diff options
Diffstat (limited to 'drivers/net/ethernet/cisco/enic')
37 files changed, 10137 insertions, 0 deletions
diff --git a/drivers/net/ethernet/cisco/enic/Kconfig b/drivers/net/ethernet/cisco/enic/Kconfig new file mode 100644 index 000000000..b63f8d8a4 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/Kconfig @@ -0,0 +1,9 @@ +# +# Cisco device configuration +# + +config ENIC + tristate "Cisco VIC Ethernet NIC Support" + depends on PCI + ---help--- + This enables the support for the Cisco VIC Ethernet card. diff --git a/drivers/net/ethernet/cisco/enic/Makefile b/drivers/net/ethernet/cisco/enic/Makefile new file mode 100644 index 000000000..aadcaf787 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/Makefile @@ -0,0 +1,6 @@ +obj-$(CONFIG_ENIC) := enic.o + +enic-y := enic_main.o vnic_cq.o vnic_intr.o vnic_wq.o \ + enic_res.o enic_dev.o enic_pp.o vnic_dev.o vnic_rq.o vnic_vic.o \ + enic_ethtool.o enic_api.o enic_clsf.o + diff --git a/drivers/net/ethernet/cisco/enic/cq_desc.h b/drivers/net/ethernet/cisco/enic/cq_desc.h new file mode 100644 index 000000000..d6dd1b4ed --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/cq_desc.h @@ -0,0 +1,80 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _CQ_DESC_H_ +#define _CQ_DESC_H_ + +/* + * Completion queue descriptor types + */ +enum cq_desc_types { + CQ_DESC_TYPE_WQ_ENET = 0, + CQ_DESC_TYPE_DESC_COPY = 1, + CQ_DESC_TYPE_WQ_EXCH = 2, + CQ_DESC_TYPE_RQ_ENET = 3, + CQ_DESC_TYPE_RQ_FCP = 4, +}; + +/* Completion queue descriptor: 16B + * + * All completion queues have this basic layout. The + * type_specfic area is unique for each completion + * queue type. + */ +struct cq_desc { + __le16 completed_index; + __le16 q_number; + u8 type_specfic[11]; + u8 type_color; +}; + +#define CQ_DESC_TYPE_BITS 4 +#define CQ_DESC_TYPE_MASK ((1 << CQ_DESC_TYPE_BITS) - 1) +#define CQ_DESC_COLOR_MASK 1 +#define CQ_DESC_COLOR_SHIFT 7 +#define CQ_DESC_Q_NUM_BITS 10 +#define CQ_DESC_Q_NUM_MASK ((1 << CQ_DESC_Q_NUM_BITS) - 1) +#define CQ_DESC_COMP_NDX_BITS 12 +#define CQ_DESC_COMP_NDX_MASK ((1 << CQ_DESC_COMP_NDX_BITS) - 1) + +static inline void cq_desc_dec(const struct cq_desc *desc_arg, + u8 *type, u8 *color, u16 *q_number, u16 *completed_index) +{ + const struct cq_desc *desc = desc_arg; + const u8 type_color = desc->type_color; + + *color = (type_color >> CQ_DESC_COLOR_SHIFT) & CQ_DESC_COLOR_MASK; + + /* + * Make sure color bit is read from desc *before* other fields + * are read from desc. Hardware guarantees color bit is last + * bit (byte) written. Adding the rmb() prevents the compiler + * and/or CPU from reordering the reads which would potentially + * result in reading stale values. + */ + + rmb(); + + *type = type_color & CQ_DESC_TYPE_MASK; + *q_number = le16_to_cpu(desc->q_number) & CQ_DESC_Q_NUM_MASK; + *completed_index = le16_to_cpu(desc->completed_index) & + CQ_DESC_COMP_NDX_MASK; +} + +#endif /* _CQ_DESC_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/cq_enet_desc.h b/drivers/net/ethernet/cisco/enic/cq_enet_desc.h new file mode 100644 index 000000000..ac37cacc6 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/cq_enet_desc.h @@ -0,0 +1,185 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _CQ_ENET_DESC_H_ +#define _CQ_ENET_DESC_H_ + +#include "cq_desc.h" + +/* Ethernet completion queue descriptor: 16B */ +struct cq_enet_wq_desc { + __le16 completed_index; + __le16 q_number; + u8 reserved[11]; + u8 type_color; +}; + +static inline void cq_enet_wq_desc_dec(struct cq_enet_wq_desc *desc, + u8 *type, u8 *color, u16 *q_number, u16 *completed_index) +{ + cq_desc_dec((struct cq_desc *)desc, type, + color, q_number, completed_index); +} + +/* Completion queue descriptor: Ethernet receive queue, 16B */ +struct cq_enet_rq_desc { + __le16 completed_index_flags; + __le16 q_number_rss_type_flags; + __le32 rss_hash; + __le16 bytes_written_flags; + __le16 vlan; + __le16 checksum_fcoe; + u8 flags; + u8 type_color; +}; + +#define CQ_ENET_RQ_DESC_FLAGS_INGRESS_PORT (0x1 << 12) +#define CQ_ENET_RQ_DESC_FLAGS_FCOE (0x1 << 13) +#define CQ_ENET_RQ_DESC_FLAGS_EOP (0x1 << 14) +#define CQ_ENET_RQ_DESC_FLAGS_SOP (0x1 << 15) + +#define CQ_ENET_RQ_DESC_RSS_TYPE_BITS 4 +#define CQ_ENET_RQ_DESC_RSS_TYPE_MASK \ + ((1 << CQ_ENET_RQ_DESC_RSS_TYPE_BITS) - 1) +#define CQ_ENET_RQ_DESC_RSS_TYPE_NONE 0 +#define CQ_ENET_RQ_DESC_RSS_TYPE_IPv4 1 +#define CQ_ENET_RQ_DESC_RSS_TYPE_TCP_IPv4 2 +#define CQ_ENET_RQ_DESC_RSS_TYPE_IPv6 3 +#define CQ_ENET_RQ_DESC_RSS_TYPE_TCP_IPv6 4 +#define CQ_ENET_RQ_DESC_RSS_TYPE_IPv6_EX 5 +#define CQ_ENET_RQ_DESC_RSS_TYPE_TCP_IPv6_EX 6 + +#define CQ_ENET_RQ_DESC_FLAGS_CSUM_NOT_CALC (0x1 << 14) + +#define CQ_ENET_RQ_DESC_BYTES_WRITTEN_BITS 14 +#define CQ_ENET_RQ_DESC_BYTES_WRITTEN_MASK \ + ((1 << CQ_ENET_RQ_DESC_BYTES_WRITTEN_BITS) - 1) +#define CQ_ENET_RQ_DESC_FLAGS_TRUNCATED (0x1 << 14) +#define CQ_ENET_RQ_DESC_FLAGS_VLAN_STRIPPED (0x1 << 15) + +#define CQ_ENET_RQ_DESC_VLAN_TCI_VLAN_BITS 12 +#define CQ_ENET_RQ_DESC_VLAN_TCI_VLAN_MASK \ + ((1 << CQ_ENET_RQ_DESC_VLAN_TCI_VLAN_BITS) - 1) +#define CQ_ENET_RQ_DESC_VLAN_TCI_CFI_MASK (0x1 << 12) +#define CQ_ENET_RQ_DESC_VLAN_TCI_USER_PRIO_BITS 3 +#define CQ_ENET_RQ_DESC_VLAN_TCI_USER_PRIO_MASK \ + ((1 << CQ_ENET_RQ_DESC_VLAN_TCI_USER_PRIO_BITS) - 1) +#define CQ_ENET_RQ_DESC_VLAN_TCI_USER_PRIO_SHIFT 13 + +#define CQ_ENET_RQ_DESC_FCOE_SOF_BITS 8 +#define CQ_ENET_RQ_DESC_FCOE_SOF_MASK \ + ((1 << CQ_ENET_RQ_DESC_FCOE_SOF_BITS) - 1) +#define CQ_ENET_RQ_DESC_FCOE_EOF_BITS 8 +#define CQ_ENET_RQ_DESC_FCOE_EOF_MASK \ + ((1 << CQ_ENET_RQ_DESC_FCOE_EOF_BITS) - 1) +#define CQ_ENET_RQ_DESC_FCOE_EOF_SHIFT 8 + +#define CQ_ENET_RQ_DESC_FLAGS_TCP_UDP_CSUM_OK (0x1 << 0) +#define CQ_ENET_RQ_DESC_FCOE_FC_CRC_OK (0x1 << 0) +#define CQ_ENET_RQ_DESC_FLAGS_UDP (0x1 << 1) +#define CQ_ENET_RQ_DESC_FCOE_ENC_ERROR (0x1 << 1) +#define CQ_ENET_RQ_DESC_FLAGS_TCP (0x1 << 2) +#define CQ_ENET_RQ_DESC_FLAGS_IPV4_CSUM_OK (0x1 << 3) +#define CQ_ENET_RQ_DESC_FLAGS_IPV6 (0x1 << 4) +#define CQ_ENET_RQ_DESC_FLAGS_IPV4 (0x1 << 5) +#define CQ_ENET_RQ_DESC_FLAGS_IPV4_FRAGMENT (0x1 << 6) +#define CQ_ENET_RQ_DESC_FLAGS_FCS_OK (0x1 << 7) + +static inline void cq_enet_rq_desc_dec(struct cq_enet_rq_desc *desc, + u8 *type, u8 *color, u16 *q_number, u16 *completed_index, + u8 *ingress_port, u8 *fcoe, u8 *eop, u8 *sop, u8 *rss_type, + u8 *csum_not_calc, u32 *rss_hash, u16 *bytes_written, u8 *packet_error, + u8 *vlan_stripped, u16 *vlan_tci, u16 *checksum, u8 *fcoe_sof, + u8 *fcoe_fc_crc_ok, u8 *fcoe_enc_error, u8 *fcoe_eof, + u8 *tcp_udp_csum_ok, u8 *udp, u8 *tcp, u8 *ipv4_csum_ok, + u8 *ipv6, u8 *ipv4, u8 *ipv4_fragment, u8 *fcs_ok) +{ + u16 completed_index_flags; + u16 q_number_rss_type_flags; + u16 bytes_written_flags; + + cq_desc_dec((struct cq_desc *)desc, type, + color, q_number, completed_index); + + completed_index_flags = le16_to_cpu(desc->completed_index_flags); + q_number_rss_type_flags = + le16_to_cpu(desc->q_number_rss_type_flags); + bytes_written_flags = le16_to_cpu(desc->bytes_written_flags); + + *ingress_port = (completed_index_flags & + CQ_ENET_RQ_DESC_FLAGS_INGRESS_PORT) ? 1 : 0; + *fcoe = (completed_index_flags & CQ_ENET_RQ_DESC_FLAGS_FCOE) ? + 1 : 0; + *eop = (completed_index_flags & CQ_ENET_RQ_DESC_FLAGS_EOP) ? + 1 : 0; + *sop = (completed_index_flags & CQ_ENET_RQ_DESC_FLAGS_SOP) ? + 1 : 0; + + *rss_type = (u8)((q_number_rss_type_flags >> CQ_DESC_Q_NUM_BITS) & + CQ_ENET_RQ_DESC_RSS_TYPE_MASK); + *csum_not_calc = (q_number_rss_type_flags & + CQ_ENET_RQ_DESC_FLAGS_CSUM_NOT_CALC) ? 1 : 0; + + *rss_hash = le32_to_cpu(desc->rss_hash); + + *bytes_written = bytes_written_flags & + CQ_ENET_RQ_DESC_BYTES_WRITTEN_MASK; + *packet_error = (bytes_written_flags & + CQ_ENET_RQ_DESC_FLAGS_TRUNCATED) ? 1 : 0; + *vlan_stripped = (bytes_written_flags & + CQ_ENET_RQ_DESC_FLAGS_VLAN_STRIPPED) ? 1 : 0; + + /* + * Tag Control Information(16) = user_priority(3) + cfi(1) + vlan(12) + */ + *vlan_tci = le16_to_cpu(desc->vlan); + + if (*fcoe) { + *fcoe_sof = (u8)(le16_to_cpu(desc->checksum_fcoe) & + CQ_ENET_RQ_DESC_FCOE_SOF_MASK); + *fcoe_fc_crc_ok = (desc->flags & + CQ_ENET_RQ_DESC_FCOE_FC_CRC_OK) ? 1 : 0; + *fcoe_enc_error = (desc->flags & + CQ_ENET_RQ_DESC_FCOE_ENC_ERROR) ? 1 : 0; + *fcoe_eof = (u8)((le16_to_cpu(desc->checksum_fcoe) >> + CQ_ENET_RQ_DESC_FCOE_EOF_SHIFT) & + CQ_ENET_RQ_DESC_FCOE_EOF_MASK); + *checksum = 0; + } else { + *fcoe_sof = 0; + *fcoe_fc_crc_ok = 0; + *fcoe_enc_error = 0; + *fcoe_eof = 0; + *checksum = le16_to_cpu(desc->checksum_fcoe); + } + + *tcp_udp_csum_ok = + (desc->flags & CQ_ENET_RQ_DESC_FLAGS_TCP_UDP_CSUM_OK) ? 1 : 0; + *udp = (desc->flags & CQ_ENET_RQ_DESC_FLAGS_UDP) ? 1 : 0; + *tcp = (desc->flags & CQ_ENET_RQ_DESC_FLAGS_TCP) ? 1 : 0; + *ipv4_csum_ok = + (desc->flags & CQ_ENET_RQ_DESC_FLAGS_IPV4_CSUM_OK) ? 1 : 0; + *ipv6 = (desc->flags & CQ_ENET_RQ_DESC_FLAGS_IPV6) ? 1 : 0; + *ipv4 = (desc->flags & CQ_ENET_RQ_DESC_FLAGS_IPV4) ? 1 : 0; + *ipv4_fragment = + (desc->flags & CQ_ENET_RQ_DESC_FLAGS_IPV4_FRAGMENT) ? 1 : 0; + *fcs_ok = (desc->flags & CQ_ENET_RQ_DESC_FLAGS_FCS_OK) ? 1 : 0; +} + +#endif /* _CQ_ENET_DESC_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/enic.h b/drivers/net/ethernet/cisco/enic/enic.h new file mode 100644 index 000000000..08cac1bfa --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic.h @@ -0,0 +1,327 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _ENIC_H_ +#define _ENIC_H_ + +#include "vnic_enet.h" +#include "vnic_dev.h" +#include "vnic_wq.h" +#include "vnic_rq.h" +#include "vnic_cq.h" +#include "vnic_intr.h" +#include "vnic_stats.h" +#include "vnic_nic.h" +#include "vnic_rss.h" +#include <linux/irq.h> + +#define DRV_NAME "enic" +#define DRV_DESCRIPTION "Cisco VIC Ethernet NIC Driver" +#define DRV_VERSION "2.3.0.53" +#define DRV_COPYRIGHT "Copyright 2008-2013 Cisco Systems, Inc" + +#define ENIC_BARS_MAX 6 + +#define ENIC_WQ_MAX 8 +#define ENIC_RQ_MAX 8 +#define ENIC_CQ_MAX (ENIC_WQ_MAX + ENIC_RQ_MAX) +#define ENIC_INTR_MAX (ENIC_CQ_MAX + 2) + +#define ENIC_WQ_NAPI_BUDGET 256 + +#define ENIC_AIC_LARGE_PKT_DIFF 3 + +struct enic_msix_entry { + int requested; + char devname[IFNAMSIZ + 8]; + irqreturn_t (*isr)(int, void *); + void *devid; + cpumask_var_t affinity_mask; +}; + +/* Store only the lower range. Higher range is given by fw. */ +struct enic_intr_mod_range { + u32 small_pkt_range_start; + u32 large_pkt_range_start; +}; + +struct enic_intr_mod_table { + u32 rx_rate; + u32 range_percent; +}; + +#define ENIC_MAX_LINK_SPEEDS 3 +#define ENIC_LINK_SPEED_10G 10000 +#define ENIC_LINK_SPEED_4G 4000 +#define ENIC_LINK_40G_INDEX 2 +#define ENIC_LINK_10G_INDEX 1 +#define ENIC_LINK_4G_INDEX 0 +#define ENIC_RX_COALESCE_RANGE_END 125 +#define ENIC_AIC_TS_BREAK 100 + +struct enic_rx_coal { + u32 small_pkt_range_start; + u32 large_pkt_range_start; + u32 range_end; + u32 use_adaptive_rx_coalesce; +}; + +/* priv_flags */ +#define ENIC_SRIOV_ENABLED (1 << 0) + +/* enic port profile set flags */ +#define ENIC_PORT_REQUEST_APPLIED (1 << 0) +#define ENIC_SET_REQUEST (1 << 1) +#define ENIC_SET_NAME (1 << 2) +#define ENIC_SET_INSTANCE (1 << 3) +#define ENIC_SET_HOST (1 << 4) + +struct enic_port_profile { + u32 set; + u8 request; + char name[PORT_PROFILE_MAX]; + u8 instance_uuid[PORT_UUID_MAX]; + u8 host_uuid[PORT_UUID_MAX]; + u8 vf_mac[ETH_ALEN]; + u8 mac_addr[ETH_ALEN]; +}; + +/* enic_rfs_fltr_node - rfs filter node in hash table + * @@keys: IPv4 5 tuple + * @flow_id: flow_id of clsf filter provided by kernel + * @fltr_id: filter id of clsf filter returned by adaptor + * @rq_id: desired rq index + * @node: hlist_node + */ +struct enic_rfs_fltr_node { + struct flow_keys keys; + u32 flow_id; + u16 fltr_id; + u16 rq_id; + struct hlist_node node; +}; + +/* enic_rfs_flw_tbl - rfs flow table + * @max: Maximum number of filters vNIC supports + * @free: Number of free filters available + * @toclean: hash table index to clean next + * @ht_head: hash table list head + * @lock: spin lock + * @rfs_may_expire: timer function for enic_rps_may_expire_flow + */ +struct enic_rfs_flw_tbl { + u16 max; + int free; + +#define ENIC_RFS_FLW_BITSHIFT (10) +#define ENIC_RFS_FLW_MASK ((1 << ENIC_RFS_FLW_BITSHIFT) - 1) + u16 toclean:ENIC_RFS_FLW_BITSHIFT; + struct hlist_head ht_head[1 << ENIC_RFS_FLW_BITSHIFT]; + spinlock_t lock; + struct timer_list rfs_may_expire; +}; + +struct vxlan_offload { + u16 vxlan_udp_port_number; + u8 patch_level; + u8 flags; +}; + +/* Per-instance private data structure */ +struct enic { + struct net_device *netdev; + struct pci_dev *pdev; + struct vnic_enet_config config; + struct vnic_dev_bar bar[ENIC_BARS_MAX]; + struct vnic_dev *vdev; + struct timer_list notify_timer; + struct work_struct reset; + struct work_struct tx_hang_reset; + struct work_struct change_mtu_work; + struct msix_entry msix_entry[ENIC_INTR_MAX]; + struct enic_msix_entry msix[ENIC_INTR_MAX]; + u32 msg_enable; + spinlock_t devcmd_lock; + u8 mac_addr[ETH_ALEN]; + unsigned int flags; + unsigned int priv_flags; + unsigned int mc_count; + unsigned int uc_count; + u32 port_mtu; + struct enic_rx_coal rx_coalesce_setting; + u32 rx_coalesce_usecs; + u32 tx_coalesce_usecs; +#ifdef CONFIG_PCI_IOV + u16 num_vfs; +#endif + spinlock_t enic_api_lock; + bool enic_api_busy; + struct enic_port_profile *pp; + + /* work queue cache line section */ + ____cacheline_aligned struct vnic_wq wq[ENIC_WQ_MAX]; + spinlock_t wq_lock[ENIC_WQ_MAX]; + unsigned int wq_count; + u16 loop_enable; + u16 loop_tag; + + /* receive queue cache line section */ + ____cacheline_aligned struct vnic_rq rq[ENIC_RQ_MAX]; + unsigned int rq_count; + struct vxlan_offload vxlan; + u64 rq_truncated_pkts; + u64 rq_bad_fcs; + struct napi_struct napi[ENIC_RQ_MAX + ENIC_WQ_MAX]; + + /* interrupt resource cache line section */ + ____cacheline_aligned struct vnic_intr intr[ENIC_INTR_MAX]; + unsigned int intr_count; + u32 __iomem *legacy_pba; /* memory-mapped */ + + /* completion queue cache line section */ + ____cacheline_aligned struct vnic_cq cq[ENIC_CQ_MAX]; + unsigned int cq_count; + struct enic_rfs_flw_tbl rfs_h; + u32 rx_copybreak; + u8 rss_key[ENIC_RSS_LEN]; + struct vnic_gen_stats gen_stats; +}; + +static inline struct net_device *vnic_get_netdev(struct vnic_dev *vdev) +{ + struct enic *enic = vdev->priv; + + return enic->netdev; +} + +/* wrappers function for kernel log + */ +#define vdev_err(vdev, fmt, ...) \ + dev_err(&(vdev)->pdev->dev, fmt, ##__VA_ARGS__) +#define vdev_warn(vdev, fmt, ...) \ + dev_warn(&(vdev)->pdev->dev, fmt, ##__VA_ARGS__) +#define vdev_info(vdev, fmt, ...) \ + dev_info(&(vdev)->pdev->dev, fmt, ##__VA_ARGS__) + +#define vdev_neterr(vdev, fmt, ...) \ + netdev_err(vnic_get_netdev(vdev), fmt, ##__VA_ARGS__) +#define vdev_netwarn(vdev, fmt, ...) \ + netdev_warn(vnic_get_netdev(vdev), fmt, ##__VA_ARGS__) +#define vdev_netinfo(vdev, fmt, ...) \ + netdev_info(vnic_get_netdev(vdev), fmt, ##__VA_ARGS__) + +static inline struct device *enic_get_dev(struct enic *enic) +{ + return &(enic->pdev->dev); +} + +static inline unsigned int enic_cq_rq(struct enic *enic, unsigned int rq) +{ + return rq; +} + +static inline unsigned int enic_cq_wq(struct enic *enic, unsigned int wq) +{ + return enic->rq_count + wq; +} + +static inline unsigned int enic_legacy_io_intr(void) +{ + return 0; +} + +static inline unsigned int enic_legacy_err_intr(void) +{ + return 1; +} + +static inline unsigned int enic_legacy_notify_intr(void) +{ + return 2; +} + +static inline unsigned int enic_msix_rq_intr(struct enic *enic, + unsigned int rq) +{ + return enic->cq[enic_cq_rq(enic, rq)].interrupt_offset; +} + +static inline unsigned int enic_msix_wq_intr(struct enic *enic, + unsigned int wq) +{ + return enic->cq[enic_cq_wq(enic, wq)].interrupt_offset; +} + +static inline unsigned int enic_msix_err_intr(struct enic *enic) +{ + return enic->rq_count + enic->wq_count; +} + +static inline unsigned int enic_msix_notify_intr(struct enic *enic) +{ + return enic->rq_count + enic->wq_count + 1; +} + +static inline bool enic_is_err_intr(struct enic *enic, int intr) +{ + switch (vnic_dev_get_intr_mode(enic->vdev)) { + case VNIC_DEV_INTR_MODE_INTX: + return intr == enic_legacy_err_intr(); + case VNIC_DEV_INTR_MODE_MSIX: + return intr == enic_msix_err_intr(enic); + case VNIC_DEV_INTR_MODE_MSI: + default: + return false; + } +} + +static inline bool enic_is_notify_intr(struct enic *enic, int intr) +{ + switch (vnic_dev_get_intr_mode(enic->vdev)) { + case VNIC_DEV_INTR_MODE_INTX: + return intr == enic_legacy_notify_intr(); + case VNIC_DEV_INTR_MODE_MSIX: + return intr == enic_msix_notify_intr(enic); + case VNIC_DEV_INTR_MODE_MSI: + default: + return false; + } +} + +static inline int enic_dma_map_check(struct enic *enic, dma_addr_t dma_addr) +{ + if (unlikely(pci_dma_mapping_error(enic->pdev, dma_addr))) { + net_warn_ratelimited("%s: PCI dma mapping failed!\n", + enic->netdev->name); + enic->gen_stats.dma_map_error++; + + return -ENOMEM; + } + + return 0; +} + +void enic_reset_addr_lists(struct enic *enic); +int enic_sriov_enabled(struct enic *enic); +int enic_is_valid_vf(struct enic *enic, int vf); +int enic_is_dynamic(struct enic *enic); +void enic_set_ethtool_ops(struct net_device *netdev); +int __enic_set_rsskey(struct enic *enic); + +#endif /* _ENIC_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/enic_api.c b/drivers/net/ethernet/cisco/enic/enic_api.c new file mode 100644 index 000000000..b028ea2de --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic_api.c @@ -0,0 +1,54 @@ +/** + * Copyright 2013 Cisco Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include <linux/netdevice.h> +#include <linux/spinlock.h> + +#include "vnic_dev.h" +#include "vnic_devcmd.h" + +#include "enic_res.h" +#include "enic.h" +#include "enic_api.h" + +int enic_api_devcmd_proxy_by_index(struct net_device *netdev, int vf, + enum vnic_devcmd_cmd cmd, u64 *a0, u64 *a1, int wait) +{ + int err; + struct enic *enic = netdev_priv(netdev); + struct vnic_dev *vdev = enic->vdev; + + spin_lock(&enic->enic_api_lock); + while (enic->enic_api_busy) { + spin_unlock(&enic->enic_api_lock); + cpu_relax(); + spin_lock(&enic->enic_api_lock); + } + + spin_lock_bh(&enic->devcmd_lock); + + vnic_dev_cmd_proxy_by_index_start(vdev, vf); + err = vnic_dev_cmd(vdev, cmd, a0, a1, wait); + vnic_dev_cmd_proxy_end(vdev); + + spin_unlock_bh(&enic->devcmd_lock); + spin_unlock(&enic->enic_api_lock); + + return err; +} +EXPORT_SYMBOL(enic_api_devcmd_proxy_by_index); diff --git a/drivers/net/ethernet/cisco/enic/enic_api.h b/drivers/net/ethernet/cisco/enic/enic_api.h new file mode 100644 index 000000000..6b9f9255a --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic_api.h @@ -0,0 +1,30 @@ +/** + * Copyright 2013 Cisco Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef __ENIC_API_H__ +#define __ENIC_API_H__ + +#include <linux/netdevice.h> + +#include "vnic_dev.h" +#include "vnic_devcmd.h" + +int enic_api_devcmd_proxy_by_index(struct net_device *netdev, int vf, + enum vnic_devcmd_cmd cmd, u64 *a0, u64 *a1, int wait); + +#endif diff --git a/drivers/net/ethernet/cisco/enic/enic_clsf.c b/drivers/net/ethernet/cisco/enic/enic_clsf.c new file mode 100644 index 000000000..99038dfc7 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic_clsf.c @@ -0,0 +1,285 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/if.h> +#include <linux/if_ether.h> +#include <linux/if_link.h> +#include <linux/netdevice.h> +#include <linux/in.h> +#include <linux/types.h> +#include <linux/skbuff.h> +#include <net/flow_dissector.h> +#include "enic_res.h" +#include "enic_clsf.h" + +/* enic_addfltr_5t - Add ipv4 5tuple filter + * @enic: enic struct of vnic + * @keys: flow_keys of ipv4 5tuple + * @rq: rq number to steer to + * + * This function returns filter_id(hardware_id) of the filter + * added. In case of error it returns a negative number. + */ +int enic_addfltr_5t(struct enic *enic, struct flow_keys *keys, u16 rq) +{ + int res; + struct filter data; + + switch (keys->basic.ip_proto) { + case IPPROTO_TCP: + data.u.ipv4.protocol = PROTO_TCP; + break; + case IPPROTO_UDP: + data.u.ipv4.protocol = PROTO_UDP; + break; + default: + return -EPROTONOSUPPORT; + }; + data.type = FILTER_IPV4_5TUPLE; + data.u.ipv4.src_addr = ntohl(keys->addrs.v4addrs.src); + data.u.ipv4.dst_addr = ntohl(keys->addrs.v4addrs.dst); + data.u.ipv4.src_port = ntohs(keys->ports.src); + data.u.ipv4.dst_port = ntohs(keys->ports.dst); + data.u.ipv4.flags = FILTER_FIELDS_IPV4_5TUPLE; + + spin_lock_bh(&enic->devcmd_lock); + res = vnic_dev_classifier(enic->vdev, CLSF_ADD, &rq, &data); + spin_unlock_bh(&enic->devcmd_lock); + res = (res == 0) ? rq : res; + + return res; +} + +/* enic_delfltr - Delete clsf filter + * @enic: enic struct of vnic + * @filter_id: filter_is(hardware_id) of filter to be deleted + * + * This function returns zero in case of success, negative number incase of + * error. + */ +int enic_delfltr(struct enic *enic, u16 filter_id) +{ + int ret; + + spin_lock_bh(&enic->devcmd_lock); + ret = vnic_dev_classifier(enic->vdev, CLSF_DEL, &filter_id, NULL); + spin_unlock_bh(&enic->devcmd_lock); + + return ret; +} + +/* enic_rfs_flw_tbl_init - initialize enic->rfs_h members + * @enic: enic data + */ +void enic_rfs_flw_tbl_init(struct enic *enic) +{ + int i; + + spin_lock_init(&enic->rfs_h.lock); + for (i = 0; i <= ENIC_RFS_FLW_MASK; i++) + INIT_HLIST_HEAD(&enic->rfs_h.ht_head[i]); + enic->rfs_h.max = enic->config.num_arfs; + enic->rfs_h.free = enic->rfs_h.max; + enic->rfs_h.toclean = 0; +} + +void enic_rfs_flw_tbl_free(struct enic *enic) +{ + int i; + + enic_rfs_timer_stop(enic); + spin_lock_bh(&enic->rfs_h.lock); + for (i = 0; i < (1 << ENIC_RFS_FLW_BITSHIFT); i++) { + struct hlist_head *hhead; + struct hlist_node *tmp; + struct enic_rfs_fltr_node *n; + + hhead = &enic->rfs_h.ht_head[i]; + hlist_for_each_entry_safe(n, tmp, hhead, node) { + enic_delfltr(enic, n->fltr_id); + hlist_del(&n->node); + kfree(n); + enic->rfs_h.free++; + } + } + spin_unlock_bh(&enic->rfs_h.lock); +} + +struct enic_rfs_fltr_node *htbl_fltr_search(struct enic *enic, u16 fltr_id) +{ + int i; + + for (i = 0; i < (1 << ENIC_RFS_FLW_BITSHIFT); i++) { + struct hlist_head *hhead; + struct hlist_node *tmp; + struct enic_rfs_fltr_node *n; + + hhead = &enic->rfs_h.ht_head[i]; + hlist_for_each_entry_safe(n, tmp, hhead, node) + if (n->fltr_id == fltr_id) + return n; + } + + return NULL; +} + +#ifdef CONFIG_RFS_ACCEL +void enic_flow_may_expire(struct timer_list *t) +{ + struct enic *enic = from_timer(enic, t, rfs_h.rfs_may_expire); + bool res; + int j; + + spin_lock_bh(&enic->rfs_h.lock); + for (j = 0; j < ENIC_CLSF_EXPIRE_COUNT; j++) { + struct hlist_head *hhead; + struct hlist_node *tmp; + struct enic_rfs_fltr_node *n; + + hhead = &enic->rfs_h.ht_head[enic->rfs_h.toclean++]; + hlist_for_each_entry_safe(n, tmp, hhead, node) { + res = rps_may_expire_flow(enic->netdev, n->rq_id, + n->flow_id, n->fltr_id); + if (res) { + res = enic_delfltr(enic, n->fltr_id); + if (unlikely(res)) + continue; + hlist_del(&n->node); + kfree(n); + enic->rfs_h.free++; + } + } + } + spin_unlock_bh(&enic->rfs_h.lock); + mod_timer(&enic->rfs_h.rfs_may_expire, jiffies + HZ/4); +} + +static struct enic_rfs_fltr_node *htbl_key_search(struct hlist_head *h, + struct flow_keys *k) +{ + struct enic_rfs_fltr_node *tpos; + + hlist_for_each_entry(tpos, h, node) + if (tpos->keys.addrs.v4addrs.src == k->addrs.v4addrs.src && + tpos->keys.addrs.v4addrs.dst == k->addrs.v4addrs.dst && + tpos->keys.ports.ports == k->ports.ports && + tpos->keys.basic.ip_proto == k->basic.ip_proto && + tpos->keys.basic.n_proto == k->basic.n_proto) + return tpos; + return NULL; +} + +int enic_rx_flow_steer(struct net_device *dev, const struct sk_buff *skb, + u16 rxq_index, u32 flow_id) +{ + struct flow_keys keys; + struct enic_rfs_fltr_node *n; + struct enic *enic; + u16 tbl_idx; + int res, i; + + enic = netdev_priv(dev); + res = skb_flow_dissect_flow_keys(skb, &keys, 0); + if (!res || keys.basic.n_proto != htons(ETH_P_IP) || + (keys.basic.ip_proto != IPPROTO_TCP && + keys.basic.ip_proto != IPPROTO_UDP)) + return -EPROTONOSUPPORT; + + tbl_idx = skb_get_hash_raw(skb) & ENIC_RFS_FLW_MASK; + spin_lock_bh(&enic->rfs_h.lock); + n = htbl_key_search(&enic->rfs_h.ht_head[tbl_idx], &keys); + + if (n) { /* entry already present */ + if (rxq_index == n->rq_id) { + res = -EEXIST; + goto ret_unlock; + } + + /* desired rq changed for the flow, we need to delete + * old fltr and add new one + * + * The moment we delete the fltr, the upcoming pkts + * are put it default rq based on rss. When we add + * new filter, upcoming pkts are put in desired queue. + * This could cause ooo pkts. + * + * Lets 1st try adding new fltr and then del old one. + */ + i = --enic->rfs_h.free; + /* clsf tbl is full, we have to del old fltr first*/ + if (unlikely(i < 0)) { + enic->rfs_h.free++; + res = enic_delfltr(enic, n->fltr_id); + if (unlikely(res < 0)) + goto ret_unlock; + res = enic_addfltr_5t(enic, &keys, rxq_index); + if (res < 0) { + hlist_del(&n->node); + enic->rfs_h.free++; + goto ret_unlock; + } + /* add new fltr 1st then del old fltr */ + } else { + int ret; + + res = enic_addfltr_5t(enic, &keys, rxq_index); + if (res < 0) { + enic->rfs_h.free++; + goto ret_unlock; + } + ret = enic_delfltr(enic, n->fltr_id); + /* deleting old fltr failed. Add old fltr to list. + * enic_flow_may_expire() will try to delete it later. + */ + if (unlikely(ret < 0)) { + struct enic_rfs_fltr_node *d; + struct hlist_head *head; + + head = &enic->rfs_h.ht_head[tbl_idx]; + d = kmalloc(sizeof(*d), GFP_ATOMIC); + if (d) { + d->fltr_id = n->fltr_id; + INIT_HLIST_NODE(&d->node); + hlist_add_head(&d->node, head); + } + } else { + enic->rfs_h.free++; + } + } + n->rq_id = rxq_index; + n->fltr_id = res; + n->flow_id = flow_id; + /* entry not present */ + } else { + i = --enic->rfs_h.free; + if (i <= 0) { + enic->rfs_h.free++; + res = -EBUSY; + goto ret_unlock; + } + + n = kmalloc(sizeof(*n), GFP_ATOMIC); + if (!n) { + res = -ENOMEM; + enic->rfs_h.free++; + goto ret_unlock; + } + + res = enic_addfltr_5t(enic, &keys, rxq_index); + if (res < 0) { + kfree(n); + enic->rfs_h.free++; + goto ret_unlock; + } + n->rq_id = rxq_index; + n->fltr_id = res; + n->flow_id = flow_id; + n->keys = keys; + INIT_HLIST_NODE(&n->node); + hlist_add_head(&n->node, &enic->rfs_h.ht_head[tbl_idx]); + } + +ret_unlock: + spin_unlock_bh(&enic->rfs_h.lock); + return res; +} + +#endif /* CONFIG_RFS_ACCEL */ diff --git a/drivers/net/ethernet/cisco/enic/enic_clsf.h b/drivers/net/ethernet/cisco/enic/enic_clsf.h new file mode 100644 index 000000000..8c4ce50da --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic_clsf.h @@ -0,0 +1,36 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _ENIC_CLSF_H_ +#define _ENIC_CLSF_H_ + +#include "vnic_dev.h" +#include "enic.h" + +#define ENIC_CLSF_EXPIRE_COUNT 128 + +int enic_addfltr_5t(struct enic *enic, struct flow_keys *keys, u16 rq); +int enic_delfltr(struct enic *enic, u16 filter_id); +void enic_rfs_flw_tbl_init(struct enic *enic); +void enic_rfs_flw_tbl_free(struct enic *enic); +struct enic_rfs_fltr_node *htbl_fltr_search(struct enic *enic, u16 fltr_id); + +#ifdef CONFIG_RFS_ACCEL +int enic_rx_flow_steer(struct net_device *dev, const struct sk_buff *skb, + u16 rxq_index, u32 flow_id); +void enic_flow_may_expire(struct timer_list *t); + +static inline void enic_rfs_timer_start(struct enic *enic) +{ + timer_setup(&enic->rfs_h.rfs_may_expire, enic_flow_may_expire, 0); + mod_timer(&enic->rfs_h.rfs_may_expire, jiffies + HZ/4); +} + +static inline void enic_rfs_timer_stop(struct enic *enic) +{ + del_timer_sync(&enic->rfs_h.rfs_may_expire); +} +#else +static inline void enic_rfs_timer_start(struct enic *enic) {} +static inline void enic_rfs_timer_stop(struct enic *enic) {} +#endif /* CONFIG_RFS_ACCEL */ + +#endif /* _ENIC_CLSF_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/enic_dev.c b/drivers/net/ethernet/cisco/enic/enic_dev.c new file mode 100644 index 000000000..f8d2a6a34 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic_dev.c @@ -0,0 +1,236 @@ +/* + * Copyright 2011 Cisco Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include <linux/pci.h> +#include <linux/etherdevice.h> + +#include "vnic_dev.h" +#include "vnic_vic.h" +#include "enic_res.h" +#include "enic.h" +#include "enic_dev.h" + +int enic_dev_fw_info(struct enic *enic, struct vnic_devcmd_fw_info **fw_info) +{ + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_fw_info(enic->vdev, fw_info); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_stats_dump(struct enic *enic, struct vnic_stats **vstats) +{ + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_stats_dump(enic->vdev, vstats); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_add_station_addr(struct enic *enic) +{ + int err; + + if (!is_valid_ether_addr(enic->netdev->dev_addr)) + return -EADDRNOTAVAIL; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_add_addr(enic->vdev, enic->netdev->dev_addr); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_del_station_addr(struct enic *enic) +{ + int err; + + if (!is_valid_ether_addr(enic->netdev->dev_addr)) + return -EADDRNOTAVAIL; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_del_addr(enic->vdev, enic->netdev->dev_addr); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_packet_filter(struct enic *enic, int directed, int multicast, + int broadcast, int promisc, int allmulti) +{ + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_packet_filter(enic->vdev, directed, + multicast, broadcast, promisc, allmulti); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_add_addr(struct enic *enic, const u8 *addr) +{ + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_add_addr(enic->vdev, addr); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_del_addr(struct enic *enic, const u8 *addr) +{ + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_del_addr(enic->vdev, addr); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_notify_unset(struct enic *enic) +{ + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_notify_unset(enic->vdev); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_hang_notify(struct enic *enic) +{ + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_hang_notify(enic->vdev); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_set_ig_vlan_rewrite_mode(struct enic *enic) +{ + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_set_ig_vlan_rewrite_mode(enic->vdev, + IG_VLAN_REWRITE_MODE_PRIORITY_TAG_DEFAULT_VLAN); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_enable(struct enic *enic) +{ + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_enable_wait(enic->vdev); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_disable(struct enic *enic) +{ + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_disable(enic->vdev); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_intr_coal_timer_info(struct enic *enic) +{ + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = vnic_dev_intr_coal_timer_info(enic->vdev); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +/* rtnl lock is held */ +int enic_vlan_rx_add_vid(struct net_device *netdev, __be16 proto, u16 vid) +{ + struct enic *enic = netdev_priv(netdev); + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = enic_add_vlan(enic, vid); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +/* rtnl lock is held */ +int enic_vlan_rx_kill_vid(struct net_device *netdev, __be16 proto, u16 vid) +{ + struct enic *enic = netdev_priv(netdev); + int err; + + spin_lock_bh(&enic->devcmd_lock); + err = enic_del_vlan(enic, vid); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +int enic_dev_status_to_errno(int devcmd_status) +{ + switch (devcmd_status) { + case ERR_SUCCESS: + return 0; + case ERR_EINVAL: + return -EINVAL; + case ERR_EFAULT: + return -EFAULT; + case ERR_EPERM: + return -EPERM; + case ERR_EBUSY: + return -EBUSY; + case ERR_ECMDUNKNOWN: + case ERR_ENOTSUPPORTED: + return -EOPNOTSUPP; + case ERR_EBADSTATE: + return -EINVAL; + case ERR_ENOMEM: + return -ENOMEM; + case ERR_ETIMEDOUT: + return -ETIMEDOUT; + case ERR_ELINKDOWN: + return -ENETDOWN; + case ERR_EINPROGRESS: + return -EINPROGRESS; + case ERR_EMAXRES: + default: + return (devcmd_status < 0) ? devcmd_status : -1; + } +} diff --git a/drivers/net/ethernet/cisco/enic/enic_dev.h b/drivers/net/ethernet/cisco/enic/enic_dev.h new file mode 100644 index 000000000..f5bb058b3 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic_dev.h @@ -0,0 +1,60 @@ +/* + * Copyright 2011 Cisco Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _ENIC_DEV_H_ +#define _ENIC_DEV_H_ + +#include "vnic_dev.h" +#include "vnic_vic.h" + +/* + * Calls the devcmd function given by argument vnicdevcmdfn. + * If vf argument is valid, it proxies the devcmd + */ +#define ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, vnicdevcmdfn, ...) \ + do { \ + spin_lock_bh(&enic->devcmd_lock); \ + if (enic_is_valid_vf(enic, vf)) { \ + vnic_dev_cmd_proxy_by_index_start(enic->vdev, vf); \ + err = vnicdevcmdfn(enic->vdev, ##__VA_ARGS__); \ + vnic_dev_cmd_proxy_end(enic->vdev); \ + } else { \ + err = vnicdevcmdfn(enic->vdev, ##__VA_ARGS__); \ + } \ + spin_unlock_bh(&enic->devcmd_lock); \ + } while (0) + +int enic_dev_fw_info(struct enic *enic, struct vnic_devcmd_fw_info **fw_info); +int enic_dev_stats_dump(struct enic *enic, struct vnic_stats **vstats); +int enic_dev_add_station_addr(struct enic *enic); +int enic_dev_del_station_addr(struct enic *enic); +int enic_dev_packet_filter(struct enic *enic, int directed, int multicast, + int broadcast, int promisc, int allmulti); +int enic_dev_add_addr(struct enic *enic, const u8 *addr); +int enic_dev_del_addr(struct enic *enic, const u8 *addr); +int enic_vlan_rx_add_vid(struct net_device *netdev, __be16 proto, u16 vid); +int enic_vlan_rx_kill_vid(struct net_device *netdev, __be16 proto, u16 vid); +int enic_dev_notify_unset(struct enic *enic); +int enic_dev_hang_notify(struct enic *enic); +int enic_dev_set_ig_vlan_rewrite_mode(struct enic *enic); +int enic_dev_enable(struct enic *enic); +int enic_dev_disable(struct enic *enic); +int enic_dev_intr_coal_timer_info(struct enic *enic); +int enic_dev_status_to_errno(int devcmd_status); + +#endif /* _ENIC_DEV_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/enic_ethtool.c b/drivers/net/ethernet/cisco/enic/enic_ethtool.c new file mode 100644 index 000000000..f42f7a6e1 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic_ethtool.c @@ -0,0 +1,663 @@ +/** + * Copyright 2013 Cisco Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include <linux/netdevice.h> +#include <linux/ethtool.h> +#include <linux/net_tstamp.h> + +#include "enic_res.h" +#include "enic.h" +#include "enic_dev.h" +#include "enic_clsf.h" +#include "vnic_rss.h" +#include "vnic_stats.h" + +struct enic_stat { + char name[ETH_GSTRING_LEN]; + unsigned int index; +}; + +#define ENIC_TX_STAT(stat) { \ + .name = #stat, \ + .index = offsetof(struct vnic_tx_stats, stat) / sizeof(u64) \ +} + +#define ENIC_RX_STAT(stat) { \ + .name = #stat, \ + .index = offsetof(struct vnic_rx_stats, stat) / sizeof(u64) \ +} + +#define ENIC_GEN_STAT(stat) { \ + .name = #stat, \ + .index = offsetof(struct vnic_gen_stats, stat) / sizeof(u64)\ +} + +static const struct enic_stat enic_tx_stats[] = { + ENIC_TX_STAT(tx_frames_ok), + ENIC_TX_STAT(tx_unicast_frames_ok), + ENIC_TX_STAT(tx_multicast_frames_ok), + ENIC_TX_STAT(tx_broadcast_frames_ok), + ENIC_TX_STAT(tx_bytes_ok), + ENIC_TX_STAT(tx_unicast_bytes_ok), + ENIC_TX_STAT(tx_multicast_bytes_ok), + ENIC_TX_STAT(tx_broadcast_bytes_ok), + ENIC_TX_STAT(tx_drops), + ENIC_TX_STAT(tx_errors), + ENIC_TX_STAT(tx_tso), +}; + +static const struct enic_stat enic_rx_stats[] = { + ENIC_RX_STAT(rx_frames_ok), + ENIC_RX_STAT(rx_frames_total), + ENIC_RX_STAT(rx_unicast_frames_ok), + ENIC_RX_STAT(rx_multicast_frames_ok), + ENIC_RX_STAT(rx_broadcast_frames_ok), + ENIC_RX_STAT(rx_bytes_ok), + ENIC_RX_STAT(rx_unicast_bytes_ok), + ENIC_RX_STAT(rx_multicast_bytes_ok), + ENIC_RX_STAT(rx_broadcast_bytes_ok), + ENIC_RX_STAT(rx_drop), + ENIC_RX_STAT(rx_no_bufs), + ENIC_RX_STAT(rx_errors), + ENIC_RX_STAT(rx_rss), + ENIC_RX_STAT(rx_crc_errors), + ENIC_RX_STAT(rx_frames_64), + ENIC_RX_STAT(rx_frames_127), + ENIC_RX_STAT(rx_frames_255), + ENIC_RX_STAT(rx_frames_511), + ENIC_RX_STAT(rx_frames_1023), + ENIC_RX_STAT(rx_frames_1518), + ENIC_RX_STAT(rx_frames_to_max), +}; + +static const struct enic_stat enic_gen_stats[] = { + ENIC_GEN_STAT(dma_map_error), +}; + +static const unsigned int enic_n_tx_stats = ARRAY_SIZE(enic_tx_stats); +static const unsigned int enic_n_rx_stats = ARRAY_SIZE(enic_rx_stats); +static const unsigned int enic_n_gen_stats = ARRAY_SIZE(enic_gen_stats); + +static void enic_intr_coal_set_rx(struct enic *enic, u32 timer) +{ + int i; + int intr; + + for (i = 0; i < enic->rq_count; i++) { + intr = enic_msix_rq_intr(enic, i); + vnic_intr_coalescing_timer_set(&enic->intr[intr], timer); + } +} + +static int enic_get_ksettings(struct net_device *netdev, + struct ethtool_link_ksettings *ecmd) +{ + struct enic *enic = netdev_priv(netdev); + struct ethtool_link_settings *base = &ecmd->base; + + ethtool_link_ksettings_add_link_mode(ecmd, supported, + 10000baseT_Full); + ethtool_link_ksettings_add_link_mode(ecmd, supported, FIBRE); + ethtool_link_ksettings_add_link_mode(ecmd, advertising, + 10000baseT_Full); + ethtool_link_ksettings_add_link_mode(ecmd, advertising, FIBRE); + base->port = PORT_FIBRE; + + if (netif_carrier_ok(netdev)) { + base->speed = vnic_dev_port_speed(enic->vdev); + base->duplex = DUPLEX_FULL; + } else { + base->speed = SPEED_UNKNOWN; + base->duplex = DUPLEX_UNKNOWN; + } + + base->autoneg = AUTONEG_DISABLE; + + return 0; +} + +static void enic_get_drvinfo(struct net_device *netdev, + struct ethtool_drvinfo *drvinfo) +{ + struct enic *enic = netdev_priv(netdev); + struct vnic_devcmd_fw_info *fw_info; + int err; + + err = enic_dev_fw_info(enic, &fw_info); + /* return only when pci_zalloc_consistent fails in vnic_dev_fw_info + * For other failures, like devcmd failure, we return previously + * recorded info. + */ + if (err == -ENOMEM) + return; + + strlcpy(drvinfo->driver, DRV_NAME, sizeof(drvinfo->driver)); + strlcpy(drvinfo->version, DRV_VERSION, sizeof(drvinfo->version)); + strlcpy(drvinfo->fw_version, fw_info->fw_version, + sizeof(drvinfo->fw_version)); + strlcpy(drvinfo->bus_info, pci_name(enic->pdev), + sizeof(drvinfo->bus_info)); +} + +static void enic_get_strings(struct net_device *netdev, u32 stringset, + u8 *data) +{ + unsigned int i; + + switch (stringset) { + case ETH_SS_STATS: + for (i = 0; i < enic_n_tx_stats; i++) { + memcpy(data, enic_tx_stats[i].name, ETH_GSTRING_LEN); + data += ETH_GSTRING_LEN; + } + for (i = 0; i < enic_n_rx_stats; i++) { + memcpy(data, enic_rx_stats[i].name, ETH_GSTRING_LEN); + data += ETH_GSTRING_LEN; + } + for (i = 0; i < enic_n_gen_stats; i++) { + memcpy(data, enic_gen_stats[i].name, ETH_GSTRING_LEN); + data += ETH_GSTRING_LEN; + } + break; + } +} + +static void enic_get_ringparam(struct net_device *netdev, + struct ethtool_ringparam *ring) +{ + struct enic *enic = netdev_priv(netdev); + struct vnic_enet_config *c = &enic->config; + + ring->rx_max_pending = ENIC_MAX_RQ_DESCS; + ring->rx_pending = c->rq_desc_count; + ring->tx_max_pending = ENIC_MAX_WQ_DESCS; + ring->tx_pending = c->wq_desc_count; +} + +static int enic_set_ringparam(struct net_device *netdev, + struct ethtool_ringparam *ring) +{ + struct enic *enic = netdev_priv(netdev); + struct vnic_enet_config *c = &enic->config; + int running = netif_running(netdev); + unsigned int rx_pending; + unsigned int tx_pending; + int err = 0; + + if (ring->rx_mini_max_pending || ring->rx_mini_pending) { + netdev_info(netdev, + "modifying mini ring params is not supported"); + return -EINVAL; + } + if (ring->rx_jumbo_max_pending || ring->rx_jumbo_pending) { + netdev_info(netdev, + "modifying jumbo ring params is not supported"); + return -EINVAL; + } + rx_pending = c->rq_desc_count; + tx_pending = c->wq_desc_count; + if (ring->rx_pending > ENIC_MAX_RQ_DESCS || + ring->rx_pending < ENIC_MIN_RQ_DESCS) { + netdev_info(netdev, "rx pending (%u) not in range [%u,%u]", + ring->rx_pending, ENIC_MIN_RQ_DESCS, + ENIC_MAX_RQ_DESCS); + return -EINVAL; + } + if (ring->tx_pending > ENIC_MAX_WQ_DESCS || + ring->tx_pending < ENIC_MIN_WQ_DESCS) { + netdev_info(netdev, "tx pending (%u) not in range [%u,%u]", + ring->tx_pending, ENIC_MIN_WQ_DESCS, + ENIC_MAX_WQ_DESCS); + return -EINVAL; + } + if (running) + dev_close(netdev); + c->rq_desc_count = + ring->rx_pending & 0xffffffe0; /* must be aligned to groups of 32 */ + c->wq_desc_count = + ring->tx_pending & 0xffffffe0; /* must be aligned to groups of 32 */ + enic_free_vnic_resources(enic); + err = enic_alloc_vnic_resources(enic); + if (err) { + netdev_err(netdev, + "Failed to alloc vNIC resources, aborting\n"); + enic_free_vnic_resources(enic); + goto err_out; + } + enic_init_vnic_resources(enic); + if (running) { + err = dev_open(netdev); + if (err) + goto err_out; + } + return 0; +err_out: + c->rq_desc_count = rx_pending; + c->wq_desc_count = tx_pending; + return err; +} + +static int enic_get_sset_count(struct net_device *netdev, int sset) +{ + switch (sset) { + case ETH_SS_STATS: + return enic_n_tx_stats + enic_n_rx_stats + enic_n_gen_stats; + default: + return -EOPNOTSUPP; + } +} + +static void enic_get_ethtool_stats(struct net_device *netdev, + struct ethtool_stats *stats, u64 *data) +{ + struct enic *enic = netdev_priv(netdev); + struct vnic_stats *vstats; + unsigned int i; + int err; + + err = enic_dev_stats_dump(enic, &vstats); + /* return only when pci_zalloc_consistent fails in vnic_dev_stats_dump + * For other failures, like devcmd failure, we return previously + * recorded stats. + */ + if (err == -ENOMEM) + return; + + for (i = 0; i < enic_n_tx_stats; i++) + *(data++) = ((u64 *)&vstats->tx)[enic_tx_stats[i].index]; + for (i = 0; i < enic_n_rx_stats; i++) + *(data++) = ((u64 *)&vstats->rx)[enic_rx_stats[i].index]; + for (i = 0; i < enic_n_gen_stats; i++) + *(data++) = ((u64 *)&enic->gen_stats)[enic_gen_stats[i].index]; +} + +static u32 enic_get_msglevel(struct net_device *netdev) +{ + struct enic *enic = netdev_priv(netdev); + return enic->msg_enable; +} + +static void enic_set_msglevel(struct net_device *netdev, u32 value) +{ + struct enic *enic = netdev_priv(netdev); + enic->msg_enable = value; +} + +static int enic_get_coalesce(struct net_device *netdev, + struct ethtool_coalesce *ecmd) +{ + struct enic *enic = netdev_priv(netdev); + struct enic_rx_coal *rxcoal = &enic->rx_coalesce_setting; + + if (vnic_dev_get_intr_mode(enic->vdev) == VNIC_DEV_INTR_MODE_MSIX) + ecmd->tx_coalesce_usecs = enic->tx_coalesce_usecs; + ecmd->rx_coalesce_usecs = enic->rx_coalesce_usecs; + if (rxcoal->use_adaptive_rx_coalesce) + ecmd->use_adaptive_rx_coalesce = 1; + ecmd->rx_coalesce_usecs_low = rxcoal->small_pkt_range_start; + ecmd->rx_coalesce_usecs_high = rxcoal->range_end; + + return 0; +} + +static int enic_coalesce_valid(struct enic *enic, + struct ethtool_coalesce *ec) +{ + u32 coalesce_usecs_max = vnic_dev_get_intr_coal_timer_max(enic->vdev); + u32 rx_coalesce_usecs_high = min_t(u32, coalesce_usecs_max, + ec->rx_coalesce_usecs_high); + u32 rx_coalesce_usecs_low = min_t(u32, coalesce_usecs_max, + ec->rx_coalesce_usecs_low); + + if (ec->rx_max_coalesced_frames || + ec->rx_coalesce_usecs_irq || + ec->rx_max_coalesced_frames_irq || + ec->tx_max_coalesced_frames || + ec->tx_coalesce_usecs_irq || + ec->tx_max_coalesced_frames_irq || + ec->stats_block_coalesce_usecs || + ec->use_adaptive_tx_coalesce || + ec->pkt_rate_low || + ec->rx_max_coalesced_frames_low || + ec->tx_coalesce_usecs_low || + ec->tx_max_coalesced_frames_low || + ec->pkt_rate_high || + ec->rx_max_coalesced_frames_high || + ec->tx_coalesce_usecs_high || + ec->tx_max_coalesced_frames_high || + ec->rate_sample_interval) + return -EINVAL; + + if ((vnic_dev_get_intr_mode(enic->vdev) != VNIC_DEV_INTR_MODE_MSIX) && + ec->tx_coalesce_usecs) + return -EINVAL; + + if ((ec->tx_coalesce_usecs > coalesce_usecs_max) || + (ec->rx_coalesce_usecs > coalesce_usecs_max) || + (ec->rx_coalesce_usecs_low > coalesce_usecs_max) || + (ec->rx_coalesce_usecs_high > coalesce_usecs_max)) + netdev_info(enic->netdev, "ethtool_set_coalesce: adaptor supports max coalesce value of %d. Setting max value.\n", + coalesce_usecs_max); + + if (ec->rx_coalesce_usecs_high && + (rx_coalesce_usecs_high < + rx_coalesce_usecs_low + ENIC_AIC_LARGE_PKT_DIFF)) + return -EINVAL; + + return 0; +} + +static int enic_set_coalesce(struct net_device *netdev, + struct ethtool_coalesce *ecmd) +{ + struct enic *enic = netdev_priv(netdev); + u32 tx_coalesce_usecs; + u32 rx_coalesce_usecs; + u32 rx_coalesce_usecs_low; + u32 rx_coalesce_usecs_high; + u32 coalesce_usecs_max; + unsigned int i, intr; + int ret; + struct enic_rx_coal *rxcoal = &enic->rx_coalesce_setting; + + ret = enic_coalesce_valid(enic, ecmd); + if (ret) + return ret; + coalesce_usecs_max = vnic_dev_get_intr_coal_timer_max(enic->vdev); + tx_coalesce_usecs = min_t(u32, ecmd->tx_coalesce_usecs, + coalesce_usecs_max); + rx_coalesce_usecs = min_t(u32, ecmd->rx_coalesce_usecs, + coalesce_usecs_max); + + rx_coalesce_usecs_low = min_t(u32, ecmd->rx_coalesce_usecs_low, + coalesce_usecs_max); + rx_coalesce_usecs_high = min_t(u32, ecmd->rx_coalesce_usecs_high, + coalesce_usecs_max); + + if (vnic_dev_get_intr_mode(enic->vdev) == VNIC_DEV_INTR_MODE_MSIX) { + for (i = 0; i < enic->wq_count; i++) { + intr = enic_msix_wq_intr(enic, i); + vnic_intr_coalescing_timer_set(&enic->intr[intr], + tx_coalesce_usecs); + } + enic->tx_coalesce_usecs = tx_coalesce_usecs; + } + rxcoal->use_adaptive_rx_coalesce = !!ecmd->use_adaptive_rx_coalesce; + if (!rxcoal->use_adaptive_rx_coalesce) + enic_intr_coal_set_rx(enic, rx_coalesce_usecs); + if (ecmd->rx_coalesce_usecs_high) { + rxcoal->range_end = rx_coalesce_usecs_high; + rxcoal->small_pkt_range_start = rx_coalesce_usecs_low; + rxcoal->large_pkt_range_start = rx_coalesce_usecs_low + + ENIC_AIC_LARGE_PKT_DIFF; + } + + enic->rx_coalesce_usecs = rx_coalesce_usecs; + + return 0; +} + +static int enic_grxclsrlall(struct enic *enic, struct ethtool_rxnfc *cmd, + u32 *rule_locs) +{ + int j, ret = 0, cnt = 0; + + cmd->data = enic->rfs_h.max - enic->rfs_h.free; + for (j = 0; j < (1 << ENIC_RFS_FLW_BITSHIFT); j++) { + struct hlist_head *hhead; + struct hlist_node *tmp; + struct enic_rfs_fltr_node *n; + + hhead = &enic->rfs_h.ht_head[j]; + hlist_for_each_entry_safe(n, tmp, hhead, node) { + if (cnt == cmd->rule_cnt) + return -EMSGSIZE; + rule_locs[cnt] = n->fltr_id; + cnt++; + } + } + cmd->rule_cnt = cnt; + + return ret; +} + +static int enic_grxclsrule(struct enic *enic, struct ethtool_rxnfc *cmd) +{ + struct ethtool_rx_flow_spec *fsp = + (struct ethtool_rx_flow_spec *)&cmd->fs; + struct enic_rfs_fltr_node *n; + + n = htbl_fltr_search(enic, (u16)fsp->location); + if (!n) + return -EINVAL; + switch (n->keys.basic.ip_proto) { + case IPPROTO_TCP: + fsp->flow_type = TCP_V4_FLOW; + break; + case IPPROTO_UDP: + fsp->flow_type = UDP_V4_FLOW; + break; + default: + return -EINVAL; + break; + } + + fsp->h_u.tcp_ip4_spec.ip4src = flow_get_u32_src(&n->keys); + fsp->m_u.tcp_ip4_spec.ip4src = (__u32)~0; + + fsp->h_u.tcp_ip4_spec.ip4dst = flow_get_u32_dst(&n->keys); + fsp->m_u.tcp_ip4_spec.ip4dst = (__u32)~0; + + fsp->h_u.tcp_ip4_spec.psrc = n->keys.ports.src; + fsp->m_u.tcp_ip4_spec.psrc = (__u16)~0; + + fsp->h_u.tcp_ip4_spec.pdst = n->keys.ports.dst; + fsp->m_u.tcp_ip4_spec.pdst = (__u16)~0; + + fsp->ring_cookie = n->rq_id; + + return 0; +} + +static int enic_get_rx_flow_hash(struct enic *enic, struct ethtool_rxnfc *cmd) +{ + u8 rss_hash_type = 0; + cmd->data = 0; + + spin_lock_bh(&enic->devcmd_lock); + (void)vnic_dev_capable_rss_hash_type(enic->vdev, &rss_hash_type); + spin_unlock_bh(&enic->devcmd_lock); + switch (cmd->flow_type) { + case TCP_V6_FLOW: + case TCP_V4_FLOW: + cmd->data |= RXH_L4_B_0_1 | RXH_L4_B_2_3 | + RXH_IP_SRC | RXH_IP_DST; + break; + case UDP_V6_FLOW: + cmd->data |= RXH_IP_SRC | RXH_IP_DST; + if (rss_hash_type & NIC_CFG_RSS_HASH_TYPE_UDP_IPV6) + cmd->data |= RXH_L4_B_0_1 | RXH_L4_B_2_3; + break; + case UDP_V4_FLOW: + cmd->data |= RXH_IP_SRC | RXH_IP_DST; + if (rss_hash_type & NIC_CFG_RSS_HASH_TYPE_UDP_IPV4) + cmd->data |= RXH_L4_B_0_1 | RXH_L4_B_2_3; + break; + case SCTP_V4_FLOW: + case AH_ESP_V4_FLOW: + case AH_V4_FLOW: + case ESP_V4_FLOW: + case SCTP_V6_FLOW: + case AH_ESP_V6_FLOW: + case AH_V6_FLOW: + case ESP_V6_FLOW: + case IPV4_FLOW: + case IPV6_FLOW: + cmd->data |= RXH_IP_SRC | RXH_IP_DST; + break; + default: + return -EINVAL; + } + + return 0; +} + +static int enic_get_rxnfc(struct net_device *dev, struct ethtool_rxnfc *cmd, + u32 *rule_locs) +{ + struct enic *enic = netdev_priv(dev); + int ret = 0; + + switch (cmd->cmd) { + case ETHTOOL_GRXRINGS: + cmd->data = enic->rq_count; + break; + case ETHTOOL_GRXCLSRLCNT: + spin_lock_bh(&enic->rfs_h.lock); + cmd->rule_cnt = enic->rfs_h.max - enic->rfs_h.free; + cmd->data = enic->rfs_h.max; + spin_unlock_bh(&enic->rfs_h.lock); + break; + case ETHTOOL_GRXCLSRLALL: + spin_lock_bh(&enic->rfs_h.lock); + ret = enic_grxclsrlall(enic, cmd, rule_locs); + spin_unlock_bh(&enic->rfs_h.lock); + break; + case ETHTOOL_GRXCLSRULE: + spin_lock_bh(&enic->rfs_h.lock); + ret = enic_grxclsrule(enic, cmd); + spin_unlock_bh(&enic->rfs_h.lock); + break; + case ETHTOOL_GRXFH: + ret = enic_get_rx_flow_hash(enic, cmd); + break; + default: + ret = -EOPNOTSUPP; + break; + } + + return ret; +} + +static int enic_get_tunable(struct net_device *dev, + const struct ethtool_tunable *tuna, void *data) +{ + struct enic *enic = netdev_priv(dev); + int ret = 0; + + switch (tuna->id) { + case ETHTOOL_RX_COPYBREAK: + *(u32 *)data = enic->rx_copybreak; + break; + default: + ret = -EINVAL; + break; + } + + return ret; +} + +static int enic_set_tunable(struct net_device *dev, + const struct ethtool_tunable *tuna, + const void *data) +{ + struct enic *enic = netdev_priv(dev); + int ret = 0; + + switch (tuna->id) { + case ETHTOOL_RX_COPYBREAK: + enic->rx_copybreak = *(u32 *)data; + break; + default: + ret = -EINVAL; + break; + } + + return ret; +} + +static u32 enic_get_rxfh_key_size(struct net_device *netdev) +{ + return ENIC_RSS_LEN; +} + +static int enic_get_rxfh(struct net_device *netdev, u32 *indir, u8 *hkey, + u8 *hfunc) +{ + struct enic *enic = netdev_priv(netdev); + + if (hkey) + memcpy(hkey, enic->rss_key, ENIC_RSS_LEN); + + if (hfunc) + *hfunc = ETH_RSS_HASH_TOP; + + return 0; +} + +static int enic_set_rxfh(struct net_device *netdev, const u32 *indir, + const u8 *hkey, const u8 hfunc) +{ + struct enic *enic = netdev_priv(netdev); + + if ((hfunc != ETH_RSS_HASH_NO_CHANGE && hfunc != ETH_RSS_HASH_TOP) || + indir) + return -EINVAL; + + if (hkey) + memcpy(enic->rss_key, hkey, ENIC_RSS_LEN); + + return __enic_set_rsskey(enic); +} + +static int enic_get_ts_info(struct net_device *netdev, + struct ethtool_ts_info *info) +{ + info->so_timestamping = SOF_TIMESTAMPING_TX_SOFTWARE | + SOF_TIMESTAMPING_RX_SOFTWARE | + SOF_TIMESTAMPING_SOFTWARE; + + return 0; +} + +static const struct ethtool_ops enic_ethtool_ops = { + .get_drvinfo = enic_get_drvinfo, + .get_msglevel = enic_get_msglevel, + .set_msglevel = enic_set_msglevel, + .get_link = ethtool_op_get_link, + .get_strings = enic_get_strings, + .get_ringparam = enic_get_ringparam, + .set_ringparam = enic_set_ringparam, + .get_sset_count = enic_get_sset_count, + .get_ethtool_stats = enic_get_ethtool_stats, + .get_coalesce = enic_get_coalesce, + .set_coalesce = enic_set_coalesce, + .get_rxnfc = enic_get_rxnfc, + .get_tunable = enic_get_tunable, + .set_tunable = enic_set_tunable, + .get_rxfh_key_size = enic_get_rxfh_key_size, + .get_rxfh = enic_get_rxfh, + .set_rxfh = enic_set_rxfh, + .get_link_ksettings = enic_get_ksettings, + .get_ts_info = enic_get_ts_info, +}; + +void enic_set_ethtool_ops(struct net_device *netdev) +{ + netdev->ethtool_ops = &enic_ethtool_ops; +} diff --git a/drivers/net/ethernet/cisco/enic/enic_main.c b/drivers/net/ethernet/cisco/enic/enic_main.c new file mode 100644 index 000000000..bfe0e8209 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic_main.c @@ -0,0 +1,3088 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include <linux/module.h> +#include <linux/kernel.h> +#include <linux/string.h> +#include <linux/errno.h> +#include <linux/types.h> +#include <linux/init.h> +#include <linux/interrupt.h> +#include <linux/workqueue.h> +#include <linux/pci.h> +#include <linux/netdevice.h> +#include <linux/etherdevice.h> +#include <linux/if.h> +#include <linux/if_ether.h> +#include <linux/if_vlan.h> +#include <linux/in.h> +#include <linux/ip.h> +#include <linux/ipv6.h> +#include <linux/tcp.h> +#include <linux/rtnetlink.h> +#include <linux/prefetch.h> +#include <net/ip6_checksum.h> +#include <linux/ktime.h> +#include <linux/numa.h> +#ifdef CONFIG_RFS_ACCEL +#include <linux/cpu_rmap.h> +#endif +#include <linux/crash_dump.h> +#include <net/busy_poll.h> +#include <net/vxlan.h> + +#include "cq_enet_desc.h" +#include "vnic_dev.h" +#include "vnic_intr.h" +#include "vnic_stats.h" +#include "vnic_vic.h" +#include "enic_res.h" +#include "enic.h" +#include "enic_dev.h" +#include "enic_pp.h" +#include "enic_clsf.h" + +#define ENIC_NOTIFY_TIMER_PERIOD (2 * HZ) +#define WQ_ENET_MAX_DESC_LEN (1 << WQ_ENET_LEN_BITS) +#define MAX_TSO (1 << 16) +#define ENIC_DESC_MAX_SPLITS (MAX_TSO / WQ_ENET_MAX_DESC_LEN + 1) + +#define PCI_DEVICE_ID_CISCO_VIC_ENET 0x0043 /* ethernet vnic */ +#define PCI_DEVICE_ID_CISCO_VIC_ENET_DYN 0x0044 /* enet dynamic vnic */ +#define PCI_DEVICE_ID_CISCO_VIC_ENET_VF 0x0071 /* enet SRIOV VF */ + +#define RX_COPYBREAK_DEFAULT 256 + +/* Supported devices */ +static const struct pci_device_id enic_id_table[] = { + { PCI_VDEVICE(CISCO, PCI_DEVICE_ID_CISCO_VIC_ENET) }, + { PCI_VDEVICE(CISCO, PCI_DEVICE_ID_CISCO_VIC_ENET_DYN) }, + { PCI_VDEVICE(CISCO, PCI_DEVICE_ID_CISCO_VIC_ENET_VF) }, + { 0, } /* end of table */ +}; + +MODULE_DESCRIPTION(DRV_DESCRIPTION); +MODULE_AUTHOR("Scott Feldman <scofeldm@cisco.com>"); +MODULE_LICENSE("GPL"); +MODULE_VERSION(DRV_VERSION); +MODULE_DEVICE_TABLE(pci, enic_id_table); + +#define ENIC_LARGE_PKT_THRESHOLD 1000 +#define ENIC_MAX_COALESCE_TIMERS 10 +/* Interrupt moderation table, which will be used to decide the + * coalescing timer values + * {rx_rate in Mbps, mapping percentage of the range} + */ +static struct enic_intr_mod_table mod_table[ENIC_MAX_COALESCE_TIMERS + 1] = { + {4000, 0}, + {4400, 10}, + {5060, 20}, + {5230, 30}, + {5540, 40}, + {5820, 50}, + {6120, 60}, + {6435, 70}, + {6745, 80}, + {7000, 90}, + {0xFFFFFFFF, 100} +}; + +/* This table helps the driver to pick different ranges for rx coalescing + * timer depending on the link speed. + */ +static struct enic_intr_mod_range mod_range[ENIC_MAX_LINK_SPEEDS] = { + {0, 0}, /* 0 - 4 Gbps */ + {0, 3}, /* 4 - 10 Gbps */ + {3, 6}, /* 10 - 40 Gbps */ +}; + +static void enic_init_affinity_hint(struct enic *enic) +{ + int numa_node = dev_to_node(&enic->pdev->dev); + int i; + + for (i = 0; i < enic->intr_count; i++) { + if (enic_is_err_intr(enic, i) || enic_is_notify_intr(enic, i) || + (cpumask_available(enic->msix[i].affinity_mask) && + !cpumask_empty(enic->msix[i].affinity_mask))) + continue; + if (zalloc_cpumask_var(&enic->msix[i].affinity_mask, + GFP_KERNEL)) + cpumask_set_cpu(cpumask_local_spread(i, numa_node), + enic->msix[i].affinity_mask); + } +} + +static void enic_free_affinity_hint(struct enic *enic) +{ + int i; + + for (i = 0; i < enic->intr_count; i++) { + if (enic_is_err_intr(enic, i) || enic_is_notify_intr(enic, i)) + continue; + free_cpumask_var(enic->msix[i].affinity_mask); + } +} + +static void enic_set_affinity_hint(struct enic *enic) +{ + int i; + int err; + + for (i = 0; i < enic->intr_count; i++) { + if (enic_is_err_intr(enic, i) || + enic_is_notify_intr(enic, i) || + !cpumask_available(enic->msix[i].affinity_mask) || + cpumask_empty(enic->msix[i].affinity_mask)) + continue; + err = irq_set_affinity_hint(enic->msix_entry[i].vector, + enic->msix[i].affinity_mask); + if (err) + netdev_warn(enic->netdev, "irq_set_affinity_hint failed, err %d\n", + err); + } + + for (i = 0; i < enic->wq_count; i++) { + int wq_intr = enic_msix_wq_intr(enic, i); + + if (cpumask_available(enic->msix[wq_intr].affinity_mask) && + !cpumask_empty(enic->msix[wq_intr].affinity_mask)) + netif_set_xps_queue(enic->netdev, + enic->msix[wq_intr].affinity_mask, + i); + } +} + +static void enic_unset_affinity_hint(struct enic *enic) +{ + int i; + + for (i = 0; i < enic->intr_count; i++) + irq_set_affinity_hint(enic->msix_entry[i].vector, NULL); +} + +static void enic_udp_tunnel_add(struct net_device *netdev, + struct udp_tunnel_info *ti) +{ + struct enic *enic = netdev_priv(netdev); + __be16 port = ti->port; + int err; + + spin_lock_bh(&enic->devcmd_lock); + + if (ti->type != UDP_TUNNEL_TYPE_VXLAN) { + netdev_info(netdev, "udp_tnl: only vxlan tunnel offload supported"); + goto error; + } + + switch (ti->sa_family) { + case AF_INET6: + if (!(enic->vxlan.flags & ENIC_VXLAN_OUTER_IPV6)) { + netdev_info(netdev, "vxlan: only IPv4 offload supported"); + goto error; + } + /* Fall through */ + case AF_INET: + break; + default: + goto error; + } + + if (enic->vxlan.vxlan_udp_port_number) { + if (ntohs(port) == enic->vxlan.vxlan_udp_port_number) + netdev_warn(netdev, "vxlan: udp port already offloaded"); + else + netdev_info(netdev, "vxlan: offload supported for only one UDP port"); + + goto error; + } + if ((vnic_dev_get_res_count(enic->vdev, RES_TYPE_WQ) != 1) && + !(enic->vxlan.flags & ENIC_VXLAN_MULTI_WQ)) { + netdev_info(netdev, "vxlan: vxlan offload with multi wq not supported on this adapter"); + goto error; + } + + err = vnic_dev_overlay_offload_cfg(enic->vdev, + OVERLAY_CFG_VXLAN_PORT_UPDATE, + ntohs(port)); + if (err) + goto error; + + err = vnic_dev_overlay_offload_ctrl(enic->vdev, OVERLAY_FEATURE_VXLAN, + enic->vxlan.patch_level); + if (err) + goto error; + + enic->vxlan.vxlan_udp_port_number = ntohs(port); + + netdev_info(netdev, "vxlan fw-vers-%d: offload enabled for udp port: %d, sa_family: %d ", + (int)enic->vxlan.patch_level, ntohs(port), ti->sa_family); + + goto unlock; + +error: + netdev_info(netdev, "failed to offload udp port: %d, sa_family: %d, type: %d", + ntohs(port), ti->sa_family, ti->type); +unlock: + spin_unlock_bh(&enic->devcmd_lock); +} + +static void enic_udp_tunnel_del(struct net_device *netdev, + struct udp_tunnel_info *ti) +{ + struct enic *enic = netdev_priv(netdev); + int err; + + spin_lock_bh(&enic->devcmd_lock); + + if ((ntohs(ti->port) != enic->vxlan.vxlan_udp_port_number) || + ti->type != UDP_TUNNEL_TYPE_VXLAN) { + netdev_info(netdev, "udp_tnl: port:%d, sa_family: %d, type: %d not offloaded", + ntohs(ti->port), ti->sa_family, ti->type); + goto unlock; + } + + err = vnic_dev_overlay_offload_ctrl(enic->vdev, OVERLAY_FEATURE_VXLAN, + OVERLAY_OFFLOAD_DISABLE); + if (err) { + netdev_err(netdev, "vxlan: del offload udp port: %d failed", + ntohs(ti->port)); + goto unlock; + } + + enic->vxlan.vxlan_udp_port_number = 0; + + netdev_info(netdev, "vxlan: del offload udp port %d, family %d\n", + ntohs(ti->port), ti->sa_family); + +unlock: + spin_unlock_bh(&enic->devcmd_lock); +} + +static netdev_features_t enic_features_check(struct sk_buff *skb, + struct net_device *dev, + netdev_features_t features) +{ + const struct ethhdr *eth = (struct ethhdr *)skb_inner_mac_header(skb); + struct enic *enic = netdev_priv(dev); + struct udphdr *udph; + u16 port = 0; + u8 proto; + + if (!skb->encapsulation) + return features; + + features = vxlan_features_check(skb, features); + + switch (vlan_get_protocol(skb)) { + case htons(ETH_P_IPV6): + if (!(enic->vxlan.flags & ENIC_VXLAN_OUTER_IPV6)) + goto out; + proto = ipv6_hdr(skb)->nexthdr; + break; + case htons(ETH_P_IP): + proto = ip_hdr(skb)->protocol; + break; + default: + goto out; + } + + switch (eth->h_proto) { + case ntohs(ETH_P_IPV6): + if (!(enic->vxlan.flags & ENIC_VXLAN_INNER_IPV6)) + goto out; + /* Fall through */ + case ntohs(ETH_P_IP): + break; + default: + goto out; + } + + + if (proto == IPPROTO_UDP) { + udph = udp_hdr(skb); + port = be16_to_cpu(udph->dest); + } + + /* HW supports offload of only one UDP port. Remove CSUM and GSO MASK + * for other UDP port tunnels + */ + if (port != enic->vxlan.vxlan_udp_port_number) + goto out; + + return features; + +out: + return features & ~(NETIF_F_CSUM_MASK | NETIF_F_GSO_MASK); +} + +int enic_is_dynamic(struct enic *enic) +{ + return enic->pdev->device == PCI_DEVICE_ID_CISCO_VIC_ENET_DYN; +} + +int enic_sriov_enabled(struct enic *enic) +{ + return (enic->priv_flags & ENIC_SRIOV_ENABLED) ? 1 : 0; +} + +static int enic_is_sriov_vf(struct enic *enic) +{ + return enic->pdev->device == PCI_DEVICE_ID_CISCO_VIC_ENET_VF; +} + +int enic_is_valid_vf(struct enic *enic, int vf) +{ +#ifdef CONFIG_PCI_IOV + return vf >= 0 && vf < enic->num_vfs; +#else + return 0; +#endif +} + +static void enic_free_wq_buf(struct vnic_wq *wq, struct vnic_wq_buf *buf) +{ + struct enic *enic = vnic_dev_priv(wq->vdev); + + if (buf->sop) + pci_unmap_single(enic->pdev, buf->dma_addr, + buf->len, PCI_DMA_TODEVICE); + else + pci_unmap_page(enic->pdev, buf->dma_addr, + buf->len, PCI_DMA_TODEVICE); + + if (buf->os_buf) + dev_kfree_skb_any(buf->os_buf); +} + +static void enic_wq_free_buf(struct vnic_wq *wq, + struct cq_desc *cq_desc, struct vnic_wq_buf *buf, void *opaque) +{ + enic_free_wq_buf(wq, buf); +} + +static int enic_wq_service(struct vnic_dev *vdev, struct cq_desc *cq_desc, + u8 type, u16 q_number, u16 completed_index, void *opaque) +{ + struct enic *enic = vnic_dev_priv(vdev); + + spin_lock(&enic->wq_lock[q_number]); + + vnic_wq_service(&enic->wq[q_number], cq_desc, + completed_index, enic_wq_free_buf, + opaque); + + if (netif_tx_queue_stopped(netdev_get_tx_queue(enic->netdev, q_number)) && + vnic_wq_desc_avail(&enic->wq[q_number]) >= + (MAX_SKB_FRAGS + ENIC_DESC_MAX_SPLITS)) + netif_wake_subqueue(enic->netdev, q_number); + + spin_unlock(&enic->wq_lock[q_number]); + + return 0; +} + +static bool enic_log_q_error(struct enic *enic) +{ + unsigned int i; + u32 error_status; + bool err = false; + + for (i = 0; i < enic->wq_count; i++) { + error_status = vnic_wq_error_status(&enic->wq[i]); + err |= error_status; + if (error_status) + netdev_err(enic->netdev, "WQ[%d] error_status %d\n", + i, error_status); + } + + for (i = 0; i < enic->rq_count; i++) { + error_status = vnic_rq_error_status(&enic->rq[i]); + err |= error_status; + if (error_status) + netdev_err(enic->netdev, "RQ[%d] error_status %d\n", + i, error_status); + } + + return err; +} + +static void enic_msglvl_check(struct enic *enic) +{ + u32 msg_enable = vnic_dev_msg_lvl(enic->vdev); + + if (msg_enable != enic->msg_enable) { + netdev_info(enic->netdev, "msg lvl changed from 0x%x to 0x%x\n", + enic->msg_enable, msg_enable); + enic->msg_enable = msg_enable; + } +} + +static void enic_mtu_check(struct enic *enic) +{ + u32 mtu = vnic_dev_mtu(enic->vdev); + struct net_device *netdev = enic->netdev; + + if (mtu && mtu != enic->port_mtu) { + enic->port_mtu = mtu; + if (enic_is_dynamic(enic) || enic_is_sriov_vf(enic)) { + mtu = max_t(int, ENIC_MIN_MTU, + min_t(int, ENIC_MAX_MTU, mtu)); + if (mtu != netdev->mtu) + schedule_work(&enic->change_mtu_work); + } else { + if (mtu < netdev->mtu) + netdev_warn(netdev, + "interface MTU (%d) set higher " + "than switch port MTU (%d)\n", + netdev->mtu, mtu); + } + } +} + +static void enic_link_check(struct enic *enic) +{ + int link_status = vnic_dev_link_status(enic->vdev); + int carrier_ok = netif_carrier_ok(enic->netdev); + + if (link_status && !carrier_ok) { + netdev_info(enic->netdev, "Link UP\n"); + netif_carrier_on(enic->netdev); + } else if (!link_status && carrier_ok) { + netdev_info(enic->netdev, "Link DOWN\n"); + netif_carrier_off(enic->netdev); + } +} + +static void enic_notify_check(struct enic *enic) +{ + enic_msglvl_check(enic); + enic_mtu_check(enic); + enic_link_check(enic); +} + +#define ENIC_TEST_INTR(pba, i) (pba & (1 << i)) + +static irqreturn_t enic_isr_legacy(int irq, void *data) +{ + struct net_device *netdev = data; + struct enic *enic = netdev_priv(netdev); + unsigned int io_intr = enic_legacy_io_intr(); + unsigned int err_intr = enic_legacy_err_intr(); + unsigned int notify_intr = enic_legacy_notify_intr(); + u32 pba; + + vnic_intr_mask(&enic->intr[io_intr]); + + pba = vnic_intr_legacy_pba(enic->legacy_pba); + if (!pba) { + vnic_intr_unmask(&enic->intr[io_intr]); + return IRQ_NONE; /* not our interrupt */ + } + + if (ENIC_TEST_INTR(pba, notify_intr)) { + enic_notify_check(enic); + vnic_intr_return_all_credits(&enic->intr[notify_intr]); + } + + if (ENIC_TEST_INTR(pba, err_intr)) { + vnic_intr_return_all_credits(&enic->intr[err_intr]); + enic_log_q_error(enic); + /* schedule recovery from WQ/RQ error */ + schedule_work(&enic->reset); + return IRQ_HANDLED; + } + + if (ENIC_TEST_INTR(pba, io_intr)) + napi_schedule_irqoff(&enic->napi[0]); + else + vnic_intr_unmask(&enic->intr[io_intr]); + + return IRQ_HANDLED; +} + +static irqreturn_t enic_isr_msi(int irq, void *data) +{ + struct enic *enic = data; + + /* With MSI, there is no sharing of interrupts, so this is + * our interrupt and there is no need to ack it. The device + * is not providing per-vector masking, so the OS will not + * write to PCI config space to mask/unmask the interrupt. + * We're using mask_on_assertion for MSI, so the device + * automatically masks the interrupt when the interrupt is + * generated. Later, when exiting polling, the interrupt + * will be unmasked (see enic_poll). + * + * Also, the device uses the same PCIe Traffic Class (TC) + * for Memory Write data and MSI, so there are no ordering + * issues; the MSI will always arrive at the Root Complex + * _after_ corresponding Memory Writes (i.e. descriptor + * writes). + */ + + napi_schedule_irqoff(&enic->napi[0]); + + return IRQ_HANDLED; +} + +static irqreturn_t enic_isr_msix(int irq, void *data) +{ + struct napi_struct *napi = data; + + napi_schedule_irqoff(napi); + + return IRQ_HANDLED; +} + +static irqreturn_t enic_isr_msix_err(int irq, void *data) +{ + struct enic *enic = data; + unsigned int intr = enic_msix_err_intr(enic); + + vnic_intr_return_all_credits(&enic->intr[intr]); + + if (enic_log_q_error(enic)) + /* schedule recovery from WQ/RQ error */ + schedule_work(&enic->reset); + + return IRQ_HANDLED; +} + +static irqreturn_t enic_isr_msix_notify(int irq, void *data) +{ + struct enic *enic = data; + unsigned int intr = enic_msix_notify_intr(enic); + + enic_notify_check(enic); + vnic_intr_return_all_credits(&enic->intr[intr]); + + return IRQ_HANDLED; +} + +static int enic_queue_wq_skb_cont(struct enic *enic, struct vnic_wq *wq, + struct sk_buff *skb, unsigned int len_left, + int loopback) +{ + const skb_frag_t *frag; + dma_addr_t dma_addr; + + /* Queue additional data fragments */ + for (frag = skb_shinfo(skb)->frags; len_left; frag++) { + len_left -= skb_frag_size(frag); + dma_addr = skb_frag_dma_map(&enic->pdev->dev, frag, 0, + skb_frag_size(frag), + DMA_TO_DEVICE); + if (unlikely(enic_dma_map_check(enic, dma_addr))) + return -ENOMEM; + enic_queue_wq_desc_cont(wq, skb, dma_addr, skb_frag_size(frag), + (len_left == 0), /* EOP? */ + loopback); + } + + return 0; +} + +static int enic_queue_wq_skb_vlan(struct enic *enic, struct vnic_wq *wq, + struct sk_buff *skb, int vlan_tag_insert, + unsigned int vlan_tag, int loopback) +{ + unsigned int head_len = skb_headlen(skb); + unsigned int len_left = skb->len - head_len; + int eop = (len_left == 0); + dma_addr_t dma_addr; + int err = 0; + + dma_addr = pci_map_single(enic->pdev, skb->data, head_len, + PCI_DMA_TODEVICE); + if (unlikely(enic_dma_map_check(enic, dma_addr))) + return -ENOMEM; + + /* Queue the main skb fragment. The fragments are no larger + * than max MTU(9000)+ETH_HDR_LEN(14) bytes, which is less + * than WQ_ENET_MAX_DESC_LEN length. So only one descriptor + * per fragment is queued. + */ + enic_queue_wq_desc(wq, skb, dma_addr, head_len, vlan_tag_insert, + vlan_tag, eop, loopback); + + if (!eop) + err = enic_queue_wq_skb_cont(enic, wq, skb, len_left, loopback); + + return err; +} + +static int enic_queue_wq_skb_csum_l4(struct enic *enic, struct vnic_wq *wq, + struct sk_buff *skb, int vlan_tag_insert, + unsigned int vlan_tag, int loopback) +{ + unsigned int head_len = skb_headlen(skb); + unsigned int len_left = skb->len - head_len; + unsigned int hdr_len = skb_checksum_start_offset(skb); + unsigned int csum_offset = hdr_len + skb->csum_offset; + int eop = (len_left == 0); + dma_addr_t dma_addr; + int err = 0; + + dma_addr = pci_map_single(enic->pdev, skb->data, head_len, + PCI_DMA_TODEVICE); + if (unlikely(enic_dma_map_check(enic, dma_addr))) + return -ENOMEM; + + /* Queue the main skb fragment. The fragments are no larger + * than max MTU(9000)+ETH_HDR_LEN(14) bytes, which is less + * than WQ_ENET_MAX_DESC_LEN length. So only one descriptor + * per fragment is queued. + */ + enic_queue_wq_desc_csum_l4(wq, skb, dma_addr, head_len, csum_offset, + hdr_len, vlan_tag_insert, vlan_tag, eop, + loopback); + + if (!eop) + err = enic_queue_wq_skb_cont(enic, wq, skb, len_left, loopback); + + return err; +} + +static void enic_preload_tcp_csum_encap(struct sk_buff *skb) +{ + const struct ethhdr *eth = (struct ethhdr *)skb_inner_mac_header(skb); + + switch (eth->h_proto) { + case ntohs(ETH_P_IP): + inner_ip_hdr(skb)->check = 0; + inner_tcp_hdr(skb)->check = + ~csum_tcpudp_magic(inner_ip_hdr(skb)->saddr, + inner_ip_hdr(skb)->daddr, 0, + IPPROTO_TCP, 0); + break; + case ntohs(ETH_P_IPV6): + inner_tcp_hdr(skb)->check = + ~csum_ipv6_magic(&inner_ipv6_hdr(skb)->saddr, + &inner_ipv6_hdr(skb)->daddr, 0, + IPPROTO_TCP, 0); + break; + default: + WARN_ONCE(1, "Non ipv4/ipv6 inner pkt for encap offload"); + break; + } +} + +static void enic_preload_tcp_csum(struct sk_buff *skb) +{ + /* Preload TCP csum field with IP pseudo hdr calculated + * with IP length set to zero. HW will later add in length + * to each TCP segment resulting from the TSO. + */ + + if (skb->protocol == cpu_to_be16(ETH_P_IP)) { + ip_hdr(skb)->check = 0; + tcp_hdr(skb)->check = ~csum_tcpudp_magic(ip_hdr(skb)->saddr, + ip_hdr(skb)->daddr, 0, IPPROTO_TCP, 0); + } else if (skb->protocol == cpu_to_be16(ETH_P_IPV6)) { + tcp_hdr(skb)->check = ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr, + &ipv6_hdr(skb)->daddr, 0, IPPROTO_TCP, 0); + } +} + +static int enic_queue_wq_skb_tso(struct enic *enic, struct vnic_wq *wq, + struct sk_buff *skb, unsigned int mss, + int vlan_tag_insert, unsigned int vlan_tag, + int loopback) +{ + unsigned int frag_len_left = skb_headlen(skb); + unsigned int len_left = skb->len - frag_len_left; + int eop = (len_left == 0); + unsigned int offset = 0; + unsigned int hdr_len; + dma_addr_t dma_addr; + unsigned int len; + skb_frag_t *frag; + + if (skb->encapsulation) { + hdr_len = skb_inner_transport_header(skb) - skb->data; + hdr_len += inner_tcp_hdrlen(skb); + enic_preload_tcp_csum_encap(skb); + } else { + hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb); + enic_preload_tcp_csum(skb); + } + + /* Queue WQ_ENET_MAX_DESC_LEN length descriptors + * for the main skb fragment + */ + while (frag_len_left) { + len = min(frag_len_left, (unsigned int)WQ_ENET_MAX_DESC_LEN); + dma_addr = pci_map_single(enic->pdev, skb->data + offset, len, + PCI_DMA_TODEVICE); + if (unlikely(enic_dma_map_check(enic, dma_addr))) + return -ENOMEM; + enic_queue_wq_desc_tso(wq, skb, dma_addr, len, mss, hdr_len, + vlan_tag_insert, vlan_tag, + eop && (len == frag_len_left), loopback); + frag_len_left -= len; + offset += len; + } + + if (eop) + return 0; + + /* Queue WQ_ENET_MAX_DESC_LEN length descriptors + * for additional data fragments + */ + for (frag = skb_shinfo(skb)->frags; len_left; frag++) { + len_left -= skb_frag_size(frag); + frag_len_left = skb_frag_size(frag); + offset = 0; + + while (frag_len_left) { + len = min(frag_len_left, + (unsigned int)WQ_ENET_MAX_DESC_LEN); + dma_addr = skb_frag_dma_map(&enic->pdev->dev, frag, + offset, len, + DMA_TO_DEVICE); + if (unlikely(enic_dma_map_check(enic, dma_addr))) + return -ENOMEM; + enic_queue_wq_desc_cont(wq, skb, dma_addr, len, + (len_left == 0) && + (len == frag_len_left),/*EOP*/ + loopback); + frag_len_left -= len; + offset += len; + } + } + + return 0; +} + +static inline int enic_queue_wq_skb_encap(struct enic *enic, struct vnic_wq *wq, + struct sk_buff *skb, + int vlan_tag_insert, + unsigned int vlan_tag, int loopback) +{ + unsigned int head_len = skb_headlen(skb); + unsigned int len_left = skb->len - head_len; + /* Hardware will overwrite the checksum fields, calculating from + * scratch and ignoring the value placed by software. + * Offload mode = 00 + * mss[2], mss[1], mss[0] bits are set + */ + unsigned int mss_or_csum = 7; + int eop = (len_left == 0); + dma_addr_t dma_addr; + int err = 0; + + dma_addr = pci_map_single(enic->pdev, skb->data, head_len, + PCI_DMA_TODEVICE); + if (unlikely(enic_dma_map_check(enic, dma_addr))) + return -ENOMEM; + + enic_queue_wq_desc_ex(wq, skb, dma_addr, head_len, mss_or_csum, 0, + vlan_tag_insert, vlan_tag, + WQ_ENET_OFFLOAD_MODE_CSUM, eop, 1 /* SOP */, eop, + loopback); + if (!eop) + err = enic_queue_wq_skb_cont(enic, wq, skb, len_left, loopback); + + return err; +} + +static inline int enic_queue_wq_skb(struct enic *enic, + struct vnic_wq *wq, struct sk_buff *skb) +{ + unsigned int mss = skb_shinfo(skb)->gso_size; + unsigned int vlan_tag = 0; + int vlan_tag_insert = 0; + int loopback = 0; + int err; + + if (skb_vlan_tag_present(skb)) { + /* VLAN tag from trunking driver */ + vlan_tag_insert = 1; + vlan_tag = skb_vlan_tag_get(skb); + } else if (enic->loop_enable) { + vlan_tag = enic->loop_tag; + loopback = 1; + } + + if (mss) + err = enic_queue_wq_skb_tso(enic, wq, skb, mss, + vlan_tag_insert, vlan_tag, + loopback); + else if (skb->encapsulation) + err = enic_queue_wq_skb_encap(enic, wq, skb, vlan_tag_insert, + vlan_tag, loopback); + else if (skb->ip_summed == CHECKSUM_PARTIAL) + err = enic_queue_wq_skb_csum_l4(enic, wq, skb, vlan_tag_insert, + vlan_tag, loopback); + else + err = enic_queue_wq_skb_vlan(enic, wq, skb, vlan_tag_insert, + vlan_tag, loopback); + if (unlikely(err)) { + struct vnic_wq_buf *buf; + + buf = wq->to_use->prev; + /* while not EOP of previous pkt && queue not empty. + * For all non EOP bufs, os_buf is NULL. + */ + while (!buf->os_buf && (buf->next != wq->to_clean)) { + enic_free_wq_buf(wq, buf); + wq->ring.desc_avail++; + buf = buf->prev; + } + wq->to_use = buf->next; + dev_kfree_skb(skb); + } + return err; +} + +/* netif_tx_lock held, process context with BHs disabled, or BH */ +static netdev_tx_t enic_hard_start_xmit(struct sk_buff *skb, + struct net_device *netdev) +{ + struct enic *enic = netdev_priv(netdev); + struct vnic_wq *wq; + unsigned int txq_map; + struct netdev_queue *txq; + + if (skb->len <= 0) { + dev_kfree_skb_any(skb); + return NETDEV_TX_OK; + } + + txq_map = skb_get_queue_mapping(skb) % enic->wq_count; + wq = &enic->wq[txq_map]; + txq = netdev_get_tx_queue(netdev, txq_map); + + /* Non-TSO sends must fit within ENIC_NON_TSO_MAX_DESC descs, + * which is very likely. In the off chance it's going to take + * more than * ENIC_NON_TSO_MAX_DESC, linearize the skb. + */ + + if (skb_shinfo(skb)->gso_size == 0 && + skb_shinfo(skb)->nr_frags + 1 > ENIC_NON_TSO_MAX_DESC && + skb_linearize(skb)) { + dev_kfree_skb_any(skb); + return NETDEV_TX_OK; + } + + spin_lock(&enic->wq_lock[txq_map]); + + if (vnic_wq_desc_avail(wq) < + skb_shinfo(skb)->nr_frags + ENIC_DESC_MAX_SPLITS) { + netif_tx_stop_queue(txq); + /* This is a hard error, log it */ + netdev_err(netdev, "BUG! Tx ring full when queue awake!\n"); + spin_unlock(&enic->wq_lock[txq_map]); + return NETDEV_TX_BUSY; + } + + if (enic_queue_wq_skb(enic, wq, skb)) + goto error; + + if (vnic_wq_desc_avail(wq) < MAX_SKB_FRAGS + ENIC_DESC_MAX_SPLITS) + netif_tx_stop_queue(txq); + skb_tx_timestamp(skb); + if (!skb->xmit_more || netif_xmit_stopped(txq)) + vnic_wq_doorbell(wq); + +error: + spin_unlock(&enic->wq_lock[txq_map]); + + return NETDEV_TX_OK; +} + +/* dev_base_lock rwlock held, nominally process context */ +static void enic_get_stats(struct net_device *netdev, + struct rtnl_link_stats64 *net_stats) +{ + struct enic *enic = netdev_priv(netdev); + struct vnic_stats *stats; + int err; + + err = enic_dev_stats_dump(enic, &stats); + /* return only when pci_zalloc_consistent fails in vnic_dev_stats_dump + * For other failures, like devcmd failure, we return previously + * recorded stats. + */ + if (err == -ENOMEM) + return; + + net_stats->tx_packets = stats->tx.tx_frames_ok; + net_stats->tx_bytes = stats->tx.tx_bytes_ok; + net_stats->tx_errors = stats->tx.tx_errors; + net_stats->tx_dropped = stats->tx.tx_drops; + + net_stats->rx_packets = stats->rx.rx_frames_ok; + net_stats->rx_bytes = stats->rx.rx_bytes_ok; + net_stats->rx_errors = stats->rx.rx_errors; + net_stats->multicast = stats->rx.rx_multicast_frames_ok; + net_stats->rx_over_errors = enic->rq_truncated_pkts; + net_stats->rx_crc_errors = enic->rq_bad_fcs; + net_stats->rx_dropped = stats->rx.rx_no_bufs + stats->rx.rx_drop; +} + +static int enic_mc_sync(struct net_device *netdev, const u8 *mc_addr) +{ + struct enic *enic = netdev_priv(netdev); + + if (enic->mc_count == ENIC_MULTICAST_PERFECT_FILTERS) { + unsigned int mc_count = netdev_mc_count(netdev); + + netdev_warn(netdev, "Registering only %d out of %d multicast addresses\n", + ENIC_MULTICAST_PERFECT_FILTERS, mc_count); + + return -ENOSPC; + } + + enic_dev_add_addr(enic, mc_addr); + enic->mc_count++; + + return 0; +} + +static int enic_mc_unsync(struct net_device *netdev, const u8 *mc_addr) +{ + struct enic *enic = netdev_priv(netdev); + + enic_dev_del_addr(enic, mc_addr); + enic->mc_count--; + + return 0; +} + +static int enic_uc_sync(struct net_device *netdev, const u8 *uc_addr) +{ + struct enic *enic = netdev_priv(netdev); + + if (enic->uc_count == ENIC_UNICAST_PERFECT_FILTERS) { + unsigned int uc_count = netdev_uc_count(netdev); + + netdev_warn(netdev, "Registering only %d out of %d unicast addresses\n", + ENIC_UNICAST_PERFECT_FILTERS, uc_count); + + return -ENOSPC; + } + + enic_dev_add_addr(enic, uc_addr); + enic->uc_count++; + + return 0; +} + +static int enic_uc_unsync(struct net_device *netdev, const u8 *uc_addr) +{ + struct enic *enic = netdev_priv(netdev); + + enic_dev_del_addr(enic, uc_addr); + enic->uc_count--; + + return 0; +} + +void enic_reset_addr_lists(struct enic *enic) +{ + struct net_device *netdev = enic->netdev; + + __dev_uc_unsync(netdev, NULL); + __dev_mc_unsync(netdev, NULL); + + enic->mc_count = 0; + enic->uc_count = 0; + enic->flags = 0; +} + +static int enic_set_mac_addr(struct net_device *netdev, char *addr) +{ + struct enic *enic = netdev_priv(netdev); + + if (enic_is_dynamic(enic) || enic_is_sriov_vf(enic)) { + if (!is_valid_ether_addr(addr) && !is_zero_ether_addr(addr)) + return -EADDRNOTAVAIL; + } else { + if (!is_valid_ether_addr(addr)) + return -EADDRNOTAVAIL; + } + + memcpy(netdev->dev_addr, addr, netdev->addr_len); + + return 0; +} + +static int enic_set_mac_address_dynamic(struct net_device *netdev, void *p) +{ + struct enic *enic = netdev_priv(netdev); + struct sockaddr *saddr = p; + char *addr = saddr->sa_data; + int err; + + if (netif_running(enic->netdev)) { + err = enic_dev_del_station_addr(enic); + if (err) + return err; + } + + err = enic_set_mac_addr(netdev, addr); + if (err) + return err; + + if (netif_running(enic->netdev)) { + err = enic_dev_add_station_addr(enic); + if (err) + return err; + } + + return err; +} + +static int enic_set_mac_address(struct net_device *netdev, void *p) +{ + struct sockaddr *saddr = p; + char *addr = saddr->sa_data; + struct enic *enic = netdev_priv(netdev); + int err; + + err = enic_dev_del_station_addr(enic); + if (err) + return err; + + err = enic_set_mac_addr(netdev, addr); + if (err) + return err; + + return enic_dev_add_station_addr(enic); +} + +/* netif_tx_lock held, BHs disabled */ +static void enic_set_rx_mode(struct net_device *netdev) +{ + struct enic *enic = netdev_priv(netdev); + int directed = 1; + int multicast = (netdev->flags & IFF_MULTICAST) ? 1 : 0; + int broadcast = (netdev->flags & IFF_BROADCAST) ? 1 : 0; + int promisc = (netdev->flags & IFF_PROMISC) || + netdev_uc_count(netdev) > ENIC_UNICAST_PERFECT_FILTERS; + int allmulti = (netdev->flags & IFF_ALLMULTI) || + netdev_mc_count(netdev) > ENIC_MULTICAST_PERFECT_FILTERS; + unsigned int flags = netdev->flags | + (allmulti ? IFF_ALLMULTI : 0) | + (promisc ? IFF_PROMISC : 0); + + if (enic->flags != flags) { + enic->flags = flags; + enic_dev_packet_filter(enic, directed, + multicast, broadcast, promisc, allmulti); + } + + if (!promisc) { + __dev_uc_sync(netdev, enic_uc_sync, enic_uc_unsync); + if (!allmulti) + __dev_mc_sync(netdev, enic_mc_sync, enic_mc_unsync); + } +} + +/* netif_tx_lock held, BHs disabled */ +static void enic_tx_timeout(struct net_device *netdev) +{ + struct enic *enic = netdev_priv(netdev); + schedule_work(&enic->tx_hang_reset); +} + +static int enic_set_vf_mac(struct net_device *netdev, int vf, u8 *mac) +{ + struct enic *enic = netdev_priv(netdev); + struct enic_port_profile *pp; + int err; + + ENIC_PP_BY_INDEX(enic, vf, pp, &err); + if (err) + return err; + + if (is_valid_ether_addr(mac) || is_zero_ether_addr(mac)) { + if (vf == PORT_SELF_VF) { + memcpy(pp->vf_mac, mac, ETH_ALEN); + return 0; + } else { + /* + * For sriov vf's set the mac in hw + */ + ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, + vnic_dev_set_mac_addr, mac); + return enic_dev_status_to_errno(err); + } + } else + return -EINVAL; +} + +static int enic_set_vf_port(struct net_device *netdev, int vf, + struct nlattr *port[]) +{ + struct enic *enic = netdev_priv(netdev); + struct enic_port_profile prev_pp; + struct enic_port_profile *pp; + int err = 0, restore_pp = 1; + + ENIC_PP_BY_INDEX(enic, vf, pp, &err); + if (err) + return err; + + if (!port[IFLA_PORT_REQUEST]) + return -EOPNOTSUPP; + + memcpy(&prev_pp, pp, sizeof(*enic->pp)); + memset(pp, 0, sizeof(*enic->pp)); + + pp->set |= ENIC_SET_REQUEST; + pp->request = nla_get_u8(port[IFLA_PORT_REQUEST]); + + if (port[IFLA_PORT_PROFILE]) { + pp->set |= ENIC_SET_NAME; + memcpy(pp->name, nla_data(port[IFLA_PORT_PROFILE]), + PORT_PROFILE_MAX); + } + + if (port[IFLA_PORT_INSTANCE_UUID]) { + pp->set |= ENIC_SET_INSTANCE; + memcpy(pp->instance_uuid, + nla_data(port[IFLA_PORT_INSTANCE_UUID]), PORT_UUID_MAX); + } + + if (port[IFLA_PORT_HOST_UUID]) { + pp->set |= ENIC_SET_HOST; + memcpy(pp->host_uuid, + nla_data(port[IFLA_PORT_HOST_UUID]), PORT_UUID_MAX); + } + + if (vf == PORT_SELF_VF) { + /* Special case handling: mac came from IFLA_VF_MAC */ + if (!is_zero_ether_addr(prev_pp.vf_mac)) + memcpy(pp->mac_addr, prev_pp.vf_mac, ETH_ALEN); + + if (is_zero_ether_addr(netdev->dev_addr)) + eth_hw_addr_random(netdev); + } else { + /* SR-IOV VF: get mac from adapter */ + ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, + vnic_dev_get_mac_addr, pp->mac_addr); + if (err) { + netdev_err(netdev, "Error getting mac for vf %d\n", vf); + memcpy(pp, &prev_pp, sizeof(*pp)); + return enic_dev_status_to_errno(err); + } + } + + err = enic_process_set_pp_request(enic, vf, &prev_pp, &restore_pp); + if (err) { + if (restore_pp) { + /* Things are still the way they were: Implicit + * DISASSOCIATE failed + */ + memcpy(pp, &prev_pp, sizeof(*pp)); + } else { + memset(pp, 0, sizeof(*pp)); + if (vf == PORT_SELF_VF) + eth_zero_addr(netdev->dev_addr); + } + } else { + /* Set flag to indicate that the port assoc/disassoc + * request has been sent out to fw + */ + pp->set |= ENIC_PORT_REQUEST_APPLIED; + + /* If DISASSOCIATE, clean up all assigned/saved macaddresses */ + if (pp->request == PORT_REQUEST_DISASSOCIATE) { + eth_zero_addr(pp->mac_addr); + if (vf == PORT_SELF_VF) + eth_zero_addr(netdev->dev_addr); + } + } + + if (vf == PORT_SELF_VF) + eth_zero_addr(pp->vf_mac); + + return err; +} + +static int enic_get_vf_port(struct net_device *netdev, int vf, + struct sk_buff *skb) +{ + struct enic *enic = netdev_priv(netdev); + u16 response = PORT_PROFILE_RESPONSE_SUCCESS; + struct enic_port_profile *pp; + int err; + + ENIC_PP_BY_INDEX(enic, vf, pp, &err); + if (err) + return err; + + if (!(pp->set & ENIC_PORT_REQUEST_APPLIED)) + return -ENODATA; + + err = enic_process_get_pp_request(enic, vf, pp->request, &response); + if (err) + return err; + + if (nla_put_u16(skb, IFLA_PORT_REQUEST, pp->request) || + nla_put_u16(skb, IFLA_PORT_RESPONSE, response) || + ((pp->set & ENIC_SET_NAME) && + nla_put(skb, IFLA_PORT_PROFILE, PORT_PROFILE_MAX, pp->name)) || + ((pp->set & ENIC_SET_INSTANCE) && + nla_put(skb, IFLA_PORT_INSTANCE_UUID, PORT_UUID_MAX, + pp->instance_uuid)) || + ((pp->set & ENIC_SET_HOST) && + nla_put(skb, IFLA_PORT_HOST_UUID, PORT_UUID_MAX, pp->host_uuid))) + goto nla_put_failure; + return 0; + +nla_put_failure: + return -EMSGSIZE; +} + +static void enic_free_rq_buf(struct vnic_rq *rq, struct vnic_rq_buf *buf) +{ + struct enic *enic = vnic_dev_priv(rq->vdev); + + if (!buf->os_buf) + return; + + pci_unmap_single(enic->pdev, buf->dma_addr, + buf->len, PCI_DMA_FROMDEVICE); + dev_kfree_skb_any(buf->os_buf); + buf->os_buf = NULL; +} + +static int enic_rq_alloc_buf(struct vnic_rq *rq) +{ + struct enic *enic = vnic_dev_priv(rq->vdev); + struct net_device *netdev = enic->netdev; + struct sk_buff *skb; + unsigned int len = netdev->mtu + VLAN_ETH_HLEN; + unsigned int os_buf_index = 0; + dma_addr_t dma_addr; + struct vnic_rq_buf *buf = rq->to_use; + + if (buf->os_buf) { + enic_queue_rq_desc(rq, buf->os_buf, os_buf_index, buf->dma_addr, + buf->len); + + return 0; + } + skb = netdev_alloc_skb_ip_align(netdev, len); + if (!skb) + return -ENOMEM; + + dma_addr = pci_map_single(enic->pdev, skb->data, len, + PCI_DMA_FROMDEVICE); + if (unlikely(enic_dma_map_check(enic, dma_addr))) { + dev_kfree_skb(skb); + return -ENOMEM; + } + + enic_queue_rq_desc(rq, skb, os_buf_index, + dma_addr, len); + + return 0; +} + +static void enic_intr_update_pkt_size(struct vnic_rx_bytes_counter *pkt_size, + u32 pkt_len) +{ + if (ENIC_LARGE_PKT_THRESHOLD <= pkt_len) + pkt_size->large_pkt_bytes_cnt += pkt_len; + else + pkt_size->small_pkt_bytes_cnt += pkt_len; +} + +static bool enic_rxcopybreak(struct net_device *netdev, struct sk_buff **skb, + struct vnic_rq_buf *buf, u16 len) +{ + struct enic *enic = netdev_priv(netdev); + struct sk_buff *new_skb; + + if (len > enic->rx_copybreak) + return false; + new_skb = netdev_alloc_skb_ip_align(netdev, len); + if (!new_skb) + return false; + pci_dma_sync_single_for_cpu(enic->pdev, buf->dma_addr, len, + DMA_FROM_DEVICE); + memcpy(new_skb->data, (*skb)->data, len); + *skb = new_skb; + + return true; +} + +static void enic_rq_indicate_buf(struct vnic_rq *rq, + struct cq_desc *cq_desc, struct vnic_rq_buf *buf, + int skipped, void *opaque) +{ + struct enic *enic = vnic_dev_priv(rq->vdev); + struct net_device *netdev = enic->netdev; + struct sk_buff *skb; + struct vnic_cq *cq = &enic->cq[enic_cq_rq(enic, rq->index)]; + + u8 type, color, eop, sop, ingress_port, vlan_stripped; + u8 fcoe, fcoe_sof, fcoe_fc_crc_ok, fcoe_enc_error, fcoe_eof; + u8 tcp_udp_csum_ok, udp, tcp, ipv4_csum_ok; + u8 ipv6, ipv4, ipv4_fragment, fcs_ok, rss_type, csum_not_calc; + u8 packet_error; + u16 q_number, completed_index, bytes_written, vlan_tci, checksum; + u32 rss_hash; + bool outer_csum_ok = true, encap = false; + + if (skipped) + return; + + skb = buf->os_buf; + + cq_enet_rq_desc_dec((struct cq_enet_rq_desc *)cq_desc, + &type, &color, &q_number, &completed_index, + &ingress_port, &fcoe, &eop, &sop, &rss_type, + &csum_not_calc, &rss_hash, &bytes_written, + &packet_error, &vlan_stripped, &vlan_tci, &checksum, + &fcoe_sof, &fcoe_fc_crc_ok, &fcoe_enc_error, + &fcoe_eof, &tcp_udp_csum_ok, &udp, &tcp, + &ipv4_csum_ok, &ipv6, &ipv4, &ipv4_fragment, + &fcs_ok); + + if (packet_error) { + + if (!fcs_ok) { + if (bytes_written > 0) + enic->rq_bad_fcs++; + else if (bytes_written == 0) + enic->rq_truncated_pkts++; + } + + pci_unmap_single(enic->pdev, buf->dma_addr, buf->len, + PCI_DMA_FROMDEVICE); + dev_kfree_skb_any(skb); + buf->os_buf = NULL; + + return; + } + + if (eop && bytes_written > 0) { + + /* Good receive + */ + + if (!enic_rxcopybreak(netdev, &skb, buf, bytes_written)) { + buf->os_buf = NULL; + pci_unmap_single(enic->pdev, buf->dma_addr, buf->len, + PCI_DMA_FROMDEVICE); + } + prefetch(skb->data - NET_IP_ALIGN); + + skb_put(skb, bytes_written); + skb->protocol = eth_type_trans(skb, netdev); + skb_record_rx_queue(skb, q_number); + if ((netdev->features & NETIF_F_RXHASH) && rss_hash && + (type == 3)) { + switch (rss_type) { + case CQ_ENET_RQ_DESC_RSS_TYPE_TCP_IPv4: + case CQ_ENET_RQ_DESC_RSS_TYPE_TCP_IPv6: + case CQ_ENET_RQ_DESC_RSS_TYPE_TCP_IPv6_EX: + skb_set_hash(skb, rss_hash, PKT_HASH_TYPE_L4); + break; + case CQ_ENET_RQ_DESC_RSS_TYPE_IPv4: + case CQ_ENET_RQ_DESC_RSS_TYPE_IPv6: + case CQ_ENET_RQ_DESC_RSS_TYPE_IPv6_EX: + skb_set_hash(skb, rss_hash, PKT_HASH_TYPE_L3); + break; + } + } + if (enic->vxlan.vxlan_udp_port_number) { + switch (enic->vxlan.patch_level) { + case 0: + if (fcoe) { + encap = true; + outer_csum_ok = fcoe_fc_crc_ok; + } + break; + case 2: + if ((type == 7) && + (rss_hash & BIT(0))) { + encap = true; + outer_csum_ok = (rss_hash & BIT(1)) && + (rss_hash & BIT(2)); + } + break; + } + } + + /* Hardware does not provide whole packet checksum. It only + * provides pseudo checksum. Since hw validates the packet + * checksum but not provide us the checksum value. use + * CHECSUM_UNNECESSARY. + * + * In case of encap pkt tcp_udp_csum_ok/tcp_udp_csum_ok is + * inner csum_ok. outer_csum_ok is set by hw when outer udp + * csum is correct or is zero. + */ + if ((netdev->features & NETIF_F_RXCSUM) && !csum_not_calc && + tcp_udp_csum_ok && outer_csum_ok && + (ipv4_csum_ok || ipv6)) { + skb->ip_summed = CHECKSUM_UNNECESSARY; + skb->csum_level = encap; + } + + if (vlan_stripped) + __vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), vlan_tci); + + skb_mark_napi_id(skb, &enic->napi[rq->index]); + if (!(netdev->features & NETIF_F_GRO)) + netif_receive_skb(skb); + else + napi_gro_receive(&enic->napi[q_number], skb); + if (enic->rx_coalesce_setting.use_adaptive_rx_coalesce) + enic_intr_update_pkt_size(&cq->pkt_size_counter, + bytes_written); + } else { + + /* Buffer overflow + */ + + pci_unmap_single(enic->pdev, buf->dma_addr, buf->len, + PCI_DMA_FROMDEVICE); + dev_kfree_skb_any(skb); + buf->os_buf = NULL; + } +} + +static int enic_rq_service(struct vnic_dev *vdev, struct cq_desc *cq_desc, + u8 type, u16 q_number, u16 completed_index, void *opaque) +{ + struct enic *enic = vnic_dev_priv(vdev); + + vnic_rq_service(&enic->rq[q_number], cq_desc, + completed_index, VNIC_RQ_RETURN_DESC, + enic_rq_indicate_buf, opaque); + + return 0; +} + +static void enic_set_int_moderation(struct enic *enic, struct vnic_rq *rq) +{ + unsigned int intr = enic_msix_rq_intr(enic, rq->index); + struct vnic_cq *cq = &enic->cq[enic_cq_rq(enic, rq->index)]; + u32 timer = cq->tobe_rx_coal_timeval; + + if (cq->tobe_rx_coal_timeval != cq->cur_rx_coal_timeval) { + vnic_intr_coalescing_timer_set(&enic->intr[intr], timer); + cq->cur_rx_coal_timeval = cq->tobe_rx_coal_timeval; + } +} + +static void enic_calc_int_moderation(struct enic *enic, struct vnic_rq *rq) +{ + struct enic_rx_coal *rx_coal = &enic->rx_coalesce_setting; + struct vnic_cq *cq = &enic->cq[enic_cq_rq(enic, rq->index)]; + struct vnic_rx_bytes_counter *pkt_size_counter = &cq->pkt_size_counter; + int index; + u32 timer; + u32 range_start; + u32 traffic; + u64 delta; + ktime_t now = ktime_get(); + + delta = ktime_us_delta(now, cq->prev_ts); + if (delta < ENIC_AIC_TS_BREAK) + return; + cq->prev_ts = now; + + traffic = pkt_size_counter->large_pkt_bytes_cnt + + pkt_size_counter->small_pkt_bytes_cnt; + /* The table takes Mbps + * traffic *= 8 => bits + * traffic *= (10^6 / delta) => bps + * traffic /= 10^6 => Mbps + * + * Combining, traffic *= (8 / delta) + */ + + traffic <<= 3; + traffic = delta > UINT_MAX ? 0 : traffic / (u32)delta; + + for (index = 0; index < ENIC_MAX_COALESCE_TIMERS; index++) + if (traffic < mod_table[index].rx_rate) + break; + range_start = (pkt_size_counter->small_pkt_bytes_cnt > + pkt_size_counter->large_pkt_bytes_cnt << 1) ? + rx_coal->small_pkt_range_start : + rx_coal->large_pkt_range_start; + timer = range_start + ((rx_coal->range_end - range_start) * + mod_table[index].range_percent / 100); + /* Damping */ + cq->tobe_rx_coal_timeval = (timer + cq->tobe_rx_coal_timeval) >> 1; + + pkt_size_counter->large_pkt_bytes_cnt = 0; + pkt_size_counter->small_pkt_bytes_cnt = 0; +} + +static int enic_poll(struct napi_struct *napi, int budget) +{ + struct net_device *netdev = napi->dev; + struct enic *enic = netdev_priv(netdev); + unsigned int cq_rq = enic_cq_rq(enic, 0); + unsigned int cq_wq = enic_cq_wq(enic, 0); + unsigned int intr = enic_legacy_io_intr(); + unsigned int rq_work_to_do = budget; + unsigned int wq_work_to_do = ENIC_WQ_NAPI_BUDGET; + unsigned int work_done, rq_work_done = 0, wq_work_done; + int err; + + wq_work_done = vnic_cq_service(&enic->cq[cq_wq], wq_work_to_do, + enic_wq_service, NULL); + + if (budget > 0) + rq_work_done = vnic_cq_service(&enic->cq[cq_rq], + rq_work_to_do, enic_rq_service, NULL); + + /* Accumulate intr event credits for this polling + * cycle. An intr event is the completion of a + * a WQ or RQ packet. + */ + + work_done = rq_work_done + wq_work_done; + + if (work_done > 0) + vnic_intr_return_credits(&enic->intr[intr], + work_done, + 0 /* don't unmask intr */, + 0 /* don't reset intr timer */); + + err = vnic_rq_fill(&enic->rq[0], enic_rq_alloc_buf); + + /* Buffer allocation failed. Stay in polling + * mode so we can try to fill the ring again. + */ + + if (err) + rq_work_done = rq_work_to_do; + if (enic->rx_coalesce_setting.use_adaptive_rx_coalesce) + /* Call the function which refreshes the intr coalescing timer + * value based on the traffic. + */ + enic_calc_int_moderation(enic, &enic->rq[0]); + + if ((rq_work_done < budget) && napi_complete_done(napi, rq_work_done)) { + + /* Some work done, but not enough to stay in polling, + * exit polling + */ + + if (enic->rx_coalesce_setting.use_adaptive_rx_coalesce) + enic_set_int_moderation(enic, &enic->rq[0]); + vnic_intr_unmask(&enic->intr[intr]); + } + + return rq_work_done; +} + +#ifdef CONFIG_RFS_ACCEL +static void enic_free_rx_cpu_rmap(struct enic *enic) +{ + free_irq_cpu_rmap(enic->netdev->rx_cpu_rmap); + enic->netdev->rx_cpu_rmap = NULL; +} + +static void enic_set_rx_cpu_rmap(struct enic *enic) +{ + int i, res; + + if (vnic_dev_get_intr_mode(enic->vdev) == VNIC_DEV_INTR_MODE_MSIX) { + enic->netdev->rx_cpu_rmap = alloc_irq_cpu_rmap(enic->rq_count); + if (unlikely(!enic->netdev->rx_cpu_rmap)) + return; + for (i = 0; i < enic->rq_count; i++) { + res = irq_cpu_rmap_add(enic->netdev->rx_cpu_rmap, + enic->msix_entry[i].vector); + if (unlikely(res)) { + enic_free_rx_cpu_rmap(enic); + return; + } + } + } +} + +#else + +static void enic_free_rx_cpu_rmap(struct enic *enic) +{ +} + +static void enic_set_rx_cpu_rmap(struct enic *enic) +{ +} + +#endif /* CONFIG_RFS_ACCEL */ + +static int enic_poll_msix_wq(struct napi_struct *napi, int budget) +{ + struct net_device *netdev = napi->dev; + struct enic *enic = netdev_priv(netdev); + unsigned int wq_index = (napi - &enic->napi[0]) - enic->rq_count; + struct vnic_wq *wq = &enic->wq[wq_index]; + unsigned int cq; + unsigned int intr; + unsigned int wq_work_to_do = ENIC_WQ_NAPI_BUDGET; + unsigned int wq_work_done; + unsigned int wq_irq; + + wq_irq = wq->index; + cq = enic_cq_wq(enic, wq_irq); + intr = enic_msix_wq_intr(enic, wq_irq); + wq_work_done = vnic_cq_service(&enic->cq[cq], wq_work_to_do, + enic_wq_service, NULL); + + vnic_intr_return_credits(&enic->intr[intr], wq_work_done, + 0 /* don't unmask intr */, + 1 /* reset intr timer */); + if (!wq_work_done) { + napi_complete(napi); + vnic_intr_unmask(&enic->intr[intr]); + return 0; + } + + return budget; +} + +static int enic_poll_msix_rq(struct napi_struct *napi, int budget) +{ + struct net_device *netdev = napi->dev; + struct enic *enic = netdev_priv(netdev); + unsigned int rq = (napi - &enic->napi[0]); + unsigned int cq = enic_cq_rq(enic, rq); + unsigned int intr = enic_msix_rq_intr(enic, rq); + unsigned int work_to_do = budget; + unsigned int work_done = 0; + int err; + + /* Service RQ + */ + + if (budget > 0) + work_done = vnic_cq_service(&enic->cq[cq], + work_to_do, enic_rq_service, NULL); + + /* Return intr event credits for this polling + * cycle. An intr event is the completion of a + * RQ packet. + */ + + if (work_done > 0) + vnic_intr_return_credits(&enic->intr[intr], + work_done, + 0 /* don't unmask intr */, + 0 /* don't reset intr timer */); + + err = vnic_rq_fill(&enic->rq[rq], enic_rq_alloc_buf); + + /* Buffer allocation failed. Stay in polling mode + * so we can try to fill the ring again. + */ + + if (err) + work_done = work_to_do; + if (enic->rx_coalesce_setting.use_adaptive_rx_coalesce) + /* Call the function which refreshes the intr coalescing timer + * value based on the traffic. + */ + enic_calc_int_moderation(enic, &enic->rq[rq]); + + if ((work_done < budget) && napi_complete_done(napi, work_done)) { + + /* Some work done, but not enough to stay in polling, + * exit polling + */ + + if (enic->rx_coalesce_setting.use_adaptive_rx_coalesce) + enic_set_int_moderation(enic, &enic->rq[rq]); + vnic_intr_unmask(&enic->intr[intr]); + } + + return work_done; +} + +static void enic_notify_timer(struct timer_list *t) +{ + struct enic *enic = from_timer(enic, t, notify_timer); + + enic_notify_check(enic); + + mod_timer(&enic->notify_timer, + round_jiffies(jiffies + ENIC_NOTIFY_TIMER_PERIOD)); +} + +static void enic_free_intr(struct enic *enic) +{ + struct net_device *netdev = enic->netdev; + unsigned int i; + + enic_free_rx_cpu_rmap(enic); + switch (vnic_dev_get_intr_mode(enic->vdev)) { + case VNIC_DEV_INTR_MODE_INTX: + free_irq(enic->pdev->irq, netdev); + break; + case VNIC_DEV_INTR_MODE_MSI: + free_irq(enic->pdev->irq, enic); + break; + case VNIC_DEV_INTR_MODE_MSIX: + for (i = 0; i < ARRAY_SIZE(enic->msix); i++) + if (enic->msix[i].requested) + free_irq(enic->msix_entry[i].vector, + enic->msix[i].devid); + break; + default: + break; + } +} + +static int enic_request_intr(struct enic *enic) +{ + struct net_device *netdev = enic->netdev; + unsigned int i, intr; + int err = 0; + + enic_set_rx_cpu_rmap(enic); + switch (vnic_dev_get_intr_mode(enic->vdev)) { + + case VNIC_DEV_INTR_MODE_INTX: + + err = request_irq(enic->pdev->irq, enic_isr_legacy, + IRQF_SHARED, netdev->name, netdev); + break; + + case VNIC_DEV_INTR_MODE_MSI: + + err = request_irq(enic->pdev->irq, enic_isr_msi, + 0, netdev->name, enic); + break; + + case VNIC_DEV_INTR_MODE_MSIX: + + for (i = 0; i < enic->rq_count; i++) { + intr = enic_msix_rq_intr(enic, i); + snprintf(enic->msix[intr].devname, + sizeof(enic->msix[intr].devname), + "%s-rx-%u", netdev->name, i); + enic->msix[intr].isr = enic_isr_msix; + enic->msix[intr].devid = &enic->napi[i]; + } + + for (i = 0; i < enic->wq_count; i++) { + int wq = enic_cq_wq(enic, i); + + intr = enic_msix_wq_intr(enic, i); + snprintf(enic->msix[intr].devname, + sizeof(enic->msix[intr].devname), + "%s-tx-%u", netdev->name, i); + enic->msix[intr].isr = enic_isr_msix; + enic->msix[intr].devid = &enic->napi[wq]; + } + + intr = enic_msix_err_intr(enic); + snprintf(enic->msix[intr].devname, + sizeof(enic->msix[intr].devname), + "%s-err", netdev->name); + enic->msix[intr].isr = enic_isr_msix_err; + enic->msix[intr].devid = enic; + + intr = enic_msix_notify_intr(enic); + snprintf(enic->msix[intr].devname, + sizeof(enic->msix[intr].devname), + "%s-notify", netdev->name); + enic->msix[intr].isr = enic_isr_msix_notify; + enic->msix[intr].devid = enic; + + for (i = 0; i < ARRAY_SIZE(enic->msix); i++) + enic->msix[i].requested = 0; + + for (i = 0; i < enic->intr_count; i++) { + err = request_irq(enic->msix_entry[i].vector, + enic->msix[i].isr, 0, + enic->msix[i].devname, + enic->msix[i].devid); + if (err) { + enic_free_intr(enic); + break; + } + enic->msix[i].requested = 1; + } + + break; + + default: + break; + } + + return err; +} + +static void enic_synchronize_irqs(struct enic *enic) +{ + unsigned int i; + + switch (vnic_dev_get_intr_mode(enic->vdev)) { + case VNIC_DEV_INTR_MODE_INTX: + case VNIC_DEV_INTR_MODE_MSI: + synchronize_irq(enic->pdev->irq); + break; + case VNIC_DEV_INTR_MODE_MSIX: + for (i = 0; i < enic->intr_count; i++) + synchronize_irq(enic->msix_entry[i].vector); + break; + default: + break; + } +} + +static void enic_set_rx_coal_setting(struct enic *enic) +{ + unsigned int speed; + int index = -1; + struct enic_rx_coal *rx_coal = &enic->rx_coalesce_setting; + + /* 1. Read the link speed from fw + * 2. Pick the default range for the speed + * 3. Update it in enic->rx_coalesce_setting + */ + speed = vnic_dev_port_speed(enic->vdev); + if (ENIC_LINK_SPEED_10G < speed) + index = ENIC_LINK_40G_INDEX; + else if (ENIC_LINK_SPEED_4G < speed) + index = ENIC_LINK_10G_INDEX; + else + index = ENIC_LINK_4G_INDEX; + + rx_coal->small_pkt_range_start = mod_range[index].small_pkt_range_start; + rx_coal->large_pkt_range_start = mod_range[index].large_pkt_range_start; + rx_coal->range_end = ENIC_RX_COALESCE_RANGE_END; + + /* Start with the value provided by UCSM */ + for (index = 0; index < enic->rq_count; index++) + enic->cq[index].cur_rx_coal_timeval = + enic->config.intr_timer_usec; + + rx_coal->use_adaptive_rx_coalesce = 1; +} + +static int enic_dev_notify_set(struct enic *enic) +{ + int err; + + spin_lock_bh(&enic->devcmd_lock); + switch (vnic_dev_get_intr_mode(enic->vdev)) { + case VNIC_DEV_INTR_MODE_INTX: + err = vnic_dev_notify_set(enic->vdev, + enic_legacy_notify_intr()); + break; + case VNIC_DEV_INTR_MODE_MSIX: + err = vnic_dev_notify_set(enic->vdev, + enic_msix_notify_intr(enic)); + break; + default: + err = vnic_dev_notify_set(enic->vdev, -1 /* no intr */); + break; + } + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +static void enic_notify_timer_start(struct enic *enic) +{ + switch (vnic_dev_get_intr_mode(enic->vdev)) { + case VNIC_DEV_INTR_MODE_MSI: + mod_timer(&enic->notify_timer, jiffies); + break; + default: + /* Using intr for notification for INTx/MSI-X */ + break; + } +} + +/* rtnl lock is held, process context */ +static int enic_open(struct net_device *netdev) +{ + struct enic *enic = netdev_priv(netdev); + unsigned int i; + int err, ret; + + err = enic_request_intr(enic); + if (err) { + netdev_err(netdev, "Unable to request irq.\n"); + return err; + } + enic_init_affinity_hint(enic); + enic_set_affinity_hint(enic); + + err = enic_dev_notify_set(enic); + if (err) { + netdev_err(netdev, + "Failed to alloc notify buffer, aborting.\n"); + goto err_out_free_intr; + } + + for (i = 0; i < enic->rq_count; i++) { + /* enable rq before updating rq desc */ + vnic_rq_enable(&enic->rq[i]); + vnic_rq_fill(&enic->rq[i], enic_rq_alloc_buf); + /* Need at least one buffer on ring to get going */ + if (vnic_rq_desc_used(&enic->rq[i]) == 0) { + netdev_err(netdev, "Unable to alloc receive buffers\n"); + err = -ENOMEM; + goto err_out_free_rq; + } + } + + for (i = 0; i < enic->wq_count; i++) + vnic_wq_enable(&enic->wq[i]); + + if (!enic_is_dynamic(enic) && !enic_is_sriov_vf(enic)) + enic_dev_add_station_addr(enic); + + enic_set_rx_mode(netdev); + + netif_tx_wake_all_queues(netdev); + + for (i = 0; i < enic->rq_count; i++) + napi_enable(&enic->napi[i]); + + if (vnic_dev_get_intr_mode(enic->vdev) == VNIC_DEV_INTR_MODE_MSIX) + for (i = 0; i < enic->wq_count; i++) + napi_enable(&enic->napi[enic_cq_wq(enic, i)]); + enic_dev_enable(enic); + + for (i = 0; i < enic->intr_count; i++) + vnic_intr_unmask(&enic->intr[i]); + + enic_notify_timer_start(enic); + enic_rfs_timer_start(enic); + + return 0; + +err_out_free_rq: + for (i = 0; i < enic->rq_count; i++) { + ret = vnic_rq_disable(&enic->rq[i]); + if (!ret) + vnic_rq_clean(&enic->rq[i], enic_free_rq_buf); + } + enic_dev_notify_unset(enic); +err_out_free_intr: + enic_unset_affinity_hint(enic); + enic_free_intr(enic); + + return err; +} + +/* rtnl lock is held, process context */ +static int enic_stop(struct net_device *netdev) +{ + struct enic *enic = netdev_priv(netdev); + unsigned int i; + int err; + + for (i = 0; i < enic->intr_count; i++) { + vnic_intr_mask(&enic->intr[i]); + (void)vnic_intr_masked(&enic->intr[i]); /* flush write */ + } + + enic_synchronize_irqs(enic); + + del_timer_sync(&enic->notify_timer); + enic_rfs_flw_tbl_free(enic); + + enic_dev_disable(enic); + + for (i = 0; i < enic->rq_count; i++) + napi_disable(&enic->napi[i]); + + netif_carrier_off(netdev); + if (vnic_dev_get_intr_mode(enic->vdev) == VNIC_DEV_INTR_MODE_MSIX) + for (i = 0; i < enic->wq_count; i++) + napi_disable(&enic->napi[enic_cq_wq(enic, i)]); + netif_tx_disable(netdev); + + if (!enic_is_dynamic(enic) && !enic_is_sriov_vf(enic)) + enic_dev_del_station_addr(enic); + + for (i = 0; i < enic->wq_count; i++) { + err = vnic_wq_disable(&enic->wq[i]); + if (err) + return err; + } + for (i = 0; i < enic->rq_count; i++) { + err = vnic_rq_disable(&enic->rq[i]); + if (err) + return err; + } + + enic_dev_notify_unset(enic); + enic_unset_affinity_hint(enic); + enic_free_intr(enic); + + for (i = 0; i < enic->wq_count; i++) + vnic_wq_clean(&enic->wq[i], enic_free_wq_buf); + for (i = 0; i < enic->rq_count; i++) + vnic_rq_clean(&enic->rq[i], enic_free_rq_buf); + for (i = 0; i < enic->cq_count; i++) + vnic_cq_clean(&enic->cq[i]); + for (i = 0; i < enic->intr_count; i++) + vnic_intr_clean(&enic->intr[i]); + + return 0; +} + +static int _enic_change_mtu(struct net_device *netdev, int new_mtu) +{ + bool running = netif_running(netdev); + int err = 0; + + ASSERT_RTNL(); + if (running) { + err = enic_stop(netdev); + if (err) + return err; + } + + netdev->mtu = new_mtu; + + if (running) { + err = enic_open(netdev); + if (err) + return err; + } + + return 0; +} + +static int enic_change_mtu(struct net_device *netdev, int new_mtu) +{ + struct enic *enic = netdev_priv(netdev); + + if (enic_is_dynamic(enic) || enic_is_sriov_vf(enic)) + return -EOPNOTSUPP; + + if (netdev->mtu > enic->port_mtu) + netdev_warn(netdev, + "interface MTU (%d) set higher than port MTU (%d)\n", + netdev->mtu, enic->port_mtu); + + return _enic_change_mtu(netdev, new_mtu); +} + +static void enic_change_mtu_work(struct work_struct *work) +{ + struct enic *enic = container_of(work, struct enic, change_mtu_work); + struct net_device *netdev = enic->netdev; + int new_mtu = vnic_dev_mtu(enic->vdev); + + rtnl_lock(); + (void)_enic_change_mtu(netdev, new_mtu); + rtnl_unlock(); + + netdev_info(netdev, "interface MTU set as %d\n", netdev->mtu); +} + +#ifdef CONFIG_NET_POLL_CONTROLLER +static void enic_poll_controller(struct net_device *netdev) +{ + struct enic *enic = netdev_priv(netdev); + struct vnic_dev *vdev = enic->vdev; + unsigned int i, intr; + + switch (vnic_dev_get_intr_mode(vdev)) { + case VNIC_DEV_INTR_MODE_MSIX: + for (i = 0; i < enic->rq_count; i++) { + intr = enic_msix_rq_intr(enic, i); + enic_isr_msix(enic->msix_entry[intr].vector, + &enic->napi[i]); + } + + for (i = 0; i < enic->wq_count; i++) { + intr = enic_msix_wq_intr(enic, i); + enic_isr_msix(enic->msix_entry[intr].vector, + &enic->napi[enic_cq_wq(enic, i)]); + } + + break; + case VNIC_DEV_INTR_MODE_MSI: + enic_isr_msi(enic->pdev->irq, enic); + break; + case VNIC_DEV_INTR_MODE_INTX: + enic_isr_legacy(enic->pdev->irq, netdev); + break; + default: + break; + } +} +#endif + +static int enic_dev_wait(struct vnic_dev *vdev, + int (*start)(struct vnic_dev *, int), + int (*finished)(struct vnic_dev *, int *), + int arg) +{ + unsigned long time; + int done; + int err; + + err = start(vdev, arg); + if (err) + return err; + + /* Wait for func to complete...2 seconds max + */ + + time = jiffies + (HZ * 2); + do { + + err = finished(vdev, &done); + if (err) + return err; + + if (done) + return 0; + + schedule_timeout_uninterruptible(HZ / 10); + + } while (time_after(time, jiffies)); + + return -ETIMEDOUT; +} + +static int enic_dev_open(struct enic *enic) +{ + int err; + u32 flags = CMD_OPENF_IG_DESCCACHE; + + err = enic_dev_wait(enic->vdev, vnic_dev_open, + vnic_dev_open_done, flags); + if (err) + dev_err(enic_get_dev(enic), "vNIC device open failed, err %d\n", + err); + + return err; +} + +static int enic_dev_soft_reset(struct enic *enic) +{ + int err; + + err = enic_dev_wait(enic->vdev, vnic_dev_soft_reset, + vnic_dev_soft_reset_done, 0); + if (err) + netdev_err(enic->netdev, "vNIC soft reset failed, err %d\n", + err); + + return err; +} + +static int enic_dev_hang_reset(struct enic *enic) +{ + int err; + + err = enic_dev_wait(enic->vdev, vnic_dev_hang_reset, + vnic_dev_hang_reset_done, 0); + if (err) + netdev_err(enic->netdev, "vNIC hang reset failed, err %d\n", + err); + + return err; +} + +int __enic_set_rsskey(struct enic *enic) +{ + union vnic_rss_key *rss_key_buf_va; + dma_addr_t rss_key_buf_pa; + int i, kidx, bidx, err; + + rss_key_buf_va = pci_zalloc_consistent(enic->pdev, + sizeof(union vnic_rss_key), + &rss_key_buf_pa); + if (!rss_key_buf_va) + return -ENOMEM; + + for (i = 0; i < ENIC_RSS_LEN; i++) { + kidx = i / ENIC_RSS_BYTES_PER_KEY; + bidx = i % ENIC_RSS_BYTES_PER_KEY; + rss_key_buf_va->key[kidx].b[bidx] = enic->rss_key[i]; + } + spin_lock_bh(&enic->devcmd_lock); + err = enic_set_rss_key(enic, + rss_key_buf_pa, + sizeof(union vnic_rss_key)); + spin_unlock_bh(&enic->devcmd_lock); + + pci_free_consistent(enic->pdev, sizeof(union vnic_rss_key), + rss_key_buf_va, rss_key_buf_pa); + + return err; +} + +static int enic_set_rsskey(struct enic *enic) +{ + netdev_rss_key_fill(enic->rss_key, ENIC_RSS_LEN); + + return __enic_set_rsskey(enic); +} + +static int enic_set_rsscpu(struct enic *enic, u8 rss_hash_bits) +{ + dma_addr_t rss_cpu_buf_pa; + union vnic_rss_cpu *rss_cpu_buf_va = NULL; + unsigned int i; + int err; + + rss_cpu_buf_va = pci_alloc_consistent(enic->pdev, + sizeof(union vnic_rss_cpu), &rss_cpu_buf_pa); + if (!rss_cpu_buf_va) + return -ENOMEM; + + for (i = 0; i < (1 << rss_hash_bits); i++) + (*rss_cpu_buf_va).cpu[i/4].b[i%4] = i % enic->rq_count; + + spin_lock_bh(&enic->devcmd_lock); + err = enic_set_rss_cpu(enic, + rss_cpu_buf_pa, + sizeof(union vnic_rss_cpu)); + spin_unlock_bh(&enic->devcmd_lock); + + pci_free_consistent(enic->pdev, sizeof(union vnic_rss_cpu), + rss_cpu_buf_va, rss_cpu_buf_pa); + + return err; +} + +static int enic_set_niccfg(struct enic *enic, u8 rss_default_cpu, + u8 rss_hash_type, u8 rss_hash_bits, u8 rss_base_cpu, u8 rss_enable) +{ + const u8 tso_ipid_split_en = 0; + const u8 ig_vlan_strip_en = 1; + int err; + + /* Enable VLAN tag stripping. + */ + + spin_lock_bh(&enic->devcmd_lock); + err = enic_set_nic_cfg(enic, + rss_default_cpu, rss_hash_type, + rss_hash_bits, rss_base_cpu, + rss_enable, tso_ipid_split_en, + ig_vlan_strip_en); + spin_unlock_bh(&enic->devcmd_lock); + + return err; +} + +static int enic_set_rss_nic_cfg(struct enic *enic) +{ + struct device *dev = enic_get_dev(enic); + const u8 rss_default_cpu = 0; + const u8 rss_hash_bits = 7; + const u8 rss_base_cpu = 0; + u8 rss_hash_type; + int res; + u8 rss_enable = ENIC_SETTING(enic, RSS) && (enic->rq_count > 1); + + spin_lock_bh(&enic->devcmd_lock); + res = vnic_dev_capable_rss_hash_type(enic->vdev, &rss_hash_type); + spin_unlock_bh(&enic->devcmd_lock); + if (res) { + /* defaults for old adapters + */ + rss_hash_type = NIC_CFG_RSS_HASH_TYPE_IPV4 | + NIC_CFG_RSS_HASH_TYPE_TCP_IPV4 | + NIC_CFG_RSS_HASH_TYPE_IPV6 | + NIC_CFG_RSS_HASH_TYPE_TCP_IPV6; + } + + if (rss_enable) { + if (!enic_set_rsskey(enic)) { + if (enic_set_rsscpu(enic, rss_hash_bits)) { + rss_enable = 0; + dev_warn(dev, "RSS disabled, " + "Failed to set RSS cpu indirection table."); + } + } else { + rss_enable = 0; + dev_warn(dev, "RSS disabled, Failed to set RSS key.\n"); + } + } + + return enic_set_niccfg(enic, rss_default_cpu, rss_hash_type, + rss_hash_bits, rss_base_cpu, rss_enable); +} + +static void enic_set_api_busy(struct enic *enic, bool busy) +{ + spin_lock(&enic->enic_api_lock); + enic->enic_api_busy = busy; + spin_unlock(&enic->enic_api_lock); +} + +static void enic_reset(struct work_struct *work) +{ + struct enic *enic = container_of(work, struct enic, reset); + + if (!netif_running(enic->netdev)) + return; + + rtnl_lock(); + + /* Stop any activity from infiniband */ + enic_set_api_busy(enic, true); + + enic_stop(enic->netdev); + enic_dev_soft_reset(enic); + enic_reset_addr_lists(enic); + enic_init_vnic_resources(enic); + enic_set_rss_nic_cfg(enic); + enic_dev_set_ig_vlan_rewrite_mode(enic); + enic_open(enic->netdev); + + /* Allow infiniband to fiddle with the device again */ + enic_set_api_busy(enic, false); + + call_netdevice_notifiers(NETDEV_REBOOT, enic->netdev); + + rtnl_unlock(); +} + +static void enic_tx_hang_reset(struct work_struct *work) +{ + struct enic *enic = container_of(work, struct enic, tx_hang_reset); + + rtnl_lock(); + + /* Stop any activity from infiniband */ + enic_set_api_busy(enic, true); + + enic_dev_hang_notify(enic); + enic_stop(enic->netdev); + enic_dev_hang_reset(enic); + enic_reset_addr_lists(enic); + enic_init_vnic_resources(enic); + enic_set_rss_nic_cfg(enic); + enic_dev_set_ig_vlan_rewrite_mode(enic); + enic_open(enic->netdev); + + /* Allow infiniband to fiddle with the device again */ + enic_set_api_busy(enic, false); + + call_netdevice_notifiers(NETDEV_REBOOT, enic->netdev); + + rtnl_unlock(); +} + +static int enic_set_intr_mode(struct enic *enic) +{ + unsigned int n = min_t(unsigned int, enic->rq_count, ENIC_RQ_MAX); + unsigned int m = min_t(unsigned int, enic->wq_count, ENIC_WQ_MAX); + unsigned int i; + + /* Set interrupt mode (INTx, MSI, MSI-X) depending + * on system capabilities. + * + * Try MSI-X first + * + * We need n RQs, m WQs, n+m CQs, and n+m+2 INTRs + * (the second to last INTR is used for WQ/RQ errors) + * (the last INTR is used for notifications) + */ + + BUG_ON(ARRAY_SIZE(enic->msix_entry) < n + m + 2); + for (i = 0; i < n + m + 2; i++) + enic->msix_entry[i].entry = i; + + /* Use multiple RQs if RSS is enabled + */ + + if (ENIC_SETTING(enic, RSS) && + enic->config.intr_mode < 1 && + enic->rq_count >= n && + enic->wq_count >= m && + enic->cq_count >= n + m && + enic->intr_count >= n + m + 2) { + + if (pci_enable_msix_range(enic->pdev, enic->msix_entry, + n + m + 2, n + m + 2) > 0) { + + enic->rq_count = n; + enic->wq_count = m; + enic->cq_count = n + m; + enic->intr_count = n + m + 2; + + vnic_dev_set_intr_mode(enic->vdev, + VNIC_DEV_INTR_MODE_MSIX); + + return 0; + } + } + + if (enic->config.intr_mode < 1 && + enic->rq_count >= 1 && + enic->wq_count >= m && + enic->cq_count >= 1 + m && + enic->intr_count >= 1 + m + 2) { + if (pci_enable_msix_range(enic->pdev, enic->msix_entry, + 1 + m + 2, 1 + m + 2) > 0) { + + enic->rq_count = 1; + enic->wq_count = m; + enic->cq_count = 1 + m; + enic->intr_count = 1 + m + 2; + + vnic_dev_set_intr_mode(enic->vdev, + VNIC_DEV_INTR_MODE_MSIX); + + return 0; + } + } + + /* Next try MSI + * + * We need 1 RQ, 1 WQ, 2 CQs, and 1 INTR + */ + + if (enic->config.intr_mode < 2 && + enic->rq_count >= 1 && + enic->wq_count >= 1 && + enic->cq_count >= 2 && + enic->intr_count >= 1 && + !pci_enable_msi(enic->pdev)) { + + enic->rq_count = 1; + enic->wq_count = 1; + enic->cq_count = 2; + enic->intr_count = 1; + + vnic_dev_set_intr_mode(enic->vdev, VNIC_DEV_INTR_MODE_MSI); + + return 0; + } + + /* Next try INTx + * + * We need 1 RQ, 1 WQ, 2 CQs, and 3 INTRs + * (the first INTR is used for WQ/RQ) + * (the second INTR is used for WQ/RQ errors) + * (the last INTR is used for notifications) + */ + + if (enic->config.intr_mode < 3 && + enic->rq_count >= 1 && + enic->wq_count >= 1 && + enic->cq_count >= 2 && + enic->intr_count >= 3) { + + enic->rq_count = 1; + enic->wq_count = 1; + enic->cq_count = 2; + enic->intr_count = 3; + + vnic_dev_set_intr_mode(enic->vdev, VNIC_DEV_INTR_MODE_INTX); + + return 0; + } + + vnic_dev_set_intr_mode(enic->vdev, VNIC_DEV_INTR_MODE_UNKNOWN); + + return -EINVAL; +} + +static void enic_clear_intr_mode(struct enic *enic) +{ + switch (vnic_dev_get_intr_mode(enic->vdev)) { + case VNIC_DEV_INTR_MODE_MSIX: + pci_disable_msix(enic->pdev); + break; + case VNIC_DEV_INTR_MODE_MSI: + pci_disable_msi(enic->pdev); + break; + default: + break; + } + + vnic_dev_set_intr_mode(enic->vdev, VNIC_DEV_INTR_MODE_UNKNOWN); +} + +static const struct net_device_ops enic_netdev_dynamic_ops = { + .ndo_open = enic_open, + .ndo_stop = enic_stop, + .ndo_start_xmit = enic_hard_start_xmit, + .ndo_get_stats64 = enic_get_stats, + .ndo_validate_addr = eth_validate_addr, + .ndo_set_rx_mode = enic_set_rx_mode, + .ndo_set_mac_address = enic_set_mac_address_dynamic, + .ndo_change_mtu = enic_change_mtu, + .ndo_vlan_rx_add_vid = enic_vlan_rx_add_vid, + .ndo_vlan_rx_kill_vid = enic_vlan_rx_kill_vid, + .ndo_tx_timeout = enic_tx_timeout, + .ndo_set_vf_port = enic_set_vf_port, + .ndo_get_vf_port = enic_get_vf_port, + .ndo_set_vf_mac = enic_set_vf_mac, +#ifdef CONFIG_NET_POLL_CONTROLLER + .ndo_poll_controller = enic_poll_controller, +#endif +#ifdef CONFIG_RFS_ACCEL + .ndo_rx_flow_steer = enic_rx_flow_steer, +#endif + .ndo_udp_tunnel_add = enic_udp_tunnel_add, + .ndo_udp_tunnel_del = enic_udp_tunnel_del, + .ndo_features_check = enic_features_check, +}; + +static const struct net_device_ops enic_netdev_ops = { + .ndo_open = enic_open, + .ndo_stop = enic_stop, + .ndo_start_xmit = enic_hard_start_xmit, + .ndo_get_stats64 = enic_get_stats, + .ndo_validate_addr = eth_validate_addr, + .ndo_set_mac_address = enic_set_mac_address, + .ndo_set_rx_mode = enic_set_rx_mode, + .ndo_change_mtu = enic_change_mtu, + .ndo_vlan_rx_add_vid = enic_vlan_rx_add_vid, + .ndo_vlan_rx_kill_vid = enic_vlan_rx_kill_vid, + .ndo_tx_timeout = enic_tx_timeout, + .ndo_set_vf_port = enic_set_vf_port, + .ndo_get_vf_port = enic_get_vf_port, + .ndo_set_vf_mac = enic_set_vf_mac, +#ifdef CONFIG_NET_POLL_CONTROLLER + .ndo_poll_controller = enic_poll_controller, +#endif +#ifdef CONFIG_RFS_ACCEL + .ndo_rx_flow_steer = enic_rx_flow_steer, +#endif + .ndo_udp_tunnel_add = enic_udp_tunnel_add, + .ndo_udp_tunnel_del = enic_udp_tunnel_del, + .ndo_features_check = enic_features_check, +}; + +static void enic_dev_deinit(struct enic *enic) +{ + unsigned int i; + + for (i = 0; i < enic->rq_count; i++) { + napi_hash_del(&enic->napi[i]); + netif_napi_del(&enic->napi[i]); + } + if (vnic_dev_get_intr_mode(enic->vdev) == VNIC_DEV_INTR_MODE_MSIX) + for (i = 0; i < enic->wq_count; i++) + netif_napi_del(&enic->napi[enic_cq_wq(enic, i)]); + + enic_free_vnic_resources(enic); + enic_clear_intr_mode(enic); + enic_free_affinity_hint(enic); +} + +static void enic_kdump_kernel_config(struct enic *enic) +{ + if (is_kdump_kernel()) { + dev_info(enic_get_dev(enic), "Running from within kdump kernel. Using minimal resources\n"); + enic->rq_count = 1; + enic->wq_count = 1; + enic->config.rq_desc_count = ENIC_MIN_RQ_DESCS; + enic->config.wq_desc_count = ENIC_MIN_WQ_DESCS; + enic->config.mtu = min_t(u16, 1500, enic->config.mtu); + } +} + +static int enic_dev_init(struct enic *enic) +{ + struct device *dev = enic_get_dev(enic); + struct net_device *netdev = enic->netdev; + unsigned int i; + int err; + + /* Get interrupt coalesce timer info */ + err = enic_dev_intr_coal_timer_info(enic); + if (err) { + dev_warn(dev, "Using default conversion factor for " + "interrupt coalesce timer\n"); + vnic_dev_intr_coal_timer_info_default(enic->vdev); + } + + /* Get vNIC configuration + */ + + err = enic_get_vnic_config(enic); + if (err) { + dev_err(dev, "Get vNIC configuration failed, aborting\n"); + return err; + } + + /* Get available resource counts + */ + + enic_get_res_counts(enic); + + /* modify resource count if we are in kdump_kernel + */ + enic_kdump_kernel_config(enic); + + /* Set interrupt mode based on resource counts and system + * capabilities + */ + + err = enic_set_intr_mode(enic); + if (err) { + dev_err(dev, "Failed to set intr mode based on resource " + "counts and system capabilities, aborting\n"); + return err; + } + + /* Allocate and configure vNIC resources + */ + + err = enic_alloc_vnic_resources(enic); + if (err) { + dev_err(dev, "Failed to alloc vNIC resources, aborting\n"); + goto err_out_free_vnic_resources; + } + + enic_init_vnic_resources(enic); + + err = enic_set_rss_nic_cfg(enic); + if (err) { + dev_err(dev, "Failed to config nic, aborting\n"); + goto err_out_free_vnic_resources; + } + + switch (vnic_dev_get_intr_mode(enic->vdev)) { + default: + netif_napi_add(netdev, &enic->napi[0], enic_poll, 64); + break; + case VNIC_DEV_INTR_MODE_MSIX: + for (i = 0; i < enic->rq_count; i++) { + netif_napi_add(netdev, &enic->napi[i], + enic_poll_msix_rq, NAPI_POLL_WEIGHT); + } + for (i = 0; i < enic->wq_count; i++) + netif_napi_add(netdev, &enic->napi[enic_cq_wq(enic, i)], + enic_poll_msix_wq, NAPI_POLL_WEIGHT); + break; + } + + return 0; + +err_out_free_vnic_resources: + enic_free_affinity_hint(enic); + enic_clear_intr_mode(enic); + enic_free_vnic_resources(enic); + + return err; +} + +static void enic_iounmap(struct enic *enic) +{ + unsigned int i; + + for (i = 0; i < ARRAY_SIZE(enic->bar); i++) + if (enic->bar[i].vaddr) + iounmap(enic->bar[i].vaddr); +} + +static int enic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) +{ + struct device *dev = &pdev->dev; + struct net_device *netdev; + struct enic *enic; + int using_dac = 0; + unsigned int i; + int err; +#ifdef CONFIG_PCI_IOV + int pos = 0; +#endif + int num_pps = 1; + + /* Allocate net device structure and initialize. Private + * instance data is initialized to zero. + */ + + netdev = alloc_etherdev_mqs(sizeof(struct enic), + ENIC_RQ_MAX, ENIC_WQ_MAX); + if (!netdev) + return -ENOMEM; + + pci_set_drvdata(pdev, netdev); + + SET_NETDEV_DEV(netdev, &pdev->dev); + + enic = netdev_priv(netdev); + enic->netdev = netdev; + enic->pdev = pdev; + + /* Setup PCI resources + */ + + err = pci_enable_device_mem(pdev); + if (err) { + dev_err(dev, "Cannot enable PCI device, aborting\n"); + goto err_out_free_netdev; + } + + err = pci_request_regions(pdev, DRV_NAME); + if (err) { + dev_err(dev, "Cannot request PCI regions, aborting\n"); + goto err_out_disable_device; + } + + pci_set_master(pdev); + + /* Query PCI controller on system for DMA addressing + * limitation for the device. Try 47-bit first, and + * fail to 32-bit. + */ + + err = pci_set_dma_mask(pdev, DMA_BIT_MASK(47)); + if (err) { + err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32)); + if (err) { + dev_err(dev, "No usable DMA configuration, aborting\n"); + goto err_out_release_regions; + } + err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32)); + if (err) { + dev_err(dev, "Unable to obtain %u-bit DMA " + "for consistent allocations, aborting\n", 32); + goto err_out_release_regions; + } + } else { + err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(47)); + if (err) { + dev_err(dev, "Unable to obtain %u-bit DMA " + "for consistent allocations, aborting\n", 47); + goto err_out_release_regions; + } + using_dac = 1; + } + + /* Map vNIC resources from BAR0-5 + */ + + for (i = 0; i < ARRAY_SIZE(enic->bar); i++) { + if (!(pci_resource_flags(pdev, i) & IORESOURCE_MEM)) + continue; + enic->bar[i].len = pci_resource_len(pdev, i); + enic->bar[i].vaddr = pci_iomap(pdev, i, enic->bar[i].len); + if (!enic->bar[i].vaddr) { + dev_err(dev, "Cannot memory-map BAR %d, aborting\n", i); + err = -ENODEV; + goto err_out_iounmap; + } + enic->bar[i].bus_addr = pci_resource_start(pdev, i); + } + + /* Register vNIC device + */ + + enic->vdev = vnic_dev_register(NULL, enic, pdev, enic->bar, + ARRAY_SIZE(enic->bar)); + if (!enic->vdev) { + dev_err(dev, "vNIC registration failed, aborting\n"); + err = -ENODEV; + goto err_out_iounmap; + } + + err = vnic_devcmd_init(enic->vdev); + + if (err) + goto err_out_vnic_unregister; + +#ifdef CONFIG_PCI_IOV + /* Get number of subvnics */ + pos = pci_find_ext_capability(pdev, PCI_EXT_CAP_ID_SRIOV); + if (pos) { + pci_read_config_word(pdev, pos + PCI_SRIOV_TOTAL_VF, + &enic->num_vfs); + if (enic->num_vfs) { + err = pci_enable_sriov(pdev, enic->num_vfs); + if (err) { + dev_err(dev, "SRIOV enable failed, aborting." + " pci_enable_sriov() returned %d\n", + err); + goto err_out_vnic_unregister; + } + enic->priv_flags |= ENIC_SRIOV_ENABLED; + num_pps = enic->num_vfs; + } + } +#endif + + /* Allocate structure for port profiles */ + enic->pp = kcalloc(num_pps, sizeof(*enic->pp), GFP_KERNEL); + if (!enic->pp) { + err = -ENOMEM; + goto err_out_disable_sriov_pp; + } + + /* Issue device open to get device in known state + */ + + err = enic_dev_open(enic); + if (err) { + dev_err(dev, "vNIC dev open failed, aborting\n"); + goto err_out_disable_sriov; + } + + /* Setup devcmd lock + */ + + spin_lock_init(&enic->devcmd_lock); + spin_lock_init(&enic->enic_api_lock); + + /* + * Set ingress vlan rewrite mode before vnic initialization + */ + + err = enic_dev_set_ig_vlan_rewrite_mode(enic); + if (err) { + dev_err(dev, + "Failed to set ingress vlan rewrite mode, aborting.\n"); + goto err_out_dev_close; + } + + /* Issue device init to initialize the vnic-to-switch link. + * We'll start with carrier off and wait for link UP + * notification later to turn on carrier. We don't need + * to wait here for the vnic-to-switch link initialization + * to complete; link UP notification is the indication that + * the process is complete. + */ + + netif_carrier_off(netdev); + + /* Do not call dev_init for a dynamic vnic. + * For a dynamic vnic, init_prov_info will be + * called later by an upper layer. + */ + + if (!enic_is_dynamic(enic)) { + err = vnic_dev_init(enic->vdev, 0); + if (err) { + dev_err(dev, "vNIC dev init failed, aborting\n"); + goto err_out_dev_close; + } + } + + err = enic_dev_init(enic); + if (err) { + dev_err(dev, "Device initialization failed, aborting\n"); + goto err_out_dev_close; + } + + netif_set_real_num_tx_queues(netdev, enic->wq_count); + netif_set_real_num_rx_queues(netdev, enic->rq_count); + + /* Setup notification timer, HW reset task, and wq locks + */ + + timer_setup(&enic->notify_timer, enic_notify_timer, 0); + + enic_rfs_flw_tbl_init(enic); + enic_set_rx_coal_setting(enic); + INIT_WORK(&enic->reset, enic_reset); + INIT_WORK(&enic->tx_hang_reset, enic_tx_hang_reset); + INIT_WORK(&enic->change_mtu_work, enic_change_mtu_work); + + for (i = 0; i < enic->wq_count; i++) + spin_lock_init(&enic->wq_lock[i]); + + /* Register net device + */ + + enic->port_mtu = enic->config.mtu; + + err = enic_set_mac_addr(netdev, enic->mac_addr); + if (err) { + dev_err(dev, "Invalid MAC address, aborting\n"); + goto err_out_dev_deinit; + } + + enic->tx_coalesce_usecs = enic->config.intr_timer_usec; + /* rx coalesce time already got initialized. This gets used + * if adaptive coal is turned off + */ + enic->rx_coalesce_usecs = enic->tx_coalesce_usecs; + + if (enic_is_dynamic(enic) || enic_is_sriov_vf(enic)) + netdev->netdev_ops = &enic_netdev_dynamic_ops; + else + netdev->netdev_ops = &enic_netdev_ops; + + netdev->watchdog_timeo = 2 * HZ; + enic_set_ethtool_ops(netdev); + + netdev->features |= NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_CTAG_RX; + if (ENIC_SETTING(enic, LOOP)) { + netdev->features &= ~NETIF_F_HW_VLAN_CTAG_TX; + enic->loop_enable = 1; + enic->loop_tag = enic->config.loop_tag; + dev_info(dev, "loopback tag=0x%04x\n", enic->loop_tag); + } + if (ENIC_SETTING(enic, TXCSUM)) + netdev->hw_features |= NETIF_F_SG | NETIF_F_HW_CSUM; + if (ENIC_SETTING(enic, TSO)) + netdev->hw_features |= NETIF_F_TSO | + NETIF_F_TSO6 | NETIF_F_TSO_ECN; + if (ENIC_SETTING(enic, RSS)) + netdev->hw_features |= NETIF_F_RXHASH; + if (ENIC_SETTING(enic, RXCSUM)) + netdev->hw_features |= NETIF_F_RXCSUM; + if (ENIC_SETTING(enic, VXLAN)) { + u64 patch_level; + u64 a1 = 0; + + netdev->hw_enc_features |= NETIF_F_RXCSUM | + NETIF_F_TSO | + NETIF_F_TSO6 | + NETIF_F_TSO_ECN | + NETIF_F_GSO_UDP_TUNNEL | + NETIF_F_HW_CSUM | + NETIF_F_GSO_UDP_TUNNEL_CSUM; + netdev->hw_features |= netdev->hw_enc_features; + /* get bit mask from hw about supported offload bit level + * BIT(0) = fw supports patch_level 0 + * fcoe bit = encap + * fcoe_fc_crc_ok = outer csum ok + * BIT(1) = always set by fw + * BIT(2) = fw supports patch_level 2 + * BIT(0) in rss_hash = encap + * BIT(1,2) in rss_hash = outer_ip_csum_ok/ + * outer_tcp_csum_ok + * used in enic_rq_indicate_buf + */ + err = vnic_dev_get_supported_feature_ver(enic->vdev, + VIC_FEATURE_VXLAN, + &patch_level, &a1); + if (err) + patch_level = 0; + enic->vxlan.flags = (u8)a1; + /* mask bits that are supported by driver + */ + patch_level &= BIT_ULL(0) | BIT_ULL(2); + patch_level = fls(patch_level); + patch_level = patch_level ? patch_level - 1 : 0; + enic->vxlan.patch_level = patch_level; + } + + netdev->features |= netdev->hw_features; + netdev->vlan_features |= netdev->features; + +#ifdef CONFIG_RFS_ACCEL + netdev->hw_features |= NETIF_F_NTUPLE; +#endif + + if (using_dac) + netdev->features |= NETIF_F_HIGHDMA; + + netdev->priv_flags |= IFF_UNICAST_FLT; + + /* MTU range: 68 - 9000 */ + netdev->min_mtu = ENIC_MIN_MTU; + netdev->max_mtu = ENIC_MAX_MTU; + netdev->mtu = enic->port_mtu; + + err = register_netdev(netdev); + if (err) { + dev_err(dev, "Cannot register net device, aborting\n"); + goto err_out_dev_deinit; + } + enic->rx_copybreak = RX_COPYBREAK_DEFAULT; + + return 0; + +err_out_dev_deinit: + enic_dev_deinit(enic); +err_out_dev_close: + vnic_dev_close(enic->vdev); +err_out_disable_sriov: + kfree(enic->pp); +err_out_disable_sriov_pp: +#ifdef CONFIG_PCI_IOV + if (enic_sriov_enabled(enic)) { + pci_disable_sriov(pdev); + enic->priv_flags &= ~ENIC_SRIOV_ENABLED; + } +#endif +err_out_vnic_unregister: + vnic_dev_unregister(enic->vdev); +err_out_iounmap: + enic_iounmap(enic); +err_out_release_regions: + pci_release_regions(pdev); +err_out_disable_device: + pci_disable_device(pdev); +err_out_free_netdev: + free_netdev(netdev); + + return err; +} + +static void enic_remove(struct pci_dev *pdev) +{ + struct net_device *netdev = pci_get_drvdata(pdev); + + if (netdev) { + struct enic *enic = netdev_priv(netdev); + + cancel_work_sync(&enic->reset); + cancel_work_sync(&enic->change_mtu_work); + unregister_netdev(netdev); + enic_dev_deinit(enic); + vnic_dev_close(enic->vdev); +#ifdef CONFIG_PCI_IOV + if (enic_sriov_enabled(enic)) { + pci_disable_sriov(pdev); + enic->priv_flags &= ~ENIC_SRIOV_ENABLED; + } +#endif + kfree(enic->pp); + vnic_dev_unregister(enic->vdev); + enic_iounmap(enic); + pci_release_regions(pdev); + pci_disable_device(pdev); + free_netdev(netdev); + } +} + +static struct pci_driver enic_driver = { + .name = DRV_NAME, + .id_table = enic_id_table, + .probe = enic_probe, + .remove = enic_remove, +}; + +static int __init enic_init_module(void) +{ + pr_info("%s, ver %s\n", DRV_DESCRIPTION, DRV_VERSION); + + return pci_register_driver(&enic_driver); +} + +static void __exit enic_cleanup_module(void) +{ + pci_unregister_driver(&enic_driver); +} + +module_init(enic_init_module); +module_exit(enic_cleanup_module); diff --git a/drivers/net/ethernet/cisco/enic/enic_pp.c b/drivers/net/ethernet/cisco/enic/enic_pp.c new file mode 100644 index 000000000..e6a83198c --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic_pp.c @@ -0,0 +1,368 @@ +/* + * Copyright 2011 Cisco Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include <linux/kernel.h> +#include <linux/string.h> +#include <linux/errno.h> +#include <linux/types.h> +#include <linux/netdevice.h> +#include <linux/etherdevice.h> +#include <linux/rtnetlink.h> +#include <net/ip.h> + +#include "vnic_vic.h" +#include "enic_res.h" +#include "enic.h" +#include "enic_dev.h" +#include "enic_pp.h" + +/* + * Checks validity of vf index that came in + * port profile request + */ +int enic_is_valid_pp_vf(struct enic *enic, int vf, int *err) +{ + if (vf != PORT_SELF_VF) { +#ifdef CONFIG_PCI_IOV + if (enic_sriov_enabled(enic)) { + if (vf < 0 || vf >= enic->num_vfs) { + *err = -EINVAL; + goto err_out; + } + } else { + *err = -EOPNOTSUPP; + goto err_out; + } +#else + *err = -EOPNOTSUPP; + goto err_out; +#endif + } + + if (vf == PORT_SELF_VF && !enic_is_dynamic(enic)) { + *err = -EOPNOTSUPP; + goto err_out; + } + + *err = 0; + return 1; + +err_out: + return 0; +} + +static int enic_set_port_profile(struct enic *enic, int vf) +{ + struct net_device *netdev = enic->netdev; + struct enic_port_profile *pp; + struct vic_provinfo *vp; + const u8 oui[3] = VIC_PROVINFO_CISCO_OUI; + const __be16 os_type = htons(VIC_GENERIC_PROV_OS_TYPE_LINUX); + char uuid_str[38]; + char client_mac_str[18]; + u8 *client_mac; + int err; + + ENIC_PP_BY_INDEX(enic, vf, pp, &err); + if (err) + return err; + + if (!(pp->set & ENIC_SET_NAME) || !strlen(pp->name)) + return -EINVAL; + + vp = vic_provinfo_alloc(GFP_KERNEL, oui, + VIC_PROVINFO_GENERIC_TYPE); + if (!vp) + return -ENOMEM; + + VIC_PROVINFO_ADD_TLV(vp, + VIC_GENERIC_PROV_TLV_PORT_PROFILE_NAME_STR, + strlen(pp->name) + 1, pp->name); + + if (!is_zero_ether_addr(pp->mac_addr)) { + client_mac = pp->mac_addr; + } else if (vf == PORT_SELF_VF) { + client_mac = netdev->dev_addr; + } else { + netdev_err(netdev, "Cannot find pp mac address " + "for VF %d\n", vf); + err = -EINVAL; + goto add_tlv_failure; + } + + VIC_PROVINFO_ADD_TLV(vp, + VIC_GENERIC_PROV_TLV_CLIENT_MAC_ADDR, + ETH_ALEN, client_mac); + + snprintf(client_mac_str, sizeof(client_mac_str), "%pM", client_mac); + VIC_PROVINFO_ADD_TLV(vp, + VIC_GENERIC_PROV_TLV_CLUSTER_PORT_UUID_STR, + sizeof(client_mac_str), client_mac_str); + + if (pp->set & ENIC_SET_INSTANCE) { + sprintf(uuid_str, "%pUB", pp->instance_uuid); + VIC_PROVINFO_ADD_TLV(vp, + VIC_GENERIC_PROV_TLV_CLIENT_UUID_STR, + sizeof(uuid_str), uuid_str); + } + + if (pp->set & ENIC_SET_HOST) { + sprintf(uuid_str, "%pUB", pp->host_uuid); + VIC_PROVINFO_ADD_TLV(vp, + VIC_GENERIC_PROV_TLV_HOST_UUID_STR, + sizeof(uuid_str), uuid_str); + } + + VIC_PROVINFO_ADD_TLV(vp, + VIC_GENERIC_PROV_TLV_OS_TYPE, + sizeof(os_type), &os_type); + + ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, vnic_dev_init_prov2, (u8 *)vp, + vic_provinfo_size(vp)); + err = enic_dev_status_to_errno(err); + +add_tlv_failure: + vic_provinfo_free(vp); + + return err; +} + +static int enic_unset_port_profile(struct enic *enic, int vf) +{ + int err; + + ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, vnic_dev_deinit); + if (err) + return enic_dev_status_to_errno(err); + + if (vf == PORT_SELF_VF) + enic_reset_addr_lists(enic); + + return 0; +} + +static int enic_are_pp_different(struct enic_port_profile *pp1, + struct enic_port_profile *pp2) +{ + return strcmp(pp1->name, pp2->name) | !!memcmp(pp1->instance_uuid, + pp2->instance_uuid, PORT_UUID_MAX) | + !!memcmp(pp1->host_uuid, pp2->host_uuid, PORT_UUID_MAX) | + !ether_addr_equal(pp1->mac_addr, pp2->mac_addr); +} + +static int enic_pp_preassociate(struct enic *enic, int vf, + struct enic_port_profile *prev_pp, int *restore_pp); +static int enic_pp_disassociate(struct enic *enic, int vf, + struct enic_port_profile *prev_pp, int *restore_pp); +static int enic_pp_preassociate_rr(struct enic *enic, int vf, + struct enic_port_profile *prev_pp, int *restore_pp); +static int enic_pp_associate(struct enic *enic, int vf, + struct enic_port_profile *prev_pp, int *restore_pp); + +static int (*enic_pp_handlers[])(struct enic *enic, int vf, + struct enic_port_profile *prev_state, + int *restore_pp) = { + [PORT_REQUEST_PREASSOCIATE] = enic_pp_preassociate, + [PORT_REQUEST_PREASSOCIATE_RR] = enic_pp_preassociate_rr, + [PORT_REQUEST_ASSOCIATE] = enic_pp_associate, + [PORT_REQUEST_DISASSOCIATE] = enic_pp_disassociate, +}; + +static const int enic_pp_handlers_count = + ARRAY_SIZE(enic_pp_handlers); + +static int enic_pp_preassociate(struct enic *enic, int vf, + struct enic_port_profile *prev_pp, int *restore_pp) +{ + return -EOPNOTSUPP; +} + +static int enic_pp_disassociate(struct enic *enic, int vf, + struct enic_port_profile *prev_pp, int *restore_pp) +{ + struct net_device *netdev = enic->netdev; + struct enic_port_profile *pp; + int err; + + ENIC_PP_BY_INDEX(enic, vf, pp, &err); + if (err) + return err; + + /* Deregister mac addresses */ + if (!is_zero_ether_addr(pp->mac_addr)) + ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, vnic_dev_del_addr, + pp->mac_addr); + else if (vf == PORT_SELF_VF && !is_zero_ether_addr(netdev->dev_addr)) + ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, vnic_dev_del_addr, + netdev->dev_addr); + + return enic_unset_port_profile(enic, vf); +} + +static int enic_pp_preassociate_rr(struct enic *enic, int vf, + struct enic_port_profile *prev_pp, int *restore_pp) +{ + struct enic_port_profile *pp; + int err; + int active = 0; + + ENIC_PP_BY_INDEX(enic, vf, pp, &err); + if (err) + return err; + + if (pp->request != PORT_REQUEST_ASSOCIATE) { + /* If pre-associate is not part of an associate. + We always disassociate first */ + err = enic_pp_handlers[PORT_REQUEST_DISASSOCIATE](enic, vf, + prev_pp, restore_pp); + if (err) + return err; + + *restore_pp = 0; + } + + *restore_pp = 0; + + err = enic_set_port_profile(enic, vf); + if (err) + return err; + + /* If pre-associate is not part of an associate. */ + if (pp->request != PORT_REQUEST_ASSOCIATE) { + /* Enable device as standby */ + ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, vnic_dev_enable2, + active); + err = enic_dev_status_to_errno(err); + } + + return err; +} + +static int enic_pp_associate(struct enic *enic, int vf, + struct enic_port_profile *prev_pp, int *restore_pp) +{ + struct net_device *netdev = enic->netdev; + struct enic_port_profile *pp; + int err; + int active = 1; + + ENIC_PP_BY_INDEX(enic, vf, pp, &err); + if (err) + return err; + + /* Check if a pre-associate was called before */ + if (prev_pp->request != PORT_REQUEST_PREASSOCIATE_RR || + (prev_pp->request == PORT_REQUEST_PREASSOCIATE_RR && + enic_are_pp_different(prev_pp, pp))) { + err = enic_pp_handlers[PORT_REQUEST_DISASSOCIATE]( + enic, vf, prev_pp, restore_pp); + if (err) + return err; + + *restore_pp = 0; + } + + err = enic_pp_handlers[PORT_REQUEST_PREASSOCIATE_RR]( + enic, vf, prev_pp, restore_pp); + if (err) + return err; + + *restore_pp = 0; + + /* Enable device as active */ + ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, vnic_dev_enable2, active); + err = enic_dev_status_to_errno(err); + if (err) + return err; + + /* Register mac address */ + if (!is_zero_ether_addr(pp->mac_addr)) + ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, vnic_dev_add_addr, + pp->mac_addr); + else if (vf == PORT_SELF_VF && !is_zero_ether_addr(netdev->dev_addr)) + ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, vnic_dev_add_addr, + netdev->dev_addr); + + return 0; +} + +int enic_process_set_pp_request(struct enic *enic, int vf, + struct enic_port_profile *prev_pp, int *restore_pp) +{ + struct enic_port_profile *pp; + int err; + + ENIC_PP_BY_INDEX(enic, vf, pp, &err); + if (err) + return err; + + if (pp->request >= enic_pp_handlers_count + || !enic_pp_handlers[pp->request]) + return -EOPNOTSUPP; + + return enic_pp_handlers[pp->request](enic, vf, prev_pp, restore_pp); +} + +int enic_process_get_pp_request(struct enic *enic, int vf, + int request, u16 *response) +{ + int err, status = ERR_SUCCESS; + + switch (request) { + + case PORT_REQUEST_PREASSOCIATE_RR: + case PORT_REQUEST_ASSOCIATE: + ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, + vnic_dev_enable2_done, &status); + break; + + case PORT_REQUEST_DISASSOCIATE: + ENIC_DEVCMD_PROXY_BY_INDEX(vf, err, enic, + vnic_dev_deinit_done, &status); + break; + + default: + return -EINVAL; + } + + if (err) + status = err; + + switch (status) { + case ERR_SUCCESS: + *response = PORT_PROFILE_RESPONSE_SUCCESS; + break; + case ERR_EINVAL: + *response = PORT_PROFILE_RESPONSE_INVALID; + break; + case ERR_EBADSTATE: + *response = PORT_PROFILE_RESPONSE_BADSTATE; + break; + case ERR_ENOMEM: + *response = PORT_PROFILE_RESPONSE_INSUFFICIENT_RESOURCES; + break; + case ERR_EINPROGRESS: + *response = PORT_PROFILE_RESPONSE_INPROGRESS; + break; + default: + *response = PORT_PROFILE_RESPONSE_ERROR; + break; + } + + return 0; +} diff --git a/drivers/net/ethernet/cisco/enic/enic_pp.h b/drivers/net/ethernet/cisco/enic/enic_pp.h new file mode 100644 index 000000000..a09ff392c --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic_pp.h @@ -0,0 +1,36 @@ +/* + * Copyright 2011 Cisco Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _ENIC_PP_H_ +#define _ENIC_PP_H_ + +#define ENIC_PP_BY_INDEX(enic, vf, pp, err) \ + do { \ + if (enic_is_valid_pp_vf(enic, vf, err)) \ + pp = (vf == PORT_SELF_VF) ? enic->pp : enic->pp + vf; \ + else \ + pp = NULL; \ + } while (0) + +int enic_process_set_pp_request(struct enic *enic, int vf, + struct enic_port_profile *prev_pp, int *restore_pp); +int enic_process_get_pp_request(struct enic *enic, int vf, + int request, u16 *response); +int enic_is_valid_pp_vf(struct enic *enic, int vf, int *err); + +#endif /* _ENIC_PP_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/enic_res.c b/drivers/net/ethernet/cisco/enic/enic_res.c new file mode 100644 index 000000000..40b20817d --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic_res.c @@ -0,0 +1,390 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/types.h> +#include <linux/pci.h> +#include <linux/netdevice.h> + +#include "wq_enet_desc.h" +#include "rq_enet_desc.h" +#include "cq_enet_desc.h" +#include "vnic_resource.h" +#include "vnic_enet.h" +#include "vnic_dev.h" +#include "vnic_wq.h" +#include "vnic_rq.h" +#include "vnic_cq.h" +#include "vnic_intr.h" +#include "vnic_stats.h" +#include "vnic_nic.h" +#include "vnic_rss.h" +#include "enic_res.h" +#include "enic.h" + +int enic_get_vnic_config(struct enic *enic) +{ + struct vnic_enet_config *c = &enic->config; + int err; + + err = vnic_dev_get_mac_addr(enic->vdev, enic->mac_addr); + if (err) { + dev_err(enic_get_dev(enic), + "Error getting MAC addr, %d\n", err); + return err; + } + +#define GET_CONFIG(m) \ + do { \ + err = vnic_dev_spec(enic->vdev, \ + offsetof(struct vnic_enet_config, m), \ + sizeof(c->m), &c->m); \ + if (err) { \ + dev_err(enic_get_dev(enic), \ + "Error getting %s, %d\n", #m, err); \ + return err; \ + } \ + } while (0) + + GET_CONFIG(flags); + GET_CONFIG(wq_desc_count); + GET_CONFIG(rq_desc_count); + GET_CONFIG(mtu); + GET_CONFIG(intr_timer_type); + GET_CONFIG(intr_mode); + GET_CONFIG(intr_timer_usec); + GET_CONFIG(loop_tag); + GET_CONFIG(num_arfs); + + c->wq_desc_count = + min_t(u32, ENIC_MAX_WQ_DESCS, + max_t(u32, ENIC_MIN_WQ_DESCS, + c->wq_desc_count)); + c->wq_desc_count &= 0xffffffe0; /* must be aligned to groups of 32 */ + + c->rq_desc_count = + min_t(u32, ENIC_MAX_RQ_DESCS, + max_t(u32, ENIC_MIN_RQ_DESCS, + c->rq_desc_count)); + c->rq_desc_count &= 0xffffffe0; /* must be aligned to groups of 32 */ + + if (c->mtu == 0) + c->mtu = 1500; + c->mtu = min_t(u16, ENIC_MAX_MTU, + max_t(u16, ENIC_MIN_MTU, + c->mtu)); + + c->intr_timer_usec = min_t(u32, c->intr_timer_usec, + vnic_dev_get_intr_coal_timer_max(enic->vdev)); + + dev_info(enic_get_dev(enic), + "vNIC MAC addr %pM wq/rq %d/%d mtu %d\n", + enic->mac_addr, c->wq_desc_count, c->rq_desc_count, c->mtu); + + dev_info(enic_get_dev(enic), "vNIC csum tx/rx %s/%s " + "tso/lro %s/%s rss %s intr mode %s type %s timer %d usec " + "loopback tag 0x%04x\n", + ENIC_SETTING(enic, TXCSUM) ? "yes" : "no", + ENIC_SETTING(enic, RXCSUM) ? "yes" : "no", + ENIC_SETTING(enic, TSO) ? "yes" : "no", + ENIC_SETTING(enic, LRO) ? "yes" : "no", + ENIC_SETTING(enic, RSS) ? "yes" : "no", + c->intr_mode == VENET_INTR_MODE_INTX ? "INTx" : + c->intr_mode == VENET_INTR_MODE_MSI ? "MSI" : + c->intr_mode == VENET_INTR_MODE_ANY ? "any" : + "unknown", + c->intr_timer_type == VENET_INTR_TYPE_MIN ? "min" : + c->intr_timer_type == VENET_INTR_TYPE_IDLE ? "idle" : + "unknown", + c->intr_timer_usec, + c->loop_tag); + + return 0; +} + +int enic_add_vlan(struct enic *enic, u16 vlanid) +{ + u64 a0 = vlanid, a1 = 0; + int wait = 1000; + int err; + + err = vnic_dev_cmd(enic->vdev, CMD_VLAN_ADD, &a0, &a1, wait); + if (err) + dev_err(enic_get_dev(enic), "Can't add vlan id, %d\n", err); + + return err; +} + +int enic_del_vlan(struct enic *enic, u16 vlanid) +{ + u64 a0 = vlanid, a1 = 0; + int wait = 1000; + int err; + + err = vnic_dev_cmd(enic->vdev, CMD_VLAN_DEL, &a0, &a1, wait); + if (err) + dev_err(enic_get_dev(enic), "Can't delete vlan id, %d\n", err); + + return err; +} + +int enic_set_nic_cfg(struct enic *enic, u8 rss_default_cpu, u8 rss_hash_type, + u8 rss_hash_bits, u8 rss_base_cpu, u8 rss_enable, u8 tso_ipid_split_en, + u8 ig_vlan_strip_en) +{ + enum vnic_devcmd_cmd cmd = CMD_NIC_CFG; + u64 a0, a1; + u32 nic_cfg; + int wait = 1000; + + vnic_set_nic_cfg(&nic_cfg, rss_default_cpu, + rss_hash_type, rss_hash_bits, rss_base_cpu, + rss_enable, tso_ipid_split_en, ig_vlan_strip_en); + + a0 = nic_cfg; + a1 = 0; + + if (rss_hash_type & (NIC_CFG_RSS_HASH_TYPE_UDP_IPV4 | + NIC_CFG_RSS_HASH_TYPE_UDP_IPV6)) + cmd = CMD_NIC_CFG_CHK; + + return vnic_dev_cmd(enic->vdev, cmd, &a0, &a1, wait); +} + +int enic_set_rss_key(struct enic *enic, dma_addr_t key_pa, u64 len) +{ + u64 a0 = (u64)key_pa, a1 = len; + int wait = 1000; + + return vnic_dev_cmd(enic->vdev, CMD_RSS_KEY, &a0, &a1, wait); +} + +int enic_set_rss_cpu(struct enic *enic, dma_addr_t cpu_pa, u64 len) +{ + u64 a0 = (u64)cpu_pa, a1 = len; + int wait = 1000; + + return vnic_dev_cmd(enic->vdev, CMD_RSS_CPU, &a0, &a1, wait); +} + +void enic_free_vnic_resources(struct enic *enic) +{ + unsigned int i; + + for (i = 0; i < enic->wq_count; i++) + vnic_wq_free(&enic->wq[i]); + for (i = 0; i < enic->rq_count; i++) + vnic_rq_free(&enic->rq[i]); + for (i = 0; i < enic->cq_count; i++) + vnic_cq_free(&enic->cq[i]); + for (i = 0; i < enic->intr_count; i++) + vnic_intr_free(&enic->intr[i]); +} + +void enic_get_res_counts(struct enic *enic) +{ + enic->wq_count = vnic_dev_get_res_count(enic->vdev, RES_TYPE_WQ); + enic->rq_count = vnic_dev_get_res_count(enic->vdev, RES_TYPE_RQ); + enic->cq_count = vnic_dev_get_res_count(enic->vdev, RES_TYPE_CQ); + enic->intr_count = vnic_dev_get_res_count(enic->vdev, + RES_TYPE_INTR_CTRL); + + dev_info(enic_get_dev(enic), + "vNIC resources avail: wq %d rq %d cq %d intr %d\n", + enic->wq_count, enic->rq_count, + enic->cq_count, enic->intr_count); +} + +void enic_init_vnic_resources(struct enic *enic) +{ + enum vnic_dev_intr_mode intr_mode; + unsigned int mask_on_assertion; + unsigned int interrupt_offset; + unsigned int error_interrupt_enable; + unsigned int error_interrupt_offset; + unsigned int cq_index; + unsigned int i; + + intr_mode = vnic_dev_get_intr_mode(enic->vdev); + + /* Init RQ/WQ resources. + * + * RQ[0 - n-1] point to CQ[0 - n-1] + * WQ[0 - m-1] point to CQ[n - n+m-1] + * + * Error interrupt is not enabled for MSI. + */ + + switch (intr_mode) { + case VNIC_DEV_INTR_MODE_INTX: + case VNIC_DEV_INTR_MODE_MSIX: + error_interrupt_enable = 1; + error_interrupt_offset = enic->intr_count - 2; + break; + default: + error_interrupt_enable = 0; + error_interrupt_offset = 0; + break; + } + + for (i = 0; i < enic->rq_count; i++) { + cq_index = i; + vnic_rq_init(&enic->rq[i], + cq_index, + error_interrupt_enable, + error_interrupt_offset); + } + + for (i = 0; i < enic->wq_count; i++) { + cq_index = enic->rq_count + i; + vnic_wq_init(&enic->wq[i], + cq_index, + error_interrupt_enable, + error_interrupt_offset); + } + + /* Init CQ resources + * + * CQ[0 - n+m-1] point to INTR[0] for INTx, MSI + * CQ[0 - n+m-1] point to INTR[0 - n+m-1] for MSI-X + */ + + for (i = 0; i < enic->cq_count; i++) { + + switch (intr_mode) { + case VNIC_DEV_INTR_MODE_MSIX: + interrupt_offset = i; + break; + default: + interrupt_offset = 0; + break; + } + + vnic_cq_init(&enic->cq[i], + 0 /* flow_control_enable */, + 1 /* color_enable */, + 0 /* cq_head */, + 0 /* cq_tail */, + 1 /* cq_tail_color */, + 1 /* interrupt_enable */, + 1 /* cq_entry_enable */, + 0 /* cq_message_enable */, + interrupt_offset, + 0 /* cq_message_addr */); + } + + /* Init INTR resources + * + * mask_on_assertion is not used for INTx due to the level- + * triggered nature of INTx + */ + + switch (intr_mode) { + case VNIC_DEV_INTR_MODE_MSI: + case VNIC_DEV_INTR_MODE_MSIX: + mask_on_assertion = 1; + break; + default: + mask_on_assertion = 0; + break; + } + + for (i = 0; i < enic->intr_count; i++) { + vnic_intr_init(&enic->intr[i], + enic->config.intr_timer_usec, + enic->config.intr_timer_type, + mask_on_assertion); + } +} + +int enic_alloc_vnic_resources(struct enic *enic) +{ + enum vnic_dev_intr_mode intr_mode; + unsigned int i; + int err; + + intr_mode = vnic_dev_get_intr_mode(enic->vdev); + + dev_info(enic_get_dev(enic), "vNIC resources used: " + "wq %d rq %d cq %d intr %d intr mode %s\n", + enic->wq_count, enic->rq_count, + enic->cq_count, enic->intr_count, + intr_mode == VNIC_DEV_INTR_MODE_INTX ? "legacy PCI INTx" : + intr_mode == VNIC_DEV_INTR_MODE_MSI ? "MSI" : + intr_mode == VNIC_DEV_INTR_MODE_MSIX ? "MSI-X" : + "unknown"); + + /* Allocate queue resources + */ + + for (i = 0; i < enic->wq_count; i++) { + err = vnic_wq_alloc(enic->vdev, &enic->wq[i], i, + enic->config.wq_desc_count, + sizeof(struct wq_enet_desc)); + if (err) + goto err_out_cleanup; + } + + for (i = 0; i < enic->rq_count; i++) { + err = vnic_rq_alloc(enic->vdev, &enic->rq[i], i, + enic->config.rq_desc_count, + sizeof(struct rq_enet_desc)); + if (err) + goto err_out_cleanup; + } + + for (i = 0; i < enic->cq_count; i++) { + if (i < enic->rq_count) + err = vnic_cq_alloc(enic->vdev, &enic->cq[i], i, + enic->config.rq_desc_count, + sizeof(struct cq_enet_rq_desc)); + else + err = vnic_cq_alloc(enic->vdev, &enic->cq[i], i, + enic->config.wq_desc_count, + sizeof(struct cq_enet_wq_desc)); + if (err) + goto err_out_cleanup; + } + + for (i = 0; i < enic->intr_count; i++) { + err = vnic_intr_alloc(enic->vdev, &enic->intr[i], i); + if (err) + goto err_out_cleanup; + } + + /* Hook remaining resource + */ + + enic->legacy_pba = vnic_dev_get_res(enic->vdev, + RES_TYPE_INTR_PBA_LEGACY, 0); + if (!enic->legacy_pba && intr_mode == VNIC_DEV_INTR_MODE_INTX) { + dev_err(enic_get_dev(enic), + "Failed to hook legacy pba resource\n"); + err = -ENODEV; + goto err_out_cleanup; + } + + return 0; + +err_out_cleanup: + enic_free_vnic_resources(enic); + + return err; +} diff --git a/drivers/net/ethernet/cisco/enic/enic_res.h b/drivers/net/ethernet/cisco/enic/enic_res.h new file mode 100644 index 000000000..81f98a8b6 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/enic_res.h @@ -0,0 +1,153 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _ENIC_RES_H_ +#define _ENIC_RES_H_ + +#include "wq_enet_desc.h" +#include "rq_enet_desc.h" +#include "vnic_wq.h" +#include "vnic_rq.h" + +#define ENIC_MIN_WQ_DESCS 64 +#define ENIC_MAX_WQ_DESCS 4096 +#define ENIC_MIN_RQ_DESCS 64 +#define ENIC_MAX_RQ_DESCS 4096 + +#define ENIC_MIN_MTU ETH_MIN_MTU +#define ENIC_MAX_MTU 9000 + +#define ENIC_MULTICAST_PERFECT_FILTERS 32 +#define ENIC_UNICAST_PERFECT_FILTERS 32 + +#define ENIC_NON_TSO_MAX_DESC 16 + +#define ENIC_SETTING(enic, f) ((enic->config.flags & VENETF_##f) ? 1 : 0) + +static inline void enic_queue_wq_desc_ex(struct vnic_wq *wq, + void *os_buf, dma_addr_t dma_addr, unsigned int len, + unsigned int mss_or_csum_offset, unsigned int hdr_len, + int vlan_tag_insert, unsigned int vlan_tag, + int offload_mode, int cq_entry, int sop, int eop, int loopback) +{ + struct wq_enet_desc *desc = vnic_wq_next_desc(wq); + u8 desc_skip_cnt = 1; + u8 compressed_send = 0; + u64 wrid = 0; + + wq_enet_desc_enc(desc, + (u64)dma_addr | VNIC_PADDR_TARGET, + (u16)len, + (u16)mss_or_csum_offset, + (u16)hdr_len, (u8)offload_mode, + (u8)eop, (u8)cq_entry, + 0, /* fcoe_encap */ + (u8)vlan_tag_insert, + (u16)vlan_tag, + (u8)loopback); + + vnic_wq_post(wq, os_buf, dma_addr, len, sop, eop, desc_skip_cnt, + (u8)cq_entry, compressed_send, wrid); +} + +static inline void enic_queue_wq_desc_cont(struct vnic_wq *wq, + void *os_buf, dma_addr_t dma_addr, unsigned int len, + int eop, int loopback) +{ + enic_queue_wq_desc_ex(wq, os_buf, dma_addr, len, + 0, 0, 0, 0, 0, + eop, 0 /* !SOP */, eop, loopback); +} + +static inline void enic_queue_wq_desc(struct vnic_wq *wq, void *os_buf, + dma_addr_t dma_addr, unsigned int len, int vlan_tag_insert, + unsigned int vlan_tag, int eop, int loopback) +{ + enic_queue_wq_desc_ex(wq, os_buf, dma_addr, len, + 0, 0, vlan_tag_insert, vlan_tag, + WQ_ENET_OFFLOAD_MODE_CSUM, + eop, 1 /* SOP */, eop, loopback); +} + +static inline void enic_queue_wq_desc_csum(struct vnic_wq *wq, + void *os_buf, dma_addr_t dma_addr, unsigned int len, + int ip_csum, int tcpudp_csum, int vlan_tag_insert, + unsigned int vlan_tag, int eop, int loopback) +{ + enic_queue_wq_desc_ex(wq, os_buf, dma_addr, len, + (ip_csum ? 1 : 0) + (tcpudp_csum ? 2 : 0), + 0, vlan_tag_insert, vlan_tag, + WQ_ENET_OFFLOAD_MODE_CSUM, + eop, 1 /* SOP */, eop, loopback); +} + +static inline void enic_queue_wq_desc_csum_l4(struct vnic_wq *wq, + void *os_buf, dma_addr_t dma_addr, unsigned int len, + unsigned int csum_offset, unsigned int hdr_len, + int vlan_tag_insert, unsigned int vlan_tag, int eop, int loopback) +{ + enic_queue_wq_desc_ex(wq, os_buf, dma_addr, len, + csum_offset, hdr_len, vlan_tag_insert, vlan_tag, + WQ_ENET_OFFLOAD_MODE_CSUM_L4, + eop, 1 /* SOP */, eop, loopback); +} + +static inline void enic_queue_wq_desc_tso(struct vnic_wq *wq, + void *os_buf, dma_addr_t dma_addr, unsigned int len, + unsigned int mss, unsigned int hdr_len, int vlan_tag_insert, + unsigned int vlan_tag, int eop, int loopback) +{ + enic_queue_wq_desc_ex(wq, os_buf, dma_addr, len, + mss, hdr_len, vlan_tag_insert, vlan_tag, + WQ_ENET_OFFLOAD_MODE_TSO, + eop, 1 /* SOP */, eop, loopback); +} + +static inline void enic_queue_rq_desc(struct vnic_rq *rq, + void *os_buf, unsigned int os_buf_index, + dma_addr_t dma_addr, unsigned int len) +{ + struct rq_enet_desc *desc = vnic_rq_next_desc(rq); + u64 wrid = 0; + u8 type = os_buf_index ? + RQ_ENET_TYPE_NOT_SOP : RQ_ENET_TYPE_ONLY_SOP; + + rq_enet_desc_enc(desc, + (u64)dma_addr | VNIC_PADDR_TARGET, + type, (u16)len); + + vnic_rq_post(rq, os_buf, os_buf_index, dma_addr, len, wrid); +} + +struct enic; + +int enic_get_vnic_config(struct enic *); +int enic_add_vlan(struct enic *enic, u16 vlanid); +int enic_del_vlan(struct enic *enic, u16 vlanid); +int enic_set_nic_cfg(struct enic *enic, u8 rss_default_cpu, u8 rss_hash_type, + u8 rss_hash_bits, u8 rss_base_cpu, u8 rss_enable, u8 tso_ipid_split_en, + u8 ig_vlan_strip_en); +int enic_set_rss_key(struct enic *enic, dma_addr_t key_pa, u64 len); +int enic_set_rss_cpu(struct enic *enic, dma_addr_t cpu_pa, u64 len); +void enic_get_res_counts(struct enic *enic); +void enic_init_vnic_resources(struct enic *enic); +int enic_alloc_vnic_resources(struct enic *); +void enic_free_vnic_resources(struct enic *); + +#endif /* _ENIC_RES_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/rq_enet_desc.h b/drivers/net/ethernet/cisco/enic/rq_enet_desc.h new file mode 100644 index 000000000..e6dd30988 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/rq_enet_desc.h @@ -0,0 +1,60 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _RQ_ENET_DESC_H_ +#define _RQ_ENET_DESC_H_ + +/* Ethernet receive queue descriptor: 16B */ +struct rq_enet_desc { + __le64 address; + __le16 length_type; + u8 reserved[6]; +}; + +enum rq_enet_type_types { + RQ_ENET_TYPE_ONLY_SOP = 0, + RQ_ENET_TYPE_NOT_SOP = 1, + RQ_ENET_TYPE_RESV2 = 2, + RQ_ENET_TYPE_RESV3 = 3, +}; + +#define RQ_ENET_ADDR_BITS 64 +#define RQ_ENET_LEN_BITS 14 +#define RQ_ENET_LEN_MASK ((1 << RQ_ENET_LEN_BITS) - 1) +#define RQ_ENET_TYPE_BITS 2 +#define RQ_ENET_TYPE_MASK ((1 << RQ_ENET_TYPE_BITS) - 1) + +static inline void rq_enet_desc_enc(struct rq_enet_desc *desc, + u64 address, u8 type, u16 length) +{ + desc->address = cpu_to_le64(address); + desc->length_type = cpu_to_le16((length & RQ_ENET_LEN_MASK) | + ((type & RQ_ENET_TYPE_MASK) << RQ_ENET_LEN_BITS)); +} + +static inline void rq_enet_desc_dec(struct rq_enet_desc *desc, + u64 *address, u8 *type, u16 *length) +{ + *address = le64_to_cpu(desc->address); + *length = le16_to_cpu(desc->length_type) & RQ_ENET_LEN_MASK; + *type = (u8)((le16_to_cpu(desc->length_type) >> RQ_ENET_LEN_BITS) & + RQ_ENET_TYPE_MASK); +} + +#endif /* _RQ_ENET_DESC_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/vnic_cq.c b/drivers/net/ethernet/cisco/enic/vnic_cq.c new file mode 100644 index 000000000..9c682aff3 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_cq.c @@ -0,0 +1,92 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/types.h> +#include <linux/pci.h> + +#include "vnic_dev.h" +#include "vnic_cq.h" +#include "enic.h" + +void vnic_cq_free(struct vnic_cq *cq) +{ + vnic_dev_free_desc_ring(cq->vdev, &cq->ring); + + cq->ctrl = NULL; +} + +int vnic_cq_alloc(struct vnic_dev *vdev, struct vnic_cq *cq, unsigned int index, + unsigned int desc_count, unsigned int desc_size) +{ + int err; + + cq->index = index; + cq->vdev = vdev; + + cq->ctrl = vnic_dev_get_res(vdev, RES_TYPE_CQ, index); + if (!cq->ctrl) { + vdev_err(vdev, "Failed to hook CQ[%d] resource\n", index); + return -EINVAL; + } + + err = vnic_dev_alloc_desc_ring(vdev, &cq->ring, desc_count, desc_size); + if (err) + return err; + + return 0; +} + +void vnic_cq_init(struct vnic_cq *cq, unsigned int flow_control_enable, + unsigned int color_enable, unsigned int cq_head, unsigned int cq_tail, + unsigned int cq_tail_color, unsigned int interrupt_enable, + unsigned int cq_entry_enable, unsigned int cq_message_enable, + unsigned int interrupt_offset, u64 cq_message_addr) +{ + u64 paddr; + + paddr = (u64)cq->ring.base_addr | VNIC_PADDR_TARGET; + writeq(paddr, &cq->ctrl->ring_base); + iowrite32(cq->ring.desc_count, &cq->ctrl->ring_size); + iowrite32(flow_control_enable, &cq->ctrl->flow_control_enable); + iowrite32(color_enable, &cq->ctrl->color_enable); + iowrite32(cq_head, &cq->ctrl->cq_head); + iowrite32(cq_tail, &cq->ctrl->cq_tail); + iowrite32(cq_tail_color, &cq->ctrl->cq_tail_color); + iowrite32(interrupt_enable, &cq->ctrl->interrupt_enable); + iowrite32(cq_entry_enable, &cq->ctrl->cq_entry_enable); + iowrite32(cq_message_enable, &cq->ctrl->cq_message_enable); + iowrite32(interrupt_offset, &cq->ctrl->interrupt_offset); + writeq(cq_message_addr, &cq->ctrl->cq_message_addr); + + cq->interrupt_offset = interrupt_offset; +} + +void vnic_cq_clean(struct vnic_cq *cq) +{ + cq->to_clean = 0; + cq->last_color = 0; + + iowrite32(0, &cq->ctrl->cq_head); + iowrite32(0, &cq->ctrl->cq_tail); + iowrite32(1, &cq->ctrl->cq_tail_color); + + vnic_dev_clear_desc_ring(&cq->ring); +} diff --git a/drivers/net/ethernet/cisco/enic/vnic_cq.h b/drivers/net/ethernet/cisco/enic/vnic_cq.h new file mode 100644 index 000000000..4e6aa6585 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_cq.h @@ -0,0 +1,123 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _VNIC_CQ_H_ +#define _VNIC_CQ_H_ + +#include "cq_desc.h" +#include "vnic_dev.h" + +/* Completion queue control */ +struct vnic_cq_ctrl { + u64 ring_base; /* 0x00 */ + u32 ring_size; /* 0x08 */ + u32 pad0; + u32 flow_control_enable; /* 0x10 */ + u32 pad1; + u32 color_enable; /* 0x18 */ + u32 pad2; + u32 cq_head; /* 0x20 */ + u32 pad3; + u32 cq_tail; /* 0x28 */ + u32 pad4; + u32 cq_tail_color; /* 0x30 */ + u32 pad5; + u32 interrupt_enable; /* 0x38 */ + u32 pad6; + u32 cq_entry_enable; /* 0x40 */ + u32 pad7; + u32 cq_message_enable; /* 0x48 */ + u32 pad8; + u32 interrupt_offset; /* 0x50 */ + u32 pad9; + u64 cq_message_addr; /* 0x58 */ + u32 pad10; +}; + +struct vnic_rx_bytes_counter { + unsigned int small_pkt_bytes_cnt; + unsigned int large_pkt_bytes_cnt; +}; + +struct vnic_cq { + unsigned int index; + struct vnic_dev *vdev; + struct vnic_cq_ctrl __iomem *ctrl; /* memory-mapped */ + struct vnic_dev_ring ring; + unsigned int to_clean; + unsigned int last_color; + unsigned int interrupt_offset; + struct vnic_rx_bytes_counter pkt_size_counter; + unsigned int cur_rx_coal_timeval; + unsigned int tobe_rx_coal_timeval; + ktime_t prev_ts; +}; + +static inline unsigned int vnic_cq_service(struct vnic_cq *cq, + unsigned int work_to_do, + int (*q_service)(struct vnic_dev *vdev, struct cq_desc *cq_desc, + u8 type, u16 q_number, u16 completed_index, void *opaque), + void *opaque) +{ + struct cq_desc *cq_desc; + unsigned int work_done = 0; + u16 q_number, completed_index; + u8 type, color; + + cq_desc = (struct cq_desc *)((u8 *)cq->ring.descs + + cq->ring.desc_size * cq->to_clean); + cq_desc_dec(cq_desc, &type, &color, + &q_number, &completed_index); + + while (color != cq->last_color) { + + if ((*q_service)(cq->vdev, cq_desc, type, + q_number, completed_index, opaque)) + break; + + cq->to_clean++; + if (cq->to_clean == cq->ring.desc_count) { + cq->to_clean = 0; + cq->last_color = cq->last_color ? 0 : 1; + } + + cq_desc = (struct cq_desc *)((u8 *)cq->ring.descs + + cq->ring.desc_size * cq->to_clean); + cq_desc_dec(cq_desc, &type, &color, + &q_number, &completed_index); + + work_done++; + if (work_done >= work_to_do) + break; + } + + return work_done; +} + +void vnic_cq_free(struct vnic_cq *cq); +int vnic_cq_alloc(struct vnic_dev *vdev, struct vnic_cq *cq, unsigned int index, + unsigned int desc_count, unsigned int desc_size); +void vnic_cq_init(struct vnic_cq *cq, unsigned int flow_control_enable, + unsigned int color_enable, unsigned int cq_head, unsigned int cq_tail, + unsigned int cq_tail_color, unsigned int interrupt_enable, + unsigned int cq_entry_enable, unsigned int message_enable, + unsigned int interrupt_offset, u64 message_addr); +void vnic_cq_clean(struct vnic_cq *cq); + +#endif /* _VNIC_CQ_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/vnic_dev.c b/drivers/net/ethernet/cisco/enic/vnic_dev.c new file mode 100644 index 000000000..901e44b0b --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_dev.c @@ -0,0 +1,1304 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/types.h> +#include <linux/pci.h> +#include <linux/delay.h> +#include <linux/if_ether.h> + +#include "vnic_resource.h" +#include "vnic_devcmd.h" +#include "vnic_dev.h" +#include "vnic_wq.h" +#include "vnic_stats.h" +#include "enic.h" + +#define VNIC_MAX_RES_HDR_SIZE \ + (sizeof(struct vnic_resource_header) + \ + sizeof(struct vnic_resource) * RES_TYPE_MAX) +#define VNIC_RES_STRIDE 128 + +void *vnic_dev_priv(struct vnic_dev *vdev) +{ + return vdev->priv; +} + +static int vnic_dev_discover_res(struct vnic_dev *vdev, + struct vnic_dev_bar *bar, unsigned int num_bars) +{ + struct vnic_resource_header __iomem *rh; + struct mgmt_barmap_hdr __iomem *mrh; + struct vnic_resource __iomem *r; + u8 type; + + if (num_bars == 0) + return -EINVAL; + + if (bar->len < VNIC_MAX_RES_HDR_SIZE) { + vdev_err(vdev, "vNIC BAR0 res hdr length error\n"); + return -EINVAL; + } + + rh = bar->vaddr; + mrh = bar->vaddr; + if (!rh) { + vdev_err(vdev, "vNIC BAR0 res hdr not mem-mapped\n"); + return -EINVAL; + } + + /* Check for mgmt vnic in addition to normal vnic */ + if ((ioread32(&rh->magic) != VNIC_RES_MAGIC) || + (ioread32(&rh->version) != VNIC_RES_VERSION)) { + if ((ioread32(&mrh->magic) != MGMTVNIC_MAGIC) || + (ioread32(&mrh->version) != MGMTVNIC_VERSION)) { + vdev_err(vdev, "vNIC BAR0 res magic/version error exp (%lx/%lx) or (%lx/%lx), curr (%x/%x)\n", + VNIC_RES_MAGIC, VNIC_RES_VERSION, + MGMTVNIC_MAGIC, MGMTVNIC_VERSION, + ioread32(&rh->magic), ioread32(&rh->version)); + return -EINVAL; + } + } + + if (ioread32(&mrh->magic) == MGMTVNIC_MAGIC) + r = (struct vnic_resource __iomem *)(mrh + 1); + else + r = (struct vnic_resource __iomem *)(rh + 1); + + + while ((type = ioread8(&r->type)) != RES_TYPE_EOL) { + + u8 bar_num = ioread8(&r->bar); + u32 bar_offset = ioread32(&r->bar_offset); + u32 count = ioread32(&r->count); + u32 len; + + r++; + + if (bar_num >= num_bars) + continue; + + if (!bar[bar_num].len || !bar[bar_num].vaddr) + continue; + + switch (type) { + case RES_TYPE_WQ: + case RES_TYPE_RQ: + case RES_TYPE_CQ: + case RES_TYPE_INTR_CTRL: + /* each count is stride bytes long */ + len = count * VNIC_RES_STRIDE; + if (len + bar_offset > bar[bar_num].len) { + vdev_err(vdev, "vNIC BAR0 resource %d out-of-bounds, offset 0x%x + size 0x%x > bar len 0x%lx\n", + type, bar_offset, len, + bar[bar_num].len); + return -EINVAL; + } + break; + case RES_TYPE_INTR_PBA_LEGACY: + case RES_TYPE_DEVCMD: + case RES_TYPE_DEVCMD2: + len = count; + break; + default: + continue; + } + + vdev->res[type].count = count; + vdev->res[type].vaddr = (char __iomem *)bar[bar_num].vaddr + + bar_offset; + vdev->res[type].bus_addr = bar[bar_num].bus_addr + bar_offset; + } + + return 0; +} + +unsigned int vnic_dev_get_res_count(struct vnic_dev *vdev, + enum vnic_res_type type) +{ + return vdev->res[type].count; +} +EXPORT_SYMBOL(vnic_dev_get_res_count); + +void __iomem *vnic_dev_get_res(struct vnic_dev *vdev, enum vnic_res_type type, + unsigned int index) +{ + if (!vdev->res[type].vaddr) + return NULL; + + switch (type) { + case RES_TYPE_WQ: + case RES_TYPE_RQ: + case RES_TYPE_CQ: + case RES_TYPE_INTR_CTRL: + return (char __iomem *)vdev->res[type].vaddr + + index * VNIC_RES_STRIDE; + default: + return (char __iomem *)vdev->res[type].vaddr; + } +} +EXPORT_SYMBOL(vnic_dev_get_res); + +static unsigned int vnic_dev_desc_ring_size(struct vnic_dev_ring *ring, + unsigned int desc_count, unsigned int desc_size) +{ + /* The base address of the desc rings must be 512 byte aligned. + * Descriptor count is aligned to groups of 32 descriptors. A + * count of 0 means the maximum 4096 descriptors. Descriptor + * size is aligned to 16 bytes. + */ + + unsigned int count_align = 32; + unsigned int desc_align = 16; + + ring->base_align = 512; + + if (desc_count == 0) + desc_count = 4096; + + ring->desc_count = ALIGN(desc_count, count_align); + + ring->desc_size = ALIGN(desc_size, desc_align); + + ring->size = ring->desc_count * ring->desc_size; + ring->size_unaligned = ring->size + ring->base_align; + + return ring->size_unaligned; +} + +void vnic_dev_clear_desc_ring(struct vnic_dev_ring *ring) +{ + memset(ring->descs, 0, ring->size); +} + +int vnic_dev_alloc_desc_ring(struct vnic_dev *vdev, struct vnic_dev_ring *ring, + unsigned int desc_count, unsigned int desc_size) +{ + vnic_dev_desc_ring_size(ring, desc_count, desc_size); + + ring->descs_unaligned = pci_alloc_consistent(vdev->pdev, + ring->size_unaligned, + &ring->base_addr_unaligned); + + if (!ring->descs_unaligned) { + vdev_err(vdev, "Failed to allocate ring (size=%d), aborting\n", + (int)ring->size); + return -ENOMEM; + } + + ring->base_addr = ALIGN(ring->base_addr_unaligned, + ring->base_align); + ring->descs = (u8 *)ring->descs_unaligned + + (ring->base_addr - ring->base_addr_unaligned); + + vnic_dev_clear_desc_ring(ring); + + ring->desc_avail = ring->desc_count - 1; + + return 0; +} + +void vnic_dev_free_desc_ring(struct vnic_dev *vdev, struct vnic_dev_ring *ring) +{ + if (ring->descs) { + pci_free_consistent(vdev->pdev, + ring->size_unaligned, + ring->descs_unaligned, + ring->base_addr_unaligned); + ring->descs = NULL; + } +} + +static int _vnic_dev_cmd(struct vnic_dev *vdev, enum vnic_devcmd_cmd cmd, + int wait) +{ + struct vnic_devcmd __iomem *devcmd = vdev->devcmd; + unsigned int i; + int delay; + u32 status; + int err; + + status = ioread32(&devcmd->status); + if (status == 0xFFFFFFFF) { + /* PCI-e target device is gone */ + return -ENODEV; + } + if (status & STAT_BUSY) { + vdev_neterr(vdev, "Busy devcmd %d\n", _CMD_N(cmd)); + return -EBUSY; + } + + if (_CMD_DIR(cmd) & _CMD_DIR_WRITE) { + for (i = 0; i < VNIC_DEVCMD_NARGS; i++) + writeq(vdev->args[i], &devcmd->args[i]); + wmb(); + } + + iowrite32(cmd, &devcmd->cmd); + + if ((_CMD_FLAGS(cmd) & _CMD_FLAGS_NOWAIT)) + return 0; + + for (delay = 0; delay < wait; delay++) { + + udelay(100); + + status = ioread32(&devcmd->status); + if (status == 0xFFFFFFFF) { + /* PCI-e target device is gone */ + return -ENODEV; + } + + if (!(status & STAT_BUSY)) { + + if (status & STAT_ERROR) { + err = (int)readq(&devcmd->args[0]); + if (err == ERR_EINVAL && + cmd == CMD_CAPABILITY) + return -err; + if (err != ERR_ECMDUNKNOWN || + cmd != CMD_CAPABILITY) + vdev_neterr(vdev, "Error %d devcmd %d\n", + err, _CMD_N(cmd)); + return -err; + } + + if (_CMD_DIR(cmd) & _CMD_DIR_READ) { + rmb(); + for (i = 0; i < VNIC_DEVCMD_NARGS; i++) + vdev->args[i] = readq(&devcmd->args[i]); + } + + return 0; + } + } + + vdev_neterr(vdev, "Timedout devcmd %d\n", _CMD_N(cmd)); + return -ETIMEDOUT; +} + +static int _vnic_dev_cmd2(struct vnic_dev *vdev, enum vnic_devcmd_cmd cmd, + int wait) +{ + struct devcmd2_controller *dc2c = vdev->devcmd2; + struct devcmd2_result *result; + u8 color; + unsigned int i; + int delay, err; + u32 fetch_index, new_posted; + u32 posted = dc2c->posted; + + fetch_index = ioread32(&dc2c->wq_ctrl->fetch_index); + + if (fetch_index == 0xFFFFFFFF) + return -ENODEV; + + new_posted = (posted + 1) % DEVCMD2_RING_SIZE; + + if (new_posted == fetch_index) { + vdev_neterr(vdev, "devcmd2 %d: wq is full. fetch index: %u, posted index: %u\n", + _CMD_N(cmd), fetch_index, posted); + return -EBUSY; + } + dc2c->cmd_ring[posted].cmd = cmd; + dc2c->cmd_ring[posted].flags = 0; + + if ((_CMD_FLAGS(cmd) & _CMD_FLAGS_NOWAIT)) + dc2c->cmd_ring[posted].flags |= DEVCMD2_FNORESULT; + if (_CMD_DIR(cmd) & _CMD_DIR_WRITE) + for (i = 0; i < VNIC_DEVCMD_NARGS; i++) + dc2c->cmd_ring[posted].args[i] = vdev->args[i]; + + /* Adding write memory barrier prevents compiler and/or CPU reordering, + * thus avoiding descriptor posting before descriptor is initialized. + * Otherwise, hardware can read stale descriptor fields. + */ + wmb(); + iowrite32(new_posted, &dc2c->wq_ctrl->posted_index); + dc2c->posted = new_posted; + + if (dc2c->cmd_ring[posted].flags & DEVCMD2_FNORESULT) + return 0; + + result = dc2c->result + dc2c->next_result; + color = dc2c->color; + + dc2c->next_result++; + if (dc2c->next_result == dc2c->result_size) { + dc2c->next_result = 0; + dc2c->color = dc2c->color ? 0 : 1; + } + + for (delay = 0; delay < wait; delay++) { + if (result->color == color) { + if (result->error) { + err = result->error; + if (err != ERR_ECMDUNKNOWN || + cmd != CMD_CAPABILITY) + vdev_neterr(vdev, "Error %d devcmd %d\n", + err, _CMD_N(cmd)); + return -err; + } + if (_CMD_DIR(cmd) & _CMD_DIR_READ) + for (i = 0; i < VNIC_DEVCMD2_NARGS; i++) + vdev->args[i] = result->results[i]; + + return 0; + } + udelay(100); + } + + vdev_neterr(vdev, "devcmd %d timed out\n", _CMD_N(cmd)); + + return -ETIMEDOUT; +} + +static int vnic_dev_init_devcmd1(struct vnic_dev *vdev) +{ + vdev->devcmd = vnic_dev_get_res(vdev, RES_TYPE_DEVCMD, 0); + if (!vdev->devcmd) + return -ENODEV; + vdev->devcmd_rtn = _vnic_dev_cmd; + + return 0; +} + +static int vnic_dev_init_devcmd2(struct vnic_dev *vdev) +{ + int err; + unsigned int fetch_index; + + if (vdev->devcmd2) + return 0; + + vdev->devcmd2 = kzalloc(sizeof(*vdev->devcmd2), GFP_KERNEL); + if (!vdev->devcmd2) + return -ENOMEM; + + vdev->devcmd2->color = 1; + vdev->devcmd2->result_size = DEVCMD2_RING_SIZE; + err = enic_wq_devcmd2_alloc(vdev, &vdev->devcmd2->wq, DEVCMD2_RING_SIZE, + DEVCMD2_DESC_SIZE); + if (err) + goto err_free_devcmd2; + + fetch_index = ioread32(&vdev->devcmd2->wq.ctrl->fetch_index); + if (fetch_index == 0xFFFFFFFF) { /* check for hardware gone */ + vdev_err(vdev, "Fatal error in devcmd2 init - hardware surprise removal\n"); + err = -ENODEV; + goto err_free_wq; + } + + enic_wq_init_start(&vdev->devcmd2->wq, 0, fetch_index, fetch_index, 0, + 0); + vdev->devcmd2->posted = fetch_index; + vnic_wq_enable(&vdev->devcmd2->wq); + + err = vnic_dev_alloc_desc_ring(vdev, &vdev->devcmd2->results_ring, + DEVCMD2_RING_SIZE, DEVCMD2_DESC_SIZE); + if (err) + goto err_disable_wq; + + vdev->devcmd2->result = vdev->devcmd2->results_ring.descs; + vdev->devcmd2->cmd_ring = vdev->devcmd2->wq.ring.descs; + vdev->devcmd2->wq_ctrl = vdev->devcmd2->wq.ctrl; + vdev->args[0] = (u64)vdev->devcmd2->results_ring.base_addr | + VNIC_PADDR_TARGET; + vdev->args[1] = DEVCMD2_RING_SIZE; + + err = _vnic_dev_cmd2(vdev, CMD_INITIALIZE_DEVCMD2, 1000); + if (err) + goto err_free_desc_ring; + + vdev->devcmd_rtn = _vnic_dev_cmd2; + + return 0; + +err_free_desc_ring: + vnic_dev_free_desc_ring(vdev, &vdev->devcmd2->results_ring); +err_disable_wq: + vnic_wq_disable(&vdev->devcmd2->wq); +err_free_wq: + vnic_wq_free(&vdev->devcmd2->wq); +err_free_devcmd2: + kfree(vdev->devcmd2); + vdev->devcmd2 = NULL; + + return err; +} + +static void vnic_dev_deinit_devcmd2(struct vnic_dev *vdev) +{ + vnic_dev_free_desc_ring(vdev, &vdev->devcmd2->results_ring); + vnic_wq_disable(&vdev->devcmd2->wq); + vnic_wq_free(&vdev->devcmd2->wq); + kfree(vdev->devcmd2); +} + +static int vnic_dev_cmd_proxy(struct vnic_dev *vdev, + enum vnic_devcmd_cmd proxy_cmd, enum vnic_devcmd_cmd cmd, + u64 *a0, u64 *a1, int wait) +{ + u32 status; + int err; + + memset(vdev->args, 0, sizeof(vdev->args)); + + vdev->args[0] = vdev->proxy_index; + vdev->args[1] = cmd; + vdev->args[2] = *a0; + vdev->args[3] = *a1; + + err = vdev->devcmd_rtn(vdev, proxy_cmd, wait); + if (err) + return err; + + status = (u32)vdev->args[0]; + if (status & STAT_ERROR) { + err = (int)vdev->args[1]; + if (err != ERR_ECMDUNKNOWN || + cmd != CMD_CAPABILITY) + vdev_neterr(vdev, "Error %d proxy devcmd %d\n", + err, _CMD_N(cmd)); + return err; + } + + *a0 = vdev->args[1]; + *a1 = vdev->args[2]; + + return 0; +} + +static int vnic_dev_cmd_no_proxy(struct vnic_dev *vdev, + enum vnic_devcmd_cmd cmd, u64 *a0, u64 *a1, int wait) +{ + int err; + + vdev->args[0] = *a0; + vdev->args[1] = *a1; + + err = vdev->devcmd_rtn(vdev, cmd, wait); + + *a0 = vdev->args[0]; + *a1 = vdev->args[1]; + + return err; +} + +void vnic_dev_cmd_proxy_by_index_start(struct vnic_dev *vdev, u16 index) +{ + vdev->proxy = PROXY_BY_INDEX; + vdev->proxy_index = index; +} + +void vnic_dev_cmd_proxy_end(struct vnic_dev *vdev) +{ + vdev->proxy = PROXY_NONE; + vdev->proxy_index = 0; +} + +int vnic_dev_cmd(struct vnic_dev *vdev, enum vnic_devcmd_cmd cmd, + u64 *a0, u64 *a1, int wait) +{ + memset(vdev->args, 0, sizeof(vdev->args)); + + switch (vdev->proxy) { + case PROXY_BY_INDEX: + return vnic_dev_cmd_proxy(vdev, CMD_PROXY_BY_INDEX, cmd, + a0, a1, wait); + case PROXY_BY_BDF: + return vnic_dev_cmd_proxy(vdev, CMD_PROXY_BY_BDF, cmd, + a0, a1, wait); + case PROXY_NONE: + default: + return vnic_dev_cmd_no_proxy(vdev, cmd, a0, a1, wait); + } +} + +static int vnic_dev_capable(struct vnic_dev *vdev, enum vnic_devcmd_cmd cmd) +{ + u64 a0 = (u32)cmd, a1 = 0; + int wait = 1000; + int err; + + err = vnic_dev_cmd(vdev, CMD_CAPABILITY, &a0, &a1, wait); + + return !(err || a0); +} + +int vnic_dev_fw_info(struct vnic_dev *vdev, + struct vnic_devcmd_fw_info **fw_info) +{ + u64 a0, a1 = 0; + int wait = 1000; + int err = 0; + + if (!vdev->fw_info) { + vdev->fw_info = pci_zalloc_consistent(vdev->pdev, + sizeof(struct vnic_devcmd_fw_info), + &vdev->fw_info_pa); + if (!vdev->fw_info) + return -ENOMEM; + + a0 = vdev->fw_info_pa; + a1 = sizeof(struct vnic_devcmd_fw_info); + + /* only get fw_info once and cache it */ + if (vnic_dev_capable(vdev, CMD_MCPU_FW_INFO)) + err = vnic_dev_cmd(vdev, CMD_MCPU_FW_INFO, + &a0, &a1, wait); + else + err = vnic_dev_cmd(vdev, CMD_MCPU_FW_INFO_OLD, + &a0, &a1, wait); + } + + *fw_info = vdev->fw_info; + + return err; +} + +int vnic_dev_spec(struct vnic_dev *vdev, unsigned int offset, unsigned int size, + void *value) +{ + u64 a0, a1; + int wait = 1000; + int err; + + a0 = offset; + a1 = size; + + err = vnic_dev_cmd(vdev, CMD_DEV_SPEC, &a0, &a1, wait); + + switch (size) { + case 1: *(u8 *)value = (u8)a0; break; + case 2: *(u16 *)value = (u16)a0; break; + case 4: *(u32 *)value = (u32)a0; break; + case 8: *(u64 *)value = a0; break; + default: BUG(); break; + } + + return err; +} + +int vnic_dev_stats_dump(struct vnic_dev *vdev, struct vnic_stats **stats) +{ + u64 a0, a1; + int wait = 1000; + + if (!vdev->stats) { + vdev->stats = pci_alloc_consistent(vdev->pdev, + sizeof(struct vnic_stats), &vdev->stats_pa); + if (!vdev->stats) + return -ENOMEM; + } + + *stats = vdev->stats; + a0 = vdev->stats_pa; + a1 = sizeof(struct vnic_stats); + + return vnic_dev_cmd(vdev, CMD_STATS_DUMP, &a0, &a1, wait); +} + +int vnic_dev_close(struct vnic_dev *vdev) +{ + u64 a0 = 0, a1 = 0; + int wait = 1000; + return vnic_dev_cmd(vdev, CMD_CLOSE, &a0, &a1, wait); +} + +int vnic_dev_enable_wait(struct vnic_dev *vdev) +{ + u64 a0 = 0, a1 = 0; + int wait = 1000; + + if (vnic_dev_capable(vdev, CMD_ENABLE_WAIT)) + return vnic_dev_cmd(vdev, CMD_ENABLE_WAIT, &a0, &a1, wait); + else + return vnic_dev_cmd(vdev, CMD_ENABLE, &a0, &a1, wait); +} + +int vnic_dev_disable(struct vnic_dev *vdev) +{ + u64 a0 = 0, a1 = 0; + int wait = 1000; + return vnic_dev_cmd(vdev, CMD_DISABLE, &a0, &a1, wait); +} + +int vnic_dev_open(struct vnic_dev *vdev, int arg) +{ + u64 a0 = (u32)arg, a1 = 0; + int wait = 1000; + return vnic_dev_cmd(vdev, CMD_OPEN, &a0, &a1, wait); +} + +int vnic_dev_open_done(struct vnic_dev *vdev, int *done) +{ + u64 a0 = 0, a1 = 0; + int wait = 1000; + int err; + + *done = 0; + + err = vnic_dev_cmd(vdev, CMD_OPEN_STATUS, &a0, &a1, wait); + if (err) + return err; + + *done = (a0 == 0); + + return 0; +} + +int vnic_dev_soft_reset(struct vnic_dev *vdev, int arg) +{ + u64 a0 = (u32)arg, a1 = 0; + int wait = 1000; + return vnic_dev_cmd(vdev, CMD_SOFT_RESET, &a0, &a1, wait); +} + +int vnic_dev_soft_reset_done(struct vnic_dev *vdev, int *done) +{ + u64 a0 = 0, a1 = 0; + int wait = 1000; + int err; + + *done = 0; + + err = vnic_dev_cmd(vdev, CMD_SOFT_RESET_STATUS, &a0, &a1, wait); + if (err) + return err; + + *done = (a0 == 0); + + return 0; +} + +int vnic_dev_hang_reset(struct vnic_dev *vdev, int arg) +{ + u64 a0 = (u32)arg, a1 = 0; + int wait = 1000; + int err; + + if (vnic_dev_capable(vdev, CMD_HANG_RESET)) { + return vnic_dev_cmd(vdev, CMD_HANG_RESET, + &a0, &a1, wait); + } else { + err = vnic_dev_soft_reset(vdev, arg); + if (err) + return err; + return vnic_dev_init(vdev, 0); + } +} + +int vnic_dev_hang_reset_done(struct vnic_dev *vdev, int *done) +{ + u64 a0 = 0, a1 = 0; + int wait = 1000; + int err; + + *done = 0; + + if (vnic_dev_capable(vdev, CMD_HANG_RESET_STATUS)) { + err = vnic_dev_cmd(vdev, CMD_HANG_RESET_STATUS, + &a0, &a1, wait); + if (err) + return err; + } else { + return vnic_dev_soft_reset_done(vdev, done); + } + + *done = (a0 == 0); + + return 0; +} + +int vnic_dev_hang_notify(struct vnic_dev *vdev) +{ + u64 a0, a1; + int wait = 1000; + return vnic_dev_cmd(vdev, CMD_HANG_NOTIFY, &a0, &a1, wait); +} + +int vnic_dev_get_mac_addr(struct vnic_dev *vdev, u8 *mac_addr) +{ + u64 a0, a1; + int wait = 1000; + int err, i; + + for (i = 0; i < ETH_ALEN; i++) + mac_addr[i] = 0; + + err = vnic_dev_cmd(vdev, CMD_GET_MAC_ADDR, &a0, &a1, wait); + if (err) + return err; + + for (i = 0; i < ETH_ALEN; i++) + mac_addr[i] = ((u8 *)&a0)[i]; + + return 0; +} + +int vnic_dev_packet_filter(struct vnic_dev *vdev, int directed, int multicast, + int broadcast, int promisc, int allmulti) +{ + u64 a0, a1 = 0; + int wait = 1000; + int err; + + a0 = (directed ? CMD_PFILTER_DIRECTED : 0) | + (multicast ? CMD_PFILTER_MULTICAST : 0) | + (broadcast ? CMD_PFILTER_BROADCAST : 0) | + (promisc ? CMD_PFILTER_PROMISCUOUS : 0) | + (allmulti ? CMD_PFILTER_ALL_MULTICAST : 0); + + err = vnic_dev_cmd(vdev, CMD_PACKET_FILTER, &a0, &a1, wait); + if (err) + vdev_neterr(vdev, "Can't set packet filter\n"); + + return err; +} + +int vnic_dev_add_addr(struct vnic_dev *vdev, const u8 *addr) +{ + u64 a0 = 0, a1 = 0; + int wait = 1000; + int err; + int i; + + for (i = 0; i < ETH_ALEN; i++) + ((u8 *)&a0)[i] = addr[i]; + + err = vnic_dev_cmd(vdev, CMD_ADDR_ADD, &a0, &a1, wait); + if (err) + vdev_neterr(vdev, "Can't add addr [%pM], %d\n", addr, err); + + return err; +} + +int vnic_dev_del_addr(struct vnic_dev *vdev, const u8 *addr) +{ + u64 a0 = 0, a1 = 0; + int wait = 1000; + int err; + int i; + + for (i = 0; i < ETH_ALEN; i++) + ((u8 *)&a0)[i] = addr[i]; + + err = vnic_dev_cmd(vdev, CMD_ADDR_DEL, &a0, &a1, wait); + if (err) + vdev_neterr(vdev, "Can't del addr [%pM], %d\n", addr, err); + + return err; +} + +int vnic_dev_set_ig_vlan_rewrite_mode(struct vnic_dev *vdev, + u8 ig_vlan_rewrite_mode) +{ + u64 a0 = ig_vlan_rewrite_mode, a1 = 0; + int wait = 1000; + + if (vnic_dev_capable(vdev, CMD_IG_VLAN_REWRITE_MODE)) + return vnic_dev_cmd(vdev, CMD_IG_VLAN_REWRITE_MODE, + &a0, &a1, wait); + else + return 0; +} + +static int vnic_dev_notify_setcmd(struct vnic_dev *vdev, + void *notify_addr, dma_addr_t notify_pa, u16 intr) +{ + u64 a0, a1; + int wait = 1000; + int r; + + memset(notify_addr, 0, sizeof(struct vnic_devcmd_notify)); + vdev->notify = notify_addr; + vdev->notify_pa = notify_pa; + + a0 = (u64)notify_pa; + a1 = ((u64)intr << 32) & 0x0000ffff00000000ULL; + a1 += sizeof(struct vnic_devcmd_notify); + + r = vnic_dev_cmd(vdev, CMD_NOTIFY, &a0, &a1, wait); + vdev->notify_sz = (r == 0) ? (u32)a1 : 0; + return r; +} + +int vnic_dev_notify_set(struct vnic_dev *vdev, u16 intr) +{ + void *notify_addr; + dma_addr_t notify_pa; + + if (vdev->notify || vdev->notify_pa) { + vdev_neterr(vdev, "notify block %p still allocated\n", + vdev->notify); + return -EINVAL; + } + + notify_addr = pci_alloc_consistent(vdev->pdev, + sizeof(struct vnic_devcmd_notify), + ¬ify_pa); + if (!notify_addr) + return -ENOMEM; + + return vnic_dev_notify_setcmd(vdev, notify_addr, notify_pa, intr); +} + +static int vnic_dev_notify_unsetcmd(struct vnic_dev *vdev) +{ + u64 a0, a1; + int wait = 1000; + int err; + + a0 = 0; /* paddr = 0 to unset notify buffer */ + a1 = 0x0000ffff00000000ULL; /* intr num = -1 to unreg for intr */ + a1 += sizeof(struct vnic_devcmd_notify); + + err = vnic_dev_cmd(vdev, CMD_NOTIFY, &a0, &a1, wait); + vdev->notify = NULL; + vdev->notify_pa = 0; + vdev->notify_sz = 0; + + return err; +} + +int vnic_dev_notify_unset(struct vnic_dev *vdev) +{ + if (vdev->notify) { + pci_free_consistent(vdev->pdev, + sizeof(struct vnic_devcmd_notify), + vdev->notify, + vdev->notify_pa); + } + + return vnic_dev_notify_unsetcmd(vdev); +} + +static int vnic_dev_notify_ready(struct vnic_dev *vdev) +{ + u32 *words; + unsigned int nwords = vdev->notify_sz / 4; + unsigned int i; + u32 csum; + + if (!vdev->notify || !vdev->notify_sz) + return 0; + + do { + csum = 0; + memcpy(&vdev->notify_copy, vdev->notify, vdev->notify_sz); + words = (u32 *)&vdev->notify_copy; + for (i = 1; i < nwords; i++) + csum += words[i]; + } while (csum != words[0]); + + return 1; +} + +int vnic_dev_init(struct vnic_dev *vdev, int arg) +{ + u64 a0 = (u32)arg, a1 = 0; + int wait = 1000; + int r = 0; + + if (vnic_dev_capable(vdev, CMD_INIT)) + r = vnic_dev_cmd(vdev, CMD_INIT, &a0, &a1, wait); + else { + vnic_dev_cmd(vdev, CMD_INIT_v1, &a0, &a1, wait); + if (a0 & CMD_INITF_DEFAULT_MAC) { + /* Emulate these for old CMD_INIT_v1 which + * didn't pass a0 so no CMD_INITF_*. + */ + vnic_dev_cmd(vdev, CMD_GET_MAC_ADDR, &a0, &a1, wait); + vnic_dev_cmd(vdev, CMD_ADDR_ADD, &a0, &a1, wait); + } + } + return r; +} + +int vnic_dev_deinit(struct vnic_dev *vdev) +{ + u64 a0 = 0, a1 = 0; + int wait = 1000; + + return vnic_dev_cmd(vdev, CMD_DEINIT, &a0, &a1, wait); +} + +void vnic_dev_intr_coal_timer_info_default(struct vnic_dev *vdev) +{ + /* Default: hardware intr coal timer is in units of 1.5 usecs */ + vdev->intr_coal_timer_info.mul = 2; + vdev->intr_coal_timer_info.div = 3; + vdev->intr_coal_timer_info.max_usec = + vnic_dev_intr_coal_timer_hw_to_usec(vdev, 0xffff); +} + +int vnic_dev_intr_coal_timer_info(struct vnic_dev *vdev) +{ + int wait = 1000; + int err; + + memset(vdev->args, 0, sizeof(vdev->args)); + + if (vnic_dev_capable(vdev, CMD_INTR_COAL_CONVERT)) + err = vdev->devcmd_rtn(vdev, CMD_INTR_COAL_CONVERT, wait); + else + err = ERR_ECMDUNKNOWN; + + /* Use defaults when firmware doesn't support the devcmd at all or + * supports it for only specific hardware + */ + if ((err == ERR_ECMDUNKNOWN) || + (!err && !(vdev->args[0] && vdev->args[1] && vdev->args[2]))) { + vdev_netwarn(vdev, "Using default conversion factor for interrupt coalesce timer\n"); + vnic_dev_intr_coal_timer_info_default(vdev); + return 0; + } + + if (!err) { + vdev->intr_coal_timer_info.mul = (u32) vdev->args[0]; + vdev->intr_coal_timer_info.div = (u32) vdev->args[1]; + vdev->intr_coal_timer_info.max_usec = (u32) vdev->args[2]; + } + + return err; +} + +int vnic_dev_link_status(struct vnic_dev *vdev) +{ + if (!vnic_dev_notify_ready(vdev)) + return 0; + + return vdev->notify_copy.link_state; +} + +u32 vnic_dev_port_speed(struct vnic_dev *vdev) +{ + if (!vnic_dev_notify_ready(vdev)) + return 0; + + return vdev->notify_copy.port_speed; +} + +u32 vnic_dev_msg_lvl(struct vnic_dev *vdev) +{ + if (!vnic_dev_notify_ready(vdev)) + return 0; + + return vdev->notify_copy.msglvl; +} + +u32 vnic_dev_mtu(struct vnic_dev *vdev) +{ + if (!vnic_dev_notify_ready(vdev)) + return 0; + + return vdev->notify_copy.mtu; +} + +void vnic_dev_set_intr_mode(struct vnic_dev *vdev, + enum vnic_dev_intr_mode intr_mode) +{ + vdev->intr_mode = intr_mode; +} + +enum vnic_dev_intr_mode vnic_dev_get_intr_mode( + struct vnic_dev *vdev) +{ + return vdev->intr_mode; +} + +u32 vnic_dev_intr_coal_timer_usec_to_hw(struct vnic_dev *vdev, u32 usec) +{ + return (usec * vdev->intr_coal_timer_info.mul) / + vdev->intr_coal_timer_info.div; +} + +u32 vnic_dev_intr_coal_timer_hw_to_usec(struct vnic_dev *vdev, u32 hw_cycles) +{ + return (hw_cycles * vdev->intr_coal_timer_info.div) / + vdev->intr_coal_timer_info.mul; +} + +u32 vnic_dev_get_intr_coal_timer_max(struct vnic_dev *vdev) +{ + return vdev->intr_coal_timer_info.max_usec; +} + +void vnic_dev_unregister(struct vnic_dev *vdev) +{ + if (vdev) { + if (vdev->notify) + pci_free_consistent(vdev->pdev, + sizeof(struct vnic_devcmd_notify), + vdev->notify, + vdev->notify_pa); + if (vdev->stats) + pci_free_consistent(vdev->pdev, + sizeof(struct vnic_stats), + vdev->stats, vdev->stats_pa); + if (vdev->fw_info) + pci_free_consistent(vdev->pdev, + sizeof(struct vnic_devcmd_fw_info), + vdev->fw_info, vdev->fw_info_pa); + if (vdev->devcmd2) + vnic_dev_deinit_devcmd2(vdev); + + kfree(vdev); + } +} +EXPORT_SYMBOL(vnic_dev_unregister); + +struct vnic_dev *vnic_dev_register(struct vnic_dev *vdev, + void *priv, struct pci_dev *pdev, struct vnic_dev_bar *bar, + unsigned int num_bars) +{ + if (!vdev) { + vdev = kzalloc(sizeof(struct vnic_dev), GFP_KERNEL); + if (!vdev) + return NULL; + } + + vdev->priv = priv; + vdev->pdev = pdev; + + if (vnic_dev_discover_res(vdev, bar, num_bars)) + goto err_out; + + return vdev; + +err_out: + vnic_dev_unregister(vdev); + return NULL; +} +EXPORT_SYMBOL(vnic_dev_register); + +struct pci_dev *vnic_dev_get_pdev(struct vnic_dev *vdev) +{ + return vdev->pdev; +} +EXPORT_SYMBOL(vnic_dev_get_pdev); + +int vnic_devcmd_init(struct vnic_dev *vdev) +{ + void __iomem *res; + int err; + + res = vnic_dev_get_res(vdev, RES_TYPE_DEVCMD2, 0); + if (res) { + err = vnic_dev_init_devcmd2(vdev); + if (err) + vdev_warn(vdev, "DEVCMD2 init failed: %d, Using DEVCMD1\n", + err); + else + return 0; + } else { + vdev_warn(vdev, "DEVCMD2 resource not found (old firmware?) Using DEVCMD1\n"); + } + err = vnic_dev_init_devcmd1(vdev); + if (err) + vdev_err(vdev, "DEVCMD1 initialization failed: %d\n", err); + + return err; +} + +int vnic_dev_init_prov2(struct vnic_dev *vdev, u8 *buf, u32 len) +{ + u64 a0, a1 = len; + int wait = 1000; + dma_addr_t prov_pa; + void *prov_buf; + int ret; + + prov_buf = pci_alloc_consistent(vdev->pdev, len, &prov_pa); + if (!prov_buf) + return -ENOMEM; + + memcpy(prov_buf, buf, len); + + a0 = prov_pa; + + ret = vnic_dev_cmd(vdev, CMD_INIT_PROV_INFO2, &a0, &a1, wait); + + pci_free_consistent(vdev->pdev, len, prov_buf, prov_pa); + + return ret; +} + +int vnic_dev_enable2(struct vnic_dev *vdev, int active) +{ + u64 a0, a1 = 0; + int wait = 1000; + + a0 = (active ? CMD_ENABLE2_ACTIVE : 0); + + return vnic_dev_cmd(vdev, CMD_ENABLE2, &a0, &a1, wait); +} + +static int vnic_dev_cmd_status(struct vnic_dev *vdev, enum vnic_devcmd_cmd cmd, + int *status) +{ + u64 a0 = cmd, a1 = 0; + int wait = 1000; + int ret; + + ret = vnic_dev_cmd(vdev, CMD_STATUS, &a0, &a1, wait); + if (!ret) + *status = (int)a0; + + return ret; +} + +int vnic_dev_enable2_done(struct vnic_dev *vdev, int *status) +{ + return vnic_dev_cmd_status(vdev, CMD_ENABLE2, status); +} + +int vnic_dev_deinit_done(struct vnic_dev *vdev, int *status) +{ + return vnic_dev_cmd_status(vdev, CMD_DEINIT, status); +} + +int vnic_dev_set_mac_addr(struct vnic_dev *vdev, u8 *mac_addr) +{ + u64 a0, a1; + int wait = 1000; + int i; + + for (i = 0; i < ETH_ALEN; i++) + ((u8 *)&a0)[i] = mac_addr[i]; + + return vnic_dev_cmd(vdev, CMD_SET_MAC_ADDR, &a0, &a1, wait); +} + +/* vnic_dev_classifier: Add/Delete classifier entries + * @vdev: vdev of the device + * @cmd: CLSF_ADD for Add filter + * CLSF_DEL for Delete filter + * @entry: In case of ADD filter, the caller passes the RQ number in this + * variable. + * + * This function stores the filter_id returned by the firmware in the + * same variable before return; + * + * In case of DEL filter, the caller passes the RQ number. Return + * value is irrelevant. + * @data: filter data + */ +int vnic_dev_classifier(struct vnic_dev *vdev, u8 cmd, u16 *entry, + struct filter *data) +{ + u64 a0, a1; + int wait = 1000; + dma_addr_t tlv_pa; + int ret = -EINVAL; + struct filter_tlv *tlv, *tlv_va; + struct filter_action *action; + u64 tlv_size; + + if (cmd == CLSF_ADD) { + tlv_size = sizeof(struct filter) + + sizeof(struct filter_action) + + 2 * sizeof(struct filter_tlv); + tlv_va = pci_alloc_consistent(vdev->pdev, tlv_size, &tlv_pa); + if (!tlv_va) + return -ENOMEM; + tlv = tlv_va; + a0 = tlv_pa; + a1 = tlv_size; + memset(tlv, 0, tlv_size); + tlv->type = CLSF_TLV_FILTER; + tlv->length = sizeof(struct filter); + *(struct filter *)&tlv->val = *data; + + tlv = (struct filter_tlv *)((char *)tlv + + sizeof(struct filter_tlv) + + sizeof(struct filter)); + + tlv->type = CLSF_TLV_ACTION; + tlv->length = sizeof(struct filter_action); + action = (struct filter_action *)&tlv->val; + action->type = FILTER_ACTION_RQ_STEERING; + action->u.rq_idx = *entry; + + ret = vnic_dev_cmd(vdev, CMD_ADD_FILTER, &a0, &a1, wait); + *entry = (u16)a0; + pci_free_consistent(vdev->pdev, tlv_size, tlv_va, tlv_pa); + } else if (cmd == CLSF_DEL) { + a0 = *entry; + ret = vnic_dev_cmd(vdev, CMD_DEL_FILTER, &a0, &a1, wait); + } + + return ret; +} + +int vnic_dev_overlay_offload_ctrl(struct vnic_dev *vdev, u8 overlay, u8 config) +{ + u64 a0 = overlay; + u64 a1 = config; + int wait = 1000; + + return vnic_dev_cmd(vdev, CMD_OVERLAY_OFFLOAD_CTRL, &a0, &a1, wait); +} + +int vnic_dev_overlay_offload_cfg(struct vnic_dev *vdev, u8 overlay, + u16 vxlan_udp_port_number) +{ + u64 a1 = vxlan_udp_port_number; + u64 a0 = overlay; + int wait = 1000; + + return vnic_dev_cmd(vdev, CMD_OVERLAY_OFFLOAD_CFG, &a0, &a1, wait); +} + +int vnic_dev_get_supported_feature_ver(struct vnic_dev *vdev, u8 feature, + u64 *supported_versions, u64 *a1) +{ + u64 a0 = feature; + int wait = 1000; + int ret; + + ret = vnic_dev_cmd(vdev, CMD_GET_SUPP_FEATURE_VER, &a0, a1, wait); + if (!ret) + *supported_versions = a0; + + return ret; +} + +int vnic_dev_capable_rss_hash_type(struct vnic_dev *vdev, u8 *rss_hash_type) +{ + u64 a0 = CMD_NIC_CFG, a1 = 0; + int wait = 1000; + int err; + + err = vnic_dev_cmd(vdev, CMD_CAPABILITY, &a0, &a1, wait); + /* rss_hash_type is valid only when a0 is 1. Adapter which does not + * support CMD_CAPABILITY for rss_hash_type has a0 = 0 + */ + if (err || (a0 != 1)) + return -EOPNOTSUPP; + + a1 = (a1 >> NIC_CFG_RSS_HASH_TYPE_SHIFT) & + NIC_CFG_RSS_HASH_TYPE_MASK_FIELD; + + *rss_hash_type = (u8)a1; + + return 0; +} diff --git a/drivers/net/ethernet/cisco/enic/vnic_dev.h b/drivers/net/ethernet/cisco/enic/vnic_dev.h new file mode 100644 index 000000000..714fc1ed7 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_dev.h @@ -0,0 +1,189 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _VNIC_DEV_H_ +#define _VNIC_DEV_H_ + +#include "vnic_resource.h" +#include "vnic_devcmd.h" + +#ifndef VNIC_PADDR_TARGET +#define VNIC_PADDR_TARGET 0x0000000000000000ULL +#endif + +#ifndef readq +static inline u64 readq(void __iomem *reg) +{ + return (((u64)readl(reg + 0x4UL) << 32) | + (u64)readl(reg)); +} + +static inline void writeq(u64 val, void __iomem *reg) +{ + writel(val & 0xffffffff, reg); + writel(val >> 32, reg + 0x4UL); +} +#endif + +#undef pr_fmt +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +enum vnic_dev_intr_mode { + VNIC_DEV_INTR_MODE_UNKNOWN, + VNIC_DEV_INTR_MODE_INTX, + VNIC_DEV_INTR_MODE_MSI, + VNIC_DEV_INTR_MODE_MSIX, +}; + +struct vnic_dev_bar { + void __iomem *vaddr; + dma_addr_t bus_addr; + unsigned long len; +}; + +struct vnic_dev_ring { + void *descs; + size_t size; + dma_addr_t base_addr; + size_t base_align; + void *descs_unaligned; + size_t size_unaligned; + dma_addr_t base_addr_unaligned; + unsigned int desc_size; + unsigned int desc_count; + unsigned int desc_avail; +}; + +enum vnic_proxy_type { + PROXY_NONE, + PROXY_BY_BDF, + PROXY_BY_INDEX, +}; + +struct vnic_res { + void __iomem *vaddr; + dma_addr_t bus_addr; + unsigned int count; +}; + +struct vnic_intr_coal_timer_info { + u32 mul; + u32 div; + u32 max_usec; +}; + +struct vnic_dev { + void *priv; + struct pci_dev *pdev; + struct vnic_res res[RES_TYPE_MAX]; + enum vnic_dev_intr_mode intr_mode; + struct vnic_devcmd __iomem *devcmd; + struct vnic_devcmd_notify *notify; + struct vnic_devcmd_notify notify_copy; + dma_addr_t notify_pa; + u32 notify_sz; + dma_addr_t linkstatus_pa; + struct vnic_stats *stats; + dma_addr_t stats_pa; + struct vnic_devcmd_fw_info *fw_info; + dma_addr_t fw_info_pa; + enum vnic_proxy_type proxy; + u32 proxy_index; + u64 args[VNIC_DEVCMD_NARGS]; + struct vnic_intr_coal_timer_info intr_coal_timer_info; + struct devcmd2_controller *devcmd2; + int (*devcmd_rtn)(struct vnic_dev *vdev, enum vnic_devcmd_cmd cmd, + int wait); +}; + +struct vnic_stats; + +void *vnic_dev_priv(struct vnic_dev *vdev); +unsigned int vnic_dev_get_res_count(struct vnic_dev *vdev, + enum vnic_res_type type); +void __iomem *vnic_dev_get_res(struct vnic_dev *vdev, enum vnic_res_type type, + unsigned int index); +void vnic_dev_clear_desc_ring(struct vnic_dev_ring *ring); +int vnic_dev_alloc_desc_ring(struct vnic_dev *vdev, struct vnic_dev_ring *ring, + unsigned int desc_count, unsigned int desc_size); +void vnic_dev_free_desc_ring(struct vnic_dev *vdev, + struct vnic_dev_ring *ring); +int vnic_dev_cmd(struct vnic_dev *vdev, enum vnic_devcmd_cmd cmd, + u64 *a0, u64 *a1, int wait); +void vnic_dev_cmd_proxy_by_index_start(struct vnic_dev *vdev, u16 index); +void vnic_dev_cmd_proxy_end(struct vnic_dev *vdev); +int vnic_dev_fw_info(struct vnic_dev *vdev, + struct vnic_devcmd_fw_info **fw_info); +int vnic_dev_spec(struct vnic_dev *vdev, unsigned int offset, unsigned int size, + void *value); +int vnic_dev_stats_dump(struct vnic_dev *vdev, struct vnic_stats **stats); +int vnic_dev_hang_notify(struct vnic_dev *vdev); +int vnic_dev_packet_filter(struct vnic_dev *vdev, int directed, int multicast, + int broadcast, int promisc, int allmulti); +int vnic_dev_add_addr(struct vnic_dev *vdev, const u8 *addr); +int vnic_dev_del_addr(struct vnic_dev *vdev, const u8 *addr); +int vnic_dev_get_mac_addr(struct vnic_dev *vdev, u8 *mac_addr); +int vnic_dev_notify_set(struct vnic_dev *vdev, u16 intr); +int vnic_dev_notify_unset(struct vnic_dev *vdev); +int vnic_dev_link_status(struct vnic_dev *vdev); +u32 vnic_dev_port_speed(struct vnic_dev *vdev); +u32 vnic_dev_msg_lvl(struct vnic_dev *vdev); +u32 vnic_dev_mtu(struct vnic_dev *vdev); +int vnic_dev_close(struct vnic_dev *vdev); +int vnic_dev_enable_wait(struct vnic_dev *vdev); +int vnic_dev_disable(struct vnic_dev *vdev); +int vnic_dev_open(struct vnic_dev *vdev, int arg); +int vnic_dev_open_done(struct vnic_dev *vdev, int *done); +int vnic_dev_init(struct vnic_dev *vdev, int arg); +int vnic_dev_deinit(struct vnic_dev *vdev); +void vnic_dev_intr_coal_timer_info_default(struct vnic_dev *vdev); +int vnic_dev_intr_coal_timer_info(struct vnic_dev *vdev); +int vnic_dev_hang_reset(struct vnic_dev *vdev, int arg); +int vnic_dev_soft_reset(struct vnic_dev *vdev, int arg); +int vnic_dev_hang_reset_done(struct vnic_dev *vdev, int *done); +int vnic_dev_soft_reset_done(struct vnic_dev *vdev, int *done); +void vnic_dev_set_intr_mode(struct vnic_dev *vdev, + enum vnic_dev_intr_mode intr_mode); +enum vnic_dev_intr_mode vnic_dev_get_intr_mode(struct vnic_dev *vdev); +u32 vnic_dev_intr_coal_timer_usec_to_hw(struct vnic_dev *vdev, u32 usec); +u32 vnic_dev_intr_coal_timer_hw_to_usec(struct vnic_dev *vdev, u32 hw_cycles); +u32 vnic_dev_get_intr_coal_timer_max(struct vnic_dev *vdev); +void vnic_dev_unregister(struct vnic_dev *vdev); +int vnic_dev_set_ig_vlan_rewrite_mode(struct vnic_dev *vdev, + u8 ig_vlan_rewrite_mode); +struct vnic_dev *vnic_dev_register(struct vnic_dev *vdev, + void *priv, struct pci_dev *pdev, struct vnic_dev_bar *bar, + unsigned int num_bars); +struct pci_dev *vnic_dev_get_pdev(struct vnic_dev *vdev); +int vnic_dev_init_prov2(struct vnic_dev *vdev, u8 *buf, u32 len); +int vnic_dev_enable2(struct vnic_dev *vdev, int active); +int vnic_dev_enable2_done(struct vnic_dev *vdev, int *status); +int vnic_dev_deinit_done(struct vnic_dev *vdev, int *status); +int vnic_dev_set_mac_addr(struct vnic_dev *vdev, u8 *mac_addr); +int vnic_dev_classifier(struct vnic_dev *vdev, u8 cmd, u16 *entry, + struct filter *data); +int vnic_devcmd_init(struct vnic_dev *vdev); +int vnic_dev_overlay_offload_ctrl(struct vnic_dev *vdev, u8 overlay, u8 config); +int vnic_dev_overlay_offload_cfg(struct vnic_dev *vdev, u8 overlay, + u16 vxlan_udp_port_number); +int vnic_dev_get_supported_feature_ver(struct vnic_dev *vdev, u8 feature, + u64 *supported_versions, u64 *a1); +int vnic_dev_capable_rss_hash_type(struct vnic_dev *vdev, u8 *rss_hash_type); + +#endif /* _VNIC_DEV_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/vnic_devcmd.h b/drivers/net/ethernet/cisco/enic/vnic_devcmd.h new file mode 100644 index 000000000..fef5a0a06 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_devcmd.h @@ -0,0 +1,734 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _VNIC_DEVCMD_H_ +#define _VNIC_DEVCMD_H_ + +#define _CMD_NBITS 14 +#define _CMD_VTYPEBITS 10 +#define _CMD_FLAGSBITS 6 +#define _CMD_DIRBITS 2 + +#define _CMD_NMASK ((1 << _CMD_NBITS)-1) +#define _CMD_VTYPEMASK ((1 << _CMD_VTYPEBITS)-1) +#define _CMD_FLAGSMASK ((1 << _CMD_FLAGSBITS)-1) +#define _CMD_DIRMASK ((1 << _CMD_DIRBITS)-1) + +#define _CMD_NSHIFT 0 +#define _CMD_VTYPESHIFT (_CMD_NSHIFT+_CMD_NBITS) +#define _CMD_FLAGSSHIFT (_CMD_VTYPESHIFT+_CMD_VTYPEBITS) +#define _CMD_DIRSHIFT (_CMD_FLAGSSHIFT+_CMD_FLAGSBITS) + +/* + * Direction bits (from host perspective). + */ +#define _CMD_DIR_NONE 0U +#define _CMD_DIR_WRITE 1U +#define _CMD_DIR_READ 2U +#define _CMD_DIR_RW (_CMD_DIR_WRITE | _CMD_DIR_READ) + +/* + * Flag bits. + */ +#define _CMD_FLAGS_NONE 0U +#define _CMD_FLAGS_NOWAIT 1U + +/* + * vNIC type bits. + */ +#define _CMD_VTYPE_NONE 0U +#define _CMD_VTYPE_ENET 1U +#define _CMD_VTYPE_FC 2U +#define _CMD_VTYPE_SCSI 4U +#define _CMD_VTYPE_ALL (_CMD_VTYPE_ENET | _CMD_VTYPE_FC | _CMD_VTYPE_SCSI) + +/* + * Used to create cmds.. +*/ +#define _CMDCF(dir, flags, vtype, nr) \ + (((dir) << _CMD_DIRSHIFT) | \ + ((flags) << _CMD_FLAGSSHIFT) | \ + ((vtype) << _CMD_VTYPESHIFT) | \ + ((nr) << _CMD_NSHIFT)) +#define _CMDC(dir, vtype, nr) _CMDCF(dir, 0, vtype, nr) +#define _CMDCNW(dir, vtype, nr) _CMDCF(dir, _CMD_FLAGS_NOWAIT, vtype, nr) + +/* + * Used to decode cmds.. +*/ +#define _CMD_DIR(cmd) (((cmd) >> _CMD_DIRSHIFT) & _CMD_DIRMASK) +#define _CMD_FLAGS(cmd) (((cmd) >> _CMD_FLAGSSHIFT) & _CMD_FLAGSMASK) +#define _CMD_VTYPE(cmd) (((cmd) >> _CMD_VTYPESHIFT) & _CMD_VTYPEMASK) +#define _CMD_N(cmd) (((cmd) >> _CMD_NSHIFT) & _CMD_NMASK) + +enum vnic_devcmd_cmd { + CMD_NONE = _CMDC(_CMD_DIR_NONE, _CMD_VTYPE_NONE, 0), + + /* + * mcpu fw info in mem: + * in: + * (u64)a0=paddr to struct vnic_devcmd_fw_info + * action: + * Fills in struct vnic_devcmd_fw_info (128 bytes) + * note: + * An old definition of CMD_MCPU_FW_INFO + */ + CMD_MCPU_FW_INFO_OLD = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 1), + + /* + * mcpu fw info in mem: + * in: + * (u64)a0=paddr to struct vnic_devcmd_fw_info + * (u16)a1=size of the structure + * out: + * (u16)a1=0 for in:a1 = 0, + * data size actually written for other values. + * action: + * Fills in first 128 bytes of vnic_devcmd_fw_info for in:a1 = 0, + * first in:a1 bytes for 0 < in:a1 <= 132, + * 132 bytes for other values of in:a1. + * note: + * CMD_MCPU_FW_INFO and CMD_MCPU_FW_INFO_OLD have the same enum 1 + * for source compatibility. + */ + CMD_MCPU_FW_INFO = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_ALL, 1), + + /* dev-specific block member: + * in: (u16)a0=offset,(u8)a1=size + * out: a0=value */ + CMD_DEV_SPEC = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_ALL, 2), + + /* stats clear */ + CMD_STATS_CLEAR = _CMDCNW(_CMD_DIR_NONE, _CMD_VTYPE_ALL, 3), + + /* stats dump in mem: (u64)a0=paddr to stats area, + * (u16)a1=sizeof stats area */ + CMD_STATS_DUMP = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 4), + + /* set Rx packet filter: (u32)a0=filters (see CMD_PFILTER_*) */ + CMD_PACKET_FILTER = _CMDCNW(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 7), + + /* set Rx packet filter for all: (u32)a0=filters (see CMD_PFILTER_*) */ + CMD_PACKET_FILTER_ALL = _CMDCNW(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 7), + + /* hang detection notification */ + CMD_HANG_NOTIFY = _CMDC(_CMD_DIR_NONE, _CMD_VTYPE_ALL, 8), + + /* MAC address in (u48)a0 */ + CMD_GET_MAC_ADDR = _CMDC(_CMD_DIR_READ, + _CMD_VTYPE_ENET | _CMD_VTYPE_FC, 9), + + /* add addr from (u48)a0 */ + CMD_ADDR_ADD = _CMDCNW(_CMD_DIR_WRITE, + _CMD_VTYPE_ENET | _CMD_VTYPE_FC, 12), + + /* del addr from (u48)a0 */ + CMD_ADDR_DEL = _CMDCNW(_CMD_DIR_WRITE, + _CMD_VTYPE_ENET | _CMD_VTYPE_FC, 13), + + /* add VLAN id in (u16)a0 */ + CMD_VLAN_ADD = _CMDCNW(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 14), + + /* del VLAN id in (u16)a0 */ + CMD_VLAN_DEL = _CMDCNW(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 15), + + /* nic_cfg (no wait, always succeeds) + * in: (u32)a0 + * + * Capability query: + * out: (u64) a0 = 1 if a1 is valid + * (u64) a1 = (NIC_CFG bits supported) | (flags << 32) + * + * flags are CMD_NIC_CFG_CAPF_xxx + */ + CMD_NIC_CFG = _CMDCNW(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 16), + /* nic_cfg_chk (will return error if flags are invalid) + * in: (u32)a0 + * + * Capability query: + * out: (u64) a0 = 1 if a1 is valid + * (u64) a1 = (NIC_CFG bits supported) | (flags << 32) + * + * flags are CMD_NIC_CFG_CAPF_xxx + */ + CMD_NIC_CFG_CHK = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 16), + + /* union vnic_rss_key in mem: (u64)a0=paddr, (u16)a1=len */ + CMD_RSS_KEY = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 17), + + /* union vnic_rss_cpu in mem: (u64)a0=paddr, (u16)a1=len */ + CMD_RSS_CPU = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 18), + + /* initiate softreset */ + CMD_SOFT_RESET = _CMDCNW(_CMD_DIR_NONE, _CMD_VTYPE_ALL, 19), + + /* softreset status: + * out: a0=0 reset complete, a0=1 reset in progress */ + CMD_SOFT_RESET_STATUS = _CMDC(_CMD_DIR_READ, _CMD_VTYPE_ALL, 20), + + /* set struct vnic_devcmd_notify buffer in mem: + * in: + * (u64)a0=paddr to notify (set paddr=0 to unset) + * (u32)a1 & 0x00000000ffffffff=sizeof(struct vnic_devcmd_notify) + * (u16)a1 & 0x0000ffff00000000=intr num (-1 for no intr) + * out: + * (u32)a1 = effective size + */ + CMD_NOTIFY = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_ALL, 21), + + /* UNDI API: (u64)a0=paddr to s_PXENV_UNDI_ struct, + * (u8)a1=PXENV_UNDI_xxx */ + CMD_UNDI = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 22), + + /* initiate open sequence (u32)a0=flags (see CMD_OPENF_*) */ + CMD_OPEN = _CMDCNW(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 23), + + /* open status: + * out: a0=0 open complete, a0=1 open in progress */ + CMD_OPEN_STATUS = _CMDC(_CMD_DIR_READ, _CMD_VTYPE_ALL, 24), + + /* close vnic */ + CMD_CLOSE = _CMDC(_CMD_DIR_NONE, _CMD_VTYPE_ALL, 25), + + /* initialize virtual link: (u32)a0=flags (see CMD_INITF_*) */ +/***** Replaced by CMD_INIT *****/ + CMD_INIT_v1 = _CMDCNW(_CMD_DIR_READ, _CMD_VTYPE_ALL, 26), + + /* variant of CMD_INIT, with provisioning info + * (u64)a0=paddr of vnic_devcmd_provinfo + * (u32)a1=sizeof provision info */ + CMD_INIT_PROV_INFO = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 27), + + /* enable virtual link */ + CMD_ENABLE = _CMDCNW(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 28), + + /* enable virtual link, waiting variant. */ + CMD_ENABLE_WAIT = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 28), + + /* disable virtual link */ + CMD_DISABLE = _CMDC(_CMD_DIR_NONE, _CMD_VTYPE_ALL, 29), + + /* stats dump sum of all vnic stats on same uplink in mem: + * (u64)a0=paddr + * (u16)a1=sizeof stats area */ + CMD_STATS_DUMP_ALL = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 30), + + /* init status: + * out: a0=0 init complete, a0=1 init in progress + * if a0=0, a1=errno */ + CMD_INIT_STATUS = _CMDC(_CMD_DIR_READ, _CMD_VTYPE_ALL, 31), + + /* INT13 API: (u64)a0=paddr to vnic_int13_params struct + * (u32)a1=INT13_CMD_xxx */ + CMD_INT13 = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_FC, 32), + + /* logical uplink enable/disable: (u64)a0: 0/1=disable/enable */ + CMD_LOGICAL_UPLINK = _CMDCNW(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 33), + + /* undo initialize of virtual link */ + CMD_DEINIT = _CMDCNW(_CMD_DIR_NONE, _CMD_VTYPE_ALL, 34), + + /* initialize virtual link: (u32)a0=flags (see CMD_INITF_*) */ + CMD_INIT = _CMDCNW(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 35), + + /* check fw capability of a cmd: + * in: (u32)a0=cmd + * out: (u32)a0=errno, 0:valid cmd, a1=supported VNIC_STF_* bits */ + CMD_CAPABILITY = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_ALL, 36), + + /* persistent binding info + * in: (u64)a0=paddr of arg + * (u32)a1=CMD_PERBI_XXX */ + CMD_PERBI = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_FC, 37), + + /* Interrupt Assert Register functionality + * in: (u16)a0=interrupt number to assert + */ + CMD_IAR = _CMDCNW(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 38), + + /* initiate hangreset, like softreset after hang detected */ + CMD_HANG_RESET = _CMDC(_CMD_DIR_NONE, _CMD_VTYPE_ALL, 39), + + /* hangreset status: + * out: a0=0 reset complete, a0=1 reset in progress */ + CMD_HANG_RESET_STATUS = _CMDC(_CMD_DIR_READ, _CMD_VTYPE_ALL, 40), + + /* + * Set hw ingress packet vlan rewrite mode: + * in: (u32)a0=new vlan rewrite mode + * out: (u32)a0=old vlan rewrite mode */ + CMD_IG_VLAN_REWRITE_MODE = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_ENET, 41), + + /* + * in: (u16)a0=bdf of target vnic + * (u32)a1=cmd to proxy + * a2-a15=args to cmd in a1 + * out: (u32)a0=status of proxied cmd + * a1-a15=out args of proxied cmd */ + CMD_PROXY_BY_BDF = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_ALL, 42), + + /* + * As for BY_BDF except a0 is index of hvnlink subordinate vnic + * or SR-IOV virtual vnic + */ + CMD_PROXY_BY_INDEX = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_ALL, 43), + + /* + * For HPP toggle: + * adapter-info-get + * in: (u64)a0=phsical address of buffer passed in from caller. + * (u16)a1=size of buffer specified in a0. + * out: (u64)a0=phsical address of buffer passed in from caller. + * (u16)a1=actual bytes from VIF-CONFIG-INFO TLV, or + * 0 if no VIF-CONFIG-INFO TLV was ever received. */ + CMD_CONFIG_INFO_GET = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_ALL, 44), + + /* INT13 API: (u64)a0=paddr to vnic_int13_params struct + * (u32)a1=INT13_CMD_xxx + */ + CMD_INT13_ALL = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 45), + + /* Set default vlan: + * in: (u16)a0=new default vlan + * (u16)a1=zero for overriding vlan with param a0, + * non-zero for resetting vlan to the default + * out: (u16)a0=old default vlan + */ + CMD_SET_DEFAULT_VLAN = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_ALL, 46), + + /* init_prov_info2: + * Variant of CMD_INIT_PROV_INFO, where it will not try to enable + * the vnic until CMD_ENABLE2 is issued. + * (u64)a0=paddr of vnic_devcmd_provinfo + * (u32)a1=sizeof provision info + */ + CMD_INIT_PROV_INFO2 = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 47), + + /* enable2: + * (u32)a0=0 ==> standby + * =CMD_ENABLE2_ACTIVE ==> active + */ + CMD_ENABLE2 = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 48), + + /* + * cmd_status: + * Returns the status of the specified command + * Input: + * a0 = command for which status is being queried. + * Possible values are: + * CMD_SOFT_RESET + * CMD_HANG_RESET + * CMD_OPEN + * CMD_INIT + * CMD_INIT_PROV_INFO + * CMD_DEINIT + * CMD_INIT_PROV_INFO2 + * CMD_ENABLE2 + * Output: + * if status == STAT_ERROR + * a0 = ERR_ENOTSUPPORTED - status for command in a0 is + * not supported + * if status == STAT_NONE + * a0 = status of the devcmd specified in a0 as follows. + * ERR_SUCCESS - command in a0 completed successfully + * ERR_EINPROGRESS - command in a0 is still in progress + */ + CMD_STATUS = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_ALL, 49), + + /* + * Returns interrupt coalescing timer conversion factors. + * After calling this devcmd, ENIC driver can convert + * interrupt coalescing timer in usec into CPU cycles as follows: + * + * intr_timer_cycles = intr_timer_usec * multiplier / divisor + * + * Interrupt coalescing timer in usecs can be obtained from + * CPU cycles as follows: + * + * intr_timer_usec = intr_timer_cycles * divisor / multiplier + * + * in: none + * out: (u32)a0 = multiplier + * (u32)a1 = divisor + * (u32)a2 = maximum timer value in usec + */ + CMD_INTR_COAL_CONVERT = _CMDC(_CMD_DIR_READ, _CMD_VTYPE_ALL, 50), + + /* + * Set the predefined mac address as default + * in: + * (u48)a0 = mac addr + */ + CMD_SET_MAC_ADDR = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 55), + + /* Update the provisioning info of the given VIF + * (u64)a0=paddr of vnic_devcmd_provinfo + * (u32)a1=sizeof provision info + */ + CMD_PROV_INFO_UPDATE = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 56), + + /* Initialization for the devcmd2 interface. + * in: (u64) a0 = host result buffer physical address + * in: (u16) a1 = number of entries in result buffer + */ + CMD_INITIALIZE_DEVCMD2 = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ALL, 57), + + /* Add a filter. + * in: (u64) a0= filter address + * (u32) a1= size of filter + * out: (u32) a0=filter identifier + */ + CMD_ADD_FILTER = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_ENET, 58), + + /* Delete a filter. + * in: (u32) a0=filter identifier + */ + CMD_DEL_FILTER = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 59), + + /* Enable a Queue Pair in User space NIC + * in: (u32) a0=Queue Pair number + * (u32) a1= command + */ + CMD_QP_ENABLE = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 60), + + /* Disable a Queue Pair in User space NIC + * in: (u32) a0=Queue Pair number + * (u32) a1= command + */ + CMD_QP_DISABLE = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 61), + + /* Stats dump Queue Pair in User space NIC + * in: (u32) a0=Queue Pair number + * (u64) a1=host buffer addr for status dump + * (u32) a2=length of the buffer + */ + CMD_QP_STATS_DUMP = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 62), + + /* Clear stats for Queue Pair in User space NIC + * in: (u32) a0=Queue Pair number + */ + CMD_QP_STATS_CLEAR = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 63), + + /* Use this devcmd for agreeing on the highest common version supported + * by both driver and fw for features who need such a facility. + * in: (u64) a0 = feature (driver requests for the supported versions + * on this feature) + * out: (u64) a0 = bitmap of all supported versions for that feature + */ + CMD_GET_SUPP_FEATURE_VER = _CMDC(_CMD_DIR_RW, _CMD_VTYPE_ENET, 69), + + /* Control (Enable/Disable) overlay offloads on the given vnic + * in: (u8) a0 = OVERLAY_FEATURE_NVGRE : NVGRE + * a0 = OVERLAY_FEATURE_VXLAN : VxLAN + * in: (u8) a1 = OVERLAY_OFFLOAD_ENABLE : Enable or + * a1 = OVERLAY_OFFLOAD_DISABLE : Disable or + * a1 = OVERLAY_OFFLOAD_ENABLE_V2 : Enable with version 2 + */ + CMD_OVERLAY_OFFLOAD_CTRL = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 72), + + /* Configuration of overlay offloads feature on a given vNIC + * in: (u8) a0 = DEVCMD_OVERLAY_NVGRE : NVGRE + * a0 = DEVCMD_OVERLAY_VXLAN : VxLAN + * in: (u8) a1 = VXLAN_PORT_UPDATE : VxLAN + * in: (u16) a2 = unsigned short int port information + */ + CMD_OVERLAY_OFFLOAD_CFG = _CMDC(_CMD_DIR_WRITE, _CMD_VTYPE_ENET, 73), +}; + +/* CMD_ENABLE2 flags */ +#define CMD_ENABLE2_STANDBY 0x0 +#define CMD_ENABLE2_ACTIVE 0x1 + +/* flags for CMD_OPEN */ +#define CMD_OPENF_OPROM 0x1 /* open coming from option rom */ +#define CMD_OPENF_IG_DESCCACHE 0x2 /* Do not flush IG DESC cache */ + +/* flags for CMD_INIT */ +#define CMD_INITF_DEFAULT_MAC 0x1 /* init with default mac addr */ + +/* flags for CMD_PACKET_FILTER */ +#define CMD_PFILTER_DIRECTED 0x01 +#define CMD_PFILTER_MULTICAST 0x02 +#define CMD_PFILTER_BROADCAST 0x04 +#define CMD_PFILTER_PROMISCUOUS 0x08 +#define CMD_PFILTER_ALL_MULTICAST 0x10 + +/* Commands for CMD_QP_ENABLE/CM_QP_DISABLE */ +#define CMD_QP_RQWQ 0x0 + +/* rewrite modes for CMD_IG_VLAN_REWRITE_MODE */ +#define IG_VLAN_REWRITE_MODE_DEFAULT_TRUNK 0 +#define IG_VLAN_REWRITE_MODE_UNTAG_DEFAULT_VLAN 1 +#define IG_VLAN_REWRITE_MODE_PRIORITY_TAG_DEFAULT_VLAN 2 +#define IG_VLAN_REWRITE_MODE_PASS_THRU 3 + +enum vnic_devcmd_status { + STAT_NONE = 0, + STAT_BUSY = 1 << 0, /* cmd in progress */ + STAT_ERROR = 1 << 1, /* last cmd caused error (code in a0) */ +}; + +enum vnic_devcmd_error { + ERR_SUCCESS = 0, + ERR_EINVAL = 1, + ERR_EFAULT = 2, + ERR_EPERM = 3, + ERR_EBUSY = 4, + ERR_ECMDUNKNOWN = 5, + ERR_EBADSTATE = 6, + ERR_ENOMEM = 7, + ERR_ETIMEDOUT = 8, + ERR_ELINKDOWN = 9, + ERR_EMAXRES = 10, + ERR_ENOTSUPPORTED = 11, + ERR_EINPROGRESS = 12, + ERR_MAX +}; + +/* + * note: hw_version and asic_rev refer to the same thing, + * but have different formats. hw_version is + * a 32-byte string (e.g. "A2") and asic_rev is + * a 16-bit integer (e.g. 0xA2). + */ +struct vnic_devcmd_fw_info { + char fw_version[32]; + char fw_build[32]; + char hw_version[32]; + char hw_serial_number[32]; + u16 asic_type; + u16 asic_rev; +}; + +struct vnic_devcmd_notify { + u32 csum; /* checksum over following words */ + + u32 link_state; /* link up == 1 */ + u32 port_speed; /* effective port speed (rate limit) */ + u32 mtu; /* MTU */ + u32 msglvl; /* requested driver msg lvl */ + u32 uif; /* uplink interface */ + u32 status; /* status bits (see VNIC_STF_*) */ + u32 error; /* error code (see ERR_*) for first ERR */ + u32 link_down_cnt; /* running count of link down transitions */ + u32 perbi_rebuild_cnt; /* running count of perbi rebuilds */ +}; +#define VNIC_STF_FATAL_ERR 0x0001 /* fatal fw error */ +#define VNIC_STF_STD_PAUSE 0x0002 /* standard link-level pause on */ +#define VNIC_STF_PFC_PAUSE 0x0004 /* priority flow control pause on */ +/* all supported status flags */ +#define VNIC_STF_ALL (VNIC_STF_FATAL_ERR |\ + VNIC_STF_STD_PAUSE |\ + VNIC_STF_PFC_PAUSE |\ + 0) + +struct vnic_devcmd_provinfo { + u8 oui[3]; + u8 type; + u8 data[0]; +}; + +/* These are used in flags field of different filters to denote + * valid fields used. + */ +#define FILTER_FIELD_VALID(fld) (1 << (fld - 1)) + +#define FILTER_FIELDS_USNIC ( \ + FILTER_FIELD_VALID(1) | \ + FILTER_FIELD_VALID(2) | \ + FILTER_FIELD_VALID(3) | \ + FILTER_FIELD_VALID(4)) + +#define FILTER_FIELDS_IPV4_5TUPLE ( \ + FILTER_FIELD_VALID(1) | \ + FILTER_FIELD_VALID(2) | \ + FILTER_FIELD_VALID(3) | \ + FILTER_FIELD_VALID(4) | \ + FILTER_FIELD_VALID(5)) + +#define FILTER_FIELDS_MAC_VLAN ( \ + FILTER_FIELD_VALID(1) | \ + FILTER_FIELD_VALID(2)) + +#define FILTER_FIELD_USNIC_VLAN FILTER_FIELD_VALID(1) +#define FILTER_FIELD_USNIC_ETHTYPE FILTER_FIELD_VALID(2) +#define FILTER_FIELD_USNIC_PROTO FILTER_FIELD_VALID(3) +#define FILTER_FIELD_USNIC_ID FILTER_FIELD_VALID(4) + +struct filter_usnic_id { + u32 flags; + u16 vlan; + u16 ethtype; + u8 proto_version; + u32 usnic_id; +} __packed; + +#define FILTER_FIELD_5TUP_PROTO FILTER_FIELD_VALID(1) +#define FILTER_FIELD_5TUP_SRC_AD FILTER_FIELD_VALID(2) +#define FILTER_FIELD_5TUP_DST_AD FILTER_FIELD_VALID(3) +#define FILTER_FIELD_5TUP_SRC_PT FILTER_FIELD_VALID(4) +#define FILTER_FIELD_5TUP_DST_PT FILTER_FIELD_VALID(5) + +/* Enums for the protocol field. */ +enum protocol_e { + PROTO_UDP = 0, + PROTO_TCP = 1, +}; + +struct filter_ipv4_5tuple { + u32 flags; + u32 protocol; + u32 src_addr; + u32 dst_addr; + u16 src_port; + u16 dst_port; +} __packed; + +#define FILTER_FIELD_VMQ_VLAN FILTER_FIELD_VALID(1) +#define FILTER_FIELD_VMQ_MAC FILTER_FIELD_VALID(2) + +struct filter_mac_vlan { + u32 flags; + u16 vlan; + u8 mac_addr[6]; +} __packed; + +/* Specifies the filter_action type. */ +enum { + FILTER_ACTION_RQ_STEERING = 0, + FILTER_ACTION_MAX +}; + +struct filter_action { + u32 type; + union { + u32 rq_idx; + } u; +} __packed; + +/* Specifies the filter type. */ +enum filter_type { + FILTER_USNIC_ID = 0, + FILTER_IPV4_5TUPLE = 1, + FILTER_MAC_VLAN = 2, + FILTER_MAX +}; + +struct filter { + u32 type; + union { + struct filter_usnic_id usnic; + struct filter_ipv4_5tuple ipv4; + struct filter_mac_vlan mac_vlan; + } u; +} __packed; + +enum { + CLSF_TLV_FILTER = 0, + CLSF_TLV_ACTION = 1, +}; + +/* Maximum size of buffer to CMD_ADD_FILTER */ +#define FILTER_MAX_BUF_SIZE 100 + +struct filter_tlv { + u_int32_t type; + u_int32_t length; + u_int32_t val[0]; +}; + +enum { + CLSF_ADD = 0, + CLSF_DEL = 1, +}; + +/* + * Writing cmd register causes STAT_BUSY to get set in status register. + * When cmd completes, STAT_BUSY will be cleared. + * + * If cmd completed successfully STAT_ERROR will be clear + * and args registers contain cmd-specific results. + * + * If cmd error, STAT_ERROR will be set and args[0] contains error code. + * + * status register is read-only. While STAT_BUSY is set, + * all other register contents are read-only. + */ + +/* Make sizeof(vnic_devcmd) a power-of-2 for I/O BAR. */ +#define VNIC_DEVCMD_NARGS 15 +struct vnic_devcmd { + u32 status; /* RO */ + u32 cmd; /* RW */ + u64 args[VNIC_DEVCMD_NARGS]; /* RW cmd args (little-endian) */ +}; + +#define DEVCMD2_FNORESULT 0x1 /* Don't copy result to host */ + +#define VNIC_DEVCMD2_NARGS VNIC_DEVCMD_NARGS +struct vnic_devcmd2 { + u16 pad; + u16 flags; + u32 cmd; + u64 args[VNIC_DEVCMD2_NARGS]; +}; + +#define VNIC_DEVCMD2_NRESULTS VNIC_DEVCMD_NARGS +struct devcmd2_result { + u64 results[VNIC_DEVCMD2_NRESULTS]; + u32 pad; + u16 completed_index; + u8 error; + u8 color; +}; + +#define DEVCMD2_RING_SIZE 32 +#define DEVCMD2_DESC_SIZE 128 + +enum overlay_feature_t { + OVERLAY_FEATURE_NVGRE = 1, + OVERLAY_FEATURE_VXLAN, + OVERLAY_FEATURE_MAX, +}; + +enum overlay_ofld_cmd { + OVERLAY_OFFLOAD_ENABLE, + OVERLAY_OFFLOAD_DISABLE, + OVERLAY_OFFLOAD_ENABLE_P2, + OVERLAY_OFFLOAD_MAX, +}; + +#define OVERLAY_CFG_VXLAN_PORT_UPDATE 0 + +#define ENIC_VXLAN_INNER_IPV6 BIT(0) +#define ENIC_VXLAN_OUTER_IPV6 BIT(1) +#define ENIC_VXLAN_MULTI_WQ BIT(2) + +/* Use this enum to get the supported versions for each of these features + * If you need to use the devcmd_get_supported_feature_version(), add + * the new feature into this enum and install function handler in devcmd.c + */ +enum vic_feature_t { + VIC_FEATURE_VXLAN, + VIC_FEATURE_RDMA, + VIC_FEATURE_VXLAN_PATCH, + VIC_FEATURE_MAX, +}; + +#endif /* _VNIC_DEVCMD_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/vnic_enet.h b/drivers/net/ethernet/cisco/enic/vnic_enet.h new file mode 100644 index 000000000..7d6fbb563 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_enet.h @@ -0,0 +1,60 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _VNIC_ENIC_H_ +#define _VNIC_ENIC_H_ + +/* Device-specific region: enet configuration */ +struct vnic_enet_config { + u32 flags; + u32 wq_desc_count; + u32 rq_desc_count; + u16 mtu; + u16 intr_timer_deprecated; + u8 intr_timer_type; + u8 intr_mode; + char devname[16]; + u32 intr_timer_usec; + u16 loop_tag; + u16 vf_rq_count; + u16 num_arfs; +}; + +#define VENETF_TSO 0x1 /* TSO enabled */ +#define VENETF_LRO 0x2 /* LRO enabled */ +#define VENETF_RXCSUM 0x4 /* RX csum enabled */ +#define VENETF_TXCSUM 0x8 /* TX csum enabled */ +#define VENETF_RSS 0x10 /* RSS enabled */ +#define VENETF_RSSHASH_IPV4 0x20 /* Hash on IPv4 fields */ +#define VENETF_RSSHASH_TCPIPV4 0x40 /* Hash on TCP + IPv4 fields */ +#define VENETF_RSSHASH_IPV6 0x80 /* Hash on IPv6 fields */ +#define VENETF_RSSHASH_TCPIPV6 0x100 /* Hash on TCP + IPv6 fields */ +#define VENETF_RSSHASH_IPV6_EX 0x200 /* Hash on IPv6 extended fields */ +#define VENETF_RSSHASH_TCPIPV6_EX 0x400 /* Hash on TCP + IPv6 ext. fields */ +#define VENETF_LOOP 0x800 /* Loopback enabled */ +#define VENETF_VXLAN 0x10000 /* VxLAN offload */ + +#define VENET_INTR_TYPE_MIN 0 /* Timer specs min interrupt spacing */ +#define VENET_INTR_TYPE_IDLE 1 /* Timer specs idle time before irq */ + +#define VENET_INTR_MODE_ANY 0 /* Try MSI-X, then MSI, then INTx */ +#define VENET_INTR_MODE_MSI 1 /* Try MSI then INTx */ +#define VENET_INTR_MODE_INTX 2 /* Try INTx only */ + +#endif /* _VNIC_ENIC_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/vnic_intr.c b/drivers/net/ethernet/cisco/enic/vnic_intr.c new file mode 100644 index 000000000..23604e3d4 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_intr.c @@ -0,0 +1,70 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/types.h> +#include <linux/pci.h> +#include <linux/delay.h> + +#include "vnic_dev.h" +#include "vnic_intr.h" +#include "enic.h" + +void vnic_intr_free(struct vnic_intr *intr) +{ + intr->ctrl = NULL; +} + +int vnic_intr_alloc(struct vnic_dev *vdev, struct vnic_intr *intr, + unsigned int index) +{ + intr->index = index; + intr->vdev = vdev; + + intr->ctrl = vnic_dev_get_res(vdev, RES_TYPE_INTR_CTRL, index); + if (!intr->ctrl) { + vdev_err(vdev, "Failed to hook INTR[%d].ctrl resource\n", + index); + return -EINVAL; + } + + return 0; +} + +void vnic_intr_init(struct vnic_intr *intr, u32 coalescing_timer, + unsigned int coalescing_type, unsigned int mask_on_assertion) +{ + vnic_intr_coalescing_timer_set(intr, coalescing_timer); + iowrite32(coalescing_type, &intr->ctrl->coalescing_type); + iowrite32(mask_on_assertion, &intr->ctrl->mask_on_assertion); + iowrite32(0, &intr->ctrl->int_credits); +} + +void vnic_intr_coalescing_timer_set(struct vnic_intr *intr, + u32 coalescing_timer) +{ + iowrite32(vnic_dev_intr_coal_timer_usec_to_hw(intr->vdev, + coalescing_timer), &intr->ctrl->coalescing_timer); +} + +void vnic_intr_clean(struct vnic_intr *intr) +{ + iowrite32(0, &intr->ctrl->int_credits); +} diff --git a/drivers/net/ethernet/cisco/enic/vnic_intr.h b/drivers/net/ethernet/cisco/enic/vnic_intr.h new file mode 100644 index 000000000..2b1636392 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_intr.h @@ -0,0 +1,111 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _VNIC_INTR_H_ +#define _VNIC_INTR_H_ + +#include <linux/pci.h> + +#include "vnic_dev.h" + +#define VNIC_INTR_TIMER_TYPE_ABS 0 +#define VNIC_INTR_TIMER_TYPE_QUIET 1 + +/* Interrupt control */ +struct vnic_intr_ctrl { + u32 coalescing_timer; /* 0x00 */ + u32 pad0; + u32 coalescing_value; /* 0x08 */ + u32 pad1; + u32 coalescing_type; /* 0x10 */ + u32 pad2; + u32 mask_on_assertion; /* 0x18 */ + u32 pad3; + u32 mask; /* 0x20 */ + u32 pad4; + u32 int_credits; /* 0x28 */ + u32 pad5; + u32 int_credit_return; /* 0x30 */ + u32 pad6; +}; + +struct vnic_intr { + unsigned int index; + struct vnic_dev *vdev; + struct vnic_intr_ctrl __iomem *ctrl; /* memory-mapped */ +}; + +static inline void vnic_intr_unmask(struct vnic_intr *intr) +{ + iowrite32(0, &intr->ctrl->mask); +} + +static inline void vnic_intr_mask(struct vnic_intr *intr) +{ + iowrite32(1, &intr->ctrl->mask); +} + +static inline int vnic_intr_masked(struct vnic_intr *intr) +{ + return ioread32(&intr->ctrl->mask); +} + +static inline void vnic_intr_return_credits(struct vnic_intr *intr, + unsigned int credits, int unmask, int reset_timer) +{ +#define VNIC_INTR_UNMASK_SHIFT 16 +#define VNIC_INTR_RESET_TIMER_SHIFT 17 + + u32 int_credit_return = (credits & 0xffff) | + (unmask ? (1 << VNIC_INTR_UNMASK_SHIFT) : 0) | + (reset_timer ? (1 << VNIC_INTR_RESET_TIMER_SHIFT) : 0); + + iowrite32(int_credit_return, &intr->ctrl->int_credit_return); +} + +static inline unsigned int vnic_intr_credits(struct vnic_intr *intr) +{ + return ioread32(&intr->ctrl->int_credits); +} + +static inline void vnic_intr_return_all_credits(struct vnic_intr *intr) +{ + unsigned int credits = vnic_intr_credits(intr); + int unmask = 1; + int reset_timer = 1; + + vnic_intr_return_credits(intr, credits, unmask, reset_timer); +} + +static inline u32 vnic_intr_legacy_pba(u32 __iomem *legacy_pba) +{ + /* read PBA without clearing */ + return ioread32(legacy_pba); +} + +void vnic_intr_free(struct vnic_intr *intr); +int vnic_intr_alloc(struct vnic_dev *vdev, struct vnic_intr *intr, + unsigned int index); +void vnic_intr_init(struct vnic_intr *intr, u32 coalescing_timer, + unsigned int coalescing_type, unsigned int mask_on_assertion); +void vnic_intr_coalescing_timer_set(struct vnic_intr *intr, + u32 coalescing_timer); +void vnic_intr_clean(struct vnic_intr *intr); + +#endif /* _VNIC_INTR_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/vnic_nic.h b/drivers/net/ethernet/cisco/enic/vnic_nic.h new file mode 100644 index 000000000..84ff8ca17 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_nic.h @@ -0,0 +1,74 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _VNIC_NIC_H_ +#define _VNIC_NIC_H_ + +#define NIC_CFG_RSS_DEFAULT_CPU_MASK_FIELD 0xffUL +#define NIC_CFG_RSS_DEFAULT_CPU_SHIFT 0 +#define NIC_CFG_RSS_HASH_TYPE (0xffUL << 8) +#define NIC_CFG_RSS_HASH_TYPE_MASK_FIELD 0xffUL +#define NIC_CFG_RSS_HASH_TYPE_SHIFT 8 +#define NIC_CFG_RSS_HASH_BITS (7UL << 16) +#define NIC_CFG_RSS_HASH_BITS_MASK_FIELD 7UL +#define NIC_CFG_RSS_HASH_BITS_SHIFT 16 +#define NIC_CFG_RSS_BASE_CPU (7UL << 19) +#define NIC_CFG_RSS_BASE_CPU_MASK_FIELD 7UL +#define NIC_CFG_RSS_BASE_CPU_SHIFT 19 +#define NIC_CFG_RSS_ENABLE (1UL << 22) +#define NIC_CFG_RSS_ENABLE_MASK_FIELD 1UL +#define NIC_CFG_RSS_ENABLE_SHIFT 22 +#define NIC_CFG_TSO_IPID_SPLIT_EN (1UL << 23) +#define NIC_CFG_TSO_IPID_SPLIT_EN_MASK_FIELD 1UL +#define NIC_CFG_TSO_IPID_SPLIT_EN_SHIFT 23 +#define NIC_CFG_IG_VLAN_STRIP_EN (1UL << 24) +#define NIC_CFG_IG_VLAN_STRIP_EN_MASK_FIELD 1UL +#define NIC_CFG_IG_VLAN_STRIP_EN_SHIFT 24 + +#define NIC_CFG_RSS_HASH_TYPE_UDP_IPV4 (1 << 0) +#define NIC_CFG_RSS_HASH_TYPE_IPV4 (1 << 1) +#define NIC_CFG_RSS_HASH_TYPE_TCP_IPV4 (1 << 2) +#define NIC_CFG_RSS_HASH_TYPE_IPV6 (1 << 3) +#define NIC_CFG_RSS_HASH_TYPE_TCP_IPV6 (1 << 4) +#define NIC_CFG_RSS_HASH_TYPE_IPV6_EX (1 << 5) +#define NIC_CFG_RSS_HASH_TYPE_TCP_IPV6_EX (1 << 6) +#define NIC_CFG_RSS_HASH_TYPE_UDP_IPV6 (1 << 7) + +static inline void vnic_set_nic_cfg(u32 *nic_cfg, + u8 rss_default_cpu, u8 rss_hash_type, + u8 rss_hash_bits, u8 rss_base_cpu, + u8 rss_enable, u8 tso_ipid_split_en, + u8 ig_vlan_strip_en) +{ + *nic_cfg = (rss_default_cpu & NIC_CFG_RSS_DEFAULT_CPU_MASK_FIELD) | + ((rss_hash_type & NIC_CFG_RSS_HASH_TYPE_MASK_FIELD) + << NIC_CFG_RSS_HASH_TYPE_SHIFT) | + ((rss_hash_bits & NIC_CFG_RSS_HASH_BITS_MASK_FIELD) + << NIC_CFG_RSS_HASH_BITS_SHIFT) | + ((rss_base_cpu & NIC_CFG_RSS_BASE_CPU_MASK_FIELD) + << NIC_CFG_RSS_BASE_CPU_SHIFT) | + ((rss_enable & NIC_CFG_RSS_ENABLE_MASK_FIELD) + << NIC_CFG_RSS_ENABLE_SHIFT) | + ((tso_ipid_split_en & NIC_CFG_TSO_IPID_SPLIT_EN_MASK_FIELD) + << NIC_CFG_TSO_IPID_SPLIT_EN_SHIFT) | + ((ig_vlan_strip_en & NIC_CFG_IG_VLAN_STRIP_EN_MASK_FIELD) + << NIC_CFG_IG_VLAN_STRIP_EN_SHIFT); +} + +#endif /* _VNIC_NIC_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/vnic_resource.h b/drivers/net/ethernet/cisco/enic/vnic_resource.h new file mode 100644 index 000000000..4e45f88ac --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_resource.h @@ -0,0 +1,83 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _VNIC_RESOURCE_H_ +#define _VNIC_RESOURCE_H_ + +#define VNIC_RES_MAGIC 0x766E6963L /* 'vnic' */ +#define VNIC_RES_VERSION 0x00000000L +#define MGMTVNIC_MAGIC 0x544d474dL /* 'MGMT' */ +#define MGMTVNIC_VERSION 0x00000000L + +/* The MAC address assigned to the CFG vNIC is fixed. */ +#define MGMTVNIC_MAC { 0x02, 0x00, 0x54, 0x4d, 0x47, 0x4d } + +/* vNIC resource types */ +enum vnic_res_type { + RES_TYPE_EOL, /* End-of-list */ + RES_TYPE_WQ, /* Work queues */ + RES_TYPE_RQ, /* Receive queues */ + RES_TYPE_CQ, /* Completion queues */ + RES_TYPE_RSVD1, + RES_TYPE_NIC_CFG, /* Enet NIC config registers */ + RES_TYPE_RSVD2, + RES_TYPE_RSVD3, + RES_TYPE_RSVD4, + RES_TYPE_RSVD5, + RES_TYPE_INTR_CTRL, /* Interrupt ctrl table */ + RES_TYPE_INTR_TABLE, /* MSI/MSI-X Interrupt table */ + RES_TYPE_INTR_PBA, /* MSI/MSI-X PBA table */ + RES_TYPE_INTR_PBA_LEGACY, /* Legacy intr status */ + RES_TYPE_RSVD6, + RES_TYPE_RSVD7, + RES_TYPE_DEVCMD, /* Device command region */ + RES_TYPE_PASS_THRU_PAGE, /* Pass-thru page */ + RES_TYPE_SUBVNIC, /* subvnic resource type */ + RES_TYPE_MQ_WQ, /* MQ Work queues */ + RES_TYPE_MQ_RQ, /* MQ Receive queues */ + RES_TYPE_MQ_CQ, /* MQ Completion queues */ + RES_TYPE_DEPRECATED1, /* Old version of devcmd 2 */ + RES_TYPE_DEPRECATED2, /* Old version of devcmd 2 */ + RES_TYPE_DEVCMD2, /* Device control region */ + + RES_TYPE_MAX, /* Count of resource types */ +}; + +struct vnic_resource_header { + u32 magic; + u32 version; +}; + +struct mgmt_barmap_hdr { + u32 magic; /* magic number */ + u32 version; /* header format version */ + u16 lif; /* loopback lif for mgmt frames */ + u16 pci_slot; /* installed pci slot */ + char serial[16]; /* card serial number */ +}; + +struct vnic_resource { + u8 type; + u8 bar; + u8 pad[2]; + u32 bar_offset; + u32 count; +}; + +#endif /* _VNIC_RESOURCE_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/vnic_rq.c b/drivers/net/ethernet/cisco/enic/vnic_rq.c new file mode 100644 index 000000000..a3e7b003a --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_rq.c @@ -0,0 +1,219 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/types.h> +#include <linux/pci.h> +#include <linux/delay.h> +#include <linux/slab.h> + +#include "vnic_dev.h" +#include "vnic_rq.h" +#include "enic.h" + +static int vnic_rq_alloc_bufs(struct vnic_rq *rq) +{ + struct vnic_rq_buf *buf; + unsigned int i, j, count = rq->ring.desc_count; + unsigned int blks = VNIC_RQ_BUF_BLKS_NEEDED(count); + + for (i = 0; i < blks; i++) { + rq->bufs[i] = kzalloc(VNIC_RQ_BUF_BLK_SZ(count), GFP_KERNEL); + if (!rq->bufs[i]) + return -ENOMEM; + } + + for (i = 0; i < blks; i++) { + buf = rq->bufs[i]; + for (j = 0; j < VNIC_RQ_BUF_BLK_ENTRIES(count); j++) { + buf->index = i * VNIC_RQ_BUF_BLK_ENTRIES(count) + j; + buf->desc = (u8 *)rq->ring.descs + + rq->ring.desc_size * buf->index; + if (buf->index + 1 == count) { + buf->next = rq->bufs[0]; + break; + } else if (j + 1 == VNIC_RQ_BUF_BLK_ENTRIES(count)) { + buf->next = rq->bufs[i + 1]; + } else { + buf->next = buf + 1; + buf++; + } + } + } + + rq->to_use = rq->to_clean = rq->bufs[0]; + + return 0; +} + +void vnic_rq_free(struct vnic_rq *rq) +{ + struct vnic_dev *vdev; + unsigned int i; + + vdev = rq->vdev; + + vnic_dev_free_desc_ring(vdev, &rq->ring); + + for (i = 0; i < VNIC_RQ_BUF_BLKS_MAX; i++) { + if (rq->bufs[i]) { + kfree(rq->bufs[i]); + rq->bufs[i] = NULL; + } + } + + rq->ctrl = NULL; +} + +int vnic_rq_alloc(struct vnic_dev *vdev, struct vnic_rq *rq, unsigned int index, + unsigned int desc_count, unsigned int desc_size) +{ + int err; + + rq->index = index; + rq->vdev = vdev; + + rq->ctrl = vnic_dev_get_res(vdev, RES_TYPE_RQ, index); + if (!rq->ctrl) { + vdev_err(vdev, "Failed to hook RQ[%d] resource\n", index); + return -EINVAL; + } + + vnic_rq_disable(rq); + + err = vnic_dev_alloc_desc_ring(vdev, &rq->ring, desc_count, desc_size); + if (err) + return err; + + err = vnic_rq_alloc_bufs(rq); + if (err) { + vnic_rq_free(rq); + return err; + } + + return 0; +} + +static void vnic_rq_init_start(struct vnic_rq *rq, unsigned int cq_index, + unsigned int fetch_index, unsigned int posted_index, + unsigned int error_interrupt_enable, + unsigned int error_interrupt_offset) +{ + u64 paddr; + unsigned int count = rq->ring.desc_count; + + paddr = (u64)rq->ring.base_addr | VNIC_PADDR_TARGET; + writeq(paddr, &rq->ctrl->ring_base); + iowrite32(count, &rq->ctrl->ring_size); + iowrite32(cq_index, &rq->ctrl->cq_index); + iowrite32(error_interrupt_enable, &rq->ctrl->error_interrupt_enable); + iowrite32(error_interrupt_offset, &rq->ctrl->error_interrupt_offset); + iowrite32(0, &rq->ctrl->dropped_packet_count); + iowrite32(0, &rq->ctrl->error_status); + iowrite32(fetch_index, &rq->ctrl->fetch_index); + iowrite32(posted_index, &rq->ctrl->posted_index); + + rq->to_use = rq->to_clean = + &rq->bufs[fetch_index / VNIC_RQ_BUF_BLK_ENTRIES(count)] + [fetch_index % VNIC_RQ_BUF_BLK_ENTRIES(count)]; +} + +void vnic_rq_init(struct vnic_rq *rq, unsigned int cq_index, + unsigned int error_interrupt_enable, + unsigned int error_interrupt_offset) +{ + vnic_rq_init_start(rq, cq_index, 0, 0, error_interrupt_enable, + error_interrupt_offset); +} + +unsigned int vnic_rq_error_status(struct vnic_rq *rq) +{ + return ioread32(&rq->ctrl->error_status); +} + +void vnic_rq_enable(struct vnic_rq *rq) +{ + iowrite32(1, &rq->ctrl->enable); +} + +int vnic_rq_disable(struct vnic_rq *rq) +{ + unsigned int wait; + struct vnic_dev *vdev = rq->vdev; + int i; + + /* Due to a race condition with clearing RQ "mini-cache" in hw, we need + * to disable the RQ twice to guarantee that stale descriptors are not + * used when this RQ is re-enabled. + */ + for (i = 0; i < 2; i++) { + iowrite32(0, &rq->ctrl->enable); + + /* Wait for HW to ACK disable request */ + for (wait = 20000; wait > 0; wait--) + if (!ioread32(&rq->ctrl->running)) + break; + if (!wait) { + vdev_neterr(vdev, "Failed to disable RQ[%d]\n", + rq->index); + + return -ETIMEDOUT; + } + } + + return 0; +} + +void vnic_rq_clean(struct vnic_rq *rq, + void (*buf_clean)(struct vnic_rq *rq, struct vnic_rq_buf *buf)) +{ + struct vnic_rq_buf *buf; + u32 fetch_index; + unsigned int count = rq->ring.desc_count; + int i; + + buf = rq->to_clean; + + for (i = 0; i < rq->ring.desc_count; i++) { + (*buf_clean)(rq, buf); + buf = buf->next; + } + rq->ring.desc_avail = rq->ring.desc_count - 1; + + /* Use current fetch_index as the ring starting point */ + fetch_index = ioread32(&rq->ctrl->fetch_index); + + if (fetch_index == 0xFFFFFFFF) { /* check for hardware gone */ + /* Hardware surprise removal: reset fetch_index */ + fetch_index = 0; + } + rq->to_use = rq->to_clean = + &rq->bufs[fetch_index / VNIC_RQ_BUF_BLK_ENTRIES(count)] + [fetch_index % VNIC_RQ_BUF_BLK_ENTRIES(count)]; + iowrite32(fetch_index, &rq->ctrl->posted_index); + + /* Anytime we write fetch_index, we need to re-write 0 to rq->enable + * to re-sync internal VIC state. + */ + iowrite32(0, &rq->ctrl->enable); + + vnic_dev_clear_desc_ring(&rq->ring); +} + diff --git a/drivers/net/ethernet/cisco/enic/vnic_rq.h b/drivers/net/ethernet/cisco/enic/vnic_rq.h new file mode 100644 index 000000000..0413103eb --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_rq.h @@ -0,0 +1,219 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _VNIC_RQ_H_ +#define _VNIC_RQ_H_ + +#include <linux/pci.h> +#include <linux/netdevice.h> + +#include "vnic_dev.h" +#include "vnic_cq.h" + +/* Receive queue control */ +struct vnic_rq_ctrl { + u64 ring_base; /* 0x00 */ + u32 ring_size; /* 0x08 */ + u32 pad0; + u32 posted_index; /* 0x10 */ + u32 pad1; + u32 cq_index; /* 0x18 */ + u32 pad2; + u32 enable; /* 0x20 */ + u32 pad3; + u32 running; /* 0x28 */ + u32 pad4; + u32 fetch_index; /* 0x30 */ + u32 pad5; + u32 error_interrupt_enable; /* 0x38 */ + u32 pad6; + u32 error_interrupt_offset; /* 0x40 */ + u32 pad7; + u32 error_status; /* 0x48 */ + u32 pad8; + u32 dropped_packet_count; /* 0x50 */ + u32 pad9; + u32 dropped_packet_count_rc; /* 0x58 */ + u32 pad10; +}; + +/* Break the vnic_rq_buf allocations into blocks of 32/64 entries */ +#define VNIC_RQ_BUF_MIN_BLK_ENTRIES 32 +#define VNIC_RQ_BUF_DFLT_BLK_ENTRIES 64 +#define VNIC_RQ_BUF_BLK_ENTRIES(entries) \ + ((unsigned int)((entries < VNIC_RQ_BUF_DFLT_BLK_ENTRIES) ? \ + VNIC_RQ_BUF_MIN_BLK_ENTRIES : VNIC_RQ_BUF_DFLT_BLK_ENTRIES)) +#define VNIC_RQ_BUF_BLK_SZ(entries) \ + (VNIC_RQ_BUF_BLK_ENTRIES(entries) * sizeof(struct vnic_rq_buf)) +#define VNIC_RQ_BUF_BLKS_NEEDED(entries) \ + DIV_ROUND_UP(entries, VNIC_RQ_BUF_BLK_ENTRIES(entries)) +#define VNIC_RQ_BUF_BLKS_MAX VNIC_RQ_BUF_BLKS_NEEDED(4096) + +struct vnic_rq_buf { + struct vnic_rq_buf *next; + dma_addr_t dma_addr; + void *os_buf; + unsigned int os_buf_index; + unsigned int len; + unsigned int index; + void *desc; + uint64_t wr_id; +}; + +enum enic_poll_state { + ENIC_POLL_STATE_IDLE, + ENIC_POLL_STATE_NAPI, + ENIC_POLL_STATE_POLL +}; + +struct vnic_rq { + unsigned int index; + struct vnic_dev *vdev; + struct vnic_rq_ctrl __iomem *ctrl; /* memory-mapped */ + struct vnic_dev_ring ring; + struct vnic_rq_buf *bufs[VNIC_RQ_BUF_BLKS_MAX]; + struct vnic_rq_buf *to_use; + struct vnic_rq_buf *to_clean; + void *os_buf_head; + unsigned int pkts_outstanding; +}; + +static inline unsigned int vnic_rq_desc_avail(struct vnic_rq *rq) +{ + /* how many does SW own? */ + return rq->ring.desc_avail; +} + +static inline unsigned int vnic_rq_desc_used(struct vnic_rq *rq) +{ + /* how many does HW own? */ + return rq->ring.desc_count - rq->ring.desc_avail - 1; +} + +static inline void *vnic_rq_next_desc(struct vnic_rq *rq) +{ + return rq->to_use->desc; +} + +static inline unsigned int vnic_rq_next_index(struct vnic_rq *rq) +{ + return rq->to_use->index; +} + +static inline void vnic_rq_post(struct vnic_rq *rq, + void *os_buf, unsigned int os_buf_index, + dma_addr_t dma_addr, unsigned int len, + uint64_t wrid) +{ + struct vnic_rq_buf *buf = rq->to_use; + + buf->os_buf = os_buf; + buf->os_buf_index = os_buf_index; + buf->dma_addr = dma_addr; + buf->len = len; + buf->wr_id = wrid; + + buf = buf->next; + rq->to_use = buf; + rq->ring.desc_avail--; + + /* Move the posted_index every nth descriptor + */ + +#ifndef VNIC_RQ_RETURN_RATE +#define VNIC_RQ_RETURN_RATE 0xf /* keep 2^n - 1 */ +#endif + + if ((buf->index & VNIC_RQ_RETURN_RATE) == 0) { + /* Adding write memory barrier prevents compiler and/or CPU + * reordering, thus avoiding descriptor posting before + * descriptor is initialized. Otherwise, hardware can read + * stale descriptor fields. + */ + wmb(); + iowrite32(buf->index, &rq->ctrl->posted_index); + } +} + +static inline void vnic_rq_return_descs(struct vnic_rq *rq, unsigned int count) +{ + rq->ring.desc_avail += count; +} + +enum desc_return_options { + VNIC_RQ_RETURN_DESC, + VNIC_RQ_DEFER_RETURN_DESC, +}; + +static inline void vnic_rq_service(struct vnic_rq *rq, + struct cq_desc *cq_desc, u16 completed_index, + int desc_return, void (*buf_service)(struct vnic_rq *rq, + struct cq_desc *cq_desc, struct vnic_rq_buf *buf, + int skipped, void *opaque), void *opaque) +{ + struct vnic_rq_buf *buf; + int skipped; + + buf = rq->to_clean; + while (1) { + + skipped = (buf->index != completed_index); + + (*buf_service)(rq, cq_desc, buf, skipped, opaque); + + if (desc_return == VNIC_RQ_RETURN_DESC) + rq->ring.desc_avail++; + + rq->to_clean = buf->next; + + if (!skipped) + break; + + buf = rq->to_clean; + } +} + +static inline int vnic_rq_fill(struct vnic_rq *rq, + int (*buf_fill)(struct vnic_rq *rq)) +{ + int err; + + while (vnic_rq_desc_avail(rq) > 0) { + + err = (*buf_fill)(rq); + if (err) + return err; + } + + return 0; +} + +void vnic_rq_free(struct vnic_rq *rq); +int vnic_rq_alloc(struct vnic_dev *vdev, struct vnic_rq *rq, unsigned int index, + unsigned int desc_count, unsigned int desc_size); +void vnic_rq_init(struct vnic_rq *rq, unsigned int cq_index, + unsigned int error_interrupt_enable, + unsigned int error_interrupt_offset); +unsigned int vnic_rq_error_status(struct vnic_rq *rq); +void vnic_rq_enable(struct vnic_rq *rq); +int vnic_rq_disable(struct vnic_rq *rq); +void vnic_rq_clean(struct vnic_rq *rq, + void (*buf_clean)(struct vnic_rq *rq, struct vnic_rq_buf *buf)); + +#endif /* _VNIC_RQ_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/vnic_rss.h b/drivers/net/ethernet/cisco/enic/vnic_rss.h new file mode 100644 index 000000000..881fa1854 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_rss.h @@ -0,0 +1,45 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +#ifndef _VNIC_RSS_H_ +#define _VNIC_RSS_H_ + +/* RSS key array */ + +#define ENIC_RSS_BYTES_PER_KEY 10 +#define ENIC_RSS_KEYS 4 +#define ENIC_RSS_LEN (ENIC_RSS_BYTES_PER_KEY * ENIC_RSS_KEYS) + +union vnic_rss_key { + struct { + u8 b[ENIC_RSS_BYTES_PER_KEY]; + u8 b_pad[6]; + } key[ENIC_RSS_KEYS]; + u64 raw[8]; +}; + +/* RSS cpu array */ +union vnic_rss_cpu { + struct { + u8 b[4] ; + u8 b_pad[4]; + } cpu[32]; + u64 raw[32]; +}; + +#endif /* _VNIC_RSS_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/vnic_stats.h b/drivers/net/ethernet/cisco/enic/vnic_stats.h new file mode 100644 index 000000000..74c81ed6f --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_stats.h @@ -0,0 +1,75 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _VNIC_STATS_H_ +#define _VNIC_STATS_H_ + +/* Tx statistics */ +struct vnic_tx_stats { + u64 tx_frames_ok; + u64 tx_unicast_frames_ok; + u64 tx_multicast_frames_ok; + u64 tx_broadcast_frames_ok; + u64 tx_bytes_ok; + u64 tx_unicast_bytes_ok; + u64 tx_multicast_bytes_ok; + u64 tx_broadcast_bytes_ok; + u64 tx_drops; + u64 tx_errors; + u64 tx_tso; + u64 rsvd[16]; +}; + +/* Rx statistics */ +struct vnic_rx_stats { + u64 rx_frames_ok; + u64 rx_frames_total; + u64 rx_unicast_frames_ok; + u64 rx_multicast_frames_ok; + u64 rx_broadcast_frames_ok; + u64 rx_bytes_ok; + u64 rx_unicast_bytes_ok; + u64 rx_multicast_bytes_ok; + u64 rx_broadcast_bytes_ok; + u64 rx_drop; + u64 rx_no_bufs; + u64 rx_errors; + u64 rx_rss; + u64 rx_crc_errors; + u64 rx_frames_64; + u64 rx_frames_127; + u64 rx_frames_255; + u64 rx_frames_511; + u64 rx_frames_1023; + u64 rx_frames_1518; + u64 rx_frames_to_max; + u64 rsvd[16]; +}; + +/* Generic statistics */ +struct vnic_gen_stats { + u64 dma_map_error; +}; + +struct vnic_stats { + struct vnic_tx_stats tx; + struct vnic_rx_stats rx; +}; + +#endif /* _VNIC_STATS_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/vnic_vic.c b/drivers/net/ethernet/cisco/enic/vnic_vic.c new file mode 100644 index 000000000..24ef8cd40 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_vic.c @@ -0,0 +1,79 @@ +/* + * Copyright 2010 Cisco Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/types.h> +#include <linux/slab.h> + +#include "vnic_vic.h" + +struct vic_provinfo *vic_provinfo_alloc(gfp_t flags, const u8 *oui, + const u8 type) +{ + struct vic_provinfo *vp; + + if (!oui) + return NULL; + + vp = kzalloc(VIC_PROVINFO_MAX_DATA, flags); + if (!vp) + return NULL; + + memcpy(vp->oui, oui, sizeof(vp->oui)); + vp->type = type; + vp->length = htonl(sizeof(vp->num_tlvs)); + + return vp; +} + +void vic_provinfo_free(struct vic_provinfo *vp) +{ + kfree(vp); +} + +int vic_provinfo_add_tlv(struct vic_provinfo *vp, u16 type, u16 length, + const void *value) +{ + struct vic_provinfo_tlv *tlv; + + if (!vp || !value) + return -EINVAL; + + if (ntohl(vp->length) + offsetof(struct vic_provinfo_tlv, value) + + length > VIC_PROVINFO_MAX_TLV_DATA) + return -ENOMEM; + + tlv = (struct vic_provinfo_tlv *)((u8 *)vp->tlv + + ntohl(vp->length) - sizeof(vp->num_tlvs)); + + tlv->type = htons(type); + tlv->length = htons(length); + memcpy(tlv->value, value, length); + + vp->num_tlvs = htonl(ntohl(vp->num_tlvs) + 1); + vp->length = htonl(ntohl(vp->length) + + offsetof(struct vic_provinfo_tlv, value) + length); + + return 0; +} + +size_t vic_provinfo_size(struct vic_provinfo *vp) +{ + return vp ? ntohl(vp->length) + sizeof(*vp) - sizeof(vp->num_tlvs) : 0; +} diff --git a/drivers/net/ethernet/cisco/enic/vnic_vic.h b/drivers/net/ethernet/cisco/enic/vnic_vic.h new file mode 100644 index 000000000..9ef81f148 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_vic.h @@ -0,0 +1,83 @@ +/* + * Copyright 2010 Cisco Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _VNIC_VIC_H_ +#define _VNIC_VIC_H_ + +/* Note: All integer fields in NETWORK byte order */ + +/* Note: String field lengths include null char */ + +#define VIC_PROVINFO_CISCO_OUI { 0x00, 0x00, 0x0c } +#define VIC_PROVINFO_GENERIC_TYPE 0x4 + +enum vic_generic_prov_tlv_type { + VIC_GENERIC_PROV_TLV_PORT_PROFILE_NAME_STR = 0, + VIC_GENERIC_PROV_TLV_CLIENT_MAC_ADDR = 1, + VIC_GENERIC_PROV_TLV_CLIENT_NAME_STR = 2, + VIC_GENERIC_PROV_TLV_CLUSTER_PORT_NAME_STR = 3, + VIC_GENERIC_PROV_TLV_CLUSTER_PORT_UUID_STR = 4, + VIC_GENERIC_PROV_TLV_CLUSTER_UUID_STR = 5, + VIC_GENERIC_PROV_TLV_CLUSTER_NAME_STR = 7, + VIC_GENERIC_PROV_TLV_HOST_UUID_STR = 8, + VIC_GENERIC_PROV_TLV_CLIENT_UUID_STR = 9, + VIC_GENERIC_PROV_TLV_INCARNATION_NUMBER = 10, + VIC_GENERIC_PROV_TLV_OS_TYPE = 11, + VIC_GENERIC_PROV_TLV_OS_VENDOR = 12, + VIC_GENERIC_PROV_TLV_CLIENT_TYPE = 15, +}; + +enum vic_generic_prov_os_type { + VIC_GENERIC_PROV_OS_TYPE_UNKNOWN = 0, + VIC_GENERIC_PROV_OS_TYPE_ESX = 1, + VIC_GENERIC_PROV_OS_TYPE_LINUX = 2, + VIC_GENERIC_PROV_OS_TYPE_WINDOWS = 3, + VIC_GENERIC_PROV_OS_TYPE_SOLARIS = 4, +}; + +struct vic_provinfo { + u8 oui[3]; /* OUI of data provider */ + u8 type; /* provider-specific type */ + u32 length; /* length of data below */ + u32 num_tlvs; /* number of tlvs */ + struct vic_provinfo_tlv { + u16 type; + u16 length; + u8 value[0]; + } tlv[0]; +} __packed; + +#define VIC_PROVINFO_ADD_TLV(vp, tlvtype, tlvlen, data) \ + do { \ + err = vic_provinfo_add_tlv(vp, tlvtype, tlvlen, data); \ + if (err) \ + goto add_tlv_failure; \ + } while (0) + +#define VIC_PROVINFO_MAX_DATA 1385 +#define VIC_PROVINFO_MAX_TLV_DATA (VIC_PROVINFO_MAX_DATA - \ + sizeof(struct vic_provinfo)) + +struct vic_provinfo *vic_provinfo_alloc(gfp_t flags, const u8 *oui, + const u8 type); +void vic_provinfo_free(struct vic_provinfo *vp); +int vic_provinfo_add_tlv(struct vic_provinfo *vp, u16 type, u16 length, + const void *value); +size_t vic_provinfo_size(struct vic_provinfo *vp); + +#endif /* _VNIC_VIC_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/vnic_wq.c b/drivers/net/ethernet/cisco/enic/vnic_wq.c new file mode 100644 index 000000000..eb7589197 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_wq.c @@ -0,0 +1,217 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/types.h> +#include <linux/pci.h> +#include <linux/delay.h> +#include <linux/slab.h> + +#include "vnic_dev.h" +#include "vnic_wq.h" +#include "enic.h" + +static int vnic_wq_alloc_bufs(struct vnic_wq *wq) +{ + struct vnic_wq_buf *buf; + unsigned int i, j, count = wq->ring.desc_count; + unsigned int blks = VNIC_WQ_BUF_BLKS_NEEDED(count); + + for (i = 0; i < blks; i++) { + wq->bufs[i] = kzalloc(VNIC_WQ_BUF_BLK_SZ(count), GFP_KERNEL); + if (!wq->bufs[i]) + return -ENOMEM; + } + + for (i = 0; i < blks; i++) { + buf = wq->bufs[i]; + for (j = 0; j < VNIC_WQ_BUF_BLK_ENTRIES(count); j++) { + buf->index = i * VNIC_WQ_BUF_BLK_ENTRIES(count) + j; + buf->desc = (u8 *)wq->ring.descs + + wq->ring.desc_size * buf->index; + if (buf->index + 1 == count) { + buf->next = wq->bufs[0]; + buf->next->prev = buf; + break; + } else if (j + 1 == VNIC_WQ_BUF_BLK_ENTRIES(count)) { + buf->next = wq->bufs[i + 1]; + buf->next->prev = buf; + } else { + buf->next = buf + 1; + buf->next->prev = buf; + buf++; + } + } + } + + wq->to_use = wq->to_clean = wq->bufs[0]; + + return 0; +} + +void vnic_wq_free(struct vnic_wq *wq) +{ + struct vnic_dev *vdev; + unsigned int i; + + vdev = wq->vdev; + + vnic_dev_free_desc_ring(vdev, &wq->ring); + + for (i = 0; i < VNIC_WQ_BUF_BLKS_MAX; i++) { + if (wq->bufs[i]) { + kfree(wq->bufs[i]); + wq->bufs[i] = NULL; + } + } + + wq->ctrl = NULL; +} + +int vnic_wq_alloc(struct vnic_dev *vdev, struct vnic_wq *wq, unsigned int index, + unsigned int desc_count, unsigned int desc_size) +{ + int err; + + wq->index = index; + wq->vdev = vdev; + + wq->ctrl = vnic_dev_get_res(vdev, RES_TYPE_WQ, index); + if (!wq->ctrl) { + vdev_err(vdev, "Failed to hook WQ[%d] resource\n", index); + return -EINVAL; + } + + vnic_wq_disable(wq); + + err = vnic_dev_alloc_desc_ring(vdev, &wq->ring, desc_count, desc_size); + if (err) + return err; + + err = vnic_wq_alloc_bufs(wq); + if (err) { + vnic_wq_free(wq); + return err; + } + + return 0; +} + +int enic_wq_devcmd2_alloc(struct vnic_dev *vdev, struct vnic_wq *wq, + unsigned int desc_count, unsigned int desc_size) +{ + int err; + + wq->index = 0; + wq->vdev = vdev; + + wq->ctrl = vnic_dev_get_res(vdev, RES_TYPE_DEVCMD2, 0); + if (!wq->ctrl) + return -EINVAL; + vnic_wq_disable(wq); + err = vnic_dev_alloc_desc_ring(vdev, &wq->ring, desc_count, desc_size); + + return err; +} + +void enic_wq_init_start(struct vnic_wq *wq, unsigned int cq_index, + unsigned int fetch_index, unsigned int posted_index, + unsigned int error_interrupt_enable, + unsigned int error_interrupt_offset) +{ + u64 paddr; + unsigned int count = wq->ring.desc_count; + + paddr = (u64)wq->ring.base_addr | VNIC_PADDR_TARGET; + writeq(paddr, &wq->ctrl->ring_base); + iowrite32(count, &wq->ctrl->ring_size); + iowrite32(fetch_index, &wq->ctrl->fetch_index); + iowrite32(posted_index, &wq->ctrl->posted_index); + iowrite32(cq_index, &wq->ctrl->cq_index); + iowrite32(error_interrupt_enable, &wq->ctrl->error_interrupt_enable); + iowrite32(error_interrupt_offset, &wq->ctrl->error_interrupt_offset); + iowrite32(0, &wq->ctrl->error_status); + + wq->to_use = wq->to_clean = + &wq->bufs[fetch_index / VNIC_WQ_BUF_BLK_ENTRIES(count)] + [fetch_index % VNIC_WQ_BUF_BLK_ENTRIES(count)]; +} + +void vnic_wq_init(struct vnic_wq *wq, unsigned int cq_index, + unsigned int error_interrupt_enable, + unsigned int error_interrupt_offset) +{ + enic_wq_init_start(wq, cq_index, 0, 0, + error_interrupt_enable, + error_interrupt_offset); +} + +unsigned int vnic_wq_error_status(struct vnic_wq *wq) +{ + return ioread32(&wq->ctrl->error_status); +} + +void vnic_wq_enable(struct vnic_wq *wq) +{ + iowrite32(1, &wq->ctrl->enable); +} + +int vnic_wq_disable(struct vnic_wq *wq) +{ + unsigned int wait; + struct vnic_dev *vdev = wq->vdev; + + iowrite32(0, &wq->ctrl->enable); + + /* Wait for HW to ACK disable request */ + for (wait = 0; wait < 1000; wait++) { + if (!(ioread32(&wq->ctrl->running))) + return 0; + udelay(10); + } + + vdev_neterr(vdev, "Failed to disable WQ[%d]\n", wq->index); + + return -ETIMEDOUT; +} + +void vnic_wq_clean(struct vnic_wq *wq, + void (*buf_clean)(struct vnic_wq *wq, struct vnic_wq_buf *buf)) +{ + struct vnic_wq_buf *buf; + + buf = wq->to_clean; + + while (vnic_wq_desc_used(wq) > 0) { + + (*buf_clean)(wq, buf); + + buf = wq->to_clean = buf->next; + wq->ring.desc_avail++; + } + + wq->to_use = wq->to_clean = wq->bufs[0]; + + iowrite32(0, &wq->ctrl->fetch_index); + iowrite32(0, &wq->ctrl->posted_index); + iowrite32(0, &wq->ctrl->error_status); + + vnic_dev_clear_desc_ring(&wq->ring); +} diff --git a/drivers/net/ethernet/cisco/enic/vnic_wq.h b/drivers/net/ethernet/cisco/enic/vnic_wq.h new file mode 100644 index 000000000..01209613d --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/vnic_wq.h @@ -0,0 +1,196 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _VNIC_WQ_H_ +#define _VNIC_WQ_H_ + +#include <linux/pci.h> + +#include "vnic_dev.h" +#include "vnic_cq.h" + +/* Work queue control */ +struct vnic_wq_ctrl { + u64 ring_base; /* 0x00 */ + u32 ring_size; /* 0x08 */ + u32 pad0; + u32 posted_index; /* 0x10 */ + u32 pad1; + u32 cq_index; /* 0x18 */ + u32 pad2; + u32 enable; /* 0x20 */ + u32 pad3; + u32 running; /* 0x28 */ + u32 pad4; + u32 fetch_index; /* 0x30 */ + u32 pad5; + u32 dca_value; /* 0x38 */ + u32 pad6; + u32 error_interrupt_enable; /* 0x40 */ + u32 pad7; + u32 error_interrupt_offset; /* 0x48 */ + u32 pad8; + u32 error_status; /* 0x50 */ + u32 pad9; +}; + +struct vnic_wq_buf { + struct vnic_wq_buf *next; + dma_addr_t dma_addr; + void *os_buf; + unsigned int len; + unsigned int index; + int sop; + void *desc; + uint64_t wr_id; /* Cookie */ + uint8_t cq_entry; /* Gets completion event from hw */ + uint8_t desc_skip_cnt; /* Num descs to occupy */ + uint8_t compressed_send; /* Both hdr and payload in one desc */ + struct vnic_wq_buf *prev; +}; + +/* Break the vnic_wq_buf allocations into blocks of 32/64 entries */ +#define VNIC_WQ_BUF_MIN_BLK_ENTRIES 32 +#define VNIC_WQ_BUF_DFLT_BLK_ENTRIES 64 +#define VNIC_WQ_BUF_BLK_ENTRIES(entries) \ + ((unsigned int)((entries < VNIC_WQ_BUF_DFLT_BLK_ENTRIES) ? \ + VNIC_WQ_BUF_MIN_BLK_ENTRIES : VNIC_WQ_BUF_DFLT_BLK_ENTRIES)) +#define VNIC_WQ_BUF_BLK_SZ(entries) \ + (VNIC_WQ_BUF_BLK_ENTRIES(entries) * sizeof(struct vnic_wq_buf)) +#define VNIC_WQ_BUF_BLKS_NEEDED(entries) \ + DIV_ROUND_UP(entries, VNIC_WQ_BUF_BLK_ENTRIES(entries)) +#define VNIC_WQ_BUF_BLKS_MAX VNIC_WQ_BUF_BLKS_NEEDED(4096) + +struct vnic_wq { + unsigned int index; + struct vnic_dev *vdev; + struct vnic_wq_ctrl __iomem *ctrl; /* memory-mapped */ + struct vnic_dev_ring ring; + struct vnic_wq_buf *bufs[VNIC_WQ_BUF_BLKS_MAX]; + struct vnic_wq_buf *to_use; + struct vnic_wq_buf *to_clean; + unsigned int pkts_outstanding; +}; + +struct devcmd2_controller { + struct vnic_wq_ctrl __iomem *wq_ctrl; + struct vnic_devcmd2 *cmd_ring; + struct devcmd2_result *result; + u16 next_result; + u16 result_size; + int color; + struct vnic_dev_ring results_ring; + struct vnic_wq wq; + u32 posted; +}; + +static inline unsigned int vnic_wq_desc_avail(struct vnic_wq *wq) +{ + /* how many does SW own? */ + return wq->ring.desc_avail; +} + +static inline unsigned int vnic_wq_desc_used(struct vnic_wq *wq) +{ + /* how many does HW own? */ + return wq->ring.desc_count - wq->ring.desc_avail - 1; +} + +static inline void *vnic_wq_next_desc(struct vnic_wq *wq) +{ + return wq->to_use->desc; +} + +static inline void vnic_wq_doorbell(struct vnic_wq *wq) +{ + /* Adding write memory barrier prevents compiler and/or CPU + * reordering, thus avoiding descriptor posting before + * descriptor is initialized. Otherwise, hardware can read + * stale descriptor fields. + */ + wmb(); + iowrite32(wq->to_use->index, &wq->ctrl->posted_index); +} + +static inline void vnic_wq_post(struct vnic_wq *wq, + void *os_buf, dma_addr_t dma_addr, + unsigned int len, int sop, int eop, + uint8_t desc_skip_cnt, uint8_t cq_entry, + uint8_t compressed_send, uint64_t wrid) +{ + struct vnic_wq_buf *buf = wq->to_use; + + buf->sop = sop; + buf->cq_entry = cq_entry; + buf->compressed_send = compressed_send; + buf->desc_skip_cnt = desc_skip_cnt; + buf->os_buf = eop ? os_buf : NULL; + buf->dma_addr = dma_addr; + buf->len = len; + buf->wr_id = wrid; + + buf = buf->next; + wq->to_use = buf; + + wq->ring.desc_avail -= desc_skip_cnt; +} + +static inline void vnic_wq_service(struct vnic_wq *wq, + struct cq_desc *cq_desc, u16 completed_index, + void (*buf_service)(struct vnic_wq *wq, + struct cq_desc *cq_desc, struct vnic_wq_buf *buf, void *opaque), + void *opaque) +{ + struct vnic_wq_buf *buf; + + buf = wq->to_clean; + while (1) { + + (*buf_service)(wq, cq_desc, buf, opaque); + + wq->ring.desc_avail++; + + wq->to_clean = buf->next; + + if (buf->index == completed_index) + break; + + buf = wq->to_clean; + } +} + +void vnic_wq_free(struct vnic_wq *wq); +int vnic_wq_alloc(struct vnic_dev *vdev, struct vnic_wq *wq, unsigned int index, + unsigned int desc_count, unsigned int desc_size); +void vnic_wq_init(struct vnic_wq *wq, unsigned int cq_index, + unsigned int error_interrupt_enable, + unsigned int error_interrupt_offset); +unsigned int vnic_wq_error_status(struct vnic_wq *wq); +void vnic_wq_enable(struct vnic_wq *wq); +int vnic_wq_disable(struct vnic_wq *wq); +void vnic_wq_clean(struct vnic_wq *wq, + void (*buf_clean)(struct vnic_wq *wq, struct vnic_wq_buf *buf)); +int enic_wq_devcmd2_alloc(struct vnic_dev *vdev, struct vnic_wq *wq, + unsigned int desc_count, unsigned int desc_size); +void enic_wq_init_start(struct vnic_wq *wq, unsigned int cq_index, + unsigned int fetch_index, unsigned int posted_index, + unsigned int error_interrupt_enable, + unsigned int error_interrupt_offset); + +#endif /* _VNIC_WQ_H_ */ diff --git a/drivers/net/ethernet/cisco/enic/wq_enet_desc.h b/drivers/net/ethernet/cisco/enic/wq_enet_desc.h new file mode 100644 index 000000000..c7021e3a6 --- /dev/null +++ b/drivers/net/ethernet/cisco/enic/wq_enet_desc.h @@ -0,0 +1,98 @@ +/* + * Copyright 2008-2010 Cisco Systems, Inc. All rights reserved. + * Copyright 2007 Nuova Systems, Inc. All rights reserved. + * + * This program is free software; you may redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 of the License. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef _WQ_ENET_DESC_H_ +#define _WQ_ENET_DESC_H_ + +/* Ethernet work queue descriptor: 16B */ +struct wq_enet_desc { + __le64 address; + __le16 length; + __le16 mss_loopback; + __le16 header_length_flags; + __le16 vlan_tag; +}; + +#define WQ_ENET_ADDR_BITS 64 +#define WQ_ENET_LEN_BITS 14 +#define WQ_ENET_LEN_MASK ((1 << WQ_ENET_LEN_BITS) - 1) +#define WQ_ENET_MSS_BITS 14 +#define WQ_ENET_MSS_MASK ((1 << WQ_ENET_MSS_BITS) - 1) +#define WQ_ENET_MSS_SHIFT 2 +#define WQ_ENET_LOOPBACK_SHIFT 1 +#define WQ_ENET_HDRLEN_BITS 10 +#define WQ_ENET_HDRLEN_MASK ((1 << WQ_ENET_HDRLEN_BITS) - 1) +#define WQ_ENET_FLAGS_OM_BITS 2 +#define WQ_ENET_FLAGS_OM_MASK ((1 << WQ_ENET_FLAGS_OM_BITS) - 1) +#define WQ_ENET_FLAGS_EOP_SHIFT 12 +#define WQ_ENET_FLAGS_CQ_ENTRY_SHIFT 13 +#define WQ_ENET_FLAGS_FCOE_ENCAP_SHIFT 14 +#define WQ_ENET_FLAGS_VLAN_TAG_INSERT_SHIFT 15 + +#define WQ_ENET_OFFLOAD_MODE_CSUM 0 +#define WQ_ENET_OFFLOAD_MODE_RESERVED 1 +#define WQ_ENET_OFFLOAD_MODE_CSUM_L4 2 +#define WQ_ENET_OFFLOAD_MODE_TSO 3 + +static inline void wq_enet_desc_enc(struct wq_enet_desc *desc, + u64 address, u16 length, u16 mss, u16 header_length, + u8 offload_mode, u8 eop, u8 cq_entry, u8 fcoe_encap, + u8 vlan_tag_insert, u16 vlan_tag, u8 loopback) +{ + desc->address = cpu_to_le64(address); + desc->length = cpu_to_le16(length & WQ_ENET_LEN_MASK); + desc->mss_loopback = cpu_to_le16((mss & WQ_ENET_MSS_MASK) << + WQ_ENET_MSS_SHIFT | (loopback & 1) << WQ_ENET_LOOPBACK_SHIFT); + desc->header_length_flags = cpu_to_le16( + (header_length & WQ_ENET_HDRLEN_MASK) | + (offload_mode & WQ_ENET_FLAGS_OM_MASK) << WQ_ENET_HDRLEN_BITS | + (eop & 1) << WQ_ENET_FLAGS_EOP_SHIFT | + (cq_entry & 1) << WQ_ENET_FLAGS_CQ_ENTRY_SHIFT | + (fcoe_encap & 1) << WQ_ENET_FLAGS_FCOE_ENCAP_SHIFT | + (vlan_tag_insert & 1) << WQ_ENET_FLAGS_VLAN_TAG_INSERT_SHIFT); + desc->vlan_tag = cpu_to_le16(vlan_tag); +} + +static inline void wq_enet_desc_dec(struct wq_enet_desc *desc, + u64 *address, u16 *length, u16 *mss, u16 *header_length, + u8 *offload_mode, u8 *eop, u8 *cq_entry, u8 *fcoe_encap, + u8 *vlan_tag_insert, u16 *vlan_tag, u8 *loopback) +{ + *address = le64_to_cpu(desc->address); + *length = le16_to_cpu(desc->length) & WQ_ENET_LEN_MASK; + *mss = (le16_to_cpu(desc->mss_loopback) >> WQ_ENET_MSS_SHIFT) & + WQ_ENET_MSS_MASK; + *loopback = (u8)((le16_to_cpu(desc->mss_loopback) >> + WQ_ENET_LOOPBACK_SHIFT) & 1); + *header_length = le16_to_cpu(desc->header_length_flags) & + WQ_ENET_HDRLEN_MASK; + *offload_mode = (u8)((le16_to_cpu(desc->header_length_flags) >> + WQ_ENET_HDRLEN_BITS) & WQ_ENET_FLAGS_OM_MASK); + *eop = (u8)((le16_to_cpu(desc->header_length_flags) >> + WQ_ENET_FLAGS_EOP_SHIFT) & 1); + *cq_entry = (u8)((le16_to_cpu(desc->header_length_flags) >> + WQ_ENET_FLAGS_CQ_ENTRY_SHIFT) & 1); + *fcoe_encap = (u8)((le16_to_cpu(desc->header_length_flags) >> + WQ_ENET_FLAGS_FCOE_ENCAP_SHIFT) & 1); + *vlan_tag_insert = (u8)((le16_to_cpu(desc->header_length_flags) >> + WQ_ENET_FLAGS_VLAN_TAG_INSERT_SHIFT) & 1); + *vlan_tag = le16_to_cpu(desc->vlan_tag); +} + +#endif /* _WQ_ENET_DESC_H_ */ |