&ptp_state->ht_gc_dw, mlxsw_sp1_ptp_ht_gc); + mlxsw_core_schedule_dw(&ptp_state->ht_gc_dw, + MLXSW_SP1_PTP_HT_GC_INTERVAL); + return &ptp_state->common; + +err_fifo_clr: + mlxsw_sp_ptp_traps_unset(mlxsw_sp); +err_ptp_traps_set: + rhltable_destroy(&ptp_state->unmatched_ht); +err_hashtable_init: + kfree(ptp_state); + return ERR_PTR(err); +} + +void mlxsw_sp1_ptp_fini(struct mlxsw_sp_ptp_state *ptp_state_common) +{ + struct mlxsw_sp *mlxsw_sp = ptp_state_common->mlxsw_sp; + struct mlxsw_sp1_ptp_state *ptp_state; + + ptp_state = mlxsw_sp1_ptp_state(mlxsw_sp); + + cancel_delayed_work_sync(&ptp_state->ht_gc_dw); + mlxsw_sp1_ptp_mtpppc_set(mlxsw_sp, 0, 0); + mlxsw_sp1_ptp_set_fifo_clr_on_trap(mlxsw_sp, false); + mlxsw_sp_ptp_traps_unset(mlxsw_sp); + rhltable_free_and_destroy(&ptp_state->unmatched_ht, + &mlxsw_sp1_ptp_unmatched_free_fn, NULL); + kfree(ptp_state); +} + +int mlxsw_sp1_ptp_hwtstamp_get(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config) +{ + *config = mlxsw_sp_port->ptp.hwtstamp_config; + return 0; +} + +static int +mlxsw_sp1_ptp_get_message_types(const struct hwtstamp_config *config, + u16 *p_ing_types, u16 *p_egr_types, + enum hwtstamp_rx_filters *p_rx_filter) +{ + enum hwtstamp_rx_filters rx_filter = config->rx_filter; + enum hwtstamp_tx_types tx_type = config->tx_type; + u16 ing_types = 0x00; + u16 egr_types = 0x00; + + switch (tx_type) { + case HWTSTAMP_TX_OFF: + egr_types = 0x00; + break; + case HWTSTAMP_TX_ON: + egr_types = 0xff; + break; + case HWTSTAMP_TX_ONESTEP_SYNC: + case HWTSTAMP_TX_ONESTEP_P2P: + return -ERANGE; + default: + return -EINVAL; + } + + switch (rx_filter) { + case HWTSTAMP_FILTER_NONE: + ing_types = 0x00; + break; + case HWTSTAMP_FILTER_PTP_V1_L4_SYNC: + case HWTSTAMP_FILTER_PTP_V2_L4_SYNC: + case HWTSTAMP_FILTER_PTP_V2_L2_SYNC: + case HWTSTAMP_FILTER_PTP_V2_SYNC: + ing_types = 0x01; + break; + case HWTSTAMP_FILTER_PTP_V1_L4_DELAY_REQ: + case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ: + case HWTSTAMP_FILTER_PTP_V2_L2_DELAY_REQ: + case HWTSTAMP_FILTER_PTP_V2_DELAY_REQ: + ing_types = 0x02; + break; + case HWTSTAMP_FILTER_PTP_V1_L4_EVENT: + case HWTSTAMP_FILTER_PTP_V2_L4_EVENT: + case HWTSTAMP_FILTER_PTP_V2_L2_EVENT: + case HWTSTAMP_FILTER_PTP_V2_EVENT: + ing_types = 0x0f; + break; + case HWTSTAMP_FILTER_ALL: + ing_types = 0xff; + break; + case HWTSTAMP_FILTER_SOME: + case HWTSTAMP_FILTER_NTP_ALL: + return -ERANGE; + default: + return -EINVAL; + } + + *p_ing_types = ing_types; + *p_egr_types = egr_types; + *p_rx_filter = rx_filter; + return 0; +} + +static int mlxsw_sp1_ptp_mtpppc_update(struct mlxsw_sp_port *mlxsw_sp_port, + u16 ing_types, u16 egr_types) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_port *tmp; + u16 orig_ing_types = 0; + u16 orig_egr_types = 0; + int err; + int i; + + /* MTPPPC configures timestamping globally, not per port. Find the + * configuration that contains all configured timestamping requests. + */ + for (i = 1; i < mlxsw_core_max_ports(mlxsw_sp->core); i++) { + tmp = mlxsw_sp->ports[i]; + if (tmp) { + orig_ing_types |= tmp->ptp.ing_types; + orig_egr_types |= tmp->ptp.egr_types; + } + if (tmp && tmp != mlxsw_sp_port) { + ing_types |= tmp->ptp.ing_types; + egr_types |= tmp->ptp.egr_types; + } + } + + if ((ing_types || egr_types) && !(orig_ing_types || orig_egr_types)) { + err = mlxsw_sp_parsing_depth_inc(mlxsw_sp); + if (err) { + netdev_err(mlxsw_sp_port->dev, "Failed to increase parsing depth"); + return err; + } + } + if (!(ing_types || egr_types) && (orig_ing_types || orig_egr_types)) + mlxsw_sp_parsing_depth_dec(mlxsw_sp); + + return mlxsw_sp1_ptp_mtpppc_set(mlxsw_sp_port->mlxsw_sp, + ing_types, egr_types); +} + +static bool mlxsw_sp1_ptp_hwtstamp_enabled(struct mlxsw_sp_port *mlxsw_sp_port) +{ + return mlxsw_sp_port->ptp.ing_types || mlxsw_sp_port->ptp.egr_types; +} + +static int +mlxsw_sp1_ptp_port_shaper_set(struct mlxsw_sp_port *mlxsw_sp_port, bool enable) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char qeec_pl[MLXSW_REG_QEEC_LEN]; + + mlxsw_reg_qeec_ptps_pack(qeec_pl, mlxsw_sp_port->local_port, enable); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qeec), qeec_pl); +} + +static int mlxsw_sp1_ptp_port_shaper_check(struct mlxsw_sp_port *mlxsw_sp_port) +{ + bool ptps = false; + int err, i; + u32 speed; + + if (!mlxsw_sp1_ptp_hwtstamp_enabled(mlxsw_sp_port)) + return mlxsw_sp1_ptp_port_shaper_set(mlxsw_sp_port, false); + + err = mlxsw_sp_port_speed_get(mlxsw_sp_port, &speed); + if (err) + return err; + + for (i = 0; i < MLXSW_SP1_PTP_SHAPER_PARAMS_LEN; i++) { + if (mlxsw_sp1_ptp_shaper_params[i].ethtool_speed == speed) { + ptps = true; + break; + } + } + + return mlxsw_sp1_ptp_port_shaper_set(mlxsw_sp_port, ptps); +} + +void mlxsw_sp1_ptp_shaper_work(struct work_struct *work) +{ + struct delayed_work *dwork = to_delayed_work(work); + struct mlxsw_sp_port *mlxsw_sp_port; + int err; + + mlxsw_sp_port = container_of(dwork, struct mlxsw_sp_port, + ptp.shaper_dw); + + if (!mlxsw_sp1_ptp_hwtstamp_enabled(mlxsw_sp_port)) + return; + + err = mlxsw_sp1_ptp_port_shaper_check(mlxsw_sp_port); + if (err) + netdev_err(mlxsw_sp_port->dev, "Failed to set up PTP shaper\n"); +} + +int mlxsw_sp1_ptp_hwtstamp_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config) +{ + enum hwtstamp_rx_filters rx_filter; + u16 ing_types; + u16 egr_types; + int err; + + err = mlxsw_sp1_ptp_get_message_types(config, &ing_types, &egr_types, + &rx_filter); + if (err) + return err; + + err = mlxsw_sp1_ptp_mtpppc_update(mlxsw_sp_port, ing_types, egr_types); + if (err) + return err; + + mlxsw_sp_port->ptp.hwtstamp_config = *config; + mlxsw_sp_port->ptp.ing_types = ing_types; + mlxsw_sp_port->ptp.egr_types = egr_types; + + err = mlxsw_sp1_ptp_port_shaper_check(mlxsw_sp_port); + if (err) + return err; + + /* Notify the ioctl caller what we are actually timestamping. */ + config->rx_filter = rx_filter; + + return 0; +} + +int mlxsw_sp1_ptp_get_ts_info(struct mlxsw_sp *mlxsw_sp, + struct ethtool_ts_info *info) +{ + info->phc_index = ptp_clock_index(mlxsw_sp->clock->ptp); + + info->so_timestamping = SOF_TIMESTAMPING_TX_HARDWARE | + SOF_TIMESTAMPING_RX_HARDWARE | + SOF_TIMESTAMPING_RAW_HARDWARE; + + info->tx_types = BIT(HWTSTAMP_TX_OFF) | + BIT(HWTSTAMP_TX_ON); + + info->rx_filters = BIT(HWTSTAMP_FILTER_NONE) | + BIT(HWTSTAMP_FILTER_ALL); + + return 0; +} + +struct mlxsw_sp_ptp_port_stat { + char str[ETH_GSTRING_LEN]; + ptrdiff_t offset; +}; + +#define MLXSW_SP_PTP_PORT_STAT(NAME, FIELD) \ + { \ + .str = NAME, \ + .offset = offsetof(struct mlxsw_sp_ptp_port_stats, \ + FIELD), \ + } + +static const struct mlxsw_sp_ptp_port_stat mlxsw_sp_ptp_port_stats[] = { + MLXSW_SP_PTP_PORT_STAT("ptp_rx_gcd_packets", rx_gcd.packets), + MLXSW_SP_PTP_PORT_STAT("ptp_rx_gcd_timestamps", rx_gcd.timestamps), + MLXSW_SP_PTP_PORT_STAT("ptp_tx_gcd_packets", tx_gcd.packets), + MLXSW_SP_PTP_PORT_STAT("ptp_tx_gcd_timestamps", tx_gcd.timestamps), +}; + +#undef MLXSW_SP_PTP_PORT_STAT + +#define MLXSW_SP_PTP_PORT_STATS_LEN \ + ARRAY_SIZE(mlxsw_sp_ptp_port_stats) + +int mlxsw_sp1_get_stats_count(void) +{ + return MLXSW_SP_PTP_PORT_STATS_LEN; +} + +void mlxsw_sp1_get_stats_strings(u8 **p) +{ + int i; + + for (i = 0; i < MLXSW_SP_PTP_PORT_STATS_LEN; i++) { + memcpy(*p, mlxsw_sp_ptp_port_stats[i].str, + ETH_GSTRING_LEN); + *p += ETH_GSTRING_LEN; + } +} + +void mlxsw_sp1_get_stats(struct mlxsw_sp_port *mlxsw_sp_port, + u64 *data, int data_index) +{ + void *stats = &mlxsw_sp_port->ptp.stats; + ptrdiff_t offset; + int i; + + data += data_index; + for (i = 0; i < MLXSW_SP_PTP_PORT_STATS_LEN; i++) { + offset = mlxsw_sp_ptp_port_stats[i].offset; + *data++ = *(u64 *)(stats + offset); + } +} + +struct mlxsw_sp_ptp_state *mlxsw_sp2_ptp_init(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp2_ptp_state *ptp_state; + int err; + + ptp_state = kzalloc(sizeof(*ptp_state), GFP_KERNEL); + if (!ptp_state) + return ERR_PTR(-ENOMEM); + + ptp_state->common.mlxsw_sp = mlxsw_sp; + + err = mlxsw_sp_ptp_traps_set(mlxsw_sp); + if (err) + goto err_ptp_traps_set; + + refcount_set(&ptp_state->ptp_port_enabled_ref, 0); + mutex_init(&ptp_state->lock); + return &ptp_state->common; + +err_ptp_traps_set: + kfree(ptp_state); + return ERR_PTR(err); +} + +void mlxsw_sp2_ptp_fini(struct mlxsw_sp_ptp_state *ptp_state_common) +{ + struct mlxsw_sp *mlxsw_sp = ptp_state_common->mlxsw_sp; + struct mlxsw_sp2_ptp_state *ptp_state; + + ptp_state = mlxsw_sp2_ptp_state(mlxsw_sp); + + mutex_destroy(&ptp_state->lock); + mlxsw_sp_ptp_traps_unset(mlxsw_sp); + kfree(ptp_state); +} + +static u32 mlxsw_ptp_utc_time_stamp_sec_get(struct mlxsw_core *mlxsw_core, + u8 cqe_ts_sec) +{ + u32 utc_sec = mlxsw_core_read_utc_sec(mlxsw_core); + + if (cqe_ts_sec > (utc_sec & 0xff)) + /* Time stamp above the last bits of UTC (UTC & 0xff) means the + * latter has wrapped after the time stamp was collected. + */ + utc_sec -= 256; + + utc_sec &= ~0xff; + utc_sec |= cqe_ts_sec; + + return utc_sec; +} + +static void mlxsw_sp2_ptp_hwtstamp_fill(struct mlxsw_core *mlxsw_core, + const struct mlxsw_skb_cb *cb, + struct skb_shared_hwtstamps *hwtstamps) +{ + u64 ts_sec, ts_nsec, nsec; + + WARN_ON_ONCE(!cb->cqe_ts.sec && !cb->cqe_ts.nsec); + + /* The time stamp in the CQE is represented by 38 bits, which is a short + * representation of UTC time. Software should create the full time + * stamp using the global UTC clock. The seconds have only 8 bits in the + * CQE, to create the full time stamp, use the current UTC time and fix + * the seconds according to the relation between UTC seconds and CQE + * seconds. + */ + ts_sec = mlxsw_ptp_utc_time_stamp_sec_get(mlxsw_core, cb->cqe_ts.sec); + ts_nsec = cb->cqe_ts.nsec; + + nsec = ts_sec * NSEC_PER_SEC + ts_nsec; + + hwtstamps->hwtstamp = ns_to_ktime(nsec); +} + +void mlxsw_sp2_ptp_receive(struct mlxsw_sp *mlxsw_sp, struct sk_buff *skb, + u16 local_port) +{ + struct skb_shared_hwtstamps hwtstamps; + + mlxsw_sp2_ptp_hwtstamp_fill(mlxsw_sp->core, mlxsw_skb_cb(skb), + &hwtstamps); + *skb_hwtstamps(skb) = hwtstamps; + mlxsw_sp_rx_listener_no_mark_func(skb, local_port, mlxsw_sp); +} + +void mlxsw_sp2_ptp_transmitted(struct mlxsw_sp *mlxsw_sp, + struct sk_buff *skb, u16 local_port) +{ + struct skb_shared_hwtstamps hwtstamps; + + mlxsw_sp2_ptp_hwtstamp_fill(mlxsw_sp->core, mlxsw_skb_cb(skb), + &hwtstamps); + skb_tstamp_tx(skb, &hwtstamps); + dev_kfree_skb_any(skb); +} + +int mlxsw_sp2_ptp_hwtstamp_get(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config) +{ + struct mlxsw_sp2_ptp_state *ptp_state; + + ptp_state = mlxsw_sp2_ptp_state(mlxsw_sp_port->mlxsw_sp); + + mutex_lock(&ptp_state->lock); + *config = ptp_state->config; + mutex_unlock(&ptp_state->lock); + + return 0; +} + +static int +mlxsw_sp2_ptp_get_message_types(const struct hwtstamp_config *config, + u16 *p_ing_types, u16 *p_egr_types, + enum hwtstamp_rx_filters *p_rx_filter) +{ + enum hwtstamp_rx_filters rx_filter = config->rx_filter; + enum hwtstamp_tx_types tx_type = config->tx_type; + u16 ing_types = 0x00; + u16 egr_types = 0x00; + + *p_rx_filter = rx_filter; + + switch (rx_filter) { + case HWTSTAMP_FILTER_NONE: + ing_types = 0x00; + break; + case HWTSTAMP_FILTER_PTP_V1_L4_SYNC: + case HWTSTAMP_FILTER_PTP_V2_L4_SYNC: + case HWTSTAMP_FILTER_PTP_V2_L2_SYNC: + case HWTSTAMP_FILTER_PTP_V2_SYNC: + case HWTSTAMP_FILTER_PTP_V1_L4_DELAY_REQ: + case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ: + case HWTSTAMP_FILTER_PTP_V2_L2_DELAY_REQ: + case HWTSTAMP_FILTER_PTP_V2_DELAY_REQ: + case HWTSTAMP_FILTER_PTP_V1_L4_EVENT: + case HWTSTAMP_FILTER_PTP_V2_L4_EVENT: + case HWTSTAMP_FILTER_PTP_V2_L2_EVENT: + case HWTSTAMP_FILTER_PTP_V2_EVENT: + /* In Spectrum-2 and above, all packets get time stamp by + * default and the driver fill the time stamp only for event + * packets. Return all event types even if only specific types + * were required. + */ + ing_types = 0x0f; + *p_rx_filter = HWTSTAMP_FILTER_SOME; + break; + case HWTSTAMP_FILTER_ALL: + case HWTSTAMP_FILTER_SOME: + case HWTSTAMP_FILTER_NTP_ALL: + return -ERANGE; + default: + return -EINVAL; + } + + switch (tx_type) { + case HWTSTAMP_TX_OFF: + egr_types = 0x00; + break; + case HWTSTAMP_TX_ON: + egr_types = 0x0f; + break; + case HWTSTAMP_TX_ONESTEP_SYNC: + case HWTSTAMP_TX_ONESTEP_P2P: + return -ERANGE; + default: + return -EINVAL; + } + + if ((ing_types && !egr_types) || (!ing_types && egr_types)) + return -EINVAL; + + *p_ing_types = ing_types; + *p_egr_types = egr_types; + return 0; +} + +static int mlxsw_sp2_ptp_mtpcpc_set(struct mlxsw_sp *mlxsw_sp, bool ptp_trap_en, + u16 ing_types, u16 egr_types) +{ + char mtpcpc_pl[MLXSW_REG_MTPCPC_LEN]; + + mlxsw_reg_mtpcpc_pack(mtpcpc_pl, false, 0, ptp_trap_en, ing_types, + egr_types); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mtpcpc), mtpcpc_pl); +} + +static int mlxsw_sp2_ptp_enable(struct mlxsw_sp *mlxsw_sp, u16 ing_types, + u16 egr_types, + struct hwtstamp_config new_config) +{ + struct mlxsw_sp2_ptp_state *ptp_state = mlxsw_sp2_ptp_state(mlxsw_sp); + int err; + + err = mlxsw_sp2_ptp_mtpcpc_set(mlxsw_sp, true, ing_types, egr_types); + if (err) + return err; + + ptp_state->config = new_config; + return 0; +} + +static int mlxsw_sp2_ptp_disable(struct mlxsw_sp *mlxsw_sp, + struct hwtstamp_config new_config) +{ + struct mlxsw_sp2_ptp_state *ptp_state = mlxsw_sp2_ptp_state(mlxsw_sp); + int err; + + err = mlxsw_sp2_ptp_mtpcpc_set(mlxsw_sp, false, 0, 0); + if (err) + return err; + + ptp_state->config = new_config; + return 0; +} + +static int mlxsw_sp2_ptp_configure_port(struct mlxsw_sp_port *mlxsw_sp_port, + u16 ing_types, u16 egr_types, + struct hwtstamp_config new_config) +{ + struct mlxsw_sp2_ptp_state *ptp_state; + int err; + + ptp_state = mlxsw_sp2_ptp_state(mlxsw_sp_port->mlxsw_sp); + + if (refcount_inc_not_zero(&ptp_state->ptp_port_enabled_ref)) + return 0; + + err = mlxsw_sp2_ptp_enable(mlxsw_sp_port->mlxsw_sp, ing_types, + egr_types, new_config); + if (err) + return err; + + refcount_set(&ptp_state->ptp_port_enabled_ref, 1); + + return 0; +} + +static int mlxsw_sp2_ptp_deconfigure_port(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config new_config) +{ + struct mlxsw_sp2_ptp_state *ptp_state; + int err; + + ptp_state = mlxsw_sp2_ptp_state(mlxsw_sp_port->mlxsw_sp); + + if (!refcount_dec_and_test(&ptp_state->ptp_port_enabled_ref)) + return 0; + + err = mlxsw_sp2_ptp_disable(mlxsw_sp_port->mlxsw_sp, new_config); + if (err) + goto err_ptp_disable; + + return 0; + +err_ptp_disable: + refcount_set(&ptp_state->ptp_port_enabled_ref, 1); + return err; +} + +int mlxsw_sp2_ptp_hwtstamp_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config) +{ + struct mlxsw_sp2_ptp_state *ptp_state; + enum hwtstamp_rx_filters rx_filter; + struct hwtstamp_config new_config; + u16 new_ing_types, new_egr_types; + bool ptp_enabled; + int err; + + ptp_state = mlxsw_sp2_ptp_state(mlxsw_sp_port->mlxsw_sp); + mutex_lock(&ptp_state->lock); + + err = mlxsw_sp2_ptp_get_message_types(config, &new_ing_types, + &new_egr_types, &rx_filter); + if (err) + goto err_get_message_types; + + new_config.flags = config->flags; + new_config.tx_type = config->tx_type; + new_config.rx_filter = rx_filter; + + ptp_enabled = mlxsw_sp_port->ptp.ing_types || + mlxsw_sp_port->ptp.egr_types; + + if ((new_ing_types || new_egr_types) && !ptp_enabled) { + err = mlxsw_sp2_ptp_configure_port(mlxsw_sp_port, new_ing_types, + new_egr_types, new_config); + if (err) + goto err_configure_port; + } else if (!new_ing_types && !new_egr_types && ptp_enabled) { + err = mlxsw_sp2_ptp_deconfigure_port(mlxsw_sp_port, new_config); + if (err) + goto err_deconfigure_port; + } + + mlxsw_sp_port->ptp.ing_types = new_ing_types; + mlxsw_sp_port->ptp.egr_types = new_egr_types; + + /* Notify the ioctl caller what we are actually timestamping. */ + config->rx_filter = rx_filter; + mutex_unlock(&ptp_state->lock); + + return 0; + +err_deconfigure_port: +err_configure_port: +err_get_message_types: + mutex_unlock(&ptp_state->lock); + return err; +} + +int mlxsw_sp2_ptp_get_ts_info(struct mlxsw_sp *mlxsw_sp, + struct ethtool_ts_info *info) +{ + info->phc_index = ptp_clock_index(mlxsw_sp->clock->ptp); + + info->so_timestamping = SOF_TIMESTAMPING_TX_HARDWARE | + SOF_TIMESTAMPING_RX_HARDWARE | + SOF_TIMESTAMPING_RAW_HARDWARE; + + info->tx_types = BIT(HWTSTAMP_TX_OFF) | + BIT(HWTSTAMP_TX_ON); + + info->rx_filters = BIT(HWTSTAMP_FILTER_NONE) | + BIT(HWTSTAMP_FILTER_PTP_V1_L4_EVENT) | + BIT(HWTSTAMP_FILTER_PTP_V2_EVENT); + + return 0; +} + +int mlxsw_sp_ptp_txhdr_construct(struct mlxsw_core *mlxsw_core, + struct mlxsw_sp_port *mlxsw_sp_port, + struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info) +{ + mlxsw_sp_txhdr_construct(skb, tx_info); + return 0; +} + +int mlxsw_sp2_ptp_txhdr_construct(struct mlxsw_core *mlxsw_core, + struct mlxsw_sp_port *mlxsw_sp_port, + struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info) +{ + /* In Spectrum-2 and Spectrum-3, in order for PTP event packets to have + * their correction field correctly set on the egress port they must be + * transmitted as data packets. Such packets ingress the ASIC via the + * CPU port and must have a VLAN tag, as the CPU port is not configured + * with a PVID. Push the default VLAN (4095), which is configured as + * egress untagged on all the ports. + */ + if (!skb_vlan_tagged(skb)) { + skb = vlan_insert_tag_set_proto(skb, htons(ETH_P_8021Q), + MLXSW_SP_DEFAULT_VID); + if (!skb) { + this_cpu_inc(mlxsw_sp_port->pcpu_stats->tx_dropped); + return -ENOMEM; + } + } + + return mlxsw_sp_txhdr_ptp_data_construct(mlxsw_core, mlxsw_sp_port, skb, + tx_info); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.h new file mode 100644 index 000000000..a8b882309 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.h @@ -0,0 +1,263 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2019 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_SPECTRUM_PTP_H +#define _MLXSW_SPECTRUM_PTP_H + +#include <linux/device.h> +#include <linux/rhashtable.h> + +struct mlxsw_sp; +struct mlxsw_sp_port; +struct mlxsw_sp_ptp_clock; + +static inline int mlxsw_sp_ptp_get_ts_info_noptp(struct ethtool_ts_info *info) +{ + info->so_timestamping = SOF_TIMESTAMPING_RX_SOFTWARE | + SOF_TIMESTAMPING_SOFTWARE; + info->phc_index = -1; + return 0; +} + +#if IS_REACHABLE(CONFIG_PTP_1588_CLOCK) + +struct mlxsw_sp_ptp_clock * +mlxsw_sp1_ptp_clock_init(struct mlxsw_sp *mlxsw_sp, struct device *dev); + +void mlxsw_sp1_ptp_clock_fini(struct mlxsw_sp_ptp_clock *clock); + +struct mlxsw_sp_ptp_state *mlxsw_sp1_ptp_init(struct mlxsw_sp *mlxsw_sp); + +void mlxsw_sp1_ptp_fini(struct mlxsw_sp_ptp_state *ptp_state); + +void mlxsw_sp1_ptp_receive(struct mlxsw_sp *mlxsw_sp, struct sk_buff *skb, + u16 local_port); + +void mlxsw_sp1_ptp_transmitted(struct mlxsw_sp *mlxsw_sp, + struct sk_buff *skb, u16 local_port); + +void mlxsw_sp1_ptp_got_timestamp(struct mlxsw_sp *mlxsw_sp, bool ingress, + u16 local_port, u8 message_type, + u8 domain_number, u16 sequence_id, + u64 timestamp); + +int mlxsw_sp1_ptp_hwtstamp_get(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config); + +int mlxsw_sp1_ptp_hwtstamp_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config); + +void mlxsw_sp1_ptp_shaper_work(struct work_struct *work); + +int mlxsw_sp1_ptp_get_ts_info(struct mlxsw_sp *mlxsw_sp, + struct ethtool_ts_info *info); + +int mlxsw_sp1_get_stats_count(void); +void mlxsw_sp1_get_stats_strings(u8 **p); +void mlxsw_sp1_get_stats(struct mlxsw_sp_port *mlxsw_sp_port, + u64 *data, int data_index); + +int mlxsw_sp_ptp_txhdr_construct(struct mlxsw_core *mlxsw_core, + struct mlxsw_sp_port *mlxsw_sp_port, + struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info); + +struct mlxsw_sp_ptp_clock * +mlxsw_sp2_ptp_clock_init(struct mlxsw_sp *mlxsw_sp, struct device *dev); + +void mlxsw_sp2_ptp_clock_fini(struct mlxsw_sp_ptp_clock *clock); + +struct mlxsw_sp_ptp_state *mlxsw_sp2_ptp_init(struct mlxsw_sp *mlxsw_sp); + +void mlxsw_sp2_ptp_fini(struct mlxsw_sp_ptp_state *ptp_state); + +void mlxsw_sp2_ptp_receive(struct mlxsw_sp *mlxsw_sp, struct sk_buff *skb, + u16 local_port); + +void mlxsw_sp2_ptp_transmitted(struct mlxsw_sp *mlxsw_sp, + struct sk_buff *skb, u16 local_port); + +int mlxsw_sp2_ptp_hwtstamp_get(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config); + +int mlxsw_sp2_ptp_hwtstamp_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config); + +int mlxsw_sp2_ptp_get_ts_info(struct mlxsw_sp *mlxsw_sp, + struct ethtool_ts_info *info); + +int mlxsw_sp2_ptp_txhdr_construct(struct mlxsw_core *mlxsw_core, + struct mlxsw_sp_port *mlxsw_sp_port, + struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info); + +#else + +static inline struct mlxsw_sp_ptp_clock * +mlxsw_sp1_ptp_clock_init(struct mlxsw_sp *mlxsw_sp, struct device *dev) +{ + return NULL; +} + +static inline void mlxsw_sp1_ptp_clock_fini(struct mlxsw_sp_ptp_clock *clock) +{ +} + +static inline struct mlxsw_sp_ptp_state * +mlxsw_sp1_ptp_init(struct mlxsw_sp *mlxsw_sp) +{ + return NULL; +} + +static inline void mlxsw_sp1_ptp_fini(struct mlxsw_sp_ptp_state *ptp_state) +{ +} + +static inline void mlxsw_sp1_ptp_receive(struct mlxsw_sp *mlxsw_sp, + struct sk_buff *skb, u16 local_port) +{ + mlxsw_sp_rx_listener_no_mark_func(skb, local_port, mlxsw_sp); +} + +static inline void mlxsw_sp1_ptp_transmitted(struct mlxsw_sp *mlxsw_sp, + struct sk_buff *skb, u16 local_port) +{ + dev_kfree_skb_any(skb); +} + +static inline void +mlxsw_sp1_ptp_got_timestamp(struct mlxsw_sp *mlxsw_sp, bool ingress, + u16 local_port, u8 message_type, + u8 domain_number, + u16 sequence_id, u64 timestamp) +{ +} + +static inline int +mlxsw_sp1_ptp_hwtstamp_get(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config) +{ + return -EOPNOTSUPP; +} + +static inline int +mlxsw_sp1_ptp_hwtstamp_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config) +{ + return -EOPNOTSUPP; +} + +static inline void mlxsw_sp1_ptp_shaper_work(struct work_struct *work) +{ +} + +static inline int mlxsw_sp1_ptp_get_ts_info(struct mlxsw_sp *mlxsw_sp, + struct ethtool_ts_info *info) +{ + return mlxsw_sp_ptp_get_ts_info_noptp(info); +} + +static inline int mlxsw_sp1_get_stats_count(void) +{ + return 0; +} + +static inline void mlxsw_sp1_get_stats_strings(u8 **p) +{ +} + +static inline void mlxsw_sp1_get_stats(struct mlxsw_sp_port *mlxsw_sp_port, + u64 *data, int data_index) +{ +} + +static inline int +mlxsw_sp_ptp_txhdr_construct(struct mlxsw_core *mlxsw_core, + struct mlxsw_sp_port *mlxsw_sp_port, + struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info) +{ + return -EOPNOTSUPP; +} + +static inline struct mlxsw_sp_ptp_clock * +mlxsw_sp2_ptp_clock_init(struct mlxsw_sp *mlxsw_sp, struct device *dev) +{ + return NULL; +} + +static inline void mlxsw_sp2_ptp_clock_fini(struct mlxsw_sp_ptp_clock *clock) +{ +} + +static inline struct mlxsw_sp_ptp_state * +mlxsw_sp2_ptp_init(struct mlxsw_sp *mlxsw_sp) +{ + return NULL; +} + +static inline void mlxsw_sp2_ptp_fini(struct mlxsw_sp_ptp_state *ptp_state) +{ +} + +static inline void mlxsw_sp2_ptp_receive(struct mlxsw_sp *mlxsw_sp, + struct sk_buff *skb, u16 local_port) +{ + mlxsw_sp_rx_listener_no_mark_func(skb, local_port, mlxsw_sp); +} + +static inline void mlxsw_sp2_ptp_transmitted(struct mlxsw_sp *mlxsw_sp, + struct sk_buff *skb, u16 local_port) +{ + dev_kfree_skb_any(skb); +} + +static inline int +mlxsw_sp2_ptp_hwtstamp_get(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config) +{ + return -EOPNOTSUPP; +} + +static inline int +mlxsw_sp2_ptp_hwtstamp_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct hwtstamp_config *config) +{ + return -EOPNOTSUPP; +} + +static inline int mlxsw_sp2_ptp_get_ts_info(struct mlxsw_sp *mlxsw_sp, + struct ethtool_ts_info *info) +{ + return mlxsw_sp_ptp_get_ts_info_noptp(info); +} + +static inline int +mlxsw_sp2_ptp_txhdr_construct(struct mlxsw_core *mlxsw_core, + struct mlxsw_sp_port *mlxsw_sp_port, + struct sk_buff *skb, + const struct mlxsw_tx_info *tx_info) +{ + return -EOPNOTSUPP; +} +#endif + +static inline void mlxsw_sp2_ptp_shaper_work(struct work_struct *work) +{ +} + +static inline int mlxsw_sp2_get_stats_count(void) +{ + return 0; +} + +static inline void mlxsw_sp2_get_stats_strings(u8 **p) +{ +} + +static inline void mlxsw_sp2_get_stats(struct mlxsw_sp_port *mlxsw_sp_port, + u64 *data, int data_index) +{ +} + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c new file mode 100644 index 000000000..4243d3b88 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c @@ -0,0 +1,2337 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/netdevice.h> +#include <net/pkt_cls.h> +#include <net/red.h> + +#include "spectrum.h" +#include "spectrum_span.h" +#include "reg.h" + +#define MLXSW_SP_PRIO_BAND_TO_TCLASS(band) (IEEE_8021QAZ_MAX_TCS - band - 1) +#define MLXSW_SP_PRIO_CHILD_TO_TCLASS(child) \ + MLXSW_SP_PRIO_BAND_TO_TCLASS((child - 1)) + +enum mlxsw_sp_qdisc_type { + MLXSW_SP_QDISC_NO_QDISC, + MLXSW_SP_QDISC_RED, + MLXSW_SP_QDISC_PRIO, + MLXSW_SP_QDISC_ETS, + MLXSW_SP_QDISC_TBF, + MLXSW_SP_QDISC_FIFO, +}; + +struct mlxsw_sp_qdisc; + +struct mlxsw_sp_qdisc_ops { + enum mlxsw_sp_qdisc_type type; + int (*check_params)(struct mlxsw_sp_port *mlxsw_sp_port, + void *params); + int (*replace)(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, void *params); + int (*destroy)(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc); + int (*get_stats)(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_qopt_offload_stats *stats_ptr); + int (*get_xstats)(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *xstats_ptr); + void (*clean_stats)(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc); + /* unoffload - to be used for a qdisc that stops being offloaded without + * being destroyed. + */ + void (*unoffload)(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, void *params); + struct mlxsw_sp_qdisc *(*find_class)(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + u32 parent); + unsigned int num_classes; + + u8 (*get_prio_bitmap)(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct mlxsw_sp_qdisc *child); + int (*get_tclass_num)(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct mlxsw_sp_qdisc *child); +}; + +struct mlxsw_sp_qdisc_ets_band { + u8 prio_bitmap; + int tclass_num; +}; + +struct mlxsw_sp_qdisc_ets_data { + struct mlxsw_sp_qdisc_ets_band bands[IEEE_8021QAZ_MAX_TCS]; +}; + +struct mlxsw_sp_qdisc { + u32 handle; + union { + struct red_stats red; + } xstats_base; + struct mlxsw_sp_qdisc_stats { + u64 tx_bytes; + u64 tx_packets; + u64 drops; + u64 overlimits; + u64 backlog; + } stats_base; + + union { + struct mlxsw_sp_qdisc_ets_data *ets_data; + }; + + struct mlxsw_sp_qdisc_ops *ops; + struct mlxsw_sp_qdisc *parent; + struct mlxsw_sp_qdisc *qdiscs; + unsigned int num_classes; +}; + +struct mlxsw_sp_qdisc_state { + struct mlxsw_sp_qdisc root_qdisc; + + /* When a PRIO or ETS are added, the invisible FIFOs in their bands are + * created first. When notifications for these FIFOs arrive, it is not + * known what qdisc their parent handle refers to. It could be a + * newly-created PRIO that will replace the currently-offloaded one, or + * it could be e.g. a RED that will be attached below it. + * + * As the notifications start to arrive, use them to note what the + * future parent handle is, and keep track of which child FIFOs were + * seen. Then when the parent is known, retroactively offload those + * FIFOs. + */ + u32 future_handle; + bool future_fifos[IEEE_8021QAZ_MAX_TCS]; + struct mutex lock; /* Protects qdisc state. */ +}; + +static bool +mlxsw_sp_qdisc_compare(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, u32 handle) +{ + return mlxsw_sp_qdisc->ops && mlxsw_sp_qdisc->handle == handle; +} + +static struct mlxsw_sp_qdisc * +mlxsw_sp_qdisc_walk(struct mlxsw_sp_qdisc *qdisc, + struct mlxsw_sp_qdisc *(*pre)(struct mlxsw_sp_qdisc *, + void *), + void *data) +{ + struct mlxsw_sp_qdisc *tmp; + unsigned int i; + + if (pre) { + tmp = pre(qdisc, data); + if (tmp) + return tmp; + } + + if (qdisc->ops) { + for (i = 0; i < qdisc->num_classes; i++) { + tmp = &qdisc->qdiscs[i]; + if (qdisc->ops) { + tmp = mlxsw_sp_qdisc_walk(tmp, pre, data); + if (tmp) + return tmp; + } + } + } + + return NULL; +} + +static struct mlxsw_sp_qdisc * +mlxsw_sp_qdisc_walk_cb_find(struct mlxsw_sp_qdisc *qdisc, void *data) +{ + u32 parent = *(u32 *)data; + + if (qdisc->ops && TC_H_MAJ(qdisc->handle) == TC_H_MAJ(parent)) { + if (qdisc->ops->find_class) + return qdisc->ops->find_class(qdisc, parent); + } + + return NULL; +} + +static struct mlxsw_sp_qdisc * +mlxsw_sp_qdisc_find(struct mlxsw_sp_port *mlxsw_sp_port, u32 parent) +{ + struct mlxsw_sp_qdisc_state *qdisc_state = mlxsw_sp_port->qdisc; + + if (!qdisc_state) + return NULL; + if (parent == TC_H_ROOT) + return &qdisc_state->root_qdisc; + return mlxsw_sp_qdisc_walk(&qdisc_state->root_qdisc, + mlxsw_sp_qdisc_walk_cb_find, &parent); +} + +static struct mlxsw_sp_qdisc * +mlxsw_sp_qdisc_walk_cb_find_by_handle(struct mlxsw_sp_qdisc *qdisc, void *data) +{ + u32 handle = *(u32 *)data; + + if (qdisc->ops && qdisc->handle == handle) + return qdisc; + return NULL; +} + +static struct mlxsw_sp_qdisc * +mlxsw_sp_qdisc_find_by_handle(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle) +{ + struct mlxsw_sp_qdisc_state *qdisc_state = mlxsw_sp_port->qdisc; + + if (!qdisc_state) + return NULL; + return mlxsw_sp_qdisc_walk(&qdisc_state->root_qdisc, + mlxsw_sp_qdisc_walk_cb_find_by_handle, + &handle); +} + +static void +mlxsw_sp_qdisc_reduce_parent_backlog(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + struct mlxsw_sp_qdisc *tmp; + + for (tmp = mlxsw_sp_qdisc->parent; tmp; tmp = tmp->parent) + tmp->stats_base.backlog -= mlxsw_sp_qdisc->stats_base.backlog; +} + +static u8 mlxsw_sp_qdisc_get_prio_bitmap(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + struct mlxsw_sp_qdisc *parent = mlxsw_sp_qdisc->parent; + + if (!parent) + return 0xff; + if (!parent->ops->get_prio_bitmap) + return mlxsw_sp_qdisc_get_prio_bitmap(mlxsw_sp_port, parent); + return parent->ops->get_prio_bitmap(parent, mlxsw_sp_qdisc); +} + +#define MLXSW_SP_PORT_DEFAULT_TCLASS 0 + +static int mlxsw_sp_qdisc_get_tclass_num(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + struct mlxsw_sp_qdisc *parent = mlxsw_sp_qdisc->parent; + + if (!parent) + return MLXSW_SP_PORT_DEFAULT_TCLASS; + if (!parent->ops->get_tclass_num) + return mlxsw_sp_qdisc_get_tclass_num(mlxsw_sp_port, parent); + return parent->ops->get_tclass_num(parent, mlxsw_sp_qdisc); +} + +static int +mlxsw_sp_qdisc_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + struct mlxsw_sp_qdisc *root_qdisc = &mlxsw_sp_port->qdisc->root_qdisc; + int err_hdroom = 0; + int err = 0; + int i; + + if (!mlxsw_sp_qdisc) + return 0; + + if (root_qdisc == mlxsw_sp_qdisc) { + struct mlxsw_sp_hdroom hdroom = *mlxsw_sp_port->hdroom; + + hdroom.mode = MLXSW_SP_HDROOM_MODE_DCB; + mlxsw_sp_hdroom_prios_reset_buf_idx(&hdroom); + mlxsw_sp_hdroom_bufs_reset_lossiness(&hdroom); + mlxsw_sp_hdroom_bufs_reset_sizes(mlxsw_sp_port, &hdroom); + err_hdroom = mlxsw_sp_hdroom_configure(mlxsw_sp_port, &hdroom); + } + + if (!mlxsw_sp_qdisc->ops) + return 0; + + for (i = 0; i < mlxsw_sp_qdisc->num_classes; i++) + mlxsw_sp_qdisc_destroy(mlxsw_sp_port, + &mlxsw_sp_qdisc->qdiscs[i]); + mlxsw_sp_qdisc_reduce_parent_backlog(mlxsw_sp_qdisc); + if (mlxsw_sp_qdisc->ops->destroy) + err = mlxsw_sp_qdisc->ops->destroy(mlxsw_sp_port, + mlxsw_sp_qdisc); + if (mlxsw_sp_qdisc->ops->clean_stats) + mlxsw_sp_qdisc->ops->clean_stats(mlxsw_sp_port, mlxsw_sp_qdisc); + + mlxsw_sp_qdisc->handle = TC_H_UNSPEC; + mlxsw_sp_qdisc->ops = NULL; + mlxsw_sp_qdisc->num_classes = 0; + kfree(mlxsw_sp_qdisc->qdiscs); + mlxsw_sp_qdisc->qdiscs = NULL; + return err_hdroom ?: err; +} + +struct mlxsw_sp_qdisc_tree_validate { + bool forbid_ets; + bool forbid_root_tbf; + bool forbid_tbf; + bool forbid_red; +}; + +static int +__mlxsw_sp_qdisc_tree_validate(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct mlxsw_sp_qdisc_tree_validate validate); + +static int +mlxsw_sp_qdisc_tree_validate_children(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct mlxsw_sp_qdisc_tree_validate validate) +{ + unsigned int i; + int err; + + for (i = 0; i < mlxsw_sp_qdisc->num_classes; i++) { + err = __mlxsw_sp_qdisc_tree_validate(&mlxsw_sp_qdisc->qdiscs[i], + validate); + if (err) + return err; + } + + return 0; +} + +static int +__mlxsw_sp_qdisc_tree_validate(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct mlxsw_sp_qdisc_tree_validate validate) +{ + if (!mlxsw_sp_qdisc->ops) + return 0; + + switch (mlxsw_sp_qdisc->ops->type) { + case MLXSW_SP_QDISC_FIFO: + break; + case MLXSW_SP_QDISC_RED: + if (validate.forbid_red) + return -EINVAL; + validate.forbid_red = true; + validate.forbid_root_tbf = true; + validate.forbid_ets = true; + break; + case MLXSW_SP_QDISC_TBF: + if (validate.forbid_root_tbf) { + if (validate.forbid_tbf) + return -EINVAL; + /* This is a TC TBF. */ + validate.forbid_tbf = true; + validate.forbid_ets = true; + } else { + /* This is root TBF. */ + validate.forbid_root_tbf = true; + } + break; + case MLXSW_SP_QDISC_PRIO: + case MLXSW_SP_QDISC_ETS: + if (validate.forbid_ets) + return -EINVAL; + validate.forbid_root_tbf = true; + validate.forbid_ets = true; + break; + default: + WARN_ON(1); + return -EINVAL; + } + + return mlxsw_sp_qdisc_tree_validate_children(mlxsw_sp_qdisc, validate); +} + +static int mlxsw_sp_qdisc_tree_validate(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp_qdisc_tree_validate validate = {}; + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc; + + mlxsw_sp_qdisc = &mlxsw_sp_port->qdisc->root_qdisc; + return __mlxsw_sp_qdisc_tree_validate(mlxsw_sp_qdisc, validate); +} + +static int mlxsw_sp_qdisc_create(struct mlxsw_sp_port *mlxsw_sp_port, + u32 handle, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct mlxsw_sp_qdisc_ops *ops, void *params) +{ + struct mlxsw_sp_qdisc *root_qdisc = &mlxsw_sp_port->qdisc->root_qdisc; + struct mlxsw_sp_hdroom orig_hdroom; + unsigned int i; + int err; + + err = ops->check_params(mlxsw_sp_port, params); + if (err) + return err; + + if (ops->num_classes) { + mlxsw_sp_qdisc->qdiscs = kcalloc(ops->num_classes, + sizeof(*mlxsw_sp_qdisc->qdiscs), + GFP_KERNEL); + if (!mlxsw_sp_qdisc->qdiscs) + return -ENOMEM; + + for (i = 0; i < ops->num_classes; i++) + mlxsw_sp_qdisc->qdiscs[i].parent = mlxsw_sp_qdisc; + } + + orig_hdroom = *mlxsw_sp_port->hdroom; + if (root_qdisc == mlxsw_sp_qdisc) { + struct mlxsw_sp_hdroom hdroom = orig_hdroom; + + hdroom.mode = MLXSW_SP_HDROOM_MODE_TC; + mlxsw_sp_hdroom_prios_reset_buf_idx(&hdroom); + mlxsw_sp_hdroom_bufs_reset_lossiness(&hdroom); + mlxsw_sp_hdroom_bufs_reset_sizes(mlxsw_sp_port, &hdroom); + + err = mlxsw_sp_hdroom_configure(mlxsw_sp_port, &hdroom); + if (err) + goto err_hdroom_configure; + } + + mlxsw_sp_qdisc->num_classes = ops->num_classes; + mlxsw_sp_qdisc->ops = ops; + mlxsw_sp_qdisc->handle = handle; + err = mlxsw_sp_qdisc_tree_validate(mlxsw_sp_port); + if (err) + goto err_replace; + + err = ops->replace(mlxsw_sp_port, handle, mlxsw_sp_qdisc, params); + if (err) + goto err_replace; + + return 0; + +err_replace: + mlxsw_sp_qdisc->handle = TC_H_UNSPEC; + mlxsw_sp_qdisc->ops = NULL; + mlxsw_sp_qdisc->num_classes = 0; + mlxsw_sp_hdroom_configure(mlxsw_sp_port, &orig_hdroom); +err_hdroom_configure: + kfree(mlxsw_sp_qdisc->qdiscs); + mlxsw_sp_qdisc->qdiscs = NULL; + return err; +} + +static int +mlxsw_sp_qdisc_change(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, void *params) +{ + struct mlxsw_sp_qdisc_ops *ops = mlxsw_sp_qdisc->ops; + int err; + + err = ops->check_params(mlxsw_sp_port, params); + if (err) + goto unoffload; + + err = ops->replace(mlxsw_sp_port, handle, mlxsw_sp_qdisc, params); + if (err) + goto unoffload; + + /* Check if the Qdisc changed. That includes a situation where an + * invisible Qdisc replaces another one, or is being added for the + * first time. + */ + if (mlxsw_sp_qdisc->handle != handle) { + if (ops->clean_stats) + ops->clean_stats(mlxsw_sp_port, mlxsw_sp_qdisc); + } + + mlxsw_sp_qdisc->handle = handle; + return 0; + +unoffload: + if (ops->unoffload) + ops->unoffload(mlxsw_sp_port, mlxsw_sp_qdisc, params); + + mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + return err; +} + +static int +mlxsw_sp_qdisc_replace(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct mlxsw_sp_qdisc_ops *ops, void *params) +{ + if (mlxsw_sp_qdisc->ops && mlxsw_sp_qdisc->ops->type != ops->type) + /* In case this location contained a different qdisc of the + * same type we can override the old qdisc configuration. + * Otherwise, we need to remove the old qdisc before setting the + * new one. + */ + mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + + if (!mlxsw_sp_qdisc->ops) + return mlxsw_sp_qdisc_create(mlxsw_sp_port, handle, + mlxsw_sp_qdisc, ops, params); + else + return mlxsw_sp_qdisc_change(mlxsw_sp_port, handle, + mlxsw_sp_qdisc, params); +} + +static int +mlxsw_sp_qdisc_get_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_qopt_offload_stats *stats_ptr) +{ + if (mlxsw_sp_qdisc && mlxsw_sp_qdisc->ops && + mlxsw_sp_qdisc->ops->get_stats) + return mlxsw_sp_qdisc->ops->get_stats(mlxsw_sp_port, + mlxsw_sp_qdisc, + stats_ptr); + + return -EOPNOTSUPP; +} + +static int +mlxsw_sp_qdisc_get_xstats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *xstats_ptr) +{ + if (mlxsw_sp_qdisc && mlxsw_sp_qdisc->ops && + mlxsw_sp_qdisc->ops->get_xstats) + return mlxsw_sp_qdisc->ops->get_xstats(mlxsw_sp_port, + mlxsw_sp_qdisc, + xstats_ptr); + + return -EOPNOTSUPP; +} + +static u64 +mlxsw_sp_xstats_backlog(struct mlxsw_sp_port_xstats *xstats, int tclass_num) +{ + return xstats->backlog[tclass_num] + + xstats->backlog[tclass_num + 8]; +} + +static u64 +mlxsw_sp_xstats_tail_drop(struct mlxsw_sp_port_xstats *xstats, int tclass_num) +{ + return xstats->tail_drop[tclass_num] + + xstats->tail_drop[tclass_num + 8]; +} + +static void +mlxsw_sp_qdisc_bstats_per_priority_get(struct mlxsw_sp_port_xstats *xstats, + u8 prio_bitmap, u64 *tx_packets, + u64 *tx_bytes) +{ + int i; + + *tx_packets = 0; + *tx_bytes = 0; + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + if (prio_bitmap & BIT(i)) { + *tx_packets += xstats->tx_packets[i]; + *tx_bytes += xstats->tx_bytes[i]; + } + } +} + +static void +mlxsw_sp_qdisc_collect_tc_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + u64 *p_tx_bytes, u64 *p_tx_packets, + u64 *p_drops, u64 *p_backlog) +{ + struct mlxsw_sp_port_xstats *xstats; + u64 tx_bytes, tx_packets; + u8 prio_bitmap; + int tclass_num; + + prio_bitmap = mlxsw_sp_qdisc_get_prio_bitmap(mlxsw_sp_port, + mlxsw_sp_qdisc); + tclass_num = mlxsw_sp_qdisc_get_tclass_num(mlxsw_sp_port, + mlxsw_sp_qdisc); + xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; + mlxsw_sp_qdisc_bstats_per_priority_get(xstats, prio_bitmap, + &tx_packets, &tx_bytes); + + *p_tx_packets += tx_packets; + *p_tx_bytes += tx_bytes; + *p_drops += xstats->wred_drop[tclass_num] + + mlxsw_sp_xstats_tail_drop(xstats, tclass_num); + *p_backlog += mlxsw_sp_xstats_backlog(xstats, tclass_num); +} + +static void +mlxsw_sp_qdisc_update_stats(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + u64 tx_bytes, u64 tx_packets, + u64 drops, u64 backlog, + struct tc_qopt_offload_stats *stats_ptr) +{ + struct mlxsw_sp_qdisc_stats *stats_base = &mlxsw_sp_qdisc->stats_base; + + tx_bytes -= stats_base->tx_bytes; + tx_packets -= stats_base->tx_packets; + drops -= stats_base->drops; + backlog -= stats_base->backlog; + + _bstats_update(stats_ptr->bstats, tx_bytes, tx_packets); + stats_ptr->qstats->drops += drops; + stats_ptr->qstats->backlog += mlxsw_sp_cells_bytes(mlxsw_sp, backlog); + + stats_base->backlog += backlog; + stats_base->drops += drops; + stats_base->tx_bytes += tx_bytes; + stats_base->tx_packets += tx_packets; +} + +static void +mlxsw_sp_qdisc_get_tc_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_qopt_offload_stats *stats_ptr) +{ + u64 tx_packets = 0; + u64 tx_bytes = 0; + u64 backlog = 0; + u64 drops = 0; + + mlxsw_sp_qdisc_collect_tc_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &tx_bytes, &tx_packets, + &drops, &backlog); + mlxsw_sp_qdisc_update_stats(mlxsw_sp_port->mlxsw_sp, mlxsw_sp_qdisc, + tx_bytes, tx_packets, drops, backlog, + stats_ptr); +} + +static int +mlxsw_sp_tclass_congestion_enable(struct mlxsw_sp_port *mlxsw_sp_port, + int tclass_num, u32 min, u32 max, + u32 probability, bool is_wred, bool is_ecn) +{ + char cwtpm_cmd[MLXSW_REG_CWTPM_LEN]; + char cwtp_cmd[MLXSW_REG_CWTP_LEN]; + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + int err; + + mlxsw_reg_cwtp_pack(cwtp_cmd, mlxsw_sp_port->local_port, tclass_num); + mlxsw_reg_cwtp_profile_pack(cwtp_cmd, MLXSW_REG_CWTP_DEFAULT_PROFILE, + roundup(min, MLXSW_REG_CWTP_MIN_VALUE), + roundup(max, MLXSW_REG_CWTP_MIN_VALUE), + probability); + + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(cwtp), cwtp_cmd); + if (err) + return err; + + mlxsw_reg_cwtpm_pack(cwtpm_cmd, mlxsw_sp_port->local_port, tclass_num, + MLXSW_REG_CWTP_DEFAULT_PROFILE, is_wred, is_ecn); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(cwtpm), cwtpm_cmd); +} + +static int +mlxsw_sp_tclass_congestion_disable(struct mlxsw_sp_port *mlxsw_sp_port, + int tclass_num) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char cwtpm_cmd[MLXSW_REG_CWTPM_LEN]; + + mlxsw_reg_cwtpm_pack(cwtpm_cmd, mlxsw_sp_port->local_port, tclass_num, + MLXSW_REG_CWTPM_RESET_PROFILE, false, false); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(cwtpm), cwtpm_cmd); +} + +static void +mlxsw_sp_setup_tc_qdisc_red_clean_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + struct mlxsw_sp_qdisc_stats *stats_base; + struct mlxsw_sp_port_xstats *xstats; + struct red_stats *red_base; + u8 prio_bitmap; + int tclass_num; + + prio_bitmap = mlxsw_sp_qdisc_get_prio_bitmap(mlxsw_sp_port, + mlxsw_sp_qdisc); + tclass_num = mlxsw_sp_qdisc_get_tclass_num(mlxsw_sp_port, + mlxsw_sp_qdisc); + xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; + stats_base = &mlxsw_sp_qdisc->stats_base; + red_base = &mlxsw_sp_qdisc->xstats_base.red; + + mlxsw_sp_qdisc_bstats_per_priority_get(xstats, prio_bitmap, + &stats_base->tx_packets, + &stats_base->tx_bytes); + red_base->prob_mark = xstats->tc_ecn[tclass_num]; + red_base->prob_drop = xstats->wred_drop[tclass_num]; + red_base->pdrop = mlxsw_sp_xstats_tail_drop(xstats, tclass_num); + + stats_base->overlimits = red_base->prob_drop + red_base->prob_mark; + stats_base->drops = red_base->prob_drop + red_base->pdrop; + + stats_base->backlog = 0; +} + +static int +mlxsw_sp_qdisc_red_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + int tclass_num = mlxsw_sp_qdisc_get_tclass_num(mlxsw_sp_port, + mlxsw_sp_qdisc); + + return mlxsw_sp_tclass_congestion_disable(mlxsw_sp_port, tclass_num); +} + +static int +mlxsw_sp_qdisc_red_check_params(struct mlxsw_sp_port *mlxsw_sp_port, + void *params) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct tc_red_qopt_offload_params *p = params; + + if (p->min > p->max) { + dev_err(mlxsw_sp->bus_info->dev, + "spectrum: RED: min %u is bigger then max %u\n", p->min, + p->max); + return -EINVAL; + } + if (p->max > MLXSW_CORE_RES_GET(mlxsw_sp->core, + GUARANTEED_SHARED_BUFFER)) { + dev_err(mlxsw_sp->bus_info->dev, + "spectrum: RED: max value %u is too big\n", p->max); + return -EINVAL; + } + if (p->min == 0 || p->max == 0) { + dev_err(mlxsw_sp->bus_info->dev, + "spectrum: RED: 0 value is illegal for min and max\n"); + return -EINVAL; + } + return 0; +} + +static int +mlxsw_sp_qdisc_future_fifo_replace(struct mlxsw_sp_port *mlxsw_sp_port, + u32 handle, unsigned int band, + struct mlxsw_sp_qdisc *child_qdisc); +static void +mlxsw_sp_qdisc_future_fifos_init(struct mlxsw_sp_port *mlxsw_sp_port, + u32 handle); + +static int +mlxsw_sp_qdisc_red_replace(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct tc_red_qopt_offload_params *p = params; + int tclass_num; + u32 min, max; + u64 prob; + int err; + + err = mlxsw_sp_qdisc_future_fifo_replace(mlxsw_sp_port, handle, 0, + &mlxsw_sp_qdisc->qdiscs[0]); + if (err) + return err; + mlxsw_sp_qdisc_future_fifos_init(mlxsw_sp_port, TC_H_UNSPEC); + + tclass_num = mlxsw_sp_qdisc_get_tclass_num(mlxsw_sp_port, + mlxsw_sp_qdisc); + + /* calculate probability in percentage */ + prob = p->probability; + prob *= 100; + prob = DIV_ROUND_UP(prob, 1 << 16); + prob = DIV_ROUND_UP(prob, 1 << 16); + min = mlxsw_sp_bytes_cells(mlxsw_sp, p->min); + max = mlxsw_sp_bytes_cells(mlxsw_sp, p->max); + return mlxsw_sp_tclass_congestion_enable(mlxsw_sp_port, tclass_num, + min, max, prob, + !p->is_nodrop, p->is_ecn); +} + +static void +mlxsw_sp_qdisc_leaf_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct gnet_stats_queue *qstats) +{ + u64 backlog; + + backlog = mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, + mlxsw_sp_qdisc->stats_base.backlog); + qstats->backlog -= backlog; + mlxsw_sp_qdisc->stats_base.backlog = 0; +} + +static void +mlxsw_sp_qdisc_red_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_red_qopt_offload_params *p = params; + + mlxsw_sp_qdisc_leaf_unoffload(mlxsw_sp_port, mlxsw_sp_qdisc, p->qstats); +} + +static int +mlxsw_sp_qdisc_get_red_xstats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *xstats_ptr) +{ + struct red_stats *xstats_base = &mlxsw_sp_qdisc->xstats_base.red; + struct mlxsw_sp_port_xstats *xstats; + struct red_stats *res = xstats_ptr; + int early_drops, marks, pdrops; + int tclass_num; + + tclass_num = mlxsw_sp_qdisc_get_tclass_num(mlxsw_sp_port, + mlxsw_sp_qdisc); + xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; + + early_drops = xstats->wred_drop[tclass_num] - xstats_base->prob_drop; + marks = xstats->tc_ecn[tclass_num] - xstats_base->prob_mark; + pdrops = mlxsw_sp_xstats_tail_drop(xstats, tclass_num) - + xstats_base->pdrop; + + res->pdrop += pdrops; + res->prob_drop += early_drops; + res->prob_mark += marks; + + xstats_base->pdrop += pdrops; + xstats_base->prob_drop += early_drops; + xstats_base->prob_mark += marks; + return 0; +} + +static int +mlxsw_sp_qdisc_get_red_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_qopt_offload_stats *stats_ptr) +{ + struct mlxsw_sp_qdisc_stats *stats_base; + struct mlxsw_sp_port_xstats *xstats; + u64 overlimits; + int tclass_num; + + tclass_num = mlxsw_sp_qdisc_get_tclass_num(mlxsw_sp_port, + mlxsw_sp_qdisc); + xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; + stats_base = &mlxsw_sp_qdisc->stats_base; + + mlxsw_sp_qdisc_get_tc_stats(mlxsw_sp_port, mlxsw_sp_qdisc, stats_ptr); + overlimits = xstats->wred_drop[tclass_num] + + xstats->tc_ecn[tclass_num] - stats_base->overlimits; + + stats_ptr->qstats->overlimits += overlimits; + stats_base->overlimits += overlimits; + + return 0; +} + +static struct mlxsw_sp_qdisc * +mlxsw_sp_qdisc_leaf_find_class(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + u32 parent) +{ + /* RED and TBF are formally classful qdiscs, but all class references, + * including X:0, just refer to the same one class. + */ + return &mlxsw_sp_qdisc->qdiscs[0]; +} + +static struct mlxsw_sp_qdisc_ops mlxsw_sp_qdisc_ops_red = { + .type = MLXSW_SP_QDISC_RED, + .check_params = mlxsw_sp_qdisc_red_check_params, + .replace = mlxsw_sp_qdisc_red_replace, + .unoffload = mlxsw_sp_qdisc_red_unoffload, + .destroy = mlxsw_sp_qdisc_red_destroy, + .get_stats = mlxsw_sp_qdisc_get_red_stats, + .get_xstats = mlxsw_sp_qdisc_get_red_xstats, + .clean_stats = mlxsw_sp_setup_tc_qdisc_red_clean_stats, + .find_class = mlxsw_sp_qdisc_leaf_find_class, + .num_classes = 1, +}; + +static int mlxsw_sp_qdisc_graft(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + u8 band, u32 child_handle); + +static int __mlxsw_sp_setup_tc_red(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_red_qopt_offload *p) +{ + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc; + + mlxsw_sp_qdisc = mlxsw_sp_qdisc_find(mlxsw_sp_port, p->parent); + if (!mlxsw_sp_qdisc) + return -EOPNOTSUPP; + + if (p->command == TC_RED_REPLACE) + return mlxsw_sp_qdisc_replace(mlxsw_sp_port, p->handle, + mlxsw_sp_qdisc, + &mlxsw_sp_qdisc_ops_red, + &p->set); + + if (!mlxsw_sp_qdisc_compare(mlxsw_sp_qdisc, p->handle)) + return -EOPNOTSUPP; + + switch (p->command) { + case TC_RED_DESTROY: + return mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + case TC_RED_XSTATS: + return mlxsw_sp_qdisc_get_xstats(mlxsw_sp_port, mlxsw_sp_qdisc, + p->xstats); + case TC_RED_STATS: + return mlxsw_sp_qdisc_get_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &p->stats); + case TC_RED_GRAFT: + return mlxsw_sp_qdisc_graft(mlxsw_sp_port, mlxsw_sp_qdisc, 0, + p->child_handle); + default: + return -EOPNOTSUPP; + } +} + +int mlxsw_sp_setup_tc_red(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_red_qopt_offload *p) +{ + int err; + + mutex_lock(&mlxsw_sp_port->qdisc->lock); + err = __mlxsw_sp_setup_tc_red(mlxsw_sp_port, p); + mutex_unlock(&mlxsw_sp_port->qdisc->lock); + + return err; +} + +static void +mlxsw_sp_setup_tc_qdisc_leaf_clean_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + u64 backlog_cells = 0; + u64 tx_packets = 0; + u64 tx_bytes = 0; + u64 drops = 0; + + mlxsw_sp_qdisc_collect_tc_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &tx_bytes, &tx_packets, + &drops, &backlog_cells); + + mlxsw_sp_qdisc->stats_base.tx_packets = tx_packets; + mlxsw_sp_qdisc->stats_base.tx_bytes = tx_bytes; + mlxsw_sp_qdisc->stats_base.drops = drops; + mlxsw_sp_qdisc->stats_base.backlog = 0; +} + +static enum mlxsw_reg_qeec_hr +mlxsw_sp_qdisc_tbf_hr(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + if (mlxsw_sp_qdisc == &mlxsw_sp_port->qdisc->root_qdisc) + return MLXSW_REG_QEEC_HR_PORT; + + /* Configure subgroup shaper, so that both UC and MC traffic is subject + * to shaping. That is unlike RED, however UC queue lengths are going to + * be different than MC ones due to different pool and quota + * configurations, so the configuration is not applicable. For shaper on + * the other hand, subjecting the overall stream to the configured + * shaper makes sense. Also note that that is what we do for + * ieee_setmaxrate(). + */ + return MLXSW_REG_QEEC_HR_SUBGROUP; +} + +static int +mlxsw_sp_qdisc_tbf_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + enum mlxsw_reg_qeec_hr hr = mlxsw_sp_qdisc_tbf_hr(mlxsw_sp_port, + mlxsw_sp_qdisc); + int tclass_num = mlxsw_sp_qdisc_get_tclass_num(mlxsw_sp_port, + mlxsw_sp_qdisc); + + return mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, hr, tclass_num, 0, + MLXSW_REG_QEEC_MAS_DIS, 0); +} + +static int +mlxsw_sp_qdisc_tbf_bs(struct mlxsw_sp_port *mlxsw_sp_port, + u32 max_size, u8 *p_burst_size) +{ + /* TBF burst size is configured in bytes. The ASIC burst size value is + * ((2 ^ bs) * 512 bits. Convert the TBF bytes to 512-bit units. + */ + u32 bs512 = max_size / 64; + u8 bs = fls(bs512); + + if (!bs) + return -EINVAL; + --bs; + + /* Demand a power of two. */ + if ((1 << bs) != bs512) + return -EINVAL; + + if (bs < mlxsw_sp_port->mlxsw_sp->lowest_shaper_bs || + bs > MLXSW_REG_QEEC_HIGHEST_SHAPER_BS) + return -EINVAL; + + *p_burst_size = bs; + return 0; +} + +static u32 +mlxsw_sp_qdisc_tbf_max_size(u8 bs) +{ + return (1U << bs) * 64; +} + +static u64 +mlxsw_sp_qdisc_tbf_rate_kbps(struct tc_tbf_qopt_offload_replace_params *p) +{ + /* TBF interface is in bytes/s, whereas Spectrum ASIC is configured in + * Kbits/s. + */ + return div_u64(p->rate.rate_bytes_ps, 1000) * 8; +} + +static int +mlxsw_sp_qdisc_tbf_check_params(struct mlxsw_sp_port *mlxsw_sp_port, + void *params) +{ + struct tc_tbf_qopt_offload_replace_params *p = params; + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u64 rate_kbps = mlxsw_sp_qdisc_tbf_rate_kbps(p); + u8 burst_size; + int err; + + if (rate_kbps >= MLXSW_REG_QEEC_MAS_DIS) { + dev_err(mlxsw_sp_port->mlxsw_sp->bus_info->dev, + "spectrum: TBF: rate of %lluKbps must be below %u\n", + rate_kbps, MLXSW_REG_QEEC_MAS_DIS); + return -EINVAL; + } + + err = mlxsw_sp_qdisc_tbf_bs(mlxsw_sp_port, p->max_size, &burst_size); + if (err) { + u8 highest_shaper_bs = MLXSW_REG_QEEC_HIGHEST_SHAPER_BS; + + dev_err(mlxsw_sp->bus_info->dev, + "spectrum: TBF: invalid burst size of %u, must be a power of two between %u and %u", + p->max_size, + mlxsw_sp_qdisc_tbf_max_size(mlxsw_sp->lowest_shaper_bs), + mlxsw_sp_qdisc_tbf_max_size(highest_shaper_bs)); + return -EINVAL; + } + + return 0; +} + +static int +mlxsw_sp_qdisc_tbf_replace(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + enum mlxsw_reg_qeec_hr hr = mlxsw_sp_qdisc_tbf_hr(mlxsw_sp_port, + mlxsw_sp_qdisc); + struct tc_tbf_qopt_offload_replace_params *p = params; + u64 rate_kbps = mlxsw_sp_qdisc_tbf_rate_kbps(p); + int tclass_num; + u8 burst_size; + int err; + + err = mlxsw_sp_qdisc_future_fifo_replace(mlxsw_sp_port, handle, 0, + &mlxsw_sp_qdisc->qdiscs[0]); + if (err) + return err; + mlxsw_sp_qdisc_future_fifos_init(mlxsw_sp_port, TC_H_UNSPEC); + + tclass_num = mlxsw_sp_qdisc_get_tclass_num(mlxsw_sp_port, + mlxsw_sp_qdisc); + + err = mlxsw_sp_qdisc_tbf_bs(mlxsw_sp_port, p->max_size, &burst_size); + if (WARN_ON_ONCE(err)) + /* check_params above was supposed to reject this value. */ + return -EINVAL; + + return mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, hr, tclass_num, 0, + rate_kbps, burst_size); +} + +static void +mlxsw_sp_qdisc_tbf_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_tbf_qopt_offload_replace_params *p = params; + + mlxsw_sp_qdisc_leaf_unoffload(mlxsw_sp_port, mlxsw_sp_qdisc, p->qstats); +} + +static int +mlxsw_sp_qdisc_get_tbf_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_qopt_offload_stats *stats_ptr) +{ + mlxsw_sp_qdisc_get_tc_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + stats_ptr); + return 0; +} + +static struct mlxsw_sp_qdisc_ops mlxsw_sp_qdisc_ops_tbf = { + .type = MLXSW_SP_QDISC_TBF, + .check_params = mlxsw_sp_qdisc_tbf_check_params, + .replace = mlxsw_sp_qdisc_tbf_replace, + .unoffload = mlxsw_sp_qdisc_tbf_unoffload, + .destroy = mlxsw_sp_qdisc_tbf_destroy, + .get_stats = mlxsw_sp_qdisc_get_tbf_stats, + .clean_stats = mlxsw_sp_setup_tc_qdisc_leaf_clean_stats, + .find_class = mlxsw_sp_qdisc_leaf_find_class, + .num_classes = 1, +}; + +static int __mlxsw_sp_setup_tc_tbf(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_tbf_qopt_offload *p) +{ + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc; + + mlxsw_sp_qdisc = mlxsw_sp_qdisc_find(mlxsw_sp_port, p->parent); + if (!mlxsw_sp_qdisc) + return -EOPNOTSUPP; + + if (p->command == TC_TBF_REPLACE) + return mlxsw_sp_qdisc_replace(mlxsw_sp_port, p->handle, + mlxsw_sp_qdisc, + &mlxsw_sp_qdisc_ops_tbf, + &p->replace_params); + + if (!mlxsw_sp_qdisc_compare(mlxsw_sp_qdisc, p->handle)) + return -EOPNOTSUPP; + + switch (p->command) { + case TC_TBF_DESTROY: + return mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + case TC_TBF_STATS: + return mlxsw_sp_qdisc_get_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &p->stats); + case TC_TBF_GRAFT: + return mlxsw_sp_qdisc_graft(mlxsw_sp_port, mlxsw_sp_qdisc, 0, + p->child_handle); + default: + return -EOPNOTSUPP; + } +} + +int mlxsw_sp_setup_tc_tbf(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_tbf_qopt_offload *p) +{ + int err; + + mutex_lock(&mlxsw_sp_port->qdisc->lock); + err = __mlxsw_sp_setup_tc_tbf(mlxsw_sp_port, p); + mutex_unlock(&mlxsw_sp_port->qdisc->lock); + + return err; +} + +static int +mlxsw_sp_qdisc_fifo_check_params(struct mlxsw_sp_port *mlxsw_sp_port, + void *params) +{ + return 0; +} + +static int +mlxsw_sp_qdisc_fifo_replace(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + return 0; +} + +static int +mlxsw_sp_qdisc_get_fifo_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_qopt_offload_stats *stats_ptr) +{ + mlxsw_sp_qdisc_get_tc_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + stats_ptr); + return 0; +} + +static struct mlxsw_sp_qdisc_ops mlxsw_sp_qdisc_ops_fifo = { + .type = MLXSW_SP_QDISC_FIFO, + .check_params = mlxsw_sp_qdisc_fifo_check_params, + .replace = mlxsw_sp_qdisc_fifo_replace, + .get_stats = mlxsw_sp_qdisc_get_fifo_stats, + .clean_stats = mlxsw_sp_setup_tc_qdisc_leaf_clean_stats, +}; + +static int +mlxsw_sp_qdisc_future_fifo_replace(struct mlxsw_sp_port *mlxsw_sp_port, + u32 handle, unsigned int band, + struct mlxsw_sp_qdisc *child_qdisc) +{ + struct mlxsw_sp_qdisc_state *qdisc_state = mlxsw_sp_port->qdisc; + + if (handle == qdisc_state->future_handle && + qdisc_state->future_fifos[band]) + return mlxsw_sp_qdisc_replace(mlxsw_sp_port, TC_H_UNSPEC, + child_qdisc, + &mlxsw_sp_qdisc_ops_fifo, + NULL); + return 0; +} + +static void +mlxsw_sp_qdisc_future_fifos_init(struct mlxsw_sp_port *mlxsw_sp_port, + u32 handle) +{ + struct mlxsw_sp_qdisc_state *qdisc_state = mlxsw_sp_port->qdisc; + + qdisc_state->future_handle = handle; + memset(qdisc_state->future_fifos, 0, sizeof(qdisc_state->future_fifos)); +} + +static int __mlxsw_sp_setup_tc_fifo(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_fifo_qopt_offload *p) +{ + struct mlxsw_sp_qdisc_state *qdisc_state = mlxsw_sp_port->qdisc; + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc; + unsigned int band; + u32 parent_handle; + + mlxsw_sp_qdisc = mlxsw_sp_qdisc_find(mlxsw_sp_port, p->parent); + if (!mlxsw_sp_qdisc && p->handle == TC_H_UNSPEC) { + parent_handle = TC_H_MAJ(p->parent); + if (parent_handle != qdisc_state->future_handle) { + /* This notifications is for a different Qdisc than + * previously. Wipe the future cache. + */ + mlxsw_sp_qdisc_future_fifos_init(mlxsw_sp_port, + parent_handle); + } + + band = TC_H_MIN(p->parent) - 1; + if (band < IEEE_8021QAZ_MAX_TCS) { + if (p->command == TC_FIFO_REPLACE) + qdisc_state->future_fifos[band] = true; + else if (p->command == TC_FIFO_DESTROY) + qdisc_state->future_fifos[band] = false; + } + } + if (!mlxsw_sp_qdisc) + return -EOPNOTSUPP; + + if (p->command == TC_FIFO_REPLACE) { + return mlxsw_sp_qdisc_replace(mlxsw_sp_port, p->handle, + mlxsw_sp_qdisc, + &mlxsw_sp_qdisc_ops_fifo, NULL); + } + + if (!mlxsw_sp_qdisc_compare(mlxsw_sp_qdisc, p->handle)) + return -EOPNOTSUPP; + + switch (p->command) { + case TC_FIFO_DESTROY: + return mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + case TC_FIFO_STATS: + return mlxsw_sp_qdisc_get_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &p->stats); + case TC_FIFO_REPLACE: /* Handled above. */ + break; + } + + return -EOPNOTSUPP; +} + +int mlxsw_sp_setup_tc_fifo(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_fifo_qopt_offload *p) +{ + int err; + + mutex_lock(&mlxsw_sp_port->qdisc->lock); + err = __mlxsw_sp_setup_tc_fifo(mlxsw_sp_port, p); + mutex_unlock(&mlxsw_sp_port->qdisc->lock); + + return err; +} + +static int __mlxsw_sp_qdisc_ets_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + int i; + + for (i = 0; i < mlxsw_sp_qdisc->num_classes; i++) { + mlxsw_sp_port_prio_tc_set(mlxsw_sp_port, i, + MLXSW_SP_PORT_DEFAULT_TCLASS); + mlxsw_sp_port_ets_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, + i, 0, false, 0); + } + + kfree(mlxsw_sp_qdisc->ets_data); + mlxsw_sp_qdisc->ets_data = NULL; + return 0; +} + +static int +mlxsw_sp_qdisc_prio_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + return __mlxsw_sp_qdisc_ets_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); +} + +static int +__mlxsw_sp_qdisc_ets_check_params(unsigned int nbands) +{ + if (nbands > IEEE_8021QAZ_MAX_TCS) + return -EOPNOTSUPP; + + return 0; +} + +static int +mlxsw_sp_qdisc_prio_check_params(struct mlxsw_sp_port *mlxsw_sp_port, + void *params) +{ + struct tc_prio_qopt_offload_params *p = params; + + return __mlxsw_sp_qdisc_ets_check_params(p->bands); +} + +static struct mlxsw_sp_qdisc * +mlxsw_sp_qdisc_walk_cb_clean_stats(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *mlxsw_sp_port) +{ + u64 backlog; + + if (mlxsw_sp_qdisc->ops) { + backlog = mlxsw_sp_qdisc->stats_base.backlog; + if (mlxsw_sp_qdisc->ops->clean_stats) + mlxsw_sp_qdisc->ops->clean_stats(mlxsw_sp_port, + mlxsw_sp_qdisc); + mlxsw_sp_qdisc->stats_base.backlog = backlog; + } + + return NULL; +} + +static void +mlxsw_sp_qdisc_tree_clean_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + mlxsw_sp_qdisc_walk(mlxsw_sp_qdisc, mlxsw_sp_qdisc_walk_cb_clean_stats, + mlxsw_sp_port); +} + +static int +__mlxsw_sp_qdisc_ets_replace(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + u32 handle, unsigned int nbands, + const unsigned int *quanta, + const unsigned int *weights, + const u8 *priomap) +{ + struct mlxsw_sp_qdisc_ets_data *ets_data = mlxsw_sp_qdisc->ets_data; + struct mlxsw_sp_qdisc_ets_band *ets_band; + struct mlxsw_sp_qdisc *child_qdisc; + u8 old_priomap, new_priomap; + int i, band; + int err; + + if (!ets_data) { + ets_data = kzalloc(sizeof(*ets_data), GFP_KERNEL); + if (!ets_data) + return -ENOMEM; + mlxsw_sp_qdisc->ets_data = ets_data; + + for (band = 0; band < mlxsw_sp_qdisc->num_classes; band++) { + int tclass_num = MLXSW_SP_PRIO_BAND_TO_TCLASS(band); + + ets_band = &ets_data->bands[band]; + ets_band->tclass_num = tclass_num; + } + } + + for (band = 0; band < nbands; band++) { + int tclass_num; + + child_qdisc = &mlxsw_sp_qdisc->qdiscs[band]; + ets_band = &ets_data->bands[band]; + + tclass_num = ets_band->tclass_num; + old_priomap = ets_band->prio_bitmap; + new_priomap = 0; + + err = mlxsw_sp_port_ets_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, + tclass_num, 0, !!quanta[band], + weights[band]); + if (err) + return err; + + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + if (priomap[i] == band) { + new_priomap |= BIT(i); + if (BIT(i) & old_priomap) + continue; + err = mlxsw_sp_port_prio_tc_set(mlxsw_sp_port, + i, tclass_num); + if (err) + return err; + } + } + + ets_band->prio_bitmap = new_priomap; + + if (old_priomap != new_priomap) + mlxsw_sp_qdisc_tree_clean_stats(mlxsw_sp_port, + child_qdisc); + + err = mlxsw_sp_qdisc_future_fifo_replace(mlxsw_sp_port, handle, + band, child_qdisc); + if (err) + return err; + } + for (; band < IEEE_8021QAZ_MAX_TCS; band++) { + ets_band = &ets_data->bands[band]; + ets_band->prio_bitmap = 0; + + child_qdisc = &mlxsw_sp_qdisc->qdiscs[band]; + mlxsw_sp_qdisc_destroy(mlxsw_sp_port, child_qdisc); + + mlxsw_sp_port_ets_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, + ets_band->tclass_num, 0, false, 0); + } + + mlxsw_sp_qdisc_future_fifos_init(mlxsw_sp_port, TC_H_UNSPEC); + return 0; +} + +static int +mlxsw_sp_qdisc_prio_replace(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_prio_qopt_offload_params *p = params; + unsigned int zeroes[TCQ_ETS_MAX_BANDS] = {0}; + + return __mlxsw_sp_qdisc_ets_replace(mlxsw_sp_port, mlxsw_sp_qdisc, + handle, p->bands, zeroes, + zeroes, p->priomap); +} + +static void +__mlxsw_sp_qdisc_ets_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct gnet_stats_queue *qstats) +{ + u64 backlog; + + backlog = mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, + mlxsw_sp_qdisc->stats_base.backlog); + qstats->backlog -= backlog; +} + +static void +mlxsw_sp_qdisc_prio_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_prio_qopt_offload_params *p = params; + + __mlxsw_sp_qdisc_ets_unoffload(mlxsw_sp_port, mlxsw_sp_qdisc, + p->qstats); +} + +static int +mlxsw_sp_qdisc_get_prio_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_qopt_offload_stats *stats_ptr) +{ + struct mlxsw_sp_qdisc *tc_qdisc; + u64 tx_packets = 0; + u64 tx_bytes = 0; + u64 backlog = 0; + u64 drops = 0; + int i; + + for (i = 0; i < mlxsw_sp_qdisc->num_classes; i++) { + tc_qdisc = &mlxsw_sp_qdisc->qdiscs[i]; + mlxsw_sp_qdisc_collect_tc_stats(mlxsw_sp_port, tc_qdisc, + &tx_bytes, &tx_packets, + &drops, &backlog); + } + + mlxsw_sp_qdisc_update_stats(mlxsw_sp_port->mlxsw_sp, mlxsw_sp_qdisc, + tx_bytes, tx_packets, drops, backlog, + stats_ptr); + return 0; +} + +static void +mlxsw_sp_setup_tc_qdisc_prio_clean_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + struct mlxsw_sp_qdisc_stats *stats_base; + struct mlxsw_sp_port_xstats *xstats; + struct rtnl_link_stats64 *stats; + int i; + + xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; + stats = &mlxsw_sp_port->periodic_hw_stats.stats; + stats_base = &mlxsw_sp_qdisc->stats_base; + + stats_base->tx_packets = stats->tx_packets; + stats_base->tx_bytes = stats->tx_bytes; + + stats_base->drops = 0; + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + stats_base->drops += mlxsw_sp_xstats_tail_drop(xstats, i); + stats_base->drops += xstats->wred_drop[i]; + } + + mlxsw_sp_qdisc->stats_base.backlog = 0; +} + +static struct mlxsw_sp_qdisc * +mlxsw_sp_qdisc_prio_find_class(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + u32 parent) +{ + int child_index = TC_H_MIN(parent); + int band = child_index - 1; + + if (band < 0 || band >= mlxsw_sp_qdisc->num_classes) + return NULL; + return &mlxsw_sp_qdisc->qdiscs[band]; +} + +static struct mlxsw_sp_qdisc_ets_band * +mlxsw_sp_qdisc_ets_get_band(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct mlxsw_sp_qdisc *child) +{ + unsigned int band = child - mlxsw_sp_qdisc->qdiscs; + + if (WARN_ON(band >= IEEE_8021QAZ_MAX_TCS)) + band = 0; + return &mlxsw_sp_qdisc->ets_data->bands[band]; +} + +static u8 +mlxsw_sp_qdisc_ets_get_prio_bitmap(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct mlxsw_sp_qdisc *child) +{ + return mlxsw_sp_qdisc_ets_get_band(mlxsw_sp_qdisc, child)->prio_bitmap; +} + +static int +mlxsw_sp_qdisc_ets_get_tclass_num(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct mlxsw_sp_qdisc *child) +{ + return mlxsw_sp_qdisc_ets_get_band(mlxsw_sp_qdisc, child)->tclass_num; +} + +static struct mlxsw_sp_qdisc_ops mlxsw_sp_qdisc_ops_prio = { + .type = MLXSW_SP_QDISC_PRIO, + .check_params = mlxsw_sp_qdisc_prio_check_params, + .replace = mlxsw_sp_qdisc_prio_replace, + .unoffload = mlxsw_sp_qdisc_prio_unoffload, + .destroy = mlxsw_sp_qdisc_prio_destroy, + .get_stats = mlxsw_sp_qdisc_get_prio_stats, + .clean_stats = mlxsw_sp_setup_tc_qdisc_prio_clean_stats, + .find_class = mlxsw_sp_qdisc_prio_find_class, + .num_classes = IEEE_8021QAZ_MAX_TCS, + .get_prio_bitmap = mlxsw_sp_qdisc_ets_get_prio_bitmap, + .get_tclass_num = mlxsw_sp_qdisc_ets_get_tclass_num, +}; + +static int +mlxsw_sp_qdisc_ets_check_params(struct mlxsw_sp_port *mlxsw_sp_port, + void *params) +{ + struct tc_ets_qopt_offload_replace_params *p = params; + + return __mlxsw_sp_qdisc_ets_check_params(p->bands); +} + +static int +mlxsw_sp_qdisc_ets_replace(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_ets_qopt_offload_replace_params *p = params; + + return __mlxsw_sp_qdisc_ets_replace(mlxsw_sp_port, mlxsw_sp_qdisc, + handle, p->bands, p->quanta, + p->weights, p->priomap); +} + +static void +mlxsw_sp_qdisc_ets_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_ets_qopt_offload_replace_params *p = params; + + __mlxsw_sp_qdisc_ets_unoffload(mlxsw_sp_port, mlxsw_sp_qdisc, + p->qstats); +} + +static int +mlxsw_sp_qdisc_ets_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + return __mlxsw_sp_qdisc_ets_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); +} + +static struct mlxsw_sp_qdisc_ops mlxsw_sp_qdisc_ops_ets = { + .type = MLXSW_SP_QDISC_ETS, + .check_params = mlxsw_sp_qdisc_ets_check_params, + .replace = mlxsw_sp_qdisc_ets_replace, + .unoffload = mlxsw_sp_qdisc_ets_unoffload, + .destroy = mlxsw_sp_qdisc_ets_destroy, + .get_stats = mlxsw_sp_qdisc_get_prio_stats, + .clean_stats = mlxsw_sp_setup_tc_qdisc_prio_clean_stats, + .find_class = mlxsw_sp_qdisc_prio_find_class, + .num_classes = IEEE_8021QAZ_MAX_TCS, + .get_prio_bitmap = mlxsw_sp_qdisc_ets_get_prio_bitmap, + .get_tclass_num = mlxsw_sp_qdisc_ets_get_tclass_num, +}; + +/* Linux allows linking of Qdiscs to arbitrary classes (so long as the resulting + * graph is free of cycles). These operations do not change the parent handle + * though, which means it can be incomplete (if there is more than one class + * where the Qdisc in question is grafted) or outright wrong (if the Qdisc was + * linked to a different class and then removed from the original class). + * + * E.g. consider this sequence of operations: + * + * # tc qdisc add dev swp1 root handle 1: prio + * # tc qdisc add dev swp1 parent 1:3 handle 13: red limit 1000000 avpkt 10000 + * RED: set bandwidth to 10Mbit + * # tc qdisc link dev swp1 handle 13: parent 1:2 + * + * At this point, both 1:2 and 1:3 have the same RED Qdisc instance as their + * child. But RED will still only claim that 1:3 is its parent. If it's removed + * from that band, its only parent will be 1:2, but it will continue to claim + * that it is in fact 1:3. + * + * The notification for child Qdisc replace (e.g. TC_RED_REPLACE) comes before + * the notification for parent graft (e.g. TC_PRIO_GRAFT). We take the replace + * notification to offload the child Qdisc, based on its parent handle, and use + * the graft operation to validate that the class where the child is actually + * grafted corresponds to the parent handle. If the two don't match, we + * unoffload the child. + */ +static int mlxsw_sp_qdisc_graft(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + u8 band, u32 child_handle) +{ + struct mlxsw_sp_qdisc *old_qdisc; + u32 parent; + + if (band < mlxsw_sp_qdisc->num_classes && + mlxsw_sp_qdisc->qdiscs[band].handle == child_handle) + return 0; + + if (!child_handle) { + /* This is an invisible FIFO replacing the original Qdisc. + * Ignore it--the original Qdisc's destroy will follow. + */ + return 0; + } + + /* See if the grafted qdisc is already offloaded on any tclass. If so, + * unoffload it. + */ + old_qdisc = mlxsw_sp_qdisc_find_by_handle(mlxsw_sp_port, + child_handle); + if (old_qdisc) + mlxsw_sp_qdisc_destroy(mlxsw_sp_port, old_qdisc); + + parent = TC_H_MAKE(mlxsw_sp_qdisc->handle, band + 1); + mlxsw_sp_qdisc = mlxsw_sp_qdisc->ops->find_class(mlxsw_sp_qdisc, + parent); + if (!WARN_ON(!mlxsw_sp_qdisc)) + mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + + return -EOPNOTSUPP; +} + +static int __mlxsw_sp_setup_tc_prio(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_prio_qopt_offload *p) +{ + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc; + + mlxsw_sp_qdisc = mlxsw_sp_qdisc_find(mlxsw_sp_port, p->parent); + if (!mlxsw_sp_qdisc) + return -EOPNOTSUPP; + + if (p->command == TC_PRIO_REPLACE) + return mlxsw_sp_qdisc_replace(mlxsw_sp_port, p->handle, + mlxsw_sp_qdisc, + &mlxsw_sp_qdisc_ops_prio, + &p->replace_params); + + if (!mlxsw_sp_qdisc_compare(mlxsw_sp_qdisc, p->handle)) + return -EOPNOTSUPP; + + switch (p->command) { + case TC_PRIO_DESTROY: + return mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + case TC_PRIO_STATS: + return mlxsw_sp_qdisc_get_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &p->stats); + case TC_PRIO_GRAFT: + return mlxsw_sp_qdisc_graft(mlxsw_sp_port, mlxsw_sp_qdisc, + p->graft_params.band, + p->graft_params.child_handle); + default: + return -EOPNOTSUPP; + } +} + +int mlxsw_sp_setup_tc_prio(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_prio_qopt_offload *p) +{ + int err; + + mutex_lock(&mlxsw_sp_port->qdisc->lock); + err = __mlxsw_sp_setup_tc_prio(mlxsw_sp_port, p); + mutex_unlock(&mlxsw_sp_port->qdisc->lock); + + return err; +} + +static int __mlxsw_sp_setup_tc_ets(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_ets_qopt_offload *p) +{ + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc; + + mlxsw_sp_qdisc = mlxsw_sp_qdisc_find(mlxsw_sp_port, p->parent); + if (!mlxsw_sp_qdisc) + return -EOPNOTSUPP; + + if (p->command == TC_ETS_REPLACE) + return mlxsw_sp_qdisc_replace(mlxsw_sp_port, p->handle, + mlxsw_sp_qdisc, + &mlxsw_sp_qdisc_ops_ets, + &p->replace_params); + + if (!mlxsw_sp_qdisc_compare(mlxsw_sp_qdisc, p->handle)) + return -EOPNOTSUPP; + + switch (p->command) { + case TC_ETS_DESTROY: + return mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + case TC_ETS_STATS: + return mlxsw_sp_qdisc_get_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &p->stats); + case TC_ETS_GRAFT: + return mlxsw_sp_qdisc_graft(mlxsw_sp_port, mlxsw_sp_qdisc, + p->graft_params.band, + p->graft_params.child_handle); + default: + return -EOPNOTSUPP; + } +} + +int mlxsw_sp_setup_tc_ets(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_ets_qopt_offload *p) +{ + int err; + + mutex_lock(&mlxsw_sp_port->qdisc->lock); + err = __mlxsw_sp_setup_tc_ets(mlxsw_sp_port, p); + mutex_unlock(&mlxsw_sp_port->qdisc->lock); + + return err; +} + +struct mlxsw_sp_qevent_block { + struct list_head binding_list; + struct list_head mall_entry_list; + struct mlxsw_sp *mlxsw_sp; +}; + +struct mlxsw_sp_qevent_binding { + struct list_head list; + struct mlxsw_sp_port *mlxsw_sp_port; + u32 handle; + int tclass_num; + enum mlxsw_sp_span_trigger span_trigger; + unsigned int action_mask; +}; + +static LIST_HEAD(mlxsw_sp_qevent_block_cb_list); + +static int mlxsw_sp_qevent_span_configure(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mall_entry *mall_entry, + struct mlxsw_sp_qevent_binding *qevent_binding, + const struct mlxsw_sp_span_agent_parms *agent_parms, + int *p_span_id) +{ + enum mlxsw_sp_span_trigger span_trigger = qevent_binding->span_trigger; + struct mlxsw_sp_port *mlxsw_sp_port = qevent_binding->mlxsw_sp_port; + struct mlxsw_sp_span_trigger_parms trigger_parms = {}; + bool ingress; + int span_id; + int err; + + err = mlxsw_sp_span_agent_get(mlxsw_sp, &span_id, agent_parms); + if (err) + return err; + + ingress = mlxsw_sp_span_trigger_is_ingress(span_trigger); + err = mlxsw_sp_span_analyzed_port_get(mlxsw_sp_port, ingress); + if (err) + goto err_analyzed_port_get; + + trigger_parms.span_id = span_id; + trigger_parms.probability_rate = 1; + err = mlxsw_sp_span_agent_bind(mlxsw_sp, span_trigger, mlxsw_sp_port, + &trigger_parms); + if (err) + goto err_agent_bind; + + err = mlxsw_sp_span_trigger_enable(mlxsw_sp_port, span_trigger, + qevent_binding->tclass_num); + if (err) + goto err_trigger_enable; + + *p_span_id = span_id; + return 0; + +err_trigger_enable: + mlxsw_sp_span_agent_unbind(mlxsw_sp, span_trigger, mlxsw_sp_port, + &trigger_parms); +err_agent_bind: + mlxsw_sp_span_analyzed_port_put(mlxsw_sp_port, ingress); +err_analyzed_port_get: + mlxsw_sp_span_agent_put(mlxsw_sp, span_id); + return err; +} + +static void mlxsw_sp_qevent_span_deconfigure(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_qevent_binding *qevent_binding, + int span_id) +{ + enum mlxsw_sp_span_trigger span_trigger = qevent_binding->span_trigger; + struct mlxsw_sp_port *mlxsw_sp_port = qevent_binding->mlxsw_sp_port; + struct mlxsw_sp_span_trigger_parms trigger_parms = { + .span_id = span_id, + }; + bool ingress; + + ingress = mlxsw_sp_span_trigger_is_ingress(span_trigger); + + mlxsw_sp_span_trigger_disable(mlxsw_sp_port, span_trigger, + qevent_binding->tclass_num); + mlxsw_sp_span_agent_unbind(mlxsw_sp, span_trigger, mlxsw_sp_port, + &trigger_parms); + mlxsw_sp_span_analyzed_port_put(mlxsw_sp_port, ingress); + mlxsw_sp_span_agent_put(mlxsw_sp, span_id); +} + +static int mlxsw_sp_qevent_mirror_configure(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mall_entry *mall_entry, + struct mlxsw_sp_qevent_binding *qevent_binding) +{ + struct mlxsw_sp_span_agent_parms agent_parms = { + .to_dev = mall_entry->mirror.to_dev, + }; + + return mlxsw_sp_qevent_span_configure(mlxsw_sp, mall_entry, qevent_binding, + &agent_parms, &mall_entry->mirror.span_id); +} + +static void mlxsw_sp_qevent_mirror_deconfigure(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mall_entry *mall_entry, + struct mlxsw_sp_qevent_binding *qevent_binding) +{ + mlxsw_sp_qevent_span_deconfigure(mlxsw_sp, qevent_binding, mall_entry->mirror.span_id); +} + +static int mlxsw_sp_qevent_trap_configure(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mall_entry *mall_entry, + struct mlxsw_sp_qevent_binding *qevent_binding) +{ + struct mlxsw_sp_span_agent_parms agent_parms = { + .session_id = MLXSW_SP_SPAN_SESSION_ID_BUFFER, + }; + int err; + + err = mlxsw_sp_trap_group_policer_hw_id_get(mlxsw_sp, + DEVLINK_TRAP_GROUP_GENERIC_ID_BUFFER_DROPS, + &agent_parms.policer_enable, + &agent_parms.policer_id); + if (err) + return err; + + return mlxsw_sp_qevent_span_configure(mlxsw_sp, mall_entry, qevent_binding, + &agent_parms, &mall_entry->trap.span_id); +} + +static void mlxsw_sp_qevent_trap_deconfigure(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mall_entry *mall_entry, + struct mlxsw_sp_qevent_binding *qevent_binding) +{ + mlxsw_sp_qevent_span_deconfigure(mlxsw_sp, qevent_binding, mall_entry->trap.span_id); +} + +static int +mlxsw_sp_qevent_entry_configure(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mall_entry *mall_entry, + struct mlxsw_sp_qevent_binding *qevent_binding, + struct netlink_ext_ack *extack) +{ + if (!(BIT(mall_entry->type) & qevent_binding->action_mask)) { + NL_SET_ERR_MSG(extack, "Action not supported at this qevent"); + return -EOPNOTSUPP; + } + + switch (mall_entry->type) { + case MLXSW_SP_MALL_ACTION_TYPE_MIRROR: + return mlxsw_sp_qevent_mirror_configure(mlxsw_sp, mall_entry, qevent_binding); + case MLXSW_SP_MALL_ACTION_TYPE_TRAP: + return mlxsw_sp_qevent_trap_configure(mlxsw_sp, mall_entry, qevent_binding); + default: + /* This should have been validated away. */ + WARN_ON(1); + return -EOPNOTSUPP; + } +} + +static void mlxsw_sp_qevent_entry_deconfigure(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mall_entry *mall_entry, + struct mlxsw_sp_qevent_binding *qevent_binding) +{ + switch (mall_entry->type) { + case MLXSW_SP_MALL_ACTION_TYPE_MIRROR: + return mlxsw_sp_qevent_mirror_deconfigure(mlxsw_sp, mall_entry, qevent_binding); + case MLXSW_SP_MALL_ACTION_TYPE_TRAP: + return mlxsw_sp_qevent_trap_deconfigure(mlxsw_sp, mall_entry, qevent_binding); + default: + WARN_ON(1); + return; + } +} + +static int +mlxsw_sp_qevent_binding_configure(struct mlxsw_sp_qevent_block *qevent_block, + struct mlxsw_sp_qevent_binding *qevent_binding, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_mall_entry *mall_entry; + int err; + + list_for_each_entry(mall_entry, &qevent_block->mall_entry_list, list) { + err = mlxsw_sp_qevent_entry_configure(qevent_block->mlxsw_sp, mall_entry, + qevent_binding, extack); + if (err) + goto err_entry_configure; + } + + return 0; + +err_entry_configure: + list_for_each_entry_continue_reverse(mall_entry, &qevent_block->mall_entry_list, list) + mlxsw_sp_qevent_entry_deconfigure(qevent_block->mlxsw_sp, mall_entry, + qevent_binding); + return err; +} + +static void mlxsw_sp_qevent_binding_deconfigure(struct mlxsw_sp_qevent_block *qevent_block, + struct mlxsw_sp_qevent_binding *qevent_binding) +{ + struct mlxsw_sp_mall_entry *mall_entry; + + list_for_each_entry(mall_entry, &qevent_block->mall_entry_list, list) + mlxsw_sp_qevent_entry_deconfigure(qevent_block->mlxsw_sp, mall_entry, + qevent_binding); +} + +static int +mlxsw_sp_qevent_block_configure(struct mlxsw_sp_qevent_block *qevent_block, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_qevent_binding *qevent_binding; + int err; + + list_for_each_entry(qevent_binding, &qevent_block->binding_list, list) { + err = mlxsw_sp_qevent_binding_configure(qevent_block, + qevent_binding, + extack); + if (err) + goto err_binding_configure; + } + + return 0; + +err_binding_configure: + list_for_each_entry_continue_reverse(qevent_binding, &qevent_block->binding_list, list) + mlxsw_sp_qevent_binding_deconfigure(qevent_block, qevent_binding); + return err; +} + +static void mlxsw_sp_qevent_block_deconfigure(struct mlxsw_sp_qevent_block *qevent_block) +{ + struct mlxsw_sp_qevent_binding *qevent_binding; + + list_for_each_entry(qevent_binding, &qevent_block->binding_list, list) + mlxsw_sp_qevent_binding_deconfigure(qevent_block, qevent_binding); +} + +static struct mlxsw_sp_mall_entry * +mlxsw_sp_qevent_mall_entry_find(struct mlxsw_sp_qevent_block *block, unsigned long cookie) +{ + struct mlxsw_sp_mall_entry *mall_entry; + + list_for_each_entry(mall_entry, &block->mall_entry_list, list) + if (mall_entry->cookie == cookie) + return mall_entry; + + return NULL; +} + +static int mlxsw_sp_qevent_mall_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_qevent_block *qevent_block, + struct tc_cls_matchall_offload *f) +{ + struct mlxsw_sp_mall_entry *mall_entry; + struct flow_action_entry *act; + int err; + + /* It should not currently be possible to replace a matchall rule. So + * this must be a new rule. + */ + if (!list_empty(&qevent_block->mall_entry_list)) { + NL_SET_ERR_MSG(f->common.extack, "At most one filter supported"); + return -EOPNOTSUPP; + } + if (f->rule->action.num_entries != 1) { + NL_SET_ERR_MSG(f->common.extack, "Only singular actions supported"); + return -EOPNOTSUPP; + } + if (f->common.chain_index) { + NL_SET_ERR_MSG(f->common.extack, "Only chain 0 is supported"); + return -EOPNOTSUPP; + } + if (f->common.protocol != htons(ETH_P_ALL)) { + NL_SET_ERR_MSG(f->common.extack, "Protocol matching not supported"); + return -EOPNOTSUPP; + } + + act = &f->rule->action.entries[0]; + if (!(act->hw_stats & FLOW_ACTION_HW_STATS_DISABLED)) { + NL_SET_ERR_MSG(f->common.extack, "HW counters not supported on qevents"); + return -EOPNOTSUPP; + } + + mall_entry = kzalloc(sizeof(*mall_entry), GFP_KERNEL); + if (!mall_entry) + return -ENOMEM; + mall_entry->cookie = f->cookie; + + if (act->id == FLOW_ACTION_MIRRED) { + mall_entry->type = MLXSW_SP_MALL_ACTION_TYPE_MIRROR; + mall_entry->mirror.to_dev = act->dev; + } else if (act->id == FLOW_ACTION_TRAP) { + mall_entry->type = MLXSW_SP_MALL_ACTION_TYPE_TRAP; + } else { + NL_SET_ERR_MSG(f->common.extack, "Unsupported action"); + err = -EOPNOTSUPP; + goto err_unsupported_action; + } + + list_add_tail(&mall_entry->list, &qevent_block->mall_entry_list); + + err = mlxsw_sp_qevent_block_configure(qevent_block, f->common.extack); + if (err) + goto err_block_configure; + + return 0; + +err_block_configure: + list_del(&mall_entry->list); +err_unsupported_action: + kfree(mall_entry); + return err; +} + +static void mlxsw_sp_qevent_mall_destroy(struct mlxsw_sp_qevent_block *qevent_block, + struct tc_cls_matchall_offload *f) +{ + struct mlxsw_sp_mall_entry *mall_entry; + + mall_entry = mlxsw_sp_qevent_mall_entry_find(qevent_block, f->cookie); + if (!mall_entry) + return; + + mlxsw_sp_qevent_block_deconfigure(qevent_block); + + list_del(&mall_entry->list); + kfree(mall_entry); +} + +static int mlxsw_sp_qevent_block_mall_cb(struct mlxsw_sp_qevent_block *qevent_block, + struct tc_cls_matchall_offload *f) +{ + struct mlxsw_sp *mlxsw_sp = qevent_block->mlxsw_sp; + + switch (f->command) { + case TC_CLSMATCHALL_REPLACE: + return mlxsw_sp_qevent_mall_replace(mlxsw_sp, qevent_block, f); + case TC_CLSMATCHALL_DESTROY: + mlxsw_sp_qevent_mall_destroy(qevent_block, f); + return 0; + default: + return -EOPNOTSUPP; + } +} + +static int mlxsw_sp_qevent_block_cb(enum tc_setup_type type, void *type_data, void *cb_priv) +{ + struct mlxsw_sp_qevent_block *qevent_block = cb_priv; + + switch (type) { + case TC_SETUP_CLSMATCHALL: + return mlxsw_sp_qevent_block_mall_cb(qevent_block, type_data); + default: + return -EOPNOTSUPP; + } +} + +static struct mlxsw_sp_qevent_block *mlxsw_sp_qevent_block_create(struct mlxsw_sp *mlxsw_sp, + struct net *net) +{ + struct mlxsw_sp_qevent_block *qevent_block; + + qevent_block = kzalloc(sizeof(*qevent_block), GFP_KERNEL); + if (!qevent_block) + return NULL; + + INIT_LIST_HEAD(&qevent_block->binding_list); + INIT_LIST_HEAD(&qevent_block->mall_entry_list); + qevent_block->mlxsw_sp = mlxsw_sp; + return qevent_block; +} + +static void +mlxsw_sp_qevent_block_destroy(struct mlxsw_sp_qevent_block *qevent_block) +{ + WARN_ON(!list_empty(&qevent_block->binding_list)); + WARN_ON(!list_empty(&qevent_block->mall_entry_list)); + kfree(qevent_block); +} + +static void mlxsw_sp_qevent_block_release(void *cb_priv) +{ + struct mlxsw_sp_qevent_block *qevent_block = cb_priv; + + mlxsw_sp_qevent_block_destroy(qevent_block); +} + +static struct mlxsw_sp_qevent_binding * +mlxsw_sp_qevent_binding_create(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, int tclass_num, + enum mlxsw_sp_span_trigger span_trigger, + unsigned int action_mask) +{ + struct mlxsw_sp_qevent_binding *binding; + + binding = kzalloc(sizeof(*binding), GFP_KERNEL); + if (!binding) + return ERR_PTR(-ENOMEM); + + binding->mlxsw_sp_port = mlxsw_sp_port; + binding->handle = handle; + binding->tclass_num = tclass_num; + binding->span_trigger = span_trigger; + binding->action_mask = action_mask; + return binding; +} + +static void +mlxsw_sp_qevent_binding_destroy(struct mlxsw_sp_qevent_binding *binding) +{ + kfree(binding); +} + +static struct mlxsw_sp_qevent_binding * +mlxsw_sp_qevent_binding_lookup(struct mlxsw_sp_qevent_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + u32 handle, + enum mlxsw_sp_span_trigger span_trigger) +{ + struct mlxsw_sp_qevent_binding *qevent_binding; + + list_for_each_entry(qevent_binding, &block->binding_list, list) + if (qevent_binding->mlxsw_sp_port == mlxsw_sp_port && + qevent_binding->handle == handle && + qevent_binding->span_trigger == span_trigger) + return qevent_binding; + return NULL; +} + +static int +mlxsw_sp_setup_tc_block_qevent_bind(struct mlxsw_sp_port *mlxsw_sp_port, + struct flow_block_offload *f, + enum mlxsw_sp_span_trigger span_trigger, + unsigned int action_mask) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_qevent_binding *qevent_binding; + struct mlxsw_sp_qevent_block *qevent_block; + struct flow_block_cb *block_cb; + struct mlxsw_sp_qdisc *qdisc; + bool register_block = false; + int tclass_num; + int err; + + block_cb = flow_block_cb_lookup(f->block, mlxsw_sp_qevent_block_cb, mlxsw_sp); + if (!block_cb) { + qevent_block = mlxsw_sp_qevent_block_create(mlxsw_sp, f->net); + if (!qevent_block) + return -ENOMEM; + block_cb = flow_block_cb_alloc(mlxsw_sp_qevent_block_cb, mlxsw_sp, qevent_block, + mlxsw_sp_qevent_block_release); + if (IS_ERR(block_cb)) { + mlxsw_sp_qevent_block_destroy(qevent_block); + return PTR_ERR(block_cb); + } + register_block = true; + } else { + qevent_block = flow_block_cb_priv(block_cb); + } + flow_block_cb_incref(block_cb); + + qdisc = mlxsw_sp_qdisc_find_by_handle(mlxsw_sp_port, f->sch->handle); + if (!qdisc) { + NL_SET_ERR_MSG(f->extack, "Qdisc not offloaded"); + err = -ENOENT; + goto err_find_qdisc; + } + + if (WARN_ON(mlxsw_sp_qevent_binding_lookup(qevent_block, mlxsw_sp_port, f->sch->handle, + span_trigger))) { + err = -EEXIST; + goto err_binding_exists; + } + + tclass_num = mlxsw_sp_qdisc_get_tclass_num(mlxsw_sp_port, qdisc); + qevent_binding = mlxsw_sp_qevent_binding_create(mlxsw_sp_port, + f->sch->handle, + tclass_num, + span_trigger, + action_mask); + if (IS_ERR(qevent_binding)) { + err = PTR_ERR(qevent_binding); + goto err_binding_create; + } + + err = mlxsw_sp_qevent_binding_configure(qevent_block, qevent_binding, + f->extack); + if (err) + goto err_binding_configure; + + list_add(&qevent_binding->list, &qevent_block->binding_list); + + if (register_block) { + flow_block_cb_add(block_cb, f); + list_add_tail(&block_cb->driver_list, &mlxsw_sp_qevent_block_cb_list); + } + + return 0; + +err_binding_configure: + mlxsw_sp_qevent_binding_destroy(qevent_binding); +err_binding_create: +err_binding_exists: +err_find_qdisc: + if (!flow_block_cb_decref(block_cb)) + flow_block_cb_free(block_cb); + return err; +} + +static void mlxsw_sp_setup_tc_block_qevent_unbind(struct mlxsw_sp_port *mlxsw_sp_port, + struct flow_block_offload *f, + enum mlxsw_sp_span_trigger span_trigger) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_qevent_binding *qevent_binding; + struct mlxsw_sp_qevent_block *qevent_block; + struct flow_block_cb *block_cb; + + block_cb = flow_block_cb_lookup(f->block, mlxsw_sp_qevent_block_cb, mlxsw_sp); + if (!block_cb) + return; + qevent_block = flow_block_cb_priv(block_cb); + + qevent_binding = mlxsw_sp_qevent_binding_lookup(qevent_block, mlxsw_sp_port, f->sch->handle, + span_trigger); + if (!qevent_binding) + return; + + list_del(&qevent_binding->list); + mlxsw_sp_qevent_binding_deconfigure(qevent_block, qevent_binding); + mlxsw_sp_qevent_binding_destroy(qevent_binding); + + if (!flow_block_cb_decref(block_cb)) { + flow_block_cb_remove(block_cb, f); + list_del(&block_cb->driver_list); + } +} + +static int +mlxsw_sp_setup_tc_block_qevent(struct mlxsw_sp_port *mlxsw_sp_port, + struct flow_block_offload *f, + enum mlxsw_sp_span_trigger span_trigger, + unsigned int action_mask) +{ + f->driver_block_list = &mlxsw_sp_qevent_block_cb_list; + + switch (f->command) { + case FLOW_BLOCK_BIND: + return mlxsw_sp_setup_tc_block_qevent_bind(mlxsw_sp_port, f, + span_trigger, + action_mask); + case FLOW_BLOCK_UNBIND: + mlxsw_sp_setup_tc_block_qevent_unbind(mlxsw_sp_port, f, span_trigger); + return 0; + default: + return -EOPNOTSUPP; + } +} + +int mlxsw_sp_setup_tc_block_qevent_early_drop(struct mlxsw_sp_port *mlxsw_sp_port, + struct flow_block_offload *f) +{ + unsigned int action_mask = BIT(MLXSW_SP_MALL_ACTION_TYPE_MIRROR) | + BIT(MLXSW_SP_MALL_ACTION_TYPE_TRAP); + + return mlxsw_sp_setup_tc_block_qevent(mlxsw_sp_port, f, + MLXSW_SP_SPAN_TRIGGER_EARLY_DROP, + action_mask); +} + +int mlxsw_sp_setup_tc_block_qevent_mark(struct mlxsw_sp_port *mlxsw_sp_port, + struct flow_block_offload *f) +{ + unsigned int action_mask = BIT(MLXSW_SP_MALL_ACTION_TYPE_MIRROR); + + return mlxsw_sp_setup_tc_block_qevent(mlxsw_sp_port, f, + MLXSW_SP_SPAN_TRIGGER_ECN, + action_mask); +} + +int mlxsw_sp_tc_qdisc_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp_qdisc_state *qdisc_state; + + qdisc_state = kzalloc(sizeof(*qdisc_state), GFP_KERNEL); + if (!qdisc_state) + return -ENOMEM; + + mutex_init(&qdisc_state->lock); + mlxsw_sp_port->qdisc = qdisc_state; + return 0; +} + +void mlxsw_sp_tc_qdisc_fini(struct mlxsw_sp_port *mlxsw_sp_port) +{ + mutex_destroy(&mlxsw_sp_port->qdisc->lock); + kfree(mlxsw_sp_port->qdisc); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c new file mode 100644 index 000000000..ab0aa1a61 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c @@ -0,0 +1,10617 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2016-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/types.h> +#include <linux/rhashtable.h> +#include <linux/bitops.h> +#include <linux/in6.h> +#include <linux/notifier.h> +#include <linux/inetdevice.h> +#include <linux/netdevice.h> +#include <linux/if_bridge.h> +#include <linux/socket.h> +#include <linux/route.h> +#include <linux/gcd.h> +#include <linux/if_macvlan.h> +#include <linux/refcount.h> +#include <linux/jhash.h> +#include <linux/net_namespace.h> +#include <linux/mutex.h> +#include <net/netevent.h> +#include <net/neighbour.h> +#include <net/arp.h> +#include <net/inet_dscp.h> +#include <net/ip_fib.h> +#include <net/ip6_fib.h> +#include <net/nexthop.h> +#include <net/fib_rules.h> +#include <net/ip_tunnels.h> +#include <net/l3mdev.h> +#include <net/addrconf.h> +#include <net/ndisc.h> +#include <net/ipv6.h> +#include <net/fib_notifier.h> +#include <net/switchdev.h> + +#include "spectrum.h" +#include "core.h" +#include "reg.h" +#include "spectrum_cnt.h" +#include "spectrum_dpipe.h" +#include "spectrum_ipip.h" +#include "spectrum_mr.h" +#include "spectrum_mr_tcam.h" +#include "spectrum_router.h" +#include "spectrum_span.h" + +struct mlxsw_sp_fib; +struct mlxsw_sp_vr; +struct mlxsw_sp_lpm_tree; +struct mlxsw_sp_rif_ops; + +struct mlxsw_sp_rif { + struct list_head nexthop_list; + struct list_head neigh_list; + struct net_device *dev; /* NULL for underlay RIF */ + struct mlxsw_sp_fid *fid; + unsigned char addr[ETH_ALEN]; + int mtu; + u16 rif_index; + u8 mac_profile_id; + u16 vr_id; + const struct mlxsw_sp_rif_ops *ops; + struct mlxsw_sp *mlxsw_sp; + + unsigned int counter_ingress; + bool counter_ingress_valid; + unsigned int counter_egress; + bool counter_egress_valid; +}; + +struct mlxsw_sp_rif_params { + struct net_device *dev; + union { + u16 system_port; + u16 lag_id; + }; + u16 vid; + bool lag; +}; + +struct mlxsw_sp_rif_subport { + struct mlxsw_sp_rif common; + refcount_t ref_count; + union { + u16 system_port; + u16 lag_id; + }; + u16 vid; + bool lag; +}; + +struct mlxsw_sp_rif_ipip_lb { + struct mlxsw_sp_rif common; + struct mlxsw_sp_rif_ipip_lb_config lb_config; + u16 ul_vr_id; /* Reserved for Spectrum-2. */ + u16 ul_rif_id; /* Reserved for Spectrum. */ +}; + +struct mlxsw_sp_rif_params_ipip_lb { + struct mlxsw_sp_rif_params common; + struct mlxsw_sp_rif_ipip_lb_config lb_config; +}; + +struct mlxsw_sp_rif_ops { + enum mlxsw_sp_rif_type type; + size_t rif_size; + + void (*setup)(struct mlxsw_sp_rif *rif, + const struct mlxsw_sp_rif_params *params); + int (*configure)(struct mlxsw_sp_rif *rif, + struct netlink_ext_ack *extack); + void (*deconfigure)(struct mlxsw_sp_rif *rif); + struct mlxsw_sp_fid * (*fid_get)(struct mlxsw_sp_rif *rif, + struct netlink_ext_ack *extack); + void (*fdb_del)(struct mlxsw_sp_rif *rif, const char *mac); +}; + +struct mlxsw_sp_rif_mac_profile { + unsigned char mac_prefix[ETH_ALEN]; + refcount_t ref_count; + u8 id; +}; + +struct mlxsw_sp_router_ops { + int (*init)(struct mlxsw_sp *mlxsw_sp); + int (*ipips_init)(struct mlxsw_sp *mlxsw_sp); +}; + +static struct mlxsw_sp_rif * +mlxsw_sp_rif_find_by_dev(const struct mlxsw_sp *mlxsw_sp, + const struct net_device *dev); +static void mlxsw_sp_rif_destroy(struct mlxsw_sp_rif *rif); +static void mlxsw_sp_lpm_tree_hold(struct mlxsw_sp_lpm_tree *lpm_tree); +static void mlxsw_sp_lpm_tree_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_lpm_tree *lpm_tree); +static int mlxsw_sp_vr_lpm_tree_bind(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_fib *fib, + u8 tree_id); +static int mlxsw_sp_vr_lpm_tree_unbind(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_fib *fib); + +static unsigned int * +mlxsw_sp_rif_p_counter_get(struct mlxsw_sp_rif *rif, + enum mlxsw_sp_rif_counter_dir dir) +{ + switch (dir) { + case MLXSW_SP_RIF_COUNTER_EGRESS: + return &rif->counter_egress; + case MLXSW_SP_RIF_COUNTER_INGRESS: + return &rif->counter_ingress; + } + return NULL; +} + +static bool +mlxsw_sp_rif_counter_valid_get(struct mlxsw_sp_rif *rif, + enum mlxsw_sp_rif_counter_dir dir) +{ + switch (dir) { + case MLXSW_SP_RIF_COUNTER_EGRESS: + return rif->counter_egress_valid; + case MLXSW_SP_RIF_COUNTER_INGRESS: + return rif->counter_ingress_valid; + } + return false; +} + +static void +mlxsw_sp_rif_counter_valid_set(struct mlxsw_sp_rif *rif, + enum mlxsw_sp_rif_counter_dir dir, + bool valid) +{ + switch (dir) { + case MLXSW_SP_RIF_COUNTER_EGRESS: + rif->counter_egress_valid = valid; + break; + case MLXSW_SP_RIF_COUNTER_INGRESS: + rif->counter_ingress_valid = valid; + break; + } +} + +static int mlxsw_sp_rif_counter_edit(struct mlxsw_sp *mlxsw_sp, u16 rif_index, + unsigned int counter_index, bool enable, + enum mlxsw_sp_rif_counter_dir dir) +{ + char ritr_pl[MLXSW_REG_RITR_LEN]; + bool is_egress = false; + int err; + + if (dir == MLXSW_SP_RIF_COUNTER_EGRESS) + is_egress = true; + mlxsw_reg_ritr_rif_pack(ritr_pl, rif_index); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); + if (err) + return err; + + mlxsw_reg_ritr_counter_pack(ritr_pl, counter_index, enable, + is_egress); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); +} + +int mlxsw_sp_rif_counter_value_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_rif *rif, + enum mlxsw_sp_rif_counter_dir dir, u64 *cnt) +{ + char ricnt_pl[MLXSW_REG_RICNT_LEN]; + unsigned int *p_counter_index; + bool valid; + int err; + + valid = mlxsw_sp_rif_counter_valid_get(rif, dir); + if (!valid) + return -EINVAL; + + p_counter_index = mlxsw_sp_rif_p_counter_get(rif, dir); + if (!p_counter_index) + return -EINVAL; + mlxsw_reg_ricnt_pack(ricnt_pl, *p_counter_index, + MLXSW_REG_RICNT_OPCODE_NOP); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ricnt), ricnt_pl); + if (err) + return err; + *cnt = mlxsw_reg_ricnt_good_unicast_packets_get(ricnt_pl); + return 0; +} + +struct mlxsw_sp_rif_counter_set_basic { + u64 good_unicast_packets; + u64 good_multicast_packets; + u64 good_broadcast_packets; + u64 good_unicast_bytes; + u64 good_multicast_bytes; + u64 good_broadcast_bytes; + u64 error_packets; + u64 discard_packets; + u64 error_bytes; + u64 discard_bytes; +}; + +static int +mlxsw_sp_rif_counter_fetch_clear(struct mlxsw_sp_rif *rif, + enum mlxsw_sp_rif_counter_dir dir, + struct mlxsw_sp_rif_counter_set_basic *set) +{ + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + char ricnt_pl[MLXSW_REG_RICNT_LEN]; + unsigned int *p_counter_index; + int err; + + if (!mlxsw_sp_rif_counter_valid_get(rif, dir)) + return -EINVAL; + + p_counter_index = mlxsw_sp_rif_p_counter_get(rif, dir); + if (!p_counter_index) + return -EINVAL; + + mlxsw_reg_ricnt_pack(ricnt_pl, *p_counter_index, + MLXSW_REG_RICNT_OPCODE_CLEAR); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ricnt), ricnt_pl); + if (err) + return err; + + if (!set) + return 0; + +#define MLXSW_SP_RIF_COUNTER_EXTRACT(NAME) \ + (set->NAME = mlxsw_reg_ricnt_ ## NAME ## _get(ricnt_pl)) + + MLXSW_SP_RIF_COUNTER_EXTRACT(good_unicast_packets); + MLXSW_SP_RIF_COUNTER_EXTRACT(good_multicast_packets); + MLXSW_SP_RIF_COUNTER_EXTRACT(good_broadcast_packets); + MLXSW_SP_RIF_COUNTER_EXTRACT(good_unicast_bytes); + MLXSW_SP_RIF_COUNTER_EXTRACT(good_multicast_bytes); + MLXSW_SP_RIF_COUNTER_EXTRACT(good_broadcast_bytes); + MLXSW_SP_RIF_COUNTER_EXTRACT(error_packets); + MLXSW_SP_RIF_COUNTER_EXTRACT(discard_packets); + MLXSW_SP_RIF_COUNTER_EXTRACT(error_bytes); + MLXSW_SP_RIF_COUNTER_EXTRACT(discard_bytes); + +#undef MLXSW_SP_RIF_COUNTER_EXTRACT + + return 0; +} + +static int mlxsw_sp_rif_counter_clear(struct mlxsw_sp *mlxsw_sp, + unsigned int counter_index) +{ + char ricnt_pl[MLXSW_REG_RICNT_LEN]; + + mlxsw_reg_ricnt_pack(ricnt_pl, counter_index, + MLXSW_REG_RICNT_OPCODE_CLEAR); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ricnt), ricnt_pl); +} + +int mlxsw_sp_rif_counter_alloc(struct mlxsw_sp_rif *rif, + enum mlxsw_sp_rif_counter_dir dir) +{ + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + unsigned int *p_counter_index; + int err; + + if (mlxsw_sp_rif_counter_valid_get(rif, dir)) + return 0; + + p_counter_index = mlxsw_sp_rif_p_counter_get(rif, dir); + if (!p_counter_index) + return -EINVAL; + + err = mlxsw_sp_counter_alloc(mlxsw_sp, MLXSW_SP_COUNTER_SUB_POOL_RIF, + p_counter_index); + if (err) + return err; + + err = mlxsw_sp_rif_counter_clear(mlxsw_sp, *p_counter_index); + if (err) + goto err_counter_clear; + + err = mlxsw_sp_rif_counter_edit(mlxsw_sp, rif->rif_index, + *p_counter_index, true, dir); + if (err) + goto err_counter_edit; + mlxsw_sp_rif_counter_valid_set(rif, dir, true); + return 0; + +err_counter_edit: +err_counter_clear: + mlxsw_sp_counter_free(mlxsw_sp, MLXSW_SP_COUNTER_SUB_POOL_RIF, + *p_counter_index); + return err; +} + +void mlxsw_sp_rif_counter_free(struct mlxsw_sp_rif *rif, + enum mlxsw_sp_rif_counter_dir dir) +{ + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + unsigned int *p_counter_index; + + if (!mlxsw_sp_rif_counter_valid_get(rif, dir)) + return; + + p_counter_index = mlxsw_sp_rif_p_counter_get(rif, dir); + if (WARN_ON(!p_counter_index)) + return; + mlxsw_sp_rif_counter_edit(mlxsw_sp, rif->rif_index, + *p_counter_index, false, dir); + mlxsw_sp_counter_free(mlxsw_sp, MLXSW_SP_COUNTER_SUB_POOL_RIF, + *p_counter_index); + mlxsw_sp_rif_counter_valid_set(rif, dir, false); +} + +static void mlxsw_sp_rif_counters_alloc(struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + struct devlink *devlink; + + devlink = priv_to_devlink(mlxsw_sp->core); + if (!devlink_dpipe_table_counter_enabled(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_ERIF)) + return; + mlxsw_sp_rif_counter_alloc(rif, MLXSW_SP_RIF_COUNTER_EGRESS); +} + +static void mlxsw_sp_rif_counters_free(struct mlxsw_sp_rif *rif) +{ + mlxsw_sp_rif_counter_free(rif, MLXSW_SP_RIF_COUNTER_EGRESS); +} + +#define MLXSW_SP_PREFIX_COUNT (sizeof(struct in6_addr) * BITS_PER_BYTE + 1) + +struct mlxsw_sp_prefix_usage { + DECLARE_BITMAP(b, MLXSW_SP_PREFIX_COUNT); +}; + +#define mlxsw_sp_prefix_usage_for_each(prefix, prefix_usage) \ + for_each_set_bit(prefix, (prefix_usage)->b, MLXSW_SP_PREFIX_COUNT) + +static bool +mlxsw_sp_prefix_usage_eq(struct mlxsw_sp_prefix_usage *prefix_usage1, + struct mlxsw_sp_prefix_usage *prefix_usage2) +{ + return !memcmp(prefix_usage1, prefix_usage2, sizeof(*prefix_usage1)); +} + +static void +mlxsw_sp_prefix_usage_cpy(struct mlxsw_sp_prefix_usage *prefix_usage1, + struct mlxsw_sp_prefix_usage *prefix_usage2) +{ + memcpy(prefix_usage1, prefix_usage2, sizeof(*prefix_usage1)); +} + +static void +mlxsw_sp_prefix_usage_set(struct mlxsw_sp_prefix_usage *prefix_usage, + unsigned char prefix_len) +{ + set_bit(prefix_len, prefix_usage->b); +} + +static void +mlxsw_sp_prefix_usage_clear(struct mlxsw_sp_prefix_usage *prefix_usage, + unsigned char prefix_len) +{ + clear_bit(prefix_len, prefix_usage->b); +} + +struct mlxsw_sp_fib_key { + unsigned char addr[sizeof(struct in6_addr)]; + unsigned char prefix_len; +}; + +enum mlxsw_sp_fib_entry_type { + MLXSW_SP_FIB_ENTRY_TYPE_REMOTE, + MLXSW_SP_FIB_ENTRY_TYPE_LOCAL, + MLXSW_SP_FIB_ENTRY_TYPE_TRAP, + MLXSW_SP_FIB_ENTRY_TYPE_BLACKHOLE, + MLXSW_SP_FIB_ENTRY_TYPE_UNREACHABLE, + + /* This is a special case of local delivery, where a packet should be + * decapsulated on reception. Note that there is no corresponding ENCAP, + * because that's a type of next hop, not of FIB entry. (There can be + * several next hops in a REMOTE entry, and some of them may be + * encapsulating entries.) + */ + MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP, + MLXSW_SP_FIB_ENTRY_TYPE_NVE_DECAP, +}; + +struct mlxsw_sp_nexthop_group_info; +struct mlxsw_sp_nexthop_group; +struct mlxsw_sp_fib_entry; + +struct mlxsw_sp_fib_node { + struct mlxsw_sp_fib_entry *fib_entry; + struct list_head list; + struct rhash_head ht_node; + struct mlxsw_sp_fib *fib; + struct mlxsw_sp_fib_key key; +}; + +struct mlxsw_sp_fib_entry_decap { + struct mlxsw_sp_ipip_entry *ipip_entry; + u32 tunnel_index; +}; + +struct mlxsw_sp_fib_entry { + struct mlxsw_sp_fib_node *fib_node; + enum mlxsw_sp_fib_entry_type type; + struct list_head nexthop_group_node; + struct mlxsw_sp_nexthop_group *nh_group; + struct mlxsw_sp_fib_entry_decap decap; /* Valid for decap entries. */ +}; + +struct mlxsw_sp_fib4_entry { + struct mlxsw_sp_fib_entry common; + struct fib_info *fi; + u32 tb_id; + dscp_t dscp; + u8 type; +}; + +struct mlxsw_sp_fib6_entry { + struct mlxsw_sp_fib_entry common; + struct list_head rt6_list; + unsigned int nrt6; +}; + +struct mlxsw_sp_rt6 { + struct list_head list; + struct fib6_info *rt; +}; + +struct mlxsw_sp_lpm_tree { + u8 id; /* tree ID */ + unsigned int ref_count; + enum mlxsw_sp_l3proto proto; + unsigned long prefix_ref_count[MLXSW_SP_PREFIX_COUNT]; + struct mlxsw_sp_prefix_usage prefix_usage; +}; + +struct mlxsw_sp_fib { + struct rhashtable ht; + struct list_head node_list; + struct mlxsw_sp_vr *vr; + struct mlxsw_sp_lpm_tree *lpm_tree; + enum mlxsw_sp_l3proto proto; +}; + +struct mlxsw_sp_vr { + u16 id; /* virtual router ID */ + u32 tb_id; /* kernel fib table id */ + unsigned int rif_count; + struct mlxsw_sp_fib *fib4; + struct mlxsw_sp_fib *fib6; + struct mlxsw_sp_mr_table *mr_table[MLXSW_SP_L3_PROTO_MAX]; + struct mlxsw_sp_rif *ul_rif; + refcount_t ul_rif_refcnt; +}; + +static const struct rhashtable_params mlxsw_sp_fib_ht_params; + +static struct mlxsw_sp_fib *mlxsw_sp_fib_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_vr *vr, + enum mlxsw_sp_l3proto proto) +{ + struct mlxsw_sp_lpm_tree *lpm_tree; + struct mlxsw_sp_fib *fib; + int err; + + lpm_tree = mlxsw_sp->router->lpm.proto_trees[proto]; + fib = kzalloc(sizeof(*fib), GFP_KERNEL); + if (!fib) + return ERR_PTR(-ENOMEM); + err = rhashtable_init(&fib->ht, &mlxsw_sp_fib_ht_params); + if (err) + goto err_rhashtable_init; + INIT_LIST_HEAD(&fib->node_list); + fib->proto = proto; + fib->vr = vr; + fib->lpm_tree = lpm_tree; + mlxsw_sp_lpm_tree_hold(lpm_tree); + err = mlxsw_sp_vr_lpm_tree_bind(mlxsw_sp, fib, lpm_tree->id); + if (err) + goto err_lpm_tree_bind; + return fib; + +err_lpm_tree_bind: + mlxsw_sp_lpm_tree_put(mlxsw_sp, lpm_tree); +err_rhashtable_init: + kfree(fib); + return ERR_PTR(err); +} + +static void mlxsw_sp_fib_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib *fib) +{ + mlxsw_sp_vr_lpm_tree_unbind(mlxsw_sp, fib); + mlxsw_sp_lpm_tree_put(mlxsw_sp, fib->lpm_tree); + WARN_ON(!list_empty(&fib->node_list)); + rhashtable_destroy(&fib->ht); + kfree(fib); +} + +static struct mlxsw_sp_lpm_tree * +mlxsw_sp_lpm_tree_find_unused(struct mlxsw_sp *mlxsw_sp) +{ + static struct mlxsw_sp_lpm_tree *lpm_tree; + int i; + + for (i = 0; i < mlxsw_sp->router->lpm.tree_count; i++) { + lpm_tree = &mlxsw_sp->router->lpm.trees[i]; + if (lpm_tree->ref_count == 0) + return lpm_tree; + } + return NULL; +} + +static int mlxsw_sp_lpm_tree_alloc(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_lpm_tree *lpm_tree) +{ + char ralta_pl[MLXSW_REG_RALTA_LEN]; + + mlxsw_reg_ralta_pack(ralta_pl, true, + (enum mlxsw_reg_ralxx_protocol) lpm_tree->proto, + lpm_tree->id); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralta), ralta_pl); +} + +static void mlxsw_sp_lpm_tree_free(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_lpm_tree *lpm_tree) +{ + char ralta_pl[MLXSW_REG_RALTA_LEN]; + + mlxsw_reg_ralta_pack(ralta_pl, false, + (enum mlxsw_reg_ralxx_protocol) lpm_tree->proto, + lpm_tree->id); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralta), ralta_pl); +} + +static int +mlxsw_sp_lpm_tree_left_struct_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_prefix_usage *prefix_usage, + struct mlxsw_sp_lpm_tree *lpm_tree) +{ + char ralst_pl[MLXSW_REG_RALST_LEN]; + u8 root_bin = 0; + u8 prefix; + u8 last_prefix = MLXSW_REG_RALST_BIN_NO_CHILD; + + mlxsw_sp_prefix_usage_for_each(prefix, prefix_usage) + root_bin = prefix; + + mlxsw_reg_ralst_pack(ralst_pl, root_bin, lpm_tree->id); + mlxsw_sp_prefix_usage_for_each(prefix, prefix_usage) { + if (prefix == 0) + continue; + mlxsw_reg_ralst_bin_pack(ralst_pl, prefix, last_prefix, + MLXSW_REG_RALST_BIN_NO_CHILD); + last_prefix = prefix; + } + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralst), ralst_pl); +} + +static struct mlxsw_sp_lpm_tree * +mlxsw_sp_lpm_tree_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_prefix_usage *prefix_usage, + enum mlxsw_sp_l3proto proto) +{ + struct mlxsw_sp_lpm_tree *lpm_tree; + int err; + + lpm_tree = mlxsw_sp_lpm_tree_find_unused(mlxsw_sp); + if (!lpm_tree) + return ERR_PTR(-EBUSY); + lpm_tree->proto = proto; + err = mlxsw_sp_lpm_tree_alloc(mlxsw_sp, lpm_tree); + if (err) + return ERR_PTR(err); + + err = mlxsw_sp_lpm_tree_left_struct_set(mlxsw_sp, prefix_usage, + lpm_tree); + if (err) + goto err_left_struct_set; + memcpy(&lpm_tree->prefix_usage, prefix_usage, + sizeof(lpm_tree->prefix_usage)); + memset(&lpm_tree->prefix_ref_count, 0, + sizeof(lpm_tree->prefix_ref_count)); + lpm_tree->ref_count = 1; + return lpm_tree; + +err_left_struct_set: + mlxsw_sp_lpm_tree_free(mlxsw_sp, lpm_tree); + return ERR_PTR(err); +} + +static void mlxsw_sp_lpm_tree_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_lpm_tree *lpm_tree) +{ + mlxsw_sp_lpm_tree_free(mlxsw_sp, lpm_tree); +} + +static struct mlxsw_sp_lpm_tree * +mlxsw_sp_lpm_tree_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_prefix_usage *prefix_usage, + enum mlxsw_sp_l3proto proto) +{ + struct mlxsw_sp_lpm_tree *lpm_tree; + int i; + + for (i = 0; i < mlxsw_sp->router->lpm.tree_count; i++) { + lpm_tree = &mlxsw_sp->router->lpm.trees[i]; + if (lpm_tree->ref_count != 0 && + lpm_tree->proto == proto && + mlxsw_sp_prefix_usage_eq(&lpm_tree->prefix_usage, + prefix_usage)) { + mlxsw_sp_lpm_tree_hold(lpm_tree); + return lpm_tree; + } + } + return mlxsw_sp_lpm_tree_create(mlxsw_sp, prefix_usage, proto); +} + +static void mlxsw_sp_lpm_tree_hold(struct mlxsw_sp_lpm_tree *lpm_tree) +{ + lpm_tree->ref_count++; +} + +static void mlxsw_sp_lpm_tree_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_lpm_tree *lpm_tree) +{ + if (--lpm_tree->ref_count == 0) + mlxsw_sp_lpm_tree_destroy(mlxsw_sp, lpm_tree); +} + +#define MLXSW_SP_LPM_TREE_MIN 1 /* tree 0 is reserved */ + +static int mlxsw_sp_lpm_init(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_prefix_usage req_prefix_usage = {{ 0 } }; + struct mlxsw_sp_lpm_tree *lpm_tree; + u64 max_trees; + int err, i; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, MAX_LPM_TREES)) + return -EIO; + + max_trees = MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_LPM_TREES); + mlxsw_sp->router->lpm.tree_count = max_trees - MLXSW_SP_LPM_TREE_MIN; + mlxsw_sp->router->lpm.trees = kcalloc(mlxsw_sp->router->lpm.tree_count, + sizeof(struct mlxsw_sp_lpm_tree), + GFP_KERNEL); + if (!mlxsw_sp->router->lpm.trees) + return -ENOMEM; + + for (i = 0; i < mlxsw_sp->router->lpm.tree_count; i++) { + lpm_tree = &mlxsw_sp->router->lpm.trees[i]; + lpm_tree->id = i + MLXSW_SP_LPM_TREE_MIN; + } + + lpm_tree = mlxsw_sp_lpm_tree_get(mlxsw_sp, &req_prefix_usage, + MLXSW_SP_L3_PROTO_IPV4); + if (IS_ERR(lpm_tree)) { + err = PTR_ERR(lpm_tree); + goto err_ipv4_tree_get; + } + mlxsw_sp->router->lpm.proto_trees[MLXSW_SP_L3_PROTO_IPV4] = lpm_tree; + + lpm_tree = mlxsw_sp_lpm_tree_get(mlxsw_sp, &req_prefix_usage, + MLXSW_SP_L3_PROTO_IPV6); + if (IS_ERR(lpm_tree)) { + err = PTR_ERR(lpm_tree); + goto err_ipv6_tree_get; + } + mlxsw_sp->router->lpm.proto_trees[MLXSW_SP_L3_PROTO_IPV6] = lpm_tree; + + return 0; + +err_ipv6_tree_get: + lpm_tree = mlxsw_sp->router->lpm.proto_trees[MLXSW_SP_L3_PROTO_IPV4]; + mlxsw_sp_lpm_tree_put(mlxsw_sp, lpm_tree); +err_ipv4_tree_get: + kfree(mlxsw_sp->router->lpm.trees); + return err; +} + +static void mlxsw_sp_lpm_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_lpm_tree *lpm_tree; + + lpm_tree = mlxsw_sp->router->lpm.proto_trees[MLXSW_SP_L3_PROTO_IPV6]; + mlxsw_sp_lpm_tree_put(mlxsw_sp, lpm_tree); + + lpm_tree = mlxsw_sp->router->lpm.proto_trees[MLXSW_SP_L3_PROTO_IPV4]; + mlxsw_sp_lpm_tree_put(mlxsw_sp, lpm_tree); + + kfree(mlxsw_sp->router->lpm.trees); +} + +static bool mlxsw_sp_vr_is_used(const struct mlxsw_sp_vr *vr) +{ + return !!vr->fib4 || !!vr->fib6 || + !!vr->mr_table[MLXSW_SP_L3_PROTO_IPV4] || + !!vr->mr_table[MLXSW_SP_L3_PROTO_IPV6]; +} + +static struct mlxsw_sp_vr *mlxsw_sp_vr_find_unused(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_vr *vr; + int i; + + for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_VRS); i++) { + vr = &mlxsw_sp->router->vrs[i]; + if (!mlxsw_sp_vr_is_used(vr)) + return vr; + } + return NULL; +} + +static int mlxsw_sp_vr_lpm_tree_bind(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_fib *fib, u8 tree_id) +{ + char raltb_pl[MLXSW_REG_RALTB_LEN]; + + mlxsw_reg_raltb_pack(raltb_pl, fib->vr->id, + (enum mlxsw_reg_ralxx_protocol) fib->proto, + tree_id); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(raltb), raltb_pl); +} + +static int mlxsw_sp_vr_lpm_tree_unbind(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_fib *fib) +{ + char raltb_pl[MLXSW_REG_RALTB_LEN]; + + /* Bind to tree 0 which is default */ + mlxsw_reg_raltb_pack(raltb_pl, fib->vr->id, + (enum mlxsw_reg_ralxx_protocol) fib->proto, 0); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(raltb), raltb_pl); +} + +static u32 mlxsw_sp_fix_tb_id(u32 tb_id) +{ + /* For our purpose, squash main, default and local tables into one */ + if (tb_id == RT_TABLE_LOCAL || tb_id == RT_TABLE_DEFAULT) + tb_id = RT_TABLE_MAIN; + return tb_id; +} + +static struct mlxsw_sp_vr *mlxsw_sp_vr_find(struct mlxsw_sp *mlxsw_sp, + u32 tb_id) +{ + struct mlxsw_sp_vr *vr; + int i; + + tb_id = mlxsw_sp_fix_tb_id(tb_id); + + for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_VRS); i++) { + vr = &mlxsw_sp->router->vrs[i]; + if (mlxsw_sp_vr_is_used(vr) && vr->tb_id == tb_id) + return vr; + } + return NULL; +} + +int mlxsw_sp_router_tb_id_vr_id(struct mlxsw_sp *mlxsw_sp, u32 tb_id, + u16 *vr_id) +{ + struct mlxsw_sp_vr *vr; + int err = 0; + + mutex_lock(&mlxsw_sp->router->lock); + vr = mlxsw_sp_vr_find(mlxsw_sp, tb_id); + if (!vr) { + err = -ESRCH; + goto out; + } + *vr_id = vr->id; +out: + mutex_unlock(&mlxsw_sp->router->lock); + return err; +} + +static struct mlxsw_sp_fib *mlxsw_sp_vr_fib(const struct mlxsw_sp_vr *vr, + enum mlxsw_sp_l3proto proto) +{ + switch (proto) { + case MLXSW_SP_L3_PROTO_IPV4: + return vr->fib4; + case MLXSW_SP_L3_PROTO_IPV6: + return vr->fib6; + } + return NULL; +} + +static struct mlxsw_sp_vr *mlxsw_sp_vr_create(struct mlxsw_sp *mlxsw_sp, + u32 tb_id, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_mr_table *mr4_table, *mr6_table; + struct mlxsw_sp_fib *fib4; + struct mlxsw_sp_fib *fib6; + struct mlxsw_sp_vr *vr; + int err; + + vr = mlxsw_sp_vr_find_unused(mlxsw_sp); + if (!vr) { + NL_SET_ERR_MSG_MOD(extack, "Exceeded number of supported virtual routers"); + return ERR_PTR(-EBUSY); + } + fib4 = mlxsw_sp_fib_create(mlxsw_sp, vr, MLXSW_SP_L3_PROTO_IPV4); + if (IS_ERR(fib4)) + return ERR_CAST(fib4); + fib6 = mlxsw_sp_fib_create(mlxsw_sp, vr, MLXSW_SP_L3_PROTO_IPV6); + if (IS_ERR(fib6)) { + err = PTR_ERR(fib6); + goto err_fib6_create; + } + mr4_table = mlxsw_sp_mr_table_create(mlxsw_sp, vr->id, + MLXSW_SP_L3_PROTO_IPV4); + if (IS_ERR(mr4_table)) { + err = PTR_ERR(mr4_table); + goto err_mr4_table_create; + } + mr6_table = mlxsw_sp_mr_table_create(mlxsw_sp, vr->id, + MLXSW_SP_L3_PROTO_IPV6); + if (IS_ERR(mr6_table)) { + err = PTR_ERR(mr6_table); + goto err_mr6_table_create; + } + + vr->fib4 = fib4; + vr->fib6 = fib6; + vr->mr_table[MLXSW_SP_L3_PROTO_IPV4] = mr4_table; + vr->mr_table[MLXSW_SP_L3_PROTO_IPV6] = mr6_table; + vr->tb_id = tb_id; + return vr; + +err_mr6_table_create: + mlxsw_sp_mr_table_destroy(mr4_table); +err_mr4_table_create: + mlxsw_sp_fib_destroy(mlxsw_sp, fib6); +err_fib6_create: + mlxsw_sp_fib_destroy(mlxsw_sp, fib4); + return ERR_PTR(err); +} + +static void mlxsw_sp_vr_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_vr *vr) +{ + mlxsw_sp_mr_table_destroy(vr->mr_table[MLXSW_SP_L3_PROTO_IPV6]); + vr->mr_table[MLXSW_SP_L3_PROTO_IPV6] = NULL; + mlxsw_sp_mr_table_destroy(vr->mr_table[MLXSW_SP_L3_PROTO_IPV4]); + vr->mr_table[MLXSW_SP_L3_PROTO_IPV4] = NULL; + mlxsw_sp_fib_destroy(mlxsw_sp, vr->fib6); + vr->fib6 = NULL; + mlxsw_sp_fib_destroy(mlxsw_sp, vr->fib4); + vr->fib4 = NULL; +} + +static struct mlxsw_sp_vr *mlxsw_sp_vr_get(struct mlxsw_sp *mlxsw_sp, u32 tb_id, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_vr *vr; + + tb_id = mlxsw_sp_fix_tb_id(tb_id); + vr = mlxsw_sp_vr_find(mlxsw_sp, tb_id); + if (!vr) + vr = mlxsw_sp_vr_create(mlxsw_sp, tb_id, extack); + return vr; +} + +static void mlxsw_sp_vr_put(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_vr *vr) +{ + if (!vr->rif_count && list_empty(&vr->fib4->node_list) && + list_empty(&vr->fib6->node_list) && + mlxsw_sp_mr_table_empty(vr->mr_table[MLXSW_SP_L3_PROTO_IPV4]) && + mlxsw_sp_mr_table_empty(vr->mr_table[MLXSW_SP_L3_PROTO_IPV6])) + mlxsw_sp_vr_destroy(mlxsw_sp, vr); +} + +static bool +mlxsw_sp_vr_lpm_tree_should_replace(struct mlxsw_sp_vr *vr, + enum mlxsw_sp_l3proto proto, u8 tree_id) +{ + struct mlxsw_sp_fib *fib = mlxsw_sp_vr_fib(vr, proto); + + if (!mlxsw_sp_vr_is_used(vr)) + return false; + if (fib->lpm_tree->id == tree_id) + return true; + return false; +} + +static int mlxsw_sp_vr_lpm_tree_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib *fib, + struct mlxsw_sp_lpm_tree *new_tree) +{ + struct mlxsw_sp_lpm_tree *old_tree = fib->lpm_tree; + int err; + + fib->lpm_tree = new_tree; + mlxsw_sp_lpm_tree_hold(new_tree); + err = mlxsw_sp_vr_lpm_tree_bind(mlxsw_sp, fib, new_tree->id); + if (err) + goto err_tree_bind; + mlxsw_sp_lpm_tree_put(mlxsw_sp, old_tree); + return 0; + +err_tree_bind: + mlxsw_sp_lpm_tree_put(mlxsw_sp, new_tree); + fib->lpm_tree = old_tree; + return err; +} + +static int mlxsw_sp_vrs_lpm_tree_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib *fib, + struct mlxsw_sp_lpm_tree *new_tree) +{ + enum mlxsw_sp_l3proto proto = fib->proto; + struct mlxsw_sp_lpm_tree *old_tree; + u8 old_id, new_id = new_tree->id; + struct mlxsw_sp_vr *vr; + int i, err; + + old_tree = mlxsw_sp->router->lpm.proto_trees[proto]; + old_id = old_tree->id; + + for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_VRS); i++) { + vr = &mlxsw_sp->router->vrs[i]; + if (!mlxsw_sp_vr_lpm_tree_should_replace(vr, proto, old_id)) + continue; + err = mlxsw_sp_vr_lpm_tree_replace(mlxsw_sp, + mlxsw_sp_vr_fib(vr, proto), + new_tree); + if (err) + goto err_tree_replace; + } + + memcpy(new_tree->prefix_ref_count, old_tree->prefix_ref_count, + sizeof(new_tree->prefix_ref_count)); + mlxsw_sp->router->lpm.proto_trees[proto] = new_tree; + mlxsw_sp_lpm_tree_put(mlxsw_sp, old_tree); + + return 0; + +err_tree_replace: + for (i--; i >= 0; i--) { + if (!mlxsw_sp_vr_lpm_tree_should_replace(vr, proto, new_id)) + continue; + mlxsw_sp_vr_lpm_tree_replace(mlxsw_sp, + mlxsw_sp_vr_fib(vr, proto), + old_tree); + } + return err; +} + +static int mlxsw_sp_vrs_init(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_vr *vr; + u64 max_vrs; + int i; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, MAX_VRS)) + return -EIO; + + max_vrs = MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_VRS); + mlxsw_sp->router->vrs = kcalloc(max_vrs, sizeof(struct mlxsw_sp_vr), + GFP_KERNEL); + if (!mlxsw_sp->router->vrs) + return -ENOMEM; + + for (i = 0; i < max_vrs; i++) { + vr = &mlxsw_sp->router->vrs[i]; + vr->id = i; + } + + return 0; +} + +static void mlxsw_sp_router_fib_flush(struct mlxsw_sp *mlxsw_sp); + +static void mlxsw_sp_vrs_fini(struct mlxsw_sp *mlxsw_sp) +{ + /* At this stage we're guaranteed not to have new incoming + * FIB notifications and the work queue is free from FIBs + * sitting on top of mlxsw netdevs. However, we can still + * have other FIBs queued. Flush the queue before flushing + * the device's tables. No need for locks, as we're the only + * writer. + */ + mlxsw_core_flush_owq(); + mlxsw_sp_router_fib_flush(mlxsw_sp); + kfree(mlxsw_sp->router->vrs); +} + +u32 mlxsw_sp_ipip_dev_ul_tb_id(const struct net_device *ol_dev) +{ + struct net_device *d; + u32 tb_id; + + rcu_read_lock(); + d = mlxsw_sp_ipip_netdev_ul_dev_get(ol_dev); + if (d) + tb_id = l3mdev_fib_table(d) ? : RT_TABLE_MAIN; + else + tb_id = RT_TABLE_MAIN; + rcu_read_unlock(); + + return tb_id; +} + +static struct mlxsw_sp_rif * +mlxsw_sp_rif_create(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_rif_params *params, + struct netlink_ext_ack *extack); + +static struct mlxsw_sp_rif_ipip_lb * +mlxsw_sp_ipip_ol_ipip_lb_create(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_ipip_type ipipt, + struct net_device *ol_dev, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_rif_params_ipip_lb lb_params; + const struct mlxsw_sp_ipip_ops *ipip_ops; + struct mlxsw_sp_rif *rif; + + ipip_ops = mlxsw_sp->router->ipip_ops_arr[ipipt]; + lb_params = (struct mlxsw_sp_rif_params_ipip_lb) { + .common.dev = ol_dev, + .common.lag = false, + .lb_config = ipip_ops->ol_loopback_config(mlxsw_sp, ol_dev), + }; + + rif = mlxsw_sp_rif_create(mlxsw_sp, &lb_params.common, extack); + if (IS_ERR(rif)) + return ERR_CAST(rif); + return container_of(rif, struct mlxsw_sp_rif_ipip_lb, common); +} + +static struct mlxsw_sp_ipip_entry * +mlxsw_sp_ipip_entry_alloc(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_ipip_type ipipt, + struct net_device *ol_dev) +{ + const struct mlxsw_sp_ipip_ops *ipip_ops; + struct mlxsw_sp_ipip_entry *ipip_entry; + struct mlxsw_sp_ipip_entry *ret = NULL; + int err; + + ipip_ops = mlxsw_sp->router->ipip_ops_arr[ipipt]; + ipip_entry = kzalloc(sizeof(*ipip_entry), GFP_KERNEL); + if (!ipip_entry) + return ERR_PTR(-ENOMEM); + + ipip_entry->ol_lb = mlxsw_sp_ipip_ol_ipip_lb_create(mlxsw_sp, ipipt, + ol_dev, NULL); + if (IS_ERR(ipip_entry->ol_lb)) { + ret = ERR_CAST(ipip_entry->ol_lb); + goto err_ol_ipip_lb_create; + } + + ipip_entry->ipipt = ipipt; + ipip_entry->ol_dev = ol_dev; + ipip_entry->parms = ipip_ops->parms_init(ol_dev); + + err = ipip_ops->rem_ip_addr_set(mlxsw_sp, ipip_entry); + if (err) { + ret = ERR_PTR(err); + goto err_rem_ip_addr_set; + } + + return ipip_entry; + +err_rem_ip_addr_set: + mlxsw_sp_rif_destroy(&ipip_entry->ol_lb->common); +err_ol_ipip_lb_create: + kfree(ipip_entry); + return ret; +} + +static void mlxsw_sp_ipip_entry_dealloc(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + const struct mlxsw_sp_ipip_ops *ipip_ops = + mlxsw_sp->router->ipip_ops_arr[ipip_entry->ipipt]; + + ipip_ops->rem_ip_addr_unset(mlxsw_sp, ipip_entry); + mlxsw_sp_rif_destroy(&ipip_entry->ol_lb->common); + kfree(ipip_entry); +} + +static bool +mlxsw_sp_ipip_entry_saddr_matches(struct mlxsw_sp *mlxsw_sp, + const enum mlxsw_sp_l3proto ul_proto, + union mlxsw_sp_l3addr saddr, + u32 ul_tb_id, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + u32 tun_ul_tb_id = mlxsw_sp_ipip_dev_ul_tb_id(ipip_entry->ol_dev); + enum mlxsw_sp_ipip_type ipipt = ipip_entry->ipipt; + union mlxsw_sp_l3addr tun_saddr; + + if (mlxsw_sp->router->ipip_ops_arr[ipipt]->ul_proto != ul_proto) + return false; + + tun_saddr = mlxsw_sp_ipip_netdev_saddr(ul_proto, ipip_entry->ol_dev); + return tun_ul_tb_id == ul_tb_id && + mlxsw_sp_l3addr_eq(&tun_saddr, &saddr); +} + +static int mlxsw_sp_ipip_decap_parsing_depth_inc(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_ipip_type ipipt) +{ + const struct mlxsw_sp_ipip_ops *ipip_ops; + + ipip_ops = mlxsw_sp->router->ipip_ops_arr[ipipt]; + + /* Not all tunnels require to increase the default pasing depth + * (96 bytes). + */ + if (ipip_ops->inc_parsing_depth) + return mlxsw_sp_parsing_depth_inc(mlxsw_sp); + + return 0; +} + +static void mlxsw_sp_ipip_decap_parsing_depth_dec(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_ipip_type ipipt) +{ + const struct mlxsw_sp_ipip_ops *ipip_ops = + mlxsw_sp->router->ipip_ops_arr[ipipt]; + + if (ipip_ops->inc_parsing_depth) + mlxsw_sp_parsing_depth_dec(mlxsw_sp); +} + +static int +mlxsw_sp_fib_entry_decap_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + u32 tunnel_index; + int err; + + err = mlxsw_sp_kvdl_alloc(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, + 1, &tunnel_index); + if (err) + return err; + + err = mlxsw_sp_ipip_decap_parsing_depth_inc(mlxsw_sp, + ipip_entry->ipipt); + if (err) + goto err_parsing_depth_inc; + + ipip_entry->decap_fib_entry = fib_entry; + fib_entry->decap.ipip_entry = ipip_entry; + fib_entry->decap.tunnel_index = tunnel_index; + + return 0; + +err_parsing_depth_inc: + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, 1, + fib_entry->decap.tunnel_index); + return err; +} + +static void mlxsw_sp_fib_entry_decap_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + enum mlxsw_sp_ipip_type ipipt = fib_entry->decap.ipip_entry->ipipt; + + /* Unlink this node from the IPIP entry that it's the decap entry of. */ + fib_entry->decap.ipip_entry->decap_fib_entry = NULL; + fib_entry->decap.ipip_entry = NULL; + mlxsw_sp_ipip_decap_parsing_depth_dec(mlxsw_sp, ipipt); + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, + 1, fib_entry->decap.tunnel_index); +} + +static struct mlxsw_sp_fib_node * +mlxsw_sp_fib_node_lookup(struct mlxsw_sp_fib *fib, const void *addr, + size_t addr_len, unsigned char prefix_len); +static int mlxsw_sp_fib_entry_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry); + +static void +mlxsw_sp_ipip_entry_demote_decap(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + struct mlxsw_sp_fib_entry *fib_entry = ipip_entry->decap_fib_entry; + + mlxsw_sp_fib_entry_decap_fini(mlxsw_sp, fib_entry); + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_TRAP; + + mlxsw_sp_fib_entry_update(mlxsw_sp, fib_entry); +} + +static void +mlxsw_sp_ipip_entry_promote_decap(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + struct mlxsw_sp_fib_entry *decap_fib_entry) +{ + if (mlxsw_sp_fib_entry_decap_init(mlxsw_sp, decap_fib_entry, + ipip_entry)) + return; + decap_fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP; + + if (mlxsw_sp_fib_entry_update(mlxsw_sp, decap_fib_entry)) + mlxsw_sp_ipip_entry_demote_decap(mlxsw_sp, ipip_entry); +} + +static struct mlxsw_sp_fib_entry * +mlxsw_sp_router_ip2me_fib_entry_find(struct mlxsw_sp *mlxsw_sp, u32 tb_id, + enum mlxsw_sp_l3proto proto, + const union mlxsw_sp_l3addr *addr, + enum mlxsw_sp_fib_entry_type type) +{ + struct mlxsw_sp_fib_node *fib_node; + unsigned char addr_prefix_len; + struct mlxsw_sp_fib *fib; + struct mlxsw_sp_vr *vr; + const void *addrp; + size_t addr_len; + u32 addr4; + + vr = mlxsw_sp_vr_find(mlxsw_sp, tb_id); + if (!vr) + return NULL; + fib = mlxsw_sp_vr_fib(vr, proto); + + switch (proto) { + case MLXSW_SP_L3_PROTO_IPV4: + addr4 = be32_to_cpu(addr->addr4); + addrp = &addr4; + addr_len = 4; + addr_prefix_len = 32; + break; + case MLXSW_SP_L3_PROTO_IPV6: + addrp = &addr->addr6; + addr_len = 16; + addr_prefix_len = 128; + break; + default: + WARN_ON(1); + return NULL; + } + + fib_node = mlxsw_sp_fib_node_lookup(fib, addrp, addr_len, + addr_prefix_len); + if (!fib_node || fib_node->fib_entry->type != type) + return NULL; + + return fib_node->fib_entry; +} + +/* Given an IPIP entry, find the corresponding decap route. */ +static struct mlxsw_sp_fib_entry * +mlxsw_sp_ipip_entry_find_decap(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + static struct mlxsw_sp_fib_node *fib_node; + const struct mlxsw_sp_ipip_ops *ipip_ops; + unsigned char saddr_prefix_len; + union mlxsw_sp_l3addr saddr; + struct mlxsw_sp_fib *ul_fib; + struct mlxsw_sp_vr *ul_vr; + const void *saddrp; + size_t saddr_len; + u32 ul_tb_id; + u32 saddr4; + + ipip_ops = mlxsw_sp->router->ipip_ops_arr[ipip_entry->ipipt]; + + ul_tb_id = mlxsw_sp_ipip_dev_ul_tb_id(ipip_entry->ol_dev); + ul_vr = mlxsw_sp_vr_find(mlxsw_sp, ul_tb_id); + if (!ul_vr) + return NULL; + + ul_fib = mlxsw_sp_vr_fib(ul_vr, ipip_ops->ul_proto); + saddr = mlxsw_sp_ipip_netdev_saddr(ipip_ops->ul_proto, + ipip_entry->ol_dev); + + switch (ipip_ops->ul_proto) { + case MLXSW_SP_L3_PROTO_IPV4: + saddr4 = be32_to_cpu(saddr.addr4); + saddrp = &saddr4; + saddr_len = 4; + saddr_prefix_len = 32; + break; + case MLXSW_SP_L3_PROTO_IPV6: + saddrp = &saddr.addr6; + saddr_len = 16; + saddr_prefix_len = 128; + break; + default: + WARN_ON(1); + return NULL; + } + + fib_node = mlxsw_sp_fib_node_lookup(ul_fib, saddrp, saddr_len, + saddr_prefix_len); + if (!fib_node || + fib_node->fib_entry->type != MLXSW_SP_FIB_ENTRY_TYPE_TRAP) + return NULL; + + return fib_node->fib_entry; +} + +static struct mlxsw_sp_ipip_entry * +mlxsw_sp_ipip_entry_create(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_ipip_type ipipt, + struct net_device *ol_dev) +{ + struct mlxsw_sp_ipip_entry *ipip_entry; + + ipip_entry = mlxsw_sp_ipip_entry_alloc(mlxsw_sp, ipipt, ol_dev); + if (IS_ERR(ipip_entry)) + return ipip_entry; + + list_add_tail(&ipip_entry->ipip_list_node, + &mlxsw_sp->router->ipip_list); + + return ipip_entry; +} + +static void +mlxsw_sp_ipip_entry_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + list_del(&ipip_entry->ipip_list_node); + mlxsw_sp_ipip_entry_dealloc(mlxsw_sp, ipip_entry); +} + +static bool +mlxsw_sp_ipip_entry_matches_decap(struct mlxsw_sp *mlxsw_sp, + const struct net_device *ul_dev, + enum mlxsw_sp_l3proto ul_proto, + union mlxsw_sp_l3addr ul_dip, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + u32 ul_tb_id = l3mdev_fib_table(ul_dev) ? : RT_TABLE_MAIN; + enum mlxsw_sp_ipip_type ipipt = ipip_entry->ipipt; + + if (mlxsw_sp->router->ipip_ops_arr[ipipt]->ul_proto != ul_proto) + return false; + + return mlxsw_sp_ipip_entry_saddr_matches(mlxsw_sp, ul_proto, ul_dip, + ul_tb_id, ipip_entry); +} + +/* Given decap parameters, find the corresponding IPIP entry. */ +static struct mlxsw_sp_ipip_entry * +mlxsw_sp_ipip_entry_find_by_decap(struct mlxsw_sp *mlxsw_sp, int ul_dev_ifindex, + enum mlxsw_sp_l3proto ul_proto, + union mlxsw_sp_l3addr ul_dip) +{ + struct mlxsw_sp_ipip_entry *ipip_entry = NULL; + struct net_device *ul_dev; + + rcu_read_lock(); + + ul_dev = dev_get_by_index_rcu(mlxsw_sp_net(mlxsw_sp), ul_dev_ifindex); + if (!ul_dev) + goto out_unlock; + + list_for_each_entry(ipip_entry, &mlxsw_sp->router->ipip_list, + ipip_list_node) + if (mlxsw_sp_ipip_entry_matches_decap(mlxsw_sp, ul_dev, + ul_proto, ul_dip, + ipip_entry)) + goto out_unlock; + + rcu_read_unlock(); + + return NULL; + +out_unlock: + rcu_read_unlock(); + return ipip_entry; +} + +static bool mlxsw_sp_netdev_ipip_type(const struct mlxsw_sp *mlxsw_sp, + const struct net_device *dev, + enum mlxsw_sp_ipip_type *p_type) +{ + struct mlxsw_sp_router *router = mlxsw_sp->router; + const struct mlxsw_sp_ipip_ops *ipip_ops; + enum mlxsw_sp_ipip_type ipipt; + + for (ipipt = 0; ipipt < MLXSW_SP_IPIP_TYPE_MAX; ++ipipt) { + ipip_ops = router->ipip_ops_arr[ipipt]; + if (dev->type == ipip_ops->dev_type) { + if (p_type) + *p_type = ipipt; + return true; + } + } + return false; +} + +static bool mlxsw_sp_netdev_is_ipip_ol(const struct mlxsw_sp *mlxsw_sp, + const struct net_device *dev) +{ + return mlxsw_sp_netdev_ipip_type(mlxsw_sp, dev, NULL); +} + +static struct mlxsw_sp_ipip_entry * +mlxsw_sp_ipip_entry_find_by_ol_dev(struct mlxsw_sp *mlxsw_sp, + const struct net_device *ol_dev) +{ + struct mlxsw_sp_ipip_entry *ipip_entry; + + list_for_each_entry(ipip_entry, &mlxsw_sp->router->ipip_list, + ipip_list_node) + if (ipip_entry->ol_dev == ol_dev) + return ipip_entry; + + return NULL; +} + +static struct mlxsw_sp_ipip_entry * +mlxsw_sp_ipip_entry_find_by_ul_dev(const struct mlxsw_sp *mlxsw_sp, + const struct net_device *ul_dev, + struct mlxsw_sp_ipip_entry *start) +{ + struct mlxsw_sp_ipip_entry *ipip_entry; + + ipip_entry = list_prepare_entry(start, &mlxsw_sp->router->ipip_list, + ipip_list_node); + list_for_each_entry_continue(ipip_entry, &mlxsw_sp->router->ipip_list, + ipip_list_node) { + struct net_device *ol_dev = ipip_entry->ol_dev; + struct net_device *ipip_ul_dev; + + rcu_read_lock(); + ipip_ul_dev = mlxsw_sp_ipip_netdev_ul_dev_get(ol_dev); + rcu_read_unlock(); + + if (ipip_ul_dev == ul_dev) + return ipip_entry; + } + + return NULL; +} + +static bool mlxsw_sp_netdev_is_ipip_ul(struct mlxsw_sp *mlxsw_sp, + const struct net_device *dev) +{ + return mlxsw_sp_ipip_entry_find_by_ul_dev(mlxsw_sp, dev, NULL); +} + +static bool mlxsw_sp_netdevice_ipip_can_offload(struct mlxsw_sp *mlxsw_sp, + const struct net_device *ol_dev, + enum mlxsw_sp_ipip_type ipipt) +{ + const struct mlxsw_sp_ipip_ops *ops + = mlxsw_sp->router->ipip_ops_arr[ipipt]; + + return ops->can_offload(mlxsw_sp, ol_dev); +} + +static int mlxsw_sp_netdevice_ipip_ol_reg_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *ol_dev) +{ + enum mlxsw_sp_ipip_type ipipt = MLXSW_SP_IPIP_TYPE_MAX; + struct mlxsw_sp_ipip_entry *ipip_entry; + enum mlxsw_sp_l3proto ul_proto; + union mlxsw_sp_l3addr saddr; + u32 ul_tb_id; + + mlxsw_sp_netdev_ipip_type(mlxsw_sp, ol_dev, &ipipt); + if (mlxsw_sp_netdevice_ipip_can_offload(mlxsw_sp, ol_dev, ipipt)) { + ul_tb_id = mlxsw_sp_ipip_dev_ul_tb_id(ol_dev); + ul_proto = mlxsw_sp->router->ipip_ops_arr[ipipt]->ul_proto; + saddr = mlxsw_sp_ipip_netdev_saddr(ul_proto, ol_dev); + if (!mlxsw_sp_ipip_demote_tunnel_by_saddr(mlxsw_sp, ul_proto, + saddr, ul_tb_id, + NULL)) { + ipip_entry = mlxsw_sp_ipip_entry_create(mlxsw_sp, ipipt, + ol_dev); + if (IS_ERR(ipip_entry)) + return PTR_ERR(ipip_entry); + } + } + + return 0; +} + +static void mlxsw_sp_netdevice_ipip_ol_unreg_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *ol_dev) +{ + struct mlxsw_sp_ipip_entry *ipip_entry; + + ipip_entry = mlxsw_sp_ipip_entry_find_by_ol_dev(mlxsw_sp, ol_dev); + if (ipip_entry) + mlxsw_sp_ipip_entry_destroy(mlxsw_sp, ipip_entry); +} + +static void +mlxsw_sp_ipip_entry_ol_up_event(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + struct mlxsw_sp_fib_entry *decap_fib_entry; + + decap_fib_entry = mlxsw_sp_ipip_entry_find_decap(mlxsw_sp, ipip_entry); + if (decap_fib_entry) + mlxsw_sp_ipip_entry_promote_decap(mlxsw_sp, ipip_entry, + decap_fib_entry); +} + +static int +mlxsw_sp_rif_ipip_lb_op(struct mlxsw_sp_rif_ipip_lb *lb_rif, u16 ul_vr_id, + u16 ul_rif_id, bool enable) +{ + struct mlxsw_sp_rif_ipip_lb_config lb_cf = lb_rif->lb_config; + enum mlxsw_reg_ritr_loopback_ipip_options ipip_options; + struct mlxsw_sp_rif *rif = &lb_rif->common; + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + char ritr_pl[MLXSW_REG_RITR_LEN]; + struct in6_addr *saddr6; + u32 saddr4; + + ipip_options = MLXSW_REG_RITR_LOOPBACK_IPIP_OPTIONS_GRE_KEY_PRESET; + switch (lb_cf.ul_protocol) { + case MLXSW_SP_L3_PROTO_IPV4: + saddr4 = be32_to_cpu(lb_cf.saddr.addr4); + mlxsw_reg_ritr_pack(ritr_pl, enable, MLXSW_REG_RITR_LOOPBACK_IF, + rif->rif_index, rif->vr_id, rif->dev->mtu); + mlxsw_reg_ritr_loopback_ipip4_pack(ritr_pl, lb_cf.lb_ipipt, + ipip_options, ul_vr_id, + ul_rif_id, saddr4, + lb_cf.okey); + break; + + case MLXSW_SP_L3_PROTO_IPV6: + saddr6 = &lb_cf.saddr.addr6; + mlxsw_reg_ritr_pack(ritr_pl, enable, MLXSW_REG_RITR_LOOPBACK_IF, + rif->rif_index, rif->vr_id, rif->dev->mtu); + mlxsw_reg_ritr_loopback_ipip6_pack(ritr_pl, lb_cf.lb_ipipt, + ipip_options, ul_vr_id, + ul_rif_id, saddr6, + lb_cf.okey); + break; + } + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); +} + +static int mlxsw_sp_netdevice_ipip_ol_update_mtu(struct mlxsw_sp *mlxsw_sp, + struct net_device *ol_dev) +{ + struct mlxsw_sp_ipip_entry *ipip_entry; + struct mlxsw_sp_rif_ipip_lb *lb_rif; + int err = 0; + + ipip_entry = mlxsw_sp_ipip_entry_find_by_ol_dev(mlxsw_sp, ol_dev); + if (ipip_entry) { + lb_rif = ipip_entry->ol_lb; + err = mlxsw_sp_rif_ipip_lb_op(lb_rif, lb_rif->ul_vr_id, + lb_rif->ul_rif_id, true); + if (err) + goto out; + lb_rif->common.mtu = ol_dev->mtu; + } + +out: + return err; +} + +static void mlxsw_sp_netdevice_ipip_ol_up_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *ol_dev) +{ + struct mlxsw_sp_ipip_entry *ipip_entry; + + ipip_entry = mlxsw_sp_ipip_entry_find_by_ol_dev(mlxsw_sp, ol_dev); + if (ipip_entry) + mlxsw_sp_ipip_entry_ol_up_event(mlxsw_sp, ipip_entry); +} + +static void +mlxsw_sp_ipip_entry_ol_down_event(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + if (ipip_entry->decap_fib_entry) + mlxsw_sp_ipip_entry_demote_decap(mlxsw_sp, ipip_entry); +} + +static void mlxsw_sp_netdevice_ipip_ol_down_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *ol_dev) +{ + struct mlxsw_sp_ipip_entry *ipip_entry; + + ipip_entry = mlxsw_sp_ipip_entry_find_by_ol_dev(mlxsw_sp, ol_dev); + if (ipip_entry) + mlxsw_sp_ipip_entry_ol_down_event(mlxsw_sp, ipip_entry); +} + +static void mlxsw_sp_nexthop_rif_migrate(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_rif *old_rif, + struct mlxsw_sp_rif *new_rif); +static int +mlxsw_sp_ipip_entry_ol_lb_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + bool keep_encap, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_rif_ipip_lb *old_lb_rif = ipip_entry->ol_lb; + struct mlxsw_sp_rif_ipip_lb *new_lb_rif; + + new_lb_rif = mlxsw_sp_ipip_ol_ipip_lb_create(mlxsw_sp, + ipip_entry->ipipt, + ipip_entry->ol_dev, + extack); + if (IS_ERR(new_lb_rif)) + return PTR_ERR(new_lb_rif); + ipip_entry->ol_lb = new_lb_rif; + + if (keep_encap) + mlxsw_sp_nexthop_rif_migrate(mlxsw_sp, &old_lb_rif->common, + &new_lb_rif->common); + + mlxsw_sp_rif_destroy(&old_lb_rif->common); + + return 0; +} + +static void mlxsw_sp_nexthop_rif_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_rif *rif); + +/** + * __mlxsw_sp_ipip_entry_update_tunnel - Update offload related to IPIP entry. + * @mlxsw_sp: mlxsw_sp. + * @ipip_entry: IPIP entry. + * @recreate_loopback: Recreates the associated loopback RIF. + * @keep_encap: Updates next hops that use the tunnel netdevice. This is only + * relevant when recreate_loopback is true. + * @update_nexthops: Updates next hops, keeping the current loopback RIF. This + * is only relevant when recreate_loopback is false. + * @extack: extack. + * + * Return: Non-zero value on failure. + */ +int __mlxsw_sp_ipip_entry_update_tunnel(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + bool recreate_loopback, + bool keep_encap, + bool update_nexthops, + struct netlink_ext_ack *extack) +{ + int err; + + /* RIFs can't be edited, so to update loopback, we need to destroy and + * recreate it. That creates a window of opportunity where RALUE and + * RATR registers end up referencing a RIF that's already gone. RATRs + * are handled in mlxsw_sp_ipip_entry_ol_lb_update(), and to take care + * of RALUE, demote the decap route back. + */ + if (ipip_entry->decap_fib_entry) + mlxsw_sp_ipip_entry_demote_decap(mlxsw_sp, ipip_entry); + + if (recreate_loopback) { + err = mlxsw_sp_ipip_entry_ol_lb_update(mlxsw_sp, ipip_entry, + keep_encap, extack); + if (err) + return err; + } else if (update_nexthops) { + mlxsw_sp_nexthop_rif_update(mlxsw_sp, + &ipip_entry->ol_lb->common); + } + + if (ipip_entry->ol_dev->flags & IFF_UP) + mlxsw_sp_ipip_entry_ol_up_event(mlxsw_sp, ipip_entry); + + return 0; +} + +static int mlxsw_sp_netdevice_ipip_ol_vrf_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *ol_dev, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_ipip_entry *ipip_entry = + mlxsw_sp_ipip_entry_find_by_ol_dev(mlxsw_sp, ol_dev); + + if (!ipip_entry) + return 0; + + return __mlxsw_sp_ipip_entry_update_tunnel(mlxsw_sp, ipip_entry, + true, false, false, extack); +} + +static int +mlxsw_sp_netdevice_ipip_ul_vrf_event(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + struct net_device *ul_dev, + bool *demote_this, + struct netlink_ext_ack *extack) +{ + u32 ul_tb_id = l3mdev_fib_table(ul_dev) ? : RT_TABLE_MAIN; + enum mlxsw_sp_l3proto ul_proto; + union mlxsw_sp_l3addr saddr; + + /* Moving underlay to a different VRF might cause local address + * conflict, and the conflicting tunnels need to be demoted. + */ + ul_proto = mlxsw_sp->router->ipip_ops_arr[ipip_entry->ipipt]->ul_proto; + saddr = mlxsw_sp_ipip_netdev_saddr(ul_proto, ipip_entry->ol_dev); + if (mlxsw_sp_ipip_demote_tunnel_by_saddr(mlxsw_sp, ul_proto, + saddr, ul_tb_id, + ipip_entry)) { + *demote_this = true; + return 0; + } + + return __mlxsw_sp_ipip_entry_update_tunnel(mlxsw_sp, ipip_entry, + true, true, false, extack); +} + +static int +mlxsw_sp_netdevice_ipip_ul_up_event(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + struct net_device *ul_dev) +{ + return __mlxsw_sp_ipip_entry_update_tunnel(mlxsw_sp, ipip_entry, + false, false, true, NULL); +} + +static int +mlxsw_sp_netdevice_ipip_ul_down_event(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + struct net_device *ul_dev) +{ + /* A down underlay device causes encapsulated packets to not be + * forwarded, but decap still works. So refresh next hops without + * touching anything else. + */ + return __mlxsw_sp_ipip_entry_update_tunnel(mlxsw_sp, ipip_entry, + false, false, true, NULL); +} + +static int +mlxsw_sp_netdevice_ipip_ol_change_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *ol_dev, + struct netlink_ext_ack *extack) +{ + const struct mlxsw_sp_ipip_ops *ipip_ops; + struct mlxsw_sp_ipip_entry *ipip_entry; + int err; + + ipip_entry = mlxsw_sp_ipip_entry_find_by_ol_dev(mlxsw_sp, ol_dev); + if (!ipip_entry) + /* A change might make a tunnel eligible for offloading, but + * that is currently not implemented. What falls to slow path + * stays there. + */ + return 0; + + /* A change might make a tunnel not eligible for offloading. */ + if (!mlxsw_sp_netdevice_ipip_can_offload(mlxsw_sp, ol_dev, + ipip_entry->ipipt)) { + mlxsw_sp_ipip_entry_demote_tunnel(mlxsw_sp, ipip_entry); + return 0; + } + + ipip_ops = mlxsw_sp->router->ipip_ops_arr[ipip_entry->ipipt]; + err = ipip_ops->ol_netdev_change(mlxsw_sp, ipip_entry, extack); + return err; +} + +void mlxsw_sp_ipip_entry_demote_tunnel(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + struct net_device *ol_dev = ipip_entry->ol_dev; + + if (ol_dev->flags & IFF_UP) + mlxsw_sp_ipip_entry_ol_down_event(mlxsw_sp, ipip_entry); + mlxsw_sp_ipip_entry_destroy(mlxsw_sp, ipip_entry); +} + +/* The configuration where several tunnels have the same local address in the + * same underlay table needs special treatment in the HW. That is currently not + * implemented in the driver. This function finds and demotes the first tunnel + * with a given source address, except the one passed in the argument + * `except'. + */ +bool +mlxsw_sp_ipip_demote_tunnel_by_saddr(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_l3proto ul_proto, + union mlxsw_sp_l3addr saddr, + u32 ul_tb_id, + const struct mlxsw_sp_ipip_entry *except) +{ + struct mlxsw_sp_ipip_entry *ipip_entry, *tmp; + + list_for_each_entry_safe(ipip_entry, tmp, &mlxsw_sp->router->ipip_list, + ipip_list_node) { + if (ipip_entry != except && + mlxsw_sp_ipip_entry_saddr_matches(mlxsw_sp, ul_proto, saddr, + ul_tb_id, ipip_entry)) { + mlxsw_sp_ipip_entry_demote_tunnel(mlxsw_sp, ipip_entry); + return true; + } + } + + return false; +} + +static void mlxsw_sp_ipip_demote_tunnel_by_ul_netdev(struct mlxsw_sp *mlxsw_sp, + struct net_device *ul_dev) +{ + struct mlxsw_sp_ipip_entry *ipip_entry, *tmp; + + list_for_each_entry_safe(ipip_entry, tmp, &mlxsw_sp->router->ipip_list, + ipip_list_node) { + struct net_device *ol_dev = ipip_entry->ol_dev; + struct net_device *ipip_ul_dev; + + rcu_read_lock(); + ipip_ul_dev = mlxsw_sp_ipip_netdev_ul_dev_get(ol_dev); + rcu_read_unlock(); + if (ipip_ul_dev == ul_dev) + mlxsw_sp_ipip_entry_demote_tunnel(mlxsw_sp, ipip_entry); + } +} + +static int mlxsw_sp_netdevice_ipip_ol_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *ol_dev, + unsigned long event, + struct netdev_notifier_info *info) +{ + struct netdev_notifier_changeupper_info *chup; + struct netlink_ext_ack *extack; + int err = 0; + + switch (event) { + case NETDEV_REGISTER: + err = mlxsw_sp_netdevice_ipip_ol_reg_event(mlxsw_sp, ol_dev); + break; + case NETDEV_UNREGISTER: + mlxsw_sp_netdevice_ipip_ol_unreg_event(mlxsw_sp, ol_dev); + break; + case NETDEV_UP: + mlxsw_sp_netdevice_ipip_ol_up_event(mlxsw_sp, ol_dev); + break; + case NETDEV_DOWN: + mlxsw_sp_netdevice_ipip_ol_down_event(mlxsw_sp, ol_dev); + break; + case NETDEV_CHANGEUPPER: + chup = container_of(info, typeof(*chup), info); + extack = info->extack; + if (netif_is_l3_master(chup->upper_dev)) + err = mlxsw_sp_netdevice_ipip_ol_vrf_event(mlxsw_sp, + ol_dev, + extack); + break; + case NETDEV_CHANGE: + extack = info->extack; + err = mlxsw_sp_netdevice_ipip_ol_change_event(mlxsw_sp, + ol_dev, extack); + break; + case NETDEV_CHANGEMTU: + err = mlxsw_sp_netdevice_ipip_ol_update_mtu(mlxsw_sp, ol_dev); + break; + } + return err; +} + +static int +__mlxsw_sp_netdevice_ipip_ul_event(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + struct net_device *ul_dev, + bool *demote_this, + unsigned long event, + struct netdev_notifier_info *info) +{ + struct netdev_notifier_changeupper_info *chup; + struct netlink_ext_ack *extack; + + switch (event) { + case NETDEV_CHANGEUPPER: + chup = container_of(info, typeof(*chup), info); + extack = info->extack; + if (netif_is_l3_master(chup->upper_dev)) + return mlxsw_sp_netdevice_ipip_ul_vrf_event(mlxsw_sp, + ipip_entry, + ul_dev, + demote_this, + extack); + break; + + case NETDEV_UP: + return mlxsw_sp_netdevice_ipip_ul_up_event(mlxsw_sp, ipip_entry, + ul_dev); + case NETDEV_DOWN: + return mlxsw_sp_netdevice_ipip_ul_down_event(mlxsw_sp, + ipip_entry, + ul_dev); + } + return 0; +} + +static int +mlxsw_sp_netdevice_ipip_ul_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *ul_dev, + unsigned long event, + struct netdev_notifier_info *info) +{ + struct mlxsw_sp_ipip_entry *ipip_entry = NULL; + int err; + + while ((ipip_entry = mlxsw_sp_ipip_entry_find_by_ul_dev(mlxsw_sp, + ul_dev, + ipip_entry))) { + struct mlxsw_sp_ipip_entry *prev; + bool demote_this = false; + + err = __mlxsw_sp_netdevice_ipip_ul_event(mlxsw_sp, ipip_entry, + ul_dev, &demote_this, + event, info); + if (err) { + mlxsw_sp_ipip_demote_tunnel_by_ul_netdev(mlxsw_sp, + ul_dev); + return err; + } + + if (demote_this) { + if (list_is_first(&ipip_entry->ipip_list_node, + &mlxsw_sp->router->ipip_list)) + prev = NULL; + else + /* This can't be cached from previous iteration, + * because that entry could be gone now. + */ + prev = list_prev_entry(ipip_entry, + ipip_list_node); + mlxsw_sp_ipip_entry_demote_tunnel(mlxsw_sp, ipip_entry); + ipip_entry = prev; + } + } + + return 0; +} + +int mlxsw_sp_router_nve_promote_decap(struct mlxsw_sp *mlxsw_sp, u32 ul_tb_id, + enum mlxsw_sp_l3proto ul_proto, + const union mlxsw_sp_l3addr *ul_sip, + u32 tunnel_index) +{ + enum mlxsw_sp_fib_entry_type type = MLXSW_SP_FIB_ENTRY_TYPE_TRAP; + struct mlxsw_sp_router *router = mlxsw_sp->router; + struct mlxsw_sp_fib_entry *fib_entry; + int err = 0; + + mutex_lock(&mlxsw_sp->router->lock); + + if (WARN_ON_ONCE(router->nve_decap_config.valid)) { + err = -EINVAL; + goto out; + } + + router->nve_decap_config.ul_tb_id = ul_tb_id; + router->nve_decap_config.tunnel_index = tunnel_index; + router->nve_decap_config.ul_proto = ul_proto; + router->nve_decap_config.ul_sip = *ul_sip; + router->nve_decap_config.valid = true; + + /* It is valid to create a tunnel with a local IP and only later + * assign this IP address to a local interface + */ + fib_entry = mlxsw_sp_router_ip2me_fib_entry_find(mlxsw_sp, ul_tb_id, + ul_proto, ul_sip, + type); + if (!fib_entry) + goto out; + + fib_entry->decap.tunnel_index = tunnel_index; + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_NVE_DECAP; + + err = mlxsw_sp_fib_entry_update(mlxsw_sp, fib_entry); + if (err) + goto err_fib_entry_update; + + goto out; + +err_fib_entry_update: + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_TRAP; + mlxsw_sp_fib_entry_update(mlxsw_sp, fib_entry); +out: + mutex_unlock(&mlxsw_sp->router->lock); + return err; +} + +void mlxsw_sp_router_nve_demote_decap(struct mlxsw_sp *mlxsw_sp, u32 ul_tb_id, + enum mlxsw_sp_l3proto ul_proto, + const union mlxsw_sp_l3addr *ul_sip) +{ + enum mlxsw_sp_fib_entry_type type = MLXSW_SP_FIB_ENTRY_TYPE_NVE_DECAP; + struct mlxsw_sp_router *router = mlxsw_sp->router; + struct mlxsw_sp_fib_entry *fib_entry; + + mutex_lock(&mlxsw_sp->router->lock); + + if (WARN_ON_ONCE(!router->nve_decap_config.valid)) + goto out; + + router->nve_decap_config.valid = false; + + fib_entry = mlxsw_sp_router_ip2me_fib_entry_find(mlxsw_sp, ul_tb_id, + ul_proto, ul_sip, + type); + if (!fib_entry) + goto out; + + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_TRAP; + mlxsw_sp_fib_entry_update(mlxsw_sp, fib_entry); +out: + mutex_unlock(&mlxsw_sp->router->lock); +} + +static bool mlxsw_sp_router_nve_is_decap(struct mlxsw_sp *mlxsw_sp, + u32 ul_tb_id, + enum mlxsw_sp_l3proto ul_proto, + const union mlxsw_sp_l3addr *ul_sip) +{ + struct mlxsw_sp_router *router = mlxsw_sp->router; + + return router->nve_decap_config.valid && + router->nve_decap_config.ul_tb_id == ul_tb_id && + router->nve_decap_config.ul_proto == ul_proto && + !memcmp(&router->nve_decap_config.ul_sip, ul_sip, + sizeof(*ul_sip)); +} + +struct mlxsw_sp_neigh_key { + struct neighbour *n; +}; + +struct mlxsw_sp_neigh_entry { + struct list_head rif_list_node; + struct rhash_head ht_node; + struct mlxsw_sp_neigh_key key; + u16 rif; + bool connected; + unsigned char ha[ETH_ALEN]; + struct list_head nexthop_list; /* list of nexthops using + * this neigh entry + */ + struct list_head nexthop_neighs_list_node; + unsigned int counter_index; + bool counter_valid; +}; + +static const struct rhashtable_params mlxsw_sp_neigh_ht_params = { + .key_offset = offsetof(struct mlxsw_sp_neigh_entry, key), + .head_offset = offsetof(struct mlxsw_sp_neigh_entry, ht_node), + .key_len = sizeof(struct mlxsw_sp_neigh_key), +}; + +struct mlxsw_sp_neigh_entry * +mlxsw_sp_rif_neigh_next(struct mlxsw_sp_rif *rif, + struct mlxsw_sp_neigh_entry *neigh_entry) +{ + if (!neigh_entry) { + if (list_empty(&rif->neigh_list)) + return NULL; + else + return list_first_entry(&rif->neigh_list, + typeof(*neigh_entry), + rif_list_node); + } + if (list_is_last(&neigh_entry->rif_list_node, &rif->neigh_list)) + return NULL; + return list_next_entry(neigh_entry, rif_list_node); +} + +int mlxsw_sp_neigh_entry_type(struct mlxsw_sp_neigh_entry *neigh_entry) +{ + return neigh_entry->key.n->tbl->family; +} + +unsigned char * +mlxsw_sp_neigh_entry_ha(struct mlxsw_sp_neigh_entry *neigh_entry) +{ + return neigh_entry->ha; +} + +u32 mlxsw_sp_neigh4_entry_dip(struct mlxsw_sp_neigh_entry *neigh_entry) +{ + struct neighbour *n; + + n = neigh_entry->key.n; + return ntohl(*((__be32 *) n->primary_key)); +} + +struct in6_addr * +mlxsw_sp_neigh6_entry_dip(struct mlxsw_sp_neigh_entry *neigh_entry) +{ + struct neighbour *n; + + n = neigh_entry->key.n; + return (struct in6_addr *) &n->primary_key; +} + +int mlxsw_sp_neigh_counter_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry, + u64 *p_counter) +{ + if (!neigh_entry->counter_valid) + return -EINVAL; + + return mlxsw_sp_flow_counter_get(mlxsw_sp, neigh_entry->counter_index, + p_counter, NULL); +} + +static struct mlxsw_sp_neigh_entry * +mlxsw_sp_neigh_entry_alloc(struct mlxsw_sp *mlxsw_sp, struct neighbour *n, + u16 rif) +{ + struct mlxsw_sp_neigh_entry *neigh_entry; + + neigh_entry = kzalloc(sizeof(*neigh_entry), GFP_KERNEL); + if (!neigh_entry) + return NULL; + + neigh_entry->key.n = n; + neigh_entry->rif = rif; + INIT_LIST_HEAD(&neigh_entry->nexthop_list); + + return neigh_entry; +} + +static void mlxsw_sp_neigh_entry_free(struct mlxsw_sp_neigh_entry *neigh_entry) +{ + kfree(neigh_entry); +} + +static int +mlxsw_sp_neigh_entry_insert(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry) +{ + return rhashtable_insert_fast(&mlxsw_sp->router->neigh_ht, + &neigh_entry->ht_node, + mlxsw_sp_neigh_ht_params); +} + +static void +mlxsw_sp_neigh_entry_remove(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry) +{ + rhashtable_remove_fast(&mlxsw_sp->router->neigh_ht, + &neigh_entry->ht_node, + mlxsw_sp_neigh_ht_params); +} + +static bool +mlxsw_sp_neigh_counter_should_alloc(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry) +{ + struct devlink *devlink; + const char *table_name; + + switch (mlxsw_sp_neigh_entry_type(neigh_entry)) { + case AF_INET: + table_name = MLXSW_SP_DPIPE_TABLE_NAME_HOST4; + break; + case AF_INET6: + table_name = MLXSW_SP_DPIPE_TABLE_NAME_HOST6; + break; + default: + WARN_ON(1); + return false; + } + + devlink = priv_to_devlink(mlxsw_sp->core); + return devlink_dpipe_table_counter_enabled(devlink, table_name); +} + +static void +mlxsw_sp_neigh_counter_alloc(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry) +{ + if (!mlxsw_sp_neigh_counter_should_alloc(mlxsw_sp, neigh_entry)) + return; + + if (mlxsw_sp_flow_counter_alloc(mlxsw_sp, &neigh_entry->counter_index)) + return; + + neigh_entry->counter_valid = true; +} + +static void +mlxsw_sp_neigh_counter_free(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry) +{ + if (!neigh_entry->counter_valid) + return; + mlxsw_sp_flow_counter_free(mlxsw_sp, + neigh_entry->counter_index); + neigh_entry->counter_valid = false; +} + +static struct mlxsw_sp_neigh_entry * +mlxsw_sp_neigh_entry_create(struct mlxsw_sp *mlxsw_sp, struct neighbour *n) +{ + struct mlxsw_sp_neigh_entry *neigh_entry; + struct mlxsw_sp_rif *rif; + int err; + + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, n->dev); + if (!rif) + return ERR_PTR(-EINVAL); + + neigh_entry = mlxsw_sp_neigh_entry_alloc(mlxsw_sp, n, rif->rif_index); + if (!neigh_entry) + return ERR_PTR(-ENOMEM); + + err = mlxsw_sp_neigh_entry_insert(mlxsw_sp, neigh_entry); + if (err) + goto err_neigh_entry_insert; + + mlxsw_sp_neigh_counter_alloc(mlxsw_sp, neigh_entry); + atomic_inc(&mlxsw_sp->router->neighs_update.neigh_count); + list_add(&neigh_entry->rif_list_node, &rif->neigh_list); + + return neigh_entry; + +err_neigh_entry_insert: + mlxsw_sp_neigh_entry_free(neigh_entry); + return ERR_PTR(err); +} + +static void +mlxsw_sp_neigh_entry_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry) +{ + list_del(&neigh_entry->rif_list_node); + atomic_dec(&mlxsw_sp->router->neighs_update.neigh_count); + mlxsw_sp_neigh_counter_free(mlxsw_sp, neigh_entry); + mlxsw_sp_neigh_entry_remove(mlxsw_sp, neigh_entry); + mlxsw_sp_neigh_entry_free(neigh_entry); +} + +static struct mlxsw_sp_neigh_entry * +mlxsw_sp_neigh_entry_lookup(struct mlxsw_sp *mlxsw_sp, struct neighbour *n) +{ + struct mlxsw_sp_neigh_key key; + + key.n = n; + return rhashtable_lookup_fast(&mlxsw_sp->router->neigh_ht, + &key, mlxsw_sp_neigh_ht_params); +} + +static void +mlxsw_sp_router_neighs_update_interval_init(struct mlxsw_sp *mlxsw_sp) +{ + unsigned long interval; + +#if IS_ENABLED(CONFIG_IPV6) + interval = min_t(unsigned long, + NEIGH_VAR(&arp_tbl.parms, DELAY_PROBE_TIME), + NEIGH_VAR(&nd_tbl.parms, DELAY_PROBE_TIME)); +#else + interval = NEIGH_VAR(&arp_tbl.parms, DELAY_PROBE_TIME); +#endif + mlxsw_sp->router->neighs_update.interval = jiffies_to_msecs(interval); +} + +static void mlxsw_sp_router_neigh_ent_ipv4_process(struct mlxsw_sp *mlxsw_sp, + char *rauhtd_pl, + int ent_index) +{ + u64 max_rifs = MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); + struct net_device *dev; + struct neighbour *n; + __be32 dipn; + u32 dip; + u16 rif; + + mlxsw_reg_rauhtd_ent_ipv4_unpack(rauhtd_pl, ent_index, &rif, &dip); + + if (WARN_ON_ONCE(rif >= max_rifs)) + return; + if (!mlxsw_sp->router->rifs[rif]) { + dev_err_ratelimited(mlxsw_sp->bus_info->dev, "Incorrect RIF in neighbour entry\n"); + return; + } + + dipn = htonl(dip); + dev = mlxsw_sp->router->rifs[rif]->dev; + n = neigh_lookup(&arp_tbl, &dipn, dev); + if (!n) + return; + + netdev_dbg(dev, "Updating neighbour with IP=%pI4h\n", &dip); + neigh_event_send(n, NULL); + neigh_release(n); +} + +#if IS_ENABLED(CONFIG_IPV6) +static void mlxsw_sp_router_neigh_ent_ipv6_process(struct mlxsw_sp *mlxsw_sp, + char *rauhtd_pl, + int rec_index) +{ + struct net_device *dev; + struct neighbour *n; + struct in6_addr dip; + u16 rif; + + mlxsw_reg_rauhtd_ent_ipv6_unpack(rauhtd_pl, rec_index, &rif, + (char *) &dip); + + if (!mlxsw_sp->router->rifs[rif]) { + dev_err_ratelimited(mlxsw_sp->bus_info->dev, "Incorrect RIF in neighbour entry\n"); + return; + } + + dev = mlxsw_sp->router->rifs[rif]->dev; + n = neigh_lookup(&nd_tbl, &dip, dev); + if (!n) + return; + + netdev_dbg(dev, "Updating neighbour with IP=%pI6c\n", &dip); + neigh_event_send(n, NULL); + neigh_release(n); +} +#else +static void mlxsw_sp_router_neigh_ent_ipv6_process(struct mlxsw_sp *mlxsw_sp, + char *rauhtd_pl, + int rec_index) +{ +} +#endif + +static void mlxsw_sp_router_neigh_rec_ipv4_process(struct mlxsw_sp *mlxsw_sp, + char *rauhtd_pl, + int rec_index) +{ + u8 num_entries; + int i; + + num_entries = mlxsw_reg_rauhtd_ipv4_rec_num_entries_get(rauhtd_pl, + rec_index); + /* Hardware starts counting at 0, so add 1. */ + num_entries++; + + /* Each record consists of several neighbour entries. */ + for (i = 0; i < num_entries; i++) { + int ent_index; + + ent_index = rec_index * MLXSW_REG_RAUHTD_IPV4_ENT_PER_REC + i; + mlxsw_sp_router_neigh_ent_ipv4_process(mlxsw_sp, rauhtd_pl, + ent_index); + } + +} + +static void mlxsw_sp_router_neigh_rec_ipv6_process(struct mlxsw_sp *mlxsw_sp, + char *rauhtd_pl, + int rec_index) +{ + /* One record contains one entry. */ + mlxsw_sp_router_neigh_ent_ipv6_process(mlxsw_sp, rauhtd_pl, + rec_index); +} + +static void mlxsw_sp_router_neigh_rec_process(struct mlxsw_sp *mlxsw_sp, + char *rauhtd_pl, int rec_index) +{ + switch (mlxsw_reg_rauhtd_rec_type_get(rauhtd_pl, rec_index)) { + case MLXSW_REG_RAUHTD_TYPE_IPV4: + mlxsw_sp_router_neigh_rec_ipv4_process(mlxsw_sp, rauhtd_pl, + rec_index); + break; + case MLXSW_REG_RAUHTD_TYPE_IPV6: + mlxsw_sp_router_neigh_rec_ipv6_process(mlxsw_sp, rauhtd_pl, + rec_index); + break; + } +} + +static bool mlxsw_sp_router_rauhtd_is_full(char *rauhtd_pl) +{ + u8 num_rec, last_rec_index, num_entries; + + num_rec = mlxsw_reg_rauhtd_num_rec_get(rauhtd_pl); + last_rec_index = num_rec - 1; + + if (num_rec < MLXSW_REG_RAUHTD_REC_MAX_NUM) + return false; + if (mlxsw_reg_rauhtd_rec_type_get(rauhtd_pl, last_rec_index) == + MLXSW_REG_RAUHTD_TYPE_IPV6) + return true; + + num_entries = mlxsw_reg_rauhtd_ipv4_rec_num_entries_get(rauhtd_pl, + last_rec_index); + if (++num_entries == MLXSW_REG_RAUHTD_IPV4_ENT_PER_REC) + return true; + return false; +} + +static int +__mlxsw_sp_router_neighs_update_rauhtd(struct mlxsw_sp *mlxsw_sp, + char *rauhtd_pl, + enum mlxsw_reg_rauhtd_type type) +{ + int i, num_rec; + int err; + + /* Ensure the RIF we read from the device does not change mid-dump. */ + mutex_lock(&mlxsw_sp->router->lock); + do { + mlxsw_reg_rauhtd_pack(rauhtd_pl, type); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(rauhtd), + rauhtd_pl); + if (err) { + dev_err_ratelimited(mlxsw_sp->bus_info->dev, "Failed to dump neighbour table\n"); + break; + } + num_rec = mlxsw_reg_rauhtd_num_rec_get(rauhtd_pl); + for (i = 0; i < num_rec; i++) + mlxsw_sp_router_neigh_rec_process(mlxsw_sp, rauhtd_pl, + i); + } while (mlxsw_sp_router_rauhtd_is_full(rauhtd_pl)); + mutex_unlock(&mlxsw_sp->router->lock); + + return err; +} + +static int mlxsw_sp_router_neighs_update_rauhtd(struct mlxsw_sp *mlxsw_sp) +{ + enum mlxsw_reg_rauhtd_type type; + char *rauhtd_pl; + int err; + + if (!atomic_read(&mlxsw_sp->router->neighs_update.neigh_count)) + return 0; + + rauhtd_pl = kmalloc(MLXSW_REG_RAUHTD_LEN, GFP_KERNEL); + if (!rauhtd_pl) + return -ENOMEM; + + type = MLXSW_REG_RAUHTD_TYPE_IPV4; + err = __mlxsw_sp_router_neighs_update_rauhtd(mlxsw_sp, rauhtd_pl, type); + if (err) + goto out; + + type = MLXSW_REG_RAUHTD_TYPE_IPV6; + err = __mlxsw_sp_router_neighs_update_rauhtd(mlxsw_sp, rauhtd_pl, type); +out: + kfree(rauhtd_pl); + return err; +} + +static void mlxsw_sp_router_neighs_update_nh(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_neigh_entry *neigh_entry; + + mutex_lock(&mlxsw_sp->router->lock); + list_for_each_entry(neigh_entry, &mlxsw_sp->router->nexthop_neighs_list, + nexthop_neighs_list_node) + /* If this neigh have nexthops, make the kernel think this neigh + * is active regardless of the traffic. + */ + neigh_event_send(neigh_entry->key.n, NULL); + mutex_unlock(&mlxsw_sp->router->lock); +} + +static void +mlxsw_sp_router_neighs_update_work_schedule(struct mlxsw_sp *mlxsw_sp) +{ + unsigned long interval = mlxsw_sp->router->neighs_update.interval; + + mlxsw_core_schedule_dw(&mlxsw_sp->router->neighs_update.dw, + msecs_to_jiffies(interval)); +} + +static void mlxsw_sp_router_neighs_update_work(struct work_struct *work) +{ + struct mlxsw_sp_router *router; + int err; + + router = container_of(work, struct mlxsw_sp_router, + neighs_update.dw.work); + err = mlxsw_sp_router_neighs_update_rauhtd(router->mlxsw_sp); + if (err) + dev_err(router->mlxsw_sp->bus_info->dev, "Could not update kernel for neigh activity"); + + mlxsw_sp_router_neighs_update_nh(router->mlxsw_sp); + + mlxsw_sp_router_neighs_update_work_schedule(router->mlxsw_sp); +} + +static void mlxsw_sp_router_probe_unresolved_nexthops(struct work_struct *work) +{ + struct mlxsw_sp_neigh_entry *neigh_entry; + struct mlxsw_sp_router *router; + + router = container_of(work, struct mlxsw_sp_router, + nexthop_probe_dw.work); + /* Iterate over nexthop neighbours, find those who are unresolved and + * send arp on them. This solves the chicken-egg problem when + * the nexthop wouldn't get offloaded until the neighbor is resolved + * but it wouldn't get resolved ever in case traffic is flowing in HW + * using different nexthop. + */ + mutex_lock(&router->lock); + list_for_each_entry(neigh_entry, &router->nexthop_neighs_list, + nexthop_neighs_list_node) + if (!neigh_entry->connected) + neigh_event_send(neigh_entry->key.n, NULL); + mutex_unlock(&router->lock); + + mlxsw_core_schedule_dw(&router->nexthop_probe_dw, + MLXSW_SP_UNRESOLVED_NH_PROBE_INTERVAL); +} + +static void +mlxsw_sp_nexthop_neigh_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry, + bool removing, bool dead); + +static enum mlxsw_reg_rauht_op mlxsw_sp_rauht_op(bool adding) +{ + return adding ? MLXSW_REG_RAUHT_OP_WRITE_ADD : + MLXSW_REG_RAUHT_OP_WRITE_DELETE; +} + +static int +mlxsw_sp_router_neigh_entry_op4(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry, + enum mlxsw_reg_rauht_op op) +{ + struct neighbour *n = neigh_entry->key.n; + u32 dip = ntohl(*((__be32 *) n->primary_key)); + char rauht_pl[MLXSW_REG_RAUHT_LEN]; + + mlxsw_reg_rauht_pack4(rauht_pl, op, neigh_entry->rif, neigh_entry->ha, + dip); + if (neigh_entry->counter_valid) + mlxsw_reg_rauht_pack_counter(rauht_pl, + neigh_entry->counter_index); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rauht), rauht_pl); +} + +static int +mlxsw_sp_router_neigh_entry_op6(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry, + enum mlxsw_reg_rauht_op op) +{ + struct neighbour *n = neigh_entry->key.n; + char rauht_pl[MLXSW_REG_RAUHT_LEN]; + const char *dip = n->primary_key; + + mlxsw_reg_rauht_pack6(rauht_pl, op, neigh_entry->rif, neigh_entry->ha, + dip); + if (neigh_entry->counter_valid) + mlxsw_reg_rauht_pack_counter(rauht_pl, + neigh_entry->counter_index); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rauht), rauht_pl); +} + +bool mlxsw_sp_neigh_ipv6_ignore(struct mlxsw_sp_neigh_entry *neigh_entry) +{ + struct neighbour *n = neigh_entry->key.n; + + /* Packets with a link-local destination address are trapped + * after LPM lookup and never reach the neighbour table, so + * there is no need to program such neighbours to the device. + */ + if (ipv6_addr_type((struct in6_addr *) &n->primary_key) & + IPV6_ADDR_LINKLOCAL) + return true; + return false; +} + +static void +mlxsw_sp_neigh_entry_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry, + bool adding) +{ + enum mlxsw_reg_rauht_op op = mlxsw_sp_rauht_op(adding); + int err; + + if (!adding && !neigh_entry->connected) + return; + neigh_entry->connected = adding; + if (neigh_entry->key.n->tbl->family == AF_INET) { + err = mlxsw_sp_router_neigh_entry_op4(mlxsw_sp, neigh_entry, + op); + if (err) + return; + } else if (neigh_entry->key.n->tbl->family == AF_INET6) { + if (mlxsw_sp_neigh_ipv6_ignore(neigh_entry)) + return; + err = mlxsw_sp_router_neigh_entry_op6(mlxsw_sp, neigh_entry, + op); + if (err) + return; + } else { + WARN_ON_ONCE(1); + return; + } + + if (adding) + neigh_entry->key.n->flags |= NTF_OFFLOADED; + else + neigh_entry->key.n->flags &= ~NTF_OFFLOADED; +} + +void +mlxsw_sp_neigh_entry_counter_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry, + bool adding) +{ + if (adding) + mlxsw_sp_neigh_counter_alloc(mlxsw_sp, neigh_entry); + else + mlxsw_sp_neigh_counter_free(mlxsw_sp, neigh_entry); + mlxsw_sp_neigh_entry_update(mlxsw_sp, neigh_entry, true); +} + +struct mlxsw_sp_netevent_work { + struct work_struct work; + struct mlxsw_sp *mlxsw_sp; + struct neighbour *n; +}; + +static void mlxsw_sp_router_neigh_event_work(struct work_struct *work) +{ + struct mlxsw_sp_netevent_work *net_work = + container_of(work, struct mlxsw_sp_netevent_work, work); + struct mlxsw_sp *mlxsw_sp = net_work->mlxsw_sp; + struct mlxsw_sp_neigh_entry *neigh_entry; + struct neighbour *n = net_work->n; + unsigned char ha[ETH_ALEN]; + bool entry_connected; + u8 nud_state, dead; + + /* If these parameters are changed after we release the lock, + * then we are guaranteed to receive another event letting us + * know about it. + */ + read_lock_bh(&n->lock); + memcpy(ha, n->ha, ETH_ALEN); + nud_state = n->nud_state; + dead = n->dead; + read_unlock_bh(&n->lock); + + mutex_lock(&mlxsw_sp->router->lock); + mlxsw_sp_span_respin(mlxsw_sp); + + entry_connected = nud_state & NUD_VALID && !dead; + neigh_entry = mlxsw_sp_neigh_entry_lookup(mlxsw_sp, n); + if (!entry_connected && !neigh_entry) + goto out; + if (!neigh_entry) { + neigh_entry = mlxsw_sp_neigh_entry_create(mlxsw_sp, n); + if (IS_ERR(neigh_entry)) + goto out; + } + + if (neigh_entry->connected && entry_connected && + !memcmp(neigh_entry->ha, ha, ETH_ALEN)) + goto out; + + memcpy(neigh_entry->ha, ha, ETH_ALEN); + mlxsw_sp_neigh_entry_update(mlxsw_sp, neigh_entry, entry_connected); + mlxsw_sp_nexthop_neigh_update(mlxsw_sp, neigh_entry, !entry_connected, + dead); + + if (!neigh_entry->connected && list_empty(&neigh_entry->nexthop_list)) + mlxsw_sp_neigh_entry_destroy(mlxsw_sp, neigh_entry); + +out: + mutex_unlock(&mlxsw_sp->router->lock); + neigh_release(n); + kfree(net_work); +} + +static int mlxsw_sp_mp_hash_init(struct mlxsw_sp *mlxsw_sp); + +static void mlxsw_sp_router_mp_hash_event_work(struct work_struct *work) +{ + struct mlxsw_sp_netevent_work *net_work = + container_of(work, struct mlxsw_sp_netevent_work, work); + struct mlxsw_sp *mlxsw_sp = net_work->mlxsw_sp; + + mlxsw_sp_mp_hash_init(mlxsw_sp); + kfree(net_work); +} + +static int __mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp); + +static void mlxsw_sp_router_update_priority_work(struct work_struct *work) +{ + struct mlxsw_sp_netevent_work *net_work = + container_of(work, struct mlxsw_sp_netevent_work, work); + struct mlxsw_sp *mlxsw_sp = net_work->mlxsw_sp; + + __mlxsw_sp_router_init(mlxsw_sp); + kfree(net_work); +} + +static int mlxsw_sp_router_schedule_work(struct net *net, + struct notifier_block *nb, + void (*cb)(struct work_struct *)) +{ + struct mlxsw_sp_netevent_work *net_work; + struct mlxsw_sp_router *router; + + router = container_of(nb, struct mlxsw_sp_router, netevent_nb); + if (!net_eq(net, mlxsw_sp_net(router->mlxsw_sp))) + return NOTIFY_DONE; + + net_work = kzalloc(sizeof(*net_work), GFP_ATOMIC); + if (!net_work) + return NOTIFY_BAD; + + INIT_WORK(&net_work->work, cb); + net_work->mlxsw_sp = router->mlxsw_sp; + mlxsw_core_schedule_work(&net_work->work); + return NOTIFY_DONE; +} + +static int mlxsw_sp_router_netevent_event(struct notifier_block *nb, + unsigned long event, void *ptr) +{ + struct mlxsw_sp_netevent_work *net_work; + struct mlxsw_sp_port *mlxsw_sp_port; + struct mlxsw_sp *mlxsw_sp; + unsigned long interval; + struct neigh_parms *p; + struct neighbour *n; + + switch (event) { + case NETEVENT_DELAY_PROBE_TIME_UPDATE: + p = ptr; + + /* We don't care about changes in the default table. */ + if (!p->dev || (p->tbl->family != AF_INET && + p->tbl->family != AF_INET6)) + return NOTIFY_DONE; + + /* We are in atomic context and can't take RTNL mutex, + * so use RCU variant to walk the device chain. + */ + mlxsw_sp_port = mlxsw_sp_port_lower_dev_hold(p->dev); + if (!mlxsw_sp_port) + return NOTIFY_DONE; + + mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + interval = jiffies_to_msecs(NEIGH_VAR(p, DELAY_PROBE_TIME)); + mlxsw_sp->router->neighs_update.interval = interval; + + mlxsw_sp_port_dev_put(mlxsw_sp_port); + break; + case NETEVENT_NEIGH_UPDATE: + n = ptr; + + if (n->tbl->family != AF_INET && n->tbl->family != AF_INET6) + return NOTIFY_DONE; + + mlxsw_sp_port = mlxsw_sp_port_lower_dev_hold(n->dev); + if (!mlxsw_sp_port) + return NOTIFY_DONE; + + net_work = kzalloc(sizeof(*net_work), GFP_ATOMIC); + if (!net_work) { + mlxsw_sp_port_dev_put(mlxsw_sp_port); + return NOTIFY_BAD; + } + + INIT_WORK(&net_work->work, mlxsw_sp_router_neigh_event_work); + net_work->mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + net_work->n = n; + + /* Take a reference to ensure the neighbour won't be + * destructed until we drop the reference in delayed + * work. + */ + neigh_clone(n); + mlxsw_core_schedule_work(&net_work->work); + mlxsw_sp_port_dev_put(mlxsw_sp_port); + break; + case NETEVENT_IPV4_MPATH_HASH_UPDATE: + case NETEVENT_IPV6_MPATH_HASH_UPDATE: + return mlxsw_sp_router_schedule_work(ptr, nb, + mlxsw_sp_router_mp_hash_event_work); + + case NETEVENT_IPV4_FWD_UPDATE_PRIORITY_UPDATE: + return mlxsw_sp_router_schedule_work(ptr, nb, + mlxsw_sp_router_update_priority_work); + } + + return NOTIFY_DONE; +} + +static int mlxsw_sp_neigh_init(struct mlxsw_sp *mlxsw_sp) +{ + int err; + + err = rhashtable_init(&mlxsw_sp->router->neigh_ht, + &mlxsw_sp_neigh_ht_params); + if (err) + return err; + + /* Initialize the polling interval according to the default + * table. + */ + mlxsw_sp_router_neighs_update_interval_init(mlxsw_sp); + + /* Create the delayed works for the activity_update */ + INIT_DELAYED_WORK(&mlxsw_sp->router->neighs_update.dw, + mlxsw_sp_router_neighs_update_work); + INIT_DELAYED_WORK(&mlxsw_sp->router->nexthop_probe_dw, + mlxsw_sp_router_probe_unresolved_nexthops); + atomic_set(&mlxsw_sp->router->neighs_update.neigh_count, 0); + mlxsw_core_schedule_dw(&mlxsw_sp->router->neighs_update.dw, 0); + mlxsw_core_schedule_dw(&mlxsw_sp->router->nexthop_probe_dw, 0); + return 0; +} + +static void mlxsw_sp_neigh_fini(struct mlxsw_sp *mlxsw_sp) +{ + cancel_delayed_work_sync(&mlxsw_sp->router->neighs_update.dw); + cancel_delayed_work_sync(&mlxsw_sp->router->nexthop_probe_dw); + rhashtable_destroy(&mlxsw_sp->router->neigh_ht); +} + +static void mlxsw_sp_neigh_rif_gone_sync(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp_neigh_entry *neigh_entry, *tmp; + + list_for_each_entry_safe(neigh_entry, tmp, &rif->neigh_list, + rif_list_node) { + mlxsw_sp_neigh_entry_update(mlxsw_sp, neigh_entry, false); + mlxsw_sp_neigh_entry_destroy(mlxsw_sp, neigh_entry); + } +} + +enum mlxsw_sp_nexthop_type { + MLXSW_SP_NEXTHOP_TYPE_ETH, + MLXSW_SP_NEXTHOP_TYPE_IPIP, +}; + +enum mlxsw_sp_nexthop_action { + /* Nexthop forwards packets to an egress RIF */ + MLXSW_SP_NEXTHOP_ACTION_FORWARD, + /* Nexthop discards packets */ + MLXSW_SP_NEXTHOP_ACTION_DISCARD, + /* Nexthop traps packets */ + MLXSW_SP_NEXTHOP_ACTION_TRAP, +}; + +struct mlxsw_sp_nexthop_key { + struct fib_nh *fib_nh; +}; + +struct mlxsw_sp_nexthop { + struct list_head neigh_list_node; /* member of neigh entry list */ + struct list_head rif_list_node; + struct list_head router_list_node; + struct mlxsw_sp_nexthop_group_info *nhgi; /* pointer back to the group + * this nexthop belongs to + */ + struct rhash_head ht_node; + struct neigh_table *neigh_tbl; + struct mlxsw_sp_nexthop_key key; + unsigned char gw_addr[sizeof(struct in6_addr)]; + int ifindex; + int nh_weight; + int norm_nh_weight; + int num_adj_entries; + struct mlxsw_sp_rif *rif; + u8 should_offload:1, /* set indicates this nexthop should be written + * to the adjacency table. + */ + offloaded:1, /* set indicates this nexthop was written to the + * adjacency table. + */ + update:1; /* set indicates this nexthop should be updated in the + * adjacency table (f.e., its MAC changed). + */ + enum mlxsw_sp_nexthop_action action; + enum mlxsw_sp_nexthop_type type; + union { + struct mlxsw_sp_neigh_entry *neigh_entry; + struct mlxsw_sp_ipip_entry *ipip_entry; + }; + unsigned int counter_index; + bool counter_valid; +}; + +enum mlxsw_sp_nexthop_group_type { + MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV4, + MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV6, + MLXSW_SP_NEXTHOP_GROUP_TYPE_OBJ, +}; + +struct mlxsw_sp_nexthop_group_info { + struct mlxsw_sp_nexthop_group *nh_grp; + u32 adj_index; + u16 ecmp_size; + u16 count; + int sum_norm_weight; + u8 adj_index_valid:1, + gateway:1, /* routes using the group use a gateway */ + is_resilient:1; + struct list_head list; /* member in nh_res_grp_list */ + struct mlxsw_sp_nexthop nexthops[0]; +#define nh_rif nexthops[0].rif +}; + +struct mlxsw_sp_nexthop_group_vr_key { + u16 vr_id; + enum mlxsw_sp_l3proto proto; +}; + +struct mlxsw_sp_nexthop_group_vr_entry { + struct list_head list; /* member in vr_list */ + struct rhash_head ht_node; /* member in vr_ht */ + refcount_t ref_count; + struct mlxsw_sp_nexthop_group_vr_key key; +}; + +struct mlxsw_sp_nexthop_group { + struct rhash_head ht_node; + struct list_head fib_list; /* list of fib entries that use this group */ + union { + struct { + struct fib_info *fi; + } ipv4; + struct { + u32 id; + } obj; + }; + struct mlxsw_sp_nexthop_group_info *nhgi; + struct list_head vr_list; + struct rhashtable vr_ht; + enum mlxsw_sp_nexthop_group_type type; + bool can_destroy; +}; + +void mlxsw_sp_nexthop_counter_alloc(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + struct devlink *devlink; + + devlink = priv_to_devlink(mlxsw_sp->core); + if (!devlink_dpipe_table_counter_enabled(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_ADJ)) + return; + + if (mlxsw_sp_flow_counter_alloc(mlxsw_sp, &nh->counter_index)) + return; + + nh->counter_valid = true; +} + +void mlxsw_sp_nexthop_counter_free(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + if (!nh->counter_valid) + return; + mlxsw_sp_flow_counter_free(mlxsw_sp, nh->counter_index); + nh->counter_valid = false; +} + +int mlxsw_sp_nexthop_counter_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh, u64 *p_counter) +{ + if (!nh->counter_valid) + return -EINVAL; + + return mlxsw_sp_flow_counter_get(mlxsw_sp, nh->counter_index, + p_counter, NULL); +} + +struct mlxsw_sp_nexthop *mlxsw_sp_nexthop_next(struct mlxsw_sp_router *router, + struct mlxsw_sp_nexthop *nh) +{ + if (!nh) { + if (list_empty(&router->nexthop_list)) + return NULL; + else + return list_first_entry(&router->nexthop_list, + typeof(*nh), router_list_node); + } + if (list_is_last(&nh->router_list_node, &router->nexthop_list)) + return NULL; + return list_next_entry(nh, router_list_node); +} + +bool mlxsw_sp_nexthop_is_forward(const struct mlxsw_sp_nexthop *nh) +{ + return nh->offloaded && nh->action == MLXSW_SP_NEXTHOP_ACTION_FORWARD; +} + +unsigned char *mlxsw_sp_nexthop_ha(struct mlxsw_sp_nexthop *nh) +{ + if (nh->type != MLXSW_SP_NEXTHOP_TYPE_ETH || + !mlxsw_sp_nexthop_is_forward(nh)) + return NULL; + return nh->neigh_entry->ha; +} + +int mlxsw_sp_nexthop_indexes(struct mlxsw_sp_nexthop *nh, u32 *p_adj_index, + u32 *p_adj_size, u32 *p_adj_hash_index) +{ + struct mlxsw_sp_nexthop_group_info *nhgi = nh->nhgi; + u32 adj_hash_index = 0; + int i; + + if (!nh->offloaded || !nhgi->adj_index_valid) + return -EINVAL; + + *p_adj_index = nhgi->adj_index; + *p_adj_size = nhgi->ecmp_size; + + for (i = 0; i < nhgi->count; i++) { + struct mlxsw_sp_nexthop *nh_iter = &nhgi->nexthops[i]; + + if (nh_iter == nh) + break; + if (nh_iter->offloaded) + adj_hash_index += nh_iter->num_adj_entries; + } + + *p_adj_hash_index = adj_hash_index; + return 0; +} + +struct mlxsw_sp_rif *mlxsw_sp_nexthop_rif(struct mlxsw_sp_nexthop *nh) +{ + return nh->rif; +} + +bool mlxsw_sp_nexthop_group_has_ipip(struct mlxsw_sp_nexthop *nh) +{ + struct mlxsw_sp_nexthop_group_info *nhgi = nh->nhgi; + int i; + + for (i = 0; i < nhgi->count; i++) { + struct mlxsw_sp_nexthop *nh_iter = &nhgi->nexthops[i]; + + if (nh_iter->type == MLXSW_SP_NEXTHOP_TYPE_IPIP) + return true; + } + return false; +} + +static const struct rhashtable_params mlxsw_sp_nexthop_group_vr_ht_params = { + .key_offset = offsetof(struct mlxsw_sp_nexthop_group_vr_entry, key), + .head_offset = offsetof(struct mlxsw_sp_nexthop_group_vr_entry, ht_node), + .key_len = sizeof(struct mlxsw_sp_nexthop_group_vr_key), + .automatic_shrinking = true, +}; + +static struct mlxsw_sp_nexthop_group_vr_entry * +mlxsw_sp_nexthop_group_vr_entry_lookup(struct mlxsw_sp_nexthop_group *nh_grp, + const struct mlxsw_sp_fib *fib) +{ + struct mlxsw_sp_nexthop_group_vr_key key; + + memset(&key, 0, sizeof(key)); + key.vr_id = fib->vr->id; + key.proto = fib->proto; + return rhashtable_lookup_fast(&nh_grp->vr_ht, &key, + mlxsw_sp_nexthop_group_vr_ht_params); +} + +static int +mlxsw_sp_nexthop_group_vr_entry_create(struct mlxsw_sp_nexthop_group *nh_grp, + const struct mlxsw_sp_fib *fib) +{ + struct mlxsw_sp_nexthop_group_vr_entry *vr_entry; + int err; + + vr_entry = kzalloc(sizeof(*vr_entry), GFP_KERNEL); + if (!vr_entry) + return -ENOMEM; + + vr_entry->key.vr_id = fib->vr->id; + vr_entry->key.proto = fib->proto; + refcount_set(&vr_entry->ref_count, 1); + + err = rhashtable_insert_fast(&nh_grp->vr_ht, &vr_entry->ht_node, + mlxsw_sp_nexthop_group_vr_ht_params); + if (err) + goto err_hashtable_insert; + + list_add(&vr_entry->list, &nh_grp->vr_list); + + return 0; + +err_hashtable_insert: + kfree(vr_entry); + return err; +} + +static void +mlxsw_sp_nexthop_group_vr_entry_destroy(struct mlxsw_sp_nexthop_group *nh_grp, + struct mlxsw_sp_nexthop_group_vr_entry *vr_entry) +{ + list_del(&vr_entry->list); + rhashtable_remove_fast(&nh_grp->vr_ht, &vr_entry->ht_node, + mlxsw_sp_nexthop_group_vr_ht_params); + kfree(vr_entry); +} + +static int +mlxsw_sp_nexthop_group_vr_link(struct mlxsw_sp_nexthop_group *nh_grp, + const struct mlxsw_sp_fib *fib) +{ + struct mlxsw_sp_nexthop_group_vr_entry *vr_entry; + + vr_entry = mlxsw_sp_nexthop_group_vr_entry_lookup(nh_grp, fib); + if (vr_entry) { + refcount_inc(&vr_entry->ref_count); + return 0; + } + + return mlxsw_sp_nexthop_group_vr_entry_create(nh_grp, fib); +} + +static void +mlxsw_sp_nexthop_group_vr_unlink(struct mlxsw_sp_nexthop_group *nh_grp, + const struct mlxsw_sp_fib *fib) +{ + struct mlxsw_sp_nexthop_group_vr_entry *vr_entry; + + vr_entry = mlxsw_sp_nexthop_group_vr_entry_lookup(nh_grp, fib); + if (WARN_ON_ONCE(!vr_entry)) + return; + + if (!refcount_dec_and_test(&vr_entry->ref_count)) + return; + + mlxsw_sp_nexthop_group_vr_entry_destroy(nh_grp, vr_entry); +} + +struct mlxsw_sp_nexthop_group_cmp_arg { + enum mlxsw_sp_nexthop_group_type type; + union { + struct fib_info *fi; + struct mlxsw_sp_fib6_entry *fib6_entry; + u32 id; + }; +}; + +static bool +mlxsw_sp_nexthop6_group_has_nexthop(const struct mlxsw_sp_nexthop_group *nh_grp, + const struct in6_addr *gw, int ifindex, + int weight) +{ + int i; + + for (i = 0; i < nh_grp->nhgi->count; i++) { + const struct mlxsw_sp_nexthop *nh; + + nh = &nh_grp->nhgi->nexthops[i]; + if (nh->ifindex == ifindex && nh->nh_weight == weight && + ipv6_addr_equal(gw, (struct in6_addr *) nh->gw_addr)) + return true; + } + + return false; +} + +static bool +mlxsw_sp_nexthop6_group_cmp(const struct mlxsw_sp_nexthop_group *nh_grp, + const struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + if (nh_grp->nhgi->count != fib6_entry->nrt6) + return false; + + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { + struct fib6_nh *fib6_nh = mlxsw_sp_rt6->rt->fib6_nh; + struct in6_addr *gw; + int ifindex, weight; + + ifindex = fib6_nh->fib_nh_dev->ifindex; + weight = fib6_nh->fib_nh_weight; + gw = &fib6_nh->fib_nh_gw6; + if (!mlxsw_sp_nexthop6_group_has_nexthop(nh_grp, gw, ifindex, + weight)) + return false; + } + + return true; +} + +static int +mlxsw_sp_nexthop_group_cmp(struct rhashtable_compare_arg *arg, const void *ptr) +{ + const struct mlxsw_sp_nexthop_group_cmp_arg *cmp_arg = arg->key; + const struct mlxsw_sp_nexthop_group *nh_grp = ptr; + + if (nh_grp->type != cmp_arg->type) + return 1; + + switch (cmp_arg->type) { + case MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV4: + return cmp_arg->fi != nh_grp->ipv4.fi; + case MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV6: + return !mlxsw_sp_nexthop6_group_cmp(nh_grp, + cmp_arg->fib6_entry); + case MLXSW_SP_NEXTHOP_GROUP_TYPE_OBJ: + return cmp_arg->id != nh_grp->obj.id; + default: + WARN_ON(1); + return 1; + } +} + +static u32 mlxsw_sp_nexthop_group_hash_obj(const void *data, u32 len, u32 seed) +{ + const struct mlxsw_sp_nexthop_group *nh_grp = data; + const struct mlxsw_sp_nexthop *nh; + struct fib_info *fi; + unsigned int val; + int i; + + switch (nh_grp->type) { + case MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV4: + fi = nh_grp->ipv4.fi; + return jhash(&fi, sizeof(fi), seed); + case MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV6: + val = nh_grp->nhgi->count; + for (i = 0; i < nh_grp->nhgi->count; i++) { + nh = &nh_grp->nhgi->nexthops[i]; + val ^= jhash(&nh->ifindex, sizeof(nh->ifindex), seed); + val ^= jhash(&nh->gw_addr, sizeof(nh->gw_addr), seed); + } + return jhash(&val, sizeof(val), seed); + case MLXSW_SP_NEXTHOP_GROUP_TYPE_OBJ: + return jhash(&nh_grp->obj.id, sizeof(nh_grp->obj.id), seed); + default: + WARN_ON(1); + return 0; + } +} + +static u32 +mlxsw_sp_nexthop6_group_hash(struct mlxsw_sp_fib6_entry *fib6_entry, u32 seed) +{ + unsigned int val = fib6_entry->nrt6; + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { + struct fib6_nh *fib6_nh = mlxsw_sp_rt6->rt->fib6_nh; + struct net_device *dev = fib6_nh->fib_nh_dev; + struct in6_addr *gw = &fib6_nh->fib_nh_gw6; + + val ^= jhash(&dev->ifindex, sizeof(dev->ifindex), seed); + val ^= jhash(gw, sizeof(*gw), seed); + } + + return jhash(&val, sizeof(val), seed); +} + +static u32 +mlxsw_sp_nexthop_group_hash(const void *data, u32 len, u32 seed) +{ + const struct mlxsw_sp_nexthop_group_cmp_arg *cmp_arg = data; + + switch (cmp_arg->type) { + case MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV4: + return jhash(&cmp_arg->fi, sizeof(cmp_arg->fi), seed); + case MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV6: + return mlxsw_sp_nexthop6_group_hash(cmp_arg->fib6_entry, seed); + case MLXSW_SP_NEXTHOP_GROUP_TYPE_OBJ: + return jhash(&cmp_arg->id, sizeof(cmp_arg->id), seed); + default: + WARN_ON(1); + return 0; + } +} + +static const struct rhashtable_params mlxsw_sp_nexthop_group_ht_params = { + .head_offset = offsetof(struct mlxsw_sp_nexthop_group, ht_node), + .hashfn = mlxsw_sp_nexthop_group_hash, + .obj_hashfn = mlxsw_sp_nexthop_group_hash_obj, + .obj_cmpfn = mlxsw_sp_nexthop_group_cmp, +}; + +static int mlxsw_sp_nexthop_group_insert(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + if (nh_grp->type == MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV6 && + !nh_grp->nhgi->gateway) + return 0; + + return rhashtable_insert_fast(&mlxsw_sp->router->nexthop_group_ht, + &nh_grp->ht_node, + mlxsw_sp_nexthop_group_ht_params); +} + +static void mlxsw_sp_nexthop_group_remove(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + if (nh_grp->type == MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV6 && + !nh_grp->nhgi->gateway) + return; + + rhashtable_remove_fast(&mlxsw_sp->router->nexthop_group_ht, + &nh_grp->ht_node, + mlxsw_sp_nexthop_group_ht_params); +} + +static struct mlxsw_sp_nexthop_group * +mlxsw_sp_nexthop4_group_lookup(struct mlxsw_sp *mlxsw_sp, + struct fib_info *fi) +{ + struct mlxsw_sp_nexthop_group_cmp_arg cmp_arg; + + cmp_arg.type = MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV4; + cmp_arg.fi = fi; + return rhashtable_lookup_fast(&mlxsw_sp->router->nexthop_group_ht, + &cmp_arg, + mlxsw_sp_nexthop_group_ht_params); +} + +static struct mlxsw_sp_nexthop_group * +mlxsw_sp_nexthop6_group_lookup(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_nexthop_group_cmp_arg cmp_arg; + + cmp_arg.type = MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV6; + cmp_arg.fib6_entry = fib6_entry; + return rhashtable_lookup_fast(&mlxsw_sp->router->nexthop_group_ht, + &cmp_arg, + mlxsw_sp_nexthop_group_ht_params); +} + +static const struct rhashtable_params mlxsw_sp_nexthop_ht_params = { + .key_offset = offsetof(struct mlxsw_sp_nexthop, key), + .head_offset = offsetof(struct mlxsw_sp_nexthop, ht_node), + .key_len = sizeof(struct mlxsw_sp_nexthop_key), +}; + +static int mlxsw_sp_nexthop_insert(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + return rhashtable_insert_fast(&mlxsw_sp->router->nexthop_ht, + &nh->ht_node, mlxsw_sp_nexthop_ht_params); +} + +static void mlxsw_sp_nexthop_remove(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + rhashtable_remove_fast(&mlxsw_sp->router->nexthop_ht, &nh->ht_node, + mlxsw_sp_nexthop_ht_params); +} + +static struct mlxsw_sp_nexthop * +mlxsw_sp_nexthop_lookup(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_key key) +{ + return rhashtable_lookup_fast(&mlxsw_sp->router->nexthop_ht, &key, + mlxsw_sp_nexthop_ht_params); +} + +static int mlxsw_sp_adj_index_mass_update_vr(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_l3proto proto, + u16 vr_id, + u32 adj_index, u16 ecmp_size, + u32 new_adj_index, + u16 new_ecmp_size) +{ + char raleu_pl[MLXSW_REG_RALEU_LEN]; + + mlxsw_reg_raleu_pack(raleu_pl, + (enum mlxsw_reg_ralxx_protocol) proto, vr_id, + adj_index, ecmp_size, new_adj_index, + new_ecmp_size); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(raleu), raleu_pl); +} + +static int mlxsw_sp_adj_index_mass_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp, + u32 old_adj_index, u16 old_ecmp_size) +{ + struct mlxsw_sp_nexthop_group_info *nhgi = nh_grp->nhgi; + struct mlxsw_sp_nexthop_group_vr_entry *vr_entry; + int err; + + list_for_each_entry(vr_entry, &nh_grp->vr_list, list) { + err = mlxsw_sp_adj_index_mass_update_vr(mlxsw_sp, + vr_entry->key.proto, + vr_entry->key.vr_id, + old_adj_index, + old_ecmp_size, + nhgi->adj_index, + nhgi->ecmp_size); + if (err) + goto err_mass_update_vr; + } + return 0; + +err_mass_update_vr: + list_for_each_entry_continue_reverse(vr_entry, &nh_grp->vr_list, list) + mlxsw_sp_adj_index_mass_update_vr(mlxsw_sp, vr_entry->key.proto, + vr_entry->key.vr_id, + nhgi->adj_index, + nhgi->ecmp_size, + old_adj_index, old_ecmp_size); + return err; +} + +static int __mlxsw_sp_nexthop_eth_update(struct mlxsw_sp *mlxsw_sp, + u32 adj_index, + struct mlxsw_sp_nexthop *nh, + bool force, char *ratr_pl) +{ + struct mlxsw_sp_neigh_entry *neigh_entry = nh->neigh_entry; + enum mlxsw_reg_ratr_op op; + u16 rif_index; + + rif_index = nh->rif ? nh->rif->rif_index : + mlxsw_sp->router->lb_rif_index; + op = force ? MLXSW_REG_RATR_OP_WRITE_WRITE_ENTRY : + MLXSW_REG_RATR_OP_WRITE_WRITE_ENTRY_ON_ACTIVITY; + mlxsw_reg_ratr_pack(ratr_pl, op, true, MLXSW_REG_RATR_TYPE_ETHERNET, + adj_index, rif_index); + switch (nh->action) { + case MLXSW_SP_NEXTHOP_ACTION_FORWARD: + mlxsw_reg_ratr_eth_entry_pack(ratr_pl, neigh_entry->ha); + break; + case MLXSW_SP_NEXTHOP_ACTION_DISCARD: + mlxsw_reg_ratr_trap_action_set(ratr_pl, + MLXSW_REG_RATR_TRAP_ACTION_DISCARD_ERRORS); + break; + case MLXSW_SP_NEXTHOP_ACTION_TRAP: + mlxsw_reg_ratr_trap_action_set(ratr_pl, + MLXSW_REG_RATR_TRAP_ACTION_TRAP); + mlxsw_reg_ratr_trap_id_set(ratr_pl, MLXSW_TRAP_ID_RTR_EGRESS0); + break; + default: + WARN_ON_ONCE(1); + return -EINVAL; + } + if (nh->counter_valid) + mlxsw_reg_ratr_counter_pack(ratr_pl, nh->counter_index, true); + else + mlxsw_reg_ratr_counter_pack(ratr_pl, 0, false); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ratr), ratr_pl); +} + +int mlxsw_sp_nexthop_eth_update(struct mlxsw_sp *mlxsw_sp, u32 adj_index, + struct mlxsw_sp_nexthop *nh, bool force, + char *ratr_pl) +{ + int i; + + for (i = 0; i < nh->num_adj_entries; i++) { + int err; + + err = __mlxsw_sp_nexthop_eth_update(mlxsw_sp, adj_index + i, + nh, force, ratr_pl); + if (err) + return err; + } + + return 0; +} + +static int __mlxsw_sp_nexthop_ipip_update(struct mlxsw_sp *mlxsw_sp, + u32 adj_index, + struct mlxsw_sp_nexthop *nh, + bool force, char *ratr_pl) +{ + const struct mlxsw_sp_ipip_ops *ipip_ops; + + ipip_ops = mlxsw_sp->router->ipip_ops_arr[nh->ipip_entry->ipipt]; + return ipip_ops->nexthop_update(mlxsw_sp, adj_index, nh->ipip_entry, + force, ratr_pl); +} + +static int mlxsw_sp_nexthop_ipip_update(struct mlxsw_sp *mlxsw_sp, + u32 adj_index, + struct mlxsw_sp_nexthop *nh, bool force, + char *ratr_pl) +{ + int i; + + for (i = 0; i < nh->num_adj_entries; i++) { + int err; + + err = __mlxsw_sp_nexthop_ipip_update(mlxsw_sp, adj_index + i, + nh, force, ratr_pl); + if (err) + return err; + } + + return 0; +} + +static int mlxsw_sp_nexthop_update(struct mlxsw_sp *mlxsw_sp, u32 adj_index, + struct mlxsw_sp_nexthop *nh, bool force, + char *ratr_pl) +{ + /* When action is discard or trap, the nexthop must be + * programmed as an Ethernet nexthop. + */ + if (nh->type == MLXSW_SP_NEXTHOP_TYPE_ETH || + nh->action == MLXSW_SP_NEXTHOP_ACTION_DISCARD || + nh->action == MLXSW_SP_NEXTHOP_ACTION_TRAP) + return mlxsw_sp_nexthop_eth_update(mlxsw_sp, adj_index, nh, + force, ratr_pl); + else + return mlxsw_sp_nexthop_ipip_update(mlxsw_sp, adj_index, nh, + force, ratr_pl); +} + +static int +mlxsw_sp_nexthop_group_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group_info *nhgi, + bool reallocate) +{ + char ratr_pl[MLXSW_REG_RATR_LEN]; + u32 adj_index = nhgi->adj_index; /* base */ + struct mlxsw_sp_nexthop *nh; + int i; + + for (i = 0; i < nhgi->count; i++) { + nh = &nhgi->nexthops[i]; + + if (!nh->should_offload) { + nh->offloaded = 0; + continue; + } + + if (nh->update || reallocate) { + int err = 0; + + err = mlxsw_sp_nexthop_update(mlxsw_sp, adj_index, nh, + true, ratr_pl); + if (err) + return err; + nh->update = 0; + nh->offloaded = 1; + } + adj_index += nh->num_adj_entries; + } + return 0; +} + +static int +mlxsw_sp_nexthop_fib_entries_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + struct mlxsw_sp_fib_entry *fib_entry; + int err; + + list_for_each_entry(fib_entry, &nh_grp->fib_list, nexthop_group_node) { + err = mlxsw_sp_fib_entry_update(mlxsw_sp, fib_entry); + if (err) + return err; + } + return 0; +} + +struct mlxsw_sp_adj_grp_size_range { + u16 start; /* Inclusive */ + u16 end; /* Inclusive */ +}; + +/* Ordered by range start value */ +static const struct mlxsw_sp_adj_grp_size_range +mlxsw_sp1_adj_grp_size_ranges[] = { + { .start = 1, .end = 64 }, + { .start = 512, .end = 512 }, + { .start = 1024, .end = 1024 }, + { .start = 2048, .end = 2048 }, + { .start = 4096, .end = 4096 }, +}; + +/* Ordered by range start value */ +static const struct mlxsw_sp_adj_grp_size_range +mlxsw_sp2_adj_grp_size_ranges[] = { + { .start = 1, .end = 128 }, + { .start = 256, .end = 256 }, + { .start = 512, .end = 512 }, + { .start = 1024, .end = 1024 }, + { .start = 2048, .end = 2048 }, + { .start = 4096, .end = 4096 }, +}; + +static void mlxsw_sp_adj_grp_size_round_up(const struct mlxsw_sp *mlxsw_sp, + u16 *p_adj_grp_size) +{ + int i; + + for (i = 0; i < mlxsw_sp->router->adj_grp_size_ranges_count; i++) { + const struct mlxsw_sp_adj_grp_size_range *size_range; + + size_range = &mlxsw_sp->router->adj_grp_size_ranges[i]; + + if (*p_adj_grp_size >= size_range->start && + *p_adj_grp_size <= size_range->end) + return; + + if (*p_adj_grp_size <= size_range->end) { + *p_adj_grp_size = size_range->end; + return; + } + } +} + +static void mlxsw_sp_adj_grp_size_round_down(const struct mlxsw_sp *mlxsw_sp, + u16 *p_adj_grp_size, + unsigned int alloc_size) +{ + int i; + + for (i = mlxsw_sp->router->adj_grp_size_ranges_count - 1; i >= 0; i--) { + const struct mlxsw_sp_adj_grp_size_range *size_range; + + size_range = &mlxsw_sp->router->adj_grp_size_ranges[i]; + + if (alloc_size >= size_range->end) { + *p_adj_grp_size = size_range->end; + return; + } + } +} + +static int mlxsw_sp_fix_adj_grp_size(struct mlxsw_sp *mlxsw_sp, + u16 *p_adj_grp_size) +{ + unsigned int alloc_size; + int err; + + /* Round up the requested group size to the next size supported + * by the device and make sure the request can be satisfied. + */ + mlxsw_sp_adj_grp_size_round_up(mlxsw_sp, p_adj_grp_size); + err = mlxsw_sp_kvdl_alloc_count_query(mlxsw_sp, + MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, + *p_adj_grp_size, &alloc_size); + if (err) + return err; + /* It is possible the allocation results in more allocated + * entries than requested. Try to use as much of them as + * possible. + */ + mlxsw_sp_adj_grp_size_round_down(mlxsw_sp, p_adj_grp_size, alloc_size); + + return 0; +} + +static void +mlxsw_sp_nexthop_group_normalize(struct mlxsw_sp_nexthop_group_info *nhgi) +{ + int i, g = 0, sum_norm_weight = 0; + struct mlxsw_sp_nexthop *nh; + + for (i = 0; i < nhgi->count; i++) { + nh = &nhgi->nexthops[i]; + + if (!nh->should_offload) + continue; + if (g > 0) + g = gcd(nh->nh_weight, g); + else + g = nh->nh_weight; + } + + for (i = 0; i < nhgi->count; i++) { + nh = &nhgi->nexthops[i]; + + if (!nh->should_offload) + continue; + nh->norm_nh_weight = nh->nh_weight / g; + sum_norm_weight += nh->norm_nh_weight; + } + + nhgi->sum_norm_weight = sum_norm_weight; +} + +static void +mlxsw_sp_nexthop_group_rebalance(struct mlxsw_sp_nexthop_group_info *nhgi) +{ + int i, weight = 0, lower_bound = 0; + int total = nhgi->sum_norm_weight; + u16 ecmp_size = nhgi->ecmp_size; + + for (i = 0; i < nhgi->count; i++) { + struct mlxsw_sp_nexthop *nh = &nhgi->nexthops[i]; + int upper_bound; + + if (!nh->should_offload) + continue; + weight += nh->norm_nh_weight; + upper_bound = DIV_ROUND_CLOSEST(ecmp_size * weight, total); + nh->num_adj_entries = upper_bound - lower_bound; + lower_bound = upper_bound; + } +} + +static struct mlxsw_sp_nexthop * +mlxsw_sp_rt6_nexthop(struct mlxsw_sp_nexthop_group *nh_grp, + const struct mlxsw_sp_rt6 *mlxsw_sp_rt6); + +static void +mlxsw_sp_nexthop4_group_offload_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + int i; + + for (i = 0; i < nh_grp->nhgi->count; i++) { + struct mlxsw_sp_nexthop *nh = &nh_grp->nhgi->nexthops[i]; + + if (nh->offloaded) + nh->key.fib_nh->fib_nh_flags |= RTNH_F_OFFLOAD; + else + nh->key.fib_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; + } +} + +static void +__mlxsw_sp_nexthop6_group_offload_refresh(struct mlxsw_sp_nexthop_group *nh_grp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { + struct fib6_nh *fib6_nh = mlxsw_sp_rt6->rt->fib6_nh; + struct mlxsw_sp_nexthop *nh; + + nh = mlxsw_sp_rt6_nexthop(nh_grp, mlxsw_sp_rt6); + if (nh && nh->offloaded) + fib6_nh->fib_nh_flags |= RTNH_F_OFFLOAD; + else + fib6_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; + } +} + +static void +mlxsw_sp_nexthop6_group_offload_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + + /* Unfortunately, in IPv6 the route and the nexthop are described by + * the same struct, so we need to iterate over all the routes using the + * nexthop group and set / clear the offload indication for them. + */ + list_for_each_entry(fib6_entry, &nh_grp->fib_list, + common.nexthop_group_node) + __mlxsw_sp_nexthop6_group_offload_refresh(nh_grp, fib6_entry); +} + +static void +mlxsw_sp_nexthop_bucket_offload_refresh(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_nexthop *nh, + u16 bucket_index) +{ + struct mlxsw_sp_nexthop_group *nh_grp = nh->nhgi->nh_grp; + bool offload = false, trap = false; + + if (nh->offloaded) { + if (nh->action == MLXSW_SP_NEXTHOP_ACTION_TRAP) + trap = true; + else + offload = true; + } + nexthop_bucket_set_hw_flags(mlxsw_sp_net(mlxsw_sp), nh_grp->obj.id, + bucket_index, offload, trap); +} + +static void +mlxsw_sp_nexthop_obj_group_offload_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + int i; + + /* Do not update the flags if the nexthop group is being destroyed + * since: + * 1. The nexthop objects is being deleted, in which case the flags are + * irrelevant. + * 2. The nexthop group was replaced by a newer group, in which case + * the flags of the nexthop object were already updated based on the + * new group. + */ + if (nh_grp->can_destroy) + return; + + nexthop_set_hw_flags(mlxsw_sp_net(mlxsw_sp), nh_grp->obj.id, + nh_grp->nhgi->adj_index_valid, false); + + /* Update flags of individual nexthop buckets in case of a resilient + * nexthop group. + */ + if (!nh_grp->nhgi->is_resilient) + return; + + for (i = 0; i < nh_grp->nhgi->count; i++) { + struct mlxsw_sp_nexthop *nh = &nh_grp->nhgi->nexthops[i]; + + mlxsw_sp_nexthop_bucket_offload_refresh(mlxsw_sp, nh, i); + } +} + +static void +mlxsw_sp_nexthop_group_offload_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + switch (nh_grp->type) { + case MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV4: + mlxsw_sp_nexthop4_group_offload_refresh(mlxsw_sp, nh_grp); + break; + case MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV6: + mlxsw_sp_nexthop6_group_offload_refresh(mlxsw_sp, nh_grp); + break; + case MLXSW_SP_NEXTHOP_GROUP_TYPE_OBJ: + mlxsw_sp_nexthop_obj_group_offload_refresh(mlxsw_sp, nh_grp); + break; + } +} + +static int +mlxsw_sp_nexthop_group_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + struct mlxsw_sp_nexthop_group_info *nhgi = nh_grp->nhgi; + u16 ecmp_size, old_ecmp_size; + struct mlxsw_sp_nexthop *nh; + bool offload_change = false; + u32 adj_index; + bool old_adj_index_valid; + u32 old_adj_index; + int i, err2, err; + + if (!nhgi->gateway) + return mlxsw_sp_nexthop_fib_entries_update(mlxsw_sp, nh_grp); + + for (i = 0; i < nhgi->count; i++) { + nh = &nhgi->nexthops[i]; + + if (nh->should_offload != nh->offloaded) { + offload_change = true; + if (nh->should_offload) + nh->update = 1; + } + } + if (!offload_change) { + /* Nothing was added or removed, so no need to reallocate. Just + * update MAC on existing adjacency indexes. + */ + err = mlxsw_sp_nexthop_group_update(mlxsw_sp, nhgi, false); + if (err) { + dev_warn(mlxsw_sp->bus_info->dev, "Failed to update neigh MAC in adjacency table.\n"); + goto set_trap; + } + /* Flags of individual nexthop buckets might need to be + * updated. + */ + mlxsw_sp_nexthop_group_offload_refresh(mlxsw_sp, nh_grp); + return 0; + } + mlxsw_sp_nexthop_group_normalize(nhgi); + if (!nhgi->sum_norm_weight) { + /* No neigh of this group is connected so we just set + * the trap and let everthing flow through kernel. + */ + err = 0; + goto set_trap; + } + + ecmp_size = nhgi->sum_norm_weight; + err = mlxsw_sp_fix_adj_grp_size(mlxsw_sp, &ecmp_size); + if (err) + /* No valid allocation size available. */ + goto set_trap; + + err = mlxsw_sp_kvdl_alloc(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, + ecmp_size, &adj_index); + if (err) { + /* We ran out of KVD linear space, just set the + * trap and let everything flow through kernel. + */ + dev_warn(mlxsw_sp->bus_info->dev, "Failed to allocate KVD linear area for nexthop group.\n"); + goto set_trap; + } + old_adj_index_valid = nhgi->adj_index_valid; + old_adj_index = nhgi->adj_index; + old_ecmp_size = nhgi->ecmp_size; + nhgi->adj_index_valid = 1; + nhgi->adj_index = adj_index; + nhgi->ecmp_size = ecmp_size; + mlxsw_sp_nexthop_group_rebalance(nhgi); + err = mlxsw_sp_nexthop_group_update(mlxsw_sp, nhgi, true); + if (err) { + dev_warn(mlxsw_sp->bus_info->dev, "Failed to update neigh MAC in adjacency table.\n"); + goto set_trap; + } + + mlxsw_sp_nexthop_group_offload_refresh(mlxsw_sp, nh_grp); + + if (!old_adj_index_valid) { + /* The trap was set for fib entries, so we have to call + * fib entry update to unset it and use adjacency index. + */ + err = mlxsw_sp_nexthop_fib_entries_update(mlxsw_sp, nh_grp); + if (err) { + dev_warn(mlxsw_sp->bus_info->dev, "Failed to add adjacency index to fib entries.\n"); + goto set_trap; + } + return 0; + } + + err = mlxsw_sp_adj_index_mass_update(mlxsw_sp, nh_grp, + old_adj_index, old_ecmp_size); + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, + old_ecmp_size, old_adj_index); + if (err) { + dev_warn(mlxsw_sp->bus_info->dev, "Failed to mass-update adjacency index for nexthop group.\n"); + goto set_trap; + } + + return 0; + +set_trap: + old_adj_index_valid = nhgi->adj_index_valid; + nhgi->adj_index_valid = 0; + for (i = 0; i < nhgi->count; i++) { + nh = &nhgi->nexthops[i]; + nh->offloaded = 0; + } + err2 = mlxsw_sp_nexthop_fib_entries_update(mlxsw_sp, nh_grp); + if (err2) + dev_warn(mlxsw_sp->bus_info->dev, "Failed to set traps for fib entries.\n"); + mlxsw_sp_nexthop_group_offload_refresh(mlxsw_sp, nh_grp); + if (old_adj_index_valid) + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, + nhgi->ecmp_size, nhgi->adj_index); + return err; +} + +static void __mlxsw_sp_nexthop_neigh_update(struct mlxsw_sp_nexthop *nh, + bool removing) +{ + if (!removing) { + nh->action = MLXSW_SP_NEXTHOP_ACTION_FORWARD; + nh->should_offload = 1; + } else if (nh->nhgi->is_resilient) { + nh->action = MLXSW_SP_NEXTHOP_ACTION_TRAP; + nh->should_offload = 1; + } else { + nh->should_offload = 0; + } + nh->update = 1; +} + +static int +mlxsw_sp_nexthop_dead_neigh_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry) +{ + struct neighbour *n, *old_n = neigh_entry->key.n; + struct mlxsw_sp_nexthop *nh; + bool entry_connected; + u8 nud_state, dead; + int err; + + nh = list_first_entry(&neigh_entry->nexthop_list, + struct mlxsw_sp_nexthop, neigh_list_node); + + n = neigh_lookup(nh->neigh_tbl, &nh->gw_addr, nh->rif->dev); + if (!n) { + n = neigh_create(nh->neigh_tbl, &nh->gw_addr, nh->rif->dev); + if (IS_ERR(n)) + return PTR_ERR(n); + neigh_event_send(n, NULL); + } + + mlxsw_sp_neigh_entry_remove(mlxsw_sp, neigh_entry); + neigh_entry->key.n = n; + err = mlxsw_sp_neigh_entry_insert(mlxsw_sp, neigh_entry); + if (err) + goto err_neigh_entry_insert; + + read_lock_bh(&n->lock); + nud_state = n->nud_state; + dead = n->dead; + read_unlock_bh(&n->lock); + entry_connected = nud_state & NUD_VALID && !dead; + + list_for_each_entry(nh, &neigh_entry->nexthop_list, + neigh_list_node) { + neigh_release(old_n); + neigh_clone(n); + __mlxsw_sp_nexthop_neigh_update(nh, !entry_connected); + mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh->nhgi->nh_grp); + } + + neigh_release(n); + + return 0; + +err_neigh_entry_insert: + neigh_entry->key.n = old_n; + mlxsw_sp_neigh_entry_insert(mlxsw_sp, neigh_entry); + neigh_release(n); + return err; +} + +static void +mlxsw_sp_nexthop_neigh_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry, + bool removing, bool dead) +{ + struct mlxsw_sp_nexthop *nh; + + if (list_empty(&neigh_entry->nexthop_list)) + return; + + if (dead) { + int err; + + err = mlxsw_sp_nexthop_dead_neigh_replace(mlxsw_sp, + neigh_entry); + if (err) + dev_err(mlxsw_sp->bus_info->dev, "Failed to replace dead neigh\n"); + return; + } + + list_for_each_entry(nh, &neigh_entry->nexthop_list, + neigh_list_node) { + __mlxsw_sp_nexthop_neigh_update(nh, removing); + mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh->nhgi->nh_grp); + } +} + +static void mlxsw_sp_nexthop_rif_init(struct mlxsw_sp_nexthop *nh, + struct mlxsw_sp_rif *rif) +{ + if (nh->rif) + return; + + nh->rif = rif; + list_add(&nh->rif_list_node, &rif->nexthop_list); +} + +static void mlxsw_sp_nexthop_rif_fini(struct mlxsw_sp_nexthop *nh) +{ + if (!nh->rif) + return; + + list_del(&nh->rif_list_node); + nh->rif = NULL; +} + +static int mlxsw_sp_nexthop_neigh_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + struct mlxsw_sp_neigh_entry *neigh_entry; + struct neighbour *n; + u8 nud_state, dead; + int err; + + if (!nh->nhgi->gateway || nh->neigh_entry) + return 0; + + /* Take a reference of neigh here ensuring that neigh would + * not be destructed before the nexthop entry is finished. + * The reference is taken either in neigh_lookup() or + * in neigh_create() in case n is not found. + */ + n = neigh_lookup(nh->neigh_tbl, &nh->gw_addr, nh->rif->dev); + if (!n) { + n = neigh_create(nh->neigh_tbl, &nh->gw_addr, nh->rif->dev); + if (IS_ERR(n)) + return PTR_ERR(n); + neigh_event_send(n, NULL); + } + neigh_entry = mlxsw_sp_neigh_entry_lookup(mlxsw_sp, n); + if (!neigh_entry) { + neigh_entry = mlxsw_sp_neigh_entry_create(mlxsw_sp, n); + if (IS_ERR(neigh_entry)) { + err = -EINVAL; + goto err_neigh_entry_create; + } + } + + /* If that is the first nexthop connected to that neigh, add to + * nexthop_neighs_list + */ + if (list_empty(&neigh_entry->nexthop_list)) + list_add_tail(&neigh_entry->nexthop_neighs_list_node, + &mlxsw_sp->router->nexthop_neighs_list); + + nh->neigh_entry = neigh_entry; + list_add_tail(&nh->neigh_list_node, &neigh_entry->nexthop_list); + read_lock_bh(&n->lock); + nud_state = n->nud_state; + dead = n->dead; + read_unlock_bh(&n->lock); + __mlxsw_sp_nexthop_neigh_update(nh, !(nud_state & NUD_VALID && !dead)); + + return 0; + +err_neigh_entry_create: + neigh_release(n); + return err; +} + +static void mlxsw_sp_nexthop_neigh_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + struct mlxsw_sp_neigh_entry *neigh_entry = nh->neigh_entry; + struct neighbour *n; + + if (!neigh_entry) + return; + n = neigh_entry->key.n; + + __mlxsw_sp_nexthop_neigh_update(nh, true); + list_del(&nh->neigh_list_node); + nh->neigh_entry = NULL; + + /* If that is the last nexthop connected to that neigh, remove from + * nexthop_neighs_list + */ + if (list_empty(&neigh_entry->nexthop_list)) + list_del(&neigh_entry->nexthop_neighs_list_node); + + if (!neigh_entry->connected && list_empty(&neigh_entry->nexthop_list)) + mlxsw_sp_neigh_entry_destroy(mlxsw_sp, neigh_entry); + + neigh_release(n); +} + +static bool mlxsw_sp_ipip_netdev_ul_up(struct net_device *ol_dev) +{ + struct net_device *ul_dev; + bool is_up; + + rcu_read_lock(); + ul_dev = mlxsw_sp_ipip_netdev_ul_dev_get(ol_dev); + is_up = ul_dev ? (ul_dev->flags & IFF_UP) : true; + rcu_read_unlock(); + + return is_up; +} + +static void mlxsw_sp_nexthop_ipip_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh, + struct mlxsw_sp_ipip_entry *ipip_entry) +{ + bool removing; + + if (!nh->nhgi->gateway || nh->ipip_entry) + return; + + nh->ipip_entry = ipip_entry; + removing = !mlxsw_sp_ipip_netdev_ul_up(ipip_entry->ol_dev); + __mlxsw_sp_nexthop_neigh_update(nh, removing); + mlxsw_sp_nexthop_rif_init(nh, &ipip_entry->ol_lb->common); +} + +static void mlxsw_sp_nexthop_ipip_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + struct mlxsw_sp_ipip_entry *ipip_entry = nh->ipip_entry; + + if (!ipip_entry) + return; + + __mlxsw_sp_nexthop_neigh_update(nh, true); + nh->ipip_entry = NULL; +} + +static bool mlxsw_sp_nexthop4_ipip_type(const struct mlxsw_sp *mlxsw_sp, + const struct fib_nh *fib_nh, + enum mlxsw_sp_ipip_type *p_ipipt) +{ + struct net_device *dev = fib_nh->fib_nh_dev; + + return dev && + fib_nh->nh_parent->fib_type == RTN_UNICAST && + mlxsw_sp_netdev_ipip_type(mlxsw_sp, dev, p_ipipt); +} + +static int mlxsw_sp_nexthop_type_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh, + const struct net_device *dev) +{ + const struct mlxsw_sp_ipip_ops *ipip_ops; + struct mlxsw_sp_ipip_entry *ipip_entry; + struct mlxsw_sp_rif *rif; + int err; + + ipip_entry = mlxsw_sp_ipip_entry_find_by_ol_dev(mlxsw_sp, dev); + if (ipip_entry) { + ipip_ops = mlxsw_sp->router->ipip_ops_arr[ipip_entry->ipipt]; + if (ipip_ops->can_offload(mlxsw_sp, dev)) { + nh->type = MLXSW_SP_NEXTHOP_TYPE_IPIP; + mlxsw_sp_nexthop_ipip_init(mlxsw_sp, nh, ipip_entry); + return 0; + } + } + + nh->type = MLXSW_SP_NEXTHOP_TYPE_ETH; + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, dev); + if (!rif) + return 0; + + mlxsw_sp_nexthop_rif_init(nh, rif); + err = mlxsw_sp_nexthop_neigh_init(mlxsw_sp, nh); + if (err) + goto err_neigh_init; + + return 0; + +err_neigh_init: + mlxsw_sp_nexthop_rif_fini(nh); + return err; +} + +static void mlxsw_sp_nexthop_type_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + switch (nh->type) { + case MLXSW_SP_NEXTHOP_TYPE_ETH: + mlxsw_sp_nexthop_neigh_fini(mlxsw_sp, nh); + mlxsw_sp_nexthop_rif_fini(nh); + break; + case MLXSW_SP_NEXTHOP_TYPE_IPIP: + mlxsw_sp_nexthop_rif_fini(nh); + mlxsw_sp_nexthop_ipip_fini(mlxsw_sp, nh); + break; + } +} + +static int mlxsw_sp_nexthop4_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp, + struct mlxsw_sp_nexthop *nh, + struct fib_nh *fib_nh) +{ + struct net_device *dev = fib_nh->fib_nh_dev; + struct in_device *in_dev; + int err; + + nh->nhgi = nh_grp->nhgi; + nh->key.fib_nh = fib_nh; +#ifdef CONFIG_IP_ROUTE_MULTIPATH + nh->nh_weight = fib_nh->fib_nh_weight; +#else + nh->nh_weight = 1; +#endif + memcpy(&nh->gw_addr, &fib_nh->fib_nh_gw4, sizeof(fib_nh->fib_nh_gw4)); + nh->neigh_tbl = &arp_tbl; + err = mlxsw_sp_nexthop_insert(mlxsw_sp, nh); + if (err) + return err; + + mlxsw_sp_nexthop_counter_alloc(mlxsw_sp, nh); + list_add_tail(&nh->router_list_node, &mlxsw_sp->router->nexthop_list); + + if (!dev) + return 0; + nh->ifindex = dev->ifindex; + + rcu_read_lock(); + in_dev = __in_dev_get_rcu(dev); + if (in_dev && IN_DEV_IGNORE_ROUTES_WITH_LINKDOWN(in_dev) && + fib_nh->fib_nh_flags & RTNH_F_LINKDOWN) { + rcu_read_unlock(); + return 0; + } + rcu_read_unlock(); + + err = mlxsw_sp_nexthop_type_init(mlxsw_sp, nh, dev); + if (err) + goto err_nexthop_neigh_init; + + return 0; + +err_nexthop_neigh_init: + list_del(&nh->router_list_node); + mlxsw_sp_nexthop_counter_free(mlxsw_sp, nh); + mlxsw_sp_nexthop_remove(mlxsw_sp, nh); + return err; +} + +static void mlxsw_sp_nexthop4_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + mlxsw_sp_nexthop_type_fini(mlxsw_sp, nh); + list_del(&nh->router_list_node); + mlxsw_sp_nexthop_counter_free(mlxsw_sp, nh); + mlxsw_sp_nexthop_remove(mlxsw_sp, nh); +} + +static void mlxsw_sp_nexthop4_event(struct mlxsw_sp *mlxsw_sp, + unsigned long event, struct fib_nh *fib_nh) +{ + struct mlxsw_sp_nexthop_key key; + struct mlxsw_sp_nexthop *nh; + + key.fib_nh = fib_nh; + nh = mlxsw_sp_nexthop_lookup(mlxsw_sp, key); + if (!nh) + return; + + switch (event) { + case FIB_EVENT_NH_ADD: + mlxsw_sp_nexthop_type_init(mlxsw_sp, nh, fib_nh->fib_nh_dev); + break; + case FIB_EVENT_NH_DEL: + mlxsw_sp_nexthop_type_fini(mlxsw_sp, nh); + break; + } + + mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh->nhgi->nh_grp); +} + +static void mlxsw_sp_nexthop_rif_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp_nexthop *nh; + bool removing; + + list_for_each_entry(nh, &rif->nexthop_list, rif_list_node) { + switch (nh->type) { + case MLXSW_SP_NEXTHOP_TYPE_ETH: + removing = false; + break; + case MLXSW_SP_NEXTHOP_TYPE_IPIP: + removing = !mlxsw_sp_ipip_netdev_ul_up(rif->dev); + break; + default: + WARN_ON(1); + continue; + } + + __mlxsw_sp_nexthop_neigh_update(nh, removing); + mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh->nhgi->nh_grp); + } +} + +static void mlxsw_sp_nexthop_rif_migrate(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_rif *old_rif, + struct mlxsw_sp_rif *new_rif) +{ + struct mlxsw_sp_nexthop *nh; + + list_splice_init(&old_rif->nexthop_list, &new_rif->nexthop_list); + list_for_each_entry(nh, &new_rif->nexthop_list, rif_list_node) + nh->rif = new_rif; + mlxsw_sp_nexthop_rif_update(mlxsw_sp, new_rif); +} + +static void mlxsw_sp_nexthop_rif_gone_sync(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp_nexthop *nh, *tmp; + + list_for_each_entry_safe(nh, tmp, &rif->nexthop_list, rif_list_node) { + mlxsw_sp_nexthop_type_fini(mlxsw_sp, nh); + mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh->nhgi->nh_grp); + } +} + +static int mlxsw_sp_adj_trap_entry_init(struct mlxsw_sp *mlxsw_sp) +{ + enum mlxsw_reg_ratr_trap_action trap_action; + char ratr_pl[MLXSW_REG_RATR_LEN]; + int err; + + err = mlxsw_sp_kvdl_alloc(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, 1, + &mlxsw_sp->router->adj_trap_index); + if (err) + return err; + + trap_action = MLXSW_REG_RATR_TRAP_ACTION_TRAP; + mlxsw_reg_ratr_pack(ratr_pl, MLXSW_REG_RATR_OP_WRITE_WRITE_ENTRY, true, + MLXSW_REG_RATR_TYPE_ETHERNET, + mlxsw_sp->router->adj_trap_index, + mlxsw_sp->router->lb_rif_index); + mlxsw_reg_ratr_trap_action_set(ratr_pl, trap_action); + mlxsw_reg_ratr_trap_id_set(ratr_pl, MLXSW_TRAP_ID_RTR_EGRESS0); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ratr), ratr_pl); + if (err) + goto err_ratr_write; + + return 0; + +err_ratr_write: + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, 1, + mlxsw_sp->router->adj_trap_index); + return err; +} + +static void mlxsw_sp_adj_trap_entry_fini(struct mlxsw_sp *mlxsw_sp) +{ + mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, 1, + mlxsw_sp->router->adj_trap_index); +} + +static int mlxsw_sp_nexthop_group_inc(struct mlxsw_sp *mlxsw_sp) +{ + int err; + + if (refcount_inc_not_zero(&mlxsw_sp->router->num_groups)) + return 0; + + err = mlxsw_sp_adj_trap_entry_init(mlxsw_sp); + if (err) + return err; + + refcount_set(&mlxsw_sp->router->num_groups, 1); + + return 0; +} + +static void mlxsw_sp_nexthop_group_dec(struct mlxsw_sp *mlxsw_sp) +{ + if (!refcount_dec_and_test(&mlxsw_sp->router->num_groups)) + return; + + mlxsw_sp_adj_trap_entry_fini(mlxsw_sp); +} + +static void +mlxsw_sp_nh_grp_activity_get(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_nexthop_group *nh_grp, + unsigned long *activity) +{ + char *ratrad_pl; + int i, err; + + ratrad_pl = kmalloc(MLXSW_REG_RATRAD_LEN, GFP_KERNEL); + if (!ratrad_pl) + return; + + mlxsw_reg_ratrad_pack(ratrad_pl, nh_grp->nhgi->adj_index, + nh_grp->nhgi->count); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ratrad), ratrad_pl); + if (err) + goto out; + + for (i = 0; i < nh_grp->nhgi->count; i++) { + if (!mlxsw_reg_ratrad_activity_vector_get(ratrad_pl, i)) + continue; + bitmap_set(activity, i, 1); + } + +out: + kfree(ratrad_pl); +} + +#define MLXSW_SP_NH_GRP_ACTIVITY_UPDATE_INTERVAL 1000 /* ms */ + +static void +mlxsw_sp_nh_grp_activity_update(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_nexthop_group *nh_grp) +{ + unsigned long *activity; + + activity = bitmap_zalloc(nh_grp->nhgi->count, GFP_KERNEL); + if (!activity) + return; + + mlxsw_sp_nh_grp_activity_get(mlxsw_sp, nh_grp, activity); + nexthop_res_grp_activity_update(mlxsw_sp_net(mlxsw_sp), nh_grp->obj.id, + nh_grp->nhgi->count, activity); + + bitmap_free(activity); +} + +static void +mlxsw_sp_nh_grp_activity_work_schedule(struct mlxsw_sp *mlxsw_sp) +{ + unsigned int interval = MLXSW_SP_NH_GRP_ACTIVITY_UPDATE_INTERVAL; + + mlxsw_core_schedule_dw(&mlxsw_sp->router->nh_grp_activity_dw, + msecs_to_jiffies(interval)); +} + +static void mlxsw_sp_nh_grp_activity_work(struct work_struct *work) +{ + struct mlxsw_sp_nexthop_group_info *nhgi; + struct mlxsw_sp_router *router; + bool reschedule = false; + + router = container_of(work, struct mlxsw_sp_router, + nh_grp_activity_dw.work); + + mutex_lock(&router->lock); + + list_for_each_entry(nhgi, &router->nh_res_grp_list, list) { + mlxsw_sp_nh_grp_activity_update(router->mlxsw_sp, nhgi->nh_grp); + reschedule = true; + } + + mutex_unlock(&router->lock); + + if (!reschedule) + return; + mlxsw_sp_nh_grp_activity_work_schedule(router->mlxsw_sp); +} + +static int +mlxsw_sp_nexthop_obj_single_validate(struct mlxsw_sp *mlxsw_sp, + const struct nh_notifier_single_info *nh, + struct netlink_ext_ack *extack) +{ + int err = -EINVAL; + + if (nh->is_fdb) + NL_SET_ERR_MSG_MOD(extack, "FDB nexthops are not supported"); + else if (nh->has_encap) + NL_SET_ERR_MSG_MOD(extack, "Encapsulating nexthops are not supported"); + else + err = 0; + + return err; +} + +static int +mlxsw_sp_nexthop_obj_group_entry_validate(struct mlxsw_sp *mlxsw_sp, + const struct nh_notifier_single_info *nh, + struct netlink_ext_ack *extack) +{ + int err; + + err = mlxsw_sp_nexthop_obj_single_validate(mlxsw_sp, nh, extack); + if (err) + return err; + + /* Device only nexthops with an IPIP device are programmed as + * encapsulating adjacency entries. + */ + if (!nh->gw_family && !nh->is_reject && + !mlxsw_sp_netdev_ipip_type(mlxsw_sp, nh->dev, NULL)) { + NL_SET_ERR_MSG_MOD(extack, "Nexthop group entry does not have a gateway"); + return -EINVAL; + } + + return 0; +} + +static int +mlxsw_sp_nexthop_obj_group_validate(struct mlxsw_sp *mlxsw_sp, + const struct nh_notifier_grp_info *nh_grp, + struct netlink_ext_ack *extack) +{ + int i; + + if (nh_grp->is_fdb) { + NL_SET_ERR_MSG_MOD(extack, "FDB nexthop groups are not supported"); + return -EINVAL; + } + + for (i = 0; i < nh_grp->num_nh; i++) { + const struct nh_notifier_single_info *nh; + int err; + + nh = &nh_grp->nh_entries[i].nh; + err = mlxsw_sp_nexthop_obj_group_entry_validate(mlxsw_sp, nh, + extack); + if (err) + return err; + } + + return 0; +} + +static int +mlxsw_sp_nexthop_obj_res_group_size_validate(struct mlxsw_sp *mlxsw_sp, + const struct nh_notifier_res_table_info *nh_res_table, + struct netlink_ext_ack *extack) +{ + unsigned int alloc_size; + bool valid_size = false; + int err, i; + + if (nh_res_table->num_nh_buckets < 32) { + NL_SET_ERR_MSG_MOD(extack, "Minimum number of buckets is 32"); + return -EINVAL; + } + + for (i = 0; i < mlxsw_sp->router->adj_grp_size_ranges_count; i++) { + const struct mlxsw_sp_adj_grp_size_range *size_range; + + size_range = &mlxsw_sp->router->adj_grp_size_ranges[i]; + + if (nh_res_table->num_nh_buckets >= size_range->start && + nh_res_table->num_nh_buckets <= size_range->end) { + valid_size = true; + break; + } + } + + if (!valid_size) { + NL_SET_ERR_MSG_MOD(extack, "Invalid number of buckets"); + return -EINVAL; + } + + err = mlxsw_sp_kvdl_alloc_count_query(mlxsw_sp, + MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, + nh_res_table->num_nh_buckets, + &alloc_size); + if (err || nh_res_table->num_nh_buckets != alloc_size) { + NL_SET_ERR_MSG_MOD(extack, "Number of buckets does not fit allocation size of any KVDL partition"); + return -EINVAL; + } + + return 0; +} + +static int +mlxsw_sp_nexthop_obj_res_group_validate(struct mlxsw_sp *mlxsw_sp, + const struct nh_notifier_res_table_info *nh_res_table, + struct netlink_ext_ack *extack) +{ + int err; + u16 i; + + err = mlxsw_sp_nexthop_obj_res_group_size_validate(mlxsw_sp, + nh_res_table, + extack); + if (err) + return err; + + for (i = 0; i < nh_res_table->num_nh_buckets; i++) { + const struct nh_notifier_single_info *nh; + int err; + + nh = &nh_res_table->nhs[i]; + err = mlxsw_sp_nexthop_obj_group_entry_validate(mlxsw_sp, nh, + extack); + if (err) + return err; + } + + return 0; +} + +static int mlxsw_sp_nexthop_obj_validate(struct mlxsw_sp *mlxsw_sp, + unsigned long event, + struct nh_notifier_info *info) +{ + struct nh_notifier_single_info *nh; + + if (event != NEXTHOP_EVENT_REPLACE && + event != NEXTHOP_EVENT_RES_TABLE_PRE_REPLACE && + event != NEXTHOP_EVENT_BUCKET_REPLACE) + return 0; + + switch (info->type) { + case NH_NOTIFIER_INFO_TYPE_SINGLE: + return mlxsw_sp_nexthop_obj_single_validate(mlxsw_sp, info->nh, + info->extack); + case NH_NOTIFIER_INFO_TYPE_GRP: + return mlxsw_sp_nexthop_obj_group_validate(mlxsw_sp, + info->nh_grp, + info->extack); + case NH_NOTIFIER_INFO_TYPE_RES_TABLE: + return mlxsw_sp_nexthop_obj_res_group_validate(mlxsw_sp, + info->nh_res_table, + info->extack); + case NH_NOTIFIER_INFO_TYPE_RES_BUCKET: + nh = &info->nh_res_bucket->new_nh; + return mlxsw_sp_nexthop_obj_group_entry_validate(mlxsw_sp, nh, + info->extack); + default: + NL_SET_ERR_MSG_MOD(info->extack, "Unsupported nexthop type"); + return -EOPNOTSUPP; + } +} + +static bool mlxsw_sp_nexthop_obj_is_gateway(struct mlxsw_sp *mlxsw_sp, + const struct nh_notifier_info *info) +{ + const struct net_device *dev; + + switch (info->type) { + case NH_NOTIFIER_INFO_TYPE_SINGLE: + dev = info->nh->dev; + return info->nh->gw_family || info->nh->is_reject || + mlxsw_sp_netdev_ipip_type(mlxsw_sp, dev, NULL); + case NH_NOTIFIER_INFO_TYPE_GRP: + case NH_NOTIFIER_INFO_TYPE_RES_TABLE: + /* Already validated earlier. */ + return true; + default: + return false; + } +} + +static void mlxsw_sp_nexthop_obj_blackhole_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + u16 lb_rif_index = mlxsw_sp->router->lb_rif_index; + + nh->action = MLXSW_SP_NEXTHOP_ACTION_DISCARD; + nh->should_offload = 1; + /* While nexthops that discard packets do not forward packets + * via an egress RIF, they still need to be programmed using a + * valid RIF, so use the loopback RIF created during init. + */ + nh->rif = mlxsw_sp->router->rifs[lb_rif_index]; +} + +static void mlxsw_sp_nexthop_obj_blackhole_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + nh->rif = NULL; + nh->should_offload = 0; +} + +static int +mlxsw_sp_nexthop_obj_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp, + struct mlxsw_sp_nexthop *nh, + struct nh_notifier_single_info *nh_obj, int weight) +{ + struct net_device *dev = nh_obj->dev; + int err; + + nh->nhgi = nh_grp->nhgi; + nh->nh_weight = weight; + + switch (nh_obj->gw_family) { + case AF_INET: + memcpy(&nh->gw_addr, &nh_obj->ipv4, sizeof(nh_obj->ipv4)); + nh->neigh_tbl = &arp_tbl; + break; + case AF_INET6: + memcpy(&nh->gw_addr, &nh_obj->ipv6, sizeof(nh_obj->ipv6)); +#if IS_ENABLED(CONFIG_IPV6) + nh->neigh_tbl = &nd_tbl; +#endif + break; + } + + mlxsw_sp_nexthop_counter_alloc(mlxsw_sp, nh); + list_add_tail(&nh->router_list_node, &mlxsw_sp->router->nexthop_list); + nh->ifindex = dev->ifindex; + + err = mlxsw_sp_nexthop_type_init(mlxsw_sp, nh, dev); + if (err) + goto err_type_init; + + if (nh_obj->is_reject) + mlxsw_sp_nexthop_obj_blackhole_init(mlxsw_sp, nh); + + /* In a resilient nexthop group, all the nexthops must be written to + * the adjacency table. Even if they do not have a valid neighbour or + * RIF. + */ + if (nh_grp->nhgi->is_resilient && !nh->should_offload) { + nh->action = MLXSW_SP_NEXTHOP_ACTION_TRAP; + nh->should_offload = 1; + } + + return 0; + +err_type_init: + list_del(&nh->router_list_node); + mlxsw_sp_nexthop_counter_free(mlxsw_sp, nh); + return err; +} + +static void mlxsw_sp_nexthop_obj_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + if (nh->action == MLXSW_SP_NEXTHOP_ACTION_DISCARD) + mlxsw_sp_nexthop_obj_blackhole_fini(mlxsw_sp, nh); + mlxsw_sp_nexthop_type_fini(mlxsw_sp, nh); + list_del(&nh->router_list_node); + mlxsw_sp_nexthop_counter_free(mlxsw_sp, nh); + nh->should_offload = 0; +} + +static int +mlxsw_sp_nexthop_obj_group_info_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp, + struct nh_notifier_info *info) +{ + struct mlxsw_sp_nexthop_group_info *nhgi; + struct mlxsw_sp_nexthop *nh; + bool is_resilient = false; + unsigned int nhs; + int err, i; + + switch (info->type) { + case NH_NOTIFIER_INFO_TYPE_SINGLE: + nhs = 1; + break; + case NH_NOTIFIER_INFO_TYPE_GRP: + nhs = info->nh_grp->num_nh; + break; + case NH_NOTIFIER_INFO_TYPE_RES_TABLE: + nhs = info->nh_res_table->num_nh_buckets; + is_resilient = true; + break; + default: + return -EINVAL; + } + + nhgi = kzalloc(struct_size(nhgi, nexthops, nhs), GFP_KERNEL); + if (!nhgi) + return -ENOMEM; + nh_grp->nhgi = nhgi; + nhgi->nh_grp = nh_grp; + nhgi->gateway = mlxsw_sp_nexthop_obj_is_gateway(mlxsw_sp, info); + nhgi->is_resilient = is_resilient; + nhgi->count = nhs; + for (i = 0; i < nhgi->count; i++) { + struct nh_notifier_single_info *nh_obj; + int weight; + + nh = &nhgi->nexthops[i]; + switch (info->type) { + case NH_NOTIFIER_INFO_TYPE_SINGLE: + nh_obj = info->nh; + weight = 1; + break; + case NH_NOTIFIER_INFO_TYPE_GRP: + nh_obj = &info->nh_grp->nh_entries[i].nh; + weight = info->nh_grp->nh_entries[i].weight; + break; + case NH_NOTIFIER_INFO_TYPE_RES_TABLE: + nh_obj = &info->nh_res_table->nhs[i]; + weight = 1; + break; + default: + err = -EINVAL; + goto err_nexthop_obj_init; + } + err = mlxsw_sp_nexthop_obj_init(mlxsw_sp, nh_grp, nh, nh_obj, + weight); + if (err) + goto err_nexthop_obj_init; + } + err = mlxsw_sp_nexthop_group_inc(mlxsw_sp); + if (err) + goto err_group_inc; + err = mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh_grp); + if (err) { + NL_SET_ERR_MSG_MOD(info->extack, "Failed to write adjacency entries to the device"); + goto err_group_refresh; + } + + /* Add resilient nexthop groups to a list so that the activity of their + * nexthop buckets will be periodically queried and cleared. + */ + if (nhgi->is_resilient) { + if (list_empty(&mlxsw_sp->router->nh_res_grp_list)) + mlxsw_sp_nh_grp_activity_work_schedule(mlxsw_sp); + list_add(&nhgi->list, &mlxsw_sp->router->nh_res_grp_list); + } + + return 0; + +err_group_refresh: + mlxsw_sp_nexthop_group_dec(mlxsw_sp); +err_group_inc: + i = nhgi->count; +err_nexthop_obj_init: + for (i--; i >= 0; i--) { + nh = &nhgi->nexthops[i]; + mlxsw_sp_nexthop_obj_fini(mlxsw_sp, nh); + } + kfree(nhgi); + return err; +} + +static void +mlxsw_sp_nexthop_obj_group_info_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + struct mlxsw_sp_nexthop_group_info *nhgi = nh_grp->nhgi; + struct mlxsw_sp_router *router = mlxsw_sp->router; + int i; + + if (nhgi->is_resilient) { + list_del(&nhgi->list); + if (list_empty(&mlxsw_sp->router->nh_res_grp_list)) + cancel_delayed_work(&router->nh_grp_activity_dw); + } + + mlxsw_sp_nexthop_group_dec(mlxsw_sp); + for (i = nhgi->count - 1; i >= 0; i--) { + struct mlxsw_sp_nexthop *nh = &nhgi->nexthops[i]; + + mlxsw_sp_nexthop_obj_fini(mlxsw_sp, nh); + } + mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh_grp); + WARN_ON_ONCE(nhgi->adj_index_valid); + kfree(nhgi); +} + +static struct mlxsw_sp_nexthop_group * +mlxsw_sp_nexthop_obj_group_create(struct mlxsw_sp *mlxsw_sp, + struct nh_notifier_info *info) +{ + struct mlxsw_sp_nexthop_group *nh_grp; + int err; + + nh_grp = kzalloc(sizeof(*nh_grp), GFP_KERNEL); + if (!nh_grp) + return ERR_PTR(-ENOMEM); + INIT_LIST_HEAD(&nh_grp->vr_list); + err = rhashtable_init(&nh_grp->vr_ht, + &mlxsw_sp_nexthop_group_vr_ht_params); + if (err) + goto err_nexthop_group_vr_ht_init; + INIT_LIST_HEAD(&nh_grp->fib_list); + nh_grp->type = MLXSW_SP_NEXTHOP_GROUP_TYPE_OBJ; + nh_grp->obj.id = info->id; + + err = mlxsw_sp_nexthop_obj_group_info_init(mlxsw_sp, nh_grp, info); + if (err) + goto err_nexthop_group_info_init; + + nh_grp->can_destroy = false; + + return nh_grp; + +err_nexthop_group_info_init: + rhashtable_destroy(&nh_grp->vr_ht); +err_nexthop_group_vr_ht_init: + kfree(nh_grp); + return ERR_PTR(err); +} + +static void +mlxsw_sp_nexthop_obj_group_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + if (!nh_grp->can_destroy) + return; + mlxsw_sp_nexthop_obj_group_info_fini(mlxsw_sp, nh_grp); + WARN_ON_ONCE(!list_empty(&nh_grp->fib_list)); + WARN_ON_ONCE(!list_empty(&nh_grp->vr_list)); + rhashtable_destroy(&nh_grp->vr_ht); + kfree(nh_grp); +} + +static struct mlxsw_sp_nexthop_group * +mlxsw_sp_nexthop_obj_group_lookup(struct mlxsw_sp *mlxsw_sp, u32 id) +{ + struct mlxsw_sp_nexthop_group_cmp_arg cmp_arg; + + cmp_arg.type = MLXSW_SP_NEXTHOP_GROUP_TYPE_OBJ; + cmp_arg.id = id; + return rhashtable_lookup_fast(&mlxsw_sp->router->nexthop_group_ht, + &cmp_arg, + mlxsw_sp_nexthop_group_ht_params); +} + +static int mlxsw_sp_nexthop_obj_group_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + return mlxsw_sp_nexthop_group_insert(mlxsw_sp, nh_grp); +} + +static int +mlxsw_sp_nexthop_obj_group_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp, + struct mlxsw_sp_nexthop_group *old_nh_grp, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_nexthop_group_info *old_nhgi = old_nh_grp->nhgi; + struct mlxsw_sp_nexthop_group_info *new_nhgi = nh_grp->nhgi; + int err; + + old_nh_grp->nhgi = new_nhgi; + new_nhgi->nh_grp = old_nh_grp; + nh_grp->nhgi = old_nhgi; + old_nhgi->nh_grp = nh_grp; + + if (old_nhgi->adj_index_valid && new_nhgi->adj_index_valid) { + /* Both the old adjacency index and the new one are valid. + * Routes are currently using the old one. Tell the device to + * replace the old adjacency index with the new one. + */ + err = mlxsw_sp_adj_index_mass_update(mlxsw_sp, old_nh_grp, + old_nhgi->adj_index, + old_nhgi->ecmp_size); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to replace old adjacency index with new one"); + goto err_out; + } + } else if (old_nhgi->adj_index_valid && !new_nhgi->adj_index_valid) { + /* The old adjacency index is valid, while the new one is not. + * Iterate over all the routes using the group and change them + * to trap packets to the CPU. + */ + err = mlxsw_sp_nexthop_fib_entries_update(mlxsw_sp, old_nh_grp); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to update routes to trap packets"); + goto err_out; + } + } else if (!old_nhgi->adj_index_valid && new_nhgi->adj_index_valid) { + /* The old adjacency index is invalid, while the new one is. + * Iterate over all the routes using the group and change them + * to forward packets using the new valid index. + */ + err = mlxsw_sp_nexthop_fib_entries_update(mlxsw_sp, old_nh_grp); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to update routes to forward packets"); + goto err_out; + } + } + + /* Make sure the flags are set / cleared based on the new nexthop group + * information. + */ + mlxsw_sp_nexthop_obj_group_offload_refresh(mlxsw_sp, old_nh_grp); + + /* At this point 'nh_grp' is just a shell that is not used by anyone + * and its nexthop group info is the old info that was just replaced + * with the new one. Remove it. + */ + nh_grp->can_destroy = true; + mlxsw_sp_nexthop_obj_group_destroy(mlxsw_sp, nh_grp); + + return 0; + +err_out: + old_nhgi->nh_grp = old_nh_grp; + nh_grp->nhgi = new_nhgi; + new_nhgi->nh_grp = nh_grp; + old_nh_grp->nhgi = old_nhgi; + return err; +} + +static int mlxsw_sp_nexthop_obj_new(struct mlxsw_sp *mlxsw_sp, + struct nh_notifier_info *info) +{ + struct mlxsw_sp_nexthop_group *nh_grp, *old_nh_grp; + struct netlink_ext_ack *extack = info->extack; + int err; + + nh_grp = mlxsw_sp_nexthop_obj_group_create(mlxsw_sp, info); + if (IS_ERR(nh_grp)) + return PTR_ERR(nh_grp); + + old_nh_grp = mlxsw_sp_nexthop_obj_group_lookup(mlxsw_sp, info->id); + if (!old_nh_grp) + err = mlxsw_sp_nexthop_obj_group_add(mlxsw_sp, nh_grp); + else + err = mlxsw_sp_nexthop_obj_group_replace(mlxsw_sp, nh_grp, + old_nh_grp, extack); + + if (err) { + nh_grp->can_destroy = true; + mlxsw_sp_nexthop_obj_group_destroy(mlxsw_sp, nh_grp); + } + + return err; +} + +static void mlxsw_sp_nexthop_obj_del(struct mlxsw_sp *mlxsw_sp, + struct nh_notifier_info *info) +{ + struct mlxsw_sp_nexthop_group *nh_grp; + + nh_grp = mlxsw_sp_nexthop_obj_group_lookup(mlxsw_sp, info->id); + if (!nh_grp) + return; + + nh_grp->can_destroy = true; + mlxsw_sp_nexthop_group_remove(mlxsw_sp, nh_grp); + + /* If the group still has routes using it, then defer the delete + * operation until the last route using it is deleted. + */ + if (!list_empty(&nh_grp->fib_list)) + return; + mlxsw_sp_nexthop_obj_group_destroy(mlxsw_sp, nh_grp); +} + +static int mlxsw_sp_nexthop_obj_bucket_query(struct mlxsw_sp *mlxsw_sp, + u32 adj_index, char *ratr_pl) +{ + MLXSW_REG_ZERO(ratr, ratr_pl); + mlxsw_reg_ratr_op_set(ratr_pl, MLXSW_REG_RATR_OP_QUERY_READ); + mlxsw_reg_ratr_adjacency_index_low_set(ratr_pl, adj_index); + mlxsw_reg_ratr_adjacency_index_high_set(ratr_pl, adj_index >> 16); + + return mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ratr), ratr_pl); +} + +static int mlxsw_sp_nexthop_obj_bucket_compare(char *ratr_pl, char *ratr_pl_new) +{ + /* Clear the opcode and activity on both the old and new payload as + * they are irrelevant for the comparison. + */ + mlxsw_reg_ratr_op_set(ratr_pl, MLXSW_REG_RATR_OP_QUERY_READ); + mlxsw_reg_ratr_a_set(ratr_pl, 0); + mlxsw_reg_ratr_op_set(ratr_pl_new, MLXSW_REG_RATR_OP_QUERY_READ); + mlxsw_reg_ratr_a_set(ratr_pl_new, 0); + + /* If the contents of the adjacency entry are consistent with the + * replacement request, then replacement was successful. + */ + if (!memcmp(ratr_pl, ratr_pl_new, MLXSW_REG_RATR_LEN)) + return 0; + + return -EINVAL; +} + +static int +mlxsw_sp_nexthop_obj_bucket_adj_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh, + struct nh_notifier_info *info) +{ + u16 bucket_index = info->nh_res_bucket->bucket_index; + struct netlink_ext_ack *extack = info->extack; + bool force = info->nh_res_bucket->force; + char ratr_pl_new[MLXSW_REG_RATR_LEN]; + char ratr_pl[MLXSW_REG_RATR_LEN]; + u32 adj_index; + int err; + + /* No point in trying an atomic replacement if the idle timer interval + * is smaller than the interval in which we query and clear activity. + */ + if (!force && info->nh_res_bucket->idle_timer_ms < + MLXSW_SP_NH_GRP_ACTIVITY_UPDATE_INTERVAL) + force = true; + + adj_index = nh->nhgi->adj_index + bucket_index; + err = mlxsw_sp_nexthop_update(mlxsw_sp, adj_index, nh, force, ratr_pl); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to overwrite nexthop bucket"); + return err; + } + + if (!force) { + err = mlxsw_sp_nexthop_obj_bucket_query(mlxsw_sp, adj_index, + ratr_pl_new); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to query nexthop bucket state after replacement. State might be inconsistent"); + return err; + } + + err = mlxsw_sp_nexthop_obj_bucket_compare(ratr_pl, ratr_pl_new); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Nexthop bucket was not replaced because it was active during replacement"); + return err; + } + } + + nh->update = 0; + nh->offloaded = 1; + mlxsw_sp_nexthop_bucket_offload_refresh(mlxsw_sp, nh, bucket_index); + + return 0; +} + +static int mlxsw_sp_nexthop_obj_bucket_replace(struct mlxsw_sp *mlxsw_sp, + struct nh_notifier_info *info) +{ + u16 bucket_index = info->nh_res_bucket->bucket_index; + struct netlink_ext_ack *extack = info->extack; + struct mlxsw_sp_nexthop_group_info *nhgi; + struct nh_notifier_single_info *nh_obj; + struct mlxsw_sp_nexthop_group *nh_grp; + struct mlxsw_sp_nexthop *nh; + int err; + + nh_grp = mlxsw_sp_nexthop_obj_group_lookup(mlxsw_sp, info->id); + if (!nh_grp) { + NL_SET_ERR_MSG_MOD(extack, "Nexthop group was not found"); + return -EINVAL; + } + + nhgi = nh_grp->nhgi; + + if (bucket_index >= nhgi->count) { + NL_SET_ERR_MSG_MOD(extack, "Nexthop bucket index out of range"); + return -EINVAL; + } + + nh = &nhgi->nexthops[bucket_index]; + mlxsw_sp_nexthop_obj_fini(mlxsw_sp, nh); + + nh_obj = &info->nh_res_bucket->new_nh; + err = mlxsw_sp_nexthop_obj_init(mlxsw_sp, nh_grp, nh, nh_obj, 1); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to initialize nexthop object for nexthop bucket replacement"); + goto err_nexthop_obj_init; + } + + err = mlxsw_sp_nexthop_obj_bucket_adj_update(mlxsw_sp, nh, info); + if (err) + goto err_nexthop_obj_bucket_adj_update; + + return 0; + +err_nexthop_obj_bucket_adj_update: + mlxsw_sp_nexthop_obj_fini(mlxsw_sp, nh); +err_nexthop_obj_init: + nh_obj = &info->nh_res_bucket->old_nh; + mlxsw_sp_nexthop_obj_init(mlxsw_sp, nh_grp, nh, nh_obj, 1); + /* The old adjacency entry was not overwritten */ + nh->update = 0; + nh->offloaded = 1; + return err; +} + +static int mlxsw_sp_nexthop_obj_event(struct notifier_block *nb, + unsigned long event, void *ptr) +{ + struct nh_notifier_info *info = ptr; + struct mlxsw_sp_router *router; + int err = 0; + + router = container_of(nb, struct mlxsw_sp_router, nexthop_nb); + err = mlxsw_sp_nexthop_obj_validate(router->mlxsw_sp, event, info); + if (err) + goto out; + + mutex_lock(&router->lock); + + switch (event) { + case NEXTHOP_EVENT_REPLACE: + err = mlxsw_sp_nexthop_obj_new(router->mlxsw_sp, info); + break; + case NEXTHOP_EVENT_DEL: + mlxsw_sp_nexthop_obj_del(router->mlxsw_sp, info); + break; + case NEXTHOP_EVENT_BUCKET_REPLACE: + err = mlxsw_sp_nexthop_obj_bucket_replace(router->mlxsw_sp, + info); + break; + default: + break; + } + + mutex_unlock(&router->lock); + +out: + return notifier_from_errno(err); +} + +static bool mlxsw_sp_fi_is_gateway(const struct mlxsw_sp *mlxsw_sp, + struct fib_info *fi) +{ + const struct fib_nh *nh = fib_info_nh(fi, 0); + + return nh->fib_nh_gw_family || + mlxsw_sp_nexthop4_ipip_type(mlxsw_sp, nh, NULL); +} + +static int +mlxsw_sp_nexthop4_group_info_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + unsigned int nhs = fib_info_num_path(nh_grp->ipv4.fi); + struct mlxsw_sp_nexthop_group_info *nhgi; + struct mlxsw_sp_nexthop *nh; + int err, i; + + nhgi = kzalloc(struct_size(nhgi, nexthops, nhs), GFP_KERNEL); + if (!nhgi) + return -ENOMEM; + nh_grp->nhgi = nhgi; + nhgi->nh_grp = nh_grp; + nhgi->gateway = mlxsw_sp_fi_is_gateway(mlxsw_sp, nh_grp->ipv4.fi); + nhgi->count = nhs; + for (i = 0; i < nhgi->count; i++) { + struct fib_nh *fib_nh; + + nh = &nhgi->nexthops[i]; + fib_nh = fib_info_nh(nh_grp->ipv4.fi, i); + err = mlxsw_sp_nexthop4_init(mlxsw_sp, nh_grp, nh, fib_nh); + if (err) + goto err_nexthop4_init; + } + err = mlxsw_sp_nexthop_group_inc(mlxsw_sp); + if (err) + goto err_group_inc; + err = mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh_grp); + if (err) + goto err_group_refresh; + + return 0; + +err_group_refresh: + mlxsw_sp_nexthop_group_dec(mlxsw_sp); +err_group_inc: + i = nhgi->count; +err_nexthop4_init: + for (i--; i >= 0; i--) { + nh = &nhgi->nexthops[i]; + mlxsw_sp_nexthop4_fini(mlxsw_sp, nh); + } + kfree(nhgi); + return err; +} + +static void +mlxsw_sp_nexthop4_group_info_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + struct mlxsw_sp_nexthop_group_info *nhgi = nh_grp->nhgi; + int i; + + mlxsw_sp_nexthop_group_dec(mlxsw_sp); + for (i = nhgi->count - 1; i >= 0; i--) { + struct mlxsw_sp_nexthop *nh = &nhgi->nexthops[i]; + + mlxsw_sp_nexthop4_fini(mlxsw_sp, nh); + } + mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh_grp); + WARN_ON_ONCE(nhgi->adj_index_valid); + kfree(nhgi); +} + +static struct mlxsw_sp_nexthop_group * +mlxsw_sp_nexthop4_group_create(struct mlxsw_sp *mlxsw_sp, struct fib_info *fi) +{ + struct mlxsw_sp_nexthop_group *nh_grp; + int err; + + nh_grp = kzalloc(sizeof(*nh_grp), GFP_KERNEL); + if (!nh_grp) + return ERR_PTR(-ENOMEM); + INIT_LIST_HEAD(&nh_grp->vr_list); + err = rhashtable_init(&nh_grp->vr_ht, + &mlxsw_sp_nexthop_group_vr_ht_params); + if (err) + goto err_nexthop_group_vr_ht_init; + INIT_LIST_HEAD(&nh_grp->fib_list); + nh_grp->type = MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV4; + nh_grp->ipv4.fi = fi; + fib_info_hold(fi); + + err = mlxsw_sp_nexthop4_group_info_init(mlxsw_sp, nh_grp); + if (err) + goto err_nexthop_group_info_init; + + err = mlxsw_sp_nexthop_group_insert(mlxsw_sp, nh_grp); + if (err) + goto err_nexthop_group_insert; + + nh_grp->can_destroy = true; + + return nh_grp; + +err_nexthop_group_insert: + mlxsw_sp_nexthop4_group_info_fini(mlxsw_sp, nh_grp); +err_nexthop_group_info_init: + fib_info_put(fi); + rhashtable_destroy(&nh_grp->vr_ht); +err_nexthop_group_vr_ht_init: + kfree(nh_grp); + return ERR_PTR(err); +} + +static void +mlxsw_sp_nexthop4_group_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + if (!nh_grp->can_destroy) + return; + mlxsw_sp_nexthop_group_remove(mlxsw_sp, nh_grp); + mlxsw_sp_nexthop4_group_info_fini(mlxsw_sp, nh_grp); + fib_info_put(nh_grp->ipv4.fi); + WARN_ON_ONCE(!list_empty(&nh_grp->vr_list)); + rhashtable_destroy(&nh_grp->vr_ht); + kfree(nh_grp); +} + +static int mlxsw_sp_nexthop4_group_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + struct fib_info *fi) +{ + struct mlxsw_sp_nexthop_group *nh_grp; + + if (fi->nh) { + nh_grp = mlxsw_sp_nexthop_obj_group_lookup(mlxsw_sp, + fi->nh->id); + if (WARN_ON_ONCE(!nh_grp)) + return -EINVAL; + goto out; + } + + nh_grp = mlxsw_sp_nexthop4_group_lookup(mlxsw_sp, fi); + if (!nh_grp) { + nh_grp = mlxsw_sp_nexthop4_group_create(mlxsw_sp, fi); + if (IS_ERR(nh_grp)) + return PTR_ERR(nh_grp); + } +out: + list_add_tail(&fib_entry->nexthop_group_node, &nh_grp->fib_list); + fib_entry->nh_group = nh_grp; + return 0; +} + +static void mlxsw_sp_nexthop4_group_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; + + list_del(&fib_entry->nexthop_group_node); + if (!list_empty(&nh_grp->fib_list)) + return; + + if (nh_grp->type == MLXSW_SP_NEXTHOP_GROUP_TYPE_OBJ) { + mlxsw_sp_nexthop_obj_group_destroy(mlxsw_sp, nh_grp); + return; + } + + mlxsw_sp_nexthop4_group_destroy(mlxsw_sp, nh_grp); +} + +static bool +mlxsw_sp_fib4_entry_should_offload(const struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_fib4_entry *fib4_entry; + + fib4_entry = container_of(fib_entry, struct mlxsw_sp_fib4_entry, + common); + return !fib4_entry->dscp; +} + +static bool +mlxsw_sp_fib_entry_should_offload(const struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_nexthop_group *nh_group = fib_entry->nh_group; + + switch (fib_entry->fib_node->fib->proto) { + case MLXSW_SP_L3_PROTO_IPV4: + if (!mlxsw_sp_fib4_entry_should_offload(fib_entry)) + return false; + break; + case MLXSW_SP_L3_PROTO_IPV6: + break; + } + + switch (fib_entry->type) { + case MLXSW_SP_FIB_ENTRY_TYPE_REMOTE: + return !!nh_group->nhgi->adj_index_valid; + case MLXSW_SP_FIB_ENTRY_TYPE_LOCAL: + return !!nh_group->nhgi->nh_rif; + case MLXSW_SP_FIB_ENTRY_TYPE_BLACKHOLE: + case MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP: + case MLXSW_SP_FIB_ENTRY_TYPE_NVE_DECAP: + return true; + default: + return false; + } +} + +static struct mlxsw_sp_nexthop * +mlxsw_sp_rt6_nexthop(struct mlxsw_sp_nexthop_group *nh_grp, + const struct mlxsw_sp_rt6 *mlxsw_sp_rt6) +{ + int i; + + for (i = 0; i < nh_grp->nhgi->count; i++) { + struct mlxsw_sp_nexthop *nh = &nh_grp->nhgi->nexthops[i]; + struct fib6_info *rt = mlxsw_sp_rt6->rt; + + if (nh->rif && nh->rif->dev == rt->fib6_nh->fib_nh_dev && + ipv6_addr_equal((const struct in6_addr *) &nh->gw_addr, + &rt->fib6_nh->fib_nh_gw6)) + return nh; + } + + return NULL; +} + +static void +mlxsw_sp_fib4_offload_failed_flag_set(struct mlxsw_sp *mlxsw_sp, + struct fib_entry_notifier_info *fen_info) +{ + u32 *p_dst = (u32 *) &fen_info->dst; + struct fib_rt_info fri; + + fri.fi = fen_info->fi; + fri.tb_id = fen_info->tb_id; + fri.dst = cpu_to_be32(*p_dst); + fri.dst_len = fen_info->dst_len; + fri.dscp = fen_info->dscp; + fri.type = fen_info->type; + fri.offload = false; + fri.trap = false; + fri.offload_failed = true; + fib_alias_hw_flags_set(mlxsw_sp_net(mlxsw_sp), &fri); +} + +static void +mlxsw_sp_fib4_entry_hw_flags_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + u32 *p_dst = (u32 *) fib_entry->fib_node->key.addr; + int dst_len = fib_entry->fib_node->key.prefix_len; + struct mlxsw_sp_fib4_entry *fib4_entry; + struct fib_rt_info fri; + bool should_offload; + + should_offload = mlxsw_sp_fib_entry_should_offload(fib_entry); + fib4_entry = container_of(fib_entry, struct mlxsw_sp_fib4_entry, + common); + fri.fi = fib4_entry->fi; + fri.tb_id = fib4_entry->tb_id; + fri.dst = cpu_to_be32(*p_dst); + fri.dst_len = dst_len; + fri.dscp = fib4_entry->dscp; + fri.type = fib4_entry->type; + fri.offload = should_offload; + fri.trap = !should_offload; + fri.offload_failed = false; + fib_alias_hw_flags_set(mlxsw_sp_net(mlxsw_sp), &fri); +} + +static void +mlxsw_sp_fib4_entry_hw_flags_clear(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + u32 *p_dst = (u32 *) fib_entry->fib_node->key.addr; + int dst_len = fib_entry->fib_node->key.prefix_len; + struct mlxsw_sp_fib4_entry *fib4_entry; + struct fib_rt_info fri; + + fib4_entry = container_of(fib_entry, struct mlxsw_sp_fib4_entry, + common); + fri.fi = fib4_entry->fi; + fri.tb_id = fib4_entry->tb_id; + fri.dst = cpu_to_be32(*p_dst); + fri.dst_len = dst_len; + fri.dscp = fib4_entry->dscp; + fri.type = fib4_entry->type; + fri.offload = false; + fri.trap = false; + fri.offload_failed = false; + fib_alias_hw_flags_set(mlxsw_sp_net(mlxsw_sp), &fri); +} + +#if IS_ENABLED(CONFIG_IPV6) +static void +mlxsw_sp_fib6_offload_failed_flag_set(struct mlxsw_sp *mlxsw_sp, + struct fib6_info **rt_arr, + unsigned int nrt6) +{ + int i; + + /* In IPv6 a multipath route is represented using multiple routes, so + * we need to set the flags on all of them. + */ + for (i = 0; i < nrt6; i++) + fib6_info_hw_flags_set(mlxsw_sp_net(mlxsw_sp), rt_arr[i], + false, false, true); +} +#else +static void +mlxsw_sp_fib6_offload_failed_flag_set(struct mlxsw_sp *mlxsw_sp, + struct fib6_info **rt_arr, + unsigned int nrt6) +{ +} +#endif + +#if IS_ENABLED(CONFIG_IPV6) +static void +mlxsw_sp_fib6_entry_hw_flags_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + bool should_offload; + + should_offload = mlxsw_sp_fib_entry_should_offload(fib_entry); + + /* In IPv6 a multipath route is represented using multiple routes, so + * we need to set the flags on all of them. + */ + fib6_entry = container_of(fib_entry, struct mlxsw_sp_fib6_entry, + common); + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) + fib6_info_hw_flags_set(mlxsw_sp_net(mlxsw_sp), mlxsw_sp_rt6->rt, + should_offload, !should_offload, false); +} +#else +static void +mlxsw_sp_fib6_entry_hw_flags_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ +} +#endif + +#if IS_ENABLED(CONFIG_IPV6) +static void +mlxsw_sp_fib6_entry_hw_flags_clear(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + fib6_entry = container_of(fib_entry, struct mlxsw_sp_fib6_entry, + common); + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) + fib6_info_hw_flags_set(mlxsw_sp_net(mlxsw_sp), mlxsw_sp_rt6->rt, + false, false, false); +} +#else +static void +mlxsw_sp_fib6_entry_hw_flags_clear(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ +} +#endif + +static void +mlxsw_sp_fib_entry_hw_flags_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + switch (fib_entry->fib_node->fib->proto) { + case MLXSW_SP_L3_PROTO_IPV4: + mlxsw_sp_fib4_entry_hw_flags_set(mlxsw_sp, fib_entry); + break; + case MLXSW_SP_L3_PROTO_IPV6: + mlxsw_sp_fib6_entry_hw_flags_set(mlxsw_sp, fib_entry); + break; + } +} + +static void +mlxsw_sp_fib_entry_hw_flags_clear(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + switch (fib_entry->fib_node->fib->proto) { + case MLXSW_SP_L3_PROTO_IPV4: + mlxsw_sp_fib4_entry_hw_flags_clear(mlxsw_sp, fib_entry); + break; + case MLXSW_SP_L3_PROTO_IPV6: + mlxsw_sp_fib6_entry_hw_flags_clear(mlxsw_sp, fib_entry); + break; + } +} + +static void +mlxsw_sp_fib_entry_hw_flags_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) +{ + switch (op) { + case MLXSW_REG_RALUE_OP_WRITE_WRITE: + mlxsw_sp_fib_entry_hw_flags_set(mlxsw_sp, fib_entry); + break; + case MLXSW_REG_RALUE_OP_WRITE_DELETE: + mlxsw_sp_fib_entry_hw_flags_clear(mlxsw_sp, fib_entry); + break; + default: + break; + } +} + +static void +mlxsw_sp_fib_entry_ralue_pack(char *ralue_pl, + const struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) +{ + struct mlxsw_sp_fib *fib = fib_entry->fib_node->fib; + enum mlxsw_reg_ralxx_protocol proto; + u32 *p_dip; + + proto = (enum mlxsw_reg_ralxx_protocol) fib->proto; + + switch (fib->proto) { + case MLXSW_SP_L3_PROTO_IPV4: + p_dip = (u32 *) fib_entry->fib_node->key.addr; + mlxsw_reg_ralue_pack4(ralue_pl, proto, op, fib->vr->id, + fib_entry->fib_node->key.prefix_len, + *p_dip); + break; + case MLXSW_SP_L3_PROTO_IPV6: + mlxsw_reg_ralue_pack6(ralue_pl, proto, op, fib->vr->id, + fib_entry->fib_node->key.prefix_len, + fib_entry->fib_node->key.addr); + break; + } +} + +static int mlxsw_sp_fib_entry_op_remote(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) +{ + struct mlxsw_sp_nexthop_group *nh_group = fib_entry->nh_group; + struct mlxsw_sp_nexthop_group_info *nhgi = nh_group->nhgi; + char ralue_pl[MLXSW_REG_RALUE_LEN]; + enum mlxsw_reg_ralue_trap_action trap_action; + u16 trap_id = 0; + u32 adjacency_index = 0; + u16 ecmp_size = 0; + + /* In case the nexthop group adjacency index is valid, use it + * with provided ECMP size. Otherwise, setup trap and pass + * traffic to kernel. + */ + if (mlxsw_sp_fib_entry_should_offload(fib_entry)) { + trap_action = MLXSW_REG_RALUE_TRAP_ACTION_NOP; + adjacency_index = nhgi->adj_index; + ecmp_size = nhgi->ecmp_size; + } else if (!nhgi->adj_index_valid && nhgi->count && nhgi->nh_rif) { + trap_action = MLXSW_REG_RALUE_TRAP_ACTION_NOP; + adjacency_index = mlxsw_sp->router->adj_trap_index; + ecmp_size = 1; + } else { + trap_action = MLXSW_REG_RALUE_TRAP_ACTION_TRAP; + trap_id = MLXSW_TRAP_ID_RTR_INGRESS0; + } + + mlxsw_sp_fib_entry_ralue_pack(ralue_pl, fib_entry, op); + mlxsw_reg_ralue_act_remote_pack(ralue_pl, trap_action, trap_id, + adjacency_index, ecmp_size); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralue), ralue_pl); +} + +static int mlxsw_sp_fib_entry_op_local(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) +{ + struct mlxsw_sp_rif *rif = fib_entry->nh_group->nhgi->nh_rif; + enum mlxsw_reg_ralue_trap_action trap_action; + char ralue_pl[MLXSW_REG_RALUE_LEN]; + u16 trap_id = 0; + u16 rif_index = 0; + + if (mlxsw_sp_fib_entry_should_offload(fib_entry)) { + trap_action = MLXSW_REG_RALUE_TRAP_ACTION_NOP; + rif_index = rif->rif_index; + } else { + trap_action = MLXSW_REG_RALUE_TRAP_ACTION_TRAP; + trap_id = MLXSW_TRAP_ID_RTR_INGRESS0; + } + + mlxsw_sp_fib_entry_ralue_pack(ralue_pl, fib_entry, op); + mlxsw_reg_ralue_act_local_pack(ralue_pl, trap_action, trap_id, + rif_index); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralue), ralue_pl); +} + +static int mlxsw_sp_fib_entry_op_trap(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) +{ + char ralue_pl[MLXSW_REG_RALUE_LEN]; + + mlxsw_sp_fib_entry_ralue_pack(ralue_pl, fib_entry, op); + mlxsw_reg_ralue_act_ip2me_pack(ralue_pl); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralue), ralue_pl); +} + +static int mlxsw_sp_fib_entry_op_blackhole(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) +{ + enum mlxsw_reg_ralue_trap_action trap_action; + char ralue_pl[MLXSW_REG_RALUE_LEN]; + + trap_action = MLXSW_REG_RALUE_TRAP_ACTION_DISCARD_ERROR; + mlxsw_sp_fib_entry_ralue_pack(ralue_pl, fib_entry, op); + mlxsw_reg_ralue_act_local_pack(ralue_pl, trap_action, 0, 0); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralue), ralue_pl); +} + +static int +mlxsw_sp_fib_entry_op_unreachable(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) +{ + enum mlxsw_reg_ralue_trap_action trap_action; + char ralue_pl[MLXSW_REG_RALUE_LEN]; + u16 trap_id; + + trap_action = MLXSW_REG_RALUE_TRAP_ACTION_TRAP; + trap_id = MLXSW_TRAP_ID_RTR_INGRESS1; + + mlxsw_sp_fib_entry_ralue_pack(ralue_pl, fib_entry, op); + mlxsw_reg_ralue_act_local_pack(ralue_pl, trap_action, trap_id, 0); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralue), ralue_pl); +} + +static int +mlxsw_sp_fib_entry_op_ipip_decap(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) +{ + struct mlxsw_sp_ipip_entry *ipip_entry = fib_entry->decap.ipip_entry; + const struct mlxsw_sp_ipip_ops *ipip_ops; + char ralue_pl[MLXSW_REG_RALUE_LEN]; + int err; + + if (WARN_ON(!ipip_entry)) + return -EINVAL; + + ipip_ops = mlxsw_sp->router->ipip_ops_arr[ipip_entry->ipipt]; + err = ipip_ops->decap_config(mlxsw_sp, ipip_entry, + fib_entry->decap.tunnel_index); + if (err) + return err; + + mlxsw_sp_fib_entry_ralue_pack(ralue_pl, fib_entry, op); + mlxsw_reg_ralue_act_ip2me_tun_pack(ralue_pl, + fib_entry->decap.tunnel_index); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralue), ralue_pl); +} + +static int mlxsw_sp_fib_entry_op_nve_decap(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) +{ + char ralue_pl[MLXSW_REG_RALUE_LEN]; + + mlxsw_sp_fib_entry_ralue_pack(ralue_pl, fib_entry, op); + mlxsw_reg_ralue_act_ip2me_tun_pack(ralue_pl, + fib_entry->decap.tunnel_index); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralue), ralue_pl); +} + +static int __mlxsw_sp_fib_entry_op(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) +{ + switch (fib_entry->type) { + case MLXSW_SP_FIB_ENTRY_TYPE_REMOTE: + return mlxsw_sp_fib_entry_op_remote(mlxsw_sp, fib_entry, op); + case MLXSW_SP_FIB_ENTRY_TYPE_LOCAL: + return mlxsw_sp_fib_entry_op_local(mlxsw_sp, fib_entry, op); + case MLXSW_SP_FIB_ENTRY_TYPE_TRAP: + return mlxsw_sp_fib_entry_op_trap(mlxsw_sp, fib_entry, op); + case MLXSW_SP_FIB_ENTRY_TYPE_BLACKHOLE: + return mlxsw_sp_fib_entry_op_blackhole(mlxsw_sp, fib_entry, op); + case MLXSW_SP_FIB_ENTRY_TYPE_UNREACHABLE: + return mlxsw_sp_fib_entry_op_unreachable(mlxsw_sp, fib_entry, + op); + case MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP: + return mlxsw_sp_fib_entry_op_ipip_decap(mlxsw_sp, + fib_entry, op); + case MLXSW_SP_FIB_ENTRY_TYPE_NVE_DECAP: + return mlxsw_sp_fib_entry_op_nve_decap(mlxsw_sp, fib_entry, op); + } + return -EINVAL; +} + +static int mlxsw_sp_fib_entry_op(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) +{ + int err = __mlxsw_sp_fib_entry_op(mlxsw_sp, fib_entry, op); + + if (err) + return err; + + mlxsw_sp_fib_entry_hw_flags_refresh(mlxsw_sp, fib_entry, op); + + return err; +} + +static int mlxsw_sp_fib_entry_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + return mlxsw_sp_fib_entry_op(mlxsw_sp, fib_entry, + MLXSW_REG_RALUE_OP_WRITE_WRITE); +} + +static int mlxsw_sp_fib_entry_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + return mlxsw_sp_fib_entry_op(mlxsw_sp, fib_entry, + MLXSW_REG_RALUE_OP_WRITE_DELETE); +} + +static int +mlxsw_sp_fib4_entry_type_set(struct mlxsw_sp *mlxsw_sp, + const struct fib_entry_notifier_info *fen_info, + struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_nexthop_group_info *nhgi = fib_entry->nh_group->nhgi; + union mlxsw_sp_l3addr dip = { .addr4 = htonl(fen_info->dst) }; + struct mlxsw_sp_router *router = mlxsw_sp->router; + u32 tb_id = mlxsw_sp_fix_tb_id(fen_info->tb_id); + int ifindex = nhgi->nexthops[0].ifindex; + struct mlxsw_sp_ipip_entry *ipip_entry; + + switch (fen_info->type) { + case RTN_LOCAL: + ipip_entry = mlxsw_sp_ipip_entry_find_by_decap(mlxsw_sp, ifindex, + MLXSW_SP_L3_PROTO_IPV4, dip); + if (ipip_entry && ipip_entry->ol_dev->flags & IFF_UP) { + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP; + return mlxsw_sp_fib_entry_decap_init(mlxsw_sp, + fib_entry, + ipip_entry); + } + if (mlxsw_sp_router_nve_is_decap(mlxsw_sp, tb_id, + MLXSW_SP_L3_PROTO_IPV4, + &dip)) { + u32 tunnel_index; + + tunnel_index = router->nve_decap_config.tunnel_index; + fib_entry->decap.tunnel_index = tunnel_index; + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_NVE_DECAP; + return 0; + } + fallthrough; + case RTN_BROADCAST: + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_TRAP; + return 0; + case RTN_BLACKHOLE: + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_BLACKHOLE; + return 0; + case RTN_UNREACHABLE: + case RTN_PROHIBIT: + /* Packets hitting these routes need to be trapped, but + * can do so with a lower priority than packets directed + * at the host, so use action type local instead of trap. + */ + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_UNREACHABLE; + return 0; + case RTN_UNICAST: + if (nhgi->gateway) + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_REMOTE; + else + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_LOCAL; + return 0; + default: + return -EINVAL; + } +} + +static void +mlxsw_sp_fib_entry_type_unset(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + switch (fib_entry->type) { + case MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP: + mlxsw_sp_fib_entry_decap_fini(mlxsw_sp, fib_entry); + break; + default: + break; + } +} + +static void +mlxsw_sp_fib4_entry_type_unset(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib4_entry *fib4_entry) +{ + mlxsw_sp_fib_entry_type_unset(mlxsw_sp, &fib4_entry->common); +} + +static struct mlxsw_sp_fib4_entry * +mlxsw_sp_fib4_entry_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node, + const struct fib_entry_notifier_info *fen_info) +{ + struct mlxsw_sp_fib4_entry *fib4_entry; + struct mlxsw_sp_fib_entry *fib_entry; + int err; + + fib4_entry = kzalloc(sizeof(*fib4_entry), GFP_KERNEL); + if (!fib4_entry) + return ERR_PTR(-ENOMEM); + fib_entry = &fib4_entry->common; + + err = mlxsw_sp_nexthop4_group_get(mlxsw_sp, fib_entry, fen_info->fi); + if (err) + goto err_nexthop4_group_get; + + err = mlxsw_sp_nexthop_group_vr_link(fib_entry->nh_group, + fib_node->fib); + if (err) + goto err_nexthop_group_vr_link; + + err = mlxsw_sp_fib4_entry_type_set(mlxsw_sp, fen_info, fib_entry); + if (err) + goto err_fib4_entry_type_set; + + fib4_entry->fi = fen_info->fi; + fib_info_hold(fib4_entry->fi); + fib4_entry->tb_id = fen_info->tb_id; + fib4_entry->type = fen_info->type; + fib4_entry->dscp = fen_info->dscp; + + fib_entry->fib_node = fib_node; + + return fib4_entry; + +err_fib4_entry_type_set: + mlxsw_sp_nexthop_group_vr_unlink(fib_entry->nh_group, fib_node->fib); +err_nexthop_group_vr_link: + mlxsw_sp_nexthop4_group_put(mlxsw_sp, &fib4_entry->common); +err_nexthop4_group_get: + kfree(fib4_entry); + return ERR_PTR(err); +} + +static void mlxsw_sp_fib4_entry_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib4_entry *fib4_entry) +{ + struct mlxsw_sp_fib_node *fib_node = fib4_entry->common.fib_node; + + fib_info_put(fib4_entry->fi); + mlxsw_sp_fib4_entry_type_unset(mlxsw_sp, fib4_entry); + mlxsw_sp_nexthop_group_vr_unlink(fib4_entry->common.nh_group, + fib_node->fib); + mlxsw_sp_nexthop4_group_put(mlxsw_sp, &fib4_entry->common); + kfree(fib4_entry); +} + +static struct mlxsw_sp_fib4_entry * +mlxsw_sp_fib4_entry_lookup(struct mlxsw_sp *mlxsw_sp, + const struct fib_entry_notifier_info *fen_info) +{ + struct mlxsw_sp_fib4_entry *fib4_entry; + struct mlxsw_sp_fib_node *fib_node; + struct mlxsw_sp_fib *fib; + struct mlxsw_sp_vr *vr; + + vr = mlxsw_sp_vr_find(mlxsw_sp, fen_info->tb_id); + if (!vr) + return NULL; + fib = mlxsw_sp_vr_fib(vr, MLXSW_SP_L3_PROTO_IPV4); + + fib_node = mlxsw_sp_fib_node_lookup(fib, &fen_info->dst, + sizeof(fen_info->dst), + fen_info->dst_len); + if (!fib_node) + return NULL; + + fib4_entry = container_of(fib_node->fib_entry, + struct mlxsw_sp_fib4_entry, common); + if (fib4_entry->tb_id == fen_info->tb_id && + fib4_entry->dscp == fen_info->dscp && + fib4_entry->type == fen_info->type && + fib4_entry->fi == fen_info->fi) + return fib4_entry; + + return NULL; +} + +static const struct rhashtable_params mlxsw_sp_fib_ht_params = { + .key_offset = offsetof(struct mlxsw_sp_fib_node, key), + .head_offset = offsetof(struct mlxsw_sp_fib_node, ht_node), + .key_len = sizeof(struct mlxsw_sp_fib_key), + .automatic_shrinking = true, +}; + +static int mlxsw_sp_fib_node_insert(struct mlxsw_sp_fib *fib, + struct mlxsw_sp_fib_node *fib_node) +{ + return rhashtable_insert_fast(&fib->ht, &fib_node->ht_node, + mlxsw_sp_fib_ht_params); +} + +static void mlxsw_sp_fib_node_remove(struct mlxsw_sp_fib *fib, + struct mlxsw_sp_fib_node *fib_node) +{ + rhashtable_remove_fast(&fib->ht, &fib_node->ht_node, + mlxsw_sp_fib_ht_params); +} + +static struct mlxsw_sp_fib_node * +mlxsw_sp_fib_node_lookup(struct mlxsw_sp_fib *fib, const void *addr, + size_t addr_len, unsigned char prefix_len) +{ + struct mlxsw_sp_fib_key key; + + memset(&key, 0, sizeof(key)); + memcpy(key.addr, addr, addr_len); + key.prefix_len = prefix_len; + return rhashtable_lookup_fast(&fib->ht, &key, mlxsw_sp_fib_ht_params); +} + +static struct mlxsw_sp_fib_node * +mlxsw_sp_fib_node_create(struct mlxsw_sp_fib *fib, const void *addr, + size_t addr_len, unsigned char prefix_len) +{ + struct mlxsw_sp_fib_node *fib_node; + + fib_node = kzalloc(sizeof(*fib_node), GFP_KERNEL); + if (!fib_node) + return NULL; + + list_add(&fib_node->list, &fib->node_list); + memcpy(fib_node->key.addr, addr, addr_len); + fib_node->key.prefix_len = prefix_len; + + return fib_node; +} + +static void mlxsw_sp_fib_node_destroy(struct mlxsw_sp_fib_node *fib_node) +{ + list_del(&fib_node->list); + kfree(fib_node); +} + +static int mlxsw_sp_fib_lpm_tree_link(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node) +{ + struct mlxsw_sp_prefix_usage req_prefix_usage; + struct mlxsw_sp_fib *fib = fib_node->fib; + struct mlxsw_sp_lpm_tree *lpm_tree; + int err; + + lpm_tree = mlxsw_sp->router->lpm.proto_trees[fib->proto]; + if (lpm_tree->prefix_ref_count[fib_node->key.prefix_len] != 0) + goto out; + + mlxsw_sp_prefix_usage_cpy(&req_prefix_usage, &lpm_tree->prefix_usage); + mlxsw_sp_prefix_usage_set(&req_prefix_usage, fib_node->key.prefix_len); + lpm_tree = mlxsw_sp_lpm_tree_get(mlxsw_sp, &req_prefix_usage, + fib->proto); + if (IS_ERR(lpm_tree)) + return PTR_ERR(lpm_tree); + + err = mlxsw_sp_vrs_lpm_tree_replace(mlxsw_sp, fib, lpm_tree); + if (err) + goto err_lpm_tree_replace; + +out: + lpm_tree->prefix_ref_count[fib_node->key.prefix_len]++; + return 0; + +err_lpm_tree_replace: + mlxsw_sp_lpm_tree_put(mlxsw_sp, lpm_tree); + return err; +} + +static void mlxsw_sp_fib_lpm_tree_unlink(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node) +{ + struct mlxsw_sp_lpm_tree *lpm_tree = fib_node->fib->lpm_tree; + struct mlxsw_sp_prefix_usage req_prefix_usage; + struct mlxsw_sp_fib *fib = fib_node->fib; + int err; + + if (--lpm_tree->prefix_ref_count[fib_node->key.prefix_len] != 0) + return; + /* Try to construct a new LPM tree from the current prefix usage + * minus the unused one. If we fail, continue using the old one. + */ + mlxsw_sp_prefix_usage_cpy(&req_prefix_usage, &lpm_tree->prefix_usage); + mlxsw_sp_prefix_usage_clear(&req_prefix_usage, + fib_node->key.prefix_len); + lpm_tree = mlxsw_sp_lpm_tree_get(mlxsw_sp, &req_prefix_usage, + fib->proto); + if (IS_ERR(lpm_tree)) + return; + + err = mlxsw_sp_vrs_lpm_tree_replace(mlxsw_sp, fib, lpm_tree); + if (err) + goto err_lpm_tree_replace; + + return; + +err_lpm_tree_replace: + mlxsw_sp_lpm_tree_put(mlxsw_sp, lpm_tree); +} + +static int mlxsw_sp_fib_node_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node, + struct mlxsw_sp_fib *fib) +{ + int err; + + err = mlxsw_sp_fib_node_insert(fib, fib_node); + if (err) + return err; + fib_node->fib = fib; + + err = mlxsw_sp_fib_lpm_tree_link(mlxsw_sp, fib_node); + if (err) + goto err_fib_lpm_tree_link; + + return 0; + +err_fib_lpm_tree_link: + fib_node->fib = NULL; + mlxsw_sp_fib_node_remove(fib, fib_node); + return err; +} + +static void mlxsw_sp_fib_node_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node) +{ + struct mlxsw_sp_fib *fib = fib_node->fib; + + mlxsw_sp_fib_lpm_tree_unlink(mlxsw_sp, fib_node); + fib_node->fib = NULL; + mlxsw_sp_fib_node_remove(fib, fib_node); +} + +static struct mlxsw_sp_fib_node * +mlxsw_sp_fib_node_get(struct mlxsw_sp *mlxsw_sp, u32 tb_id, const void *addr, + size_t addr_len, unsigned char prefix_len, + enum mlxsw_sp_l3proto proto) +{ + struct mlxsw_sp_fib_node *fib_node; + struct mlxsw_sp_fib *fib; + struct mlxsw_sp_vr *vr; + int err; + + vr = mlxsw_sp_vr_get(mlxsw_sp, tb_id, NULL); + if (IS_ERR(vr)) + return ERR_CAST(vr); + fib = mlxsw_sp_vr_fib(vr, proto); + + fib_node = mlxsw_sp_fib_node_lookup(fib, addr, addr_len, prefix_len); + if (fib_node) + return fib_node; + + fib_node = mlxsw_sp_fib_node_create(fib, addr, addr_len, prefix_len); + if (!fib_node) { + err = -ENOMEM; + goto err_fib_node_create; + } + + err = mlxsw_sp_fib_node_init(mlxsw_sp, fib_node, fib); + if (err) + goto err_fib_node_init; + + return fib_node; + +err_fib_node_init: + mlxsw_sp_fib_node_destroy(fib_node); +err_fib_node_create: + mlxsw_sp_vr_put(mlxsw_sp, vr); + return ERR_PTR(err); +} + +static void mlxsw_sp_fib_node_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node) +{ + struct mlxsw_sp_vr *vr = fib_node->fib->vr; + + if (fib_node->fib_entry) + return; + mlxsw_sp_fib_node_fini(mlxsw_sp, fib_node); + mlxsw_sp_fib_node_destroy(fib_node); + mlxsw_sp_vr_put(mlxsw_sp, vr); +} + +static int mlxsw_sp_fib_node_entry_link(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_fib_node *fib_node = fib_entry->fib_node; + int err; + + fib_node->fib_entry = fib_entry; + + err = mlxsw_sp_fib_entry_update(mlxsw_sp, fib_entry); + if (err) + goto err_fib_entry_update; + + return 0; + +err_fib_entry_update: + fib_node->fib_entry = NULL; + return err; +} + +static void +mlxsw_sp_fib_node_entry_unlink(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_fib_node *fib_node = fib_entry->fib_node; + + mlxsw_sp_fib_entry_del(mlxsw_sp, fib_entry); + fib_node->fib_entry = NULL; +} + +static bool mlxsw_sp_fib4_allow_replace(struct mlxsw_sp_fib4_entry *fib4_entry) +{ + struct mlxsw_sp_fib_node *fib_node = fib4_entry->common.fib_node; + struct mlxsw_sp_fib4_entry *fib4_replaced; + + if (!fib_node->fib_entry) + return true; + + fib4_replaced = container_of(fib_node->fib_entry, + struct mlxsw_sp_fib4_entry, common); + if (fib4_entry->tb_id == RT_TABLE_MAIN && + fib4_replaced->tb_id == RT_TABLE_LOCAL) + return false; + + return true; +} + +static int +mlxsw_sp_router_fib4_replace(struct mlxsw_sp *mlxsw_sp, + const struct fib_entry_notifier_info *fen_info) +{ + struct mlxsw_sp_fib4_entry *fib4_entry, *fib4_replaced; + struct mlxsw_sp_fib_entry *replaced; + struct mlxsw_sp_fib_node *fib_node; + int err; + + if (fen_info->fi->nh && + !mlxsw_sp_nexthop_obj_group_lookup(mlxsw_sp, fen_info->fi->nh->id)) + return 0; + + fib_node = mlxsw_sp_fib_node_get(mlxsw_sp, fen_info->tb_id, + &fen_info->dst, sizeof(fen_info->dst), + fen_info->dst_len, + MLXSW_SP_L3_PROTO_IPV4); + if (IS_ERR(fib_node)) { + dev_warn(mlxsw_sp->bus_info->dev, "Failed to get FIB node\n"); + return PTR_ERR(fib_node); + } + + fib4_entry = mlxsw_sp_fib4_entry_create(mlxsw_sp, fib_node, fen_info); + if (IS_ERR(fib4_entry)) { + dev_warn(mlxsw_sp->bus_info->dev, "Failed to create FIB entry\n"); + err = PTR_ERR(fib4_entry); + goto err_fib4_entry_create; + } + + if (!mlxsw_sp_fib4_allow_replace(fib4_entry)) { + mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_entry); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); + return 0; + } + + replaced = fib_node->fib_entry; + err = mlxsw_sp_fib_node_entry_link(mlxsw_sp, &fib4_entry->common); + if (err) { + dev_warn(mlxsw_sp->bus_info->dev, "Failed to link FIB entry to node\n"); + goto err_fib_node_entry_link; + } + + /* Nothing to replace */ + if (!replaced) + return 0; + + mlxsw_sp_fib_entry_hw_flags_clear(mlxsw_sp, replaced); + fib4_replaced = container_of(replaced, struct mlxsw_sp_fib4_entry, + common); + mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_replaced); + + return 0; + +err_fib_node_entry_link: + fib_node->fib_entry = replaced; + mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_entry); +err_fib4_entry_create: + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); + return err; +} + +static void mlxsw_sp_router_fib4_del(struct mlxsw_sp *mlxsw_sp, + struct fib_entry_notifier_info *fen_info) +{ + struct mlxsw_sp_fib4_entry *fib4_entry; + struct mlxsw_sp_fib_node *fib_node; + + fib4_entry = mlxsw_sp_fib4_entry_lookup(mlxsw_sp, fen_info); + if (!fib4_entry) + return; + fib_node = fib4_entry->common.fib_node; + + mlxsw_sp_fib_node_entry_unlink(mlxsw_sp, &fib4_entry->common); + mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_entry); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); +} + +static bool mlxsw_sp_fib6_rt_should_ignore(const struct fib6_info *rt) +{ + /* Multicast routes aren't supported, so ignore them. Neighbour + * Discovery packets are specifically trapped. + */ + if (ipv6_addr_type(&rt->fib6_dst.addr) & IPV6_ADDR_MULTICAST) + return true; + + /* Cloned routes are irrelevant in the forwarding path. */ + if (rt->fib6_flags & RTF_CACHE) + return true; + + return false; +} + +static struct mlxsw_sp_rt6 *mlxsw_sp_rt6_create(struct fib6_info *rt) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + mlxsw_sp_rt6 = kzalloc(sizeof(*mlxsw_sp_rt6), GFP_KERNEL); + if (!mlxsw_sp_rt6) + return ERR_PTR(-ENOMEM); + + /* In case of route replace, replaced route is deleted with + * no notification. Take reference to prevent accessing freed + * memory. + */ + mlxsw_sp_rt6->rt = rt; + fib6_info_hold(rt); + + return mlxsw_sp_rt6; +} + +#if IS_ENABLED(CONFIG_IPV6) +static void mlxsw_sp_rt6_release(struct fib6_info *rt) +{ + fib6_info_release(rt); +} +#else +static void mlxsw_sp_rt6_release(struct fib6_info *rt) +{ +} +#endif + +static void mlxsw_sp_rt6_destroy(struct mlxsw_sp_rt6 *mlxsw_sp_rt6) +{ + struct fib6_nh *fib6_nh = mlxsw_sp_rt6->rt->fib6_nh; + + if (!mlxsw_sp_rt6->rt->nh) + fib6_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; + mlxsw_sp_rt6_release(mlxsw_sp_rt6->rt); + kfree(mlxsw_sp_rt6); +} + +static struct fib6_info * +mlxsw_sp_fib6_entry_rt(const struct mlxsw_sp_fib6_entry *fib6_entry) +{ + return list_first_entry(&fib6_entry->rt6_list, struct mlxsw_sp_rt6, + list)->rt; +} + +static struct mlxsw_sp_rt6 * +mlxsw_sp_fib6_entry_rt_find(const struct mlxsw_sp_fib6_entry *fib6_entry, + const struct fib6_info *rt) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { + if (mlxsw_sp_rt6->rt == rt) + return mlxsw_sp_rt6; + } + + return NULL; +} + +static bool mlxsw_sp_nexthop6_ipip_type(const struct mlxsw_sp *mlxsw_sp, + const struct fib6_info *rt, + enum mlxsw_sp_ipip_type *ret) +{ + return rt->fib6_nh->fib_nh_dev && + mlxsw_sp_netdev_ipip_type(mlxsw_sp, rt->fib6_nh->fib_nh_dev, ret); +} + +static int mlxsw_sp_nexthop6_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp, + struct mlxsw_sp_nexthop *nh, + const struct fib6_info *rt) +{ + struct net_device *dev = rt->fib6_nh->fib_nh_dev; + int err; + + nh->nhgi = nh_grp->nhgi; + nh->nh_weight = rt->fib6_nh->fib_nh_weight; + memcpy(&nh->gw_addr, &rt->fib6_nh->fib_nh_gw6, sizeof(nh->gw_addr)); +#if IS_ENABLED(CONFIG_IPV6) + nh->neigh_tbl = &nd_tbl; +#endif + mlxsw_sp_nexthop_counter_alloc(mlxsw_sp, nh); + + list_add_tail(&nh->router_list_node, &mlxsw_sp->router->nexthop_list); + + if (!dev) + return 0; + nh->ifindex = dev->ifindex; + + err = mlxsw_sp_nexthop_type_init(mlxsw_sp, nh, dev); + if (err) + goto err_nexthop_type_init; + + return 0; + +err_nexthop_type_init: + list_del(&nh->router_list_node); + mlxsw_sp_nexthop_counter_free(mlxsw_sp, nh); + return err; +} + +static void mlxsw_sp_nexthop6_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + mlxsw_sp_nexthop_type_fini(mlxsw_sp, nh); + list_del(&nh->router_list_node); + mlxsw_sp_nexthop_counter_free(mlxsw_sp, nh); +} + +static bool mlxsw_sp_rt6_is_gateway(const struct mlxsw_sp *mlxsw_sp, + const struct fib6_info *rt) +{ + return rt->fib6_nh->fib_nh_gw_family || + mlxsw_sp_nexthop6_ipip_type(mlxsw_sp, rt, NULL); +} + +static int +mlxsw_sp_nexthop6_group_info_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_nexthop_group_info *nhgi; + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + struct mlxsw_sp_nexthop *nh; + int err, i; + + nhgi = kzalloc(struct_size(nhgi, nexthops, fib6_entry->nrt6), + GFP_KERNEL); + if (!nhgi) + return -ENOMEM; + nh_grp->nhgi = nhgi; + nhgi->nh_grp = nh_grp; + mlxsw_sp_rt6 = list_first_entry(&fib6_entry->rt6_list, + struct mlxsw_sp_rt6, list); + nhgi->gateway = mlxsw_sp_rt6_is_gateway(mlxsw_sp, mlxsw_sp_rt6->rt); + nhgi->count = fib6_entry->nrt6; + for (i = 0; i < nhgi->count; i++) { + struct fib6_info *rt = mlxsw_sp_rt6->rt; + + nh = &nhgi->nexthops[i]; + err = mlxsw_sp_nexthop6_init(mlxsw_sp, nh_grp, nh, rt); + if (err) + goto err_nexthop6_init; + mlxsw_sp_rt6 = list_next_entry(mlxsw_sp_rt6, list); + } + nh_grp->nhgi = nhgi; + err = mlxsw_sp_nexthop_group_inc(mlxsw_sp); + if (err) + goto err_group_inc; + err = mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh_grp); + if (err) + goto err_group_refresh; + + return 0; + +err_group_refresh: + mlxsw_sp_nexthop_group_dec(mlxsw_sp); +err_group_inc: + i = nhgi->count; +err_nexthop6_init: + for (i--; i >= 0; i--) { + nh = &nhgi->nexthops[i]; + mlxsw_sp_nexthop6_fini(mlxsw_sp, nh); + } + kfree(nhgi); + return err; +} + +static void +mlxsw_sp_nexthop6_group_info_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + struct mlxsw_sp_nexthop_group_info *nhgi = nh_grp->nhgi; + int i; + + mlxsw_sp_nexthop_group_dec(mlxsw_sp); + for (i = nhgi->count - 1; i >= 0; i--) { + struct mlxsw_sp_nexthop *nh = &nhgi->nexthops[i]; + + mlxsw_sp_nexthop6_fini(mlxsw_sp, nh); + } + mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh_grp); + WARN_ON_ONCE(nhgi->adj_index_valid); + kfree(nhgi); +} + +static struct mlxsw_sp_nexthop_group * +mlxsw_sp_nexthop6_group_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_nexthop_group *nh_grp; + int err; + + nh_grp = kzalloc(sizeof(*nh_grp), GFP_KERNEL); + if (!nh_grp) + return ERR_PTR(-ENOMEM); + INIT_LIST_HEAD(&nh_grp->vr_list); + err = rhashtable_init(&nh_grp->vr_ht, + &mlxsw_sp_nexthop_group_vr_ht_params); + if (err) + goto err_nexthop_group_vr_ht_init; + INIT_LIST_HEAD(&nh_grp->fib_list); + nh_grp->type = MLXSW_SP_NEXTHOP_GROUP_TYPE_IPV6; + + err = mlxsw_sp_nexthop6_group_info_init(mlxsw_sp, nh_grp, fib6_entry); + if (err) + goto err_nexthop_group_info_init; + + err = mlxsw_sp_nexthop_group_insert(mlxsw_sp, nh_grp); + if (err) + goto err_nexthop_group_insert; + + nh_grp->can_destroy = true; + + return nh_grp; + +err_nexthop_group_insert: + mlxsw_sp_nexthop6_group_info_fini(mlxsw_sp, nh_grp); +err_nexthop_group_info_init: + rhashtable_destroy(&nh_grp->vr_ht); +err_nexthop_group_vr_ht_init: + kfree(nh_grp); + return ERR_PTR(err); +} + +static void +mlxsw_sp_nexthop6_group_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + if (!nh_grp->can_destroy) + return; + mlxsw_sp_nexthop_group_remove(mlxsw_sp, nh_grp); + mlxsw_sp_nexthop6_group_info_fini(mlxsw_sp, nh_grp); + WARN_ON_ONCE(!list_empty(&nh_grp->vr_list)); + rhashtable_destroy(&nh_grp->vr_ht); + kfree(nh_grp); +} + +static int mlxsw_sp_nexthop6_group_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct fib6_info *rt = mlxsw_sp_fib6_entry_rt(fib6_entry); + struct mlxsw_sp_nexthop_group *nh_grp; + + if (rt->nh) { + nh_grp = mlxsw_sp_nexthop_obj_group_lookup(mlxsw_sp, + rt->nh->id); + if (WARN_ON_ONCE(!nh_grp)) + return -EINVAL; + goto out; + } + + nh_grp = mlxsw_sp_nexthop6_group_lookup(mlxsw_sp, fib6_entry); + if (!nh_grp) { + nh_grp = mlxsw_sp_nexthop6_group_create(mlxsw_sp, fib6_entry); + if (IS_ERR(nh_grp)) + return PTR_ERR(nh_grp); + } + + /* The route and the nexthop are described by the same struct, so we + * need to the update the nexthop offload indication for the new route. + */ + __mlxsw_sp_nexthop6_group_offload_refresh(nh_grp, fib6_entry); + +out: + list_add_tail(&fib6_entry->common.nexthop_group_node, + &nh_grp->fib_list); + fib6_entry->common.nh_group = nh_grp; + + return 0; +} + +static void mlxsw_sp_nexthop6_group_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; + + list_del(&fib_entry->nexthop_group_node); + if (!list_empty(&nh_grp->fib_list)) + return; + + if (nh_grp->type == MLXSW_SP_NEXTHOP_GROUP_TYPE_OBJ) { + mlxsw_sp_nexthop_obj_group_destroy(mlxsw_sp, nh_grp); + return; + } + + mlxsw_sp_nexthop6_group_destroy(mlxsw_sp, nh_grp); +} + +static int +mlxsw_sp_nexthop6_group_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_nexthop_group *old_nh_grp = fib6_entry->common.nh_group; + struct mlxsw_sp_fib_node *fib_node = fib6_entry->common.fib_node; + int err; + + mlxsw_sp_nexthop_group_vr_unlink(old_nh_grp, fib_node->fib); + fib6_entry->common.nh_group = NULL; + list_del(&fib6_entry->common.nexthop_group_node); + + err = mlxsw_sp_nexthop6_group_get(mlxsw_sp, fib6_entry); + if (err) + goto err_nexthop6_group_get; + + err = mlxsw_sp_nexthop_group_vr_link(fib6_entry->common.nh_group, + fib_node->fib); + if (err) + goto err_nexthop_group_vr_link; + + /* In case this entry is offloaded, then the adjacency index + * currently associated with it in the device's table is that + * of the old group. Start using the new one instead. + */ + err = mlxsw_sp_fib_entry_update(mlxsw_sp, &fib6_entry->common); + if (err) + goto err_fib_entry_update; + + if (list_empty(&old_nh_grp->fib_list)) + mlxsw_sp_nexthop6_group_destroy(mlxsw_sp, old_nh_grp); + + return 0; + +err_fib_entry_update: + mlxsw_sp_nexthop_group_vr_unlink(fib6_entry->common.nh_group, + fib_node->fib); +err_nexthop_group_vr_link: + mlxsw_sp_nexthop6_group_put(mlxsw_sp, &fib6_entry->common); +err_nexthop6_group_get: + list_add_tail(&fib6_entry->common.nexthop_group_node, + &old_nh_grp->fib_list); + fib6_entry->common.nh_group = old_nh_grp; + mlxsw_sp_nexthop_group_vr_link(old_nh_grp, fib_node->fib); + return err; +} + +static int +mlxsw_sp_fib6_entry_nexthop_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry, + struct fib6_info **rt_arr, unsigned int nrt6) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + int err, i; + + for (i = 0; i < nrt6; i++) { + mlxsw_sp_rt6 = mlxsw_sp_rt6_create(rt_arr[i]); + if (IS_ERR(mlxsw_sp_rt6)) { + err = PTR_ERR(mlxsw_sp_rt6); + goto err_rt6_unwind; + } + + list_add_tail(&mlxsw_sp_rt6->list, &fib6_entry->rt6_list); + fib6_entry->nrt6++; + } + + err = mlxsw_sp_nexthop6_group_update(mlxsw_sp, fib6_entry); + if (err) + goto err_rt6_unwind; + + return 0; + +err_rt6_unwind: + for (; i > 0; i--) { + fib6_entry->nrt6--; + mlxsw_sp_rt6 = list_last_entry(&fib6_entry->rt6_list, + struct mlxsw_sp_rt6, list); + list_del(&mlxsw_sp_rt6->list); + mlxsw_sp_rt6_destroy(mlxsw_sp_rt6); + } + return err; +} + +static void +mlxsw_sp_fib6_entry_nexthop_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry, + struct fib6_info **rt_arr, unsigned int nrt6) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + int i; + + for (i = 0; i < nrt6; i++) { + mlxsw_sp_rt6 = mlxsw_sp_fib6_entry_rt_find(fib6_entry, + rt_arr[i]); + if (WARN_ON_ONCE(!mlxsw_sp_rt6)) + continue; + + fib6_entry->nrt6--; + list_del(&mlxsw_sp_rt6->list); + mlxsw_sp_rt6_destroy(mlxsw_sp_rt6); + } + + mlxsw_sp_nexthop6_group_update(mlxsw_sp, fib6_entry); +} + +static int +mlxsw_sp_fib6_entry_type_set_local(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + const struct fib6_info *rt) +{ + struct mlxsw_sp_nexthop_group_info *nhgi = fib_entry->nh_group->nhgi; + union mlxsw_sp_l3addr dip = { .addr6 = rt->fib6_dst.addr }; + u32 tb_id = mlxsw_sp_fix_tb_id(rt->fib6_table->tb6_id); + struct mlxsw_sp_router *router = mlxsw_sp->router; + int ifindex = nhgi->nexthops[0].ifindex; + struct mlxsw_sp_ipip_entry *ipip_entry; + + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_TRAP; + ipip_entry = mlxsw_sp_ipip_entry_find_by_decap(mlxsw_sp, ifindex, + MLXSW_SP_L3_PROTO_IPV6, + dip); + + if (ipip_entry && ipip_entry->ol_dev->flags & IFF_UP) { + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP; + return mlxsw_sp_fib_entry_decap_init(mlxsw_sp, fib_entry, + ipip_entry); + } + if (mlxsw_sp_router_nve_is_decap(mlxsw_sp, tb_id, + MLXSW_SP_L3_PROTO_IPV6, &dip)) { + u32 tunnel_index; + + tunnel_index = router->nve_decap_config.tunnel_index; + fib_entry->decap.tunnel_index = tunnel_index; + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_NVE_DECAP; + } + + return 0; +} + +static int mlxsw_sp_fib6_entry_type_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + const struct fib6_info *rt) +{ + if (rt->fib6_flags & RTF_LOCAL) + return mlxsw_sp_fib6_entry_type_set_local(mlxsw_sp, fib_entry, + rt); + if (rt->fib6_flags & RTF_ANYCAST) + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_TRAP; + else if (rt->fib6_type == RTN_BLACKHOLE) + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_BLACKHOLE; + else if (rt->fib6_flags & RTF_REJECT) + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_UNREACHABLE; + else if (fib_entry->nh_group->nhgi->gateway) + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_REMOTE; + else + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_LOCAL; + + return 0; +} + +static void +mlxsw_sp_fib6_entry_rt_destroy_all(struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6, *tmp; + + list_for_each_entry_safe(mlxsw_sp_rt6, tmp, &fib6_entry->rt6_list, + list) { + fib6_entry->nrt6--; + list_del(&mlxsw_sp_rt6->list); + mlxsw_sp_rt6_destroy(mlxsw_sp_rt6); + } +} + +static struct mlxsw_sp_fib6_entry * +mlxsw_sp_fib6_entry_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node, + struct fib6_info **rt_arr, unsigned int nrt6) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + struct mlxsw_sp_fib_entry *fib_entry; + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + int err, i; + + fib6_entry = kzalloc(sizeof(*fib6_entry), GFP_KERNEL); + if (!fib6_entry) + return ERR_PTR(-ENOMEM); + fib_entry = &fib6_entry->common; + + INIT_LIST_HEAD(&fib6_entry->rt6_list); + + for (i = 0; i < nrt6; i++) { + mlxsw_sp_rt6 = mlxsw_sp_rt6_create(rt_arr[i]); + if (IS_ERR(mlxsw_sp_rt6)) { + err = PTR_ERR(mlxsw_sp_rt6); + goto err_rt6_unwind; + } + list_add_tail(&mlxsw_sp_rt6->list, &fib6_entry->rt6_list); + fib6_entry->nrt6++; + } + + err = mlxsw_sp_nexthop6_group_get(mlxsw_sp, fib6_entry); + if (err) + goto err_rt6_unwind; + + err = mlxsw_sp_nexthop_group_vr_link(fib_entry->nh_group, + fib_node->fib); + if (err) + goto err_nexthop_group_vr_link; + + err = mlxsw_sp_fib6_entry_type_set(mlxsw_sp, fib_entry, rt_arr[0]); + if (err) + goto err_fib6_entry_type_set; + + fib_entry->fib_node = fib_node; + + return fib6_entry; + +err_fib6_entry_type_set: + mlxsw_sp_nexthop_group_vr_unlink(fib_entry->nh_group, fib_node->fib); +err_nexthop_group_vr_link: + mlxsw_sp_nexthop6_group_put(mlxsw_sp, fib_entry); +err_rt6_unwind: + for (; i > 0; i--) { + fib6_entry->nrt6--; + mlxsw_sp_rt6 = list_last_entry(&fib6_entry->rt6_list, + struct mlxsw_sp_rt6, list); + list_del(&mlxsw_sp_rt6->list); + mlxsw_sp_rt6_destroy(mlxsw_sp_rt6); + } + kfree(fib6_entry); + return ERR_PTR(err); +} + +static void +mlxsw_sp_fib6_entry_type_unset(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + mlxsw_sp_fib_entry_type_unset(mlxsw_sp, &fib6_entry->common); +} + +static void mlxsw_sp_fib6_entry_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_fib_node *fib_node = fib6_entry->common.fib_node; + + mlxsw_sp_fib6_entry_type_unset(mlxsw_sp, fib6_entry); + mlxsw_sp_nexthop_group_vr_unlink(fib6_entry->common.nh_group, + fib_node->fib); + mlxsw_sp_nexthop6_group_put(mlxsw_sp, &fib6_entry->common); + mlxsw_sp_fib6_entry_rt_destroy_all(fib6_entry); + WARN_ON(fib6_entry->nrt6); + kfree(fib6_entry); +} + +static struct mlxsw_sp_fib6_entry * +mlxsw_sp_fib6_entry_lookup(struct mlxsw_sp *mlxsw_sp, + const struct fib6_info *rt) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + struct mlxsw_sp_fib_node *fib_node; + struct mlxsw_sp_fib *fib; + struct fib6_info *cmp_rt; + struct mlxsw_sp_vr *vr; + + vr = mlxsw_sp_vr_find(mlxsw_sp, rt->fib6_table->tb6_id); + if (!vr) + return NULL; + fib = mlxsw_sp_vr_fib(vr, MLXSW_SP_L3_PROTO_IPV6); + + fib_node = mlxsw_sp_fib_node_lookup(fib, &rt->fib6_dst.addr, + sizeof(rt->fib6_dst.addr), + rt->fib6_dst.plen); + if (!fib_node) + return NULL; + + fib6_entry = container_of(fib_node->fib_entry, + struct mlxsw_sp_fib6_entry, common); + cmp_rt = mlxsw_sp_fib6_entry_rt(fib6_entry); + if (rt->fib6_table->tb6_id == cmp_rt->fib6_table->tb6_id && + rt->fib6_metric == cmp_rt->fib6_metric && + mlxsw_sp_fib6_entry_rt_find(fib6_entry, rt)) + return fib6_entry; + + return NULL; +} + +static bool mlxsw_sp_fib6_allow_replace(struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_fib_node *fib_node = fib6_entry->common.fib_node; + struct mlxsw_sp_fib6_entry *fib6_replaced; + struct fib6_info *rt, *rt_replaced; + + if (!fib_node->fib_entry) + return true; + + fib6_replaced = container_of(fib_node->fib_entry, + struct mlxsw_sp_fib6_entry, + common); + rt = mlxsw_sp_fib6_entry_rt(fib6_entry); + rt_replaced = mlxsw_sp_fib6_entry_rt(fib6_replaced); + if (rt->fib6_table->tb6_id == RT_TABLE_MAIN && + rt_replaced->fib6_table->tb6_id == RT_TABLE_LOCAL) + return false; + + return true; +} + +static int mlxsw_sp_router_fib6_replace(struct mlxsw_sp *mlxsw_sp, + struct fib6_info **rt_arr, + unsigned int nrt6) +{ + struct mlxsw_sp_fib6_entry *fib6_entry, *fib6_replaced; + struct mlxsw_sp_fib_entry *replaced; + struct mlxsw_sp_fib_node *fib_node; + struct fib6_info *rt = rt_arr[0]; + int err; + + if (rt->fib6_src.plen) + return -EINVAL; + + if (mlxsw_sp_fib6_rt_should_ignore(rt)) + return 0; + + if (rt->nh && !mlxsw_sp_nexthop_obj_group_lookup(mlxsw_sp, rt->nh->id)) + return 0; + + fib_node = mlxsw_sp_fib_node_get(mlxsw_sp, rt->fib6_table->tb6_id, + &rt->fib6_dst.addr, + sizeof(rt->fib6_dst.addr), + rt->fib6_dst.plen, + MLXSW_SP_L3_PROTO_IPV6); + if (IS_ERR(fib_node)) + return PTR_ERR(fib_node); + + fib6_entry = mlxsw_sp_fib6_entry_create(mlxsw_sp, fib_node, rt_arr, + nrt6); + if (IS_ERR(fib6_entry)) { + err = PTR_ERR(fib6_entry); + goto err_fib6_entry_create; + } + + if (!mlxsw_sp_fib6_allow_replace(fib6_entry)) { + mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_entry); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); + return 0; + } + + replaced = fib_node->fib_entry; + err = mlxsw_sp_fib_node_entry_link(mlxsw_sp, &fib6_entry->common); + if (err) + goto err_fib_node_entry_link; + + /* Nothing to replace */ + if (!replaced) + return 0; + + mlxsw_sp_fib_entry_hw_flags_clear(mlxsw_sp, replaced); + fib6_replaced = container_of(replaced, struct mlxsw_sp_fib6_entry, + common); + mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_replaced); + + return 0; + +err_fib_node_entry_link: + fib_node->fib_entry = replaced; + mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_entry); +err_fib6_entry_create: + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); + return err; +} + +static int mlxsw_sp_router_fib6_append(struct mlxsw_sp *mlxsw_sp, + struct fib6_info **rt_arr, + unsigned int nrt6) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + struct mlxsw_sp_fib_node *fib_node; + struct fib6_info *rt = rt_arr[0]; + int err; + + if (rt->fib6_src.plen) + return -EINVAL; + + if (mlxsw_sp_fib6_rt_should_ignore(rt)) + return 0; + + fib_node = mlxsw_sp_fib_node_get(mlxsw_sp, rt->fib6_table->tb6_id, + &rt->fib6_dst.addr, + sizeof(rt->fib6_dst.addr), + rt->fib6_dst.plen, + MLXSW_SP_L3_PROTO_IPV6); + if (IS_ERR(fib_node)) + return PTR_ERR(fib_node); + + if (WARN_ON_ONCE(!fib_node->fib_entry)) { + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); + return -EINVAL; + } + + fib6_entry = container_of(fib_node->fib_entry, + struct mlxsw_sp_fib6_entry, common); + err = mlxsw_sp_fib6_entry_nexthop_add(mlxsw_sp, fib6_entry, rt_arr, + nrt6); + if (err) + goto err_fib6_entry_nexthop_add; + + return 0; + +err_fib6_entry_nexthop_add: + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); + return err; +} + +static void mlxsw_sp_router_fib6_del(struct mlxsw_sp *mlxsw_sp, + struct fib6_info **rt_arr, + unsigned int nrt6) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + struct mlxsw_sp_fib_node *fib_node; + struct fib6_info *rt = rt_arr[0]; + + if (mlxsw_sp_fib6_rt_should_ignore(rt)) + return; + + /* Multipath routes are first added to the FIB trie and only then + * notified. If we vetoed the addition, we will get a delete + * notification for a route we do not have. Therefore, do not warn if + * route was not found. + */ + fib6_entry = mlxsw_sp_fib6_entry_lookup(mlxsw_sp, rt); + if (!fib6_entry) + return; + + /* If not all the nexthops are deleted, then only reduce the nexthop + * group. + */ + if (nrt6 != fib6_entry->nrt6) { + mlxsw_sp_fib6_entry_nexthop_del(mlxsw_sp, fib6_entry, rt_arr, + nrt6); + return; + } + + fib_node = fib6_entry->common.fib_node; + + mlxsw_sp_fib_node_entry_unlink(mlxsw_sp, &fib6_entry->common); + mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_entry); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); +} + +static struct mlxsw_sp_mr_table * +mlxsw_sp_router_fibmr_family_to_table(struct mlxsw_sp_vr *vr, int family) +{ + if (family == RTNL_FAMILY_IPMR) + return vr->mr_table[MLXSW_SP_L3_PROTO_IPV4]; + else + return vr->mr_table[MLXSW_SP_L3_PROTO_IPV6]; +} + +static int mlxsw_sp_router_fibmr_add(struct mlxsw_sp *mlxsw_sp, + struct mfc_entry_notifier_info *men_info, + bool replace) +{ + struct mlxsw_sp_mr_table *mrt; + struct mlxsw_sp_vr *vr; + + vr = mlxsw_sp_vr_get(mlxsw_sp, men_info->tb_id, NULL); + if (IS_ERR(vr)) + return PTR_ERR(vr); + + mrt = mlxsw_sp_router_fibmr_family_to_table(vr, men_info->info.family); + return mlxsw_sp_mr_route_add(mrt, men_info->mfc, replace); +} + +static void mlxsw_sp_router_fibmr_del(struct mlxsw_sp *mlxsw_sp, + struct mfc_entry_notifier_info *men_info) +{ + struct mlxsw_sp_mr_table *mrt; + struct mlxsw_sp_vr *vr; + + vr = mlxsw_sp_vr_find(mlxsw_sp, men_info->tb_id); + if (WARN_ON(!vr)) + return; + + mrt = mlxsw_sp_router_fibmr_family_to_table(vr, men_info->info.family); + mlxsw_sp_mr_route_del(mrt, men_info->mfc); + mlxsw_sp_vr_put(mlxsw_sp, vr); +} + +static int +mlxsw_sp_router_fibmr_vif_add(struct mlxsw_sp *mlxsw_sp, + struct vif_entry_notifier_info *ven_info) +{ + struct mlxsw_sp_mr_table *mrt; + struct mlxsw_sp_rif *rif; + struct mlxsw_sp_vr *vr; + + vr = mlxsw_sp_vr_get(mlxsw_sp, ven_info->tb_id, NULL); + if (IS_ERR(vr)) + return PTR_ERR(vr); + + mrt = mlxsw_sp_router_fibmr_family_to_table(vr, ven_info->info.family); + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, ven_info->dev); + return mlxsw_sp_mr_vif_add(mrt, ven_info->dev, + ven_info->vif_index, + ven_info->vif_flags, rif); +} + +static void +mlxsw_sp_router_fibmr_vif_del(struct mlxsw_sp *mlxsw_sp, + struct vif_entry_notifier_info *ven_info) +{ + struct mlxsw_sp_mr_table *mrt; + struct mlxsw_sp_vr *vr; + + vr = mlxsw_sp_vr_find(mlxsw_sp, ven_info->tb_id); + if (WARN_ON(!vr)) + return; + + mrt = mlxsw_sp_router_fibmr_family_to_table(vr, ven_info->info.family); + mlxsw_sp_mr_vif_del(mrt, ven_info->vif_index); + mlxsw_sp_vr_put(mlxsw_sp, vr); +} + +static void mlxsw_sp_fib4_node_flush(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node) +{ + struct mlxsw_sp_fib4_entry *fib4_entry; + + fib4_entry = container_of(fib_node->fib_entry, + struct mlxsw_sp_fib4_entry, common); + mlxsw_sp_fib_node_entry_unlink(mlxsw_sp, fib_node->fib_entry); + mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_entry); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); +} + +static void mlxsw_sp_fib6_node_flush(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + + fib6_entry = container_of(fib_node->fib_entry, + struct mlxsw_sp_fib6_entry, common); + mlxsw_sp_fib_node_entry_unlink(mlxsw_sp, fib_node->fib_entry); + mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_entry); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); +} + +static void mlxsw_sp_fib_node_flush(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node) +{ + switch (fib_node->fib->proto) { + case MLXSW_SP_L3_PROTO_IPV4: + mlxsw_sp_fib4_node_flush(mlxsw_sp, fib_node); + break; + case MLXSW_SP_L3_PROTO_IPV6: + mlxsw_sp_fib6_node_flush(mlxsw_sp, fib_node); + break; + } +} + +static void mlxsw_sp_vr_fib_flush(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_vr *vr, + enum mlxsw_sp_l3proto proto) +{ + struct mlxsw_sp_fib *fib = mlxsw_sp_vr_fib(vr, proto); + struct mlxsw_sp_fib_node *fib_node, *tmp; + + list_for_each_entry_safe(fib_node, tmp, &fib->node_list, list) { + bool do_break = &tmp->list == &fib->node_list; + + mlxsw_sp_fib_node_flush(mlxsw_sp, fib_node); + if (do_break) + break; + } +} + +static void mlxsw_sp_router_fib_flush(struct mlxsw_sp *mlxsw_sp) +{ + int i, j; + + for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_VRS); i++) { + struct mlxsw_sp_vr *vr = &mlxsw_sp->router->vrs[i]; + + if (!mlxsw_sp_vr_is_used(vr)) + continue; + + for (j = 0; j < MLXSW_SP_L3_PROTO_MAX; j++) + mlxsw_sp_mr_table_flush(vr->mr_table[j]); + mlxsw_sp_vr_fib_flush(mlxsw_sp, vr, MLXSW_SP_L3_PROTO_IPV4); + + /* If virtual router was only used for IPv4, then it's no + * longer used. + */ + if (!mlxsw_sp_vr_is_used(vr)) + continue; + mlxsw_sp_vr_fib_flush(mlxsw_sp, vr, MLXSW_SP_L3_PROTO_IPV6); + } +} + +struct mlxsw_sp_fib6_event_work { + struct fib6_info **rt_arr; + unsigned int nrt6; +}; + +struct mlxsw_sp_fib_event_work { + struct work_struct work; + union { + struct mlxsw_sp_fib6_event_work fib6_work; + struct fib_entry_notifier_info fen_info; + struct fib_rule_notifier_info fr_info; + struct fib_nh_notifier_info fnh_info; + struct mfc_entry_notifier_info men_info; + struct vif_entry_notifier_info ven_info; + }; + struct mlxsw_sp *mlxsw_sp; + unsigned long event; +}; + +static int +mlxsw_sp_router_fib6_work_init(struct mlxsw_sp_fib6_event_work *fib6_work, + struct fib6_entry_notifier_info *fen6_info) +{ + struct fib6_info *rt = fen6_info->rt; + struct fib6_info **rt_arr; + struct fib6_info *iter; + unsigned int nrt6; + int i = 0; + + nrt6 = fen6_info->nsiblings + 1; + + rt_arr = kcalloc(nrt6, sizeof(struct fib6_info *), GFP_ATOMIC); + if (!rt_arr) + return -ENOMEM; + + fib6_work->rt_arr = rt_arr; + fib6_work->nrt6 = nrt6; + + rt_arr[0] = rt; + fib6_info_hold(rt); + + if (!fen6_info->nsiblings) + return 0; + + list_for_each_entry(iter, &rt->fib6_siblings, fib6_siblings) { + if (i == fen6_info->nsiblings) + break; + + rt_arr[i + 1] = iter; + fib6_info_hold(iter); + i++; + } + WARN_ON_ONCE(i != fen6_info->nsiblings); + + return 0; +} + +static void +mlxsw_sp_router_fib6_work_fini(struct mlxsw_sp_fib6_event_work *fib6_work) +{ + int i; + + for (i = 0; i < fib6_work->nrt6; i++) + mlxsw_sp_rt6_release(fib6_work->rt_arr[i]); + kfree(fib6_work->rt_arr); +} + +static void mlxsw_sp_router_fib4_event_work(struct work_struct *work) +{ + struct mlxsw_sp_fib_event_work *fib_work = + container_of(work, struct mlxsw_sp_fib_event_work, work); + struct mlxsw_sp *mlxsw_sp = fib_work->mlxsw_sp; + int err; + + mutex_lock(&mlxsw_sp->router->lock); + mlxsw_sp_span_respin(mlxsw_sp); + + switch (fib_work->event) { + case FIB_EVENT_ENTRY_REPLACE: + err = mlxsw_sp_router_fib4_replace(mlxsw_sp, + &fib_work->fen_info); + if (err) { + dev_warn(mlxsw_sp->bus_info->dev, "FIB replace failed.\n"); + mlxsw_sp_fib4_offload_failed_flag_set(mlxsw_sp, + &fib_work->fen_info); + } + fib_info_put(fib_work->fen_info.fi); + break; + case FIB_EVENT_ENTRY_DEL: + mlxsw_sp_router_fib4_del(mlxsw_sp, &fib_work->fen_info); + fib_info_put(fib_work->fen_info.fi); + break; + case FIB_EVENT_NH_ADD: + case FIB_EVENT_NH_DEL: + mlxsw_sp_nexthop4_event(mlxsw_sp, fib_work->event, + fib_work->fnh_info.fib_nh); + fib_info_put(fib_work->fnh_info.fib_nh->nh_parent); + break; + } + mutex_unlock(&mlxsw_sp->router->lock); + kfree(fib_work); +} + +static void mlxsw_sp_router_fib6_event_work(struct work_struct *work) +{ + struct mlxsw_sp_fib_event_work *fib_work = + container_of(work, struct mlxsw_sp_fib_event_work, work); + struct mlxsw_sp_fib6_event_work *fib6_work = &fib_work->fib6_work; + struct mlxsw_sp *mlxsw_sp = fib_work->mlxsw_sp; + int err; + + mutex_lock(&mlxsw_sp->router->lock); + mlxsw_sp_span_respin(mlxsw_sp); + + switch (fib_work->event) { + case FIB_EVENT_ENTRY_REPLACE: + err = mlxsw_sp_router_fib6_replace(mlxsw_sp, + fib6_work->rt_arr, + fib6_work->nrt6); + if (err) { + dev_warn(mlxsw_sp->bus_info->dev, "FIB replace failed.\n"); + mlxsw_sp_fib6_offload_failed_flag_set(mlxsw_sp, + fib6_work->rt_arr, + fib6_work->nrt6); + } + mlxsw_sp_router_fib6_work_fini(fib6_work); + break; + case FIB_EVENT_ENTRY_APPEND: + err = mlxsw_sp_router_fib6_append(mlxsw_sp, + fib6_work->rt_arr, + fib6_work->nrt6); + if (err) { + dev_warn(mlxsw_sp->bus_info->dev, "FIB append failed.\n"); + mlxsw_sp_fib6_offload_failed_flag_set(mlxsw_sp, + fib6_work->rt_arr, + fib6_work->nrt6); + } + mlxsw_sp_router_fib6_work_fini(fib6_work); + break; + case FIB_EVENT_ENTRY_DEL: + mlxsw_sp_router_fib6_del(mlxsw_sp, + fib6_work->rt_arr, + fib6_work->nrt6); + mlxsw_sp_router_fib6_work_fini(fib6_work); + break; + } + mutex_unlock(&mlxsw_sp->router->lock); + kfree(fib_work); +} + +static void mlxsw_sp_router_fibmr_event_work(struct work_struct *work) +{ + struct mlxsw_sp_fib_event_work *fib_work = + container_of(work, struct mlxsw_sp_fib_event_work, work); + struct mlxsw_sp *mlxsw_sp = fib_work->mlxsw_sp; + bool replace; + int err; + + rtnl_lock(); + mutex_lock(&mlxsw_sp->router->lock); + switch (fib_work->event) { + case FIB_EVENT_ENTRY_REPLACE: + case FIB_EVENT_ENTRY_ADD: + replace = fib_work->event == FIB_EVENT_ENTRY_REPLACE; + + err = mlxsw_sp_router_fibmr_add(mlxsw_sp, &fib_work->men_info, + replace); + if (err) + dev_warn(mlxsw_sp->bus_info->dev, "MR entry add failed.\n"); + mr_cache_put(fib_work->men_info.mfc); + break; + case FIB_EVENT_ENTRY_DEL: + mlxsw_sp_router_fibmr_del(mlxsw_sp, &fib_work->men_info); + mr_cache_put(fib_work->men_info.mfc); + break; + case FIB_EVENT_VIF_ADD: + err = mlxsw_sp_router_fibmr_vif_add(mlxsw_sp, + &fib_work->ven_info); + if (err) + dev_warn(mlxsw_sp->bus_info->dev, "MR VIF add failed.\n"); + dev_put(fib_work->ven_info.dev); + break; + case FIB_EVENT_VIF_DEL: + mlxsw_sp_router_fibmr_vif_del(mlxsw_sp, + &fib_work->ven_info); + dev_put(fib_work->ven_info.dev); + break; + } + mutex_unlock(&mlxsw_sp->router->lock); + rtnl_unlock(); + kfree(fib_work); +} + +static void mlxsw_sp_router_fib4_event(struct mlxsw_sp_fib_event_work *fib_work, + struct fib_notifier_info *info) +{ + struct fib_entry_notifier_info *fen_info; + struct fib_nh_notifier_info *fnh_info; + + switch (fib_work->event) { + case FIB_EVENT_ENTRY_REPLACE: + case FIB_EVENT_ENTRY_DEL: + fen_info = container_of(info, struct fib_entry_notifier_info, + info); + fib_work->fen_info = *fen_info; + /* Take reference on fib_info to prevent it from being + * freed while work is queued. Release it afterwards. + */ + fib_info_hold(fib_work->fen_info.fi); + break; + case FIB_EVENT_NH_ADD: + case FIB_EVENT_NH_DEL: + fnh_info = container_of(info, struct fib_nh_notifier_info, + info); + fib_work->fnh_info = *fnh_info; + fib_info_hold(fib_work->fnh_info.fib_nh->nh_parent); + break; + } +} + +static int mlxsw_sp_router_fib6_event(struct mlxsw_sp_fib_event_work *fib_work, + struct fib_notifier_info *info) +{ + struct fib6_entry_notifier_info *fen6_info; + int err; + + switch (fib_work->event) { + case FIB_EVENT_ENTRY_REPLACE: + case FIB_EVENT_ENTRY_APPEND: + case FIB_EVENT_ENTRY_DEL: + fen6_info = container_of(info, struct fib6_entry_notifier_info, + info); + err = mlxsw_sp_router_fib6_work_init(&fib_work->fib6_work, + fen6_info); + if (err) + return err; + break; + } + + return 0; +} + +static void +mlxsw_sp_router_fibmr_event(struct mlxsw_sp_fib_event_work *fib_work, + struct fib_notifier_info *info) +{ + switch (fib_work->event) { + case FIB_EVENT_ENTRY_REPLACE: + case FIB_EVENT_ENTRY_ADD: + case FIB_EVENT_ENTRY_DEL: + memcpy(&fib_work->men_info, info, sizeof(fib_work->men_info)); + mr_cache_hold(fib_work->men_info.mfc); + break; + case FIB_EVENT_VIF_ADD: + case FIB_EVENT_VIF_DEL: + memcpy(&fib_work->ven_info, info, sizeof(fib_work->ven_info)); + dev_hold(fib_work->ven_info.dev); + break; + } +} + +static int mlxsw_sp_router_fib_rule_event(unsigned long event, + struct fib_notifier_info *info, + struct mlxsw_sp *mlxsw_sp) +{ + struct netlink_ext_ack *extack = info->extack; + struct fib_rule_notifier_info *fr_info; + struct fib_rule *rule; + int err = 0; + + /* nothing to do at the moment */ + if (event == FIB_EVENT_RULE_DEL) + return 0; + + fr_info = container_of(info, struct fib_rule_notifier_info, info); + rule = fr_info->rule; + + /* Rule only affects locally generated traffic */ + if (rule->iifindex == mlxsw_sp_net(mlxsw_sp)->loopback_dev->ifindex) + return 0; + + switch (info->family) { + case AF_INET: + if (!fib4_rule_default(rule) && !rule->l3mdev) + err = -EOPNOTSUPP; + break; + case AF_INET6: + if (!fib6_rule_default(rule) && !rule->l3mdev) + err = -EOPNOTSUPP; + break; + case RTNL_FAMILY_IPMR: + if (!ipmr_rule_default(rule) && !rule->l3mdev) + err = -EOPNOTSUPP; + break; + case RTNL_FAMILY_IP6MR: + if (!ip6mr_rule_default(rule) && !rule->l3mdev) + err = -EOPNOTSUPP; + break; + } + + if (err < 0) + NL_SET_ERR_MSG_MOD(extack, "FIB rules not supported"); + + return err; +} + +/* Called with rcu_read_lock() */ +static int mlxsw_sp_router_fib_event(struct notifier_block *nb, + unsigned long event, void *ptr) +{ + struct mlxsw_sp_fib_event_work *fib_work; + struct fib_notifier_info *info = ptr; + struct mlxsw_sp_router *router; + int err; + + if ((info->family != AF_INET && info->family != AF_INET6 && + info->family != RTNL_FAMILY_IPMR && + info->family != RTNL_FAMILY_IP6MR)) + return NOTIFY_DONE; + + router = container_of(nb, struct mlxsw_sp_router, fib_nb); + + switch (event) { + case FIB_EVENT_RULE_ADD: + case FIB_EVENT_RULE_DEL: + err = mlxsw_sp_router_fib_rule_event(event, info, + router->mlxsw_sp); + return notifier_from_errno(err); + case FIB_EVENT_ENTRY_ADD: + case FIB_EVENT_ENTRY_REPLACE: + case FIB_EVENT_ENTRY_APPEND: + if (info->family == AF_INET) { + struct fib_entry_notifier_info *fen_info = ptr; + + if (fen_info->fi->fib_nh_is_v6) { + NL_SET_ERR_MSG_MOD(info->extack, "IPv6 gateway with IPv4 route is not supported"); + return notifier_from_errno(-EINVAL); + } + } + break; + } + + fib_work = kzalloc(sizeof(*fib_work), GFP_ATOMIC); + if (!fib_work) + return NOTIFY_BAD; + + fib_work->mlxsw_sp = router->mlxsw_sp; + fib_work->event = event; + + switch (info->family) { + case AF_INET: + INIT_WORK(&fib_work->work, mlxsw_sp_router_fib4_event_work); + mlxsw_sp_router_fib4_event(fib_work, info); + break; + case AF_INET6: + INIT_WORK(&fib_work->work, mlxsw_sp_router_fib6_event_work); + err = mlxsw_sp_router_fib6_event(fib_work, info); + if (err) + goto err_fib_event; + break; + case RTNL_FAMILY_IP6MR: + case RTNL_FAMILY_IPMR: + INIT_WORK(&fib_work->work, mlxsw_sp_router_fibmr_event_work); + mlxsw_sp_router_fibmr_event(fib_work, info); + break; + } + + mlxsw_core_schedule_work(&fib_work->work); + + return NOTIFY_DONE; + +err_fib_event: + kfree(fib_work); + return NOTIFY_BAD; +} + +static struct mlxsw_sp_rif * +mlxsw_sp_rif_find_by_dev(const struct mlxsw_sp *mlxsw_sp, + const struct net_device *dev) +{ + int i; + + for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); i++) + if (mlxsw_sp->router->rifs[i] && + mlxsw_sp->router->rifs[i]->dev == dev) + return mlxsw_sp->router->rifs[i]; + + return NULL; +} + +bool mlxsw_sp_rif_exists(struct mlxsw_sp *mlxsw_sp, + const struct net_device *dev) +{ + struct mlxsw_sp_rif *rif; + + mutex_lock(&mlxsw_sp->router->lock); + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, dev); + mutex_unlock(&mlxsw_sp->router->lock); + + return rif; +} + +u16 mlxsw_sp_rif_vid(struct mlxsw_sp *mlxsw_sp, const struct net_device *dev) +{ + struct mlxsw_sp_rif *rif; + u16 vid = 0; + + mutex_lock(&mlxsw_sp->router->lock); + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, dev); + if (!rif) + goto out; + + /* We only return the VID for VLAN RIFs. Otherwise we return an + * invalid value (0). + */ + if (rif->ops->type != MLXSW_SP_RIF_TYPE_VLAN) + goto out; + + vid = mlxsw_sp_fid_8021q_vid(rif->fid); + +out: + mutex_unlock(&mlxsw_sp->router->lock); + return vid; +} + +static int mlxsw_sp_router_rif_disable(struct mlxsw_sp *mlxsw_sp, u16 rif) +{ + char ritr_pl[MLXSW_REG_RITR_LEN]; + int err; + + mlxsw_reg_ritr_rif_pack(ritr_pl, rif); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); + if (err) + return err; + + mlxsw_reg_ritr_enable_set(ritr_pl, false); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); +} + +static void mlxsw_sp_router_rif_gone_sync(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_rif *rif) +{ + mlxsw_sp_router_rif_disable(mlxsw_sp, rif->rif_index); + mlxsw_sp_nexthop_rif_gone_sync(mlxsw_sp, rif); + mlxsw_sp_neigh_rif_gone_sync(mlxsw_sp, rif); +} + +static bool +mlxsw_sp_rif_should_config(struct mlxsw_sp_rif *rif, struct net_device *dev, + unsigned long event) +{ + struct inet6_dev *inet6_dev; + bool addr_list_empty = true; + struct in_device *idev; + + switch (event) { + case NETDEV_UP: + return rif == NULL; + case NETDEV_DOWN: + rcu_read_lock(); + idev = __in_dev_get_rcu(dev); + if (idev && idev->ifa_list) + addr_list_empty = false; + + inet6_dev = __in6_dev_get(dev); + if (addr_list_empty && inet6_dev && + !list_empty(&inet6_dev->addr_list)) + addr_list_empty = false; + rcu_read_unlock(); + + /* macvlans do not have a RIF, but rather piggy back on the + * RIF of their lower device. + */ + if (netif_is_macvlan(dev) && addr_list_empty) + return true; + + if (rif && addr_list_empty && + !netif_is_l3_slave(rif->dev)) + return true; + /* It is possible we already removed the RIF ourselves + * if it was assigned to a netdev that is now a bridge + * or LAG slave. + */ + return false; + } + + return false; +} + +static enum mlxsw_sp_rif_type +mlxsw_sp_dev_rif_type(const struct mlxsw_sp *mlxsw_sp, + const struct net_device *dev) +{ + enum mlxsw_sp_fid_type type; + + if (mlxsw_sp_netdev_ipip_type(mlxsw_sp, dev, NULL)) + return MLXSW_SP_RIF_TYPE_IPIP_LB; + + /* Otherwise RIF type is derived from the type of the underlying FID. */ + if (is_vlan_dev(dev) && netif_is_bridge_master(vlan_dev_real_dev(dev))) + type = MLXSW_SP_FID_TYPE_8021Q; + else if (netif_is_bridge_master(dev) && br_vlan_enabled(dev)) + type = MLXSW_SP_FID_TYPE_8021Q; + else if (netif_is_bridge_master(dev)) + type = MLXSW_SP_FID_TYPE_8021D; + else + type = MLXSW_SP_FID_TYPE_RFID; + + return mlxsw_sp_fid_type_rif_type(mlxsw_sp, type); +} + +static int mlxsw_sp_rif_index_alloc(struct mlxsw_sp *mlxsw_sp, u16 *p_rif_index) +{ + int i; + + for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); i++) { + if (!mlxsw_sp->router->rifs[i]) { + *p_rif_index = i; + return 0; + } + } + + return -ENOBUFS; +} + +static struct mlxsw_sp_rif *mlxsw_sp_rif_alloc(size_t rif_size, u16 rif_index, + u16 vr_id, + struct net_device *l3_dev) +{ + struct mlxsw_sp_rif *rif; + + rif = kzalloc(rif_size, GFP_KERNEL); + if (!rif) + return NULL; + + INIT_LIST_HEAD(&rif->nexthop_list); + INIT_LIST_HEAD(&rif->neigh_list); + if (l3_dev) { + ether_addr_copy(rif->addr, l3_dev->dev_addr); + rif->mtu = l3_dev->mtu; + rif->dev = l3_dev; + } + rif->vr_id = vr_id; + rif->rif_index = rif_index; + + return rif; +} + +struct mlxsw_sp_rif *mlxsw_sp_rif_by_index(const struct mlxsw_sp *mlxsw_sp, + u16 rif_index) +{ + return mlxsw_sp->router->rifs[rif_index]; +} + +u16 mlxsw_sp_rif_index(const struct mlxsw_sp_rif *rif) +{ + return rif->rif_index; +} + +u16 mlxsw_sp_ipip_lb_rif_index(const struct mlxsw_sp_rif_ipip_lb *lb_rif) +{ + return lb_rif->common.rif_index; +} + +u16 mlxsw_sp_ipip_lb_ul_vr_id(const struct mlxsw_sp_rif_ipip_lb *lb_rif) +{ + u32 ul_tb_id = mlxsw_sp_ipip_dev_ul_tb_id(lb_rif->common.dev); + struct mlxsw_sp_vr *ul_vr; + + ul_vr = mlxsw_sp_vr_get(lb_rif->common.mlxsw_sp, ul_tb_id, NULL); + if (WARN_ON(IS_ERR(ul_vr))) + return 0; + + return ul_vr->id; +} + +u16 mlxsw_sp_ipip_lb_ul_rif_id(const struct mlxsw_sp_rif_ipip_lb *lb_rif) +{ + return lb_rif->ul_rif_id; +} + +static bool +mlxsw_sp_router_port_l3_stats_enabled(struct mlxsw_sp_rif *rif) +{ + return mlxsw_sp_rif_counter_valid_get(rif, + MLXSW_SP_RIF_COUNTER_EGRESS) && + mlxsw_sp_rif_counter_valid_get(rif, + MLXSW_SP_RIF_COUNTER_INGRESS); +} + +static int +mlxsw_sp_router_port_l3_stats_enable(struct mlxsw_sp_rif *rif) +{ + int err; + + err = mlxsw_sp_rif_counter_alloc(rif, MLXSW_SP_RIF_COUNTER_INGRESS); + if (err) + return err; + + /* Clear stale data. */ + err = mlxsw_sp_rif_counter_fetch_clear(rif, + MLXSW_SP_RIF_COUNTER_INGRESS, + NULL); + if (err) + goto err_clear_ingress; + + err = mlxsw_sp_rif_counter_alloc(rif, MLXSW_SP_RIF_COUNTER_EGRESS); + if (err) + goto err_alloc_egress; + + /* Clear stale data. */ + err = mlxsw_sp_rif_counter_fetch_clear(rif, + MLXSW_SP_RIF_COUNTER_EGRESS, + NULL); + if (err) + goto err_clear_egress; + + return 0; + +err_clear_egress: + mlxsw_sp_rif_counter_free(rif, MLXSW_SP_RIF_COUNTER_EGRESS); +err_alloc_egress: +err_clear_ingress: + mlxsw_sp_rif_counter_free(rif, MLXSW_SP_RIF_COUNTER_INGRESS); + return err; +} + +static void +mlxsw_sp_router_port_l3_stats_disable(struct mlxsw_sp_rif *rif) +{ + mlxsw_sp_rif_counter_free(rif, MLXSW_SP_RIF_COUNTER_EGRESS); + mlxsw_sp_rif_counter_free(rif, MLXSW_SP_RIF_COUNTER_INGRESS); +} + +static void +mlxsw_sp_router_port_l3_stats_report_used(struct mlxsw_sp_rif *rif, + struct netdev_notifier_offload_xstats_info *info) +{ + if (!mlxsw_sp_router_port_l3_stats_enabled(rif)) + return; + netdev_offload_xstats_report_used(info->report_used); +} + +static int +mlxsw_sp_router_port_l3_stats_fetch(struct mlxsw_sp_rif *rif, + struct rtnl_hw_stats64 *p_stats) +{ + struct mlxsw_sp_rif_counter_set_basic ingress; + struct mlxsw_sp_rif_counter_set_basic egress; + int err; + + err = mlxsw_sp_rif_counter_fetch_clear(rif, + MLXSW_SP_RIF_COUNTER_INGRESS, + &ingress); + if (err) + return err; + + err = mlxsw_sp_rif_counter_fetch_clear(rif, + MLXSW_SP_RIF_COUNTER_EGRESS, + &egress); + if (err) + return err; + +#define MLXSW_SP_ROUTER_ALL_GOOD(SET, SFX) \ + ((SET.good_unicast_ ## SFX) + \ + (SET.good_multicast_ ## SFX) + \ + (SET.good_broadcast_ ## SFX)) + + p_stats->rx_packets = MLXSW_SP_ROUTER_ALL_GOOD(ingress, packets); + p_stats->tx_packets = MLXSW_SP_ROUTER_ALL_GOOD(egress, packets); + p_stats->rx_bytes = MLXSW_SP_ROUTER_ALL_GOOD(ingress, bytes); + p_stats->tx_bytes = MLXSW_SP_ROUTER_ALL_GOOD(egress, bytes); + p_stats->rx_errors = ingress.error_packets; + p_stats->tx_errors = egress.error_packets; + p_stats->rx_dropped = ingress.discard_packets; + p_stats->tx_dropped = egress.discard_packets; + p_stats->multicast = ingress.good_multicast_packets + + ingress.good_broadcast_packets; + +#undef MLXSW_SP_ROUTER_ALL_GOOD + + return 0; +} + +static int +mlxsw_sp_router_port_l3_stats_report_delta(struct mlxsw_sp_rif *rif, + struct netdev_notifier_offload_xstats_info *info) +{ + struct rtnl_hw_stats64 stats = {}; + int err; + + if (!mlxsw_sp_router_port_l3_stats_enabled(rif)) + return 0; + + err = mlxsw_sp_router_port_l3_stats_fetch(rif, &stats); + if (err) + return err; + + netdev_offload_xstats_report_delta(info->report_delta, &stats); + return 0; +} + +struct mlxsw_sp_router_hwstats_notify_work { + struct work_struct work; + struct net_device *dev; +}; + +static void mlxsw_sp_router_hwstats_notify_work(struct work_struct *work) +{ + struct mlxsw_sp_router_hwstats_notify_work *hws_work = + container_of(work, struct mlxsw_sp_router_hwstats_notify_work, + work); + + rtnl_lock(); + rtnl_offload_xstats_notify(hws_work->dev); + rtnl_unlock(); + dev_put(hws_work->dev); + kfree(hws_work); +} + +static void +mlxsw_sp_router_hwstats_notify_schedule(struct net_device *dev) +{ + struct mlxsw_sp_router_hwstats_notify_work *hws_work; + + /* To collect notification payload, the core ends up sending another + * notifier block message, which would deadlock on the attempt to + * acquire the router lock again. Just postpone the notification until + * later. + */ + + hws_work = kzalloc(sizeof(*hws_work), GFP_KERNEL); + if (!hws_work) + return; + + INIT_WORK(&hws_work->work, mlxsw_sp_router_hwstats_notify_work); + dev_hold(dev); + hws_work->dev = dev; + mlxsw_core_schedule_work(&hws_work->work); +} + +int mlxsw_sp_rif_dev_ifindex(const struct mlxsw_sp_rif *rif) +{ + return rif->dev->ifindex; +} + +const struct net_device *mlxsw_sp_rif_dev(const struct mlxsw_sp_rif *rif) +{ + return rif->dev; +} + +static void mlxsw_sp_rif_push_l3_stats(struct mlxsw_sp_rif *rif) +{ + struct rtnl_hw_stats64 stats = {}; + + if (!mlxsw_sp_router_port_l3_stats_fetch(rif, &stats)) + netdev_offload_xstats_push_delta(rif->dev, + NETDEV_OFFLOAD_XSTATS_TYPE_L3, + &stats); +} + +static struct mlxsw_sp_rif * +mlxsw_sp_rif_create(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_rif_params *params, + struct netlink_ext_ack *extack) +{ + u32 tb_id = l3mdev_fib_table(params->dev); + const struct mlxsw_sp_rif_ops *ops; + struct mlxsw_sp_fid *fid = NULL; + enum mlxsw_sp_rif_type type; + struct mlxsw_sp_rif *rif; + struct mlxsw_sp_vr *vr; + u16 rif_index; + int i, err; + + type = mlxsw_sp_dev_rif_type(mlxsw_sp, params->dev); + ops = mlxsw_sp->router->rif_ops_arr[type]; + + vr = mlxsw_sp_vr_get(mlxsw_sp, tb_id ? : RT_TABLE_MAIN, extack); + if (IS_ERR(vr)) + return ERR_CAST(vr); + vr->rif_count++; + + err = mlxsw_sp_rif_index_alloc(mlxsw_sp, &rif_index); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Exceeded number of supported router interfaces"); + goto err_rif_index_alloc; + } + + rif = mlxsw_sp_rif_alloc(ops->rif_size, rif_index, vr->id, params->dev); + if (!rif) { + err = -ENOMEM; + goto err_rif_alloc; + } + dev_hold(rif->dev); + mlxsw_sp->router->rifs[rif_index] = rif; + rif->mlxsw_sp = mlxsw_sp; + rif->ops = ops; + + if (ops->fid_get) { + fid = ops->fid_get(rif, extack); + if (IS_ERR(fid)) { + err = PTR_ERR(fid); + goto err_fid_get; + } + rif->fid = fid; + } + + if (ops->setup) + ops->setup(rif, params); + + err = ops->configure(rif, extack); + if (err) + goto err_configure; + + for (i = 0; i < MLXSW_SP_L3_PROTO_MAX; i++) { + err = mlxsw_sp_mr_rif_add(vr->mr_table[i], rif); + if (err) + goto err_mr_rif_add; + } + + if (netdev_offload_xstats_enabled(rif->dev, + NETDEV_OFFLOAD_XSTATS_TYPE_L3)) { + err = mlxsw_sp_router_port_l3_stats_enable(rif); + if (err) + goto err_stats_enable; + mlxsw_sp_router_hwstats_notify_schedule(rif->dev); + } else { + mlxsw_sp_rif_counters_alloc(rif); + } + + atomic_inc(&mlxsw_sp->router->rifs_count); + return rif; + +err_stats_enable: +err_mr_rif_add: + for (i--; i >= 0; i--) + mlxsw_sp_mr_rif_del(vr->mr_table[i], rif); + ops->deconfigure(rif); +err_configure: + if (fid) + mlxsw_sp_fid_put(fid); +err_fid_get: + mlxsw_sp->router->rifs[rif_index] = NULL; + dev_put(rif->dev); + kfree(rif); +err_rif_alloc: +err_rif_index_alloc: + vr->rif_count--; + mlxsw_sp_vr_put(mlxsw_sp, vr); + return ERR_PTR(err); +} + +static void mlxsw_sp_rif_destroy(struct mlxsw_sp_rif *rif) +{ + const struct mlxsw_sp_rif_ops *ops = rif->ops; + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + struct mlxsw_sp_fid *fid = rif->fid; + struct mlxsw_sp_vr *vr; + int i; + + atomic_dec(&mlxsw_sp->router->rifs_count); + mlxsw_sp_router_rif_gone_sync(mlxsw_sp, rif); + vr = &mlxsw_sp->router->vrs[rif->vr_id]; + + if (netdev_offload_xstats_enabled(rif->dev, + NETDEV_OFFLOAD_XSTATS_TYPE_L3)) { + mlxsw_sp_rif_push_l3_stats(rif); + mlxsw_sp_router_port_l3_stats_disable(rif); + mlxsw_sp_router_hwstats_notify_schedule(rif->dev); + } else { + mlxsw_sp_rif_counters_free(rif); + } + + for (i = 0; i < MLXSW_SP_L3_PROTO_MAX; i++) + mlxsw_sp_mr_rif_del(vr->mr_table[i], rif); + ops->deconfigure(rif); + if (fid) + /* Loopback RIFs are not associated with a FID. */ + mlxsw_sp_fid_put(fid); + mlxsw_sp->router->rifs[rif->rif_index] = NULL; + dev_put(rif->dev); + kfree(rif); + vr->rif_count--; + mlxsw_sp_vr_put(mlxsw_sp, vr); +} + +void mlxsw_sp_rif_destroy_by_dev(struct mlxsw_sp *mlxsw_sp, + struct net_device *dev) +{ + struct mlxsw_sp_rif *rif; + + mutex_lock(&mlxsw_sp->router->lock); + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, dev); + if (!rif) + goto out; + mlxsw_sp_rif_destroy(rif); +out: + mutex_unlock(&mlxsw_sp->router->lock); +} + +static void +mlxsw_sp_rif_subport_params_init(struct mlxsw_sp_rif_params *params, + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan) +{ + struct mlxsw_sp_port *mlxsw_sp_port = mlxsw_sp_port_vlan->mlxsw_sp_port; + + params->vid = mlxsw_sp_port_vlan->vid; + params->lag = mlxsw_sp_port->lagged; + if (params->lag) + params->lag_id = mlxsw_sp_port->lag_id; + else + params->system_port = mlxsw_sp_port->local_port; +} + +static struct mlxsw_sp_rif_subport * +mlxsw_sp_rif_subport_rif(const struct mlxsw_sp_rif *rif) +{ + return container_of(rif, struct mlxsw_sp_rif_subport, common); +} + +static struct mlxsw_sp_rif * +mlxsw_sp_rif_subport_get(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_rif_params *params, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_rif_subport *rif_subport; + struct mlxsw_sp_rif *rif; + + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, params->dev); + if (!rif) + return mlxsw_sp_rif_create(mlxsw_sp, params, extack); + + rif_subport = mlxsw_sp_rif_subport_rif(rif); + refcount_inc(&rif_subport->ref_count); + return rif; +} + +static void mlxsw_sp_rif_subport_put(struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp_rif_subport *rif_subport; + + rif_subport = mlxsw_sp_rif_subport_rif(rif); + if (!refcount_dec_and_test(&rif_subport->ref_count)) + return; + + mlxsw_sp_rif_destroy(rif); +} + +static int mlxsw_sp_rif_mac_profile_index_alloc(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_rif_mac_profile *profile, + struct netlink_ext_ack *extack) +{ + u8 max_rif_mac_profiles = mlxsw_sp->router->max_rif_mac_profile; + struct mlxsw_sp_router *router = mlxsw_sp->router; + int id; + + id = idr_alloc(&router->rif_mac_profiles_idr, profile, 0, + max_rif_mac_profiles, GFP_KERNEL); + + if (id >= 0) { + profile->id = id; + return 0; + } + + if (id == -ENOSPC) + NL_SET_ERR_MSG_MOD(extack, + "Exceeded number of supported router interface MAC profiles"); + + return id; +} + +static struct mlxsw_sp_rif_mac_profile * +mlxsw_sp_rif_mac_profile_index_free(struct mlxsw_sp *mlxsw_sp, u8 mac_profile) +{ + struct mlxsw_sp_rif_mac_profile *profile; + + profile = idr_remove(&mlxsw_sp->router->rif_mac_profiles_idr, + mac_profile); + WARN_ON(!profile); + return profile; +} + +static struct mlxsw_sp_rif_mac_profile * +mlxsw_sp_rif_mac_profile_alloc(const char *mac) +{ + struct mlxsw_sp_rif_mac_profile *profile; + + profile = kzalloc(sizeof(*profile), GFP_KERNEL); + if (!profile) + return NULL; + + ether_addr_copy(profile->mac_prefix, mac); + refcount_set(&profile->ref_count, 1); + return profile; +} + +static struct mlxsw_sp_rif_mac_profile * +mlxsw_sp_rif_mac_profile_find(const struct mlxsw_sp *mlxsw_sp, const char *mac) +{ + struct mlxsw_sp_router *router = mlxsw_sp->router; + struct mlxsw_sp_rif_mac_profile *profile; + int id; + + idr_for_each_entry(&router->rif_mac_profiles_idr, profile, id) { + if (ether_addr_equal_masked(profile->mac_prefix, mac, + mlxsw_sp->mac_mask)) + return profile; + } + + return NULL; +} + +static u64 mlxsw_sp_rif_mac_profiles_occ_get(void *priv) +{ + const struct mlxsw_sp *mlxsw_sp = priv; + + return atomic_read(&mlxsw_sp->router->rif_mac_profiles_count); +} + +static u64 mlxsw_sp_rifs_occ_get(void *priv) +{ + const struct mlxsw_sp *mlxsw_sp = priv; + + return atomic_read(&mlxsw_sp->router->rifs_count); +} + +static struct mlxsw_sp_rif_mac_profile * +mlxsw_sp_rif_mac_profile_create(struct mlxsw_sp *mlxsw_sp, const char *mac, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_rif_mac_profile *profile; + int err; + + profile = mlxsw_sp_rif_mac_profile_alloc(mac); + if (!profile) + return ERR_PTR(-ENOMEM); + + err = mlxsw_sp_rif_mac_profile_index_alloc(mlxsw_sp, profile, extack); + if (err) + goto profile_index_alloc_err; + + atomic_inc(&mlxsw_sp->router->rif_mac_profiles_count); + return profile; + +profile_index_alloc_err: + kfree(profile); + return ERR_PTR(err); +} + +static void mlxsw_sp_rif_mac_profile_destroy(struct mlxsw_sp *mlxsw_sp, + u8 mac_profile) +{ + struct mlxsw_sp_rif_mac_profile *profile; + + atomic_dec(&mlxsw_sp->router->rif_mac_profiles_count); + profile = mlxsw_sp_rif_mac_profile_index_free(mlxsw_sp, mac_profile); + kfree(profile); +} + +static int mlxsw_sp_rif_mac_profile_get(struct mlxsw_sp *mlxsw_sp, + const char *mac, u8 *p_mac_profile, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_rif_mac_profile *profile; + + profile = mlxsw_sp_rif_mac_profile_find(mlxsw_sp, mac); + if (profile) { + refcount_inc(&profile->ref_count); + goto out; + } + + profile = mlxsw_sp_rif_mac_profile_create(mlxsw_sp, mac, extack); + if (IS_ERR(profile)) + return PTR_ERR(profile); + +out: + *p_mac_profile = profile->id; + return 0; +} + +static void mlxsw_sp_rif_mac_profile_put(struct mlxsw_sp *mlxsw_sp, + u8 mac_profile) +{ + struct mlxsw_sp_rif_mac_profile *profile; + + profile = idr_find(&mlxsw_sp->router->rif_mac_profiles_idr, + mac_profile); + if (WARN_ON(!profile)) + return; + + if (!refcount_dec_and_test(&profile->ref_count)) + return; + + mlxsw_sp_rif_mac_profile_destroy(mlxsw_sp, mac_profile); +} + +static bool mlxsw_sp_rif_mac_profile_is_shared(const struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + struct mlxsw_sp_rif_mac_profile *profile; + + profile = idr_find(&mlxsw_sp->router->rif_mac_profiles_idr, + rif->mac_profile_id); + if (WARN_ON(!profile)) + return false; + + return refcount_read(&profile->ref_count) > 1; +} + +static int mlxsw_sp_rif_mac_profile_edit(struct mlxsw_sp_rif *rif, + const char *new_mac) +{ + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + struct mlxsw_sp_rif_mac_profile *profile; + + profile = idr_find(&mlxsw_sp->router->rif_mac_profiles_idr, + rif->mac_profile_id); + if (WARN_ON(!profile)) + return -EINVAL; + + ether_addr_copy(profile->mac_prefix, new_mac); + return 0; +} + +static int +mlxsw_sp_rif_mac_profile_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_rif *rif, + const char *new_mac, + struct netlink_ext_ack *extack) +{ + u8 mac_profile; + int err; + + if (!mlxsw_sp_rif_mac_profile_is_shared(rif) && + !mlxsw_sp_rif_mac_profile_find(mlxsw_sp, new_mac)) + return mlxsw_sp_rif_mac_profile_edit(rif, new_mac); + + err = mlxsw_sp_rif_mac_profile_get(mlxsw_sp, new_mac, + &mac_profile, extack); + if (err) + return err; + + mlxsw_sp_rif_mac_profile_put(mlxsw_sp, rif->mac_profile_id); + rif->mac_profile_id = mac_profile; + return 0; +} + +static int +__mlxsw_sp_port_vlan_router_join(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan, + struct net_device *l3_dev, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port *mlxsw_sp_port = mlxsw_sp_port_vlan->mlxsw_sp_port; + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_rif_params params = { + .dev = l3_dev, + }; + u16 vid = mlxsw_sp_port_vlan->vid; + struct mlxsw_sp_rif *rif; + struct mlxsw_sp_fid *fid; + int err; + + mlxsw_sp_rif_subport_params_init(¶ms, mlxsw_sp_port_vlan); + rif = mlxsw_sp_rif_subport_get(mlxsw_sp, ¶ms, extack); + if (IS_ERR(rif)) + return PTR_ERR(rif); + + /* FID was already created, just take a reference */ + fid = rif->ops->fid_get(rif, extack); + err = mlxsw_sp_fid_port_vid_map(fid, mlxsw_sp_port, vid); + if (err) + goto err_fid_port_vid_map; + + err = mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, vid, false); + if (err) + goto err_port_vid_learning_set; + + err = mlxsw_sp_port_vid_stp_set(mlxsw_sp_port, vid, + BR_STATE_FORWARDING); + if (err) + goto err_port_vid_stp_set; + + mlxsw_sp_port_vlan->fid = fid; + + return 0; + +err_port_vid_stp_set: + mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, vid, true); +err_port_vid_learning_set: + mlxsw_sp_fid_port_vid_unmap(fid, mlxsw_sp_port, vid); +err_fid_port_vid_map: + mlxsw_sp_fid_put(fid); + mlxsw_sp_rif_subport_put(rif); + return err; +} + +static void +__mlxsw_sp_port_vlan_router_leave(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan) +{ + struct mlxsw_sp_port *mlxsw_sp_port = mlxsw_sp_port_vlan->mlxsw_sp_port; + struct mlxsw_sp_fid *fid = mlxsw_sp_port_vlan->fid; + struct mlxsw_sp_rif *rif = mlxsw_sp_fid_rif(fid); + u16 vid = mlxsw_sp_port_vlan->vid; + + if (WARN_ON(mlxsw_sp_fid_type(fid) != MLXSW_SP_FID_TYPE_RFID)) + return; + + mlxsw_sp_port_vlan->fid = NULL; + mlxsw_sp_port_vid_stp_set(mlxsw_sp_port, vid, BR_STATE_BLOCKING); + mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, vid, true); + mlxsw_sp_fid_port_vid_unmap(fid, mlxsw_sp_port, vid); + mlxsw_sp_fid_put(fid); + mlxsw_sp_rif_subport_put(rif); +} + +int +mlxsw_sp_port_vlan_router_join(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan, + struct net_device *l3_dev, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port_vlan->mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_rif *rif; + int err = 0; + + mutex_lock(&mlxsw_sp->router->lock); + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, l3_dev); + if (!rif) + goto out; + + err = __mlxsw_sp_port_vlan_router_join(mlxsw_sp_port_vlan, l3_dev, + extack); +out: + mutex_unlock(&mlxsw_sp->router->lock); + return err; +} + +void +mlxsw_sp_port_vlan_router_leave(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port_vlan->mlxsw_sp_port->mlxsw_sp; + + mutex_lock(&mlxsw_sp->router->lock); + __mlxsw_sp_port_vlan_router_leave(mlxsw_sp_port_vlan); + mutex_unlock(&mlxsw_sp->router->lock); +} + +static int mlxsw_sp_inetaddr_port_vlan_event(struct net_device *l3_dev, + struct net_device *port_dev, + unsigned long event, u16 vid, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(port_dev); + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_find_by_vid(mlxsw_sp_port, vid); + if (WARN_ON(!mlxsw_sp_port_vlan)) + return -EINVAL; + + switch (event) { + case NETDEV_UP: + return __mlxsw_sp_port_vlan_router_join(mlxsw_sp_port_vlan, + l3_dev, extack); + case NETDEV_DOWN: + __mlxsw_sp_port_vlan_router_leave(mlxsw_sp_port_vlan); + break; + } + + return 0; +} + +static int mlxsw_sp_inetaddr_port_event(struct net_device *port_dev, + unsigned long event, + struct netlink_ext_ack *extack) +{ + if (netif_is_any_bridge_port(port_dev) || netif_is_lag_port(port_dev)) + return 0; + + return mlxsw_sp_inetaddr_port_vlan_event(port_dev, port_dev, event, + MLXSW_SP_DEFAULT_VID, extack); +} + +static int __mlxsw_sp_inetaddr_lag_event(struct net_device *l3_dev, + struct net_device *lag_dev, + unsigned long event, u16 vid, + struct netlink_ext_ack *extack) +{ + struct net_device *port_dev; + struct list_head *iter; + int err; + + netdev_for_each_lower_dev(lag_dev, port_dev, iter) { + if (mlxsw_sp_port_dev_check(port_dev)) { + err = mlxsw_sp_inetaddr_port_vlan_event(l3_dev, + port_dev, + event, vid, + extack); + if (err) + return err; + } + } + + return 0; +} + +static int mlxsw_sp_inetaddr_lag_event(struct net_device *lag_dev, + unsigned long event, + struct netlink_ext_ack *extack) +{ + if (netif_is_bridge_port(lag_dev)) + return 0; + + return __mlxsw_sp_inetaddr_lag_event(lag_dev, lag_dev, event, + MLXSW_SP_DEFAULT_VID, extack); +} + +static int mlxsw_sp_inetaddr_bridge_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *l3_dev, + unsigned long event, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_rif_params params = { + .dev = l3_dev, + }; + struct mlxsw_sp_rif *rif; + + switch (event) { + case NETDEV_UP: + if (netif_is_bridge_master(l3_dev) && br_vlan_enabled(l3_dev)) { + u16 proto; + + br_vlan_get_proto(l3_dev, &proto); + if (proto == ETH_P_8021AD) { + NL_SET_ERR_MSG_MOD(extack, "Adding an IP address to 802.1ad bridge is not supported"); + return -EOPNOTSUPP; + } + } + rif = mlxsw_sp_rif_create(mlxsw_sp, ¶ms, extack); + if (IS_ERR(rif)) + return PTR_ERR(rif); + break; + case NETDEV_DOWN: + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, l3_dev); + mlxsw_sp_rif_destroy(rif); + break; + } + + return 0; +} + +static int mlxsw_sp_inetaddr_vlan_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *vlan_dev, + unsigned long event, + struct netlink_ext_ack *extack) +{ + struct net_device *real_dev = vlan_dev_real_dev(vlan_dev); + u16 vid = vlan_dev_vlan_id(vlan_dev); + + if (netif_is_bridge_port(vlan_dev)) + return 0; + + if (mlxsw_sp_port_dev_check(real_dev)) + return mlxsw_sp_inetaddr_port_vlan_event(vlan_dev, real_dev, + event, vid, extack); + else if (netif_is_lag_master(real_dev)) + return __mlxsw_sp_inetaddr_lag_event(vlan_dev, real_dev, event, + vid, extack); + else if (netif_is_bridge_master(real_dev) && br_vlan_enabled(real_dev)) + return mlxsw_sp_inetaddr_bridge_event(mlxsw_sp, vlan_dev, event, + extack); + + return 0; +} + +static bool mlxsw_sp_rif_macvlan_is_vrrp4(const u8 *mac) +{ + u8 vrrp4[ETH_ALEN] = { 0x00, 0x00, 0x5e, 0x00, 0x01, 0x00 }; + u8 mask[ETH_ALEN] = { 0xff, 0xff, 0xff, 0xff, 0xff, 0x00 }; + + return ether_addr_equal_masked(mac, vrrp4, mask); +} + +static bool mlxsw_sp_rif_macvlan_is_vrrp6(const u8 *mac) +{ + u8 vrrp6[ETH_ALEN] = { 0x00, 0x00, 0x5e, 0x00, 0x02, 0x00 }; + u8 mask[ETH_ALEN] = { 0xff, 0xff, 0xff, 0xff, 0xff, 0x00 }; + + return ether_addr_equal_masked(mac, vrrp6, mask); +} + +static int mlxsw_sp_rif_vrrp_op(struct mlxsw_sp *mlxsw_sp, u16 rif_index, + const u8 *mac, bool adding) +{ + char ritr_pl[MLXSW_REG_RITR_LEN]; + u8 vrrp_id = adding ? mac[5] : 0; + int err; + + if (!mlxsw_sp_rif_macvlan_is_vrrp4(mac) && + !mlxsw_sp_rif_macvlan_is_vrrp6(mac)) + return 0; + + mlxsw_reg_ritr_rif_pack(ritr_pl, rif_index); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); + if (err) + return err; + + if (mlxsw_sp_rif_macvlan_is_vrrp4(mac)) + mlxsw_reg_ritr_if_vrrp_id_ipv4_set(ritr_pl, vrrp_id); + else + mlxsw_reg_ritr_if_vrrp_id_ipv6_set(ritr_pl, vrrp_id); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); +} + +static int mlxsw_sp_rif_macvlan_add(struct mlxsw_sp *mlxsw_sp, + const struct net_device *macvlan_dev, + struct netlink_ext_ack *extack) +{ + struct macvlan_dev *vlan = netdev_priv(macvlan_dev); + struct mlxsw_sp_rif *rif; + int err; + + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, vlan->lowerdev); + if (!rif) { + NL_SET_ERR_MSG_MOD(extack, "macvlan is only supported on top of router interfaces"); + return -EOPNOTSUPP; + } + + err = mlxsw_sp_rif_fdb_op(mlxsw_sp, macvlan_dev->dev_addr, + mlxsw_sp_fid_index(rif->fid), true); + if (err) + return err; + + err = mlxsw_sp_rif_vrrp_op(mlxsw_sp, rif->rif_index, + macvlan_dev->dev_addr, true); + if (err) + goto err_rif_vrrp_add; + + /* Make sure the bridge driver does not have this MAC pointing at + * some other port. + */ + if (rif->ops->fdb_del) + rif->ops->fdb_del(rif, macvlan_dev->dev_addr); + + return 0; + +err_rif_vrrp_add: + mlxsw_sp_rif_fdb_op(mlxsw_sp, macvlan_dev->dev_addr, + mlxsw_sp_fid_index(rif->fid), false); + return err; +} + +static void __mlxsw_sp_rif_macvlan_del(struct mlxsw_sp *mlxsw_sp, + const struct net_device *macvlan_dev) +{ + struct macvlan_dev *vlan = netdev_priv(macvlan_dev); + struct mlxsw_sp_rif *rif; + + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, vlan->lowerdev); + /* If we do not have a RIF, then we already took care of + * removing the macvlan's MAC during RIF deletion. + */ + if (!rif) + return; + mlxsw_sp_rif_vrrp_op(mlxsw_sp, rif->rif_index, macvlan_dev->dev_addr, + false); + mlxsw_sp_rif_fdb_op(mlxsw_sp, macvlan_dev->dev_addr, + mlxsw_sp_fid_index(rif->fid), false); +} + +void mlxsw_sp_rif_macvlan_del(struct mlxsw_sp *mlxsw_sp, + const struct net_device *macvlan_dev) +{ + mutex_lock(&mlxsw_sp->router->lock); + __mlxsw_sp_rif_macvlan_del(mlxsw_sp, macvlan_dev); + mutex_unlock(&mlxsw_sp->router->lock); +} + +static int mlxsw_sp_inetaddr_macvlan_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *macvlan_dev, + unsigned long event, + struct netlink_ext_ack *extack) +{ + switch (event) { + case NETDEV_UP: + return mlxsw_sp_rif_macvlan_add(mlxsw_sp, macvlan_dev, extack); + case NETDEV_DOWN: + __mlxsw_sp_rif_macvlan_del(mlxsw_sp, macvlan_dev); + break; + } + + return 0; +} + +static int __mlxsw_sp_inetaddr_event(struct mlxsw_sp *mlxsw_sp, + struct net_device *dev, + unsigned long event, + struct netlink_ext_ack *extack) +{ + if (mlxsw_sp_port_dev_check(dev)) + return mlxsw_sp_inetaddr_port_event(dev, event, extack); + else if (netif_is_lag_master(dev)) + return mlxsw_sp_inetaddr_lag_event(dev, event, extack); + else if (netif_is_bridge_master(dev)) + return mlxsw_sp_inetaddr_bridge_event(mlxsw_sp, dev, event, + extack); + else if (is_vlan_dev(dev)) + return mlxsw_sp_inetaddr_vlan_event(mlxsw_sp, dev, event, + extack); + else if (netif_is_macvlan(dev)) + return mlxsw_sp_inetaddr_macvlan_event(mlxsw_sp, dev, event, + extack); + else + return 0; +} + +static int mlxsw_sp_inetaddr_event(struct notifier_block *nb, + unsigned long event, void *ptr) +{ + struct in_ifaddr *ifa = (struct in_ifaddr *) ptr; + struct net_device *dev = ifa->ifa_dev->dev; + struct mlxsw_sp_router *router; + struct mlxsw_sp_rif *rif; + int err = 0; + + /* NETDEV_UP event is handled by mlxsw_sp_inetaddr_valid_event */ + if (event == NETDEV_UP) + return NOTIFY_DONE; + + router = container_of(nb, struct mlxsw_sp_router, inetaddr_nb); + mutex_lock(&router->lock); + rif = mlxsw_sp_rif_find_by_dev(router->mlxsw_sp, dev); + if (!mlxsw_sp_rif_should_config(rif, dev, event)) + goto out; + + err = __mlxsw_sp_inetaddr_event(router->mlxsw_sp, dev, event, NULL); +out: + mutex_unlock(&router->lock); + return notifier_from_errno(err); +} + +int mlxsw_sp_inetaddr_valid_event(struct notifier_block *unused, + unsigned long event, void *ptr) +{ + struct in_validator_info *ivi = (struct in_validator_info *) ptr; + struct net_device *dev = ivi->ivi_dev->dev; + struct mlxsw_sp *mlxsw_sp; + struct mlxsw_sp_rif *rif; + int err = 0; + + mlxsw_sp = mlxsw_sp_lower_get(dev); + if (!mlxsw_sp) + return NOTIFY_DONE; + + mutex_lock(&mlxsw_sp->router->lock); + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, dev); + if (!mlxsw_sp_rif_should_config(rif, dev, event)) + goto out; + + err = __mlxsw_sp_inetaddr_event(mlxsw_sp, dev, event, ivi->extack); +out: + mutex_unlock(&mlxsw_sp->router->lock); + return notifier_from_errno(err); +} + +struct mlxsw_sp_inet6addr_event_work { + struct work_struct work; + struct mlxsw_sp *mlxsw_sp; + struct net_device *dev; + unsigned long event; +}; + +static void mlxsw_sp_inet6addr_event_work(struct work_struct *work) +{ + struct mlxsw_sp_inet6addr_event_work *inet6addr_work = + container_of(work, struct mlxsw_sp_inet6addr_event_work, work); + struct mlxsw_sp *mlxsw_sp = inet6addr_work->mlxsw_sp; + struct net_device *dev = inet6addr_work->dev; + unsigned long event = inet6addr_work->event; + struct mlxsw_sp_rif *rif; + + rtnl_lock(); + mutex_lock(&mlxsw_sp->router->lock); + + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, dev); + if (!mlxsw_sp_rif_should_config(rif, dev, event)) + goto out; + + __mlxsw_sp_inetaddr_event(mlxsw_sp, dev, event, NULL); +out: + mutex_unlock(&mlxsw_sp->router->lock); + rtnl_unlock(); + dev_put(dev); + kfree(inet6addr_work); +} + +/* Called with rcu_read_lock() */ +static int mlxsw_sp_inet6addr_event(struct notifier_block *nb, + unsigned long event, void *ptr) +{ + struct inet6_ifaddr *if6 = (struct inet6_ifaddr *) ptr; + struct mlxsw_sp_inet6addr_event_work *inet6addr_work; + struct net_device *dev = if6->idev->dev; + struct mlxsw_sp_router *router; + + /* NETDEV_UP event is handled by mlxsw_sp_inet6addr_valid_event */ + if (event == NETDEV_UP) + return NOTIFY_DONE; + + inet6addr_work = kzalloc(sizeof(*inet6addr_work), GFP_ATOMIC); + if (!inet6addr_work) + return NOTIFY_BAD; + + router = container_of(nb, struct mlxsw_sp_router, inet6addr_nb); + INIT_WORK(&inet6addr_work->work, mlxsw_sp_inet6addr_event_work); + inet6addr_work->mlxsw_sp = router->mlxsw_sp; + inet6addr_work->dev = dev; + inet6addr_work->event = event; + dev_hold(dev); + mlxsw_core_schedule_work(&inet6addr_work->work); + + return NOTIFY_DONE; +} + +int mlxsw_sp_inet6addr_valid_event(struct notifier_block *unused, + unsigned long event, void *ptr) +{ + struct in6_validator_info *i6vi = (struct in6_validator_info *) ptr; + struct net_device *dev = i6vi->i6vi_dev->dev; + struct mlxsw_sp *mlxsw_sp; + struct mlxsw_sp_rif *rif; + int err = 0; + + mlxsw_sp = mlxsw_sp_lower_get(dev); + if (!mlxsw_sp) + return NOTIFY_DONE; + + mutex_lock(&mlxsw_sp->router->lock); + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, dev); + if (!mlxsw_sp_rif_should_config(rif, dev, event)) + goto out; + + err = __mlxsw_sp_inetaddr_event(mlxsw_sp, dev, event, i6vi->extack); +out: + mutex_unlock(&mlxsw_sp->router->lock); + return notifier_from_errno(err); +} + +static int mlxsw_sp_rif_edit(struct mlxsw_sp *mlxsw_sp, u16 rif_index, + const char *mac, int mtu, u8 mac_profile) +{ + char ritr_pl[MLXSW_REG_RITR_LEN]; + int err; + + mlxsw_reg_ritr_rif_pack(ritr_pl, rif_index); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); + if (err) + return err; + + mlxsw_reg_ritr_mtu_set(ritr_pl, mtu); + mlxsw_reg_ritr_if_mac_memcpy_to(ritr_pl, mac); + mlxsw_reg_ritr_if_mac_profile_id_set(ritr_pl, mac_profile); + mlxsw_reg_ritr_op_set(ritr_pl, MLXSW_REG_RITR_RIF_CREATE); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); +} + +static int +mlxsw_sp_router_port_change_event(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_rif *rif, + struct netlink_ext_ack *extack) +{ + struct net_device *dev = rif->dev; + u8 old_mac_profile; + u16 fid_index; + int err; + + fid_index = mlxsw_sp_fid_index(rif->fid); + + err = mlxsw_sp_rif_fdb_op(mlxsw_sp, rif->addr, fid_index, false); + if (err) + return err; + + old_mac_profile = rif->mac_profile_id; + err = mlxsw_sp_rif_mac_profile_replace(mlxsw_sp, rif, dev->dev_addr, + extack); + if (err) + goto err_rif_mac_profile_replace; + + err = mlxsw_sp_rif_edit(mlxsw_sp, rif->rif_index, dev->dev_addr, + dev->mtu, rif->mac_profile_id); + if (err) + goto err_rif_edit; + + err = mlxsw_sp_rif_fdb_op(mlxsw_sp, dev->dev_addr, fid_index, true); + if (err) + goto err_rif_fdb_op; + + if (rif->mtu != dev->mtu) { + struct mlxsw_sp_vr *vr; + int i; + + /* The RIF is relevant only to its mr_table instance, as unlike + * unicast routing, in multicast routing a RIF cannot be shared + * between several multicast routing tables. + */ + vr = &mlxsw_sp->router->vrs[rif->vr_id]; + for (i = 0; i < MLXSW_SP_L3_PROTO_MAX; i++) + mlxsw_sp_mr_rif_mtu_update(vr->mr_table[i], + rif, dev->mtu); + } + + ether_addr_copy(rif->addr, dev->dev_addr); + rif->mtu = dev->mtu; + + netdev_dbg(dev, "Updated RIF=%d\n", rif->rif_index); + + return 0; + +err_rif_fdb_op: + mlxsw_sp_rif_edit(mlxsw_sp, rif->rif_index, rif->addr, rif->mtu, + old_mac_profile); +err_rif_edit: + mlxsw_sp_rif_mac_profile_replace(mlxsw_sp, rif, rif->addr, extack); +err_rif_mac_profile_replace: + mlxsw_sp_rif_fdb_op(mlxsw_sp, rif->addr, fid_index, true); + return err; +} + +static int mlxsw_sp_router_port_pre_changeaddr_event(struct mlxsw_sp_rif *rif, + struct netdev_notifier_pre_changeaddr_info *info) +{ + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + struct mlxsw_sp_rif_mac_profile *profile; + struct netlink_ext_ack *extack; + u8 max_rif_mac_profiles; + u64 occ; + + extack = netdev_notifier_info_to_extack(&info->info); + + profile = mlxsw_sp_rif_mac_profile_find(mlxsw_sp, info->dev_addr); + if (profile) + return 0; + + max_rif_mac_profiles = mlxsw_sp->router->max_rif_mac_profile; + occ = mlxsw_sp_rif_mac_profiles_occ_get(mlxsw_sp); + if (occ < max_rif_mac_profiles) + return 0; + + if (!mlxsw_sp_rif_mac_profile_is_shared(rif)) + return 0; + + NL_SET_ERR_MSG_MOD(extack, "Exceeded number of supported router interface MAC profiles"); + return -ENOBUFS; +} + +static bool mlxsw_sp_is_offload_xstats_event(unsigned long event) +{ + switch (event) { + case NETDEV_OFFLOAD_XSTATS_ENABLE: + case NETDEV_OFFLOAD_XSTATS_DISABLE: + case NETDEV_OFFLOAD_XSTATS_REPORT_USED: + case NETDEV_OFFLOAD_XSTATS_REPORT_DELTA: + return true; + } + + return false; +} + +static int +mlxsw_sp_router_port_offload_xstats_cmd(struct mlxsw_sp_rif *rif, + unsigned long event, + struct netdev_notifier_offload_xstats_info *info) +{ + switch (info->type) { + case NETDEV_OFFLOAD_XSTATS_TYPE_L3: + break; + default: + return 0; + } + + switch (event) { + case NETDEV_OFFLOAD_XSTATS_ENABLE: + return mlxsw_sp_router_port_l3_stats_enable(rif); + case NETDEV_OFFLOAD_XSTATS_DISABLE: + mlxsw_sp_router_port_l3_stats_disable(rif); + return 0; + case NETDEV_OFFLOAD_XSTATS_REPORT_USED: + mlxsw_sp_router_port_l3_stats_report_used(rif, info); + return 0; + case NETDEV_OFFLOAD_XSTATS_REPORT_DELTA: + return mlxsw_sp_router_port_l3_stats_report_delta(rif, info); + } + + WARN_ON_ONCE(1); + return 0; +} + +static int +mlxsw_sp_netdevice_offload_xstats_cmd(struct mlxsw_sp *mlxsw_sp, + struct net_device *dev, + unsigned long event, + struct netdev_notifier_offload_xstats_info *info) +{ + struct mlxsw_sp_rif *rif; + + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, dev); + if (!rif) + return 0; + + return mlxsw_sp_router_port_offload_xstats_cmd(rif, event, info); +} + +static bool mlxsw_sp_is_router_event(unsigned long event) +{ + switch (event) { + case NETDEV_PRE_CHANGEADDR: + case NETDEV_CHANGEADDR: + case NETDEV_CHANGEMTU: + return true; + default: + return false; + } +} + +static int mlxsw_sp_netdevice_router_port_event(struct net_device *dev, + unsigned long event, void *ptr) +{ + struct netlink_ext_ack *extack = netdev_notifier_info_to_extack(ptr); + struct mlxsw_sp *mlxsw_sp; + struct mlxsw_sp_rif *rif; + + mlxsw_sp = mlxsw_sp_lower_get(dev); + if (!mlxsw_sp) + return 0; + + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, dev); + if (!rif) + return 0; + + switch (event) { + case NETDEV_CHANGEMTU: + case NETDEV_CHANGEADDR: + return mlxsw_sp_router_port_change_event(mlxsw_sp, rif, extack); + case NETDEV_PRE_CHANGEADDR: + return mlxsw_sp_router_port_pre_changeaddr_event(rif, ptr); + default: + WARN_ON_ONCE(1); + break; + } + + return 0; +} + +static int mlxsw_sp_port_vrf_join(struct mlxsw_sp *mlxsw_sp, + struct net_device *l3_dev, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_rif *rif; + + /* If netdev is already associated with a RIF, then we need to + * destroy it and create a new one with the new virtual router ID. + */ + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, l3_dev); + if (rif) + __mlxsw_sp_inetaddr_event(mlxsw_sp, l3_dev, NETDEV_DOWN, + extack); + + return __mlxsw_sp_inetaddr_event(mlxsw_sp, l3_dev, NETDEV_UP, extack); +} + +static void mlxsw_sp_port_vrf_leave(struct mlxsw_sp *mlxsw_sp, + struct net_device *l3_dev) +{ + struct mlxsw_sp_rif *rif; + + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, l3_dev); + if (!rif) + return; + __mlxsw_sp_inetaddr_event(mlxsw_sp, l3_dev, NETDEV_DOWN, NULL); +} + +static bool mlxsw_sp_is_vrf_event(unsigned long event, void *ptr) +{ + struct netdev_notifier_changeupper_info *info = ptr; + + if (event != NETDEV_PRECHANGEUPPER && event != NETDEV_CHANGEUPPER) + return false; + return netif_is_l3_master(info->upper_dev); +} + +static int +mlxsw_sp_netdevice_vrf_event(struct net_device *l3_dev, unsigned long event, + struct netdev_notifier_changeupper_info *info) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_lower_get(l3_dev); + int err = 0; + + /* We do not create a RIF for a macvlan, but only use it to + * direct more MAC addresses to the router. + */ + if (!mlxsw_sp || netif_is_macvlan(l3_dev)) + return 0; + + switch (event) { + case NETDEV_PRECHANGEUPPER: + break; + case NETDEV_CHANGEUPPER: + if (info->linking) { + struct netlink_ext_ack *extack; + + extack = netdev_notifier_info_to_extack(&info->info); + err = mlxsw_sp_port_vrf_join(mlxsw_sp, l3_dev, extack); + } else { + mlxsw_sp_port_vrf_leave(mlxsw_sp, l3_dev); + } + break; + } + + return err; +} + +static int mlxsw_sp_router_netdevice_event(struct notifier_block *nb, + unsigned long event, void *ptr) +{ + struct net_device *dev = netdev_notifier_info_to_dev(ptr); + struct mlxsw_sp_router *router; + struct mlxsw_sp *mlxsw_sp; + int err = 0; + + router = container_of(nb, struct mlxsw_sp_router, netdevice_nb); + mlxsw_sp = router->mlxsw_sp; + + mutex_lock(&mlxsw_sp->router->lock); + + if (mlxsw_sp_is_offload_xstats_event(event)) + err = mlxsw_sp_netdevice_offload_xstats_cmd(mlxsw_sp, dev, + event, ptr); + else if (mlxsw_sp_netdev_is_ipip_ol(mlxsw_sp, dev)) + err = mlxsw_sp_netdevice_ipip_ol_event(mlxsw_sp, dev, + event, ptr); + else if (mlxsw_sp_netdev_is_ipip_ul(mlxsw_sp, dev)) + err = mlxsw_sp_netdevice_ipip_ul_event(mlxsw_sp, dev, + event, ptr); + else if (mlxsw_sp_is_router_event(event)) + err = mlxsw_sp_netdevice_router_port_event(dev, event, ptr); + else if (mlxsw_sp_is_vrf_event(event, ptr)) + err = mlxsw_sp_netdevice_vrf_event(dev, event, ptr); + + mutex_unlock(&mlxsw_sp->router->lock); + + return notifier_from_errno(err); +} + +static int __mlxsw_sp_rif_macvlan_flush(struct net_device *dev, + struct netdev_nested_priv *priv) +{ + struct mlxsw_sp_rif *rif = (struct mlxsw_sp_rif *)priv->data; + + if (!netif_is_macvlan(dev)) + return 0; + + return mlxsw_sp_rif_fdb_op(rif->mlxsw_sp, dev->dev_addr, + mlxsw_sp_fid_index(rif->fid), false); +} + +static int mlxsw_sp_rif_macvlan_flush(struct mlxsw_sp_rif *rif) +{ + struct netdev_nested_priv priv = { + .data = (void *)rif, + }; + + if (!netif_is_macvlan_port(rif->dev)) + return 0; + + netdev_warn(rif->dev, "Router interface is deleted. Upper macvlans will not work\n"); + return netdev_walk_all_upper_dev_rcu(rif->dev, + __mlxsw_sp_rif_macvlan_flush, &priv); +} + +static void mlxsw_sp_rif_subport_setup(struct mlxsw_sp_rif *rif, + const struct mlxsw_sp_rif_params *params) +{ + struct mlxsw_sp_rif_subport *rif_subport; + + rif_subport = mlxsw_sp_rif_subport_rif(rif); + refcount_set(&rif_subport->ref_count, 1); + rif_subport->vid = params->vid; + rif_subport->lag = params->lag; + if (params->lag) + rif_subport->lag_id = params->lag_id; + else + rif_subport->system_port = params->system_port; +} + +static int mlxsw_sp_rif_subport_op(struct mlxsw_sp_rif *rif, bool enable) +{ + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + struct mlxsw_sp_rif_subport *rif_subport; + char ritr_pl[MLXSW_REG_RITR_LEN]; + u16 efid; + + rif_subport = mlxsw_sp_rif_subport_rif(rif); + mlxsw_reg_ritr_pack(ritr_pl, enable, MLXSW_REG_RITR_SP_IF, + rif->rif_index, rif->vr_id, rif->dev->mtu); + mlxsw_reg_ritr_mac_pack(ritr_pl, rif->dev->dev_addr); + mlxsw_reg_ritr_if_mac_profile_id_set(ritr_pl, rif->mac_profile_id); + efid = mlxsw_sp_fid_index(rif->fid); + mlxsw_reg_ritr_sp_if_pack(ritr_pl, rif_subport->lag, + rif_subport->lag ? rif_subport->lag_id : + rif_subport->system_port, + efid, 0); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); +} + +static int mlxsw_sp_rif_subport_configure(struct mlxsw_sp_rif *rif, + struct netlink_ext_ack *extack) +{ + u8 mac_profile; + int err; + + err = mlxsw_sp_rif_mac_profile_get(rif->mlxsw_sp, rif->addr, + &mac_profile, extack); + if (err) + return err; + rif->mac_profile_id = mac_profile; + + err = mlxsw_sp_rif_subport_op(rif, true); + if (err) + goto err_rif_subport_op; + + err = mlxsw_sp_rif_fdb_op(rif->mlxsw_sp, rif->dev->dev_addr, + mlxsw_sp_fid_index(rif->fid), true); + if (err) + goto err_rif_fdb_op; + + err = mlxsw_sp_fid_rif_set(rif->fid, rif); + if (err) + goto err_fid_rif_set; + + return 0; + +err_fid_rif_set: + mlxsw_sp_rif_fdb_op(rif->mlxsw_sp, rif->dev->dev_addr, + mlxsw_sp_fid_index(rif->fid), false); +err_rif_fdb_op: + mlxsw_sp_rif_subport_op(rif, false); +err_rif_subport_op: + mlxsw_sp_rif_mac_profile_put(rif->mlxsw_sp, mac_profile); + return err; +} + +static void mlxsw_sp_rif_subport_deconfigure(struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp_fid *fid = rif->fid; + + mlxsw_sp_fid_rif_unset(fid); + mlxsw_sp_rif_fdb_op(rif->mlxsw_sp, rif->dev->dev_addr, + mlxsw_sp_fid_index(fid), false); + mlxsw_sp_rif_macvlan_flush(rif); + mlxsw_sp_rif_subport_op(rif, false); + mlxsw_sp_rif_mac_profile_put(rif->mlxsw_sp, rif->mac_profile_id); +} + +static struct mlxsw_sp_fid * +mlxsw_sp_rif_subport_fid_get(struct mlxsw_sp_rif *rif, + struct netlink_ext_ack *extack) +{ + return mlxsw_sp_fid_rfid_get(rif->mlxsw_sp, rif->rif_index); +} + +static const struct mlxsw_sp_rif_ops mlxsw_sp_rif_subport_ops = { + .type = MLXSW_SP_RIF_TYPE_SUBPORT, + .rif_size = sizeof(struct mlxsw_sp_rif_subport), + .setup = mlxsw_sp_rif_subport_setup, + .configure = mlxsw_sp_rif_subport_configure, + .deconfigure = mlxsw_sp_rif_subport_deconfigure, + .fid_get = mlxsw_sp_rif_subport_fid_get, +}; + +static int mlxsw_sp_rif_fid_op(struct mlxsw_sp_rif *rif, u16 fid, bool enable) +{ + enum mlxsw_reg_ritr_if_type type = MLXSW_REG_RITR_FID_IF; + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + char ritr_pl[MLXSW_REG_RITR_LEN]; + + mlxsw_reg_ritr_pack(ritr_pl, enable, type, rif->rif_index, rif->vr_id, + rif->dev->mtu); + mlxsw_reg_ritr_mac_pack(ritr_pl, rif->dev->dev_addr); + mlxsw_reg_ritr_if_mac_profile_id_set(ritr_pl, rif->mac_profile_id); + mlxsw_reg_ritr_fid_if_fid_set(ritr_pl, fid); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); +} + +u16 mlxsw_sp_router_port(const struct mlxsw_sp *mlxsw_sp) +{ + return mlxsw_core_max_ports(mlxsw_sp->core) + 1; +} + +static int mlxsw_sp_rif_fid_configure(struct mlxsw_sp_rif *rif, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + u16 fid_index = mlxsw_sp_fid_index(rif->fid); + u8 mac_profile; + int err; + + err = mlxsw_sp_rif_mac_profile_get(mlxsw_sp, rif->addr, + &mac_profile, extack); + if (err) + return err; + rif->mac_profile_id = mac_profile; + + err = mlxsw_sp_rif_fid_op(rif, fid_index, true); + if (err) + goto err_rif_fid_op; + + err = mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_MC, + mlxsw_sp_router_port(mlxsw_sp), true); + if (err) + goto err_fid_mc_flood_set; + + err = mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_BC, + mlxsw_sp_router_port(mlxsw_sp), true); + if (err) + goto err_fid_bc_flood_set; + + err = mlxsw_sp_rif_fdb_op(rif->mlxsw_sp, rif->dev->dev_addr, + mlxsw_sp_fid_index(rif->fid), true); + if (err) + goto err_rif_fdb_op; + + err = mlxsw_sp_fid_rif_set(rif->fid, rif); + if (err) + goto err_fid_rif_set; + + return 0; + +err_fid_rif_set: + mlxsw_sp_rif_fdb_op(rif->mlxsw_sp, rif->dev->dev_addr, + mlxsw_sp_fid_index(rif->fid), false); +err_rif_fdb_op: + mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_BC, + mlxsw_sp_router_port(mlxsw_sp), false); +err_fid_bc_flood_set: + mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_MC, + mlxsw_sp_router_port(mlxsw_sp), false); +err_fid_mc_flood_set: + mlxsw_sp_rif_fid_op(rif, fid_index, false); +err_rif_fid_op: + mlxsw_sp_rif_mac_profile_put(mlxsw_sp, mac_profile); + return err; +} + +static void mlxsw_sp_rif_fid_deconfigure(struct mlxsw_sp_rif *rif) +{ + u16 fid_index = mlxsw_sp_fid_index(rif->fid); + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + struct mlxsw_sp_fid *fid = rif->fid; + + mlxsw_sp_fid_rif_unset(fid); + mlxsw_sp_rif_fdb_op(rif->mlxsw_sp, rif->dev->dev_addr, + mlxsw_sp_fid_index(fid), false); + mlxsw_sp_rif_macvlan_flush(rif); + mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_BC, + mlxsw_sp_router_port(mlxsw_sp), false); + mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_MC, + mlxsw_sp_router_port(mlxsw_sp), false); + mlxsw_sp_rif_fid_op(rif, fid_index, false); + mlxsw_sp_rif_mac_profile_put(rif->mlxsw_sp, rif->mac_profile_id); +} + +static struct mlxsw_sp_fid * +mlxsw_sp_rif_fid_fid_get(struct mlxsw_sp_rif *rif, + struct netlink_ext_ack *extack) +{ + return mlxsw_sp_fid_8021d_get(rif->mlxsw_sp, rif->dev->ifindex); +} + +static void mlxsw_sp_rif_fid_fdb_del(struct mlxsw_sp_rif *rif, const char *mac) +{ + struct switchdev_notifier_fdb_info info = {}; + struct net_device *dev; + + dev = br_fdb_find_port(rif->dev, mac, 0); + if (!dev) + return; + + info.addr = mac; + info.vid = 0; + call_switchdev_notifiers(SWITCHDEV_FDB_DEL_TO_BRIDGE, dev, &info.info, + NULL); +} + +static const struct mlxsw_sp_rif_ops mlxsw_sp_rif_fid_ops = { + .type = MLXSW_SP_RIF_TYPE_FID, + .rif_size = sizeof(struct mlxsw_sp_rif), + .configure = mlxsw_sp_rif_fid_configure, + .deconfigure = mlxsw_sp_rif_fid_deconfigure, + .fid_get = mlxsw_sp_rif_fid_fid_get, + .fdb_del = mlxsw_sp_rif_fid_fdb_del, +}; + +static struct mlxsw_sp_fid * +mlxsw_sp_rif_vlan_fid_get(struct mlxsw_sp_rif *rif, + struct netlink_ext_ack *extack) +{ + struct net_device *br_dev; + u16 vid; + int err; + + if (is_vlan_dev(rif->dev)) { + vid = vlan_dev_vlan_id(rif->dev); + br_dev = vlan_dev_real_dev(rif->dev); + if (WARN_ON(!netif_is_bridge_master(br_dev))) + return ERR_PTR(-EINVAL); + } else { + err = br_vlan_get_pvid(rif->dev, &vid); + if (err < 0 || !vid) { + NL_SET_ERR_MSG_MOD(extack, "Couldn't determine bridge PVID"); + return ERR_PTR(-EINVAL); + } + } + + return mlxsw_sp_fid_8021q_get(rif->mlxsw_sp, vid); +} + +static void mlxsw_sp_rif_vlan_fdb_del(struct mlxsw_sp_rif *rif, const char *mac) +{ + struct switchdev_notifier_fdb_info info = {}; + u16 vid = mlxsw_sp_fid_8021q_vid(rif->fid); + struct net_device *br_dev; + struct net_device *dev; + + br_dev = is_vlan_dev(rif->dev) ? vlan_dev_real_dev(rif->dev) : rif->dev; + dev = br_fdb_find_port(br_dev, mac, vid); + if (!dev) + return; + + info.addr = mac; + info.vid = vid; + call_switchdev_notifiers(SWITCHDEV_FDB_DEL_TO_BRIDGE, dev, &info.info, + NULL); +} + +static int mlxsw_sp_rif_vlan_op(struct mlxsw_sp_rif *rif, u16 vid, u16 efid, + bool enable) +{ + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + char ritr_pl[MLXSW_REG_RITR_LEN]; + + mlxsw_reg_ritr_vlan_if_pack(ritr_pl, enable, rif->rif_index, rif->vr_id, + rif->dev->mtu, rif->dev->dev_addr, + rif->mac_profile_id, vid, efid); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); +} + +static int mlxsw_sp_rif_vlan_configure(struct mlxsw_sp_rif *rif, u16 efid, + struct netlink_ext_ack *extack) +{ + u16 vid = mlxsw_sp_fid_8021q_vid(rif->fid); + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + u8 mac_profile; + int err; + + err = mlxsw_sp_rif_mac_profile_get(mlxsw_sp, rif->addr, + &mac_profile, extack); + if (err) + return err; + rif->mac_profile_id = mac_profile; + + err = mlxsw_sp_rif_vlan_op(rif, vid, efid, true); + if (err) + goto err_rif_vlan_fid_op; + + err = mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_MC, + mlxsw_sp_router_port(mlxsw_sp), true); + if (err) + goto err_fid_mc_flood_set; + + err = mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_BC, + mlxsw_sp_router_port(mlxsw_sp), true); + if (err) + goto err_fid_bc_flood_set; + + err = mlxsw_sp_rif_fdb_op(rif->mlxsw_sp, rif->dev->dev_addr, + mlxsw_sp_fid_index(rif->fid), true); + if (err) + goto err_rif_fdb_op; + + err = mlxsw_sp_fid_rif_set(rif->fid, rif); + if (err) + goto err_fid_rif_set; + + return 0; + +err_fid_rif_set: + mlxsw_sp_rif_fdb_op(rif->mlxsw_sp, rif->dev->dev_addr, + mlxsw_sp_fid_index(rif->fid), false); +err_rif_fdb_op: + mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_BC, + mlxsw_sp_router_port(mlxsw_sp), false); +err_fid_bc_flood_set: + mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_MC, + mlxsw_sp_router_port(mlxsw_sp), false); +err_fid_mc_flood_set: + mlxsw_sp_rif_vlan_op(rif, vid, 0, false); +err_rif_vlan_fid_op: + mlxsw_sp_rif_mac_profile_put(mlxsw_sp, mac_profile); + return err; +} + +static void mlxsw_sp_rif_vlan_deconfigure(struct mlxsw_sp_rif *rif) +{ + u16 vid = mlxsw_sp_fid_8021q_vid(rif->fid); + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + + mlxsw_sp_fid_rif_unset(rif->fid); + mlxsw_sp_rif_fdb_op(rif->mlxsw_sp, rif->dev->dev_addr, + mlxsw_sp_fid_index(rif->fid), false); + mlxsw_sp_rif_macvlan_flush(rif); + mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_BC, + mlxsw_sp_router_port(mlxsw_sp), false); + mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_MC, + mlxsw_sp_router_port(mlxsw_sp), false); + mlxsw_sp_rif_vlan_op(rif, vid, 0, false); + mlxsw_sp_rif_mac_profile_put(rif->mlxsw_sp, rif->mac_profile_id); +} + +static int mlxsw_sp1_rif_vlan_configure(struct mlxsw_sp_rif *rif, + struct netlink_ext_ack *extack) +{ + return mlxsw_sp_rif_vlan_configure(rif, 0, extack); +} + +static const struct mlxsw_sp_rif_ops mlxsw_sp1_rif_vlan_ops = { + .type = MLXSW_SP_RIF_TYPE_VLAN, + .rif_size = sizeof(struct mlxsw_sp_rif), + .configure = mlxsw_sp1_rif_vlan_configure, + .deconfigure = mlxsw_sp_rif_vlan_deconfigure, + .fid_get = mlxsw_sp_rif_vlan_fid_get, + .fdb_del = mlxsw_sp_rif_vlan_fdb_del, +}; + +static int mlxsw_sp2_rif_vlan_configure(struct mlxsw_sp_rif *rif, + struct netlink_ext_ack *extack) +{ + u16 efid = mlxsw_sp_fid_index(rif->fid); + + return mlxsw_sp_rif_vlan_configure(rif, efid, extack); +} + +static const struct mlxsw_sp_rif_ops mlxsw_sp2_rif_vlan_ops = { + .type = MLXSW_SP_RIF_TYPE_VLAN, + .rif_size = sizeof(struct mlxsw_sp_rif), + .configure = mlxsw_sp2_rif_vlan_configure, + .deconfigure = mlxsw_sp_rif_vlan_deconfigure, + .fid_get = mlxsw_sp_rif_vlan_fid_get, + .fdb_del = mlxsw_sp_rif_vlan_fdb_del, +}; + +static struct mlxsw_sp_rif_ipip_lb * +mlxsw_sp_rif_ipip_lb_rif(struct mlxsw_sp_rif *rif) +{ + return container_of(rif, struct mlxsw_sp_rif_ipip_lb, common); +} + +static void +mlxsw_sp_rif_ipip_lb_setup(struct mlxsw_sp_rif *rif, + const struct mlxsw_sp_rif_params *params) +{ + struct mlxsw_sp_rif_params_ipip_lb *params_lb; + struct mlxsw_sp_rif_ipip_lb *rif_lb; + + params_lb = container_of(params, struct mlxsw_sp_rif_params_ipip_lb, + common); + rif_lb = mlxsw_sp_rif_ipip_lb_rif(rif); + rif_lb->lb_config = params_lb->lb_config; +} + +static int +mlxsw_sp1_rif_ipip_lb_configure(struct mlxsw_sp_rif *rif, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_rif_ipip_lb *lb_rif = mlxsw_sp_rif_ipip_lb_rif(rif); + u32 ul_tb_id = mlxsw_sp_ipip_dev_ul_tb_id(rif->dev); + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + struct mlxsw_sp_vr *ul_vr; + int err; + + ul_vr = mlxsw_sp_vr_get(mlxsw_sp, ul_tb_id, NULL); + if (IS_ERR(ul_vr)) + return PTR_ERR(ul_vr); + + err = mlxsw_sp_rif_ipip_lb_op(lb_rif, ul_vr->id, 0, true); + if (err) + goto err_loopback_op; + + lb_rif->ul_vr_id = ul_vr->id; + lb_rif->ul_rif_id = 0; + ++ul_vr->rif_count; + return 0; + +err_loopback_op: + mlxsw_sp_vr_put(mlxsw_sp, ul_vr); + return err; +} + +static void mlxsw_sp1_rif_ipip_lb_deconfigure(struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp_rif_ipip_lb *lb_rif = mlxsw_sp_rif_ipip_lb_rif(rif); + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + struct mlxsw_sp_vr *ul_vr; + + ul_vr = &mlxsw_sp->router->vrs[lb_rif->ul_vr_id]; + mlxsw_sp_rif_ipip_lb_op(lb_rif, ul_vr->id, 0, false); + + --ul_vr->rif_count; + mlxsw_sp_vr_put(mlxsw_sp, ul_vr); +} + +static const struct mlxsw_sp_rif_ops mlxsw_sp1_rif_ipip_lb_ops = { + .type = MLXSW_SP_RIF_TYPE_IPIP_LB, + .rif_size = sizeof(struct mlxsw_sp_rif_ipip_lb), + .setup = mlxsw_sp_rif_ipip_lb_setup, + .configure = mlxsw_sp1_rif_ipip_lb_configure, + .deconfigure = mlxsw_sp1_rif_ipip_lb_deconfigure, +}; + +static const struct mlxsw_sp_rif_ops *mlxsw_sp1_rif_ops_arr[] = { + [MLXSW_SP_RIF_TYPE_SUBPORT] = &mlxsw_sp_rif_subport_ops, + [MLXSW_SP_RIF_TYPE_VLAN] = &mlxsw_sp1_rif_vlan_ops, + [MLXSW_SP_RIF_TYPE_FID] = &mlxsw_sp_rif_fid_ops, + [MLXSW_SP_RIF_TYPE_IPIP_LB] = &mlxsw_sp1_rif_ipip_lb_ops, +}; + +static int +mlxsw_sp_rif_ipip_lb_ul_rif_op(struct mlxsw_sp_rif *ul_rif, bool enable) +{ + struct mlxsw_sp *mlxsw_sp = ul_rif->mlxsw_sp; + char ritr_pl[MLXSW_REG_RITR_LEN]; + + mlxsw_reg_ritr_pack(ritr_pl, enable, MLXSW_REG_RITR_LOOPBACK_IF, + ul_rif->rif_index, ul_rif->vr_id, IP_MAX_MTU); + mlxsw_reg_ritr_loopback_protocol_set(ritr_pl, + MLXSW_REG_RITR_LOOPBACK_GENERIC); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl); +} + +static struct mlxsw_sp_rif * +mlxsw_sp_ul_rif_create(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_vr *vr, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_rif *ul_rif; + u16 rif_index; + int err; + + err = mlxsw_sp_rif_index_alloc(mlxsw_sp, &rif_index); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Exceeded number of supported router interfaces"); + return ERR_PTR(err); + } + + ul_rif = mlxsw_sp_rif_alloc(sizeof(*ul_rif), rif_index, vr->id, NULL); + if (!ul_rif) + return ERR_PTR(-ENOMEM); + + mlxsw_sp->router->rifs[rif_index] = ul_rif; + ul_rif->mlxsw_sp = mlxsw_sp; + err = mlxsw_sp_rif_ipip_lb_ul_rif_op(ul_rif, true); + if (err) + goto ul_rif_op_err; + + atomic_inc(&mlxsw_sp->router->rifs_count); + return ul_rif; + +ul_rif_op_err: + mlxsw_sp->router->rifs[rif_index] = NULL; + kfree(ul_rif); + return ERR_PTR(err); +} + +static void mlxsw_sp_ul_rif_destroy(struct mlxsw_sp_rif *ul_rif) +{ + struct mlxsw_sp *mlxsw_sp = ul_rif->mlxsw_sp; + + atomic_dec(&mlxsw_sp->router->rifs_count); + mlxsw_sp_rif_ipip_lb_ul_rif_op(ul_rif, false); + mlxsw_sp->router->rifs[ul_rif->rif_index] = NULL; + kfree(ul_rif); +} + +static struct mlxsw_sp_rif * +mlxsw_sp_ul_rif_get(struct mlxsw_sp *mlxsw_sp, u32 tb_id, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_vr *vr; + int err; + + vr = mlxsw_sp_vr_get(mlxsw_sp, tb_id, extack); + if (IS_ERR(vr)) + return ERR_CAST(vr); + + if (refcount_inc_not_zero(&vr->ul_rif_refcnt)) + return vr->ul_rif; + + vr->ul_rif = mlxsw_sp_ul_rif_create(mlxsw_sp, vr, extack); + if (IS_ERR(vr->ul_rif)) { + err = PTR_ERR(vr->ul_rif); + goto err_ul_rif_create; + } + + vr->rif_count++; + refcount_set(&vr->ul_rif_refcnt, 1); + + return vr->ul_rif; + +err_ul_rif_create: + mlxsw_sp_vr_put(mlxsw_sp, vr); + return ERR_PTR(err); +} + +static void mlxsw_sp_ul_rif_put(struct mlxsw_sp_rif *ul_rif) +{ + struct mlxsw_sp *mlxsw_sp = ul_rif->mlxsw_sp; + struct mlxsw_sp_vr *vr; + + vr = &mlxsw_sp->router->vrs[ul_rif->vr_id]; + + if (!refcount_dec_and_test(&vr->ul_rif_refcnt)) + return; + + vr->rif_count--; + mlxsw_sp_ul_rif_destroy(ul_rif); + mlxsw_sp_vr_put(mlxsw_sp, vr); +} + +int mlxsw_sp_router_ul_rif_get(struct mlxsw_sp *mlxsw_sp, u32 ul_tb_id, + u16 *ul_rif_index) +{ + struct mlxsw_sp_rif *ul_rif; + int err = 0; + + mutex_lock(&mlxsw_sp->router->lock); + ul_rif = mlxsw_sp_ul_rif_get(mlxsw_sp, ul_tb_id, NULL); + if (IS_ERR(ul_rif)) { + err = PTR_ERR(ul_rif); + goto out; + } + *ul_rif_index = ul_rif->rif_index; +out: + mutex_unlock(&mlxsw_sp->router->lock); + return err; +} + +void mlxsw_sp_router_ul_rif_put(struct mlxsw_sp *mlxsw_sp, u16 ul_rif_index) +{ + struct mlxsw_sp_rif *ul_rif; + + mutex_lock(&mlxsw_sp->router->lock); + ul_rif = mlxsw_sp->router->rifs[ul_rif_index]; + if (WARN_ON(!ul_rif)) + goto out; + + mlxsw_sp_ul_rif_put(ul_rif); +out: + mutex_unlock(&mlxsw_sp->router->lock); +} + +static int +mlxsw_sp2_rif_ipip_lb_configure(struct mlxsw_sp_rif *rif, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_rif_ipip_lb *lb_rif = mlxsw_sp_rif_ipip_lb_rif(rif); + u32 ul_tb_id = mlxsw_sp_ipip_dev_ul_tb_id(rif->dev); + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + struct mlxsw_sp_rif *ul_rif; + int err; + + ul_rif = mlxsw_sp_ul_rif_get(mlxsw_sp, ul_tb_id, NULL); + if (IS_ERR(ul_rif)) + return PTR_ERR(ul_rif); + + err = mlxsw_sp_rif_ipip_lb_op(lb_rif, 0, ul_rif->rif_index, true); + if (err) + goto err_loopback_op; + + lb_rif->ul_vr_id = 0; + lb_rif->ul_rif_id = ul_rif->rif_index; + + return 0; + +err_loopback_op: + mlxsw_sp_ul_rif_put(ul_rif); + return err; +} + +static void mlxsw_sp2_rif_ipip_lb_deconfigure(struct mlxsw_sp_rif *rif) +{ + struct mlxsw_sp_rif_ipip_lb *lb_rif = mlxsw_sp_rif_ipip_lb_rif(rif); + struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp; + struct mlxsw_sp_rif *ul_rif; + + ul_rif = mlxsw_sp_rif_by_index(mlxsw_sp, lb_rif->ul_rif_id); + mlxsw_sp_rif_ipip_lb_op(lb_rif, 0, lb_rif->ul_rif_id, false); + mlxsw_sp_ul_rif_put(ul_rif); +} + +static const struct mlxsw_sp_rif_ops mlxsw_sp2_rif_ipip_lb_ops = { + .type = MLXSW_SP_RIF_TYPE_IPIP_LB, + .rif_size = sizeof(struct mlxsw_sp_rif_ipip_lb), + .setup = mlxsw_sp_rif_ipip_lb_setup, + .configure = mlxsw_sp2_rif_ipip_lb_configure, + .deconfigure = mlxsw_sp2_rif_ipip_lb_deconfigure, +}; + +static const struct mlxsw_sp_rif_ops *mlxsw_sp2_rif_ops_arr[] = { + [MLXSW_SP_RIF_TYPE_SUBPORT] = &mlxsw_sp_rif_subport_ops, + [MLXSW_SP_RIF_TYPE_VLAN] = &mlxsw_sp2_rif_vlan_ops, + [MLXSW_SP_RIF_TYPE_FID] = &mlxsw_sp_rif_fid_ops, + [MLXSW_SP_RIF_TYPE_IPIP_LB] = &mlxsw_sp2_rif_ipip_lb_ops, +}; + +static int mlxsw_sp_rifs_init(struct mlxsw_sp *mlxsw_sp) +{ + u64 max_rifs = MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + struct mlxsw_core *core = mlxsw_sp->core; + + if (!MLXSW_CORE_RES_VALID(core, MAX_RIF_MAC_PROFILES)) + return -EIO; + mlxsw_sp->router->max_rif_mac_profile = + MLXSW_CORE_RES_GET(core, MAX_RIF_MAC_PROFILES); + + mlxsw_sp->router->rifs = kcalloc(max_rifs, + sizeof(struct mlxsw_sp_rif *), + GFP_KERNEL); + if (!mlxsw_sp->router->rifs) + return -ENOMEM; + + idr_init(&mlxsw_sp->router->rif_mac_profiles_idr); + atomic_set(&mlxsw_sp->router->rif_mac_profiles_count, 0); + atomic_set(&mlxsw_sp->router->rifs_count, 0); + devl_resource_occ_get_register(devlink, + MLXSW_SP_RESOURCE_RIF_MAC_PROFILES, + mlxsw_sp_rif_mac_profiles_occ_get, + mlxsw_sp); + devl_resource_occ_get_register(devlink, + MLXSW_SP_RESOURCE_RIFS, + mlxsw_sp_rifs_occ_get, + mlxsw_sp); + + return 0; +} + +static void mlxsw_sp_rifs_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + int i; + + WARN_ON_ONCE(atomic_read(&mlxsw_sp->router->rifs_count)); + for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); i++) + WARN_ON_ONCE(mlxsw_sp->router->rifs[i]); + + devl_resource_occ_get_unregister(devlink, MLXSW_SP_RESOURCE_RIFS); + devl_resource_occ_get_unregister(devlink, + MLXSW_SP_RESOURCE_RIF_MAC_PROFILES); + WARN_ON(!idr_is_empty(&mlxsw_sp->router->rif_mac_profiles_idr)); + idr_destroy(&mlxsw_sp->router->rif_mac_profiles_idr); + kfree(mlxsw_sp->router->rifs); +} + +static int +mlxsw_sp_ipip_config_tigcr(struct mlxsw_sp *mlxsw_sp) +{ + char tigcr_pl[MLXSW_REG_TIGCR_LEN]; + + mlxsw_reg_tigcr_pack(tigcr_pl, true, 0); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tigcr), tigcr_pl); +} + +static int mlxsw_sp_ipips_init(struct mlxsw_sp *mlxsw_sp) +{ + int err; + + INIT_LIST_HEAD(&mlxsw_sp->router->ipip_list); + + err = mlxsw_sp_ipip_ecn_encap_init(mlxsw_sp); + if (err) + return err; + err = mlxsw_sp_ipip_ecn_decap_init(mlxsw_sp); + if (err) + return err; + + return mlxsw_sp_ipip_config_tigcr(mlxsw_sp); +} + +static int mlxsw_sp1_ipips_init(struct mlxsw_sp *mlxsw_sp) +{ + mlxsw_sp->router->ipip_ops_arr = mlxsw_sp1_ipip_ops_arr; + return mlxsw_sp_ipips_init(mlxsw_sp); +} + +static int mlxsw_sp2_ipips_init(struct mlxsw_sp *mlxsw_sp) +{ + mlxsw_sp->router->ipip_ops_arr = mlxsw_sp2_ipip_ops_arr; + return mlxsw_sp_ipips_init(mlxsw_sp); +} + +static void mlxsw_sp_ipips_fini(struct mlxsw_sp *mlxsw_sp) +{ + WARN_ON(!list_empty(&mlxsw_sp->router->ipip_list)); +} + +static void mlxsw_sp_router_fib_dump_flush(struct notifier_block *nb) +{ + struct mlxsw_sp_router *router; + + /* Flush pending FIB notifications and then flush the device's + * table before requesting another dump. The FIB notification + * block is unregistered, so no need to take RTNL. + */ + mlxsw_core_flush_owq(); + router = container_of(nb, struct mlxsw_sp_router, fib_nb); + mlxsw_sp_router_fib_flush(router->mlxsw_sp); +} + +#ifdef CONFIG_IP_ROUTE_MULTIPATH +struct mlxsw_sp_mp_hash_config { + DECLARE_BITMAP(headers, __MLXSW_REG_RECR2_HEADER_CNT); + DECLARE_BITMAP(fields, __MLXSW_REG_RECR2_FIELD_CNT); + DECLARE_BITMAP(inner_headers, __MLXSW_REG_RECR2_HEADER_CNT); + DECLARE_BITMAP(inner_fields, __MLXSW_REG_RECR2_INNER_FIELD_CNT); + bool inc_parsing_depth; +}; + +#define MLXSW_SP_MP_HASH_HEADER_SET(_headers, _header) \ + bitmap_set(_headers, MLXSW_REG_RECR2_##_header, 1) + +#define MLXSW_SP_MP_HASH_FIELD_SET(_fields, _field) \ + bitmap_set(_fields, MLXSW_REG_RECR2_##_field, 1) + +#define MLXSW_SP_MP_HASH_FIELD_RANGE_SET(_fields, _field, _nr) \ + bitmap_set(_fields, MLXSW_REG_RECR2_##_field, _nr) + +static void mlxsw_sp_mp_hash_inner_l3(struct mlxsw_sp_mp_hash_config *config) +{ + unsigned long *inner_headers = config->inner_headers; + unsigned long *inner_fields = config->inner_fields; + + /* IPv4 inner */ + MLXSW_SP_MP_HASH_HEADER_SET(inner_headers, IPV4_EN_NOT_TCP_NOT_UDP); + MLXSW_SP_MP_HASH_HEADER_SET(inner_headers, IPV4_EN_TCP_UDP); + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(inner_fields, INNER_IPV4_SIP0, 4); + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(inner_fields, INNER_IPV4_DIP0, 4); + /* IPv6 inner */ + MLXSW_SP_MP_HASH_HEADER_SET(inner_headers, IPV6_EN_NOT_TCP_NOT_UDP); + MLXSW_SP_MP_HASH_HEADER_SET(inner_headers, IPV6_EN_TCP_UDP); + MLXSW_SP_MP_HASH_FIELD_SET(inner_fields, INNER_IPV6_SIP0_7); + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(inner_fields, INNER_IPV6_SIP8, 8); + MLXSW_SP_MP_HASH_FIELD_SET(inner_fields, INNER_IPV6_DIP0_7); + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(inner_fields, INNER_IPV6_DIP8, 8); + MLXSW_SP_MP_HASH_FIELD_SET(inner_fields, INNER_IPV6_NEXT_HEADER); + MLXSW_SP_MP_HASH_FIELD_SET(inner_fields, INNER_IPV6_FLOW_LABEL); +} + +static void mlxsw_sp_mp4_hash_outer_addr(struct mlxsw_sp_mp_hash_config *config) +{ + unsigned long *headers = config->headers; + unsigned long *fields = config->fields; + + MLXSW_SP_MP_HASH_HEADER_SET(headers, IPV4_EN_NOT_TCP_NOT_UDP); + MLXSW_SP_MP_HASH_HEADER_SET(headers, IPV4_EN_TCP_UDP); + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(fields, IPV4_SIP0, 4); + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(fields, IPV4_DIP0, 4); +} + +static void +mlxsw_sp_mp_hash_inner_custom(struct mlxsw_sp_mp_hash_config *config, + u32 hash_fields) +{ + unsigned long *inner_headers = config->inner_headers; + unsigned long *inner_fields = config->inner_fields; + + /* IPv4 Inner */ + MLXSW_SP_MP_HASH_HEADER_SET(inner_headers, IPV4_EN_NOT_TCP_NOT_UDP); + MLXSW_SP_MP_HASH_HEADER_SET(inner_headers, IPV4_EN_TCP_UDP); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_INNER_SRC_IP) + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(inner_fields, INNER_IPV4_SIP0, 4); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_INNER_DST_IP) + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(inner_fields, INNER_IPV4_DIP0, 4); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_INNER_IP_PROTO) + MLXSW_SP_MP_HASH_FIELD_SET(inner_fields, INNER_IPV4_PROTOCOL); + /* IPv6 inner */ + MLXSW_SP_MP_HASH_HEADER_SET(inner_headers, IPV6_EN_NOT_TCP_NOT_UDP); + MLXSW_SP_MP_HASH_HEADER_SET(inner_headers, IPV6_EN_TCP_UDP); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_INNER_SRC_IP) { + MLXSW_SP_MP_HASH_FIELD_SET(inner_fields, INNER_IPV6_SIP0_7); + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(inner_fields, INNER_IPV6_SIP8, 8); + } + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_INNER_DST_IP) { + MLXSW_SP_MP_HASH_FIELD_SET(inner_fields, INNER_IPV6_DIP0_7); + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(inner_fields, INNER_IPV6_DIP8, 8); + } + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_INNER_IP_PROTO) + MLXSW_SP_MP_HASH_FIELD_SET(inner_fields, INNER_IPV6_NEXT_HEADER); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_INNER_FLOWLABEL) + MLXSW_SP_MP_HASH_FIELD_SET(inner_fields, INNER_IPV6_FLOW_LABEL); + /* L4 inner */ + MLXSW_SP_MP_HASH_HEADER_SET(inner_headers, TCP_UDP_EN_IPV4); + MLXSW_SP_MP_HASH_HEADER_SET(inner_headers, TCP_UDP_EN_IPV6); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_INNER_SRC_PORT) + MLXSW_SP_MP_HASH_FIELD_SET(inner_fields, INNER_TCP_UDP_SPORT); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_INNER_DST_PORT) + MLXSW_SP_MP_HASH_FIELD_SET(inner_fields, INNER_TCP_UDP_DPORT); +} + +static void mlxsw_sp_mp4_hash_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mp_hash_config *config) +{ + struct net *net = mlxsw_sp_net(mlxsw_sp); + unsigned long *headers = config->headers; + unsigned long *fields = config->fields; + u32 hash_fields; + + switch (READ_ONCE(net->ipv4.sysctl_fib_multipath_hash_policy)) { + case 0: + mlxsw_sp_mp4_hash_outer_addr(config); + break; + case 1: + mlxsw_sp_mp4_hash_outer_addr(config); + MLXSW_SP_MP_HASH_HEADER_SET(headers, TCP_UDP_EN_IPV4); + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV4_PROTOCOL); + MLXSW_SP_MP_HASH_FIELD_SET(fields, TCP_UDP_SPORT); + MLXSW_SP_MP_HASH_FIELD_SET(fields, TCP_UDP_DPORT); + break; + case 2: + /* Outer */ + mlxsw_sp_mp4_hash_outer_addr(config); + /* Inner */ + mlxsw_sp_mp_hash_inner_l3(config); + break; + case 3: + hash_fields = READ_ONCE(net->ipv4.sysctl_fib_multipath_hash_fields); + /* Outer */ + MLXSW_SP_MP_HASH_HEADER_SET(headers, IPV4_EN_NOT_TCP_NOT_UDP); + MLXSW_SP_MP_HASH_HEADER_SET(headers, IPV4_EN_TCP_UDP); + MLXSW_SP_MP_HASH_HEADER_SET(headers, TCP_UDP_EN_IPV4); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_SRC_IP) + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(fields, IPV4_SIP0, 4); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_DST_IP) + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(fields, IPV4_DIP0, 4); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_IP_PROTO) + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV4_PROTOCOL); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_SRC_PORT) + MLXSW_SP_MP_HASH_FIELD_SET(fields, TCP_UDP_SPORT); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_DST_PORT) + MLXSW_SP_MP_HASH_FIELD_SET(fields, TCP_UDP_DPORT); + /* Inner */ + mlxsw_sp_mp_hash_inner_custom(config, hash_fields); + break; + } +} + +static void mlxsw_sp_mp6_hash_outer_addr(struct mlxsw_sp_mp_hash_config *config) +{ + unsigned long *headers = config->headers; + unsigned long *fields = config->fields; + + MLXSW_SP_MP_HASH_HEADER_SET(headers, IPV6_EN_NOT_TCP_NOT_UDP); + MLXSW_SP_MP_HASH_HEADER_SET(headers, IPV6_EN_TCP_UDP); + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV6_SIP0_7); + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(fields, IPV6_SIP8, 8); + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV6_DIP0_7); + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(fields, IPV6_DIP8, 8); +} + +static void mlxsw_sp_mp6_hash_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mp_hash_config *config) +{ + u32 hash_fields = ip6_multipath_hash_fields(mlxsw_sp_net(mlxsw_sp)); + unsigned long *headers = config->headers; + unsigned long *fields = config->fields; + + switch (ip6_multipath_hash_policy(mlxsw_sp_net(mlxsw_sp))) { + case 0: + mlxsw_sp_mp6_hash_outer_addr(config); + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV6_NEXT_HEADER); + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV6_FLOW_LABEL); + break; + case 1: + mlxsw_sp_mp6_hash_outer_addr(config); + MLXSW_SP_MP_HASH_HEADER_SET(headers, TCP_UDP_EN_IPV6); + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV6_NEXT_HEADER); + MLXSW_SP_MP_HASH_FIELD_SET(fields, TCP_UDP_SPORT); + MLXSW_SP_MP_HASH_FIELD_SET(fields, TCP_UDP_DPORT); + break; + case 2: + /* Outer */ + mlxsw_sp_mp6_hash_outer_addr(config); + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV6_NEXT_HEADER); + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV6_FLOW_LABEL); + /* Inner */ + mlxsw_sp_mp_hash_inner_l3(config); + config->inc_parsing_depth = true; + break; + case 3: + /* Outer */ + MLXSW_SP_MP_HASH_HEADER_SET(headers, IPV6_EN_NOT_TCP_NOT_UDP); + MLXSW_SP_MP_HASH_HEADER_SET(headers, IPV6_EN_TCP_UDP); + MLXSW_SP_MP_HASH_HEADER_SET(headers, TCP_UDP_EN_IPV6); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_SRC_IP) { + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV6_SIP0_7); + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(fields, IPV6_SIP8, 8); + } + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_DST_IP) { + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV6_DIP0_7); + MLXSW_SP_MP_HASH_FIELD_RANGE_SET(fields, IPV6_DIP8, 8); + } + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_IP_PROTO) + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV6_NEXT_HEADER); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_FLOWLABEL) + MLXSW_SP_MP_HASH_FIELD_SET(fields, IPV6_FLOW_LABEL); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_SRC_PORT) + MLXSW_SP_MP_HASH_FIELD_SET(fields, TCP_UDP_SPORT); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_DST_PORT) + MLXSW_SP_MP_HASH_FIELD_SET(fields, TCP_UDP_DPORT); + /* Inner */ + mlxsw_sp_mp_hash_inner_custom(config, hash_fields); + if (hash_fields & FIB_MULTIPATH_HASH_FIELD_INNER_MASK) + config->inc_parsing_depth = true; + break; + } +} + +static int mlxsw_sp_mp_hash_parsing_depth_adjust(struct mlxsw_sp *mlxsw_sp, + bool old_inc_parsing_depth, + bool new_inc_parsing_depth) +{ + int err; + + if (!old_inc_parsing_depth && new_inc_parsing_depth) { + err = mlxsw_sp_parsing_depth_inc(mlxsw_sp); + if (err) + return err; + mlxsw_sp->router->inc_parsing_depth = true; + } else if (old_inc_parsing_depth && !new_inc_parsing_depth) { + mlxsw_sp_parsing_depth_dec(mlxsw_sp); + mlxsw_sp->router->inc_parsing_depth = false; + } + + return 0; +} + +static int mlxsw_sp_mp_hash_init(struct mlxsw_sp *mlxsw_sp) +{ + bool old_inc_parsing_depth, new_inc_parsing_depth; + struct mlxsw_sp_mp_hash_config config = {}; + char recr2_pl[MLXSW_REG_RECR2_LEN]; + unsigned long bit; + u32 seed; + int err; + + seed = jhash(mlxsw_sp->base_mac, sizeof(mlxsw_sp->base_mac), 0); + mlxsw_reg_recr2_pack(recr2_pl, seed); + mlxsw_sp_mp4_hash_init(mlxsw_sp, &config); + mlxsw_sp_mp6_hash_init(mlxsw_sp, &config); + + old_inc_parsing_depth = mlxsw_sp->router->inc_parsing_depth; + new_inc_parsing_depth = config.inc_parsing_depth; + err = mlxsw_sp_mp_hash_parsing_depth_adjust(mlxsw_sp, + old_inc_parsing_depth, + new_inc_parsing_depth); + if (err) + return err; + + for_each_set_bit(bit, config.headers, __MLXSW_REG_RECR2_HEADER_CNT) + mlxsw_reg_recr2_outer_header_enables_set(recr2_pl, bit, 1); + for_each_set_bit(bit, config.fields, __MLXSW_REG_RECR2_FIELD_CNT) + mlxsw_reg_recr2_outer_header_fields_enable_set(recr2_pl, bit, 1); + for_each_set_bit(bit, config.inner_headers, __MLXSW_REG_RECR2_HEADER_CNT) + mlxsw_reg_recr2_inner_header_enables_set(recr2_pl, bit, 1); + for_each_set_bit(bit, config.inner_fields, __MLXSW_REG_RECR2_INNER_FIELD_CNT) + mlxsw_reg_recr2_inner_header_fields_enable_set(recr2_pl, bit, 1); + + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(recr2), recr2_pl); + if (err) + goto err_reg_write; + + return 0; + +err_reg_write: + mlxsw_sp_mp_hash_parsing_depth_adjust(mlxsw_sp, new_inc_parsing_depth, + old_inc_parsing_depth); + return err; +} + +static void mlxsw_sp_mp_hash_fini(struct mlxsw_sp *mlxsw_sp) +{ + bool old_inc_parsing_depth = mlxsw_sp->router->inc_parsing_depth; + + mlxsw_sp_mp_hash_parsing_depth_adjust(mlxsw_sp, old_inc_parsing_depth, + false); +} +#else +static int mlxsw_sp_mp_hash_init(struct mlxsw_sp *mlxsw_sp) +{ + return 0; +} + +static void mlxsw_sp_mp_hash_fini(struct mlxsw_sp *mlxsw_sp) +{ +} +#endif + +static int mlxsw_sp_dscp_init(struct mlxsw_sp *mlxsw_sp) +{ + char rdpm_pl[MLXSW_REG_RDPM_LEN]; + unsigned int i; + + MLXSW_REG_ZERO(rdpm, rdpm_pl); + + /* HW is determining switch priority based on DSCP-bits, but the + * kernel is still doing that based on the ToS. Since there's a + * mismatch in bits we need to make sure to translate the right + * value ToS would observe, skipping the 2 least-significant ECN bits. + */ + for (i = 0; i < MLXSW_REG_RDPM_DSCP_ENTRY_REC_MAX_COUNT; i++) + mlxsw_reg_rdpm_pack(rdpm_pl, i, rt_tos2priority(i << 2)); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rdpm), rdpm_pl); +} + +static int __mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp) +{ + struct net *net = mlxsw_sp_net(mlxsw_sp); + char rgcr_pl[MLXSW_REG_RGCR_LEN]; + u64 max_rifs; + bool usp; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, MAX_RIFS)) + return -EIO; + max_rifs = MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); + usp = READ_ONCE(net->ipv4.sysctl_ip_fwd_update_priority); + + mlxsw_reg_rgcr_pack(rgcr_pl, true, true); + mlxsw_reg_rgcr_max_router_interfaces_set(rgcr_pl, max_rifs); + mlxsw_reg_rgcr_usp_set(rgcr_pl, usp); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rgcr), rgcr_pl); +} + +static void __mlxsw_sp_router_fini(struct mlxsw_sp *mlxsw_sp) +{ + char rgcr_pl[MLXSW_REG_RGCR_LEN]; + + mlxsw_reg_rgcr_pack(rgcr_pl, false, false); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rgcr), rgcr_pl); +} + +static int mlxsw_sp_lb_rif_init(struct mlxsw_sp *mlxsw_sp) +{ + u16 lb_rif_index; + int err; + + /* Create a generic loopback RIF associated with the main table + * (default VRF). Any table can be used, but the main table exists + * anyway, so we do not waste resources. + */ + err = mlxsw_sp_router_ul_rif_get(mlxsw_sp, RT_TABLE_MAIN, + &lb_rif_index); + if (err) + return err; + + mlxsw_sp->router->lb_rif_index = lb_rif_index; + + return 0; +} + +static void mlxsw_sp_lb_rif_fini(struct mlxsw_sp *mlxsw_sp) +{ + mlxsw_sp_router_ul_rif_put(mlxsw_sp, mlxsw_sp->router->lb_rif_index); +} + +static int mlxsw_sp1_router_init(struct mlxsw_sp *mlxsw_sp) +{ + size_t size_ranges_count = ARRAY_SIZE(mlxsw_sp1_adj_grp_size_ranges); + + mlxsw_sp->router->rif_ops_arr = mlxsw_sp1_rif_ops_arr; + mlxsw_sp->router->adj_grp_size_ranges = mlxsw_sp1_adj_grp_size_ranges; + mlxsw_sp->router->adj_grp_size_ranges_count = size_ranges_count; + + return 0; +} + +const struct mlxsw_sp_router_ops mlxsw_sp1_router_ops = { + .init = mlxsw_sp1_router_init, + .ipips_init = mlxsw_sp1_ipips_init, +}; + +static int mlxsw_sp2_router_init(struct mlxsw_sp *mlxsw_sp) +{ + size_t size_ranges_count = ARRAY_SIZE(mlxsw_sp2_adj_grp_size_ranges); + + mlxsw_sp->router->rif_ops_arr = mlxsw_sp2_rif_ops_arr; + mlxsw_sp->router->adj_grp_size_ranges = mlxsw_sp2_adj_grp_size_ranges; + mlxsw_sp->router->adj_grp_size_ranges_count = size_ranges_count; + + return 0; +} + +const struct mlxsw_sp_router_ops mlxsw_sp2_router_ops = { + .init = mlxsw_sp2_router_init, + .ipips_init = mlxsw_sp2_ipips_init, +}; + +int mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_router *router; + int err; + + router = kzalloc(sizeof(*mlxsw_sp->router), GFP_KERNEL); + if (!router) + return -ENOMEM; + mutex_init(&router->lock); + mlxsw_sp->router = router; + router->mlxsw_sp = mlxsw_sp; + + err = mlxsw_sp->router_ops->init(mlxsw_sp); + if (err) + goto err_router_ops_init; + + INIT_LIST_HEAD(&mlxsw_sp->router->nh_res_grp_list); + INIT_DELAYED_WORK(&mlxsw_sp->router->nh_grp_activity_dw, + mlxsw_sp_nh_grp_activity_work); + INIT_LIST_HEAD(&mlxsw_sp->router->nexthop_neighs_list); + err = __mlxsw_sp_router_init(mlxsw_sp); + if (err) + goto err_router_init; + + err = mlxsw_sp_rifs_init(mlxsw_sp); + if (err) + goto err_rifs_init; + + err = mlxsw_sp->router_ops->ipips_init(mlxsw_sp); + if (err) + goto err_ipips_init; + + err = rhashtable_init(&mlxsw_sp->router->nexthop_ht, + &mlxsw_sp_nexthop_ht_params); + if (err) + goto err_nexthop_ht_init; + + err = rhashtable_init(&mlxsw_sp->router->nexthop_group_ht, + &mlxsw_sp_nexthop_group_ht_params); + if (err) + goto err_nexthop_group_ht_init; + + INIT_LIST_HEAD(&mlxsw_sp->router->nexthop_list); + err = mlxsw_sp_lpm_init(mlxsw_sp); + if (err) + goto err_lpm_init; + + err = mlxsw_sp_mr_init(mlxsw_sp, &mlxsw_sp_mr_tcam_ops); + if (err) + goto err_mr_init; + + err = mlxsw_sp_vrs_init(mlxsw_sp); + if (err) + goto err_vrs_init; + + err = mlxsw_sp_lb_rif_init(mlxsw_sp); + if (err) + goto err_lb_rif_init; + + err = mlxsw_sp_neigh_init(mlxsw_sp); + if (err) + goto err_neigh_init; + + err = mlxsw_sp_mp_hash_init(mlxsw_sp); + if (err) + goto err_mp_hash_init; + + err = mlxsw_sp_dscp_init(mlxsw_sp); + if (err) + goto err_dscp_init; + + router->inetaddr_nb.notifier_call = mlxsw_sp_inetaddr_event; + err = register_inetaddr_notifier(&router->inetaddr_nb); + if (err) + goto err_register_inetaddr_notifier; + + router->inet6addr_nb.notifier_call = mlxsw_sp_inet6addr_event; + err = register_inet6addr_notifier(&router->inet6addr_nb); + if (err) + goto err_register_inet6addr_notifier; + + mlxsw_sp->router->netevent_nb.notifier_call = + mlxsw_sp_router_netevent_event; + err = register_netevent_notifier(&mlxsw_sp->router->netevent_nb); + if (err) + goto err_register_netevent_notifier; + + mlxsw_sp->router->nexthop_nb.notifier_call = + mlxsw_sp_nexthop_obj_event; + err = register_nexthop_notifier(mlxsw_sp_net(mlxsw_sp), + &mlxsw_sp->router->nexthop_nb, + extack); + if (err) + goto err_register_nexthop_notifier; + + mlxsw_sp->router->fib_nb.notifier_call = mlxsw_sp_router_fib_event; + err = register_fib_notifier(mlxsw_sp_net(mlxsw_sp), + &mlxsw_sp->router->fib_nb, + mlxsw_sp_router_fib_dump_flush, extack); + if (err) + goto err_register_fib_notifier; + + mlxsw_sp->router->netdevice_nb.notifier_call = + mlxsw_sp_router_netdevice_event; + err = register_netdevice_notifier_net(mlxsw_sp_net(mlxsw_sp), + &mlxsw_sp->router->netdevice_nb); + if (err) + goto err_register_netdev_notifier; + + return 0; + +err_register_netdev_notifier: + unregister_fib_notifier(mlxsw_sp_net(mlxsw_sp), + &mlxsw_sp->router->fib_nb); +err_register_fib_notifier: + unregister_nexthop_notifier(mlxsw_sp_net(mlxsw_sp), + &mlxsw_sp->router->nexthop_nb); +err_register_nexthop_notifier: + unregister_netevent_notifier(&mlxsw_sp->router->netevent_nb); +err_register_netevent_notifier: + unregister_inet6addr_notifier(&router->inet6addr_nb); +err_register_inet6addr_notifier: + unregister_inetaddr_notifier(&router->inetaddr_nb); +err_register_inetaddr_notifier: + mlxsw_core_flush_owq(); +err_dscp_init: + mlxsw_sp_mp_hash_fini(mlxsw_sp); +err_mp_hash_init: + mlxsw_sp_neigh_fini(mlxsw_sp); +err_neigh_init: + mlxsw_sp_lb_rif_fini(mlxsw_sp); +err_lb_rif_init: + mlxsw_sp_vrs_fini(mlxsw_sp); +err_vrs_init: + mlxsw_sp_mr_fini(mlxsw_sp); +err_mr_init: + mlxsw_sp_lpm_fini(mlxsw_sp); +err_lpm_init: + rhashtable_destroy(&mlxsw_sp->router->nexthop_group_ht); +err_nexthop_group_ht_init: + rhashtable_destroy(&mlxsw_sp->router->nexthop_ht); +err_nexthop_ht_init: + mlxsw_sp_ipips_fini(mlxsw_sp); +err_ipips_init: + mlxsw_sp_rifs_fini(mlxsw_sp); +err_rifs_init: + __mlxsw_sp_router_fini(mlxsw_sp); +err_router_init: + cancel_delayed_work_sync(&mlxsw_sp->router->nh_grp_activity_dw); +err_router_ops_init: + mutex_destroy(&mlxsw_sp->router->lock); + kfree(mlxsw_sp->router); + return err; +} + +void mlxsw_sp_router_fini(struct mlxsw_sp *mlxsw_sp) +{ + unregister_netdevice_notifier_net(mlxsw_sp_net(mlxsw_sp), + &mlxsw_sp->router->netdevice_nb); + unregister_fib_notifier(mlxsw_sp_net(mlxsw_sp), + &mlxsw_sp->router->fib_nb); + unregister_nexthop_notifier(mlxsw_sp_net(mlxsw_sp), + &mlxsw_sp->router->nexthop_nb); + unregister_netevent_notifier(&mlxsw_sp->router->netevent_nb); + unregister_inet6addr_notifier(&mlxsw_sp->router->inet6addr_nb); + unregister_inetaddr_notifier(&mlxsw_sp->router->inetaddr_nb); + mlxsw_core_flush_owq(); + mlxsw_sp_mp_hash_fini(mlxsw_sp); + mlxsw_sp_neigh_fini(mlxsw_sp); + mlxsw_sp_lb_rif_fini(mlxsw_sp); + mlxsw_sp_vrs_fini(mlxsw_sp); + mlxsw_sp_mr_fini(mlxsw_sp); + mlxsw_sp_lpm_fini(mlxsw_sp); + rhashtable_destroy(&mlxsw_sp->router->nexthop_group_ht); + rhashtable_destroy(&mlxsw_sp->router->nexthop_ht); + mlxsw_sp_ipips_fini(mlxsw_sp); + mlxsw_sp_rifs_fini(mlxsw_sp); + __mlxsw_sp_router_fini(mlxsw_sp); + cancel_delayed_work_sync(&mlxsw_sp->router->nh_grp_activity_dw); + mutex_destroy(&mlxsw_sp->router->lock); + kfree(mlxsw_sp->router); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h new file mode 100644 index 000000000..c5dfb972b --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h @@ -0,0 +1,166 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2017-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_ROUTER_H_ +#define _MLXSW_ROUTER_H_ + +#include "spectrum.h" +#include "reg.h" + +struct mlxsw_sp_router_nve_decap { + u32 ul_tb_id; + u32 tunnel_index; + enum mlxsw_sp_l3proto ul_proto; + union mlxsw_sp_l3addr ul_sip; + u8 valid:1; +}; + +struct mlxsw_sp_router { + struct mlxsw_sp *mlxsw_sp; + struct mlxsw_sp_rif **rifs; + struct idr rif_mac_profiles_idr; + atomic_t rif_mac_profiles_count; + atomic_t rifs_count; + u8 max_rif_mac_profile; + struct mlxsw_sp_vr *vrs; + struct rhashtable neigh_ht; + struct rhashtable nexthop_group_ht; + struct rhashtable nexthop_ht; + struct list_head nexthop_list; + struct { + /* One tree for each protocol: IPv4 and IPv6 */ + struct mlxsw_sp_lpm_tree *proto_trees[2]; + struct mlxsw_sp_lpm_tree *trees; + unsigned int tree_count; + } lpm; + struct { + struct delayed_work dw; + unsigned long interval; /* ms */ + atomic_t neigh_count; + } neighs_update; + struct delayed_work nexthop_probe_dw; +#define MLXSW_SP_UNRESOLVED_NH_PROBE_INTERVAL 5000 /* ms */ + struct list_head nexthop_neighs_list; + struct list_head ipip_list; + struct notifier_block nexthop_nb; + struct notifier_block fib_nb; + struct notifier_block netevent_nb; + struct notifier_block inetaddr_nb; + struct notifier_block inet6addr_nb; + struct notifier_block netdevice_nb; + const struct mlxsw_sp_rif_ops **rif_ops_arr; + const struct mlxsw_sp_ipip_ops **ipip_ops_arr; + struct mlxsw_sp_router_nve_decap nve_decap_config; + struct mutex lock; /* Protects shared router resources */ + struct mlxsw_sp_fib_entry_op_ctx *ll_op_ctx; + u16 lb_rif_index; + const struct mlxsw_sp_adj_grp_size_range *adj_grp_size_ranges; + size_t adj_grp_size_ranges_count; + struct delayed_work nh_grp_activity_dw; + struct list_head nh_res_grp_list; + bool inc_parsing_depth; + refcount_t num_groups; + u32 adj_trap_index; +}; + +struct mlxsw_sp_rif_ipip_lb; +struct mlxsw_sp_rif_ipip_lb_config { + enum mlxsw_reg_ritr_loopback_ipip_type lb_ipipt; + u32 okey; + enum mlxsw_sp_l3proto ul_protocol; /* Underlay. */ + union mlxsw_sp_l3addr saddr; +}; + +enum mlxsw_sp_rif_counter_dir { + MLXSW_SP_RIF_COUNTER_INGRESS, + MLXSW_SP_RIF_COUNTER_EGRESS, +}; + +struct mlxsw_sp_neigh_entry; +struct mlxsw_sp_nexthop; +struct mlxsw_sp_ipip_entry; + +struct mlxsw_sp_rif *mlxsw_sp_rif_by_index(const struct mlxsw_sp *mlxsw_sp, + u16 rif_index); +u16 mlxsw_sp_ipip_lb_rif_index(const struct mlxsw_sp_rif_ipip_lb *rif); +u16 mlxsw_sp_ipip_lb_ul_vr_id(const struct mlxsw_sp_rif_ipip_lb *rif); +u16 mlxsw_sp_ipip_lb_ul_rif_id(const struct mlxsw_sp_rif_ipip_lb *lb_rif); +u32 mlxsw_sp_ipip_dev_ul_tb_id(const struct net_device *ol_dev); +int mlxsw_sp_rif_dev_ifindex(const struct mlxsw_sp_rif *rif); +const struct net_device *mlxsw_sp_rif_dev(const struct mlxsw_sp_rif *rif); +int mlxsw_sp_rif_counter_value_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_rif *rif, + enum mlxsw_sp_rif_counter_dir dir, + u64 *cnt); +void mlxsw_sp_rif_counter_free(struct mlxsw_sp_rif *rif, + enum mlxsw_sp_rif_counter_dir dir); +int mlxsw_sp_rif_counter_alloc(struct mlxsw_sp_rif *rif, + enum mlxsw_sp_rif_counter_dir dir); +struct mlxsw_sp_neigh_entry * +mlxsw_sp_rif_neigh_next(struct mlxsw_sp_rif *rif, + struct mlxsw_sp_neigh_entry *neigh_entry); +int mlxsw_sp_neigh_entry_type(struct mlxsw_sp_neigh_entry *neigh_entry); +unsigned char * +mlxsw_sp_neigh_entry_ha(struct mlxsw_sp_neigh_entry *neigh_entry); +u32 mlxsw_sp_neigh4_entry_dip(struct mlxsw_sp_neigh_entry *neigh_entry); +struct in6_addr * +mlxsw_sp_neigh6_entry_dip(struct mlxsw_sp_neigh_entry *neigh_entry); + +#define mlxsw_sp_rif_neigh_for_each(neigh_entry, rif) \ + for (neigh_entry = mlxsw_sp_rif_neigh_next(rif, NULL); neigh_entry; \ + neigh_entry = mlxsw_sp_rif_neigh_next(rif, neigh_entry)) +int mlxsw_sp_neigh_counter_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry, + u64 *p_counter); +void +mlxsw_sp_neigh_entry_counter_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry, + bool adding); +bool mlxsw_sp_neigh_ipv6_ignore(struct mlxsw_sp_neigh_entry *neigh_entry); +int __mlxsw_sp_ipip_entry_update_tunnel(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry, + bool recreate_loopback, + bool keep_encap, + bool update_nexthops, + struct netlink_ext_ack *extack); +void mlxsw_sp_ipip_entry_demote_tunnel(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ipip_entry *ipip_entry); +bool +mlxsw_sp_ipip_demote_tunnel_by_saddr(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_l3proto ul_proto, + union mlxsw_sp_l3addr saddr, + u32 ul_tb_id, + const struct mlxsw_sp_ipip_entry *except); +struct mlxsw_sp_nexthop *mlxsw_sp_nexthop_next(struct mlxsw_sp_router *router, + struct mlxsw_sp_nexthop *nh); +bool mlxsw_sp_nexthop_is_forward(const struct mlxsw_sp_nexthop *nh); +unsigned char *mlxsw_sp_nexthop_ha(struct mlxsw_sp_nexthop *nh); +int mlxsw_sp_nexthop_indexes(struct mlxsw_sp_nexthop *nh, u32 *p_adj_index, + u32 *p_adj_size, u32 *p_adj_hash_index); +struct mlxsw_sp_rif *mlxsw_sp_nexthop_rif(struct mlxsw_sp_nexthop *nh); +bool mlxsw_sp_nexthop_group_has_ipip(struct mlxsw_sp_nexthop *nh); +#define mlxsw_sp_nexthop_for_each(nh, router) \ + for (nh = mlxsw_sp_nexthop_next(router, NULL); nh; \ + nh = mlxsw_sp_nexthop_next(router, nh)) +int mlxsw_sp_nexthop_counter_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh, u64 *p_counter); +int mlxsw_sp_nexthop_eth_update(struct mlxsw_sp *mlxsw_sp, u32 adj_index, + struct mlxsw_sp_nexthop *nh, bool force, + char *ratr_pl); +void mlxsw_sp_nexthop_counter_alloc(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh); +void mlxsw_sp_nexthop_counter_free(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh); + +static inline bool mlxsw_sp_l3addr_eq(const union mlxsw_sp_l3addr *addr1, + const union mlxsw_sp_l3addr *addr2) +{ + return !memcmp(addr1, addr2, sizeof(*addr1)); +} + +int mlxsw_sp_ipip_ecn_encap_init(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_ipip_ecn_decap_init(struct mlxsw_sp *mlxsw_sp); +struct net_device * +mlxsw_sp_ipip_netdev_ul_dev_get(const struct net_device *ol_dev); + +#endif /* _MLXSW_ROUTER_H_*/ diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.c new file mode 100644 index 000000000..b3472fb94 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.c @@ -0,0 +1,1741 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#include <linux/if_bridge.h> +#include <linux/list.h> +#include <linux/mutex.h> +#include <linux/refcount.h> +#include <linux/rtnetlink.h> +#include <linux/workqueue.h> +#include <net/arp.h> +#include <net/gre.h> +#include <net/lag.h> +#include <net/ndisc.h> +#include <net/ip6_tunnel.h> + +#include "spectrum.h" +#include "spectrum_ipip.h" +#include "spectrum_span.h" +#include "spectrum_switchdev.h" + +struct mlxsw_sp_span { + struct work_struct work; + struct mlxsw_sp *mlxsw_sp; + const struct mlxsw_sp_span_trigger_ops **span_trigger_ops_arr; + const struct mlxsw_sp_span_entry_ops **span_entry_ops_arr; + size_t span_entry_ops_arr_size; + struct list_head analyzed_ports_list; + struct mutex analyzed_ports_lock; /* Protects analyzed_ports_list */ + struct list_head trigger_entries_list; + u16 policer_id_base; + refcount_t policer_id_base_ref_count; + atomic_t active_entries_count; + int entries_count; + struct mlxsw_sp_span_entry entries[]; +}; + +struct mlxsw_sp_span_analyzed_port { + struct list_head list; /* Member of analyzed_ports_list */ + refcount_t ref_count; + u16 local_port; + bool ingress; +}; + +struct mlxsw_sp_span_trigger_entry { + struct list_head list; /* Member of trigger_entries_list */ + struct mlxsw_sp_span *span; + const struct mlxsw_sp_span_trigger_ops *ops; + refcount_t ref_count; + u16 local_port; + enum mlxsw_sp_span_trigger trigger; + struct mlxsw_sp_span_trigger_parms parms; +}; + +enum mlxsw_sp_span_trigger_type { + MLXSW_SP_SPAN_TRIGGER_TYPE_PORT, + MLXSW_SP_SPAN_TRIGGER_TYPE_GLOBAL, +}; + +struct mlxsw_sp_span_trigger_ops { + int (*bind)(struct mlxsw_sp_span_trigger_entry *trigger_entry); + void (*unbind)(struct mlxsw_sp_span_trigger_entry *trigger_entry); + bool (*matches)(struct mlxsw_sp_span_trigger_entry *trigger_entry, + enum mlxsw_sp_span_trigger trigger, + struct mlxsw_sp_port *mlxsw_sp_port); + int (*enable)(struct mlxsw_sp_span_trigger_entry *trigger_entry, + struct mlxsw_sp_port *mlxsw_sp_port, u8 tc); + void (*disable)(struct mlxsw_sp_span_trigger_entry *trigger_entry, + struct mlxsw_sp_port *mlxsw_sp_port, u8 tc); +}; + +static void mlxsw_sp_span_respin_work(struct work_struct *work); + +static u64 mlxsw_sp_span_occ_get(void *priv) +{ + const struct mlxsw_sp *mlxsw_sp = priv; + + return atomic_read(&mlxsw_sp->span->active_entries_count); +} + +int mlxsw_sp_span_init(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + struct mlxsw_sp_span *span; + int i, entries_count, err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, MAX_SPAN)) + return -EIO; + + entries_count = MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_SPAN); + span = kzalloc(struct_size(span, entries, entries_count), GFP_KERNEL); + if (!span) + return -ENOMEM; + refcount_set(&span->policer_id_base_ref_count, 0); + span->entries_count = entries_count; + atomic_set(&span->active_entries_count, 0); + mutex_init(&span->analyzed_ports_lock); + INIT_LIST_HEAD(&span->analyzed_ports_list); + INIT_LIST_HEAD(&span->trigger_entries_list); + span->mlxsw_sp = mlxsw_sp; + mlxsw_sp->span = span; + + for (i = 0; i < mlxsw_sp->span->entries_count; i++) + mlxsw_sp->span->entries[i].id = i; + + err = mlxsw_sp->span_ops->init(mlxsw_sp); + if (err) + goto err_init; + + devl_resource_occ_get_register(devlink, MLXSW_SP_RESOURCE_SPAN, + mlxsw_sp_span_occ_get, mlxsw_sp); + INIT_WORK(&span->work, mlxsw_sp_span_respin_work); + + return 0; + +err_init: + mutex_destroy(&mlxsw_sp->span->analyzed_ports_lock); + kfree(mlxsw_sp->span); + return err; +} + +void mlxsw_sp_span_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + + cancel_work_sync(&mlxsw_sp->span->work); + devl_resource_occ_get_unregister(devlink, MLXSW_SP_RESOURCE_SPAN); + + WARN_ON_ONCE(!list_empty(&mlxsw_sp->span->trigger_entries_list)); + WARN_ON_ONCE(!list_empty(&mlxsw_sp->span->analyzed_ports_list)); + mutex_destroy(&mlxsw_sp->span->analyzed_ports_lock); + kfree(mlxsw_sp->span); +} + +static bool mlxsw_sp1_span_cpu_can_handle(const struct net_device *dev) +{ + return !dev; +} + +static int mlxsw_sp1_span_entry_cpu_parms(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev, + struct mlxsw_sp_span_parms *sparmsp) +{ + return -EOPNOTSUPP; +} + +static int +mlxsw_sp1_span_entry_cpu_configure(struct mlxsw_sp_span_entry *span_entry, + struct mlxsw_sp_span_parms sparms) +{ + return -EOPNOTSUPP; +} + +static void +mlxsw_sp1_span_entry_cpu_deconfigure(struct mlxsw_sp_span_entry *span_entry) +{ +} + +static const +struct mlxsw_sp_span_entry_ops mlxsw_sp1_span_entry_ops_cpu = { + .is_static = true, + .can_handle = mlxsw_sp1_span_cpu_can_handle, + .parms_set = mlxsw_sp1_span_entry_cpu_parms, + .configure = mlxsw_sp1_span_entry_cpu_configure, + .deconfigure = mlxsw_sp1_span_entry_cpu_deconfigure, +}; + +static int +mlxsw_sp_span_entry_phys_parms(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev, + struct mlxsw_sp_span_parms *sparmsp) +{ + sparmsp->dest_port = netdev_priv(to_dev); + return 0; +} + +static int +mlxsw_sp_span_entry_phys_configure(struct mlxsw_sp_span_entry *span_entry, + struct mlxsw_sp_span_parms sparms) +{ + struct mlxsw_sp_port *dest_port = sparms.dest_port; + struct mlxsw_sp *mlxsw_sp = dest_port->mlxsw_sp; + u16 local_port = dest_port->local_port; + char mpat_pl[MLXSW_REG_MPAT_LEN]; + int pa_id = span_entry->id; + + /* Create a new port analayzer entry for local_port. */ + mlxsw_reg_mpat_pack(mpat_pl, pa_id, local_port, true, + MLXSW_REG_MPAT_SPAN_TYPE_LOCAL_ETH); + mlxsw_reg_mpat_session_id_set(mpat_pl, sparms.session_id); + mlxsw_reg_mpat_pide_set(mpat_pl, sparms.policer_enable); + mlxsw_reg_mpat_pid_set(mpat_pl, sparms.policer_id); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mpat), mpat_pl); +} + +static void +mlxsw_sp_span_entry_deconfigure_common(struct mlxsw_sp_span_entry *span_entry, + enum mlxsw_reg_mpat_span_type span_type) +{ + struct mlxsw_sp_port *dest_port = span_entry->parms.dest_port; + struct mlxsw_sp *mlxsw_sp = dest_port->mlxsw_sp; + u16 local_port = dest_port->local_port; + char mpat_pl[MLXSW_REG_MPAT_LEN]; + int pa_id = span_entry->id; + + mlxsw_reg_mpat_pack(mpat_pl, pa_id, local_port, false, span_type); + mlxsw_reg_mpat_session_id_set(mpat_pl, span_entry->parms.session_id); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mpat), mpat_pl); +} + +static void +mlxsw_sp_span_entry_phys_deconfigure(struct mlxsw_sp_span_entry *span_entry) +{ + mlxsw_sp_span_entry_deconfigure_common(span_entry, + MLXSW_REG_MPAT_SPAN_TYPE_LOCAL_ETH); +} + +static const +struct mlxsw_sp_span_entry_ops mlxsw_sp_span_entry_ops_phys = { + .is_static = true, + .can_handle = mlxsw_sp_port_dev_check, + .parms_set = mlxsw_sp_span_entry_phys_parms, + .configure = mlxsw_sp_span_entry_phys_configure, + .deconfigure = mlxsw_sp_span_entry_phys_deconfigure, +}; + +static int mlxsw_sp_span_dmac(struct neigh_table *tbl, + const void *pkey, + struct net_device *dev, + unsigned char dmac[ETH_ALEN]) +{ + struct neighbour *neigh = neigh_lookup(tbl, pkey, dev); + int err = 0; + + if (!neigh) { + neigh = neigh_create(tbl, pkey, dev); + if (IS_ERR(neigh)) + return PTR_ERR(neigh); + } + + neigh_event_send(neigh, NULL); + + read_lock_bh(&neigh->lock); + if ((neigh->nud_state & NUD_VALID) && !neigh->dead) + memcpy(dmac, neigh->ha, ETH_ALEN); + else + err = -ENOENT; + read_unlock_bh(&neigh->lock); + + neigh_release(neigh); + return err; +} + +static int +mlxsw_sp_span_entry_unoffloadable(struct mlxsw_sp_span_parms *sparmsp) +{ + sparmsp->dest_port = NULL; + return 0; +} + +static struct net_device * +mlxsw_sp_span_entry_bridge_8021q(const struct net_device *br_dev, + unsigned char *dmac, + u16 *p_vid) +{ + struct bridge_vlan_info vinfo; + struct net_device *edev; + u16 vid = *p_vid; + + if (!vid && WARN_ON(br_vlan_get_pvid(br_dev, &vid))) + return NULL; + if (!vid || br_vlan_get_info(br_dev, vid, &vinfo) || + !(vinfo.flags & BRIDGE_VLAN_INFO_BRENTRY)) + return NULL; + + edev = br_fdb_find_port(br_dev, dmac, vid); + if (!edev) + return NULL; + + if (br_vlan_get_info(edev, vid, &vinfo)) + return NULL; + if (vinfo.flags & BRIDGE_VLAN_INFO_UNTAGGED) + *p_vid = 0; + else + *p_vid = vid; + return edev; +} + +static struct net_device * +mlxsw_sp_span_entry_bridge_8021d(const struct net_device *br_dev, + unsigned char *dmac) +{ + return br_fdb_find_port(br_dev, dmac, 0); +} + +static struct net_device * +mlxsw_sp_span_entry_bridge(const struct net_device *br_dev, + unsigned char dmac[ETH_ALEN], + u16 *p_vid) +{ + struct mlxsw_sp_bridge_port *bridge_port; + enum mlxsw_reg_spms_state spms_state; + struct net_device *dev = NULL; + struct mlxsw_sp_port *port; + u8 stp_state; + + if (br_vlan_enabled(br_dev)) + dev = mlxsw_sp_span_entry_bridge_8021q(br_dev, dmac, p_vid); + else if (!*p_vid) + dev = mlxsw_sp_span_entry_bridge_8021d(br_dev, dmac); + if (!dev) + return NULL; + + port = mlxsw_sp_port_dev_lower_find(dev); + if (!port) + return NULL; + + bridge_port = mlxsw_sp_bridge_port_find(port->mlxsw_sp->bridge, dev); + if (!bridge_port) + return NULL; + + stp_state = mlxsw_sp_bridge_port_stp_state(bridge_port); + spms_state = mlxsw_sp_stp_spms_state(stp_state); + if (spms_state != MLXSW_REG_SPMS_STATE_FORWARDING) + return NULL; + + return dev; +} + +static struct net_device * +mlxsw_sp_span_entry_vlan(const struct net_device *vlan_dev, + u16 *p_vid) +{ + *p_vid = vlan_dev_vlan_id(vlan_dev); + return vlan_dev_real_dev(vlan_dev); +} + +static struct net_device * +mlxsw_sp_span_entry_lag(struct net_device *lag_dev) +{ + struct net_device *dev; + struct list_head *iter; + + netdev_for_each_lower_dev(lag_dev, dev, iter) + if (netif_carrier_ok(dev) && + net_lag_port_dev_txable(dev) && + mlxsw_sp_port_dev_check(dev)) + return dev; + + return NULL; +} + +static __maybe_unused int +mlxsw_sp_span_entry_tunnel_parms_common(struct net_device *edev, + union mlxsw_sp_l3addr saddr, + union mlxsw_sp_l3addr daddr, + union mlxsw_sp_l3addr gw, + __u8 ttl, + struct neigh_table *tbl, + struct mlxsw_sp_span_parms *sparmsp) +{ + unsigned char dmac[ETH_ALEN]; + u16 vid = 0; + + if (mlxsw_sp_l3addr_is_zero(gw)) + gw = daddr; + + if (!edev || mlxsw_sp_span_dmac(tbl, &gw, edev, dmac)) + goto unoffloadable; + + if (is_vlan_dev(edev)) + edev = mlxsw_sp_span_entry_vlan(edev, &vid); + + if (netif_is_bridge_master(edev)) { + edev = mlxsw_sp_span_entry_bridge(edev, dmac, &vid); + if (!edev) + goto unoffloadable; + } + + if (is_vlan_dev(edev)) { + if (vid || !(edev->flags & IFF_UP)) + goto unoffloadable; + edev = mlxsw_sp_span_entry_vlan(edev, &vid); + } + + if (netif_is_lag_master(edev)) { + if (!(edev->flags & IFF_UP)) + goto unoffloadable; + edev = mlxsw_sp_span_entry_lag(edev); + if (!edev) + goto unoffloadable; + } + + if (!mlxsw_sp_port_dev_check(edev)) + goto unoffloadable; + + sparmsp->dest_port = netdev_priv(edev); + sparmsp->ttl = ttl; + memcpy(sparmsp->dmac, dmac, ETH_ALEN); + memcpy(sparmsp->smac, edev->dev_addr, ETH_ALEN); + sparmsp->saddr = saddr; + sparmsp->daddr = daddr; + sparmsp->vid = vid; + return 0; + +unoffloadable: + return mlxsw_sp_span_entry_unoffloadable(sparmsp); +} + +#if IS_ENABLED(CONFIG_NET_IPGRE) +static struct net_device * +mlxsw_sp_span_gretap4_route(const struct net_device *to_dev, + __be32 *saddrp, __be32 *daddrp) +{ + struct ip_tunnel *tun = netdev_priv(to_dev); + struct net_device *dev = NULL; + struct ip_tunnel_parm parms; + struct rtable *rt = NULL; + struct flowi4 fl4; + + /* We assume "dev" stays valid after rt is put. */ + ASSERT_RTNL(); + + parms = mlxsw_sp_ipip_netdev_parms4(to_dev); + ip_tunnel_init_flow(&fl4, parms.iph.protocol, *daddrp, *saddrp, + 0, 0, dev_net(to_dev), parms.link, tun->fwmark, 0, + 0); + + rt = ip_route_output_key(tun->net, &fl4); + if (IS_ERR(rt)) + return NULL; + + if (rt->rt_type != RTN_UNICAST) + goto out; + + dev = rt->dst.dev; + *saddrp = fl4.saddr; + if (rt->rt_gw_family == AF_INET) + *daddrp = rt->rt_gw4; + /* can not offload if route has an IPv6 gateway */ + else if (rt->rt_gw_family == AF_INET6) + dev = NULL; + +out: + ip_rt_put(rt); + return dev; +} + +static int +mlxsw_sp_span_entry_gretap4_parms(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev, + struct mlxsw_sp_span_parms *sparmsp) +{ + struct ip_tunnel_parm tparm = mlxsw_sp_ipip_netdev_parms4(to_dev); + union mlxsw_sp_l3addr saddr = { .addr4 = tparm.iph.saddr }; + union mlxsw_sp_l3addr daddr = { .addr4 = tparm.iph.daddr }; + bool inherit_tos = tparm.iph.tos & 0x1; + bool inherit_ttl = !tparm.iph.ttl; + union mlxsw_sp_l3addr gw = daddr; + struct net_device *l3edev; + + if (!(to_dev->flags & IFF_UP) || + /* Reject tunnels with GRE keys, checksums, etc. */ + tparm.i_flags || tparm.o_flags || + /* Require a fixed TTL and a TOS copied from the mirrored packet. */ + inherit_ttl || !inherit_tos || + /* A destination address may not be "any". */ + mlxsw_sp_l3addr_is_zero(daddr)) + return mlxsw_sp_span_entry_unoffloadable(sparmsp); + + l3edev = mlxsw_sp_span_gretap4_route(to_dev, &saddr.addr4, &gw.addr4); + return mlxsw_sp_span_entry_tunnel_parms_common(l3edev, saddr, daddr, gw, + tparm.iph.ttl, + &arp_tbl, sparmsp); +} + +static int +mlxsw_sp_span_entry_gretap4_configure(struct mlxsw_sp_span_entry *span_entry, + struct mlxsw_sp_span_parms sparms) +{ + struct mlxsw_sp_port *dest_port = sparms.dest_port; + struct mlxsw_sp *mlxsw_sp = dest_port->mlxsw_sp; + u16 local_port = dest_port->local_port; + char mpat_pl[MLXSW_REG_MPAT_LEN]; + int pa_id = span_entry->id; + + /* Create a new port analayzer entry for local_port. */ + mlxsw_reg_mpat_pack(mpat_pl, pa_id, local_port, true, + MLXSW_REG_MPAT_SPAN_TYPE_REMOTE_ETH_L3); + mlxsw_reg_mpat_pide_set(mpat_pl, sparms.policer_enable); + mlxsw_reg_mpat_pid_set(mpat_pl, sparms.policer_id); + mlxsw_reg_mpat_eth_rspan_pack(mpat_pl, sparms.vid); + mlxsw_reg_mpat_eth_rspan_l2_pack(mpat_pl, + MLXSW_REG_MPAT_ETH_RSPAN_VERSION_NO_HEADER, + sparms.dmac, !!sparms.vid); + mlxsw_reg_mpat_eth_rspan_l3_ipv4_pack(mpat_pl, + sparms.ttl, sparms.smac, + be32_to_cpu(sparms.saddr.addr4), + be32_to_cpu(sparms.daddr.addr4)); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mpat), mpat_pl); +} + +static void +mlxsw_sp_span_entry_gretap4_deconfigure(struct mlxsw_sp_span_entry *span_entry) +{ + mlxsw_sp_span_entry_deconfigure_common(span_entry, + MLXSW_REG_MPAT_SPAN_TYPE_REMOTE_ETH_L3); +} + +static const struct mlxsw_sp_span_entry_ops mlxsw_sp_span_entry_ops_gretap4 = { + .can_handle = netif_is_gretap, + .parms_set = mlxsw_sp_span_entry_gretap4_parms, + .configure = mlxsw_sp_span_entry_gretap4_configure, + .deconfigure = mlxsw_sp_span_entry_gretap4_deconfigure, +}; +#endif + +#if IS_ENABLED(CONFIG_IPV6_GRE) +static struct net_device * +mlxsw_sp_span_gretap6_route(const struct net_device *to_dev, + struct in6_addr *saddrp, + struct in6_addr *daddrp) +{ + struct ip6_tnl *t = netdev_priv(to_dev); + struct flowi6 fl6 = t->fl.u.ip6; + struct net_device *dev = NULL; + struct dst_entry *dst; + struct rt6_info *rt6; + + /* We assume "dev" stays valid after dst is released. */ + ASSERT_RTNL(); + + fl6.flowi6_mark = t->parms.fwmark; + if (!ip6_tnl_xmit_ctl(t, &fl6.saddr, &fl6.daddr)) + return NULL; + + dst = ip6_route_output(t->net, NULL, &fl6); + if (!dst || dst->error) + goto out; + + rt6 = container_of(dst, struct rt6_info, dst); + + dev = dst->dev; + *saddrp = fl6.saddr; + *daddrp = rt6->rt6i_gateway; + +out: + dst_release(dst); + return dev; +} + +static int +mlxsw_sp_span_entry_gretap6_parms(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev, + struct mlxsw_sp_span_parms *sparmsp) +{ + struct __ip6_tnl_parm tparm = mlxsw_sp_ipip_netdev_parms6(to_dev); + bool inherit_tos = tparm.flags & IP6_TNL_F_USE_ORIG_TCLASS; + union mlxsw_sp_l3addr saddr = { .addr6 = tparm.laddr }; + union mlxsw_sp_l3addr daddr = { .addr6 = tparm.raddr }; + bool inherit_ttl = !tparm.hop_limit; + union mlxsw_sp_l3addr gw = daddr; + struct net_device *l3edev; + + if (!(to_dev->flags & IFF_UP) || + /* Reject tunnels with GRE keys, checksums, etc. */ + tparm.i_flags || tparm.o_flags || + /* Require a fixed TTL and a TOS copied from the mirrored packet. */ + inherit_ttl || !inherit_tos || + /* A destination address may not be "any". */ + mlxsw_sp_l3addr_is_zero(daddr)) + return mlxsw_sp_span_entry_unoffloadable(sparmsp); + + l3edev = mlxsw_sp_span_gretap6_route(to_dev, &saddr.addr6, &gw.addr6); + return mlxsw_sp_span_entry_tunnel_parms_common(l3edev, saddr, daddr, gw, + tparm.hop_limit, + &nd_tbl, sparmsp); +} + +static int +mlxsw_sp_span_entry_gretap6_configure(struct mlxsw_sp_span_entry *span_entry, + struct mlxsw_sp_span_parms sparms) +{ + struct mlxsw_sp_port *dest_port = sparms.dest_port; + struct mlxsw_sp *mlxsw_sp = dest_port->mlxsw_sp; + u16 local_port = dest_port->local_port; + char mpat_pl[MLXSW_REG_MPAT_LEN]; + int pa_id = span_entry->id; + + /* Create a new port analayzer entry for local_port. */ + mlxsw_reg_mpat_pack(mpat_pl, pa_id, local_port, true, + MLXSW_REG_MPAT_SPAN_TYPE_REMOTE_ETH_L3); + mlxsw_reg_mpat_pide_set(mpat_pl, sparms.policer_enable); + mlxsw_reg_mpat_pid_set(mpat_pl, sparms.policer_id); + mlxsw_reg_mpat_eth_rspan_pack(mpat_pl, sparms.vid); + mlxsw_reg_mpat_eth_rspan_l2_pack(mpat_pl, + MLXSW_REG_MPAT_ETH_RSPAN_VERSION_NO_HEADER, + sparms.dmac, !!sparms.vid); + mlxsw_reg_mpat_eth_rspan_l3_ipv6_pack(mpat_pl, sparms.ttl, sparms.smac, + sparms.saddr.addr6, + sparms.daddr.addr6); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mpat), mpat_pl); +} + +static void +mlxsw_sp_span_entry_gretap6_deconfigure(struct mlxsw_sp_span_entry *span_entry) +{ + mlxsw_sp_span_entry_deconfigure_common(span_entry, + MLXSW_REG_MPAT_SPAN_TYPE_REMOTE_ETH_L3); +} + +static const +struct mlxsw_sp_span_entry_ops mlxsw_sp_span_entry_ops_gretap6 = { + .can_handle = netif_is_ip6gretap, + .parms_set = mlxsw_sp_span_entry_gretap6_parms, + .configure = mlxsw_sp_span_entry_gretap6_configure, + .deconfigure = mlxsw_sp_span_entry_gretap6_deconfigure, +}; +#endif + +static bool +mlxsw_sp_span_vlan_can_handle(const struct net_device *dev) +{ + return is_vlan_dev(dev) && + mlxsw_sp_port_dev_check(vlan_dev_real_dev(dev)); +} + +static int +mlxsw_sp_span_entry_vlan_parms(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev, + struct mlxsw_sp_span_parms *sparmsp) +{ + struct net_device *real_dev; + u16 vid; + + if (!(to_dev->flags & IFF_UP)) + return mlxsw_sp_span_entry_unoffloadable(sparmsp); + + real_dev = mlxsw_sp_span_entry_vlan(to_dev, &vid); + sparmsp->dest_port = netdev_priv(real_dev); + sparmsp->vid = vid; + return 0; +} + +static int +mlxsw_sp_span_entry_vlan_configure(struct mlxsw_sp_span_entry *span_entry, + struct mlxsw_sp_span_parms sparms) +{ + struct mlxsw_sp_port *dest_port = sparms.dest_port; + struct mlxsw_sp *mlxsw_sp = dest_port->mlxsw_sp; + u16 local_port = dest_port->local_port; + char mpat_pl[MLXSW_REG_MPAT_LEN]; + int pa_id = span_entry->id; + + mlxsw_reg_mpat_pack(mpat_pl, pa_id, local_port, true, + MLXSW_REG_MPAT_SPAN_TYPE_REMOTE_ETH); + mlxsw_reg_mpat_pide_set(mpat_pl, sparms.policer_enable); + mlxsw_reg_mpat_pid_set(mpat_pl, sparms.policer_id); + mlxsw_reg_mpat_eth_rspan_pack(mpat_pl, sparms.vid); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mpat), mpat_pl); +} + +static void +mlxsw_sp_span_entry_vlan_deconfigure(struct mlxsw_sp_span_entry *span_entry) +{ + mlxsw_sp_span_entry_deconfigure_common(span_entry, + MLXSW_REG_MPAT_SPAN_TYPE_REMOTE_ETH); +} + +static const +struct mlxsw_sp_span_entry_ops mlxsw_sp_span_entry_ops_vlan = { + .can_handle = mlxsw_sp_span_vlan_can_handle, + .parms_set = mlxsw_sp_span_entry_vlan_parms, + .configure = mlxsw_sp_span_entry_vlan_configure, + .deconfigure = mlxsw_sp_span_entry_vlan_deconfigure, +}; + +static const +struct mlxsw_sp_span_entry_ops *mlxsw_sp1_span_entry_ops_arr[] = { + &mlxsw_sp1_span_entry_ops_cpu, + &mlxsw_sp_span_entry_ops_phys, +#if IS_ENABLED(CONFIG_NET_IPGRE) + &mlxsw_sp_span_entry_ops_gretap4, +#endif +#if IS_ENABLED(CONFIG_IPV6_GRE) + &mlxsw_sp_span_entry_ops_gretap6, +#endif + &mlxsw_sp_span_entry_ops_vlan, +}; + +static bool mlxsw_sp2_span_cpu_can_handle(const struct net_device *dev) +{ + return !dev; +} + +static int mlxsw_sp2_span_entry_cpu_parms(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev, + struct mlxsw_sp_span_parms *sparmsp) +{ + sparmsp->dest_port = mlxsw_sp->ports[MLXSW_PORT_CPU_PORT]; + return 0; +} + +static int +mlxsw_sp2_span_entry_cpu_configure(struct mlxsw_sp_span_entry *span_entry, + struct mlxsw_sp_span_parms sparms) +{ + /* Mirroring to the CPU port is like mirroring to any other physical + * port. Its local port is used instead of that of the physical port. + */ + return mlxsw_sp_span_entry_phys_configure(span_entry, sparms); +} + +static void +mlxsw_sp2_span_entry_cpu_deconfigure(struct mlxsw_sp_span_entry *span_entry) +{ + enum mlxsw_reg_mpat_span_type span_type; + + span_type = MLXSW_REG_MPAT_SPAN_TYPE_LOCAL_ETH; + mlxsw_sp_span_entry_deconfigure_common(span_entry, span_type); +} + +static const +struct mlxsw_sp_span_entry_ops mlxsw_sp2_span_entry_ops_cpu = { + .is_static = true, + .can_handle = mlxsw_sp2_span_cpu_can_handle, + .parms_set = mlxsw_sp2_span_entry_cpu_parms, + .configure = mlxsw_sp2_span_entry_cpu_configure, + .deconfigure = mlxsw_sp2_span_entry_cpu_deconfigure, +}; + +static const +struct mlxsw_sp_span_entry_ops *mlxsw_sp2_span_entry_ops_arr[] = { + &mlxsw_sp2_span_entry_ops_cpu, + &mlxsw_sp_span_entry_ops_phys, +#if IS_ENABLED(CONFIG_NET_IPGRE) + &mlxsw_sp_span_entry_ops_gretap4, +#endif +#if IS_ENABLED(CONFIG_IPV6_GRE) + &mlxsw_sp_span_entry_ops_gretap6, +#endif + &mlxsw_sp_span_entry_ops_vlan, +}; + +static int +mlxsw_sp_span_entry_nop_parms(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev, + struct mlxsw_sp_span_parms *sparmsp) +{ + return mlxsw_sp_span_entry_unoffloadable(sparmsp); +} + +static int +mlxsw_sp_span_entry_nop_configure(struct mlxsw_sp_span_entry *span_entry, + struct mlxsw_sp_span_parms sparms) +{ + return 0; +} + +static void +mlxsw_sp_span_entry_nop_deconfigure(struct mlxsw_sp_span_entry *span_entry) +{ +} + +static const struct mlxsw_sp_span_entry_ops mlxsw_sp_span_entry_ops_nop = { + .parms_set = mlxsw_sp_span_entry_nop_parms, + .configure = mlxsw_sp_span_entry_nop_configure, + .deconfigure = mlxsw_sp_span_entry_nop_deconfigure, +}; + +static void +mlxsw_sp_span_entry_configure(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_span_entry *span_entry, + struct mlxsw_sp_span_parms sparms) +{ + int err; + + if (!sparms.dest_port) + goto set_parms; + + if (sparms.dest_port->mlxsw_sp != mlxsw_sp) { + dev_err(mlxsw_sp->bus_info->dev, + "Cannot mirror to a port which belongs to a different mlxsw instance\n"); + sparms.dest_port = NULL; + goto set_parms; + } + + err = span_entry->ops->configure(span_entry, sparms); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to offload mirror\n"); + sparms.dest_port = NULL; + goto set_parms; + } + +set_parms: + span_entry->parms = sparms; +} + +static void +mlxsw_sp_span_entry_deconfigure(struct mlxsw_sp_span_entry *span_entry) +{ + if (span_entry->parms.dest_port) + span_entry->ops->deconfigure(span_entry); +} + +static int mlxsw_sp_span_policer_id_base_set(struct mlxsw_sp_span *span, + u16 policer_id) +{ + struct mlxsw_sp *mlxsw_sp = span->mlxsw_sp; + u16 policer_id_base; + int err; + + /* Policers set on SPAN agents must be in the range of + * `policer_id_base .. policer_id_base + max_span_agents - 1`. If the + * base is set and the new policer is not within the range, then we + * must error out. + */ + if (refcount_read(&span->policer_id_base_ref_count)) { + if (policer_id < span->policer_id_base || + policer_id >= span->policer_id_base + span->entries_count) + return -EINVAL; + + refcount_inc(&span->policer_id_base_ref_count); + return 0; + } + + /* Base must be even. */ + policer_id_base = policer_id % 2 == 0 ? policer_id : policer_id - 1; + err = mlxsw_sp->span_ops->policer_id_base_set(mlxsw_sp, + policer_id_base); + if (err) + return err; + + span->policer_id_base = policer_id_base; + refcount_set(&span->policer_id_base_ref_count, 1); + + return 0; +} + +static void mlxsw_sp_span_policer_id_base_unset(struct mlxsw_sp_span *span) +{ + if (refcount_dec_and_test(&span->policer_id_base_ref_count)) + span->policer_id_base = 0; +} + +static struct mlxsw_sp_span_entry * +mlxsw_sp_span_entry_create(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev, + const struct mlxsw_sp_span_entry_ops *ops, + struct mlxsw_sp_span_parms sparms) +{ + struct mlxsw_sp_span_entry *span_entry = NULL; + int i; + + /* find a free entry to use */ + for (i = 0; i < mlxsw_sp->span->entries_count; i++) { + if (!refcount_read(&mlxsw_sp->span->entries[i].ref_count)) { + span_entry = &mlxsw_sp->span->entries[i]; + break; + } + } + if (!span_entry) + return NULL; + + if (sparms.policer_enable) { + int err; + + err = mlxsw_sp_span_policer_id_base_set(mlxsw_sp->span, + sparms.policer_id); + if (err) + return NULL; + } + + atomic_inc(&mlxsw_sp->span->active_entries_count); + span_entry->ops = ops; + refcount_set(&span_entry->ref_count, 1); + span_entry->to_dev = to_dev; + mlxsw_sp_span_entry_configure(mlxsw_sp, span_entry, sparms); + + return span_entry; +} + +static void mlxsw_sp_span_entry_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_span_entry *span_entry) +{ + mlxsw_sp_span_entry_deconfigure(span_entry); + atomic_dec(&mlxsw_sp->span->active_entries_count); + if (span_entry->parms.policer_enable) + mlxsw_sp_span_policer_id_base_unset(mlxsw_sp->span); +} + +struct mlxsw_sp_span_entry * +mlxsw_sp_span_entry_find_by_port(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev) +{ + int i; + + for (i = 0; i < mlxsw_sp->span->entries_count; i++) { + struct mlxsw_sp_span_entry *curr = &mlxsw_sp->span->entries[i]; + + if (refcount_read(&curr->ref_count) && curr->to_dev == to_dev) + return curr; + } + return NULL; +} + +void mlxsw_sp_span_entry_invalidate(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_span_entry *span_entry) +{ + mlxsw_sp_span_entry_deconfigure(span_entry); + span_entry->ops = &mlxsw_sp_span_entry_ops_nop; +} + +static struct mlxsw_sp_span_entry * +mlxsw_sp_span_entry_find_by_id(struct mlxsw_sp *mlxsw_sp, int span_id) +{ + int i; + + for (i = 0; i < mlxsw_sp->span->entries_count; i++) { + struct mlxsw_sp_span_entry *curr = &mlxsw_sp->span->entries[i]; + + if (refcount_read(&curr->ref_count) && curr->id == span_id) + return curr; + } + return NULL; +} + +static struct mlxsw_sp_span_entry * +mlxsw_sp_span_entry_find_by_parms(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev, + const struct mlxsw_sp_span_parms *sparms) +{ + int i; + + for (i = 0; i < mlxsw_sp->span->entries_count; i++) { + struct mlxsw_sp_span_entry *curr = &mlxsw_sp->span->entries[i]; + + if (refcount_read(&curr->ref_count) && curr->to_dev == to_dev && + curr->parms.policer_enable == sparms->policer_enable && + curr->parms.policer_id == sparms->policer_id && + curr->parms.session_id == sparms->session_id) + return curr; + } + return NULL; +} + +static struct mlxsw_sp_span_entry * +mlxsw_sp_span_entry_get(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev, + const struct mlxsw_sp_span_entry_ops *ops, + struct mlxsw_sp_span_parms sparms) +{ + struct mlxsw_sp_span_entry *span_entry; + + span_entry = mlxsw_sp_span_entry_find_by_parms(mlxsw_sp, to_dev, + &sparms); + if (span_entry) { + /* Already exists, just take a reference */ + refcount_inc(&span_entry->ref_count); + return span_entry; + } + + return mlxsw_sp_span_entry_create(mlxsw_sp, to_dev, ops, sparms); +} + +static int mlxsw_sp_span_entry_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_span_entry *span_entry) +{ + if (refcount_dec_and_test(&span_entry->ref_count)) + mlxsw_sp_span_entry_destroy(mlxsw_sp, span_entry); + return 0; +} + +static int mlxsw_sp_span_port_buffer_update(struct mlxsw_sp_port *mlxsw_sp_port, bool enable) +{ + struct mlxsw_sp_hdroom hdroom; + + hdroom = *mlxsw_sp_port->hdroom; + hdroom.int_buf.enable = enable; + mlxsw_sp_hdroom_bufs_reset_sizes(mlxsw_sp_port, &hdroom); + + return mlxsw_sp_hdroom_configure(mlxsw_sp_port, &hdroom); +} + +static int +mlxsw_sp_span_port_buffer_enable(struct mlxsw_sp_port *mlxsw_sp_port) +{ + return mlxsw_sp_span_port_buffer_update(mlxsw_sp_port, true); +} + +static void mlxsw_sp_span_port_buffer_disable(struct mlxsw_sp_port *mlxsw_sp_port) +{ + mlxsw_sp_span_port_buffer_update(mlxsw_sp_port, false); +} + +static struct mlxsw_sp_span_analyzed_port * +mlxsw_sp_span_analyzed_port_find(struct mlxsw_sp_span *span, u16 local_port, + bool ingress) +{ + struct mlxsw_sp_span_analyzed_port *analyzed_port; + + list_for_each_entry(analyzed_port, &span->analyzed_ports_list, list) { + if (analyzed_port->local_port == local_port && + analyzed_port->ingress == ingress) + return analyzed_port; + } + + return NULL; +} + +static const struct mlxsw_sp_span_entry_ops * +mlxsw_sp_span_entry_ops(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev) +{ + struct mlxsw_sp_span *span = mlxsw_sp->span; + size_t i; + + for (i = 0; i < span->span_entry_ops_arr_size; ++i) + if (span->span_entry_ops_arr[i]->can_handle(to_dev)) + return span->span_entry_ops_arr[i]; + + return NULL; +} + +static void mlxsw_sp_span_respin_work(struct work_struct *work) +{ + struct mlxsw_sp_span *span; + struct mlxsw_sp *mlxsw_sp; + int i, err; + + span = container_of(work, struct mlxsw_sp_span, work); + mlxsw_sp = span->mlxsw_sp; + + rtnl_lock(); + for (i = 0; i < mlxsw_sp->span->entries_count; i++) { + struct mlxsw_sp_span_entry *curr = &mlxsw_sp->span->entries[i]; + struct mlxsw_sp_span_parms sparms = {NULL}; + + if (!refcount_read(&curr->ref_count)) + continue; + + if (curr->ops->is_static) + continue; + + err = curr->ops->parms_set(mlxsw_sp, curr->to_dev, &sparms); + if (err) + continue; + + if (memcmp(&sparms, &curr->parms, sizeof(sparms))) { + mlxsw_sp_span_entry_deconfigure(curr); + mlxsw_sp_span_entry_configure(mlxsw_sp, curr, sparms); + } + } + rtnl_unlock(); +} + +void mlxsw_sp_span_respin(struct mlxsw_sp *mlxsw_sp) +{ + if (atomic_read(&mlxsw_sp->span->active_entries_count) == 0) + return; + mlxsw_core_schedule_work(&mlxsw_sp->span->work); +} + +int mlxsw_sp_span_agent_get(struct mlxsw_sp *mlxsw_sp, int *p_span_id, + const struct mlxsw_sp_span_agent_parms *parms) +{ + const struct net_device *to_dev = parms->to_dev; + const struct mlxsw_sp_span_entry_ops *ops; + struct mlxsw_sp_span_entry *span_entry; + struct mlxsw_sp_span_parms sparms; + int err; + + ASSERT_RTNL(); + + ops = mlxsw_sp_span_entry_ops(mlxsw_sp, to_dev); + if (!ops) { + dev_err(mlxsw_sp->bus_info->dev, "Cannot mirror to requested destination\n"); + return -EOPNOTSUPP; + } + + memset(&sparms, 0, sizeof(sparms)); + err = ops->parms_set(mlxsw_sp, to_dev, &sparms); + if (err) + return err; + + sparms.policer_id = parms->policer_id; + sparms.policer_enable = parms->policer_enable; + sparms.session_id = parms->session_id; + span_entry = mlxsw_sp_span_entry_get(mlxsw_sp, to_dev, ops, sparms); + if (!span_entry) + return -ENOBUFS; + + *p_span_id = span_entry->id; + + return 0; +} + +void mlxsw_sp_span_agent_put(struct mlxsw_sp *mlxsw_sp, int span_id) +{ + struct mlxsw_sp_span_entry *span_entry; + + ASSERT_RTNL(); + + span_entry = mlxsw_sp_span_entry_find_by_id(mlxsw_sp, span_id); + if (WARN_ON_ONCE(!span_entry)) + return; + + mlxsw_sp_span_entry_put(mlxsw_sp, span_entry); +} + +static struct mlxsw_sp_span_analyzed_port * +mlxsw_sp_span_analyzed_port_create(struct mlxsw_sp_span *span, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + struct mlxsw_sp_span_analyzed_port *analyzed_port; + int err; + + analyzed_port = kzalloc(sizeof(*analyzed_port), GFP_KERNEL); + if (!analyzed_port) + return ERR_PTR(-ENOMEM); + + refcount_set(&analyzed_port->ref_count, 1); + analyzed_port->local_port = mlxsw_sp_port->local_port; + analyzed_port->ingress = ingress; + list_add_tail(&analyzed_port->list, &span->analyzed_ports_list); + + /* An egress mirror buffer should be allocated on the egress port which + * does the mirroring. + */ + if (!ingress) { + err = mlxsw_sp_span_port_buffer_enable(mlxsw_sp_port); + if (err) + goto err_buffer_update; + } + + return analyzed_port; + +err_buffer_update: + list_del(&analyzed_port->list); + kfree(analyzed_port); + return ERR_PTR(err); +} + +static void +mlxsw_sp_span_analyzed_port_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_span_analyzed_port * + analyzed_port) +{ + /* Remove egress mirror buffer now that port is no longer analyzed + * at egress. + */ + if (!analyzed_port->ingress) + mlxsw_sp_span_port_buffer_disable(mlxsw_sp_port); + + list_del(&analyzed_port->list); + kfree(analyzed_port); +} + +int mlxsw_sp_span_analyzed_port_get(struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_span_analyzed_port *analyzed_port; + u16 local_port = mlxsw_sp_port->local_port; + int err = 0; + + mutex_lock(&mlxsw_sp->span->analyzed_ports_lock); + + analyzed_port = mlxsw_sp_span_analyzed_port_find(mlxsw_sp->span, + local_port, ingress); + if (analyzed_port) { + refcount_inc(&analyzed_port->ref_count); + goto out_unlock; + } + + analyzed_port = mlxsw_sp_span_analyzed_port_create(mlxsw_sp->span, + mlxsw_sp_port, + ingress); + if (IS_ERR(analyzed_port)) + err = PTR_ERR(analyzed_port); + +out_unlock: + mutex_unlock(&mlxsw_sp->span->analyzed_ports_lock); + return err; +} + +void mlxsw_sp_span_analyzed_port_put(struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_span_analyzed_port *analyzed_port; + u16 local_port = mlxsw_sp_port->local_port; + + mutex_lock(&mlxsw_sp->span->analyzed_ports_lock); + + analyzed_port = mlxsw_sp_span_analyzed_port_find(mlxsw_sp->span, + local_port, ingress); + if (WARN_ON_ONCE(!analyzed_port)) + goto out_unlock; + + if (!refcount_dec_and_test(&analyzed_port->ref_count)) + goto out_unlock; + + mlxsw_sp_span_analyzed_port_destroy(mlxsw_sp_port, analyzed_port); + +out_unlock: + mutex_unlock(&mlxsw_sp->span->analyzed_ports_lock); +} + +static int +__mlxsw_sp_span_trigger_port_bind(struct mlxsw_sp_span *span, + struct mlxsw_sp_span_trigger_entry * + trigger_entry, bool enable) +{ + char mpar_pl[MLXSW_REG_MPAR_LEN]; + enum mlxsw_reg_mpar_i_e i_e; + + switch (trigger_entry->trigger) { + case MLXSW_SP_SPAN_TRIGGER_INGRESS: + i_e = MLXSW_REG_MPAR_TYPE_INGRESS; + break; + case MLXSW_SP_SPAN_TRIGGER_EGRESS: + i_e = MLXSW_REG_MPAR_TYPE_EGRESS; + break; + default: + WARN_ON_ONCE(1); + return -EINVAL; + } + + if (trigger_entry->parms.probability_rate > MLXSW_REG_MPAR_RATE_MAX) + return -EINVAL; + + mlxsw_reg_mpar_pack(mpar_pl, trigger_entry->local_port, i_e, enable, + trigger_entry->parms.span_id, + trigger_entry->parms.probability_rate); + return mlxsw_reg_write(span->mlxsw_sp->core, MLXSW_REG(mpar), mpar_pl); +} + +static int +mlxsw_sp_span_trigger_port_bind(struct mlxsw_sp_span_trigger_entry * + trigger_entry) +{ + return __mlxsw_sp_span_trigger_port_bind(trigger_entry->span, + trigger_entry, true); +} + +static void +mlxsw_sp_span_trigger_port_unbind(struct mlxsw_sp_span_trigger_entry * + trigger_entry) +{ + __mlxsw_sp_span_trigger_port_bind(trigger_entry->span, trigger_entry, + false); +} + +static bool +mlxsw_sp_span_trigger_port_matches(struct mlxsw_sp_span_trigger_entry * + trigger_entry, + enum mlxsw_sp_span_trigger trigger, + struct mlxsw_sp_port *mlxsw_sp_port) +{ + return trigger_entry->trigger == trigger && + trigger_entry->local_port == mlxsw_sp_port->local_port; +} + +static int +mlxsw_sp_span_trigger_port_enable(struct mlxsw_sp_span_trigger_entry * + trigger_entry, + struct mlxsw_sp_port *mlxsw_sp_port, u8 tc) +{ + /* Port trigger are enabled during binding. */ + return 0; +} + +static void +mlxsw_sp_span_trigger_port_disable(struct mlxsw_sp_span_trigger_entry * + trigger_entry, + struct mlxsw_sp_port *mlxsw_sp_port, u8 tc) +{ +} + +static const struct mlxsw_sp_span_trigger_ops +mlxsw_sp_span_trigger_port_ops = { + .bind = mlxsw_sp_span_trigger_port_bind, + .unbind = mlxsw_sp_span_trigger_port_unbind, + .matches = mlxsw_sp_span_trigger_port_matches, + .enable = mlxsw_sp_span_trigger_port_enable, + .disable = mlxsw_sp_span_trigger_port_disable, +}; + +static int +mlxsw_sp1_span_trigger_global_bind(struct mlxsw_sp_span_trigger_entry * + trigger_entry) +{ + return -EOPNOTSUPP; +} + +static void +mlxsw_sp1_span_trigger_global_unbind(struct mlxsw_sp_span_trigger_entry * + trigger_entry) +{ +} + +static bool +mlxsw_sp1_span_trigger_global_matches(struct mlxsw_sp_span_trigger_entry * + trigger_entry, + enum mlxsw_sp_span_trigger trigger, + struct mlxsw_sp_port *mlxsw_sp_port) +{ + WARN_ON_ONCE(1); + return false; +} + +static int +mlxsw_sp1_span_trigger_global_enable(struct mlxsw_sp_span_trigger_entry * + trigger_entry, + struct mlxsw_sp_port *mlxsw_sp_port, + u8 tc) +{ + return -EOPNOTSUPP; +} + +static void +mlxsw_sp1_span_trigger_global_disable(struct mlxsw_sp_span_trigger_entry * + trigger_entry, + struct mlxsw_sp_port *mlxsw_sp_port, + u8 tc) +{ +} + +static const struct mlxsw_sp_span_trigger_ops +mlxsw_sp1_span_trigger_global_ops = { + .bind = mlxsw_sp1_span_trigger_global_bind, + .unbind = mlxsw_sp1_span_trigger_global_unbind, + .matches = mlxsw_sp1_span_trigger_global_matches, + .enable = mlxsw_sp1_span_trigger_global_enable, + .disable = mlxsw_sp1_span_trigger_global_disable, +}; + +static const struct mlxsw_sp_span_trigger_ops * +mlxsw_sp1_span_trigger_ops_arr[] = { + [MLXSW_SP_SPAN_TRIGGER_TYPE_PORT] = &mlxsw_sp_span_trigger_port_ops, + [MLXSW_SP_SPAN_TRIGGER_TYPE_GLOBAL] = + &mlxsw_sp1_span_trigger_global_ops, +}; + +static int +mlxsw_sp2_span_trigger_global_bind(struct mlxsw_sp_span_trigger_entry * + trigger_entry) +{ + struct mlxsw_sp *mlxsw_sp = trigger_entry->span->mlxsw_sp; + enum mlxsw_reg_mpagr_trigger trigger; + char mpagr_pl[MLXSW_REG_MPAGR_LEN]; + + switch (trigger_entry->trigger) { + case MLXSW_SP_SPAN_TRIGGER_TAIL_DROP: + trigger = MLXSW_REG_MPAGR_TRIGGER_INGRESS_SHARED_BUFFER; + break; + case MLXSW_SP_SPAN_TRIGGER_EARLY_DROP: + trigger = MLXSW_REG_MPAGR_TRIGGER_INGRESS_WRED; + break; + case MLXSW_SP_SPAN_TRIGGER_ECN: + trigger = MLXSW_REG_MPAGR_TRIGGER_EGRESS_ECN; + break; + default: + WARN_ON_ONCE(1); + return -EINVAL; + } + + if (trigger_entry->parms.probability_rate > MLXSW_REG_MPAGR_RATE_MAX) + return -EINVAL; + + mlxsw_reg_mpagr_pack(mpagr_pl, trigger, trigger_entry->parms.span_id, + trigger_entry->parms.probability_rate); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mpagr), mpagr_pl); +} + +static void +mlxsw_sp2_span_trigger_global_unbind(struct mlxsw_sp_span_trigger_entry * + trigger_entry) +{ + /* There is no unbinding for global triggers. The trigger should be + * disabled on all ports by now. + */ +} + +static bool +mlxsw_sp2_span_trigger_global_matches(struct mlxsw_sp_span_trigger_entry * + trigger_entry, + enum mlxsw_sp_span_trigger trigger, + struct mlxsw_sp_port *mlxsw_sp_port) +{ + return trigger_entry->trigger == trigger; +} + +static int +__mlxsw_sp2_span_trigger_global_enable(struct mlxsw_sp_span_trigger_entry * + trigger_entry, + struct mlxsw_sp_port *mlxsw_sp_port, + u8 tc, bool enable) +{ + struct mlxsw_sp *mlxsw_sp = trigger_entry->span->mlxsw_sp; + char momte_pl[MLXSW_REG_MOMTE_LEN]; + enum mlxsw_reg_momte_type type; + int err; + + switch (trigger_entry->trigger) { + case MLXSW_SP_SPAN_TRIGGER_TAIL_DROP: + type = MLXSW_REG_MOMTE_TYPE_SHARED_BUFFER_TCLASS; + break; + case MLXSW_SP_SPAN_TRIGGER_EARLY_DROP: + type = MLXSW_REG_MOMTE_TYPE_WRED; + break; + case MLXSW_SP_SPAN_TRIGGER_ECN: + type = MLXSW_REG_MOMTE_TYPE_ECN; + break; + default: + WARN_ON_ONCE(1); + return -EINVAL; + } + + /* Query existing configuration in order to only change the state of + * the specified traffic class. + */ + mlxsw_reg_momte_pack(momte_pl, mlxsw_sp_port->local_port, type); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(momte), momte_pl); + if (err) + return err; + + mlxsw_reg_momte_tclass_en_set(momte_pl, tc, enable); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(momte), momte_pl); +} + +static int +mlxsw_sp2_span_trigger_global_enable(struct mlxsw_sp_span_trigger_entry * + trigger_entry, + struct mlxsw_sp_port *mlxsw_sp_port, + u8 tc) +{ + return __mlxsw_sp2_span_trigger_global_enable(trigger_entry, + mlxsw_sp_port, tc, true); +} + +static void +mlxsw_sp2_span_trigger_global_disable(struct mlxsw_sp_span_trigger_entry * + trigger_entry, + struct mlxsw_sp_port *mlxsw_sp_port, + u8 tc) +{ + __mlxsw_sp2_span_trigger_global_enable(trigger_entry, mlxsw_sp_port, tc, + false); +} + +static const struct mlxsw_sp_span_trigger_ops +mlxsw_sp2_span_trigger_global_ops = { + .bind = mlxsw_sp2_span_trigger_global_bind, + .unbind = mlxsw_sp2_span_trigger_global_unbind, + .matches = mlxsw_sp2_span_trigger_global_matches, + .enable = mlxsw_sp2_span_trigger_global_enable, + .disable = mlxsw_sp2_span_trigger_global_disable, +}; + +static const struct mlxsw_sp_span_trigger_ops * +mlxsw_sp2_span_trigger_ops_arr[] = { + [MLXSW_SP_SPAN_TRIGGER_TYPE_PORT] = &mlxsw_sp_span_trigger_port_ops, + [MLXSW_SP_SPAN_TRIGGER_TYPE_GLOBAL] = + &mlxsw_sp2_span_trigger_global_ops, +}; + +static void +mlxsw_sp_span_trigger_ops_set(struct mlxsw_sp_span_trigger_entry *trigger_entry) +{ + struct mlxsw_sp_span *span = trigger_entry->span; + enum mlxsw_sp_span_trigger_type type; + + switch (trigger_entry->trigger) { + case MLXSW_SP_SPAN_TRIGGER_INGRESS: + case MLXSW_SP_SPAN_TRIGGER_EGRESS: + type = MLXSW_SP_SPAN_TRIGGER_TYPE_PORT; + break; + case MLXSW_SP_SPAN_TRIGGER_TAIL_DROP: + case MLXSW_SP_SPAN_TRIGGER_EARLY_DROP: + case MLXSW_SP_SPAN_TRIGGER_ECN: + type = MLXSW_SP_SPAN_TRIGGER_TYPE_GLOBAL; + break; + default: + WARN_ON_ONCE(1); + return; + } + + trigger_entry->ops = span->span_trigger_ops_arr[type]; +} + +static struct mlxsw_sp_span_trigger_entry * +mlxsw_sp_span_trigger_entry_create(struct mlxsw_sp_span *span, + enum mlxsw_sp_span_trigger trigger, + struct mlxsw_sp_port *mlxsw_sp_port, + const struct mlxsw_sp_span_trigger_parms + *parms) +{ + struct mlxsw_sp_span_trigger_entry *trigger_entry; + int err; + + trigger_entry = kzalloc(sizeof(*trigger_entry), GFP_KERNEL); + if (!trigger_entry) + return ERR_PTR(-ENOMEM); + + refcount_set(&trigger_entry->ref_count, 1); + trigger_entry->local_port = mlxsw_sp_port ? mlxsw_sp_port->local_port : + 0; + trigger_entry->trigger = trigger; + memcpy(&trigger_entry->parms, parms, sizeof(trigger_entry->parms)); + trigger_entry->span = span; + mlxsw_sp_span_trigger_ops_set(trigger_entry); + list_add_tail(&trigger_entry->list, &span->trigger_entries_list); + + err = trigger_entry->ops->bind(trigger_entry); + if (err) + goto err_trigger_entry_bind; + + return trigger_entry; + +err_trigger_entry_bind: + list_del(&trigger_entry->list); + kfree(trigger_entry); + return ERR_PTR(err); +} + +static void +mlxsw_sp_span_trigger_entry_destroy(struct mlxsw_sp_span *span, + struct mlxsw_sp_span_trigger_entry * + trigger_entry) +{ + trigger_entry->ops->unbind(trigger_entry); + list_del(&trigger_entry->list); + kfree(trigger_entry); +} + +static struct mlxsw_sp_span_trigger_entry * +mlxsw_sp_span_trigger_entry_find(struct mlxsw_sp_span *span, + enum mlxsw_sp_span_trigger trigger, + struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp_span_trigger_entry *trigger_entry; + + list_for_each_entry(trigger_entry, &span->trigger_entries_list, list) { + if (trigger_entry->ops->matches(trigger_entry, trigger, + mlxsw_sp_port)) + return trigger_entry; + } + + return NULL; +} + +int mlxsw_sp_span_agent_bind(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_span_trigger trigger, + struct mlxsw_sp_port *mlxsw_sp_port, + const struct mlxsw_sp_span_trigger_parms *parms) +{ + struct mlxsw_sp_span_trigger_entry *trigger_entry; + int err = 0; + + ASSERT_RTNL(); + + if (!mlxsw_sp_span_entry_find_by_id(mlxsw_sp, parms->span_id)) + return -EINVAL; + + trigger_entry = mlxsw_sp_span_trigger_entry_find(mlxsw_sp->span, + trigger, + mlxsw_sp_port); + if (trigger_entry) { + if (trigger_entry->parms.span_id != parms->span_id || + trigger_entry->parms.probability_rate != + parms->probability_rate) + return -EINVAL; + refcount_inc(&trigger_entry->ref_count); + goto out; + } + + trigger_entry = mlxsw_sp_span_trigger_entry_create(mlxsw_sp->span, + trigger, + mlxsw_sp_port, + parms); + if (IS_ERR(trigger_entry)) + err = PTR_ERR(trigger_entry); + +out: + return err; +} + +void mlxsw_sp_span_agent_unbind(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_span_trigger trigger, + struct mlxsw_sp_port *mlxsw_sp_port, + const struct mlxsw_sp_span_trigger_parms *parms) +{ + struct mlxsw_sp_span_trigger_entry *trigger_entry; + + ASSERT_RTNL(); + + if (WARN_ON_ONCE(!mlxsw_sp_span_entry_find_by_id(mlxsw_sp, + parms->span_id))) + return; + + trigger_entry = mlxsw_sp_span_trigger_entry_find(mlxsw_sp->span, + trigger, + mlxsw_sp_port); + if (WARN_ON_ONCE(!trigger_entry)) + return; + + if (!refcount_dec_and_test(&trigger_entry->ref_count)) + return; + + mlxsw_sp_span_trigger_entry_destroy(mlxsw_sp->span, trigger_entry); +} + +int mlxsw_sp_span_trigger_enable(struct mlxsw_sp_port *mlxsw_sp_port, + enum mlxsw_sp_span_trigger trigger, u8 tc) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_span_trigger_entry *trigger_entry; + + ASSERT_RTNL(); + + trigger_entry = mlxsw_sp_span_trigger_entry_find(mlxsw_sp->span, + trigger, + mlxsw_sp_port); + if (WARN_ON_ONCE(!trigger_entry)) + return -EINVAL; + + return trigger_entry->ops->enable(trigger_entry, mlxsw_sp_port, tc); +} + +void mlxsw_sp_span_trigger_disable(struct mlxsw_sp_port *mlxsw_sp_port, + enum mlxsw_sp_span_trigger trigger, u8 tc) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_span_trigger_entry *trigger_entry; + + ASSERT_RTNL(); + + trigger_entry = mlxsw_sp_span_trigger_entry_find(mlxsw_sp->span, + trigger, + mlxsw_sp_port); + if (WARN_ON_ONCE(!trigger_entry)) + return; + + return trigger_entry->ops->disable(trigger_entry, mlxsw_sp_port, tc); +} + +bool mlxsw_sp_span_trigger_is_ingress(enum mlxsw_sp_span_trigger trigger) +{ + switch (trigger) { + case MLXSW_SP_SPAN_TRIGGER_INGRESS: + case MLXSW_SP_SPAN_TRIGGER_EARLY_DROP: + case MLXSW_SP_SPAN_TRIGGER_TAIL_DROP: + return true; + case MLXSW_SP_SPAN_TRIGGER_EGRESS: + case MLXSW_SP_SPAN_TRIGGER_ECN: + return false; + } + + WARN_ON_ONCE(1); + return false; +} + +static int mlxsw_sp1_span_init(struct mlxsw_sp *mlxsw_sp) +{ + size_t arr_size = ARRAY_SIZE(mlxsw_sp1_span_entry_ops_arr); + + /* Must be first to avoid NULL pointer dereference by subsequent + * can_handle() callbacks. + */ + if (WARN_ON(mlxsw_sp1_span_entry_ops_arr[0] != + &mlxsw_sp1_span_entry_ops_cpu)) + return -EINVAL; + + mlxsw_sp->span->span_trigger_ops_arr = mlxsw_sp1_span_trigger_ops_arr; + mlxsw_sp->span->span_entry_ops_arr = mlxsw_sp1_span_entry_ops_arr; + mlxsw_sp->span->span_entry_ops_arr_size = arr_size; + + return 0; +} + +static int mlxsw_sp1_span_policer_id_base_set(struct mlxsw_sp *mlxsw_sp, + u16 policer_id_base) +{ + return -EOPNOTSUPP; +} + +const struct mlxsw_sp_span_ops mlxsw_sp1_span_ops = { + .init = mlxsw_sp1_span_init, + .policer_id_base_set = mlxsw_sp1_span_policer_id_base_set, +}; + +static int mlxsw_sp2_span_init(struct mlxsw_sp *mlxsw_sp) +{ + size_t arr_size = ARRAY_SIZE(mlxsw_sp2_span_entry_ops_arr); + + /* Must be first to avoid NULL pointer dereference by subsequent + * can_handle() callbacks. + */ + if (WARN_ON(mlxsw_sp2_span_entry_ops_arr[0] != + &mlxsw_sp2_span_entry_ops_cpu)) + return -EINVAL; + + mlxsw_sp->span->span_trigger_ops_arr = mlxsw_sp2_span_trigger_ops_arr; + mlxsw_sp->span->span_entry_ops_arr = mlxsw_sp2_span_entry_ops_arr; + mlxsw_sp->span->span_entry_ops_arr_size = arr_size; + + return 0; +} + +#define MLXSW_SP2_SPAN_EG_MIRROR_BUFFER_FACTOR 38 +#define MLXSW_SP3_SPAN_EG_MIRROR_BUFFER_FACTOR 50 + +static int mlxsw_sp2_span_policer_id_base_set(struct mlxsw_sp *mlxsw_sp, + u16 policer_id_base) +{ + char mogcr_pl[MLXSW_REG_MOGCR_LEN]; + int err; + + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(mogcr), mogcr_pl); + if (err) + return err; + + mlxsw_reg_mogcr_mirroring_pid_base_set(mogcr_pl, policer_id_base); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mogcr), mogcr_pl); +} + +const struct mlxsw_sp_span_ops mlxsw_sp2_span_ops = { + .init = mlxsw_sp2_span_init, + .policer_id_base_set = mlxsw_sp2_span_policer_id_base_set, +}; + +const struct mlxsw_sp_span_ops mlxsw_sp3_span_ops = { + .init = mlxsw_sp2_span_init, + .policer_id_base_set = mlxsw_sp2_span_policer_id_base_set, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.h new file mode 100644 index 000000000..82e711afb --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.h @@ -0,0 +1,129 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_SPECTRUM_SPAN_H +#define _MLXSW_SPECTRUM_SPAN_H + +#include <linux/types.h> +#include <linux/if_ether.h> +#include <linux/refcount.h> + +#include "spectrum_router.h" + +struct mlxsw_sp; +struct mlxsw_sp_port; + +/* SPAN session identifiers that correspond to MLXSW_TRAP_ID_MIRROR_SESSION<i> + * trap identifiers. The session identifier is an attribute of the SPAN agent, + * which determines the trap identifier of packets that are mirrored to the + * CPU. Packets that are trapped to the CPU for the same logical reason (e.g., + * buffer drops) should use the same session identifier. + */ +enum mlxsw_sp_span_session_id { + MLXSW_SP_SPAN_SESSION_ID_BUFFER, + MLXSW_SP_SPAN_SESSION_ID_SAMPLING, + + __MLXSW_SP_SPAN_SESSION_ID_MAX = 8, +}; + +struct mlxsw_sp_span_parms { + struct mlxsw_sp_port *dest_port; /* NULL for unoffloaded SPAN. */ + unsigned int ttl; + unsigned char dmac[ETH_ALEN]; + unsigned char smac[ETH_ALEN]; + union mlxsw_sp_l3addr daddr; + union mlxsw_sp_l3addr saddr; + u16 vid; + u16 policer_id; + bool policer_enable; + enum mlxsw_sp_span_session_id session_id; +}; + +enum mlxsw_sp_span_trigger { + MLXSW_SP_SPAN_TRIGGER_INGRESS, + MLXSW_SP_SPAN_TRIGGER_EGRESS, + MLXSW_SP_SPAN_TRIGGER_TAIL_DROP, + MLXSW_SP_SPAN_TRIGGER_EARLY_DROP, + MLXSW_SP_SPAN_TRIGGER_ECN, +}; + +struct mlxsw_sp_span_trigger_parms { + int span_id; + u32 probability_rate; +}; + +struct mlxsw_sp_span_agent_parms { + const struct net_device *to_dev; + u16 policer_id; + bool policer_enable; + enum mlxsw_sp_span_session_id session_id; +}; + +struct mlxsw_sp_span_entry_ops; + +struct mlxsw_sp_span_ops { + int (*init)(struct mlxsw_sp *mlxsw_sp); + int (*policer_id_base_set)(struct mlxsw_sp *mlxsw_sp, + u16 policer_id_base); +}; + +struct mlxsw_sp_span_entry { + const struct net_device *to_dev; + const struct mlxsw_sp_span_entry_ops *ops; + struct mlxsw_sp_span_parms parms; + refcount_t ref_count; + int id; +}; + +struct mlxsw_sp_span_entry_ops { + bool is_static; + bool (*can_handle)(const struct net_device *to_dev); + int (*parms_set)(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev, + struct mlxsw_sp_span_parms *sparmsp); + int (*configure)(struct mlxsw_sp_span_entry *span_entry, + struct mlxsw_sp_span_parms sparms); + void (*deconfigure)(struct mlxsw_sp_span_entry *span_entry); +}; + +int mlxsw_sp_span_init(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_span_fini(struct mlxsw_sp *mlxsw_sp); +void mlxsw_sp_span_respin(struct mlxsw_sp *mlxsw_sp); + +struct mlxsw_sp_span_entry * +mlxsw_sp_span_entry_find_by_port(struct mlxsw_sp *mlxsw_sp, + const struct net_device *to_dev); + +void mlxsw_sp_span_entry_invalidate(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_span_entry *span_entry); + +int mlxsw_sp_span_port_mtu_update(struct mlxsw_sp_port *port, u16 mtu); +void mlxsw_sp_span_speed_update_work(struct work_struct *work); + +int mlxsw_sp_span_agent_get(struct mlxsw_sp *mlxsw_sp, int *p_span_id, + const struct mlxsw_sp_span_agent_parms *parms); +void mlxsw_sp_span_agent_put(struct mlxsw_sp *mlxsw_sp, int span_id); +int mlxsw_sp_span_analyzed_port_get(struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress); +void mlxsw_sp_span_analyzed_port_put(struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress); +int mlxsw_sp_span_agent_bind(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_span_trigger trigger, + struct mlxsw_sp_port *mlxsw_sp_port, + const struct mlxsw_sp_span_trigger_parms *parms); +void +mlxsw_sp_span_agent_unbind(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_sp_span_trigger trigger, + struct mlxsw_sp_port *mlxsw_sp_port, + const struct mlxsw_sp_span_trigger_parms *parms); +int mlxsw_sp_span_trigger_enable(struct mlxsw_sp_port *mlxsw_sp_port, + enum mlxsw_sp_span_trigger trigger, u8 tc); +void mlxsw_sp_span_trigger_disable(struct mlxsw_sp_port *mlxsw_sp_port, + enum mlxsw_sp_span_trigger trigger, u8 tc); +bool mlxsw_sp_span_trigger_is_ingress(enum mlxsw_sp_span_trigger trigger); + +extern const struct mlxsw_sp_span_ops mlxsw_sp1_span_ops; +extern const struct mlxsw_sp_span_ops mlxsw_sp2_span_ops; +extern const struct mlxsw_sp_span_ops mlxsw_sp3_span_ops; + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c new file mode 100644 index 000000000..1290b2d3e --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c @@ -0,0 +1,4015 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#include <linux/kernel.h> +#include <linux/types.h> +#include <linux/netdevice.h> +#include <linux/etherdevice.h> +#include <linux/slab.h> +#include <linux/device.h> +#include <linux/skbuff.h> +#include <linux/if_vlan.h> +#include <linux/if_bridge.h> +#include <linux/workqueue.h> +#include <linux/jiffies.h> +#include <linux/rtnetlink.h> +#include <linux/netlink.h> +#include <net/switchdev.h> +#include <net/vxlan.h> + +#include "spectrum_span.h" +#include "spectrum_switchdev.h" +#include "spectrum.h" +#include "core.h" +#include "reg.h" + +struct mlxsw_sp_bridge_ops; + +struct mlxsw_sp_bridge { + struct mlxsw_sp *mlxsw_sp; + struct { + struct delayed_work dw; +#define MLXSW_SP_DEFAULT_LEARNING_INTERVAL 100 + unsigned int interval; /* ms */ + } fdb_notify; +#define MLXSW_SP_MIN_AGEING_TIME 10 +#define MLXSW_SP_MAX_AGEING_TIME 1000000 +#define MLXSW_SP_DEFAULT_AGEING_TIME 300 + u32 ageing_time; + bool vlan_enabled_exists; + struct list_head bridges_list; + DECLARE_BITMAP(mids_bitmap, MLXSW_SP_MID_MAX); + const struct mlxsw_sp_bridge_ops *bridge_8021q_ops; + const struct mlxsw_sp_bridge_ops *bridge_8021d_ops; + const struct mlxsw_sp_bridge_ops *bridge_8021ad_ops; +}; + +struct mlxsw_sp_bridge_device { + struct net_device *dev; + struct list_head list; + struct list_head ports_list; + struct list_head mdb_list; + struct rhashtable mdb_ht; + u8 vlan_enabled:1, + multicast_enabled:1, + mrouter:1; + const struct mlxsw_sp_bridge_ops *ops; +}; + +struct mlxsw_sp_bridge_port { + struct net_device *dev; + struct mlxsw_sp_bridge_device *bridge_device; + struct list_head list; + struct list_head vlans_list; + unsigned int ref_count; + u8 stp_state; + unsigned long flags; + bool mrouter; + bool lagged; + union { + u16 lag_id; + u16 system_port; + }; +}; + +struct mlxsw_sp_bridge_vlan { + struct list_head list; + struct list_head port_vlan_list; + u16 vid; +}; + +struct mlxsw_sp_bridge_ops { + int (*port_join)(struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp_bridge_port *bridge_port, + struct mlxsw_sp_port *mlxsw_sp_port, + struct netlink_ext_ack *extack); + void (*port_leave)(struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp_bridge_port *bridge_port, + struct mlxsw_sp_port *mlxsw_sp_port); + int (*vxlan_join)(struct mlxsw_sp_bridge_device *bridge_device, + const struct net_device *vxlan_dev, u16 vid, + struct netlink_ext_ack *extack); + struct mlxsw_sp_fid * + (*fid_get)(struct mlxsw_sp_bridge_device *bridge_device, + u16 vid, struct netlink_ext_ack *extack); + struct mlxsw_sp_fid * + (*fid_lookup)(struct mlxsw_sp_bridge_device *bridge_device, + u16 vid); + u16 (*fid_vid)(struct mlxsw_sp_bridge_device *bridge_device, + const struct mlxsw_sp_fid *fid); +}; + +struct mlxsw_sp_switchdev_ops { + void (*init)(struct mlxsw_sp *mlxsw_sp); +}; + +struct mlxsw_sp_mdb_entry_key { + unsigned char addr[ETH_ALEN]; + u16 fid; +}; + +struct mlxsw_sp_mdb_entry { + struct list_head list; + struct rhash_head ht_node; + struct mlxsw_sp_mdb_entry_key key; + u16 mid; + struct list_head ports_list; + u16 ports_count; +}; + +struct mlxsw_sp_mdb_entry_port { + struct list_head list; /* Member of 'ports_list'. */ + u16 local_port; + refcount_t refcount; + bool mrouter; +}; + +static const struct rhashtable_params mlxsw_sp_mdb_ht_params = { + .key_offset = offsetof(struct mlxsw_sp_mdb_entry, key), + .head_offset = offsetof(struct mlxsw_sp_mdb_entry, ht_node), + .key_len = sizeof(struct mlxsw_sp_mdb_entry_key), +}; + +static int +mlxsw_sp_bridge_port_fdb_flush(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_bridge_port *bridge_port, + u16 fid_index); + +static void +mlxsw_sp_bridge_port_mdb_flush(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_bridge_port *bridge_port, + u16 fid_index); + +static int +mlxsw_sp_bridge_mdb_mc_enable_sync(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_bridge_device + *bridge_device, bool mc_enabled); + +static void +mlxsw_sp_port_mrouter_update_mdb(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_bridge_port *bridge_port, + bool add); + +static struct mlxsw_sp_bridge_device * +mlxsw_sp_bridge_device_find(const struct mlxsw_sp_bridge *bridge, + const struct net_device *br_dev) +{ + struct mlxsw_sp_bridge_device *bridge_device; + + list_for_each_entry(bridge_device, &bridge->bridges_list, list) + if (bridge_device->dev == br_dev) + return bridge_device; + + return NULL; +} + +bool mlxsw_sp_bridge_device_is_offloaded(const struct mlxsw_sp *mlxsw_sp, + const struct net_device *br_dev) +{ + return !!mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, br_dev); +} + +static int mlxsw_sp_bridge_device_upper_rif_destroy(struct net_device *dev, + struct netdev_nested_priv *priv) +{ + struct mlxsw_sp *mlxsw_sp = priv->data; + + mlxsw_sp_rif_destroy_by_dev(mlxsw_sp, dev); + return 0; +} + +static void mlxsw_sp_bridge_device_rifs_destroy(struct mlxsw_sp *mlxsw_sp, + struct net_device *dev) +{ + struct netdev_nested_priv priv = { + .data = (void *)mlxsw_sp, + }; + + mlxsw_sp_rif_destroy_by_dev(mlxsw_sp, dev); + netdev_walk_all_upper_dev_rcu(dev, + mlxsw_sp_bridge_device_upper_rif_destroy, + &priv); +} + +static int mlxsw_sp_bridge_device_vxlan_init(struct mlxsw_sp_bridge *bridge, + struct net_device *br_dev, + struct netlink_ext_ack *extack) +{ + struct net_device *dev, *stop_dev; + struct list_head *iter; + int err; + + netdev_for_each_lower_dev(br_dev, dev, iter) { + if (netif_is_vxlan(dev) && netif_running(dev)) { + err = mlxsw_sp_bridge_vxlan_join(bridge->mlxsw_sp, + br_dev, dev, 0, + extack); + if (err) { + stop_dev = dev; + goto err_vxlan_join; + } + } + } + + return 0; + +err_vxlan_join: + netdev_for_each_lower_dev(br_dev, dev, iter) { + if (netif_is_vxlan(dev) && netif_running(dev)) { + if (stop_dev == dev) + break; + mlxsw_sp_bridge_vxlan_leave(bridge->mlxsw_sp, dev); + } + } + return err; +} + +static void mlxsw_sp_bridge_device_vxlan_fini(struct mlxsw_sp_bridge *bridge, + struct net_device *br_dev) +{ + struct net_device *dev; + struct list_head *iter; + + netdev_for_each_lower_dev(br_dev, dev, iter) { + if (netif_is_vxlan(dev) && netif_running(dev)) + mlxsw_sp_bridge_vxlan_leave(bridge->mlxsw_sp, dev); + } +} + +static void mlxsw_sp_fdb_notify_work_schedule(struct mlxsw_sp *mlxsw_sp, + bool no_delay) +{ + struct mlxsw_sp_bridge *bridge = mlxsw_sp->bridge; + unsigned int interval = no_delay ? 0 : bridge->fdb_notify.interval; + + mlxsw_core_schedule_dw(&bridge->fdb_notify.dw, + msecs_to_jiffies(interval)); +} + +static struct mlxsw_sp_bridge_device * +mlxsw_sp_bridge_device_create(struct mlxsw_sp_bridge *bridge, + struct net_device *br_dev, + struct netlink_ext_ack *extack) +{ + struct device *dev = bridge->mlxsw_sp->bus_info->dev; + struct mlxsw_sp_bridge_device *bridge_device; + bool vlan_enabled = br_vlan_enabled(br_dev); + int err; + + if (vlan_enabled && bridge->vlan_enabled_exists) { + dev_err(dev, "Only one VLAN-aware bridge is supported\n"); + NL_SET_ERR_MSG_MOD(extack, "Only one VLAN-aware bridge is supported"); + return ERR_PTR(-EINVAL); + } + + bridge_device = kzalloc(sizeof(*bridge_device), GFP_KERNEL); + if (!bridge_device) + return ERR_PTR(-ENOMEM); + + err = rhashtable_init(&bridge_device->mdb_ht, &mlxsw_sp_mdb_ht_params); + if (err) + goto err_mdb_rhashtable_init; + + bridge_device->dev = br_dev; + bridge_device->vlan_enabled = vlan_enabled; + bridge_device->multicast_enabled = br_multicast_enabled(br_dev); + bridge_device->mrouter = br_multicast_router(br_dev); + INIT_LIST_HEAD(&bridge_device->ports_list); + if (vlan_enabled) { + u16 proto; + + bridge->vlan_enabled_exists = true; + br_vlan_get_proto(br_dev, &proto); + if (proto == ETH_P_8021AD) + bridge_device->ops = bridge->bridge_8021ad_ops; + else + bridge_device->ops = bridge->bridge_8021q_ops; + } else { + bridge_device->ops = bridge->bridge_8021d_ops; + } + INIT_LIST_HEAD(&bridge_device->mdb_list); + + if (list_empty(&bridge->bridges_list)) + mlxsw_sp_fdb_notify_work_schedule(bridge->mlxsw_sp, false); + list_add(&bridge_device->list, &bridge->bridges_list); + + /* It is possible we already have VXLAN devices enslaved to the bridge. + * In which case, we need to replay their configuration as if they were + * just now enslaved to the bridge. + */ + err = mlxsw_sp_bridge_device_vxlan_init(bridge, br_dev, extack); + if (err) + goto err_vxlan_init; + + return bridge_device; + +err_vxlan_init: + list_del(&bridge_device->list); + if (bridge_device->vlan_enabled) + bridge->vlan_enabled_exists = false; + rhashtable_destroy(&bridge_device->mdb_ht); +err_mdb_rhashtable_init: + kfree(bridge_device); + return ERR_PTR(err); +} + +static void +mlxsw_sp_bridge_device_destroy(struct mlxsw_sp_bridge *bridge, + struct mlxsw_sp_bridge_device *bridge_device) +{ + mlxsw_sp_bridge_device_vxlan_fini(bridge, bridge_device->dev); + mlxsw_sp_bridge_device_rifs_destroy(bridge->mlxsw_sp, + bridge_device->dev); + list_del(&bridge_device->list); + if (list_empty(&bridge->bridges_list)) + cancel_delayed_work(&bridge->fdb_notify.dw); + if (bridge_device->vlan_enabled) + bridge->vlan_enabled_exists = false; + WARN_ON(!list_empty(&bridge_device->ports_list)); + WARN_ON(!list_empty(&bridge_device->mdb_list)); + rhashtable_destroy(&bridge_device->mdb_ht); + kfree(bridge_device); +} + +static struct mlxsw_sp_bridge_device * +mlxsw_sp_bridge_device_get(struct mlxsw_sp_bridge *bridge, + struct net_device *br_dev, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_bridge_device *bridge_device; + + bridge_device = mlxsw_sp_bridge_device_find(bridge, br_dev); + if (bridge_device) + return bridge_device; + + return mlxsw_sp_bridge_device_create(bridge, br_dev, extack); +} + +static void +mlxsw_sp_bridge_device_put(struct mlxsw_sp_bridge *bridge, + struct mlxsw_sp_bridge_device *bridge_device) +{ + if (list_empty(&bridge_device->ports_list)) + mlxsw_sp_bridge_device_destroy(bridge, bridge_device); +} + +static struct mlxsw_sp_bridge_port * +__mlxsw_sp_bridge_port_find(const struct mlxsw_sp_bridge_device *bridge_device, + const struct net_device *brport_dev) +{ + struct mlxsw_sp_bridge_port *bridge_port; + + list_for_each_entry(bridge_port, &bridge_device->ports_list, list) { + if (bridge_port->dev == brport_dev) + return bridge_port; + } + + return NULL; +} + +struct mlxsw_sp_bridge_port * +mlxsw_sp_bridge_port_find(struct mlxsw_sp_bridge *bridge, + struct net_device *brport_dev) +{ + struct net_device *br_dev = netdev_master_upper_dev_get(brport_dev); + struct mlxsw_sp_bridge_device *bridge_device; + + if (!br_dev) + return NULL; + + bridge_device = mlxsw_sp_bridge_device_find(bridge, br_dev); + if (!bridge_device) + return NULL; + + return __mlxsw_sp_bridge_port_find(bridge_device, brport_dev); +} + +static struct mlxsw_sp_bridge_port * +mlxsw_sp_bridge_port_create(struct mlxsw_sp_bridge_device *bridge_device, + struct net_device *brport_dev, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_bridge_port *bridge_port; + struct mlxsw_sp_port *mlxsw_sp_port; + int err; + + bridge_port = kzalloc(sizeof(*bridge_port), GFP_KERNEL); + if (!bridge_port) + return ERR_PTR(-ENOMEM); + + mlxsw_sp_port = mlxsw_sp_port_dev_lower_find(brport_dev); + bridge_port->lagged = mlxsw_sp_port->lagged; + if (bridge_port->lagged) + bridge_port->lag_id = mlxsw_sp_port->lag_id; + else + bridge_port->system_port = mlxsw_sp_port->local_port; + bridge_port->dev = brport_dev; + bridge_port->bridge_device = bridge_device; + bridge_port->stp_state = BR_STATE_DISABLED; + bridge_port->flags = BR_LEARNING | BR_FLOOD | BR_LEARNING_SYNC | + BR_MCAST_FLOOD; + INIT_LIST_HEAD(&bridge_port->vlans_list); + list_add(&bridge_port->list, &bridge_device->ports_list); + bridge_port->ref_count = 1; + + err = switchdev_bridge_port_offload(brport_dev, mlxsw_sp_port->dev, + NULL, NULL, NULL, false, extack); + if (err) + goto err_switchdev_offload; + + return bridge_port; + +err_switchdev_offload: + list_del(&bridge_port->list); + kfree(bridge_port); + return ERR_PTR(err); +} + +static void +mlxsw_sp_bridge_port_destroy(struct mlxsw_sp_bridge_port *bridge_port) +{ + switchdev_bridge_port_unoffload(bridge_port->dev, NULL, NULL, NULL); + list_del(&bridge_port->list); + WARN_ON(!list_empty(&bridge_port->vlans_list)); + kfree(bridge_port); +} + +static struct mlxsw_sp_bridge_port * +mlxsw_sp_bridge_port_get(struct mlxsw_sp_bridge *bridge, + struct net_device *brport_dev, + struct netlink_ext_ack *extack) +{ + struct net_device *br_dev = netdev_master_upper_dev_get(brport_dev); + struct mlxsw_sp_bridge_device *bridge_device; + struct mlxsw_sp_bridge_port *bridge_port; + int err; + + bridge_port = mlxsw_sp_bridge_port_find(bridge, brport_dev); + if (bridge_port) { + bridge_port->ref_count++; + return bridge_port; + } + + bridge_device = mlxsw_sp_bridge_device_get(bridge, br_dev, extack); + if (IS_ERR(bridge_device)) + return ERR_CAST(bridge_device); + + bridge_port = mlxsw_sp_bridge_port_create(bridge_device, brport_dev, + extack); + if (IS_ERR(bridge_port)) { + err = PTR_ERR(bridge_port); + goto err_bridge_port_create; + } + + return bridge_port; + +err_bridge_port_create: + mlxsw_sp_bridge_device_put(bridge, bridge_device); + return ERR_PTR(err); +} + +static void mlxsw_sp_bridge_port_put(struct mlxsw_sp_bridge *bridge, + struct mlxsw_sp_bridge_port *bridge_port) +{ + struct mlxsw_sp_bridge_device *bridge_device; + + if (--bridge_port->ref_count != 0) + return; + bridge_device = bridge_port->bridge_device; + mlxsw_sp_bridge_port_destroy(bridge_port); + mlxsw_sp_bridge_device_put(bridge, bridge_device); +} + +static struct mlxsw_sp_port_vlan * +mlxsw_sp_port_vlan_find_by_bridge(struct mlxsw_sp_port *mlxsw_sp_port, + const struct mlxsw_sp_bridge_device * + bridge_device, + u16 vid) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + + list_for_each_entry(mlxsw_sp_port_vlan, &mlxsw_sp_port->vlans_list, + list) { + if (!mlxsw_sp_port_vlan->bridge_port) + continue; + if (mlxsw_sp_port_vlan->bridge_port->bridge_device != + bridge_device) + continue; + if (bridge_device->vlan_enabled && + mlxsw_sp_port_vlan->vid != vid) + continue; + return mlxsw_sp_port_vlan; + } + + return NULL; +} + +static struct mlxsw_sp_port_vlan* +mlxsw_sp_port_vlan_find_by_fid(struct mlxsw_sp_port *mlxsw_sp_port, + u16 fid_index) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + + list_for_each_entry(mlxsw_sp_port_vlan, &mlxsw_sp_port->vlans_list, + list) { + struct mlxsw_sp_fid *fid = mlxsw_sp_port_vlan->fid; + + if (fid && mlxsw_sp_fid_index(fid) == fid_index) + return mlxsw_sp_port_vlan; + } + + return NULL; +} + +static struct mlxsw_sp_bridge_vlan * +mlxsw_sp_bridge_vlan_find(const struct mlxsw_sp_bridge_port *bridge_port, + u16 vid) +{ + struct mlxsw_sp_bridge_vlan *bridge_vlan; + + list_for_each_entry(bridge_vlan, &bridge_port->vlans_list, list) { + if (bridge_vlan->vid == vid) + return bridge_vlan; + } + + return NULL; +} + +static struct mlxsw_sp_bridge_vlan * +mlxsw_sp_bridge_vlan_create(struct mlxsw_sp_bridge_port *bridge_port, u16 vid) +{ + struct mlxsw_sp_bridge_vlan *bridge_vlan; + + bridge_vlan = kzalloc(sizeof(*bridge_vlan), GFP_KERNEL); + if (!bridge_vlan) + return NULL; + + INIT_LIST_HEAD(&bridge_vlan->port_vlan_list); + bridge_vlan->vid = vid; + list_add(&bridge_vlan->list, &bridge_port->vlans_list); + + return bridge_vlan; +} + +static void +mlxsw_sp_bridge_vlan_destroy(struct mlxsw_sp_bridge_vlan *bridge_vlan) +{ + list_del(&bridge_vlan->list); + WARN_ON(!list_empty(&bridge_vlan->port_vlan_list)); + kfree(bridge_vlan); +} + +static struct mlxsw_sp_bridge_vlan * +mlxsw_sp_bridge_vlan_get(struct mlxsw_sp_bridge_port *bridge_port, u16 vid) +{ + struct mlxsw_sp_bridge_vlan *bridge_vlan; + + bridge_vlan = mlxsw_sp_bridge_vlan_find(bridge_port, vid); + if (bridge_vlan) + return bridge_vlan; + + return mlxsw_sp_bridge_vlan_create(bridge_port, vid); +} + +static void mlxsw_sp_bridge_vlan_put(struct mlxsw_sp_bridge_vlan *bridge_vlan) +{ + if (list_empty(&bridge_vlan->port_vlan_list)) + mlxsw_sp_bridge_vlan_destroy(bridge_vlan); +} + +static int +mlxsw_sp_port_bridge_vlan_stp_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_bridge_vlan *bridge_vlan, + u8 state) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + + list_for_each_entry(mlxsw_sp_port_vlan, &bridge_vlan->port_vlan_list, + bridge_vlan_node) { + if (mlxsw_sp_port_vlan->mlxsw_sp_port != mlxsw_sp_port) + continue; + return mlxsw_sp_port_vid_stp_set(mlxsw_sp_port, + bridge_vlan->vid, state); + } + + return 0; +} + +static int mlxsw_sp_port_attr_stp_state_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *orig_dev, + u8 state) +{ + struct mlxsw_sp_bridge_port *bridge_port; + struct mlxsw_sp_bridge_vlan *bridge_vlan; + int err; + + /* It's possible we failed to enslave the port, yet this + * operation is executed due to it being deferred. + */ + bridge_port = mlxsw_sp_bridge_port_find(mlxsw_sp_port->mlxsw_sp->bridge, + orig_dev); + if (!bridge_port) + return 0; + + list_for_each_entry(bridge_vlan, &bridge_port->vlans_list, list) { + err = mlxsw_sp_port_bridge_vlan_stp_set(mlxsw_sp_port, + bridge_vlan, state); + if (err) + goto err_port_bridge_vlan_stp_set; + } + + bridge_port->stp_state = state; + + return 0; + +err_port_bridge_vlan_stp_set: + list_for_each_entry_continue_reverse(bridge_vlan, + &bridge_port->vlans_list, list) + mlxsw_sp_port_bridge_vlan_stp_set(mlxsw_sp_port, bridge_vlan, + bridge_port->stp_state); + return err; +} + +static int +mlxsw_sp_port_bridge_vlan_flood_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_bridge_vlan *bridge_vlan, + enum mlxsw_sp_flood_type packet_type, + bool member) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + + list_for_each_entry(mlxsw_sp_port_vlan, &bridge_vlan->port_vlan_list, + bridge_vlan_node) { + if (mlxsw_sp_port_vlan->mlxsw_sp_port != mlxsw_sp_port) + continue; + return mlxsw_sp_fid_flood_set(mlxsw_sp_port_vlan->fid, + packet_type, + mlxsw_sp_port->local_port, + member); + } + + return 0; +} + +static int +mlxsw_sp_bridge_port_flood_table_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_bridge_port *bridge_port, + enum mlxsw_sp_flood_type packet_type, + bool member) +{ + struct mlxsw_sp_bridge_vlan *bridge_vlan; + int err; + + list_for_each_entry(bridge_vlan, &bridge_port->vlans_list, list) { + err = mlxsw_sp_port_bridge_vlan_flood_set(mlxsw_sp_port, + bridge_vlan, + packet_type, + member); + if (err) + goto err_port_bridge_vlan_flood_set; + } + + return 0; + +err_port_bridge_vlan_flood_set: + list_for_each_entry_continue_reverse(bridge_vlan, + &bridge_port->vlans_list, list) + mlxsw_sp_port_bridge_vlan_flood_set(mlxsw_sp_port, bridge_vlan, + packet_type, !member); + return err; +} + +static int +mlxsw_sp_bridge_vlans_flood_set(struct mlxsw_sp_bridge_vlan *bridge_vlan, + enum mlxsw_sp_flood_type packet_type, + bool member) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + int err; + + list_for_each_entry(mlxsw_sp_port_vlan, &bridge_vlan->port_vlan_list, + bridge_vlan_node) { + u16 local_port = mlxsw_sp_port_vlan->mlxsw_sp_port->local_port; + + err = mlxsw_sp_fid_flood_set(mlxsw_sp_port_vlan->fid, + packet_type, local_port, member); + if (err) + goto err_fid_flood_set; + } + + return 0; + +err_fid_flood_set: + list_for_each_entry_continue_reverse(mlxsw_sp_port_vlan, + &bridge_vlan->port_vlan_list, + list) { + u16 local_port = mlxsw_sp_port_vlan->mlxsw_sp_port->local_port; + + mlxsw_sp_fid_flood_set(mlxsw_sp_port_vlan->fid, packet_type, + local_port, !member); + } + + return err; +} + +static int +mlxsw_sp_bridge_ports_flood_table_set(struct mlxsw_sp_bridge_port *bridge_port, + enum mlxsw_sp_flood_type packet_type, + bool member) +{ + struct mlxsw_sp_bridge_vlan *bridge_vlan; + int err; + + list_for_each_entry(bridge_vlan, &bridge_port->vlans_list, list) { + err = mlxsw_sp_bridge_vlans_flood_set(bridge_vlan, packet_type, + member); + if (err) + goto err_bridge_vlans_flood_set; + } + + return 0; + +err_bridge_vlans_flood_set: + list_for_each_entry_continue_reverse(bridge_vlan, + &bridge_port->vlans_list, list) + mlxsw_sp_bridge_vlans_flood_set(bridge_vlan, packet_type, + !member); + return err; +} + +static int +mlxsw_sp_port_bridge_vlan_learning_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_bridge_vlan *bridge_vlan, + bool set) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + u16 vid = bridge_vlan->vid; + + list_for_each_entry(mlxsw_sp_port_vlan, &bridge_vlan->port_vlan_list, + bridge_vlan_node) { + if (mlxsw_sp_port_vlan->mlxsw_sp_port != mlxsw_sp_port) + continue; + return mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, vid, set); + } + + return 0; +} + +static int +mlxsw_sp_bridge_port_learning_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_bridge_port *bridge_port, + bool set) +{ + struct mlxsw_sp_bridge_vlan *bridge_vlan; + int err; + + list_for_each_entry(bridge_vlan, &bridge_port->vlans_list, list) { + err = mlxsw_sp_port_bridge_vlan_learning_set(mlxsw_sp_port, + bridge_vlan, set); + if (err) + goto err_port_bridge_vlan_learning_set; + } + + return 0; + +err_port_bridge_vlan_learning_set: + list_for_each_entry_continue_reverse(bridge_vlan, + &bridge_port->vlans_list, list) + mlxsw_sp_port_bridge_vlan_learning_set(mlxsw_sp_port, + bridge_vlan, !set); + return err; +} + +static int +mlxsw_sp_port_attr_br_pre_flags_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct switchdev_brport_flags flags) +{ + if (flags.mask & ~(BR_LEARNING | BR_FLOOD | BR_MCAST_FLOOD)) + return -EINVAL; + + return 0; +} + +static int mlxsw_sp_port_attr_br_flags_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *orig_dev, + struct switchdev_brport_flags flags) +{ + struct mlxsw_sp_bridge_port *bridge_port; + int err; + + bridge_port = mlxsw_sp_bridge_port_find(mlxsw_sp_port->mlxsw_sp->bridge, + orig_dev); + if (!bridge_port) + return 0; + + if (flags.mask & BR_FLOOD) { + err = mlxsw_sp_bridge_port_flood_table_set(mlxsw_sp_port, + bridge_port, + MLXSW_SP_FLOOD_TYPE_UC, + flags.val & BR_FLOOD); + if (err) + return err; + } + + if (flags.mask & BR_LEARNING) { + err = mlxsw_sp_bridge_port_learning_set(mlxsw_sp_port, + bridge_port, + flags.val & BR_LEARNING); + if (err) + return err; + } + + if (bridge_port->bridge_device->multicast_enabled) + goto out; + + if (flags.mask & BR_MCAST_FLOOD) { + err = mlxsw_sp_bridge_port_flood_table_set(mlxsw_sp_port, + bridge_port, + MLXSW_SP_FLOOD_TYPE_MC, + flags.val & BR_MCAST_FLOOD); + if (err) + return err; + } + +out: + memcpy(&bridge_port->flags, &flags.val, sizeof(flags.val)); + return 0; +} + +static int mlxsw_sp_ageing_set(struct mlxsw_sp *mlxsw_sp, u32 ageing_time) +{ + char sfdat_pl[MLXSW_REG_SFDAT_LEN]; + int err; + + mlxsw_reg_sfdat_pack(sfdat_pl, ageing_time); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sfdat), sfdat_pl); + if (err) + return err; + mlxsw_sp->bridge->ageing_time = ageing_time; + return 0; +} + +static int mlxsw_sp_port_attr_br_ageing_set(struct mlxsw_sp_port *mlxsw_sp_port, + unsigned long ageing_clock_t) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + unsigned long ageing_jiffies = clock_t_to_jiffies(ageing_clock_t); + u32 ageing_time = jiffies_to_msecs(ageing_jiffies) / 1000; + + if (ageing_time < MLXSW_SP_MIN_AGEING_TIME || + ageing_time > MLXSW_SP_MAX_AGEING_TIME) + return -ERANGE; + + return mlxsw_sp_ageing_set(mlxsw_sp, ageing_time); +} + +static int mlxsw_sp_port_attr_br_vlan_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *orig_dev, + bool vlan_enabled) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_bridge_device *bridge_device; + + bridge_device = mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, orig_dev); + if (WARN_ON(!bridge_device)) + return -EINVAL; + + if (bridge_device->vlan_enabled == vlan_enabled) + return 0; + + netdev_err(bridge_device->dev, "VLAN filtering can't be changed for existing bridge\n"); + return -EINVAL; +} + +static int mlxsw_sp_port_attr_br_vlan_proto_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *orig_dev, + u16 vlan_proto) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_bridge_device *bridge_device; + + bridge_device = mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, orig_dev); + if (WARN_ON(!bridge_device)) + return -EINVAL; + + netdev_err(bridge_device->dev, "VLAN protocol can't be changed on existing bridge\n"); + return -EINVAL; +} + +static int mlxsw_sp_port_attr_mrouter_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *orig_dev, + bool is_port_mrouter) +{ + struct mlxsw_sp_bridge_port *bridge_port; + int err; + + bridge_port = mlxsw_sp_bridge_port_find(mlxsw_sp_port->mlxsw_sp->bridge, + orig_dev); + if (!bridge_port) + return 0; + + mlxsw_sp_port_mrouter_update_mdb(mlxsw_sp_port, bridge_port, + is_port_mrouter); + + if (!bridge_port->bridge_device->multicast_enabled) + goto out; + + err = mlxsw_sp_bridge_port_flood_table_set(mlxsw_sp_port, bridge_port, + MLXSW_SP_FLOOD_TYPE_MC, + is_port_mrouter); + if (err) + return err; + +out: + bridge_port->mrouter = is_port_mrouter; + return 0; +} + +static bool mlxsw_sp_mc_flood(const struct mlxsw_sp_bridge_port *bridge_port) +{ + const struct mlxsw_sp_bridge_device *bridge_device; + + bridge_device = bridge_port->bridge_device; + return bridge_device->multicast_enabled ? bridge_port->mrouter : + bridge_port->flags & BR_MCAST_FLOOD; +} + +static int mlxsw_sp_port_mc_disabled_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *orig_dev, + bool mc_disabled) +{ + enum mlxsw_sp_flood_type packet_type = MLXSW_SP_FLOOD_TYPE_MC; + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_bridge_device *bridge_device; + struct mlxsw_sp_bridge_port *bridge_port; + int err; + + /* It's possible we failed to enslave the port, yet this + * operation is executed due to it being deferred. + */ + bridge_device = mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, orig_dev); + if (!bridge_device) + return 0; + + if (bridge_device->multicast_enabled == !mc_disabled) + return 0; + + bridge_device->multicast_enabled = !mc_disabled; + err = mlxsw_sp_bridge_mdb_mc_enable_sync(mlxsw_sp, bridge_device, + !mc_disabled); + if (err) + goto err_mc_enable_sync; + + list_for_each_entry(bridge_port, &bridge_device->ports_list, list) { + bool member = mlxsw_sp_mc_flood(bridge_port); + + err = mlxsw_sp_bridge_ports_flood_table_set(bridge_port, + packet_type, + member); + if (err) + goto err_flood_table_set; + } + + return 0; + +err_flood_table_set: + list_for_each_entry_continue_reverse(bridge_port, + &bridge_device->ports_list, list) { + bool member = mlxsw_sp_mc_flood(bridge_port); + + mlxsw_sp_bridge_ports_flood_table_set(bridge_port, packet_type, + !member); + } + mlxsw_sp_bridge_mdb_mc_enable_sync(mlxsw_sp, bridge_device, + mc_disabled); +err_mc_enable_sync: + bridge_device->multicast_enabled = mc_disabled; + return err; +} + +static struct mlxsw_sp_mdb_entry_port * +mlxsw_sp_mdb_entry_port_lookup(struct mlxsw_sp_mdb_entry *mdb_entry, + u16 local_port) +{ + struct mlxsw_sp_mdb_entry_port *mdb_entry_port; + + list_for_each_entry(mdb_entry_port, &mdb_entry->ports_list, list) { + if (mdb_entry_port->local_port == local_port) + return mdb_entry_port; + } + + return NULL; +} + +static struct mlxsw_sp_mdb_entry_port * +mlxsw_sp_mdb_entry_port_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mdb_entry *mdb_entry, + u16 local_port) +{ + struct mlxsw_sp_mdb_entry_port *mdb_entry_port; + int err; + + mdb_entry_port = mlxsw_sp_mdb_entry_port_lookup(mdb_entry, local_port); + if (mdb_entry_port) { + if (mdb_entry_port->mrouter && + refcount_read(&mdb_entry_port->refcount) == 1) + mdb_entry->ports_count++; + + refcount_inc(&mdb_entry_port->refcount); + return mdb_entry_port; + } + + err = mlxsw_sp_pgt_entry_port_set(mlxsw_sp, mdb_entry->mid, + mdb_entry->key.fid, local_port, true); + if (err) + return ERR_PTR(err); + + mdb_entry_port = kzalloc(sizeof(*mdb_entry_port), GFP_KERNEL); + if (!mdb_entry_port) { + err = -ENOMEM; + goto err_mdb_entry_port_alloc; + } + + mdb_entry_port->local_port = local_port; + refcount_set(&mdb_entry_port->refcount, 1); + list_add(&mdb_entry_port->list, &mdb_entry->ports_list); + mdb_entry->ports_count++; + + return mdb_entry_port; + +err_mdb_entry_port_alloc: + mlxsw_sp_pgt_entry_port_set(mlxsw_sp, mdb_entry->mid, + mdb_entry->key.fid, local_port, false); + return ERR_PTR(err); +} + +static void +mlxsw_sp_mdb_entry_port_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mdb_entry *mdb_entry, + u16 local_port, bool force) +{ + struct mlxsw_sp_mdb_entry_port *mdb_entry_port; + + mdb_entry_port = mlxsw_sp_mdb_entry_port_lookup(mdb_entry, local_port); + if (!mdb_entry_port) + return; + + if (!force && !refcount_dec_and_test(&mdb_entry_port->refcount)) { + if (mdb_entry_port->mrouter && + refcount_read(&mdb_entry_port->refcount) == 1) + mdb_entry->ports_count--; + return; + } + + mdb_entry->ports_count--; + list_del(&mdb_entry_port->list); + kfree(mdb_entry_port); + mlxsw_sp_pgt_entry_port_set(mlxsw_sp, mdb_entry->mid, + mdb_entry->key.fid, local_port, false); +} + +static __always_unused struct mlxsw_sp_mdb_entry_port * +mlxsw_sp_mdb_entry_mrouter_port_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mdb_entry *mdb_entry, + u16 local_port) +{ + struct mlxsw_sp_mdb_entry_port *mdb_entry_port; + int err; + + mdb_entry_port = mlxsw_sp_mdb_entry_port_lookup(mdb_entry, local_port); + if (mdb_entry_port) { + if (!mdb_entry_port->mrouter) + refcount_inc(&mdb_entry_port->refcount); + return mdb_entry_port; + } + + err = mlxsw_sp_pgt_entry_port_set(mlxsw_sp, mdb_entry->mid, + mdb_entry->key.fid, local_port, true); + if (err) + return ERR_PTR(err); + + mdb_entry_port = kzalloc(sizeof(*mdb_entry_port), GFP_KERNEL); + if (!mdb_entry_port) { + err = -ENOMEM; + goto err_mdb_entry_port_alloc; + } + + mdb_entry_port->local_port = local_port; + refcount_set(&mdb_entry_port->refcount, 1); + mdb_entry_port->mrouter = true; + list_add(&mdb_entry_port->list, &mdb_entry->ports_list); + + return mdb_entry_port; + +err_mdb_entry_port_alloc: + mlxsw_sp_pgt_entry_port_set(mlxsw_sp, mdb_entry->mid, + mdb_entry->key.fid, local_port, false); + return ERR_PTR(err); +} + +static __always_unused void +mlxsw_sp_mdb_entry_mrouter_port_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mdb_entry *mdb_entry, + u16 local_port) +{ + struct mlxsw_sp_mdb_entry_port *mdb_entry_port; + + mdb_entry_port = mlxsw_sp_mdb_entry_port_lookup(mdb_entry, local_port); + if (!mdb_entry_port) + return; + + if (!mdb_entry_port->mrouter) + return; + + mdb_entry_port->mrouter = false; + if (!refcount_dec_and_test(&mdb_entry_port->refcount)) + return; + + list_del(&mdb_entry_port->list); + kfree(mdb_entry_port); + mlxsw_sp_pgt_entry_port_set(mlxsw_sp, mdb_entry->mid, + mdb_entry->key.fid, local_port, false); +} + +static void +mlxsw_sp_bridge_mrouter_update_mdb(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_bridge_device *bridge_device, + bool add) +{ + u16 local_port = mlxsw_sp_router_port(mlxsw_sp); + struct mlxsw_sp_mdb_entry *mdb_entry; + + list_for_each_entry(mdb_entry, &bridge_device->mdb_list, list) { + if (add) + mlxsw_sp_mdb_entry_mrouter_port_get(mlxsw_sp, mdb_entry, + local_port); + else + mlxsw_sp_mdb_entry_mrouter_port_put(mlxsw_sp, mdb_entry, + local_port); + } +} + +static int +mlxsw_sp_port_attr_br_mrouter_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *orig_dev, + bool is_mrouter) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_bridge_device *bridge_device; + + /* It's possible we failed to enslave the port, yet this + * operation is executed due to it being deferred. + */ + bridge_device = mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, orig_dev); + if (!bridge_device) + return 0; + + if (bridge_device->mrouter != is_mrouter) + mlxsw_sp_bridge_mrouter_update_mdb(mlxsw_sp, bridge_device, + is_mrouter); + bridge_device->mrouter = is_mrouter; + return 0; +} + +static int mlxsw_sp_port_attr_set(struct net_device *dev, const void *ctx, + const struct switchdev_attr *attr, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + int err; + + switch (attr->id) { + case SWITCHDEV_ATTR_ID_PORT_STP_STATE: + err = mlxsw_sp_port_attr_stp_state_set(mlxsw_sp_port, + attr->orig_dev, + attr->u.stp_state); + break; + case SWITCHDEV_ATTR_ID_PORT_PRE_BRIDGE_FLAGS: + err = mlxsw_sp_port_attr_br_pre_flags_set(mlxsw_sp_port, + attr->u.brport_flags); + break; + case SWITCHDEV_ATTR_ID_PORT_BRIDGE_FLAGS: + err = mlxsw_sp_port_attr_br_flags_set(mlxsw_sp_port, + attr->orig_dev, + attr->u.brport_flags); + break; + case SWITCHDEV_ATTR_ID_BRIDGE_AGEING_TIME: + err = mlxsw_sp_port_attr_br_ageing_set(mlxsw_sp_port, + attr->u.ageing_time); + break; + case SWITCHDEV_ATTR_ID_BRIDGE_VLAN_FILTERING: + err = mlxsw_sp_port_attr_br_vlan_set(mlxsw_sp_port, + attr->orig_dev, + attr->u.vlan_filtering); + break; + case SWITCHDEV_ATTR_ID_BRIDGE_VLAN_PROTOCOL: + err = mlxsw_sp_port_attr_br_vlan_proto_set(mlxsw_sp_port, + attr->orig_dev, + attr->u.vlan_protocol); + break; + case SWITCHDEV_ATTR_ID_PORT_MROUTER: + err = mlxsw_sp_port_attr_mrouter_set(mlxsw_sp_port, + attr->orig_dev, + attr->u.mrouter); + break; + case SWITCHDEV_ATTR_ID_BRIDGE_MC_DISABLED: + err = mlxsw_sp_port_mc_disabled_set(mlxsw_sp_port, + attr->orig_dev, + attr->u.mc_disabled); + break; + case SWITCHDEV_ATTR_ID_BRIDGE_MROUTER: + err = mlxsw_sp_port_attr_br_mrouter_set(mlxsw_sp_port, + attr->orig_dev, + attr->u.mrouter); + break; + default: + err = -EOPNOTSUPP; + break; + } + + mlxsw_sp_span_respin(mlxsw_sp_port->mlxsw_sp); + + return err; +} + +static int +mlxsw_sp_port_vlan_fid_join(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan, + struct mlxsw_sp_bridge_port *bridge_port, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port *mlxsw_sp_port = mlxsw_sp_port_vlan->mlxsw_sp_port; + struct mlxsw_sp_bridge_device *bridge_device; + u16 local_port = mlxsw_sp_port->local_port; + u16 vid = mlxsw_sp_port_vlan->vid; + struct mlxsw_sp_fid *fid; + int err; + + bridge_device = bridge_port->bridge_device; + fid = bridge_device->ops->fid_get(bridge_device, vid, extack); + if (IS_ERR(fid)) + return PTR_ERR(fid); + + err = mlxsw_sp_fid_flood_set(fid, MLXSW_SP_FLOOD_TYPE_UC, local_port, + bridge_port->flags & BR_FLOOD); + if (err) + goto err_fid_uc_flood_set; + + err = mlxsw_sp_fid_flood_set(fid, MLXSW_SP_FLOOD_TYPE_MC, local_port, + mlxsw_sp_mc_flood(bridge_port)); + if (err) + goto err_fid_mc_flood_set; + + err = mlxsw_sp_fid_flood_set(fid, MLXSW_SP_FLOOD_TYPE_BC, local_port, + true); + if (err) + goto err_fid_bc_flood_set; + + err = mlxsw_sp_fid_port_vid_map(fid, mlxsw_sp_port, vid); + if (err) + goto err_fid_port_vid_map; + + mlxsw_sp_port_vlan->fid = fid; + + return 0; + +err_fid_port_vid_map: + mlxsw_sp_fid_flood_set(fid, MLXSW_SP_FLOOD_TYPE_BC, local_port, false); +err_fid_bc_flood_set: + mlxsw_sp_fid_flood_set(fid, MLXSW_SP_FLOOD_TYPE_MC, local_port, false); +err_fid_mc_flood_set: + mlxsw_sp_fid_flood_set(fid, MLXSW_SP_FLOOD_TYPE_UC, local_port, false); +err_fid_uc_flood_set: + mlxsw_sp_fid_put(fid); + return err; +} + +static void +mlxsw_sp_port_vlan_fid_leave(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan) +{ + struct mlxsw_sp_port *mlxsw_sp_port = mlxsw_sp_port_vlan->mlxsw_sp_port; + struct mlxsw_sp_fid *fid = mlxsw_sp_port_vlan->fid; + u16 local_port = mlxsw_sp_port->local_port; + u16 vid = mlxsw_sp_port_vlan->vid; + + mlxsw_sp_port_vlan->fid = NULL; + mlxsw_sp_fid_port_vid_unmap(fid, mlxsw_sp_port, vid); + mlxsw_sp_fid_flood_set(fid, MLXSW_SP_FLOOD_TYPE_BC, local_port, false); + mlxsw_sp_fid_flood_set(fid, MLXSW_SP_FLOOD_TYPE_MC, local_port, false); + mlxsw_sp_fid_flood_set(fid, MLXSW_SP_FLOOD_TYPE_UC, local_port, false); + mlxsw_sp_fid_put(fid); +} + +static u16 +mlxsw_sp_port_pvid_determine(const struct mlxsw_sp_port *mlxsw_sp_port, + u16 vid, bool is_pvid) +{ + if (is_pvid) + return vid; + else if (mlxsw_sp_port->pvid == vid) + return 0; /* Dis-allow untagged packets */ + else + return mlxsw_sp_port->pvid; +} + +static int +mlxsw_sp_port_vlan_bridge_join(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan, + struct mlxsw_sp_bridge_port *bridge_port, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port *mlxsw_sp_port = mlxsw_sp_port_vlan->mlxsw_sp_port; + struct mlxsw_sp_bridge_vlan *bridge_vlan; + u16 vid = mlxsw_sp_port_vlan->vid; + int err; + + /* No need to continue if only VLAN flags were changed */ + if (mlxsw_sp_port_vlan->bridge_port) + return 0; + + err = mlxsw_sp_port_vlan_fid_join(mlxsw_sp_port_vlan, bridge_port, + extack); + if (err) + return err; + + err = mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, vid, + bridge_port->flags & BR_LEARNING); + if (err) + goto err_port_vid_learning_set; + + err = mlxsw_sp_port_vid_stp_set(mlxsw_sp_port, vid, + bridge_port->stp_state); + if (err) + goto err_port_vid_stp_set; + + bridge_vlan = mlxsw_sp_bridge_vlan_get(bridge_port, vid); + if (!bridge_vlan) { + err = -ENOMEM; + goto err_bridge_vlan_get; + } + + list_add(&mlxsw_sp_port_vlan->bridge_vlan_node, + &bridge_vlan->port_vlan_list); + + mlxsw_sp_bridge_port_get(mlxsw_sp_port->mlxsw_sp->bridge, + bridge_port->dev, extack); + mlxsw_sp_port_vlan->bridge_port = bridge_port; + + return 0; + +err_bridge_vlan_get: + mlxsw_sp_port_vid_stp_set(mlxsw_sp_port, vid, BR_STATE_DISABLED); +err_port_vid_stp_set: + mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, vid, false); +err_port_vid_learning_set: + mlxsw_sp_port_vlan_fid_leave(mlxsw_sp_port_vlan); + return err; +} + +void +mlxsw_sp_port_vlan_bridge_leave(struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan) +{ + struct mlxsw_sp_port *mlxsw_sp_port = mlxsw_sp_port_vlan->mlxsw_sp_port; + struct mlxsw_sp_fid *fid = mlxsw_sp_port_vlan->fid; + struct mlxsw_sp_bridge_vlan *bridge_vlan; + struct mlxsw_sp_bridge_port *bridge_port; + u16 vid = mlxsw_sp_port_vlan->vid; + bool last_port; + + if (WARN_ON(mlxsw_sp_fid_type(fid) != MLXSW_SP_FID_TYPE_8021Q && + mlxsw_sp_fid_type(fid) != MLXSW_SP_FID_TYPE_8021D)) + return; + + bridge_port = mlxsw_sp_port_vlan->bridge_port; + bridge_vlan = mlxsw_sp_bridge_vlan_find(bridge_port, vid); + last_port = list_is_singular(&bridge_vlan->port_vlan_list); + + list_del(&mlxsw_sp_port_vlan->bridge_vlan_node); + mlxsw_sp_bridge_vlan_put(bridge_vlan); + mlxsw_sp_port_vid_stp_set(mlxsw_sp_port, vid, BR_STATE_DISABLED); + mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, vid, false); + if (last_port) + mlxsw_sp_bridge_port_fdb_flush(mlxsw_sp_port->mlxsw_sp, + bridge_port, + mlxsw_sp_fid_index(fid)); + + mlxsw_sp_bridge_port_mdb_flush(mlxsw_sp_port, bridge_port, + mlxsw_sp_fid_index(fid)); + + mlxsw_sp_port_vlan_fid_leave(mlxsw_sp_port_vlan); + + mlxsw_sp_bridge_port_put(mlxsw_sp_port->mlxsw_sp->bridge, bridge_port); + mlxsw_sp_port_vlan->bridge_port = NULL; +} + +static int +mlxsw_sp_bridge_port_vlan_add(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_bridge_port *bridge_port, + u16 vid, bool is_untagged, bool is_pvid, + struct netlink_ext_ack *extack) +{ + u16 pvid = mlxsw_sp_port_pvid_determine(mlxsw_sp_port, vid, is_pvid); + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + u16 old_pvid = mlxsw_sp_port->pvid; + u16 proto; + int err; + + /* The only valid scenario in which a port-vlan already exists, is if + * the VLAN flags were changed and the port-vlan is associated with the + * correct bridge port + */ + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_find_by_vid(mlxsw_sp_port, vid); + if (mlxsw_sp_port_vlan && + mlxsw_sp_port_vlan->bridge_port != bridge_port) + return -EEXIST; + + if (!mlxsw_sp_port_vlan) { + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_create(mlxsw_sp_port, + vid); + if (IS_ERR(mlxsw_sp_port_vlan)) + return PTR_ERR(mlxsw_sp_port_vlan); + } + + err = mlxsw_sp_port_vlan_set(mlxsw_sp_port, vid, vid, true, + is_untagged); + if (err) + goto err_port_vlan_set; + + br_vlan_get_proto(bridge_port->bridge_device->dev, &proto); + err = mlxsw_sp_port_pvid_set(mlxsw_sp_port, pvid, proto); + if (err) + goto err_port_pvid_set; + + err = mlxsw_sp_port_vlan_bridge_join(mlxsw_sp_port_vlan, bridge_port, + extack); + if (err) + goto err_port_vlan_bridge_join; + + return 0; + +err_port_vlan_bridge_join: + mlxsw_sp_port_pvid_set(mlxsw_sp_port, old_pvid, proto); +err_port_pvid_set: + mlxsw_sp_port_vlan_set(mlxsw_sp_port, vid, vid, false, false); +err_port_vlan_set: + mlxsw_sp_port_vlan_destroy(mlxsw_sp_port_vlan); + return err; +} + +static int +mlxsw_sp_br_ban_rif_pvid_change(struct mlxsw_sp *mlxsw_sp, + const struct net_device *br_dev, + const struct switchdev_obj_port_vlan *vlan) +{ + u16 pvid; + + pvid = mlxsw_sp_rif_vid(mlxsw_sp, br_dev); + if (!pvid) + return 0; + + if (vlan->flags & BRIDGE_VLAN_INFO_PVID) { + if (vlan->vid != pvid) { + netdev_err(br_dev, "Can't change PVID, it's used by router interface\n"); + return -EBUSY; + } + } else { + if (vlan->vid == pvid) { + netdev_err(br_dev, "Can't remove PVID, it's used by router interface\n"); + return -EBUSY; + } + } + + return 0; +} + +static int mlxsw_sp_port_vlans_add(struct mlxsw_sp_port *mlxsw_sp_port, + const struct switchdev_obj_port_vlan *vlan, + struct netlink_ext_ack *extack) +{ + bool flag_untagged = vlan->flags & BRIDGE_VLAN_INFO_UNTAGGED; + bool flag_pvid = vlan->flags & BRIDGE_VLAN_INFO_PVID; + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct net_device *orig_dev = vlan->obj.orig_dev; + struct mlxsw_sp_bridge_port *bridge_port; + + if (netif_is_bridge_master(orig_dev)) { + int err = 0; + + if (br_vlan_enabled(orig_dev)) + err = mlxsw_sp_br_ban_rif_pvid_change(mlxsw_sp, + orig_dev, vlan); + if (!err) + err = -EOPNOTSUPP; + return err; + } + + bridge_port = mlxsw_sp_bridge_port_find(mlxsw_sp->bridge, orig_dev); + if (WARN_ON(!bridge_port)) + return -EINVAL; + + if (!bridge_port->bridge_device->vlan_enabled) + return 0; + + return mlxsw_sp_bridge_port_vlan_add(mlxsw_sp_port, bridge_port, + vlan->vid, flag_untagged, + flag_pvid, extack); +} + +static enum mlxsw_reg_sfdf_flush_type mlxsw_sp_fdb_flush_type(bool lagged) +{ + return lagged ? MLXSW_REG_SFDF_FLUSH_PER_LAG_AND_FID : + MLXSW_REG_SFDF_FLUSH_PER_PORT_AND_FID; +} + +static int +mlxsw_sp_bridge_port_fdb_flush(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_bridge_port *bridge_port, + u16 fid_index) +{ + bool lagged = bridge_port->lagged; + char sfdf_pl[MLXSW_REG_SFDF_LEN]; + u16 system_port; + + system_port = lagged ? bridge_port->lag_id : bridge_port->system_port; + mlxsw_reg_sfdf_pack(sfdf_pl, mlxsw_sp_fdb_flush_type(lagged)); + mlxsw_reg_sfdf_fid_set(sfdf_pl, fid_index); + mlxsw_reg_sfdf_port_fid_system_port_set(sfdf_pl, system_port); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sfdf), sfdf_pl); +} + +static enum mlxsw_reg_sfd_rec_policy mlxsw_sp_sfd_rec_policy(bool dynamic) +{ + return dynamic ? MLXSW_REG_SFD_REC_POLICY_DYNAMIC_ENTRY_INGRESS : + MLXSW_REG_SFD_REC_POLICY_DYNAMIC_ENTRY_MLAG; +} + +static enum mlxsw_reg_sfd_op mlxsw_sp_sfd_op(bool adding) +{ + return adding ? MLXSW_REG_SFD_OP_WRITE_EDIT : + MLXSW_REG_SFD_OP_WRITE_REMOVE; +} + +static int +mlxsw_sp_port_fdb_tun_uc_op4(struct mlxsw_sp *mlxsw_sp, bool dynamic, + const char *mac, u16 fid, __be32 addr, bool adding) +{ + char *sfd_pl; + u8 num_rec; + u32 uip; + int err; + + sfd_pl = kmalloc(MLXSW_REG_SFD_LEN, GFP_KERNEL); + if (!sfd_pl) + return -ENOMEM; + + uip = be32_to_cpu(addr); + mlxsw_reg_sfd_pack(sfd_pl, mlxsw_sp_sfd_op(adding), 0); + mlxsw_reg_sfd_uc_tunnel_pack4(sfd_pl, 0, + mlxsw_sp_sfd_rec_policy(dynamic), mac, + fid, MLXSW_REG_SFD_REC_ACTION_NOP, uip); + num_rec = mlxsw_reg_sfd_num_rec_get(sfd_pl); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sfd), sfd_pl); + if (err) + goto out; + + if (num_rec != mlxsw_reg_sfd_num_rec_get(sfd_pl)) + err = -EBUSY; + +out: + kfree(sfd_pl); + return err; +} + +static int mlxsw_sp_port_fdb_tun_uc_op6_sfd_write(struct mlxsw_sp *mlxsw_sp, + const char *mac, u16 fid, + u32 kvdl_index, bool adding) +{ + char *sfd_pl; + u8 num_rec; + int err; + + sfd_pl = kmalloc(MLXSW_REG_SFD_LEN, GFP_KERNEL); + if (!sfd_pl) + return -ENOMEM; + + mlxsw_reg_sfd_pack(sfd_pl, mlxsw_sp_sfd_op(adding), 0); + mlxsw_reg_sfd_uc_tunnel_pack6(sfd_pl, 0, mac, fid, + MLXSW_REG_SFD_REC_ACTION_NOP, kvdl_index); + num_rec = mlxsw_reg_sfd_num_rec_get(sfd_pl); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sfd), sfd_pl); + if (err) + goto out; + + if (num_rec != mlxsw_reg_sfd_num_rec_get(sfd_pl)) + err = -EBUSY; + +out: + kfree(sfd_pl); + return err; +} + +static int mlxsw_sp_port_fdb_tun_uc_op6_add(struct mlxsw_sp *mlxsw_sp, + const char *mac, u16 fid, + const struct in6_addr *addr) +{ + u32 kvdl_index; + int err; + + err = mlxsw_sp_nve_ipv6_addr_kvdl_set(mlxsw_sp, addr, &kvdl_index); + if (err) + return err; + + err = mlxsw_sp_port_fdb_tun_uc_op6_sfd_write(mlxsw_sp, mac, fid, + kvdl_index, true); + if (err) + goto err_sfd_write; + + err = mlxsw_sp_nve_ipv6_addr_map_replace(mlxsw_sp, mac, fid, addr); + if (err) + /* Replace can fail only for creating new mapping, so removing + * the FDB entry in the error path is OK. + */ + goto err_addr_replace; + + return 0; + +err_addr_replace: + mlxsw_sp_port_fdb_tun_uc_op6_sfd_write(mlxsw_sp, mac, fid, kvdl_index, + false); +err_sfd_write: + mlxsw_sp_nve_ipv6_addr_kvdl_unset(mlxsw_sp, addr); + return err; +} + +static void mlxsw_sp_port_fdb_tun_uc_op6_del(struct mlxsw_sp *mlxsw_sp, + const char *mac, u16 fid, + const struct in6_addr *addr) +{ + mlxsw_sp_nve_ipv6_addr_map_del(mlxsw_sp, mac, fid); + mlxsw_sp_port_fdb_tun_uc_op6_sfd_write(mlxsw_sp, mac, fid, 0, false); + mlxsw_sp_nve_ipv6_addr_kvdl_unset(mlxsw_sp, addr); +} + +static int +mlxsw_sp_port_fdb_tun_uc_op6(struct mlxsw_sp *mlxsw_sp, const char *mac, + u16 fid, const struct in6_addr *addr, bool adding) +{ + if (adding) + return mlxsw_sp_port_fdb_tun_uc_op6_add(mlxsw_sp, mac, fid, + addr); + + mlxsw_sp_port_fdb_tun_uc_op6_del(mlxsw_sp, mac, fid, addr); + return 0; +} + +static int mlxsw_sp_port_fdb_tunnel_uc_op(struct mlxsw_sp *mlxsw_sp, + const char *mac, u16 fid, + enum mlxsw_sp_l3proto proto, + const union mlxsw_sp_l3addr *addr, + bool adding, bool dynamic) +{ + switch (proto) { + case MLXSW_SP_L3_PROTO_IPV4: + return mlxsw_sp_port_fdb_tun_uc_op4(mlxsw_sp, dynamic, mac, fid, + addr->addr4, adding); + case MLXSW_SP_L3_PROTO_IPV6: + return mlxsw_sp_port_fdb_tun_uc_op6(mlxsw_sp, mac, fid, + &addr->addr6, adding); + default: + WARN_ON(1); + return -EOPNOTSUPP; + } +} + +static int __mlxsw_sp_port_fdb_uc_op(struct mlxsw_sp *mlxsw_sp, u16 local_port, + const char *mac, u16 fid, u16 vid, + bool adding, + enum mlxsw_reg_sfd_rec_action action, + enum mlxsw_reg_sfd_rec_policy policy) +{ + char *sfd_pl; + u8 num_rec; + int err; + + sfd_pl = kmalloc(MLXSW_REG_SFD_LEN, GFP_KERNEL); + if (!sfd_pl) + return -ENOMEM; + + mlxsw_reg_sfd_pack(sfd_pl, mlxsw_sp_sfd_op(adding), 0); + mlxsw_reg_sfd_uc_pack(sfd_pl, 0, policy, mac, fid, vid, action, + local_port); + num_rec = mlxsw_reg_sfd_num_rec_get(sfd_pl); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sfd), sfd_pl); + if (err) + goto out; + + if (num_rec != mlxsw_reg_sfd_num_rec_get(sfd_pl)) + err = -EBUSY; + +out: + kfree(sfd_pl); + return err; +} + +static int mlxsw_sp_port_fdb_uc_op(struct mlxsw_sp *mlxsw_sp, u16 local_port, + const char *mac, u16 fid, u16 vid, + bool adding, bool dynamic) +{ + return __mlxsw_sp_port_fdb_uc_op(mlxsw_sp, local_port, mac, fid, vid, + adding, MLXSW_REG_SFD_REC_ACTION_NOP, + mlxsw_sp_sfd_rec_policy(dynamic)); +} + +int mlxsw_sp_rif_fdb_op(struct mlxsw_sp *mlxsw_sp, const char *mac, u16 fid, + bool adding) +{ + return __mlxsw_sp_port_fdb_uc_op(mlxsw_sp, 0, mac, fid, 0, adding, + MLXSW_REG_SFD_REC_ACTION_FORWARD_IP_ROUTER, + MLXSW_REG_SFD_REC_POLICY_STATIC_ENTRY); +} + +static int mlxsw_sp_port_fdb_uc_lag_op(struct mlxsw_sp *mlxsw_sp, u16 lag_id, + const char *mac, u16 fid, u16 lag_vid, + bool adding, bool dynamic) +{ + char *sfd_pl; + u8 num_rec; + int err; + + sfd_pl = kmalloc(MLXSW_REG_SFD_LEN, GFP_KERNEL); + if (!sfd_pl) + return -ENOMEM; + + mlxsw_reg_sfd_pack(sfd_pl, mlxsw_sp_sfd_op(adding), 0); + mlxsw_reg_sfd_uc_lag_pack(sfd_pl, 0, mlxsw_sp_sfd_rec_policy(dynamic), + mac, fid, MLXSW_REG_SFD_REC_ACTION_NOP, + lag_vid, lag_id); + num_rec = mlxsw_reg_sfd_num_rec_get(sfd_pl); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sfd), sfd_pl); + if (err) + goto out; + + if (num_rec != mlxsw_reg_sfd_num_rec_get(sfd_pl)) + err = -EBUSY; + +out: + kfree(sfd_pl); + return err; +} + +static int +mlxsw_sp_port_fdb_set(struct mlxsw_sp_port *mlxsw_sp_port, + struct switchdev_notifier_fdb_info *fdb_info, bool adding) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct net_device *orig_dev = fdb_info->info.dev; + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + struct mlxsw_sp_bridge_device *bridge_device; + struct mlxsw_sp_bridge_port *bridge_port; + u16 fid_index, vid; + + bridge_port = mlxsw_sp_bridge_port_find(mlxsw_sp->bridge, orig_dev); + if (!bridge_port) + return -EINVAL; + + bridge_device = bridge_port->bridge_device; + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_find_by_bridge(mlxsw_sp_port, + bridge_device, + fdb_info->vid); + if (!mlxsw_sp_port_vlan) + return 0; + + fid_index = mlxsw_sp_fid_index(mlxsw_sp_port_vlan->fid); + vid = mlxsw_sp_port_vlan->vid; + + if (!bridge_port->lagged) + return mlxsw_sp_port_fdb_uc_op(mlxsw_sp, + bridge_port->system_port, + fdb_info->addr, fid_index, vid, + adding, false); + else + return mlxsw_sp_port_fdb_uc_lag_op(mlxsw_sp, + bridge_port->lag_id, + fdb_info->addr, fid_index, + vid, adding, false); +} + +static int mlxsw_sp_mdb_entry_write(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_mdb_entry *mdb_entry, + bool adding) +{ + char *sfd_pl; + u8 num_rec; + int err; + + sfd_pl = kmalloc(MLXSW_REG_SFD_LEN, GFP_KERNEL); + if (!sfd_pl) + return -ENOMEM; + + mlxsw_reg_sfd_pack(sfd_pl, mlxsw_sp_sfd_op(adding), 0); + mlxsw_reg_sfd_mc_pack(sfd_pl, 0, mdb_entry->key.addr, + mdb_entry->key.fid, MLXSW_REG_SFD_REC_ACTION_NOP, + mdb_entry->mid); + num_rec = mlxsw_reg_sfd_num_rec_get(sfd_pl); + err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sfd), sfd_pl); + if (err) + goto out; + + if (num_rec != mlxsw_reg_sfd_num_rec_get(sfd_pl)) + err = -EBUSY; + +out: + kfree(sfd_pl); + return err; +} + +static void +mlxsw_sp_bridge_port_get_ports_bitmap(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_bridge_port *bridge_port, + struct mlxsw_sp_ports_bitmap *ports_bm) +{ + struct mlxsw_sp_port *mlxsw_sp_port; + u64 max_lag_members, i; + int lag_id; + + if (!bridge_port->lagged) { + set_bit(bridge_port->system_port, ports_bm->bitmap); + } else { + max_lag_members = MLXSW_CORE_RES_GET(mlxsw_sp->core, + MAX_LAG_MEMBERS); + lag_id = bridge_port->lag_id; + for (i = 0; i < max_lag_members; i++) { + mlxsw_sp_port = mlxsw_sp_port_lagged_get(mlxsw_sp, + lag_id, i); + if (mlxsw_sp_port) + set_bit(mlxsw_sp_port->local_port, + ports_bm->bitmap); + } + } +} + +static void +mlxsw_sp_mc_get_mrouters_bitmap(struct mlxsw_sp_ports_bitmap *flood_bm, + struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_bridge_port *bridge_port; + + list_for_each_entry(bridge_port, &bridge_device->ports_list, list) { + if (bridge_port->mrouter) { + mlxsw_sp_bridge_port_get_ports_bitmap(mlxsw_sp, + bridge_port, + flood_bm); + } + } +} + +static int mlxsw_sp_mc_mdb_mrouters_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ports_bitmap *ports_bm, + struct mlxsw_sp_mdb_entry *mdb_entry) +{ + struct mlxsw_sp_mdb_entry_port *mdb_entry_port; + unsigned int nbits = ports_bm->nbits; + int i; + + for_each_set_bit(i, ports_bm->bitmap, nbits) { + mdb_entry_port = mlxsw_sp_mdb_entry_mrouter_port_get(mlxsw_sp, + mdb_entry, + i); + if (IS_ERR(mdb_entry_port)) { + nbits = i; + goto err_mrouter_port_get; + } + } + + return 0; + +err_mrouter_port_get: + for_each_set_bit(i, ports_bm->bitmap, nbits) + mlxsw_sp_mdb_entry_mrouter_port_put(mlxsw_sp, mdb_entry, i); + return PTR_ERR(mdb_entry_port); +} + +static void mlxsw_sp_mc_mdb_mrouters_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_ports_bitmap *ports_bm, + struct mlxsw_sp_mdb_entry *mdb_entry) +{ + int i; + + for_each_set_bit(i, ports_bm->bitmap, ports_bm->nbits) + mlxsw_sp_mdb_entry_mrouter_port_put(mlxsw_sp, mdb_entry, i); +} + +static int +mlxsw_sp_mc_mdb_mrouters_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp_mdb_entry *mdb_entry, bool add) +{ + struct mlxsw_sp_ports_bitmap ports_bm; + int err; + + err = mlxsw_sp_port_bitmap_init(mlxsw_sp, &ports_bm); + if (err) + return err; + + mlxsw_sp_mc_get_mrouters_bitmap(&ports_bm, bridge_device, mlxsw_sp); + + if (add) + err = mlxsw_sp_mc_mdb_mrouters_add(mlxsw_sp, &ports_bm, + mdb_entry); + else + mlxsw_sp_mc_mdb_mrouters_del(mlxsw_sp, &ports_bm, mdb_entry); + + mlxsw_sp_port_bitmap_fini(&ports_bm); + return err; +} + +static struct mlxsw_sp_mdb_entry * +mlxsw_sp_mc_mdb_entry_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_bridge_device *bridge_device, + const unsigned char *addr, u16 fid, u16 local_port) +{ + struct mlxsw_sp_mdb_entry_port *mdb_entry_port; + struct mlxsw_sp_mdb_entry *mdb_entry; + int err; + + mdb_entry = kzalloc(sizeof(*mdb_entry), GFP_KERNEL); + if (!mdb_entry) + return ERR_PTR(-ENOMEM); + + ether_addr_copy(mdb_entry->key.addr, addr); + mdb_entry->key.fid = fid; + err = mlxsw_sp_pgt_mid_alloc(mlxsw_sp, &mdb_entry->mid); + if (err) + goto err_pgt_mid_alloc; + + INIT_LIST_HEAD(&mdb_entry->ports_list); + + err = mlxsw_sp_mc_mdb_mrouters_set(mlxsw_sp, bridge_device, mdb_entry, + true); + if (err) + goto err_mdb_mrouters_set; + + mdb_entry_port = mlxsw_sp_mdb_entry_port_get(mlxsw_sp, mdb_entry, + local_port); + if (IS_ERR(mdb_entry_port)) { + err = PTR_ERR(mdb_entry_port); + goto err_mdb_entry_port_get; + } + + if (bridge_device->multicast_enabled) { + err = mlxsw_sp_mdb_entry_write(mlxsw_sp, mdb_entry, true); + if (err) + goto err_mdb_entry_write; + } + + err = rhashtable_insert_fast(&bridge_device->mdb_ht, + &mdb_entry->ht_node, + mlxsw_sp_mdb_ht_params); + if (err) + goto err_rhashtable_insert; + + list_add_tail(&mdb_entry->list, &bridge_device->mdb_list); + + return mdb_entry; + +err_rhashtable_insert: + if (bridge_device->multicast_enabled) + mlxsw_sp_mdb_entry_write(mlxsw_sp, mdb_entry, false); +err_mdb_entry_write: + mlxsw_sp_mdb_entry_port_put(mlxsw_sp, mdb_entry, local_port, false); +err_mdb_entry_port_get: + mlxsw_sp_mc_mdb_mrouters_set(mlxsw_sp, bridge_device, mdb_entry, false); +err_mdb_mrouters_set: + mlxsw_sp_pgt_mid_free(mlxsw_sp, mdb_entry->mid); +err_pgt_mid_alloc: + kfree(mdb_entry); + return ERR_PTR(err); +} + +static void +mlxsw_sp_mc_mdb_entry_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_mdb_entry *mdb_entry, + struct mlxsw_sp_bridge_device *bridge_device, + u16 local_port, bool force) +{ + list_del(&mdb_entry->list); + rhashtable_remove_fast(&bridge_device->mdb_ht, &mdb_entry->ht_node, + mlxsw_sp_mdb_ht_params); + if (bridge_device->multicast_enabled) + mlxsw_sp_mdb_entry_write(mlxsw_sp, mdb_entry, false); + mlxsw_sp_mdb_entry_port_put(mlxsw_sp, mdb_entry, local_port, force); + mlxsw_sp_mc_mdb_mrouters_set(mlxsw_sp, bridge_device, mdb_entry, false); + WARN_ON(!list_empty(&mdb_entry->ports_list)); + mlxsw_sp_pgt_mid_free(mlxsw_sp, mdb_entry->mid); + kfree(mdb_entry); +} + +static struct mlxsw_sp_mdb_entry * +mlxsw_sp_mc_mdb_entry_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_bridge_device *bridge_device, + const unsigned char *addr, u16 fid, u16 local_port) +{ + struct mlxsw_sp_mdb_entry_key key = {}; + struct mlxsw_sp_mdb_entry *mdb_entry; + + ether_addr_copy(key.addr, addr); + key.fid = fid; + mdb_entry = rhashtable_lookup_fast(&bridge_device->mdb_ht, &key, + mlxsw_sp_mdb_ht_params); + if (mdb_entry) { + struct mlxsw_sp_mdb_entry_port *mdb_entry_port; + + mdb_entry_port = mlxsw_sp_mdb_entry_port_get(mlxsw_sp, + mdb_entry, + local_port); + if (IS_ERR(mdb_entry_port)) + return ERR_CAST(mdb_entry_port); + + return mdb_entry; + } + + return mlxsw_sp_mc_mdb_entry_init(mlxsw_sp, bridge_device, addr, fid, + local_port); +} + +static bool +mlxsw_sp_mc_mdb_entry_remove(struct mlxsw_sp_mdb_entry *mdb_entry, + struct mlxsw_sp_mdb_entry_port *removed_entry_port, + bool force) +{ + if (mdb_entry->ports_count > 1) + return false; + + if (force) + return true; + + if (!removed_entry_port->mrouter && + refcount_read(&removed_entry_port->refcount) > 1) + return false; + + if (removed_entry_port->mrouter && + refcount_read(&removed_entry_port->refcount) > 2) + return false; + + return true; +} + +static void +mlxsw_sp_mc_mdb_entry_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp_mdb_entry *mdb_entry, u16 local_port, + bool force) +{ + struct mlxsw_sp_mdb_entry_port *mdb_entry_port; + + mdb_entry_port = mlxsw_sp_mdb_entry_port_lookup(mdb_entry, local_port); + if (!mdb_entry_port) + return; + + /* Avoid a temporary situation in which the MDB entry points to an empty + * PGT entry, as otherwise packets will be temporarily dropped instead + * of being flooded. Instead, in this situation, call + * mlxsw_sp_mc_mdb_entry_fini(), which first deletes the MDB entry and + * then releases the PGT entry. + */ + if (mlxsw_sp_mc_mdb_entry_remove(mdb_entry, mdb_entry_port, force)) + mlxsw_sp_mc_mdb_entry_fini(mlxsw_sp, mdb_entry, bridge_device, + local_port, force); + else + mlxsw_sp_mdb_entry_port_put(mlxsw_sp, mdb_entry, local_port, + force); +} + +static int mlxsw_sp_port_mdb_add(struct mlxsw_sp_port *mlxsw_sp_port, + const struct switchdev_obj_port_mdb *mdb) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct net_device *orig_dev = mdb->obj.orig_dev; + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + struct mlxsw_sp_bridge_device *bridge_device; + struct mlxsw_sp_bridge_port *bridge_port; + struct mlxsw_sp_mdb_entry *mdb_entry; + u16 fid_index; + + bridge_port = mlxsw_sp_bridge_port_find(mlxsw_sp->bridge, orig_dev); + if (!bridge_port) + return 0; + + bridge_device = bridge_port->bridge_device; + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_find_by_bridge(mlxsw_sp_port, + bridge_device, + mdb->vid); + if (!mlxsw_sp_port_vlan) + return 0; + + fid_index = mlxsw_sp_fid_index(mlxsw_sp_port_vlan->fid); + + mdb_entry = mlxsw_sp_mc_mdb_entry_get(mlxsw_sp, bridge_device, + mdb->addr, fid_index, + mlxsw_sp_port->local_port); + if (IS_ERR(mdb_entry)) + return PTR_ERR(mdb_entry); + + return 0; +} + +static int +mlxsw_sp_bridge_mdb_mc_enable_sync(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_bridge_device *bridge_device, + bool mc_enabled) +{ + struct mlxsw_sp_mdb_entry *mdb_entry; + int err; + + list_for_each_entry(mdb_entry, &bridge_device->mdb_list, list) { + err = mlxsw_sp_mdb_entry_write(mlxsw_sp, mdb_entry, mc_enabled); + if (err) + goto err_mdb_entry_write; + } + return 0; + +err_mdb_entry_write: + list_for_each_entry_continue_reverse(mdb_entry, + &bridge_device->mdb_list, list) + mlxsw_sp_mdb_entry_write(mlxsw_sp, mdb_entry, !mc_enabled); + return err; +} + +static void +mlxsw_sp_port_mrouter_update_mdb(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_bridge_port *bridge_port, + bool add) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_bridge_device *bridge_device; + u16 local_port = mlxsw_sp_port->local_port; + struct mlxsw_sp_mdb_entry *mdb_entry; + + bridge_device = bridge_port->bridge_device; + + list_for_each_entry(mdb_entry, &bridge_device->mdb_list, list) { + if (add) + mlxsw_sp_mdb_entry_mrouter_port_get(mlxsw_sp, mdb_entry, + local_port); + else + mlxsw_sp_mdb_entry_mrouter_port_put(mlxsw_sp, mdb_entry, + local_port); + } +} + +static int mlxsw_sp_port_obj_add(struct net_device *dev, const void *ctx, + const struct switchdev_obj *obj, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + const struct switchdev_obj_port_vlan *vlan; + int err = 0; + + switch (obj->id) { + case SWITCHDEV_OBJ_ID_PORT_VLAN: + vlan = SWITCHDEV_OBJ_PORT_VLAN(obj); + + err = mlxsw_sp_port_vlans_add(mlxsw_sp_port, vlan, extack); + + /* The event is emitted before the changes are actually + * applied to the bridge. Therefore schedule the respin + * call for later, so that the respin logic sees the + * updated bridge state. + */ + mlxsw_sp_span_respin(mlxsw_sp_port->mlxsw_sp); + break; + case SWITCHDEV_OBJ_ID_PORT_MDB: + err = mlxsw_sp_port_mdb_add(mlxsw_sp_port, + SWITCHDEV_OBJ_PORT_MDB(obj)); + break; + default: + err = -EOPNOTSUPP; + break; + } + + return err; +} + +static void +mlxsw_sp_bridge_port_vlan_del(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_bridge_port *bridge_port, u16 vid) +{ + u16 pvid = mlxsw_sp_port->pvid == vid ? 0 : mlxsw_sp_port->pvid; + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + u16 proto; + + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_find_by_vid(mlxsw_sp_port, vid); + if (WARN_ON(!mlxsw_sp_port_vlan)) + return; + + mlxsw_sp_port_vlan_bridge_leave(mlxsw_sp_port_vlan); + br_vlan_get_proto(bridge_port->bridge_device->dev, &proto); + mlxsw_sp_port_pvid_set(mlxsw_sp_port, pvid, proto); + mlxsw_sp_port_vlan_set(mlxsw_sp_port, vid, vid, false, false); + mlxsw_sp_port_vlan_destroy(mlxsw_sp_port_vlan); +} + +static int mlxsw_sp_port_vlans_del(struct mlxsw_sp_port *mlxsw_sp_port, + const struct switchdev_obj_port_vlan *vlan) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct net_device *orig_dev = vlan->obj.orig_dev; + struct mlxsw_sp_bridge_port *bridge_port; + + if (netif_is_bridge_master(orig_dev)) + return -EOPNOTSUPP; + + bridge_port = mlxsw_sp_bridge_port_find(mlxsw_sp->bridge, orig_dev); + if (WARN_ON(!bridge_port)) + return -EINVAL; + + if (!bridge_port->bridge_device->vlan_enabled) + return 0; + + mlxsw_sp_bridge_port_vlan_del(mlxsw_sp_port, bridge_port, vlan->vid); + + return 0; +} + +static int mlxsw_sp_port_mdb_del(struct mlxsw_sp_port *mlxsw_sp_port, + const struct switchdev_obj_port_mdb *mdb) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct net_device *orig_dev = mdb->obj.orig_dev; + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + struct mlxsw_sp_bridge_device *bridge_device; + struct net_device *dev = mlxsw_sp_port->dev; + struct mlxsw_sp_bridge_port *bridge_port; + struct mlxsw_sp_mdb_entry_key key = {}; + struct mlxsw_sp_mdb_entry *mdb_entry; + u16 fid_index; + + bridge_port = mlxsw_sp_bridge_port_find(mlxsw_sp->bridge, orig_dev); + if (!bridge_port) + return 0; + + bridge_device = bridge_port->bridge_device; + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_find_by_bridge(mlxsw_sp_port, + bridge_device, + mdb->vid); + if (!mlxsw_sp_port_vlan) + return 0; + + fid_index = mlxsw_sp_fid_index(mlxsw_sp_port_vlan->fid); + + ether_addr_copy(key.addr, mdb->addr); + key.fid = fid_index; + mdb_entry = rhashtable_lookup_fast(&bridge_device->mdb_ht, &key, + mlxsw_sp_mdb_ht_params); + if (!mdb_entry) { + netdev_err(dev, "Unable to remove port from MC DB\n"); + return -EINVAL; + } + + mlxsw_sp_mc_mdb_entry_put(mlxsw_sp, bridge_device, mdb_entry, + mlxsw_sp_port->local_port, false); + return 0; +} + +static void +mlxsw_sp_bridge_port_mdb_flush(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_bridge_port *bridge_port, + u16 fid_index) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_bridge_device *bridge_device; + struct mlxsw_sp_mdb_entry *mdb_entry, *tmp; + u16 local_port = mlxsw_sp_port->local_port; + + bridge_device = bridge_port->bridge_device; + + list_for_each_entry_safe(mdb_entry, tmp, &bridge_device->mdb_list, + list) { + if (mdb_entry->key.fid != fid_index) + continue; + + if (bridge_port->mrouter) + mlxsw_sp_mdb_entry_mrouter_port_put(mlxsw_sp, + mdb_entry, + local_port); + + mlxsw_sp_mc_mdb_entry_put(mlxsw_sp, bridge_device, mdb_entry, + local_port, true); + } +} + +static int mlxsw_sp_port_obj_del(struct net_device *dev, const void *ctx, + const struct switchdev_obj *obj) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + int err = 0; + + switch (obj->id) { + case SWITCHDEV_OBJ_ID_PORT_VLAN: + err = mlxsw_sp_port_vlans_del(mlxsw_sp_port, + SWITCHDEV_OBJ_PORT_VLAN(obj)); + break; + case SWITCHDEV_OBJ_ID_PORT_MDB: + err = mlxsw_sp_port_mdb_del(mlxsw_sp_port, + SWITCHDEV_OBJ_PORT_MDB(obj)); + break; + default: + err = -EOPNOTSUPP; + break; + } + + mlxsw_sp_span_respin(mlxsw_sp_port->mlxsw_sp); + + return err; +} + +static struct mlxsw_sp_port *mlxsw_sp_lag_rep_port(struct mlxsw_sp *mlxsw_sp, + u16 lag_id) +{ + struct mlxsw_sp_port *mlxsw_sp_port; + u64 max_lag_members; + int i; + + max_lag_members = MLXSW_CORE_RES_GET(mlxsw_sp->core, + MAX_LAG_MEMBERS); + for (i = 0; i < max_lag_members; i++) { + mlxsw_sp_port = mlxsw_sp_port_lagged_get(mlxsw_sp, lag_id, i); + if (mlxsw_sp_port) + return mlxsw_sp_port; + } + return NULL; +} + +static int +mlxsw_sp_bridge_vlan_aware_port_join(struct mlxsw_sp_bridge_port *bridge_port, + struct mlxsw_sp_port *mlxsw_sp_port, + struct netlink_ext_ack *extack) +{ + if (is_vlan_dev(bridge_port->dev)) { + NL_SET_ERR_MSG_MOD(extack, "Can not enslave a VLAN device to a VLAN-aware bridge"); + return -EINVAL; + } + + /* Port is no longer usable as a router interface */ + if (mlxsw_sp_port->default_vlan->fid) + mlxsw_sp_port_vlan_router_leave(mlxsw_sp_port->default_vlan); + + return 0; +} + +static int +mlxsw_sp_bridge_8021q_port_join(struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp_bridge_port *bridge_port, + struct mlxsw_sp_port *mlxsw_sp_port, + struct netlink_ext_ack *extack) +{ + return mlxsw_sp_bridge_vlan_aware_port_join(bridge_port, mlxsw_sp_port, + extack); +} + +static void +mlxsw_sp_bridge_vlan_aware_port_leave(struct mlxsw_sp_port *mlxsw_sp_port) +{ + /* Make sure untagged frames are allowed to ingress */ + mlxsw_sp_port_pvid_set(mlxsw_sp_port, MLXSW_SP_DEFAULT_VID, + ETH_P_8021Q); +} + +static void +mlxsw_sp_bridge_8021q_port_leave(struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp_bridge_port *bridge_port, + struct mlxsw_sp_port *mlxsw_sp_port) +{ + mlxsw_sp_bridge_vlan_aware_port_leave(mlxsw_sp_port); +} + +static int +mlxsw_sp_bridge_vlan_aware_vxlan_join(struct mlxsw_sp_bridge_device *bridge_device, + const struct net_device *vxlan_dev, + u16 vid, u16 ethertype, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_lower_get(bridge_device->dev); + struct vxlan_dev *vxlan = netdev_priv(vxlan_dev); + struct mlxsw_sp_nve_params params = { + .type = MLXSW_SP_NVE_TYPE_VXLAN, + .vni = vxlan->cfg.vni, + .dev = vxlan_dev, + .ethertype = ethertype, + }; + struct mlxsw_sp_fid *fid; + int err; + + /* If the VLAN is 0, we need to find the VLAN that is configured as + * PVID and egress untagged on the bridge port of the VxLAN device. + * It is possible no such VLAN exists + */ + if (!vid) { + err = mlxsw_sp_vxlan_mapped_vid(vxlan_dev, &vid); + if (err || !vid) + return err; + } + + fid = mlxsw_sp_fid_8021q_get(mlxsw_sp, vid); + if (IS_ERR(fid)) { + NL_SET_ERR_MSG_MOD(extack, "Failed to create 802.1Q FID"); + return PTR_ERR(fid); + } + + if (mlxsw_sp_fid_vni_is_set(fid)) { + NL_SET_ERR_MSG_MOD(extack, "VNI is already set on FID"); + err = -EINVAL; + goto err_vni_exists; + } + + err = mlxsw_sp_nve_fid_enable(mlxsw_sp, fid, ¶ms, extack); + if (err) + goto err_nve_fid_enable; + + return 0; + +err_nve_fid_enable: +err_vni_exists: + mlxsw_sp_fid_put(fid); + return err; +} + +static int +mlxsw_sp_bridge_8021q_vxlan_join(struct mlxsw_sp_bridge_device *bridge_device, + const struct net_device *vxlan_dev, u16 vid, + struct netlink_ext_ack *extack) +{ + return mlxsw_sp_bridge_vlan_aware_vxlan_join(bridge_device, vxlan_dev, + vid, ETH_P_8021Q, extack); +} + +static struct net_device * +mlxsw_sp_bridge_8021q_vxlan_dev_find(struct net_device *br_dev, u16 vid) +{ + struct net_device *dev; + struct list_head *iter; + + netdev_for_each_lower_dev(br_dev, dev, iter) { + u16 pvid; + int err; + + if (!netif_is_vxlan(dev)) + continue; + + err = mlxsw_sp_vxlan_mapped_vid(dev, &pvid); + if (err || pvid != vid) + continue; + + return dev; + } + + return NULL; +} + +static struct mlxsw_sp_fid * +mlxsw_sp_bridge_8021q_fid_get(struct mlxsw_sp_bridge_device *bridge_device, + u16 vid, struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_lower_get(bridge_device->dev); + + return mlxsw_sp_fid_8021q_get(mlxsw_sp, vid); +} + +static struct mlxsw_sp_fid * +mlxsw_sp_bridge_8021q_fid_lookup(struct mlxsw_sp_bridge_device *bridge_device, + u16 vid) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_lower_get(bridge_device->dev); + + return mlxsw_sp_fid_8021q_lookup(mlxsw_sp, vid); +} + +static u16 +mlxsw_sp_bridge_8021q_fid_vid(struct mlxsw_sp_bridge_device *bridge_device, + const struct mlxsw_sp_fid *fid) +{ + return mlxsw_sp_fid_8021q_vid(fid); +} + +static const struct mlxsw_sp_bridge_ops mlxsw_sp_bridge_8021q_ops = { + .port_join = mlxsw_sp_bridge_8021q_port_join, + .port_leave = mlxsw_sp_bridge_8021q_port_leave, + .vxlan_join = mlxsw_sp_bridge_8021q_vxlan_join, + .fid_get = mlxsw_sp_bridge_8021q_fid_get, + .fid_lookup = mlxsw_sp_bridge_8021q_fid_lookup, + .fid_vid = mlxsw_sp_bridge_8021q_fid_vid, +}; + +static bool +mlxsw_sp_port_is_br_member(const struct mlxsw_sp_port *mlxsw_sp_port, + const struct net_device *br_dev) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + + list_for_each_entry(mlxsw_sp_port_vlan, &mlxsw_sp_port->vlans_list, + list) { + if (mlxsw_sp_port_vlan->bridge_port && + mlxsw_sp_port_vlan->bridge_port->bridge_device->dev == + br_dev) + return true; + } + + return false; +} + +static int +mlxsw_sp_bridge_8021d_port_join(struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp_bridge_port *bridge_port, + struct mlxsw_sp_port *mlxsw_sp_port, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + struct net_device *dev = bridge_port->dev; + u16 vid; + + vid = is_vlan_dev(dev) ? vlan_dev_vlan_id(dev) : MLXSW_SP_DEFAULT_VID; + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_find_by_vid(mlxsw_sp_port, vid); + if (WARN_ON(!mlxsw_sp_port_vlan)) + return -EINVAL; + + if (mlxsw_sp_port_is_br_member(mlxsw_sp_port, bridge_device->dev)) { + NL_SET_ERR_MSG_MOD(extack, "Can not bridge VLAN uppers of the same port"); + return -EINVAL; + } + + /* Port is no longer usable as a router interface */ + if (mlxsw_sp_port_vlan->fid) + mlxsw_sp_port_vlan_router_leave(mlxsw_sp_port_vlan); + + return mlxsw_sp_port_vlan_bridge_join(mlxsw_sp_port_vlan, bridge_port, + extack); +} + +static void +mlxsw_sp_bridge_8021d_port_leave(struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp_bridge_port *bridge_port, + struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + struct net_device *dev = bridge_port->dev; + u16 vid; + + vid = is_vlan_dev(dev) ? vlan_dev_vlan_id(dev) : MLXSW_SP_DEFAULT_VID; + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_find_by_vid(mlxsw_sp_port, vid); + if (!mlxsw_sp_port_vlan || !mlxsw_sp_port_vlan->bridge_port) + return; + + mlxsw_sp_port_vlan_bridge_leave(mlxsw_sp_port_vlan); +} + +static int +mlxsw_sp_bridge_8021d_vxlan_join(struct mlxsw_sp_bridge_device *bridge_device, + const struct net_device *vxlan_dev, u16 vid, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_lower_get(bridge_device->dev); + struct vxlan_dev *vxlan = netdev_priv(vxlan_dev); + struct mlxsw_sp_nve_params params = { + .type = MLXSW_SP_NVE_TYPE_VXLAN, + .vni = vxlan->cfg.vni, + .dev = vxlan_dev, + .ethertype = ETH_P_8021Q, + }; + struct mlxsw_sp_fid *fid; + int err; + + fid = mlxsw_sp_fid_8021d_get(mlxsw_sp, bridge_device->dev->ifindex); + if (IS_ERR(fid)) { + NL_SET_ERR_MSG_MOD(extack, "Failed to create 802.1D FID"); + return -EINVAL; + } + + if (mlxsw_sp_fid_vni_is_set(fid)) { + NL_SET_ERR_MSG_MOD(extack, "VNI is already set on FID"); + err = -EINVAL; + goto err_vni_exists; + } + + err = mlxsw_sp_nve_fid_enable(mlxsw_sp, fid, ¶ms, extack); + if (err) + goto err_nve_fid_enable; + + return 0; + +err_nve_fid_enable: +err_vni_exists: + mlxsw_sp_fid_put(fid); + return err; +} + +static struct mlxsw_sp_fid * +mlxsw_sp_bridge_8021d_fid_get(struct mlxsw_sp_bridge_device *bridge_device, + u16 vid, struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_lower_get(bridge_device->dev); + + return mlxsw_sp_fid_8021d_get(mlxsw_sp, bridge_device->dev->ifindex); +} + +static struct mlxsw_sp_fid * +mlxsw_sp_bridge_8021d_fid_lookup(struct mlxsw_sp_bridge_device *bridge_device, + u16 vid) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_lower_get(bridge_device->dev); + + /* The only valid VLAN for a VLAN-unaware bridge is 0 */ + if (vid) + return NULL; + + return mlxsw_sp_fid_8021d_lookup(mlxsw_sp, bridge_device->dev->ifindex); +} + +static u16 +mlxsw_sp_bridge_8021d_fid_vid(struct mlxsw_sp_bridge_device *bridge_device, + const struct mlxsw_sp_fid *fid) +{ + return 0; +} + +static const struct mlxsw_sp_bridge_ops mlxsw_sp_bridge_8021d_ops = { + .port_join = mlxsw_sp_bridge_8021d_port_join, + .port_leave = mlxsw_sp_bridge_8021d_port_leave, + .vxlan_join = mlxsw_sp_bridge_8021d_vxlan_join, + .fid_get = mlxsw_sp_bridge_8021d_fid_get, + .fid_lookup = mlxsw_sp_bridge_8021d_fid_lookup, + .fid_vid = mlxsw_sp_bridge_8021d_fid_vid, +}; + +static int +mlxsw_sp_bridge_8021ad_port_join(struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp_bridge_port *bridge_port, + struct mlxsw_sp_port *mlxsw_sp_port, + struct netlink_ext_ack *extack) +{ + int err; + + err = mlxsw_sp_port_vlan_classification_set(mlxsw_sp_port, true, false); + if (err) + return err; + + err = mlxsw_sp_bridge_vlan_aware_port_join(bridge_port, mlxsw_sp_port, + extack); + if (err) + goto err_bridge_vlan_aware_port_join; + + return 0; + +err_bridge_vlan_aware_port_join: + mlxsw_sp_port_vlan_classification_set(mlxsw_sp_port, false, true); + return err; +} + +static void +mlxsw_sp_bridge_8021ad_port_leave(struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp_bridge_port *bridge_port, + struct mlxsw_sp_port *mlxsw_sp_port) +{ + mlxsw_sp_bridge_vlan_aware_port_leave(mlxsw_sp_port); + mlxsw_sp_port_vlan_classification_set(mlxsw_sp_port, false, true); +} + +static int +mlxsw_sp_bridge_8021ad_vxlan_join(struct mlxsw_sp_bridge_device *bridge_device, + const struct net_device *vxlan_dev, u16 vid, + struct netlink_ext_ack *extack) +{ + return mlxsw_sp_bridge_vlan_aware_vxlan_join(bridge_device, vxlan_dev, + vid, ETH_P_8021AD, extack); +} + +static const struct mlxsw_sp_bridge_ops mlxsw_sp1_bridge_8021ad_ops = { + .port_join = mlxsw_sp_bridge_8021ad_port_join, + .port_leave = mlxsw_sp_bridge_8021ad_port_leave, + .vxlan_join = mlxsw_sp_bridge_8021ad_vxlan_join, + .fid_get = mlxsw_sp_bridge_8021q_fid_get, + .fid_lookup = mlxsw_sp_bridge_8021q_fid_lookup, + .fid_vid = mlxsw_sp_bridge_8021q_fid_vid, +}; + +static int +mlxsw_sp2_bridge_8021ad_port_join(struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp_bridge_port *bridge_port, + struct mlxsw_sp_port *mlxsw_sp_port, + struct netlink_ext_ack *extack) +{ + int err; + + /* The EtherType of decapsulated packets is determined at the egress + * port to allow 802.1d and 802.1ad bridges with VXLAN devices to + * co-exist. + */ + err = mlxsw_sp_port_egress_ethtype_set(mlxsw_sp_port, ETH_P_8021AD); + if (err) + return err; + + err = mlxsw_sp_bridge_8021ad_port_join(bridge_device, bridge_port, + mlxsw_sp_port, extack); + if (err) + goto err_bridge_8021ad_port_join; + + return 0; + +err_bridge_8021ad_port_join: + mlxsw_sp_port_egress_ethtype_set(mlxsw_sp_port, ETH_P_8021Q); + return err; +} + +static void +mlxsw_sp2_bridge_8021ad_port_leave(struct mlxsw_sp_bridge_device *bridge_device, + struct mlxsw_sp_bridge_port *bridge_port, + struct mlxsw_sp_port *mlxsw_sp_port) +{ + mlxsw_sp_bridge_8021ad_port_leave(bridge_device, bridge_port, + mlxsw_sp_port); + mlxsw_sp_port_egress_ethtype_set(mlxsw_sp_port, ETH_P_8021Q); +} + +static const struct mlxsw_sp_bridge_ops mlxsw_sp2_bridge_8021ad_ops = { + .port_join = mlxsw_sp2_bridge_8021ad_port_join, + .port_leave = mlxsw_sp2_bridge_8021ad_port_leave, + .vxlan_join = mlxsw_sp_bridge_8021ad_vxlan_join, + .fid_get = mlxsw_sp_bridge_8021q_fid_get, + .fid_lookup = mlxsw_sp_bridge_8021q_fid_lookup, + .fid_vid = mlxsw_sp_bridge_8021q_fid_vid, +}; + +int mlxsw_sp_port_bridge_join(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *brport_dev, + struct net_device *br_dev, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_bridge_device *bridge_device; + struct mlxsw_sp_bridge_port *bridge_port; + int err; + + bridge_port = mlxsw_sp_bridge_port_get(mlxsw_sp->bridge, brport_dev, + extack); + if (IS_ERR(bridge_port)) + return PTR_ERR(bridge_port); + bridge_device = bridge_port->bridge_device; + + err = bridge_device->ops->port_join(bridge_device, bridge_port, + mlxsw_sp_port, extack); + if (err) + goto err_port_join; + + return 0; + +err_port_join: + mlxsw_sp_bridge_port_put(mlxsw_sp->bridge, bridge_port); + return err; +} + +void mlxsw_sp_port_bridge_leave(struct mlxsw_sp_port *mlxsw_sp_port, + struct net_device *brport_dev, + struct net_device *br_dev) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_bridge_device *bridge_device; + struct mlxsw_sp_bridge_port *bridge_port; + + bridge_device = mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, br_dev); + if (!bridge_device) + return; + bridge_port = __mlxsw_sp_bridge_port_find(bridge_device, brport_dev); + if (!bridge_port) + return; + + bridge_device->ops->port_leave(bridge_device, bridge_port, + mlxsw_sp_port); + mlxsw_sp_bridge_port_put(mlxsw_sp->bridge, bridge_port); +} + +int mlxsw_sp_bridge_vxlan_join(struct mlxsw_sp *mlxsw_sp, + const struct net_device *br_dev, + const struct net_device *vxlan_dev, u16 vid, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp_bridge_device *bridge_device; + + bridge_device = mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, br_dev); + if (WARN_ON(!bridge_device)) + return -EINVAL; + + return bridge_device->ops->vxlan_join(bridge_device, vxlan_dev, vid, + extack); +} + +void mlxsw_sp_bridge_vxlan_leave(struct mlxsw_sp *mlxsw_sp, + const struct net_device *vxlan_dev) +{ + struct vxlan_dev *vxlan = netdev_priv(vxlan_dev); + struct mlxsw_sp_fid *fid; + + /* If the VxLAN device is down, then the FID does not have a VNI */ + fid = mlxsw_sp_fid_lookup_by_vni(mlxsw_sp, vxlan->cfg.vni); + if (!fid) + return; + + mlxsw_sp_nve_fid_disable(mlxsw_sp, fid); + /* Drop both the reference we just took during lookup and the reference + * the VXLAN device took. + */ + mlxsw_sp_fid_put(fid); + mlxsw_sp_fid_put(fid); +} + +static void +mlxsw_sp_switchdev_vxlan_addr_convert(const union vxlan_addr *vxlan_addr, + enum mlxsw_sp_l3proto *proto, + union mlxsw_sp_l3addr *addr) +{ + if (vxlan_addr->sa.sa_family == AF_INET) { + addr->addr4 = vxlan_addr->sin.sin_addr.s_addr; + *proto = MLXSW_SP_L3_PROTO_IPV4; + } else { + addr->addr6 = vxlan_addr->sin6.sin6_addr; + *proto = MLXSW_SP_L3_PROTO_IPV6; + } +} + +static void +mlxsw_sp_switchdev_addr_vxlan_convert(enum mlxsw_sp_l3proto proto, + const union mlxsw_sp_l3addr *addr, + union vxlan_addr *vxlan_addr) +{ + switch (proto) { + case MLXSW_SP_L3_PROTO_IPV4: + vxlan_addr->sa.sa_family = AF_INET; + vxlan_addr->sin.sin_addr.s_addr = addr->addr4; + break; + case MLXSW_SP_L3_PROTO_IPV6: + vxlan_addr->sa.sa_family = AF_INET6; + vxlan_addr->sin6.sin6_addr = addr->addr6; + break; + } +} + +static void mlxsw_sp_fdb_vxlan_call_notifiers(struct net_device *dev, + const char *mac, + enum mlxsw_sp_l3proto proto, + union mlxsw_sp_l3addr *addr, + __be32 vni, bool adding) +{ + struct switchdev_notifier_vxlan_fdb_info info; + struct vxlan_dev *vxlan = netdev_priv(dev); + enum switchdev_notifier_type type; + + type = adding ? SWITCHDEV_VXLAN_FDB_ADD_TO_BRIDGE : + SWITCHDEV_VXLAN_FDB_DEL_TO_BRIDGE; + mlxsw_sp_switchdev_addr_vxlan_convert(proto, addr, &info.remote_ip); + info.remote_port = vxlan->cfg.dst_port; + info.remote_vni = vni; + info.remote_ifindex = 0; + ether_addr_copy(info.eth_addr, mac); + info.vni = vni; + info.offloaded = adding; + call_switchdev_notifiers(type, dev, &info.info, NULL); +} + +static void mlxsw_sp_fdb_nve_call_notifiers(struct net_device *dev, + const char *mac, + enum mlxsw_sp_l3proto proto, + union mlxsw_sp_l3addr *addr, + __be32 vni, + bool adding) +{ + if (netif_is_vxlan(dev)) + mlxsw_sp_fdb_vxlan_call_notifiers(dev, mac, proto, addr, vni, + adding); +} + +static void +mlxsw_sp_fdb_call_notifiers(enum switchdev_notifier_type type, + const char *mac, u16 vid, + struct net_device *dev, bool offloaded) +{ + struct switchdev_notifier_fdb_info info = {}; + + info.addr = mac; + info.vid = vid; + info.offloaded = offloaded; + call_switchdev_notifiers(type, dev, &info.info, NULL); +} + +static void mlxsw_sp_fdb_notify_mac_process(struct mlxsw_sp *mlxsw_sp, + char *sfn_pl, int rec_index, + bool adding) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + struct mlxsw_sp_bridge_device *bridge_device; + struct mlxsw_sp_bridge_port *bridge_port; + struct mlxsw_sp_port *mlxsw_sp_port; + u16 local_port, vid, fid, evid = 0; + enum switchdev_notifier_type type; + char mac[ETH_ALEN]; + bool do_notification = true; + int err; + + mlxsw_reg_sfn_mac_unpack(sfn_pl, rec_index, mac, &fid, &local_port); + + if (WARN_ON_ONCE(!mlxsw_sp_local_port_is_valid(mlxsw_sp, local_port))) + return; + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + if (!mlxsw_sp_port) { + dev_err_ratelimited(mlxsw_sp->bus_info->dev, "Incorrect local port in FDB notification\n"); + goto just_remove; + } + + if (mlxsw_sp_fid_is_dummy(mlxsw_sp, fid)) + goto just_remove; + + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_find_by_fid(mlxsw_sp_port, fid); + if (!mlxsw_sp_port_vlan) { + netdev_err(mlxsw_sp_port->dev, "Failed to find a matching {Port, VID} following FDB notification\n"); + goto just_remove; + } + + bridge_port = mlxsw_sp_port_vlan->bridge_port; + if (!bridge_port) { + netdev_err(mlxsw_sp_port->dev, "{Port, VID} not associated with a bridge\n"); + goto just_remove; + } + + bridge_device = bridge_port->bridge_device; + vid = bridge_device->vlan_enabled ? mlxsw_sp_port_vlan->vid : 0; + evid = mlxsw_sp_port_vlan->vid; + +do_fdb_op: + err = mlxsw_sp_port_fdb_uc_op(mlxsw_sp, local_port, mac, fid, evid, + adding, true); + if (err) { + dev_err_ratelimited(mlxsw_sp->bus_info->dev, "Failed to set FDB entry\n"); + return; + } + + if (!do_notification) + return; + type = adding ? SWITCHDEV_FDB_ADD_TO_BRIDGE : SWITCHDEV_FDB_DEL_TO_BRIDGE; + mlxsw_sp_fdb_call_notifiers(type, mac, vid, bridge_port->dev, adding); + + return; + +just_remove: + adding = false; + do_notification = false; + goto do_fdb_op; +} + +static void mlxsw_sp_fdb_notify_mac_lag_process(struct mlxsw_sp *mlxsw_sp, + char *sfn_pl, int rec_index, + bool adding) +{ + struct mlxsw_sp_port_vlan *mlxsw_sp_port_vlan; + struct mlxsw_sp_bridge_device *bridge_device; + struct mlxsw_sp_bridge_port *bridge_port; + struct mlxsw_sp_port *mlxsw_sp_port; + enum switchdev_notifier_type type; + char mac[ETH_ALEN]; + u16 lag_vid = 0; + u16 lag_id; + u16 vid, fid; + bool do_notification = true; + int err; + + mlxsw_reg_sfn_mac_lag_unpack(sfn_pl, rec_index, mac, &fid, &lag_id); + mlxsw_sp_port = mlxsw_sp_lag_rep_port(mlxsw_sp, lag_id); + if (!mlxsw_sp_port) { + dev_err_ratelimited(mlxsw_sp->bus_info->dev, "Cannot find port representor for LAG\n"); + goto just_remove; + } + + if (mlxsw_sp_fid_is_dummy(mlxsw_sp, fid)) + goto just_remove; + + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_find_by_fid(mlxsw_sp_port, fid); + if (!mlxsw_sp_port_vlan) { + netdev_err(mlxsw_sp_port->dev, "Failed to find a matching {Port, VID} following FDB notification\n"); + goto just_remove; + } + + bridge_port = mlxsw_sp_port_vlan->bridge_port; + if (!bridge_port) { + netdev_err(mlxsw_sp_port->dev, "{Port, VID} not associated with a bridge\n"); + goto just_remove; + } + + bridge_device = bridge_port->bridge_device; + vid = bridge_device->vlan_enabled ? mlxsw_sp_port_vlan->vid : 0; + lag_vid = mlxsw_sp_port_vlan->vid; + +do_fdb_op: + err = mlxsw_sp_port_fdb_uc_lag_op(mlxsw_sp, lag_id, mac, fid, lag_vid, + adding, true); + if (err) { + dev_err_ratelimited(mlxsw_sp->bus_info->dev, "Failed to set FDB entry\n"); + return; + } + + if (!do_notification) + return; + type = adding ? SWITCHDEV_FDB_ADD_TO_BRIDGE : SWITCHDEV_FDB_DEL_TO_BRIDGE; + mlxsw_sp_fdb_call_notifiers(type, mac, vid, bridge_port->dev, adding); + + return; + +just_remove: + adding = false; + do_notification = false; + goto do_fdb_op; +} + +static int +__mlxsw_sp_fdb_notify_mac_uc_tunnel_process(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_fid *fid, + bool adding, + struct net_device **nve_dev, + u16 *p_vid, __be32 *p_vni) +{ + struct mlxsw_sp_bridge_device *bridge_device; + struct net_device *br_dev, *dev; + int nve_ifindex; + int err; + + err = mlxsw_sp_fid_nve_ifindex(fid, &nve_ifindex); + if (err) + return err; + + err = mlxsw_sp_fid_vni(fid, p_vni); + if (err) + return err; + + dev = __dev_get_by_index(mlxsw_sp_net(mlxsw_sp), nve_ifindex); + if (!dev) + return -EINVAL; + *nve_dev = dev; + + if (!netif_running(dev)) + return -EINVAL; + + if (adding && !br_port_flag_is_set(dev, BR_LEARNING)) + return -EINVAL; + + if (adding && netif_is_vxlan(dev)) { + struct vxlan_dev *vxlan = netdev_priv(dev); + + if (!(vxlan->cfg.flags & VXLAN_F_LEARN)) + return -EINVAL; + } + + br_dev = netdev_master_upper_dev_get(dev); + if (!br_dev) + return -EINVAL; + + bridge_device = mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, br_dev); + if (!bridge_device) + return -EINVAL; + + *p_vid = bridge_device->ops->fid_vid(bridge_device, fid); + + return 0; +} + +static void mlxsw_sp_fdb_notify_mac_uc_tunnel_process(struct mlxsw_sp *mlxsw_sp, + char *sfn_pl, + int rec_index, + bool adding) +{ + enum mlxsw_reg_sfn_uc_tunnel_protocol sfn_proto; + enum switchdev_notifier_type type; + struct net_device *nve_dev; + union mlxsw_sp_l3addr addr; + struct mlxsw_sp_fid *fid; + char mac[ETH_ALEN]; + u16 fid_index, vid; + __be32 vni; + u32 uip; + int err; + + mlxsw_reg_sfn_uc_tunnel_unpack(sfn_pl, rec_index, mac, &fid_index, + &uip, &sfn_proto); + + fid = mlxsw_sp_fid_lookup_by_index(mlxsw_sp, fid_index); + if (!fid) + goto err_fid_lookup; + + err = mlxsw_sp_nve_learned_ip_resolve(mlxsw_sp, uip, + (enum mlxsw_sp_l3proto) sfn_proto, + &addr); + if (err) + goto err_ip_resolve; + + err = __mlxsw_sp_fdb_notify_mac_uc_tunnel_process(mlxsw_sp, fid, adding, + &nve_dev, &vid, &vni); + if (err) + goto err_fdb_process; + + err = mlxsw_sp_port_fdb_tunnel_uc_op(mlxsw_sp, mac, fid_index, + (enum mlxsw_sp_l3proto) sfn_proto, + &addr, adding, true); + if (err) + goto err_fdb_op; + + mlxsw_sp_fdb_nve_call_notifiers(nve_dev, mac, + (enum mlxsw_sp_l3proto) sfn_proto, + &addr, vni, adding); + + type = adding ? SWITCHDEV_FDB_ADD_TO_BRIDGE : + SWITCHDEV_FDB_DEL_TO_BRIDGE; + mlxsw_sp_fdb_call_notifiers(type, mac, vid, nve_dev, adding); + + mlxsw_sp_fid_put(fid); + + return; + +err_fdb_op: +err_fdb_process: +err_ip_resolve: + mlxsw_sp_fid_put(fid); +err_fid_lookup: + /* Remove an FDB entry in case we cannot process it. Otherwise the + * device will keep sending the same notification over and over again. + */ + mlxsw_sp_port_fdb_tunnel_uc_op(mlxsw_sp, mac, fid_index, + (enum mlxsw_sp_l3proto) sfn_proto, &addr, + false, true); +} + +static void mlxsw_sp_fdb_notify_rec_process(struct mlxsw_sp *mlxsw_sp, + char *sfn_pl, int rec_index) +{ + switch (mlxsw_reg_sfn_rec_type_get(sfn_pl, rec_index)) { + case MLXSW_REG_SFN_REC_TYPE_LEARNED_MAC: + mlxsw_sp_fdb_notify_mac_process(mlxsw_sp, sfn_pl, + rec_index, true); + break; + case MLXSW_REG_SFN_REC_TYPE_AGED_OUT_MAC: + mlxsw_sp_fdb_notify_mac_process(mlxsw_sp, sfn_pl, + rec_index, false); + break; + case MLXSW_REG_SFN_REC_TYPE_LEARNED_MAC_LAG: + mlxsw_sp_fdb_notify_mac_lag_process(mlxsw_sp, sfn_pl, + rec_index, true); + break; + case MLXSW_REG_SFN_REC_TYPE_AGED_OUT_MAC_LAG: + mlxsw_sp_fdb_notify_mac_lag_process(mlxsw_sp, sfn_pl, + rec_index, false); + break; + case MLXSW_REG_SFN_REC_TYPE_LEARNED_UNICAST_TUNNEL: + mlxsw_sp_fdb_notify_mac_uc_tunnel_process(mlxsw_sp, sfn_pl, + rec_index, true); + break; + case MLXSW_REG_SFN_REC_TYPE_AGED_OUT_UNICAST_TUNNEL: + mlxsw_sp_fdb_notify_mac_uc_tunnel_process(mlxsw_sp, sfn_pl, + rec_index, false); + break; + } +} + +#define MLXSW_SP_FDB_SFN_QUERIES_PER_SESSION 10 + +static void mlxsw_sp_fdb_notify_work(struct work_struct *work) +{ + struct mlxsw_sp_bridge *bridge; + struct mlxsw_sp *mlxsw_sp; + bool reschedule = false; + char *sfn_pl; + int queries; + u8 num_rec; + int i; + int err; + + sfn_pl = kmalloc(MLXSW_REG_SFN_LEN, GFP_KERNEL); + if (!sfn_pl) + return; + + bridge = container_of(work, struct mlxsw_sp_bridge, fdb_notify.dw.work); + mlxsw_sp = bridge->mlxsw_sp; + + rtnl_lock(); + if (list_empty(&bridge->bridges_list)) + goto out; + reschedule = true; + queries = MLXSW_SP_FDB_SFN_QUERIES_PER_SESSION; + while (queries > 0) { + mlxsw_reg_sfn_pack(sfn_pl); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(sfn), sfn_pl); + if (err) { + dev_err_ratelimited(mlxsw_sp->bus_info->dev, "Failed to get FDB notifications\n"); + goto out; + } + num_rec = mlxsw_reg_sfn_num_rec_get(sfn_pl); + for (i = 0; i < num_rec; i++) + mlxsw_sp_fdb_notify_rec_process(mlxsw_sp, sfn_pl, i); + if (num_rec != MLXSW_REG_SFN_REC_MAX_COUNT) + goto out; + queries--; + } + +out: + rtnl_unlock(); + kfree(sfn_pl); + if (!reschedule) + return; + mlxsw_sp_fdb_notify_work_schedule(mlxsw_sp, !queries); +} + +struct mlxsw_sp_switchdev_event_work { + struct work_struct work; + union { + struct switchdev_notifier_fdb_info fdb_info; + struct switchdev_notifier_vxlan_fdb_info vxlan_fdb_info; + }; + struct net_device *dev; + unsigned long event; +}; + +static void +mlxsw_sp_switchdev_bridge_vxlan_fdb_event(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_switchdev_event_work * + switchdev_work, + struct mlxsw_sp_fid *fid, __be32 vni) +{ + struct switchdev_notifier_vxlan_fdb_info vxlan_fdb_info; + struct switchdev_notifier_fdb_info *fdb_info; + struct net_device *dev = switchdev_work->dev; + enum mlxsw_sp_l3proto proto; + union mlxsw_sp_l3addr addr; + int err; + + fdb_info = &switchdev_work->fdb_info; + err = vxlan_fdb_find_uc(dev, fdb_info->addr, vni, &vxlan_fdb_info); + if (err) + return; + + mlxsw_sp_switchdev_vxlan_addr_convert(&vxlan_fdb_info.remote_ip, + &proto, &addr); + + switch (switchdev_work->event) { + case SWITCHDEV_FDB_ADD_TO_DEVICE: + err = mlxsw_sp_port_fdb_tunnel_uc_op(mlxsw_sp, + vxlan_fdb_info.eth_addr, + mlxsw_sp_fid_index(fid), + proto, &addr, true, false); + if (err) + return; + vxlan_fdb_info.offloaded = true; + call_switchdev_notifiers(SWITCHDEV_VXLAN_FDB_OFFLOADED, dev, + &vxlan_fdb_info.info, NULL); + mlxsw_sp_fdb_call_notifiers(SWITCHDEV_FDB_OFFLOADED, + vxlan_fdb_info.eth_addr, + fdb_info->vid, dev, true); + break; + case SWITCHDEV_FDB_DEL_TO_DEVICE: + err = mlxsw_sp_port_fdb_tunnel_uc_op(mlxsw_sp, + vxlan_fdb_info.eth_addr, + mlxsw_sp_fid_index(fid), + proto, &addr, false, + false); + vxlan_fdb_info.offloaded = false; + call_switchdev_notifiers(SWITCHDEV_VXLAN_FDB_OFFLOADED, dev, + &vxlan_fdb_info.info, NULL); + break; + } +} + +static void +mlxsw_sp_switchdev_bridge_nve_fdb_event(struct mlxsw_sp_switchdev_event_work * + switchdev_work) +{ + struct mlxsw_sp_bridge_device *bridge_device; + struct net_device *dev = switchdev_work->dev; + struct net_device *br_dev; + struct mlxsw_sp *mlxsw_sp; + struct mlxsw_sp_fid *fid; + __be32 vni; + int err; + + if (switchdev_work->event != SWITCHDEV_FDB_ADD_TO_DEVICE && + switchdev_work->event != SWITCHDEV_FDB_DEL_TO_DEVICE) + return; + + if (switchdev_work->event == SWITCHDEV_FDB_ADD_TO_DEVICE && + (!switchdev_work->fdb_info.added_by_user || + switchdev_work->fdb_info.is_local)) + return; + + if (!netif_running(dev)) + return; + br_dev = netdev_master_upper_dev_get(dev); + if (!br_dev) + return; + if (!netif_is_bridge_master(br_dev)) + return; + mlxsw_sp = mlxsw_sp_lower_get(br_dev); + if (!mlxsw_sp) + return; + bridge_device = mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, br_dev); + if (!bridge_device) + return; + + fid = bridge_device->ops->fid_lookup(bridge_device, + switchdev_work->fdb_info.vid); + if (!fid) + return; + + err = mlxsw_sp_fid_vni(fid, &vni); + if (err) + goto out; + + mlxsw_sp_switchdev_bridge_vxlan_fdb_event(mlxsw_sp, switchdev_work, fid, + vni); + +out: + mlxsw_sp_fid_put(fid); +} + +static void mlxsw_sp_switchdev_bridge_fdb_event_work(struct work_struct *work) +{ + struct mlxsw_sp_switchdev_event_work *switchdev_work = + container_of(work, struct mlxsw_sp_switchdev_event_work, work); + struct net_device *dev = switchdev_work->dev; + struct switchdev_notifier_fdb_info *fdb_info; + struct mlxsw_sp_port *mlxsw_sp_port; + int err; + + rtnl_lock(); + if (netif_is_vxlan(dev)) { + mlxsw_sp_switchdev_bridge_nve_fdb_event(switchdev_work); + goto out; + } + + mlxsw_sp_port = mlxsw_sp_port_dev_lower_find(dev); + if (!mlxsw_sp_port) + goto out; + + switch (switchdev_work->event) { + case SWITCHDEV_FDB_ADD_TO_DEVICE: + fdb_info = &switchdev_work->fdb_info; + if (!fdb_info->added_by_user || fdb_info->is_local) + break; + err = mlxsw_sp_port_fdb_set(mlxsw_sp_port, fdb_info, true); + if (err) + break; + mlxsw_sp_fdb_call_notifiers(SWITCHDEV_FDB_OFFLOADED, + fdb_info->addr, + fdb_info->vid, dev, true); + break; + case SWITCHDEV_FDB_DEL_TO_DEVICE: + fdb_info = &switchdev_work->fdb_info; + mlxsw_sp_port_fdb_set(mlxsw_sp_port, fdb_info, false); + break; + case SWITCHDEV_FDB_ADD_TO_BRIDGE: + case SWITCHDEV_FDB_DEL_TO_BRIDGE: + /* These events are only used to potentially update an existing + * SPAN mirror. + */ + break; + } + + mlxsw_sp_span_respin(mlxsw_sp_port->mlxsw_sp); + +out: + rtnl_unlock(); + kfree(switchdev_work->fdb_info.addr); + kfree(switchdev_work); + dev_put(dev); +} + +static void +mlxsw_sp_switchdev_vxlan_fdb_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_switchdev_event_work * + switchdev_work) +{ + struct switchdev_notifier_vxlan_fdb_info *vxlan_fdb_info; + struct mlxsw_sp_bridge_device *bridge_device; + struct net_device *dev = switchdev_work->dev; + u8 all_zeros_mac[ETH_ALEN] = { 0 }; + enum mlxsw_sp_l3proto proto; + union mlxsw_sp_l3addr addr; + struct net_device *br_dev; + struct mlxsw_sp_fid *fid; + u16 vid; + int err; + + vxlan_fdb_info = &switchdev_work->vxlan_fdb_info; + br_dev = netdev_master_upper_dev_get(dev); + + bridge_device = mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, br_dev); + if (!bridge_device) + return; + + fid = mlxsw_sp_fid_lookup_by_vni(mlxsw_sp, vxlan_fdb_info->vni); + if (!fid) + return; + + mlxsw_sp_switchdev_vxlan_addr_convert(&vxlan_fdb_info->remote_ip, + &proto, &addr); + + if (ether_addr_equal(vxlan_fdb_info->eth_addr, all_zeros_mac)) { + err = mlxsw_sp_nve_flood_ip_add(mlxsw_sp, fid, proto, &addr); + if (err) { + mlxsw_sp_fid_put(fid); + return; + } + vxlan_fdb_info->offloaded = true; + call_switchdev_notifiers(SWITCHDEV_VXLAN_FDB_OFFLOADED, dev, + &vxlan_fdb_info->info, NULL); + mlxsw_sp_fid_put(fid); + return; + } + + /* The device has a single FDB table, whereas Linux has two - one + * in the bridge driver and another in the VxLAN driver. We only + * program an entry to the device if the MAC points to the VxLAN + * device in the bridge's FDB table + */ + vid = bridge_device->ops->fid_vid(bridge_device, fid); + if (br_fdb_find_port(br_dev, vxlan_fdb_info->eth_addr, vid) != dev) + goto err_br_fdb_find; + + err = mlxsw_sp_port_fdb_tunnel_uc_op(mlxsw_sp, vxlan_fdb_info->eth_addr, + mlxsw_sp_fid_index(fid), proto, + &addr, true, false); + if (err) + goto err_fdb_tunnel_uc_op; + vxlan_fdb_info->offloaded = true; + call_switchdev_notifiers(SWITCHDEV_VXLAN_FDB_OFFLOADED, dev, + &vxlan_fdb_info->info, NULL); + mlxsw_sp_fdb_call_notifiers(SWITCHDEV_FDB_OFFLOADED, + vxlan_fdb_info->eth_addr, vid, dev, true); + + mlxsw_sp_fid_put(fid); + + return; + +err_fdb_tunnel_uc_op: +err_br_fdb_find: + mlxsw_sp_fid_put(fid); +} + +static void +mlxsw_sp_switchdev_vxlan_fdb_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_switchdev_event_work * + switchdev_work) +{ + struct switchdev_notifier_vxlan_fdb_info *vxlan_fdb_info; + struct mlxsw_sp_bridge_device *bridge_device; + struct net_device *dev = switchdev_work->dev; + struct net_device *br_dev = netdev_master_upper_dev_get(dev); + u8 all_zeros_mac[ETH_ALEN] = { 0 }; + enum mlxsw_sp_l3proto proto; + union mlxsw_sp_l3addr addr; + struct mlxsw_sp_fid *fid; + u16 vid; + + vxlan_fdb_info = &switchdev_work->vxlan_fdb_info; + if (!vxlan_fdb_info->offloaded) + return; + + bridge_device = mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, br_dev); + if (!bridge_device) + return; + + fid = mlxsw_sp_fid_lookup_by_vni(mlxsw_sp, vxlan_fdb_info->vni); + if (!fid) + return; + + mlxsw_sp_switchdev_vxlan_addr_convert(&vxlan_fdb_info->remote_ip, + &proto, &addr); + + if (ether_addr_equal(vxlan_fdb_info->eth_addr, all_zeros_mac)) { + mlxsw_sp_nve_flood_ip_del(mlxsw_sp, fid, proto, &addr); + mlxsw_sp_fid_put(fid); + return; + } + + mlxsw_sp_port_fdb_tunnel_uc_op(mlxsw_sp, vxlan_fdb_info->eth_addr, + mlxsw_sp_fid_index(fid), proto, &addr, + false, false); + vid = bridge_device->ops->fid_vid(bridge_device, fid); + mlxsw_sp_fdb_call_notifiers(SWITCHDEV_FDB_OFFLOADED, + vxlan_fdb_info->eth_addr, vid, dev, false); + + mlxsw_sp_fid_put(fid); +} + +static void mlxsw_sp_switchdev_vxlan_fdb_event_work(struct work_struct *work) +{ + struct mlxsw_sp_switchdev_event_work *switchdev_work = + container_of(work, struct mlxsw_sp_switchdev_event_work, work); + struct net_device *dev = switchdev_work->dev; + struct mlxsw_sp *mlxsw_sp; + struct net_device *br_dev; + + rtnl_lock(); + + if (!netif_running(dev)) + goto out; + br_dev = netdev_master_upper_dev_get(dev); + if (!br_dev) + goto out; + if (!netif_is_bridge_master(br_dev)) + goto out; + mlxsw_sp = mlxsw_sp_lower_get(br_dev); + if (!mlxsw_sp) + goto out; + + switch (switchdev_work->event) { + case SWITCHDEV_VXLAN_FDB_ADD_TO_DEVICE: + mlxsw_sp_switchdev_vxlan_fdb_add(mlxsw_sp, switchdev_work); + break; + case SWITCHDEV_VXLAN_FDB_DEL_TO_DEVICE: + mlxsw_sp_switchdev_vxlan_fdb_del(mlxsw_sp, switchdev_work); + break; + } + +out: + rtnl_unlock(); + kfree(switchdev_work); + dev_put(dev); +} + +static int +mlxsw_sp_switchdev_vxlan_work_prepare(struct mlxsw_sp_switchdev_event_work * + switchdev_work, + struct switchdev_notifier_info *info) +{ + struct vxlan_dev *vxlan = netdev_priv(switchdev_work->dev); + struct switchdev_notifier_vxlan_fdb_info *vxlan_fdb_info; + struct vxlan_config *cfg = &vxlan->cfg; + struct netlink_ext_ack *extack; + + extack = switchdev_notifier_info_to_extack(info); + vxlan_fdb_info = container_of(info, + struct switchdev_notifier_vxlan_fdb_info, + info); + + if (vxlan_fdb_info->remote_port != cfg->dst_port) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: FDB: Non-default remote port is not supported"); + return -EOPNOTSUPP; + } + if (vxlan_fdb_info->remote_vni != cfg->vni || + vxlan_fdb_info->vni != cfg->vni) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: FDB: Non-default VNI is not supported"); + return -EOPNOTSUPP; + } + if (vxlan_fdb_info->remote_ifindex) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: FDB: Local interface is not supported"); + return -EOPNOTSUPP; + } + if (is_multicast_ether_addr(vxlan_fdb_info->eth_addr)) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: FDB: Multicast MAC addresses not supported"); + return -EOPNOTSUPP; + } + if (vxlan_addr_multicast(&vxlan_fdb_info->remote_ip)) { + NL_SET_ERR_MSG_MOD(extack, "VxLAN: FDB: Multicast destination IP is not supported"); + return -EOPNOTSUPP; + } + + switchdev_work->vxlan_fdb_info = *vxlan_fdb_info; + + return 0; +} + +/* Called under rcu_read_lock() */ +static int mlxsw_sp_switchdev_event(struct notifier_block *unused, + unsigned long event, void *ptr) +{ + struct net_device *dev = switchdev_notifier_info_to_dev(ptr); + struct mlxsw_sp_switchdev_event_work *switchdev_work; + struct switchdev_notifier_fdb_info *fdb_info; + struct switchdev_notifier_info *info = ptr; + struct net_device *br_dev; + int err; + + if (event == SWITCHDEV_PORT_ATTR_SET) { + err = switchdev_handle_port_attr_set(dev, ptr, + mlxsw_sp_port_dev_check, + mlxsw_sp_port_attr_set); + return notifier_from_errno(err); + } + + /* Tunnel devices are not our uppers, so check their master instead */ + br_dev = netdev_master_upper_dev_get_rcu(dev); + if (!br_dev) + return NOTIFY_DONE; + if (!netif_is_bridge_master(br_dev)) + return NOTIFY_DONE; + if (!mlxsw_sp_port_dev_lower_find_rcu(br_dev)) + return NOTIFY_DONE; + + switchdev_work = kzalloc(sizeof(*switchdev_work), GFP_ATOMIC); + if (!switchdev_work) + return NOTIFY_BAD; + + switchdev_work->dev = dev; + switchdev_work->event = event; + + switch (event) { + case SWITCHDEV_FDB_ADD_TO_DEVICE: + case SWITCHDEV_FDB_DEL_TO_DEVICE: + case SWITCHDEV_FDB_ADD_TO_BRIDGE: + case SWITCHDEV_FDB_DEL_TO_BRIDGE: + fdb_info = container_of(info, + struct switchdev_notifier_fdb_info, + info); + INIT_WORK(&switchdev_work->work, + mlxsw_sp_switchdev_bridge_fdb_event_work); + memcpy(&switchdev_work->fdb_info, ptr, + sizeof(switchdev_work->fdb_info)); + switchdev_work->fdb_info.addr = kzalloc(ETH_ALEN, GFP_ATOMIC); + if (!switchdev_work->fdb_info.addr) + goto err_addr_alloc; + ether_addr_copy((u8 *)switchdev_work->fdb_info.addr, + fdb_info->addr); + /* Take a reference on the device. This can be either + * upper device containig mlxsw_sp_port or just a + * mlxsw_sp_port + */ + dev_hold(dev); + break; + case SWITCHDEV_VXLAN_FDB_ADD_TO_DEVICE: + case SWITCHDEV_VXLAN_FDB_DEL_TO_DEVICE: + INIT_WORK(&switchdev_work->work, + mlxsw_sp_switchdev_vxlan_fdb_event_work); + err = mlxsw_sp_switchdev_vxlan_work_prepare(switchdev_work, + info); + if (err) + goto err_vxlan_work_prepare; + dev_hold(dev); + break; + default: + kfree(switchdev_work); + return NOTIFY_DONE; + } + + mlxsw_core_schedule_work(&switchdev_work->work); + + return NOTIFY_DONE; + +err_vxlan_work_prepare: +err_addr_alloc: + kfree(switchdev_work); + return NOTIFY_BAD; +} + +struct notifier_block mlxsw_sp_switchdev_notifier = { + .notifier_call = mlxsw_sp_switchdev_event, +}; + +static int +mlxsw_sp_switchdev_vxlan_vlan_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_bridge_device *bridge_device, + const struct net_device *vxlan_dev, u16 vid, + bool flag_untagged, bool flag_pvid, + struct netlink_ext_ack *extack) +{ + struct vxlan_dev *vxlan = netdev_priv(vxlan_dev); + __be32 vni = vxlan->cfg.vni; + struct mlxsw_sp_fid *fid; + u16 old_vid; + int err; + + /* We cannot have the same VLAN as PVID and egress untagged on multiple + * VxLAN devices. Note that we get this notification before the VLAN is + * actually added to the bridge's database, so it is not possible for + * the lookup function to return 'vxlan_dev' + */ + if (flag_untagged && flag_pvid && + mlxsw_sp_bridge_8021q_vxlan_dev_find(bridge_device->dev, vid)) { + NL_SET_ERR_MSG_MOD(extack, "VLAN already mapped to a different VNI"); + return -EINVAL; + } + + if (!netif_running(vxlan_dev)) + return 0; + + /* First case: FID is not associated with this VNI, but the new VLAN + * is both PVID and egress untagged. Need to enable NVE on the FID, if + * it exists + */ + fid = mlxsw_sp_fid_lookup_by_vni(mlxsw_sp, vni); + if (!fid) { + if (!flag_untagged || !flag_pvid) + return 0; + return bridge_device->ops->vxlan_join(bridge_device, vxlan_dev, + vid, extack); + } + + /* Second case: FID is associated with the VNI and the VLAN associated + * with the FID is the same as the notified VLAN. This means the flags + * (PVID / egress untagged) were toggled and that NVE should be + * disabled on the FID + */ + old_vid = mlxsw_sp_fid_8021q_vid(fid); + if (vid == old_vid) { + if (WARN_ON(flag_untagged && flag_pvid)) { + mlxsw_sp_fid_put(fid); + return -EINVAL; + } + mlxsw_sp_bridge_vxlan_leave(mlxsw_sp, vxlan_dev); + mlxsw_sp_fid_put(fid); + return 0; + } + + /* Third case: A new VLAN was configured on the VxLAN device, but this + * VLAN is not PVID, so there is nothing to do. + */ + if (!flag_pvid) { + mlxsw_sp_fid_put(fid); + return 0; + } + + /* Fourth case: Thew new VLAN is PVID, which means the VLAN currently + * mapped to the VNI should be unmapped + */ + mlxsw_sp_bridge_vxlan_leave(mlxsw_sp, vxlan_dev); + mlxsw_sp_fid_put(fid); + + /* Fifth case: The new VLAN is also egress untagged, which means the + * VLAN needs to be mapped to the VNI + */ + if (!flag_untagged) + return 0; + + err = bridge_device->ops->vxlan_join(bridge_device, vxlan_dev, vid, extack); + if (err) + goto err_vxlan_join; + + return 0; + +err_vxlan_join: + bridge_device->ops->vxlan_join(bridge_device, vxlan_dev, old_vid, NULL); + return err; +} + +static void +mlxsw_sp_switchdev_vxlan_vlan_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_bridge_device *bridge_device, + const struct net_device *vxlan_dev, u16 vid) +{ + struct vxlan_dev *vxlan = netdev_priv(vxlan_dev); + __be32 vni = vxlan->cfg.vni; + struct mlxsw_sp_fid *fid; + + if (!netif_running(vxlan_dev)) + return; + + fid = mlxsw_sp_fid_lookup_by_vni(mlxsw_sp, vni); + if (!fid) + return; + + /* A different VLAN than the one mapped to the VNI is deleted */ + if (mlxsw_sp_fid_8021q_vid(fid) != vid) + goto out; + + mlxsw_sp_bridge_vxlan_leave(mlxsw_sp, vxlan_dev); + +out: + mlxsw_sp_fid_put(fid); +} + +static int +mlxsw_sp_switchdev_vxlan_vlans_add(struct net_device *vxlan_dev, + struct switchdev_notifier_port_obj_info * + port_obj_info) +{ + struct switchdev_obj_port_vlan *vlan = + SWITCHDEV_OBJ_PORT_VLAN(port_obj_info->obj); + bool flag_untagged = vlan->flags & BRIDGE_VLAN_INFO_UNTAGGED; + bool flag_pvid = vlan->flags & BRIDGE_VLAN_INFO_PVID; + struct mlxsw_sp_bridge_device *bridge_device; + struct netlink_ext_ack *extack; + struct mlxsw_sp *mlxsw_sp; + struct net_device *br_dev; + + extack = switchdev_notifier_info_to_extack(&port_obj_info->info); + br_dev = netdev_master_upper_dev_get(vxlan_dev); + if (!br_dev) + return 0; + + mlxsw_sp = mlxsw_sp_lower_get(br_dev); + if (!mlxsw_sp) + return 0; + + port_obj_info->handled = true; + + bridge_device = mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, br_dev); + if (!bridge_device) + return -EINVAL; + + if (!bridge_device->vlan_enabled) + return 0; + + return mlxsw_sp_switchdev_vxlan_vlan_add(mlxsw_sp, bridge_device, + vxlan_dev, vlan->vid, + flag_untagged, + flag_pvid, extack); +} + +static void +mlxsw_sp_switchdev_vxlan_vlans_del(struct net_device *vxlan_dev, + struct switchdev_notifier_port_obj_info * + port_obj_info) +{ + struct switchdev_obj_port_vlan *vlan = + SWITCHDEV_OBJ_PORT_VLAN(port_obj_info->obj); + struct mlxsw_sp_bridge_device *bridge_device; + struct mlxsw_sp *mlxsw_sp; + struct net_device *br_dev; + + br_dev = netdev_master_upper_dev_get(vxlan_dev); + if (!br_dev) + return; + + mlxsw_sp = mlxsw_sp_lower_get(br_dev); + if (!mlxsw_sp) + return; + + port_obj_info->handled = true; + + bridge_device = mlxsw_sp_bridge_device_find(mlxsw_sp->bridge, br_dev); + if (!bridge_device) + return; + + if (!bridge_device->vlan_enabled) + return; + + mlxsw_sp_switchdev_vxlan_vlan_del(mlxsw_sp, bridge_device, vxlan_dev, + vlan->vid); +} + +static int +mlxsw_sp_switchdev_handle_vxlan_obj_add(struct net_device *vxlan_dev, + struct switchdev_notifier_port_obj_info * + port_obj_info) +{ + int err = 0; + + switch (port_obj_info->obj->id) { + case SWITCHDEV_OBJ_ID_PORT_VLAN: + err = mlxsw_sp_switchdev_vxlan_vlans_add(vxlan_dev, + port_obj_info); + break; + default: + break; + } + + return err; +} + +static void +mlxsw_sp_switchdev_handle_vxlan_obj_del(struct net_device *vxlan_dev, + struct switchdev_notifier_port_obj_info * + port_obj_info) +{ + switch (port_obj_info->obj->id) { + case SWITCHDEV_OBJ_ID_PORT_VLAN: + mlxsw_sp_switchdev_vxlan_vlans_del(vxlan_dev, port_obj_info); + break; + default: + break; + } +} + +static int mlxsw_sp_switchdev_blocking_event(struct notifier_block *unused, + unsigned long event, void *ptr) +{ + struct net_device *dev = switchdev_notifier_info_to_dev(ptr); + int err = 0; + + switch (event) { + case SWITCHDEV_PORT_OBJ_ADD: + if (netif_is_vxlan(dev)) + err = mlxsw_sp_switchdev_handle_vxlan_obj_add(dev, ptr); + else + err = switchdev_handle_port_obj_add(dev, ptr, + mlxsw_sp_port_dev_check, + mlxsw_sp_port_obj_add); + return notifier_from_errno(err); + case SWITCHDEV_PORT_OBJ_DEL: + if (netif_is_vxlan(dev)) + mlxsw_sp_switchdev_handle_vxlan_obj_del(dev, ptr); + else + err = switchdev_handle_port_obj_del(dev, ptr, + mlxsw_sp_port_dev_check, + mlxsw_sp_port_obj_del); + return notifier_from_errno(err); + case SWITCHDEV_PORT_ATTR_SET: + err = switchdev_handle_port_attr_set(dev, ptr, + mlxsw_sp_port_dev_check, + mlxsw_sp_port_attr_set); + return notifier_from_errno(err); + } + + return NOTIFY_DONE; +} + +static struct notifier_block mlxsw_sp_switchdev_blocking_notifier = { + .notifier_call = mlxsw_sp_switchdev_blocking_event, +}; + +u8 +mlxsw_sp_bridge_port_stp_state(struct mlxsw_sp_bridge_port *bridge_port) +{ + return bridge_port->stp_state; +} + +static int mlxsw_sp_fdb_init(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_bridge *bridge = mlxsw_sp->bridge; + struct notifier_block *nb; + int err; + + err = mlxsw_sp_ageing_set(mlxsw_sp, MLXSW_SP_DEFAULT_AGEING_TIME); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to set default ageing time\n"); + return err; + } + + err = register_switchdev_notifier(&mlxsw_sp_switchdev_notifier); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to register switchdev notifier\n"); + return err; + } + + nb = &mlxsw_sp_switchdev_blocking_notifier; + err = register_switchdev_blocking_notifier(nb); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to register switchdev blocking notifier\n"); + goto err_register_switchdev_blocking_notifier; + } + + INIT_DELAYED_WORK(&bridge->fdb_notify.dw, mlxsw_sp_fdb_notify_work); + bridge->fdb_notify.interval = MLXSW_SP_DEFAULT_LEARNING_INTERVAL; + return 0; + +err_register_switchdev_blocking_notifier: + unregister_switchdev_notifier(&mlxsw_sp_switchdev_notifier); + return err; +} + +static void mlxsw_sp_fdb_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct notifier_block *nb; + + cancel_delayed_work_sync(&mlxsw_sp->bridge->fdb_notify.dw); + + nb = &mlxsw_sp_switchdev_blocking_notifier; + unregister_switchdev_blocking_notifier(nb); + + unregister_switchdev_notifier(&mlxsw_sp_switchdev_notifier); +} + +static void mlxsw_sp1_switchdev_init(struct mlxsw_sp *mlxsw_sp) +{ + mlxsw_sp->bridge->bridge_8021ad_ops = &mlxsw_sp1_bridge_8021ad_ops; +} + +const struct mlxsw_sp_switchdev_ops mlxsw_sp1_switchdev_ops = { + .init = mlxsw_sp1_switchdev_init, +}; + +static void mlxsw_sp2_switchdev_init(struct mlxsw_sp *mlxsw_sp) +{ + mlxsw_sp->bridge->bridge_8021ad_ops = &mlxsw_sp2_bridge_8021ad_ops; +} + +const struct mlxsw_sp_switchdev_ops mlxsw_sp2_switchdev_ops = { + .init = mlxsw_sp2_switchdev_init, +}; + +int mlxsw_sp_switchdev_init(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_bridge *bridge; + + bridge = kzalloc(sizeof(*mlxsw_sp->bridge), GFP_KERNEL); + if (!bridge) + return -ENOMEM; + mlxsw_sp->bridge = bridge; + bridge->mlxsw_sp = mlxsw_sp; + + INIT_LIST_HEAD(&mlxsw_sp->bridge->bridges_list); + + bridge->bridge_8021q_ops = &mlxsw_sp_bridge_8021q_ops; + bridge->bridge_8021d_ops = &mlxsw_sp_bridge_8021d_ops; + + mlxsw_sp->switchdev_ops->init(mlxsw_sp); + + return mlxsw_sp_fdb_init(mlxsw_sp); +} + +void mlxsw_sp_switchdev_fini(struct mlxsw_sp *mlxsw_sp) +{ + mlxsw_sp_fdb_fini(mlxsw_sp); + WARN_ON(!list_empty(&mlxsw_sp->bridge->bridges_list)); + kfree(mlxsw_sp->bridge); +} + diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.h new file mode 100644 index 000000000..c218e10bd --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.h @@ -0,0 +1,13 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2018 Mellanox Technologies. All rights reserved */ + +#include <linux/netdevice.h> + +struct mlxsw_sp_bridge; +struct mlxsw_sp_bridge_port; + +struct mlxsw_sp_bridge_port * +mlxsw_sp_bridge_port_find(struct mlxsw_sp_bridge *bridge, + struct net_device *brport_dev); + +u8 mlxsw_sp_bridge_port_stp_state(struct mlxsw_sp_bridge_port *bridge_port); diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.c new file mode 100644 index 000000000..f4bfdb6da --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.c @@ -0,0 +1,1969 @@ +// SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 +/* Copyright (c) 2019 Mellanox Technologies. All rights reserved */ + +#include <linux/bitops.h> +#include <linux/kernel.h> +#include <linux/netlink.h> +#include <net/devlink.h> +#include <uapi/linux/devlink.h> + +#include "core.h" +#include "reg.h" +#include "spectrum.h" +#include "spectrum_trap.h" + +struct mlxsw_sp_trap_policer_item { + struct devlink_trap_policer policer; + u16 hw_id; +}; + +struct mlxsw_sp_trap_group_item { + struct devlink_trap_group group; + u16 hw_group_id; + u8 priority; + u8 fixed_policer:1; /* Whether policer binding can change */ +}; + +#define MLXSW_SP_TRAP_LISTENERS_MAX 3 + +struct mlxsw_sp_trap_item { + struct devlink_trap trap; + struct mlxsw_listener listeners_arr[MLXSW_SP_TRAP_LISTENERS_MAX]; + u8 is_source:1; +}; + +/* All driver-specific traps must be documented in + * Documentation/networking/devlink/mlxsw.rst + */ +enum { + DEVLINK_MLXSW_TRAP_ID_BASE = DEVLINK_TRAP_GENERIC_ID_MAX, + DEVLINK_MLXSW_TRAP_ID_IRIF_DISABLED, + DEVLINK_MLXSW_TRAP_ID_ERIF_DISABLED, +}; + +#define DEVLINK_MLXSW_TRAP_NAME_IRIF_DISABLED \ + "irif_disabled" +#define DEVLINK_MLXSW_TRAP_NAME_ERIF_DISABLED \ + "erif_disabled" + +#define MLXSW_SP_TRAP_METADATA DEVLINK_TRAP_METADATA_TYPE_F_IN_PORT + +enum { + /* Packet was mirrored from ingress. */ + MLXSW_SP_MIRROR_REASON_INGRESS = 1, + /* Packet was mirrored from policy engine. */ + MLXSW_SP_MIRROR_REASON_POLICY_ENGINE = 2, + /* Packet was early dropped. */ + MLXSW_SP_MIRROR_REASON_INGRESS_WRED = 9, + /* Packet was mirrored from egress. */ + MLXSW_SP_MIRROR_REASON_EGRESS = 14, +}; + +static int mlxsw_sp_rx_listener(struct mlxsw_sp *mlxsw_sp, struct sk_buff *skb, + u16 local_port, + struct mlxsw_sp_port *mlxsw_sp_port) +{ + struct mlxsw_sp_port_pcpu_stats *pcpu_stats; + + if (unlikely(!mlxsw_sp_port)) { + dev_warn_ratelimited(mlxsw_sp->bus_info->dev, "Port %d: skb received for non-existent port\n", + local_port); + kfree_skb(skb); + return -EINVAL; + } + + skb->dev = mlxsw_sp_port->dev; + + pcpu_stats = this_cpu_ptr(mlxsw_sp_port->pcpu_stats); + u64_stats_update_begin(&pcpu_stats->syncp); + pcpu_stats->rx_packets++; + pcpu_stats->rx_bytes += skb->len; + u64_stats_update_end(&pcpu_stats->syncp); + + skb->protocol = eth_type_trans(skb, skb->dev); + + return 0; +} + +static void mlxsw_sp_rx_drop_listener(struct sk_buff *skb, u16 local_port, + void *trap_ctx) +{ + struct devlink_port *in_devlink_port; + struct mlxsw_sp_port *mlxsw_sp_port; + struct mlxsw_sp *mlxsw_sp; + struct devlink *devlink; + int err; + + mlxsw_sp = devlink_trap_ctx_priv(trap_ctx); + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + + err = mlxsw_sp_rx_listener(mlxsw_sp, skb, local_port, mlxsw_sp_port); + if (err) + return; + + devlink = priv_to_devlink(mlxsw_sp->core); + in_devlink_port = mlxsw_core_port_devlink_port_get(mlxsw_sp->core, + local_port); + skb_push(skb, ETH_HLEN); + devlink_trap_report(devlink, skb, trap_ctx, in_devlink_port, NULL); + consume_skb(skb); +} + +static void mlxsw_sp_rx_acl_drop_listener(struct sk_buff *skb, u16 local_port, + void *trap_ctx) +{ + u32 cookie_index = mlxsw_skb_cb(skb)->rx_md_info.cookie_index; + const struct flow_action_cookie *fa_cookie; + struct devlink_port *in_devlink_port; + struct mlxsw_sp_port *mlxsw_sp_port; + struct mlxsw_sp *mlxsw_sp; + struct devlink *devlink; + int err; + + mlxsw_sp = devlink_trap_ctx_priv(trap_ctx); + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + + err = mlxsw_sp_rx_listener(mlxsw_sp, skb, local_port, mlxsw_sp_port); + if (err) + return; + + devlink = priv_to_devlink(mlxsw_sp->core); + in_devlink_port = mlxsw_core_port_devlink_port_get(mlxsw_sp->core, + local_port); + skb_push(skb, ETH_HLEN); + rcu_read_lock(); + fa_cookie = mlxsw_sp_acl_act_cookie_lookup(mlxsw_sp, cookie_index); + devlink_trap_report(devlink, skb, trap_ctx, in_devlink_port, fa_cookie); + rcu_read_unlock(); + consume_skb(skb); +} + +static int __mlxsw_sp_rx_no_mark_listener(struct sk_buff *skb, u16 local_port, + void *trap_ctx) +{ + struct devlink_port *in_devlink_port; + struct mlxsw_sp_port *mlxsw_sp_port; + struct mlxsw_sp *mlxsw_sp; + struct devlink *devlink; + int err; + + mlxsw_sp = devlink_trap_ctx_priv(trap_ctx); + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + + err = mlxsw_sp_rx_listener(mlxsw_sp, skb, local_port, mlxsw_sp_port); + if (err) + return err; + + devlink = priv_to_devlink(mlxsw_sp->core); + in_devlink_port = mlxsw_core_port_devlink_port_get(mlxsw_sp->core, + local_port); + skb_push(skb, ETH_HLEN); + devlink_trap_report(devlink, skb, trap_ctx, in_devlink_port, NULL); + skb_pull(skb, ETH_HLEN); + + return 0; +} + +static void mlxsw_sp_rx_no_mark_listener(struct sk_buff *skb, u16 local_port, + void *trap_ctx) +{ + int err; + + err = __mlxsw_sp_rx_no_mark_listener(skb, local_port, trap_ctx); + if (err) + return; + + netif_receive_skb(skb); +} + +static void mlxsw_sp_rx_mark_listener(struct sk_buff *skb, u16 local_port, + void *trap_ctx) +{ + skb->offload_fwd_mark = 1; + mlxsw_sp_rx_no_mark_listener(skb, local_port, trap_ctx); +} + +static void mlxsw_sp_rx_l3_mark_listener(struct sk_buff *skb, u16 local_port, + void *trap_ctx) +{ + skb->offload_l3_fwd_mark = 1; + skb->offload_fwd_mark = 1; + mlxsw_sp_rx_no_mark_listener(skb, local_port, trap_ctx); +} + +static void mlxsw_sp_rx_ptp_listener(struct sk_buff *skb, u16 local_port, + void *trap_ctx) +{ + struct mlxsw_sp *mlxsw_sp = devlink_trap_ctx_priv(trap_ctx); + int err; + + err = __mlxsw_sp_rx_no_mark_listener(skb, local_port, trap_ctx); + if (err) + return; + + /* The PTP handler expects skb->data to point to the start of the + * Ethernet header. + */ + skb_push(skb, ETH_HLEN); + mlxsw_sp_ptp_receive(mlxsw_sp, skb, local_port); +} + +static struct mlxsw_sp_port * +mlxsw_sp_sample_tx_port_get(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_rx_md_info *rx_md_info) +{ + u16 local_port; + + if (!rx_md_info->tx_port_valid) + return NULL; + + if (rx_md_info->tx_port_is_lag) + local_port = mlxsw_core_lag_mapping_get(mlxsw_sp->core, + rx_md_info->tx_lag_id, + rx_md_info->tx_lag_port_index); + else + local_port = rx_md_info->tx_sys_port; + + if (local_port >= mlxsw_core_max_ports(mlxsw_sp->core)) + return NULL; + + return mlxsw_sp->ports[local_port]; +} + +/* The latency units are determined according to MOGCR.mirror_latency_units. It + * defaults to 64 nanoseconds. + */ +#define MLXSW_SP_MIRROR_LATENCY_SHIFT 6 + +static void mlxsw_sp_psample_md_init(struct mlxsw_sp *mlxsw_sp, + struct psample_metadata *md, + struct sk_buff *skb, int in_ifindex, + bool truncate, u32 trunc_size) +{ + struct mlxsw_rx_md_info *rx_md_info = &mlxsw_skb_cb(skb)->rx_md_info; + struct mlxsw_sp_port *mlxsw_sp_port; + + md->trunc_size = truncate ? trunc_size : skb->len; + md->in_ifindex = in_ifindex; + mlxsw_sp_port = mlxsw_sp_sample_tx_port_get(mlxsw_sp, rx_md_info); + md->out_ifindex = mlxsw_sp_port && mlxsw_sp_port->dev ? + mlxsw_sp_port->dev->ifindex : 0; + md->out_tc_valid = rx_md_info->tx_tc_valid; + md->out_tc = rx_md_info->tx_tc; + md->out_tc_occ_valid = rx_md_info->tx_congestion_valid; + md->out_tc_occ = rx_md_info->tx_congestion; + md->latency_valid = rx_md_info->latency_valid; + md->latency = rx_md_info->latency; + md->latency <<= MLXSW_SP_MIRROR_LATENCY_SHIFT; +} + +static void mlxsw_sp_rx_sample_listener(struct sk_buff *skb, u16 local_port, + void *trap_ctx) +{ + struct mlxsw_sp *mlxsw_sp = devlink_trap_ctx_priv(trap_ctx); + struct mlxsw_sp_sample_trigger trigger; + struct mlxsw_sp_sample_params *params; + struct mlxsw_sp_port *mlxsw_sp_port; + struct psample_metadata md = {}; + int err; + + err = __mlxsw_sp_rx_no_mark_listener(skb, local_port, trap_ctx); + if (err) + return; + + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + if (!mlxsw_sp_port) + goto out; + + trigger.type = MLXSW_SP_SAMPLE_TRIGGER_TYPE_INGRESS; + trigger.local_port = local_port; + params = mlxsw_sp_sample_trigger_params_lookup(mlxsw_sp, &trigger); + if (!params) + goto out; + + /* The psample module expects skb->data to point to the start of the + * Ethernet header. + */ + skb_push(skb, ETH_HLEN); + mlxsw_sp_psample_md_init(mlxsw_sp, &md, skb, + mlxsw_sp_port->dev->ifindex, params->truncate, + params->trunc_size); + psample_sample_packet(params->psample_group, skb, params->rate, &md); +out: + consume_skb(skb); +} + +static void mlxsw_sp_rx_sample_tx_listener(struct sk_buff *skb, u16 local_port, + void *trap_ctx) +{ + struct mlxsw_rx_md_info *rx_md_info = &mlxsw_skb_cb(skb)->rx_md_info; + struct mlxsw_sp *mlxsw_sp = devlink_trap_ctx_priv(trap_ctx); + struct mlxsw_sp_port *mlxsw_sp_port, *mlxsw_sp_port_tx; + struct mlxsw_sp_sample_trigger trigger; + struct mlxsw_sp_sample_params *params; + struct psample_metadata md = {}; + int err; + + /* Locally generated packets are not reported from the policy engine + * trigger, so do not report them from the egress trigger as well. + */ + if (local_port == MLXSW_PORT_CPU_PORT) + goto out; + + err = __mlxsw_sp_rx_no_mark_listener(skb, local_port, trap_ctx); + if (err) + return; + + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + if (!mlxsw_sp_port) + goto out; + + /* Packet was sampled from Tx, so we need to retrieve the sample + * parameters based on the Tx port and not the Rx port. + */ + mlxsw_sp_port_tx = mlxsw_sp_sample_tx_port_get(mlxsw_sp, rx_md_info); + if (!mlxsw_sp_port_tx) + goto out; + + trigger.type = MLXSW_SP_SAMPLE_TRIGGER_TYPE_EGRESS; + trigger.local_port = mlxsw_sp_port_tx->local_port; + params = mlxsw_sp_sample_trigger_params_lookup(mlxsw_sp, &trigger); + if (!params) + goto out; + + /* The psample module expects skb->data to point to the start of the + * Ethernet header. + */ + skb_push(skb, ETH_HLEN); + mlxsw_sp_psample_md_init(mlxsw_sp, &md, skb, + mlxsw_sp_port->dev->ifindex, params->truncate, + params->trunc_size); + psample_sample_packet(params->psample_group, skb, params->rate, &md); +out: + consume_skb(skb); +} + +static void mlxsw_sp_rx_sample_acl_listener(struct sk_buff *skb, u16 local_port, + void *trap_ctx) +{ + struct mlxsw_sp *mlxsw_sp = devlink_trap_ctx_priv(trap_ctx); + struct mlxsw_sp_sample_trigger trigger = { + .type = MLXSW_SP_SAMPLE_TRIGGER_TYPE_POLICY_ENGINE, + }; + struct mlxsw_sp_sample_params *params; + struct mlxsw_sp_port *mlxsw_sp_port; + struct psample_metadata md = {}; + int err; + + err = __mlxsw_sp_rx_no_mark_listener(skb, local_port, trap_ctx); + if (err) + return; + + mlxsw_sp_port = mlxsw_sp->ports[local_port]; + if (!mlxsw_sp_port) + goto out; + + params = mlxsw_sp_sample_trigger_params_lookup(mlxsw_sp, &trigger); + if (!params) + goto out; + + /* The psample module expects skb->data to point to the start of the + * Ethernet header. + */ + skb_push(skb, ETH_HLEN); + mlxsw_sp_psample_md_init(mlxsw_sp, &md, skb, + mlxsw_sp_port->dev->ifindex, params->truncate, + params->trunc_size); + psample_sample_packet(params->psample_group, skb, params->rate, &md); +out: + consume_skb(skb); +} + +#define MLXSW_SP_TRAP_DROP(_id, _group_id) \ + DEVLINK_TRAP_GENERIC(DROP, DROP, _id, \ + DEVLINK_TRAP_GROUP_GENERIC_ID_##_group_id, \ + MLXSW_SP_TRAP_METADATA) + +#define MLXSW_SP_TRAP_DROP_EXT(_id, _group_id, _metadata) \ + DEVLINK_TRAP_GENERIC(DROP, DROP, _id, \ + DEVLINK_TRAP_GROUP_GENERIC_ID_##_group_id, \ + MLXSW_SP_TRAP_METADATA | (_metadata)) + +#define MLXSW_SP_TRAP_BUFFER_DROP(_id) \ + DEVLINK_TRAP_GENERIC(DROP, TRAP, _id, \ + DEVLINK_TRAP_GROUP_GENERIC_ID_BUFFER_DROPS, \ + MLXSW_SP_TRAP_METADATA) + +#define MLXSW_SP_TRAP_DRIVER_DROP(_id, _group_id) \ + DEVLINK_TRAP_DRIVER(DROP, DROP, DEVLINK_MLXSW_TRAP_ID_##_id, \ + DEVLINK_MLXSW_TRAP_NAME_##_id, \ + DEVLINK_TRAP_GROUP_GENERIC_ID_##_group_id, \ + MLXSW_SP_TRAP_METADATA) + +#define MLXSW_SP_TRAP_EXCEPTION(_id, _group_id) \ + DEVLINK_TRAP_GENERIC(EXCEPTION, TRAP, _id, \ + DEVLINK_TRAP_GROUP_GENERIC_ID_##_group_id, \ + MLXSW_SP_TRAP_METADATA) + +#define MLXSW_SP_TRAP_CONTROL(_id, _group_id, _action) \ + DEVLINK_TRAP_GENERIC(CONTROL, _action, _id, \ + DEVLINK_TRAP_GROUP_GENERIC_ID_##_group_id, \ + MLXSW_SP_TRAP_METADATA) + +#define MLXSW_SP_RXL_DISCARD(_id, _group_id) \ + MLXSW_RXL_DIS(mlxsw_sp_rx_drop_listener, DISCARD_##_id, \ + TRAP_EXCEPTION_TO_CPU, false, SP_##_group_id, \ + SET_FW_DEFAULT, SP_##_group_id) + +#define MLXSW_SP_RXL_ACL_DISCARD(_id, _en_group_id, _dis_group_id) \ + MLXSW_RXL_DIS(mlxsw_sp_rx_acl_drop_listener, DISCARD_##_id, \ + TRAP_EXCEPTION_TO_CPU, false, SP_##_en_group_id, \ + SET_FW_DEFAULT, SP_##_dis_group_id) + +#define MLXSW_SP_RXL_BUFFER_DISCARD(_mirror_reason) \ + MLXSW_RXL_MIRROR(mlxsw_sp_rx_drop_listener, 0, SP_BUFFER_DISCARDS, \ + MLXSW_SP_MIRROR_REASON_##_mirror_reason) + +#define MLXSW_SP_RXL_EXCEPTION(_id, _group_id, _action) \ + MLXSW_RXL(mlxsw_sp_rx_mark_listener, _id, \ + _action, false, SP_##_group_id, SET_FW_DEFAULT) + +#define MLXSW_SP_RXL_NO_MARK(_id, _group_id, _action, _is_ctrl) \ + MLXSW_RXL(mlxsw_sp_rx_no_mark_listener, _id, _action, \ + _is_ctrl, SP_##_group_id, DISCARD) + +#define MLXSW_SP_RXL_MARK(_id, _group_id, _action, _is_ctrl) \ + MLXSW_RXL(mlxsw_sp_rx_mark_listener, _id, _action, _is_ctrl, \ + SP_##_group_id, DISCARD) + +#define MLXSW_SP_RXL_L3_MARK(_id, _group_id, _action, _is_ctrl) \ + MLXSW_RXL(mlxsw_sp_rx_l3_mark_listener, _id, _action, _is_ctrl, \ + SP_##_group_id, DISCARD) + +#define MLXSW_SP_TRAP_POLICER(_id, _rate, _burst) \ + DEVLINK_TRAP_POLICER(_id, _rate, _burst, \ + MLXSW_REG_QPCR_HIGHEST_CIR, \ + MLXSW_REG_QPCR_LOWEST_CIR, \ + 1 << MLXSW_REG_QPCR_HIGHEST_CBS, \ + 1 << MLXSW_REG_QPCR_LOWEST_CBS) + +/* Ordered by policer identifier */ +static const struct mlxsw_sp_trap_policer_item +mlxsw_sp_trap_policer_items_arr[] = { + { + .policer = MLXSW_SP_TRAP_POLICER(1, 10 * 1024, 4096), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(2, 128, 128), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(3, 128, 128), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(4, 128, 128), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(5, 16 * 1024, 8192), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(6, 128, 128), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(7, 1024, 512), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(8, 20 * 1024, 8192), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(9, 128, 128), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(10, 1024, 512), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(11, 256, 128), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(12, 128, 128), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(13, 128, 128), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(14, 1024, 512), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(15, 1024, 512), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(16, 24 * 1024, 16384), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(17, 19 * 1024, 8192), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(18, 1024, 512), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(19, 1024, 512), + }, + { + .policer = MLXSW_SP_TRAP_POLICER(20, 10240, 4096), + }, +}; + +static const struct mlxsw_sp_trap_group_item mlxsw_sp_trap_group_items_arr[] = { + { + .group = DEVLINK_TRAP_GROUP_GENERIC(L2_DROPS, 1), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_L2_DISCARDS, + .priority = 0, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(L3_DROPS, 1), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_L3_DISCARDS, + .priority = 0, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(L3_EXCEPTIONS, 1), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_L3_EXCEPTIONS, + .priority = 2, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(TUNNEL_DROPS, 1), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_TUNNEL_DISCARDS, + .priority = 0, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(ACL_DROPS, 1), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_ACL_DISCARDS, + .priority = 0, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(STP, 2), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_STP, + .priority = 5, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(LACP, 3), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_LACP, + .priority = 5, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(LLDP, 4), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_LLDP, + .priority = 5, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(MC_SNOOPING, 5), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_MC_SNOOPING, + .priority = 3, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(DHCP, 6), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_DHCP, + .priority = 2, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(NEIGH_DISCOVERY, 7), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_NEIGH_DISCOVERY, + .priority = 2, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(BFD, 8), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_BFD, + .priority = 5, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(OSPF, 9), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_OSPF, + .priority = 5, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(BGP, 10), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_BGP, + .priority = 4, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(VRRP, 11), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_VRRP, + .priority = 5, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(PIM, 12), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_PIM, + .priority = 5, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(UC_LB, 13), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_LBERROR, + .priority = 0, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(LOCAL_DELIVERY, 14), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_IP2ME, + .priority = 2, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(EXTERNAL_DELIVERY, 19), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_EXTERNAL_ROUTE, + .priority = 1, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(IPV6, 15), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_IPV6, + .priority = 2, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(PTP_EVENT, 16), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_PTP0, + .priority = 5, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(PTP_GENERAL, 17), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_PTP1, + .priority = 2, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(ACL_TRAP, 18), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_FLOW_LOGGING, + .priority = 4, + }, +}; + +static const struct mlxsw_sp_trap_item mlxsw_sp_trap_items_arr[] = { + { + .trap = MLXSW_SP_TRAP_DROP(SMAC_MC, L2_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ING_PACKET_SMAC_MC, L2_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(VLAN_TAG_MISMATCH, L2_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ING_SWITCH_VTAG_ALLOW, + L2_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(INGRESS_VLAN_FILTER, L2_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ING_SWITCH_VLAN, L2_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(INGRESS_STP_FILTER, L2_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ING_SWITCH_STP, L2_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(EMPTY_TX_LIST, L2_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(LOOKUP_SWITCH_UC, L2_DISCARDS), + MLXSW_SP_RXL_DISCARD(LOOKUP_SWITCH_MC_NULL, L2_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(PORT_LOOPBACK_FILTER, L2_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(LOOKUP_SWITCH_LB, L2_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(BLACKHOLE_ROUTE, L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ROUTER2, L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(NON_IP_PACKET, L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ING_ROUTER_NON_IP_PACKET, + L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(UC_DIP_MC_DMAC, L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ING_ROUTER_UC_DIP_MC_DMAC, + L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(DIP_LB, L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ING_ROUTER_DIP_LB, L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(SIP_MC, L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ING_ROUTER_SIP_MC, L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(SIP_LB, L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ING_ROUTER_SIP_LB, L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(CORRUPTED_IP_HDR, L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ING_ROUTER_CORRUPTED_IP_HDR, + L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(IPV4_SIP_BC, L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ING_ROUTER_IPV4_SIP_BC, + L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(IPV6_MC_DIP_RESERVED_SCOPE, + L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(IPV6_MC_DIP_RESERVED_SCOPE, + L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(IPV6_MC_DIP_INTERFACE_LOCAL_SCOPE, + L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(IPV6_MC_DIP_INTERFACE_LOCAL_SCOPE, + L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_EXCEPTION(MTU_ERROR, L3_EXCEPTIONS), + .listeners_arr = { + MLXSW_SP_RXL_EXCEPTION(MTUERROR, L3_EXCEPTIONS, + TRAP_TO_CPU), + }, + }, + { + .trap = MLXSW_SP_TRAP_EXCEPTION(TTL_ERROR, L3_EXCEPTIONS), + .listeners_arr = { + MLXSW_SP_RXL_EXCEPTION(TTLERROR, L3_EXCEPTIONS, + TRAP_TO_CPU), + }, + }, + { + .trap = MLXSW_SP_TRAP_EXCEPTION(RPF, L3_EXCEPTIONS), + .listeners_arr = { + MLXSW_SP_RXL_EXCEPTION(RPF, L3_EXCEPTIONS, TRAP_TO_CPU), + }, + }, + { + .trap = MLXSW_SP_TRAP_EXCEPTION(REJECT_ROUTE, L3_EXCEPTIONS), + .listeners_arr = { + MLXSW_SP_RXL_EXCEPTION(RTR_INGRESS1, L3_EXCEPTIONS, + TRAP_TO_CPU), + }, + }, + { + .trap = MLXSW_SP_TRAP_EXCEPTION(UNRESOLVED_NEIGH, + L3_EXCEPTIONS), + .listeners_arr = { + MLXSW_SP_RXL_EXCEPTION(HOST_MISS_IPV4, L3_EXCEPTIONS, + TRAP_TO_CPU), + MLXSW_SP_RXL_EXCEPTION(HOST_MISS_IPV6, L3_EXCEPTIONS, + TRAP_TO_CPU), + MLXSW_SP_RXL_EXCEPTION(RTR_EGRESS0, L3_EXCEPTIONS, + TRAP_EXCEPTION_TO_CPU), + }, + }, + { + .trap = MLXSW_SP_TRAP_EXCEPTION(IPV4_LPM_UNICAST_MISS, + L3_EXCEPTIONS), + .listeners_arr = { + MLXSW_SP_RXL_EXCEPTION(DISCARD_ROUTER_LPM4, + L3_EXCEPTIONS, + TRAP_EXCEPTION_TO_CPU), + }, + }, + { + .trap = MLXSW_SP_TRAP_EXCEPTION(IPV6_LPM_UNICAST_MISS, + L3_EXCEPTIONS), + .listeners_arr = { + MLXSW_SP_RXL_EXCEPTION(DISCARD_ROUTER_LPM6, + L3_EXCEPTIONS, + TRAP_EXCEPTION_TO_CPU), + }, + }, + { + .trap = MLXSW_SP_TRAP_DRIVER_DROP(IRIF_DISABLED, L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ROUTER_IRIF_EN, L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DRIVER_DROP(ERIF_DISABLED, L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ROUTER_ERIF_EN, L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(NON_ROUTABLE, L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(NON_ROUTABLE, L3_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_EXCEPTION(DECAP_ERROR, TUNNEL_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_EXCEPTION(DECAP_ECN0, TUNNEL_DISCARDS, + TRAP_EXCEPTION_TO_CPU), + MLXSW_SP_RXL_EXCEPTION(IPIP_DECAP_ERROR, + TUNNEL_DISCARDS, + TRAP_EXCEPTION_TO_CPU), + MLXSW_SP_RXL_EXCEPTION(DISCARD_DEC_PKT, TUNNEL_DISCARDS, + TRAP_EXCEPTION_TO_CPU), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(OVERLAY_SMAC_MC, TUNNEL_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(OVERLAY_SMAC_MC, TUNNEL_DISCARDS), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP_EXT(INGRESS_FLOW_ACTION_DROP, + ACL_DROPS, + DEVLINK_TRAP_METADATA_TYPE_F_FA_COOKIE), + .listeners_arr = { + MLXSW_SP_RXL_ACL_DISCARD(INGRESS_ACL, ACL_DISCARDS, + DUMMY), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP_EXT(EGRESS_FLOW_ACTION_DROP, + ACL_DROPS, + DEVLINK_TRAP_METADATA_TYPE_F_FA_COOKIE), + .listeners_arr = { + MLXSW_SP_RXL_ACL_DISCARD(EGRESS_ACL, ACL_DISCARDS, + DUMMY), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(STP, STP, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_NO_MARK(STP, STP, TRAP_TO_CPU, true), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(LACP, LACP, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_NO_MARK(LACP, LACP, TRAP_TO_CPU, true), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(LLDP, LLDP, TRAP), + .listeners_arr = { + MLXSW_RXL(mlxsw_sp_rx_ptp_listener, LLDP, TRAP_TO_CPU, + true, SP_LLDP, DISCARD), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IGMP_QUERY, MC_SNOOPING, MIRROR), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IGMP_QUERY, MC_SNOOPING, + MIRROR_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IGMP_V1_REPORT, MC_SNOOPING, + TRAP), + .listeners_arr = { + MLXSW_SP_RXL_NO_MARK(IGMP_V1_REPORT, MC_SNOOPING, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IGMP_V2_REPORT, MC_SNOOPING, + TRAP), + .listeners_arr = { + MLXSW_SP_RXL_NO_MARK(IGMP_V2_REPORT, MC_SNOOPING, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IGMP_V3_REPORT, MC_SNOOPING, + TRAP), + .listeners_arr = { + MLXSW_SP_RXL_NO_MARK(IGMP_V3_REPORT, MC_SNOOPING, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IGMP_V2_LEAVE, MC_SNOOPING, + TRAP), + .listeners_arr = { + MLXSW_SP_RXL_NO_MARK(IGMP_V2_LEAVE, MC_SNOOPING, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(MLD_QUERY, MC_SNOOPING, MIRROR), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV6_MLDV12_LISTENER_QUERY, + MC_SNOOPING, MIRROR_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(MLD_V1_REPORT, MC_SNOOPING, + TRAP), + .listeners_arr = { + MLXSW_SP_RXL_NO_MARK(IPV6_MLDV1_LISTENER_REPORT, + MC_SNOOPING, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(MLD_V2_REPORT, MC_SNOOPING, + TRAP), + .listeners_arr = { + MLXSW_SP_RXL_NO_MARK(IPV6_MLDV2_LISTENER_REPORT, + MC_SNOOPING, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(MLD_V1_DONE, MC_SNOOPING, + TRAP), + .listeners_arr = { + MLXSW_SP_RXL_NO_MARK(IPV6_MLDV1_LISTENER_DONE, + MC_SNOOPING, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV4_DHCP, DHCP, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV4_DHCP, DHCP, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_DHCP, DHCP, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV6_DHCP, DHCP, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(ARP_REQUEST, NEIGH_DISCOVERY, + MIRROR), + .listeners_arr = { + MLXSW_SP_RXL_MARK(ROUTER_ARPBC, NEIGH_DISCOVERY, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(ARP_RESPONSE, NEIGH_DISCOVERY, + MIRROR), + .listeners_arr = { + MLXSW_SP_RXL_MARK(ROUTER_ARPUC, NEIGH_DISCOVERY, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(ARP_OVERLAY, NEIGH_DISCOVERY, + TRAP), + .listeners_arr = { + MLXSW_SP_RXL_NO_MARK(NVE_DECAP_ARP, NEIGH_DISCOVERY, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_NEIGH_SOLICIT, + NEIGH_DISCOVERY, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(L3_IPV6_NEIGHBOR_SOLICITATION, + NEIGH_DISCOVERY, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_NEIGH_ADVERT, + NEIGH_DISCOVERY, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(L3_IPV6_NEIGHBOR_ADVERTISEMENT, + NEIGH_DISCOVERY, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV4_BFD, BFD, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV4_BFD, BFD, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_BFD, BFD, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV6_BFD, BFD, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV4_OSPF, OSPF, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV4_OSPF, OSPF, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_OSPF, OSPF, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV6_OSPF, OSPF, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV4_BGP, BGP, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV4_BGP, BGP, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_BGP, BGP, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV6_BGP, BGP, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV4_VRRP, VRRP, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV4_VRRP, VRRP, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_VRRP, VRRP, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV6_VRRP, VRRP, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV4_PIM, PIM, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV4_PIM, PIM, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_PIM, PIM, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV6_PIM, PIM, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(UC_LB, UC_LB, MIRROR), + .listeners_arr = { + MLXSW_SP_RXL_L3_MARK(LBERROR, LBERROR, MIRROR_TO_CPU, + false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(LOCAL_ROUTE, LOCAL_DELIVERY, + TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IP2ME, IP2ME, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(EXTERNAL_ROUTE, EXTERNAL_DELIVERY, + TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(RTR_INGRESS0, EXTERNAL_ROUTE, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_UC_DIP_LINK_LOCAL_SCOPE, + LOCAL_DELIVERY, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV6_LINK_LOCAL_DEST, IP2ME, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV4_ROUTER_ALERT, LOCAL_DELIVERY, + TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(ROUTER_ALERT_IPV4, IP2ME, TRAP_TO_CPU, + false), + }, + }, + { + /* IPV6_ROUTER_ALERT is defined in uAPI as 22, but it is not + * used in this file, so undefine it. + */ + #undef IPV6_ROUTER_ALERT + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_ROUTER_ALERT, LOCAL_DELIVERY, + TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(ROUTER_ALERT_IPV6, IP2ME, TRAP_TO_CPU, + false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_DIP_ALL_NODES, IPV6, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV6_ALL_NODES_LINK, IPV6, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_DIP_ALL_ROUTERS, IPV6, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(IPV6_ALL_ROUTERS_LINK, IPV6, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_ROUTER_SOLICIT, IPV6, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(L3_IPV6_ROUTER_SOLICITATION, IPV6, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_ROUTER_ADVERT, IPV6, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(L3_IPV6_ROUTER_ADVERTISEMENT, IPV6, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(IPV6_REDIRECT, IPV6, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_MARK(L3_IPV6_REDIRECTION, IPV6, + TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(PTP_EVENT, PTP_EVENT, TRAP), + .listeners_arr = { + MLXSW_RXL(mlxsw_sp_rx_ptp_listener, PTP0, TRAP_TO_CPU, + false, SP_PTP0, DISCARD), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(PTP_GENERAL, PTP_GENERAL, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_NO_MARK(PTP1, PTP1, TRAP_TO_CPU, false), + }, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(FLOW_ACTION_TRAP, ACL_TRAP, TRAP), + .listeners_arr = { + MLXSW_SP_RXL_NO_MARK(ACL0, FLOW_LOGGING, TRAP_TO_CPU, + false), + }, + }, + { + .trap = MLXSW_SP_TRAP_DROP(BLACKHOLE_NEXTHOP, L3_DROPS), + .listeners_arr = { + MLXSW_SP_RXL_DISCARD(ROUTER3, L3_DISCARDS), + }, + }, +}; + +static struct mlxsw_sp_trap_policer_item * +mlxsw_sp_trap_policer_item_lookup(struct mlxsw_sp *mlxsw_sp, u32 id) +{ + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + int i; + + for (i = 0; i < trap->policers_count; i++) { + if (trap->policer_items_arr[i].policer.id == id) + return &trap->policer_items_arr[i]; + } + + return NULL; +} + +static struct mlxsw_sp_trap_group_item * +mlxsw_sp_trap_group_item_lookup(struct mlxsw_sp *mlxsw_sp, u16 id) +{ + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + int i; + + for (i = 0; i < trap->groups_count; i++) { + if (trap->group_items_arr[i].group.id == id) + return &trap->group_items_arr[i]; + } + + return NULL; +} + +static struct mlxsw_sp_trap_item * +mlxsw_sp_trap_item_lookup(struct mlxsw_sp *mlxsw_sp, u16 id) +{ + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + int i; + + for (i = 0; i < trap->traps_count; i++) { + if (trap->trap_items_arr[i].trap.id == id) + return &trap->trap_items_arr[i]; + } + + return NULL; +} + +static int mlxsw_sp_trap_cpu_policers_set(struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + char qpcr_pl[MLXSW_REG_QPCR_LEN]; + u16 hw_id; + + /* The purpose of "thin" policer is to drop as many packets + * as possible. The dummy group is using it. + */ + hw_id = find_first_zero_bit(trap->policers_usage, trap->max_policers); + if (WARN_ON(hw_id == trap->max_policers)) + return -ENOBUFS; + + __set_bit(hw_id, trap->policers_usage); + trap->thin_policer_hw_id = hw_id; + mlxsw_reg_qpcr_pack(qpcr_pl, hw_id, MLXSW_REG_QPCR_IR_UNITS_M, + false, 1, 4); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qpcr), qpcr_pl); +} + +static int mlxsw_sp_trap_dummy_group_init(struct mlxsw_sp *mlxsw_sp) +{ + char htgt_pl[MLXSW_REG_HTGT_LEN]; + + mlxsw_reg_htgt_pack(htgt_pl, MLXSW_REG_HTGT_TRAP_GROUP_SP_DUMMY, + mlxsw_sp->trap->thin_policer_hw_id, 0, 1); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(htgt), htgt_pl); +} + +static int mlxsw_sp_trap_policer_items_arr_init(struct mlxsw_sp *mlxsw_sp) +{ + size_t arr_size = ARRAY_SIZE(mlxsw_sp_trap_policer_items_arr); + size_t elem_size = sizeof(struct mlxsw_sp_trap_policer_item); + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + size_t free_policers = 0; + u32 last_id; + int i; + + for_each_clear_bit(i, trap->policers_usage, trap->max_policers) + free_policers++; + + if (arr_size > free_policers) { + dev_err(mlxsw_sp->bus_info->dev, "Exceeded number of supported packet trap policers\n"); + return -ENOBUFS; + } + + trap->policer_items_arr = kcalloc(free_policers, elem_size, GFP_KERNEL); + if (!trap->policer_items_arr) + return -ENOMEM; + + trap->policers_count = free_policers; + + /* Initialize policer items array with pre-defined policers. */ + memcpy(trap->policer_items_arr, mlxsw_sp_trap_policer_items_arr, + elem_size * arr_size); + + /* Initialize policer items array with the rest of the available + * policers. + */ + last_id = mlxsw_sp_trap_policer_items_arr[arr_size - 1].policer.id; + for (i = arr_size; i < trap->policers_count; i++) { + const struct mlxsw_sp_trap_policer_item *policer_item; + + /* Use parameters set for first policer and override + * relevant ones. + */ + policer_item = &mlxsw_sp_trap_policer_items_arr[0]; + trap->policer_items_arr[i] = *policer_item; + trap->policer_items_arr[i].policer.id = ++last_id; + trap->policer_items_arr[i].policer.init_rate = 1; + trap->policer_items_arr[i].policer.init_burst = 16; + } + + return 0; +} + +static void mlxsw_sp_trap_policer_items_arr_fini(struct mlxsw_sp *mlxsw_sp) +{ + kfree(mlxsw_sp->trap->policer_items_arr); +} + +static int mlxsw_sp_trap_policers_init(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + const struct mlxsw_sp_trap_policer_item *policer_item; + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + int err, i; + + err = mlxsw_sp_trap_policer_items_arr_init(mlxsw_sp); + if (err) + return err; + + for (i = 0; i < trap->policers_count; i++) { + policer_item = &trap->policer_items_arr[i]; + err = devl_trap_policers_register(devlink, + &policer_item->policer, 1); + if (err) + goto err_trap_policer_register; + } + + return 0; + +err_trap_policer_register: + for (i--; i >= 0; i--) { + policer_item = &trap->policer_items_arr[i]; + devl_trap_policers_unregister(devlink, + &policer_item->policer, 1); + } + mlxsw_sp_trap_policer_items_arr_fini(mlxsw_sp); + return err; +} + +static void mlxsw_sp_trap_policers_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + const struct mlxsw_sp_trap_policer_item *policer_item; + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + int i; + + for (i = trap->policers_count - 1; i >= 0; i--) { + policer_item = &trap->policer_items_arr[i]; + devl_trap_policers_unregister(devlink, + &policer_item->policer, 1); + } + mlxsw_sp_trap_policer_items_arr_fini(mlxsw_sp); +} + +static int mlxsw_sp_trap_group_items_arr_init(struct mlxsw_sp *mlxsw_sp) +{ + size_t common_groups_count = ARRAY_SIZE(mlxsw_sp_trap_group_items_arr); + const struct mlxsw_sp_trap_group_item *spec_group_items_arr; + size_t elem_size = sizeof(struct mlxsw_sp_trap_group_item); + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + size_t groups_count, spec_groups_count; + int err; + + err = mlxsw_sp->trap_ops->groups_init(mlxsw_sp, &spec_group_items_arr, + &spec_groups_count); + if (err) + return err; + + /* The group items array is created by concatenating the common trap + * group items and the ASIC-specific trap group items. + */ + groups_count = common_groups_count + spec_groups_count; + trap->group_items_arr = kcalloc(groups_count, elem_size, GFP_KERNEL); + if (!trap->group_items_arr) + return -ENOMEM; + + memcpy(trap->group_items_arr, mlxsw_sp_trap_group_items_arr, + elem_size * common_groups_count); + memcpy(trap->group_items_arr + common_groups_count, + spec_group_items_arr, elem_size * spec_groups_count); + + trap->groups_count = groups_count; + + return 0; +} + +static void mlxsw_sp_trap_group_items_arr_fini(struct mlxsw_sp *mlxsw_sp) +{ + kfree(mlxsw_sp->trap->group_items_arr); +} + +static int mlxsw_sp_trap_groups_init(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + const struct mlxsw_sp_trap_group_item *group_item; + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + int err, i; + + err = mlxsw_sp_trap_group_items_arr_init(mlxsw_sp); + if (err) + return err; + + for (i = 0; i < trap->groups_count; i++) { + group_item = &trap->group_items_arr[i]; + err = devl_trap_groups_register(devlink, &group_item->group, 1); + if (err) + goto err_trap_group_register; + } + + return 0; + +err_trap_group_register: + for (i--; i >= 0; i--) { + group_item = &trap->group_items_arr[i]; + devl_trap_groups_unregister(devlink, &group_item->group, 1); + } + mlxsw_sp_trap_group_items_arr_fini(mlxsw_sp); + return err; +} + +static void mlxsw_sp_trap_groups_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + int i; + + for (i = trap->groups_count - 1; i >= 0; i--) { + const struct mlxsw_sp_trap_group_item *group_item; + + group_item = &trap->group_items_arr[i]; + devl_trap_groups_unregister(devlink, &group_item->group, 1); + } + mlxsw_sp_trap_group_items_arr_fini(mlxsw_sp); +} + +static bool +mlxsw_sp_trap_listener_is_valid(const struct mlxsw_listener *listener) +{ + return listener->trap_id != 0; +} + +static int mlxsw_sp_trap_items_arr_init(struct mlxsw_sp *mlxsw_sp) +{ + size_t common_traps_count = ARRAY_SIZE(mlxsw_sp_trap_items_arr); + const struct mlxsw_sp_trap_item *spec_trap_items_arr; + size_t elem_size = sizeof(struct mlxsw_sp_trap_item); + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + size_t traps_count, spec_traps_count; + int err; + + err = mlxsw_sp->trap_ops->traps_init(mlxsw_sp, &spec_trap_items_arr, + &spec_traps_count); + if (err) + return err; + + /* The trap items array is created by concatenating the common trap + * items and the ASIC-specific trap items. + */ + traps_count = common_traps_count + spec_traps_count; + trap->trap_items_arr = kcalloc(traps_count, elem_size, GFP_KERNEL); + if (!trap->trap_items_arr) + return -ENOMEM; + + memcpy(trap->trap_items_arr, mlxsw_sp_trap_items_arr, + elem_size * common_traps_count); + memcpy(trap->trap_items_arr + common_traps_count, + spec_trap_items_arr, elem_size * spec_traps_count); + + trap->traps_count = traps_count; + + return 0; +} + +static void mlxsw_sp_trap_items_arr_fini(struct mlxsw_sp *mlxsw_sp) +{ + kfree(mlxsw_sp->trap->trap_items_arr); +} + +static int mlxsw_sp_traps_init(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + const struct mlxsw_sp_trap_item *trap_item; + int err, i; + + err = mlxsw_sp_trap_items_arr_init(mlxsw_sp); + if (err) + return err; + + for (i = 0; i < trap->traps_count; i++) { + trap_item = &trap->trap_items_arr[i]; + err = devl_traps_register(devlink, &trap_item->trap, 1, + mlxsw_sp); + if (err) + goto err_trap_register; + } + + return 0; + +err_trap_register: + for (i--; i >= 0; i--) { + trap_item = &trap->trap_items_arr[i]; + devl_traps_unregister(devlink, &trap_item->trap, 1); + } + mlxsw_sp_trap_items_arr_fini(mlxsw_sp); + return err; +} + +static void mlxsw_sp_traps_fini(struct mlxsw_sp *mlxsw_sp) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + int i; + + for (i = trap->traps_count - 1; i >= 0; i--) { + const struct mlxsw_sp_trap_item *trap_item; + + trap_item = &trap->trap_items_arr[i]; + devl_traps_unregister(devlink, &trap_item->trap, 1); + } + mlxsw_sp_trap_items_arr_fini(mlxsw_sp); +} + +int mlxsw_sp_devlink_traps_init(struct mlxsw_sp *mlxsw_sp) +{ + int err; + + err = mlxsw_sp_trap_cpu_policers_set(mlxsw_sp); + if (err) + return err; + + err = mlxsw_sp_trap_dummy_group_init(mlxsw_sp); + if (err) + return err; + + err = mlxsw_sp_trap_policers_init(mlxsw_sp); + if (err) + return err; + + err = mlxsw_sp_trap_groups_init(mlxsw_sp); + if (err) + goto err_trap_groups_init; + + err = mlxsw_sp_traps_init(mlxsw_sp); + if (err) + goto err_traps_init; + + return 0; + +err_traps_init: + mlxsw_sp_trap_groups_fini(mlxsw_sp); +err_trap_groups_init: + mlxsw_sp_trap_policers_fini(mlxsw_sp); + return err; +} + +void mlxsw_sp_devlink_traps_fini(struct mlxsw_sp *mlxsw_sp) +{ + mlxsw_sp_traps_fini(mlxsw_sp); + mlxsw_sp_trap_groups_fini(mlxsw_sp); + mlxsw_sp_trap_policers_fini(mlxsw_sp); +} + +int mlxsw_sp_trap_init(struct mlxsw_core *mlxsw_core, + const struct devlink_trap *trap, void *trap_ctx) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + const struct mlxsw_sp_trap_item *trap_item; + int i; + + trap_item = mlxsw_sp_trap_item_lookup(mlxsw_sp, trap->id); + if (WARN_ON(!trap_item)) + return -EINVAL; + + for (i = 0; i < MLXSW_SP_TRAP_LISTENERS_MAX; i++) { + const struct mlxsw_listener *listener; + int err; + + listener = &trap_item->listeners_arr[i]; + if (!mlxsw_sp_trap_listener_is_valid(listener)) + continue; + err = mlxsw_core_trap_register(mlxsw_core, listener, trap_ctx); + if (err) + return err; + } + + return 0; +} + +void mlxsw_sp_trap_fini(struct mlxsw_core *mlxsw_core, + const struct devlink_trap *trap, void *trap_ctx) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + const struct mlxsw_sp_trap_item *trap_item; + int i; + + trap_item = mlxsw_sp_trap_item_lookup(mlxsw_sp, trap->id); + if (WARN_ON(!trap_item)) + return; + + for (i = MLXSW_SP_TRAP_LISTENERS_MAX - 1; i >= 0; i--) { + const struct mlxsw_listener *listener; + + listener = &trap_item->listeners_arr[i]; + if (!mlxsw_sp_trap_listener_is_valid(listener)) + continue; + mlxsw_core_trap_unregister(mlxsw_core, listener, trap_ctx); + } +} + +int mlxsw_sp_trap_action_set(struct mlxsw_core *mlxsw_core, + const struct devlink_trap *trap, + enum devlink_trap_action action, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + const struct mlxsw_sp_trap_item *trap_item; + int i; + + trap_item = mlxsw_sp_trap_item_lookup(mlxsw_sp, trap->id); + if (WARN_ON(!trap_item)) + return -EINVAL; + + if (trap_item->is_source) { + NL_SET_ERR_MSG_MOD(extack, "Changing the action of source traps is not supported"); + return -EOPNOTSUPP; + } + + for (i = 0; i < MLXSW_SP_TRAP_LISTENERS_MAX; i++) { + const struct mlxsw_listener *listener; + bool enabled; + int err; + + listener = &trap_item->listeners_arr[i]; + if (!mlxsw_sp_trap_listener_is_valid(listener)) + continue; + + switch (action) { + case DEVLINK_TRAP_ACTION_DROP: + enabled = false; + break; + case DEVLINK_TRAP_ACTION_TRAP: + enabled = true; + break; + default: + return -EINVAL; + } + err = mlxsw_core_trap_state_set(mlxsw_core, listener, enabled); + if (err) + return err; + } + + return 0; +} + +static int +__mlxsw_sp_trap_group_init(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_group *group, + u32 policer_id, struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + u16 hw_policer_id = MLXSW_REG_HTGT_INVALID_POLICER; + const struct mlxsw_sp_trap_group_item *group_item; + char htgt_pl[MLXSW_REG_HTGT_LEN]; + + group_item = mlxsw_sp_trap_group_item_lookup(mlxsw_sp, group->id); + if (WARN_ON(!group_item)) + return -EINVAL; + + if (group_item->fixed_policer && policer_id != group->init_policer_id) { + NL_SET_ERR_MSG_MOD(extack, "Changing the policer binding of this group is not supported"); + return -EOPNOTSUPP; + } + + if (policer_id) { + struct mlxsw_sp_trap_policer_item *policer_item; + + policer_item = mlxsw_sp_trap_policer_item_lookup(mlxsw_sp, + policer_id); + if (WARN_ON(!policer_item)) + return -EINVAL; + hw_policer_id = policer_item->hw_id; + } + + mlxsw_reg_htgt_pack(htgt_pl, group_item->hw_group_id, hw_policer_id, + group_item->priority, group_item->priority); + return mlxsw_reg_write(mlxsw_core, MLXSW_REG(htgt), htgt_pl); +} + +int mlxsw_sp_trap_group_init(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_group *group) +{ + return __mlxsw_sp_trap_group_init(mlxsw_core, group, + group->init_policer_id, NULL); +} + +int mlxsw_sp_trap_group_set(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_group *group, + const struct devlink_trap_policer *policer, + struct netlink_ext_ack *extack) +{ + u32 policer_id = policer ? policer->id : 0; + + return __mlxsw_sp_trap_group_init(mlxsw_core, group, policer_id, + extack); +} + +static int +mlxsw_sp_trap_policer_item_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_trap_policer_item *policer_item) +{ + struct mlxsw_sp_trap *trap = mlxsw_sp->trap; + u16 hw_id; + + /* We should be able to allocate a policer because the number of + * policers we registered with devlink is in according with the number + * of available policers. + */ + hw_id = find_first_zero_bit(trap->policers_usage, trap->max_policers); + if (WARN_ON(hw_id == trap->max_policers)) + return -ENOBUFS; + + __set_bit(hw_id, trap->policers_usage); + policer_item->hw_id = hw_id; + + return 0; +} + +static void +mlxsw_sp_trap_policer_item_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_trap_policer_item *policer_item) +{ + __clear_bit(policer_item->hw_id, mlxsw_sp->trap->policers_usage); +} + +static int mlxsw_sp_trap_policer_bs(u64 burst, u8 *p_burst_size, + struct netlink_ext_ack *extack) +{ + int bs = fls64(burst) - 1; + + if (burst != (BIT_ULL(bs))) { + NL_SET_ERR_MSG_MOD(extack, "Policer burst size is not power of two"); + return -EINVAL; + } + + *p_burst_size = bs; + + return 0; +} + +static int __mlxsw_sp_trap_policer_set(struct mlxsw_sp *mlxsw_sp, u16 hw_id, + u64 rate, u64 burst, bool clear_counter, + struct netlink_ext_ack *extack) +{ + char qpcr_pl[MLXSW_REG_QPCR_LEN]; + u8 burst_size; + int err; + + err = mlxsw_sp_trap_policer_bs(burst, &burst_size, extack); + if (err) + return err; + + mlxsw_reg_qpcr_pack(qpcr_pl, hw_id, MLXSW_REG_QPCR_IR_UNITS_M, false, + rate, burst_size); + mlxsw_reg_qpcr_clear_counter_set(qpcr_pl, clear_counter); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qpcr), qpcr_pl); +} + +int mlxsw_sp_trap_policer_init(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_policer *policer) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + struct mlxsw_sp_trap_policer_item *policer_item; + int err; + + policer_item = mlxsw_sp_trap_policer_item_lookup(mlxsw_sp, policer->id); + if (WARN_ON(!policer_item)) + return -EINVAL; + + err = mlxsw_sp_trap_policer_item_init(mlxsw_sp, policer_item); + if (err) + return err; + + err = __mlxsw_sp_trap_policer_set(mlxsw_sp, policer_item->hw_id, + policer->init_rate, + policer->init_burst, true, NULL); + if (err) + goto err_trap_policer_set; + + return 0; + +err_trap_policer_set: + mlxsw_sp_trap_policer_item_fini(mlxsw_sp, policer_item); + return err; +} + +void mlxsw_sp_trap_policer_fini(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_policer *policer) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + struct mlxsw_sp_trap_policer_item *policer_item; + + policer_item = mlxsw_sp_trap_policer_item_lookup(mlxsw_sp, policer->id); + if (WARN_ON(!policer_item)) + return; + + mlxsw_sp_trap_policer_item_fini(mlxsw_sp, policer_item); +} + +int mlxsw_sp_trap_policer_set(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_policer *policer, + u64 rate, u64 burst, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + struct mlxsw_sp_trap_policer_item *policer_item; + + policer_item = mlxsw_sp_trap_policer_item_lookup(mlxsw_sp, policer->id); + if (WARN_ON(!policer_item)) + return -EINVAL; + + return __mlxsw_sp_trap_policer_set(mlxsw_sp, policer_item->hw_id, + rate, burst, false, extack); +} + +int +mlxsw_sp_trap_policer_counter_get(struct mlxsw_core *mlxsw_core, + const struct devlink_trap_policer *policer, + u64 *p_drops) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + struct mlxsw_sp_trap_policer_item *policer_item; + char qpcr_pl[MLXSW_REG_QPCR_LEN]; + int err; + + policer_item = mlxsw_sp_trap_policer_item_lookup(mlxsw_sp, policer->id); + if (WARN_ON(!policer_item)) + return -EINVAL; + + mlxsw_reg_qpcr_pack(qpcr_pl, policer_item->hw_id, + MLXSW_REG_QPCR_IR_UNITS_M, false, 0, 0); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(qpcr), qpcr_pl); + if (err) + return err; + + *p_drops = mlxsw_reg_qpcr_violate_count_get(qpcr_pl); + + return 0; +} + +int mlxsw_sp_trap_group_policer_hw_id_get(struct mlxsw_sp *mlxsw_sp, u16 id, + bool *p_enabled, u16 *p_hw_id) +{ + struct mlxsw_sp_trap_policer_item *pol_item; + struct mlxsw_sp_trap_group_item *gr_item; + u32 pol_id; + + gr_item = mlxsw_sp_trap_group_item_lookup(mlxsw_sp, id); + if (!gr_item) + return -ENOENT; + + pol_id = gr_item->group.init_policer_id; + if (!pol_id) { + *p_enabled = false; + return 0; + } + + pol_item = mlxsw_sp_trap_policer_item_lookup(mlxsw_sp, pol_id); + if (WARN_ON(!pol_item)) + return -ENOENT; + + *p_enabled = true; + *p_hw_id = pol_item->hw_id; + return 0; +} + +static const struct mlxsw_sp_trap_group_item +mlxsw_sp1_trap_group_items_arr[] = { + { + .group = DEVLINK_TRAP_GROUP_GENERIC(ACL_SAMPLE, 0), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_PKT_SAMPLE, + .priority = 0, + }, +}; + +static const struct mlxsw_sp_trap_item +mlxsw_sp1_trap_items_arr[] = { + { + .trap = MLXSW_SP_TRAP_CONTROL(FLOW_ACTION_SAMPLE, ACL_SAMPLE, + MIRROR), + .listeners_arr = { + MLXSW_RXL(mlxsw_sp_rx_sample_listener, PKT_SAMPLE, + MIRROR_TO_CPU, false, SP_PKT_SAMPLE, DISCARD), + }, + }, +}; + +static int +mlxsw_sp1_trap_groups_init(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_trap_group_item **arr, + size_t *p_groups_count) +{ + *arr = mlxsw_sp1_trap_group_items_arr; + *p_groups_count = ARRAY_SIZE(mlxsw_sp1_trap_group_items_arr); + + return 0; +} + +static int mlxsw_sp1_traps_init(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_trap_item **arr, + size_t *p_traps_count) +{ + *arr = mlxsw_sp1_trap_items_arr; + *p_traps_count = ARRAY_SIZE(mlxsw_sp1_trap_items_arr); + + return 0; +} + +const struct mlxsw_sp_trap_ops mlxsw_sp1_trap_ops = { + .groups_init = mlxsw_sp1_trap_groups_init, + .traps_init = mlxsw_sp1_traps_init, +}; + +static const struct mlxsw_sp_trap_group_item +mlxsw_sp2_trap_group_items_arr[] = { + { + .group = DEVLINK_TRAP_GROUP_GENERIC(BUFFER_DROPS, 20), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_BUFFER_DISCARDS, + .priority = 0, + .fixed_policer = true, + }, + { + .group = DEVLINK_TRAP_GROUP_GENERIC(ACL_SAMPLE, 0), + .hw_group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_PKT_SAMPLE, + .priority = 0, + .fixed_policer = true, + }, +}; + +static const struct mlxsw_sp_trap_item +mlxsw_sp2_trap_items_arr[] = { + { + .trap = MLXSW_SP_TRAP_BUFFER_DROP(EARLY_DROP), + .listeners_arr = { + MLXSW_SP_RXL_BUFFER_DISCARD(INGRESS_WRED), + }, + .is_source = true, + }, + { + .trap = MLXSW_SP_TRAP_CONTROL(FLOW_ACTION_SAMPLE, ACL_SAMPLE, + MIRROR), + .listeners_arr = { + MLXSW_RXL_MIRROR(mlxsw_sp_rx_sample_listener, 1, + SP_PKT_SAMPLE, + MLXSW_SP_MIRROR_REASON_INGRESS), + MLXSW_RXL_MIRROR(mlxsw_sp_rx_sample_tx_listener, 1, + SP_PKT_SAMPLE, + MLXSW_SP_MIRROR_REASON_EGRESS), + MLXSW_RXL_MIRROR(mlxsw_sp_rx_sample_acl_listener, 1, + SP_PKT_SAMPLE, + MLXSW_SP_MIRROR_REASON_POLICY_ENGINE), + }, + }, +}; + +static int +mlxsw_sp2_trap_groups_init(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_trap_group_item **arr, + size_t *p_groups_count) +{ + *arr = mlxsw_sp2_trap_group_items_arr; + *p_groups_count = ARRAY_SIZE(mlxsw_sp2_trap_group_items_arr); + + return 0; +} + +static int mlxsw_sp2_traps_init(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_trap_item **arr, + size_t *p_traps_count) +{ + *arr = mlxsw_sp2_trap_items_arr; + *p_traps_count = ARRAY_SIZE(mlxsw_sp2_trap_items_arr); + + return 0; +} + +const struct mlxsw_sp_trap_ops mlxsw_sp2_trap_ops = { + .groups_init = mlxsw_sp2_trap_groups_init, + .traps_init = mlxsw_sp2_traps_init, +}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.h new file mode 100644 index 000000000..b8df684be --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.h @@ -0,0 +1,38 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2020 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_SPECTRUM_TRAP_H +#define _MLXSW_SPECTRUM_TRAP_H + +#include <linux/list.h> +#include <net/devlink.h> + +struct mlxsw_sp_trap { + struct mlxsw_sp_trap_policer_item *policer_items_arr; + size_t policers_count; /* Number of registered policers */ + + struct mlxsw_sp_trap_group_item *group_items_arr; + size_t groups_count; /* Number of registered groups */ + + struct mlxsw_sp_trap_item *trap_items_arr; + size_t traps_count; /* Number of registered traps */ + + u16 thin_policer_hw_id; + + u64 max_policers; + unsigned long policers_usage[]; /* Usage bitmap */ +}; + +struct mlxsw_sp_trap_ops { + int (*groups_init)(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_trap_group_item **arr, + size_t *p_groups_count); + int (*traps_init)(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_trap_item **arr, + size_t *p_traps_count); +}; + +extern const struct mlxsw_sp_trap_ops mlxsw_sp1_trap_ops; +extern const struct mlxsw_sp_trap_ops mlxsw_sp2_trap_ops; + +#endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/trap.h b/drivers/net/ethernet/mellanox/mlxsw/trap.h new file mode 100644 index 000000000..8da169663 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/trap.h @@ -0,0 +1,144 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_TRAP_H +#define _MLXSW_TRAP_H + +enum { + /* Ethernet EMAD and FDB miss */ + MLXSW_TRAP_ID_FDB_MC = 0x01, + MLXSW_TRAP_ID_ETHEMAD = 0x05, + /* L2 traps for specific packet types */ + MLXSW_TRAP_ID_STP = 0x10, + MLXSW_TRAP_ID_LACP = 0x11, + MLXSW_TRAP_ID_EAPOL = 0x12, + MLXSW_TRAP_ID_LLDP = 0x13, + MLXSW_TRAP_ID_MMRP = 0x14, + MLXSW_TRAP_ID_MVRP = 0x15, + MLXSW_TRAP_ID_RPVST = 0x16, + MLXSW_TRAP_ID_DHCP = 0x19, + MLXSW_TRAP_ID_PTP0 = 0x28, + MLXSW_TRAP_ID_PTP1 = 0x29, + MLXSW_TRAP_ID_IGMP_QUERY = 0x30, + MLXSW_TRAP_ID_IGMP_V1_REPORT = 0x31, + MLXSW_TRAP_ID_IGMP_V2_REPORT = 0x32, + MLXSW_TRAP_ID_IGMP_V2_LEAVE = 0x33, + MLXSW_TRAP_ID_IGMP_V3_REPORT = 0x34, + MLXSW_TRAP_ID_PKT_SAMPLE = 0x38, + MLXSW_TRAP_ID_FID_MISS = 0x3D, + MLXSW_TRAP_ID_DECAP_ECN0 = 0x40, + MLXSW_TRAP_ID_MTUERROR = 0x52, + MLXSW_TRAP_ID_TTLERROR = 0x53, + MLXSW_TRAP_ID_LBERROR = 0x54, + MLXSW_TRAP_ID_IPV4_OSPF = 0x55, + MLXSW_TRAP_ID_IPV4_PIM = 0x58, + MLXSW_TRAP_ID_IPV4_VRRP = 0x59, + MLXSW_TRAP_ID_RPF = 0x5C, + MLXSW_TRAP_ID_IP2ME = 0x5F, + MLXSW_TRAP_ID_IPV6_UNSPECIFIED_ADDRESS = 0x60, + MLXSW_TRAP_ID_IPV6_LINK_LOCAL_DEST = 0x61, + MLXSW_TRAP_ID_IPV6_LINK_LOCAL_SRC = 0x62, + MLXSW_TRAP_ID_IPV6_ALL_NODES_LINK = 0x63, + MLXSW_TRAP_ID_IPV6_OSPF = 0x64, + MLXSW_TRAP_ID_IPV6_MLDV12_LISTENER_QUERY = 0x65, + MLXSW_TRAP_ID_IPV6_MLDV1_LISTENER_REPORT = 0x66, + MLXSW_TRAP_ID_IPV6_MLDV1_LISTENER_DONE = 0x67, + MLXSW_TRAP_ID_IPV6_MLDV2_LISTENER_REPORT = 0x68, + MLXSW_TRAP_ID_IPV6_DHCP = 0x69, + MLXSW_TRAP_ID_IPV6_ALL_ROUTERS_LINK = 0x6F, + MLXSW_TRAP_ID_RTR_INGRESS0 = 0x70, + MLXSW_TRAP_ID_RTR_INGRESS1 = 0x71, + MLXSW_TRAP_ID_IPV6_PIM = 0x79, + MLXSW_TRAP_ID_IPV6_VRRP = 0x7A, + MLXSW_TRAP_ID_RTR_EGRESS0 = 0x80, + MLXSW_TRAP_ID_IPV4_BGP = 0x88, + MLXSW_TRAP_ID_IPV6_BGP = 0x89, + MLXSW_TRAP_ID_L3_IPV6_ROUTER_SOLICITATION = 0x8A, + MLXSW_TRAP_ID_L3_IPV6_ROUTER_ADVERTISEMENT = 0x8B, + MLXSW_TRAP_ID_L3_IPV6_NEIGHBOR_SOLICITATION = 0x8C, + MLXSW_TRAP_ID_L3_IPV6_NEIGHBOR_ADVERTISEMENT = 0x8D, + MLXSW_TRAP_ID_L3_IPV6_REDIRECTION = 0x8E, + MLXSW_TRAP_ID_IPV4_DHCP = 0x8F, + MLXSW_TRAP_ID_HOST_MISS_IPV4 = 0x90, + MLXSW_TRAP_ID_IPV6_MC_LINK_LOCAL_DEST = 0x91, + MLXSW_TRAP_ID_HOST_MISS_IPV6 = 0x92, + MLXSW_TRAP_ID_IPIP_DECAP_ERROR = 0xB1, + MLXSW_TRAP_ID_NVE_DECAP_ARP = 0xB8, + MLXSW_TRAP_ID_NVE_ENCAP_ARP = 0xBD, + MLXSW_TRAP_ID_IPV4_BFD = 0xD0, + MLXSW_TRAP_ID_IPV6_BFD = 0xD1, + MLXSW_TRAP_ID_ROUTER_ALERT_IPV4 = 0xD6, + MLXSW_TRAP_ID_ROUTER_ALERT_IPV6 = 0xD7, + MLXSW_TRAP_ID_ROUTER_ARPBC = 0xE0, + MLXSW_TRAP_ID_ROUTER_ARPUC = 0xE1, + MLXSW_TRAP_ID_DISCARD_NON_ROUTABLE = 0x11A, + MLXSW_TRAP_ID_DISCARD_ROUTER2 = 0x130, + MLXSW_TRAP_ID_DISCARD_ROUTER3 = 0x131, + MLXSW_TRAP_ID_DISCARD_ING_PACKET_SMAC_MC = 0x140, + MLXSW_TRAP_ID_DISCARD_ING_SWITCH_VTAG_ALLOW = 0x148, + MLXSW_TRAP_ID_DISCARD_ING_SWITCH_VLAN = 0x149, + MLXSW_TRAP_ID_DISCARD_ING_SWITCH_STP = 0x14A, + MLXSW_TRAP_ID_DISCARD_LOOKUP_SWITCH_UC = 0x150, + MLXSW_TRAP_ID_DISCARD_LOOKUP_SWITCH_MC_NULL = 0x151, + MLXSW_TRAP_ID_DISCARD_LOOKUP_SWITCH_LB = 0x152, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_NON_IP_PACKET = 0x160, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_UC_DIP_MC_DMAC = 0x161, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_DIP_LB = 0x162, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_SIP_MC = 0x163, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_SIP_CLASS_E = 0x164, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_SIP_LB = 0x165, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_CORRUPTED_IP_HDR = 0x167, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_MC_DMAC = 0x168, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_SIP_DIP = 0x169, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_IPV4_SIP_BC = 0x16A, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_IPV4_DIP_LOCAL_NET = 0x16B, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_DIP_LINK_LOCAL = 0x16C, + MLXSW_TRAP_ID_DISCARD_ROUTER_IRIF_EN = 0x178, + MLXSW_TRAP_ID_DISCARD_ROUTER_ERIF_EN = 0x179, + MLXSW_TRAP_ID_DISCARD_ROUTER_LPM4 = 0x17B, + MLXSW_TRAP_ID_DISCARD_ROUTER_LPM6 = 0x17C, + MLXSW_TRAP_ID_DISCARD_DEC_PKT = 0x188, + MLXSW_TRAP_ID_DISCARD_OVERLAY_SMAC_MC = 0x190, + MLXSW_TRAP_ID_DISCARD_IPV6_MC_DIP_RESERVED_SCOPE = 0x1B0, + MLXSW_TRAP_ID_DISCARD_IPV6_MC_DIP_INTERFACE_LOCAL_SCOPE = 0x1B1, + MLXSW_TRAP_ID_ACL0 = 0x1C0, + /* Multicast trap used for routes with trap action */ + MLXSW_TRAP_ID_ACL1 = 0x1C1, + /* Multicast trap used for routes with trap-and-forward action */ + MLXSW_TRAP_ID_ACL2 = 0x1C2, + MLXSW_TRAP_ID_DISCARD_INGRESS_ACL = 0x1C3, + MLXSW_TRAP_ID_DISCARD_EGRESS_ACL = 0x1C4, + MLXSW_TRAP_ID_MIRROR_SESSION0 = 0x220, + MLXSW_TRAP_ID_MIRROR_SESSION1 = 0x221, + MLXSW_TRAP_ID_MIRROR_SESSION2 = 0x222, + MLXSW_TRAP_ID_MIRROR_SESSION3 = 0x223, + MLXSW_TRAP_ID_MIRROR_SESSION4 = 0x224, + MLXSW_TRAP_ID_MIRROR_SESSION5 = 0x225, + MLXSW_TRAP_ID_MIRROR_SESSION6 = 0x226, + MLXSW_TRAP_ID_MIRROR_SESSION7 = 0x227, + + MLXSW_TRAP_ID_MAX = 0x3FF, +}; + +enum mlxsw_event_trap_id { + /* Fatal Event generated by FW */ + MLXSW_TRAP_ID_MFDE = 0x3, + /* Port Up/Down event generated by hardware */ + MLXSW_TRAP_ID_PUDE = 0x8, + /* Port Module Plug/Unplug Event generated by hardware */ + MLXSW_TRAP_ID_PMPE = 0x9, + /* Temperature Warning event generated by hardware */ + MLXSW_TRAP_ID_MTWE = 0xC, + /* PTP Ingress FIFO has a new entry */ + MLXSW_TRAP_ID_PTP_ING_FIFO = 0x2D, + /* PTP Egress FIFO has a new entry */ + MLXSW_TRAP_ID_PTP_EGR_FIFO = 0x2E, + /* Downstream Device Status Change */ + MLXSW_TRAP_ID_DSDSC = 0x321, + /* Binary Code Transfer Operation Executed Event */ + MLXSW_TRAP_ID_BCTOE = 0x322, + /* Port mapping change */ + MLXSW_TRAP_ID_PMLPE = 0x32E, +}; + +#endif /* _MLXSW_TRAP_H */ diff --git a/drivers/net/ethernet/mellanox/mlxsw/txheader.h b/drivers/net/ethernet/mellanox/mlxsw/txheader.h new file mode 100644 index 000000000..da51dd9d5 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlxsw/txheader.h @@ -0,0 +1,49 @@ +/* SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 */ +/* Copyright (c) 2015-2018 Mellanox Technologies. All rights reserved */ + +#ifndef _MLXSW_TXHEADER_H +#define _MLXSW_TXHEADER_H + +#define MLXSW_TXHDR_LEN 0x10 +#define MLXSW_TXHDR_VERSION_0 0 +#define MLXSW_TXHDR_VERSION_1 1 + +enum { + MLXSW_TXHDR_ETH_CTL, + MLXSW_TXHDR_ETH_DATA, +}; + +#define MLXSW_TXHDR_PROTO_ETH 1 + +enum { + MLXSW_TXHDR_ETCLASS_0, + MLXSW_TXHDR_ETCLASS_1, + MLXSW_TXHDR_ETCLASS_2, + MLXSW_TXHDR_ETCLASS_3, + MLXSW_TXHDR_ETCLASS_4, + MLXSW_TXHDR_ETCLASS_5, + MLXSW_TXHDR_ETCLASS_6, + MLXSW_TXHDR_ETCLASS_7, +}; + +enum { + MLXSW_TXHDR_RDQ_OTHER, + MLXSW_TXHDR_RDQ_EMAD = 0x1f, +}; + +#define MLXSW_TXHDR_CTCLASS3 0 +#define MLXSW_TXHDR_CPU_SIG 0 +#define MLXSW_TXHDR_SIG 0xE0E0 +#define MLXSW_TXHDR_STCLASS_NONE 0 + +enum { + MLXSW_TXHDR_NOT_EMAD, + MLXSW_TXHDR_EMAD, +}; + +enum { + MLXSW_TXHDR_TYPE_DATA, + MLXSW_TXHDR_TYPE_CONTROL = 6, +}; + +#endif |