diff options
Diffstat (limited to '')
| -rw-r--r-- | drivers/net/xen-netback/Makefile | 4 | ||||
| -rw-r--r-- | drivers/net/xen-netback/common.h | 442 | ||||
| -rw-r--r-- | drivers/net/xen-netback/hash.c | 468 | ||||
| -rw-r--r-- | drivers/net/xen-netback/interface.c | 856 | ||||
| -rw-r--r-- | drivers/net/xen-netback/netback.c | 1797 | ||||
| -rw-r--r-- | drivers/net/xen-netback/rx.c | 680 | ||||
| -rw-r--r-- | drivers/net/xen-netback/xenbus.c | 1181 |
7 files changed, 5428 insertions, 0 deletions
diff --git a/drivers/net/xen-netback/Makefile b/drivers/net/xen-netback/Makefile new file mode 100644 index 000000000..84e9cbc36 --- /dev/null +++ b/drivers/net/xen-netback/Makefile @@ -0,0 +1,4 @@ +# SPDX-License-Identifier: GPL-2.0-only +obj-$(CONFIG_XEN_NETDEV_BACKEND) := xen-netback.o + +xen-netback-y := netback.o xenbus.o interface.o hash.o rx.o diff --git a/drivers/net/xen-netback/common.h b/drivers/net/xen-netback/common.h new file mode 100644 index 000000000..fe99439ad --- /dev/null +++ b/drivers/net/xen-netback/common.h @@ -0,0 +1,442 @@ +/* + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * 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 __XEN_NETBACK__COMMON_H__ +#define __XEN_NETBACK__COMMON_H__ + +#define pr_fmt(fmt) KBUILD_MODNAME ":%s: " fmt, __func__ + +#include <linux/module.h> +#include <linux/interrupt.h> +#include <linux/slab.h> +#include <linux/ip.h> +#include <linux/in.h> +#include <linux/io.h> +#include <linux/netdevice.h> +#include <linux/etherdevice.h> +#include <linux/wait.h> +#include <linux/sched.h> + +#include <xen/interface/io/netif.h> +#include <xen/interface/grant_table.h> +#include <xen/grant_table.h> +#include <xen/xenbus.h> +#include <xen/page.h> +#include <linux/debugfs.h> + +typedef unsigned int pending_ring_idx_t; + +struct pending_tx_info { + struct xen_netif_tx_request req; /* tx request */ + unsigned int extra_count; + /* Callback data for released SKBs. The callback is always + * xenvif_zerocopy_callback, desc contains the pending_idx, which is + * also an index in pending_tx_info array. It is initialized in + * xenvif_alloc and it never changes. + * skb_shinfo(skb)->destructor_arg points to the first mapped slot's + * callback_struct in this array of struct pending_tx_info's, then ctx + * to the next, or NULL if there is no more slot for this skb. + * ubuf_to_vif is a helper which finds the struct xenvif from a pointer + * to this field. + */ + struct ubuf_info callback_struct; +}; + +#define XEN_NETIF_TX_RING_SIZE __CONST_RING_SIZE(xen_netif_tx, XEN_PAGE_SIZE) +#define XEN_NETIF_RX_RING_SIZE __CONST_RING_SIZE(xen_netif_rx, XEN_PAGE_SIZE) + +struct xenvif_rx_meta { + int id; + int size; + int gso_type; + int gso_size; +}; + +#define GSO_BIT(type) \ + (1 << XEN_NETIF_GSO_TYPE_ ## type) + +/* Discriminate from any valid pending_idx value. */ +#define INVALID_PENDING_IDX 0xFFFF + +#define MAX_PENDING_REQS XEN_NETIF_TX_RING_SIZE + +/* The maximum number of frags is derived from the size of a grant (same + * as a Xen page size for now). + */ +#define MAX_XEN_SKB_FRAGS (65536 / XEN_PAGE_SIZE + 1) + +#define NETBACK_INVALID_HANDLE -1 + +/* To avoid confusion, we define XEN_NETBK_LEGACY_SLOTS_MAX indicating + * the maximum slots a valid packet can use. Now this value is defined + * to be XEN_NETIF_NR_SLOTS_MIN, which is supposed to be supported by + * all backend. + */ +#define XEN_NETBK_LEGACY_SLOTS_MAX XEN_NETIF_NR_SLOTS_MIN + +/* Queue name is interface name with "-qNNN" appended */ +#define QUEUE_NAME_SIZE (IFNAMSIZ + 5) + +/* IRQ name is queue name with "-tx" or "-rx" appended */ +#define IRQ_NAME_SIZE (QUEUE_NAME_SIZE + 3) + +struct xenvif; + +struct xenvif_stats { + /* Stats fields to be updated per-queue. + * A subset of struct net_device_stats that contains only the + * fields that are updated in netback.c for each queue. + */ + u64 rx_bytes; + u64 rx_packets; + u64 tx_bytes; + u64 tx_packets; + + /* Additional stats used by xenvif */ + unsigned long rx_gso_checksum_fixup; + unsigned long tx_zerocopy_sent; + unsigned long tx_zerocopy_success; + unsigned long tx_zerocopy_fail; + unsigned long tx_frag_overflow; +}; + +#define COPY_BATCH_SIZE 64 + +struct xenvif_copy_state { + struct gnttab_copy op[COPY_BATCH_SIZE]; + RING_IDX idx[COPY_BATCH_SIZE]; + unsigned int num; + struct sk_buff_head *completed; +}; + +struct xenvif_queue { /* Per-queue data for xenvif */ + unsigned int id; /* Queue ID, 0-based */ + char name[QUEUE_NAME_SIZE]; /* DEVNAME-qN */ + struct xenvif *vif; /* Parent VIF */ + + /* + * TX/RX common EOI handling. + * When feature-split-event-channels = 0, interrupt handler sets + * NETBK_COMMON_EOI, otherwise NETBK_RX_EOI and NETBK_TX_EOI are set + * by the RX and TX interrupt handlers. + * RX and TX handler threads will issue an EOI when either + * NETBK_COMMON_EOI or their specific bits (NETBK_RX_EOI or + * NETBK_TX_EOI) are set and they will reset those bits. + */ + atomic_t eoi_pending; +#define NETBK_RX_EOI 0x01 +#define NETBK_TX_EOI 0x02 +#define NETBK_COMMON_EOI 0x04 + + /* Use NAPI for guest TX */ + struct napi_struct napi; + /* When feature-split-event-channels = 0, tx_irq = rx_irq. */ + unsigned int tx_irq; + /* Only used when feature-split-event-channels = 1 */ + char tx_irq_name[IRQ_NAME_SIZE]; /* DEVNAME-qN-tx */ + struct xen_netif_tx_back_ring tx; + struct sk_buff_head tx_queue; + struct page *mmap_pages[MAX_PENDING_REQS]; + pending_ring_idx_t pending_prod; + pending_ring_idx_t pending_cons; + u16 pending_ring[MAX_PENDING_REQS]; + struct pending_tx_info pending_tx_info[MAX_PENDING_REQS]; + grant_handle_t grant_tx_handle[MAX_PENDING_REQS]; + + struct gnttab_copy tx_copy_ops[2 * MAX_PENDING_REQS]; + struct gnttab_map_grant_ref tx_map_ops[MAX_PENDING_REQS]; + struct gnttab_unmap_grant_ref tx_unmap_ops[MAX_PENDING_REQS]; + /* passed to gnttab_[un]map_refs with pages under (un)mapping */ + struct page *pages_to_map[MAX_PENDING_REQS]; + struct page *pages_to_unmap[MAX_PENDING_REQS]; + + /* This prevents zerocopy callbacks to race over dealloc_ring */ + spinlock_t callback_lock; + /* This prevents dealloc thread and NAPI instance to race over response + * creation and pending_ring in xenvif_idx_release. In xenvif_tx_err + * it only protect response creation + */ + spinlock_t response_lock; + pending_ring_idx_t dealloc_prod; + pending_ring_idx_t dealloc_cons; + u16 dealloc_ring[MAX_PENDING_REQS]; + struct task_struct *dealloc_task; + wait_queue_head_t dealloc_wq; + atomic_t inflight_packets; + + /* Use kthread for guest RX */ + struct task_struct *task; + wait_queue_head_t wq; + /* When feature-split-event-channels = 0, tx_irq = rx_irq. */ + unsigned int rx_irq; + /* Only used when feature-split-event-channels = 1 */ + char rx_irq_name[IRQ_NAME_SIZE]; /* DEVNAME-qN-rx */ + struct xen_netif_rx_back_ring rx; + struct sk_buff_head rx_queue; + + unsigned int rx_queue_max; + unsigned int rx_queue_len; + unsigned long last_rx_time; + unsigned int rx_slots_needed; + bool stalled; + + struct xenvif_copy_state rx_copy; + + /* Transmit shaping: allow 'credit_bytes' every 'credit_usec'. */ + unsigned long credit_bytes; + unsigned long credit_usec; + unsigned long remaining_credit; + struct timer_list credit_timeout; + u64 credit_window_start; + bool rate_limited; + + /* Statistics */ + struct xenvif_stats stats; +}; + +enum state_bit_shift { + /* This bit marks that the vif is connected */ + VIF_STATUS_CONNECTED, +}; + +struct xenvif_mcast_addr { + struct list_head entry; + struct rcu_head rcu; + u8 addr[6]; +}; + +#define XEN_NETBK_MCAST_MAX 64 + +#define XEN_NETBK_MAX_HASH_KEY_SIZE 40 +#define XEN_NETBK_MAX_HASH_MAPPING_SIZE 128 +#define XEN_NETBK_HASH_TAG_SIZE 40 + +struct xenvif_hash_cache_entry { + struct list_head link; + struct rcu_head rcu; + u8 tag[XEN_NETBK_HASH_TAG_SIZE]; + unsigned int len; + u32 val; + int seq; +}; + +struct xenvif_hash_cache { + spinlock_t lock; + struct list_head list; + unsigned int count; + atomic_t seq; +}; + +struct xenvif_hash { + unsigned int alg; + u32 flags; + bool mapping_sel; + u8 key[XEN_NETBK_MAX_HASH_KEY_SIZE]; + u32 mapping[2][XEN_NETBK_MAX_HASH_MAPPING_SIZE]; + unsigned int size; + struct xenvif_hash_cache cache; +}; + +struct backend_info { + struct xenbus_device *dev; + struct xenvif *vif; + + /* This is the state that will be reflected in xenstore when any + * active hotplug script completes. + */ + enum xenbus_state state; + + enum xenbus_state frontend_state; + struct xenbus_watch hotplug_status_watch; + u8 have_hotplug_status_watch:1; + + const char *hotplug_script; +}; + +struct xenvif { + /* Unique identifier for this interface. */ + domid_t domid; + unsigned int handle; + + u8 fe_dev_addr[6]; + struct list_head fe_mcast_addr; + unsigned int fe_mcast_count; + + /* Frontend feature information. */ + int gso_mask; + + u8 can_sg:1; + u8 ip_csum:1; + u8 ipv6_csum:1; + u8 multicast_control:1; + + /* headroom requested by xen-netfront */ + u16 xdp_headroom; + + /* Is this interface disabled? True when backend discovers + * frontend is rogue. + */ + bool disabled; + unsigned long status; + unsigned long drain_timeout; + unsigned long stall_timeout; + + /* Queues */ + struct xenvif_queue *queues; + unsigned int num_queues; /* active queues, resource allocated */ + unsigned int stalled_queues; + + struct xenvif_hash hash; + + struct xenbus_watch credit_watch; + struct xenbus_watch mcast_ctrl_watch; + + struct backend_info *be; + + spinlock_t lock; + +#ifdef CONFIG_DEBUG_FS + struct dentry *xenvif_dbg_root; +#endif + + struct xen_netif_ctrl_back_ring ctrl; + unsigned int ctrl_irq; + + /* Miscellaneous private stuff. */ + struct net_device *dev; +}; + +struct xenvif_rx_cb { + unsigned long expires; + int meta_slots_used; +}; + +#define XENVIF_RX_CB(skb) ((struct xenvif_rx_cb *)(skb)->cb) + +static inline struct xenbus_device *xenvif_to_xenbus_device(struct xenvif *vif) +{ + return to_xenbus_device(vif->dev->dev.parent); +} + +void xenvif_tx_credit_callback(struct timer_list *t); + +struct xenvif *xenvif_alloc(struct device *parent, + domid_t domid, + unsigned int handle); + +int xenvif_init_queue(struct xenvif_queue *queue); +void xenvif_deinit_queue(struct xenvif_queue *queue); + +int xenvif_connect_data(struct xenvif_queue *queue, + unsigned long tx_ring_ref, + unsigned long rx_ring_ref, + unsigned int tx_evtchn, + unsigned int rx_evtchn); +void xenvif_disconnect_data(struct xenvif *vif); +int xenvif_connect_ctrl(struct xenvif *vif, grant_ref_t ring_ref, + unsigned int evtchn); +void xenvif_disconnect_ctrl(struct xenvif *vif); +void xenvif_free(struct xenvif *vif); + +int xenvif_xenbus_init(void); +void xenvif_xenbus_fini(void); + +/* (Un)Map communication rings. */ +void xenvif_unmap_frontend_data_rings(struct xenvif_queue *queue); +int xenvif_map_frontend_data_rings(struct xenvif_queue *queue, + grant_ref_t tx_ring_ref, + grant_ref_t rx_ring_ref); + +/* Check for SKBs from frontend and schedule backend processing */ +void xenvif_napi_schedule_or_enable_events(struct xenvif_queue *queue); + +/* Prevent the device from generating any further traffic. */ +void xenvif_carrier_off(struct xenvif *vif); + +int xenvif_tx_action(struct xenvif_queue *queue, int budget); + +int xenvif_kthread_guest_rx(void *data); +void xenvif_kick_thread(struct xenvif_queue *queue); + +int xenvif_dealloc_kthread(void *data); + +irqreturn_t xenvif_ctrl_irq_fn(int irq, void *data); + +bool xenvif_have_rx_work(struct xenvif_queue *queue, bool test_kthread); +bool xenvif_rx_queue_tail(struct xenvif_queue *queue, struct sk_buff *skb); + +void xenvif_carrier_on(struct xenvif *vif); + +/* Callback from stack when TX packet can be released */ +void xenvif_zerocopy_callback(struct ubuf_info *ubuf, bool zerocopy_success); + +static inline pending_ring_idx_t nr_pending_reqs(struct xenvif_queue *queue) +{ + return MAX_PENDING_REQS - + queue->pending_prod + queue->pending_cons; +} + +irqreturn_t xenvif_interrupt(int irq, void *dev_id); + +extern bool separate_tx_rx_irq; +extern bool provides_xdp_headroom; + +extern unsigned int rx_drain_timeout_msecs; +extern unsigned int rx_stall_timeout_msecs; +extern unsigned int xenvif_max_queues; +extern unsigned int xenvif_hash_cache_size; + +#ifdef CONFIG_DEBUG_FS +extern struct dentry *xen_netback_dbg_root; +#endif + +void xenvif_skb_zerocopy_prepare(struct xenvif_queue *queue, + struct sk_buff *skb); +void xenvif_skb_zerocopy_complete(struct xenvif_queue *queue); + +/* Multicast control */ +bool xenvif_mcast_match(struct xenvif *vif, const u8 *addr); +void xenvif_mcast_addr_list_free(struct xenvif *vif); + +/* Hash */ +void xenvif_init_hash(struct xenvif *vif); +void xenvif_deinit_hash(struct xenvif *vif); + +u32 xenvif_set_hash_alg(struct xenvif *vif, u32 alg); +u32 xenvif_get_hash_flags(struct xenvif *vif, u32 *flags); +u32 xenvif_set_hash_flags(struct xenvif *vif, u32 flags); +u32 xenvif_set_hash_key(struct xenvif *vif, u32 gref, u32 len); +u32 xenvif_set_hash_mapping_size(struct xenvif *vif, u32 size); +u32 xenvif_set_hash_mapping(struct xenvif *vif, u32 gref, u32 len, + u32 off); + +void xenvif_set_skb_hash(struct xenvif *vif, struct sk_buff *skb); + +#ifdef CONFIG_DEBUG_FS +void xenvif_dump_hash_info(struct xenvif *vif, struct seq_file *m); +#endif + +#endif /* __XEN_NETBACK__COMMON_H__ */ diff --git a/drivers/net/xen-netback/hash.c b/drivers/net/xen-netback/hash.c new file mode 100644 index 000000000..ff96f2264 --- /dev/null +++ b/drivers/net/xen-netback/hash.c @@ -0,0 +1,468 @@ +/* + * Copyright (c) 2016 Citrix Systems Inc. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Softare Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * 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. + */ + +#define XEN_NETIF_DEFINE_TOEPLITZ + +#include "common.h" +#include <linux/vmalloc.h> +#include <linux/rculist.h> + +static void xenvif_add_hash(struct xenvif *vif, const u8 *tag, + unsigned int len, u32 val) +{ + struct xenvif_hash_cache_entry *new, *entry, *oldest; + unsigned long flags; + bool found; + + new = kmalloc(sizeof(*entry), GFP_ATOMIC); + if (!new) + return; + + memcpy(new->tag, tag, len); + new->len = len; + new->val = val; + + spin_lock_irqsave(&vif->hash.cache.lock, flags); + + found = false; + oldest = NULL; + list_for_each_entry_rcu(entry, &vif->hash.cache.list, link, + lockdep_is_held(&vif->hash.cache.lock)) { + /* Make sure we don't add duplicate entries */ + if (entry->len == len && + memcmp(entry->tag, tag, len) == 0) + found = true; + if (!oldest || entry->seq < oldest->seq) + oldest = entry; + } + + if (!found) { + new->seq = atomic_inc_return(&vif->hash.cache.seq); + list_add_rcu(&new->link, &vif->hash.cache.list); + + if (++vif->hash.cache.count > xenvif_hash_cache_size) { + list_del_rcu(&oldest->link); + vif->hash.cache.count--; + kfree_rcu(oldest, rcu); + } + } + + spin_unlock_irqrestore(&vif->hash.cache.lock, flags); + + if (found) + kfree(new); +} + +static u32 xenvif_new_hash(struct xenvif *vif, const u8 *data, + unsigned int len) +{ + u32 val; + + val = xen_netif_toeplitz_hash(vif->hash.key, + sizeof(vif->hash.key), + data, len); + + if (xenvif_hash_cache_size != 0) + xenvif_add_hash(vif, data, len, val); + + return val; +} + +static void xenvif_flush_hash(struct xenvif *vif) +{ + struct xenvif_hash_cache_entry *entry; + unsigned long flags; + + if (xenvif_hash_cache_size == 0) + return; + + spin_lock_irqsave(&vif->hash.cache.lock, flags); + + list_for_each_entry_rcu(entry, &vif->hash.cache.list, link, + lockdep_is_held(&vif->hash.cache.lock)) { + list_del_rcu(&entry->link); + vif->hash.cache.count--; + kfree_rcu(entry, rcu); + } + + spin_unlock_irqrestore(&vif->hash.cache.lock, flags); +} + +static u32 xenvif_find_hash(struct xenvif *vif, const u8 *data, + unsigned int len) +{ + struct xenvif_hash_cache_entry *entry; + u32 val; + bool found; + + if (len >= XEN_NETBK_HASH_TAG_SIZE) + return 0; + + if (xenvif_hash_cache_size == 0) + return xenvif_new_hash(vif, data, len); + + rcu_read_lock(); + + found = false; + + list_for_each_entry_rcu(entry, &vif->hash.cache.list, link) { + if (entry->len == len && + memcmp(entry->tag, data, len) == 0) { + val = entry->val; + entry->seq = atomic_inc_return(&vif->hash.cache.seq); + found = true; + break; + } + } + + rcu_read_unlock(); + + if (!found) + val = xenvif_new_hash(vif, data, len); + + return val; +} + +void xenvif_set_skb_hash(struct xenvif *vif, struct sk_buff *skb) +{ + struct flow_keys flow; + u32 hash = 0; + enum pkt_hash_types type = PKT_HASH_TYPE_NONE; + u32 flags = vif->hash.flags; + bool has_tcp_hdr; + + /* Quick rejection test: If the network protocol doesn't + * correspond to any enabled hash type then there's no point + * in parsing the packet header. + */ + switch (skb->protocol) { + case htons(ETH_P_IP): + if (flags & (XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP | + XEN_NETIF_CTRL_HASH_TYPE_IPV4)) + break; + + goto done; + + case htons(ETH_P_IPV6): + if (flags & (XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP | + XEN_NETIF_CTRL_HASH_TYPE_IPV6)) + break; + + goto done; + + default: + goto done; + } + + memset(&flow, 0, sizeof(flow)); + if (!skb_flow_dissect_flow_keys(skb, &flow, 0)) + goto done; + + has_tcp_hdr = (flow.basic.ip_proto == IPPROTO_TCP) && + !(flow.control.flags & FLOW_DIS_IS_FRAGMENT); + + switch (skb->protocol) { + case htons(ETH_P_IP): + if (has_tcp_hdr && + (flags & XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP)) { + u8 data[12]; + + memcpy(&data[0], &flow.addrs.v4addrs.src, 4); + memcpy(&data[4], &flow.addrs.v4addrs.dst, 4); + memcpy(&data[8], &flow.ports.src, 2); + memcpy(&data[10], &flow.ports.dst, 2); + + hash = xenvif_find_hash(vif, data, sizeof(data)); + type = PKT_HASH_TYPE_L4; + } else if (flags & XEN_NETIF_CTRL_HASH_TYPE_IPV4) { + u8 data[8]; + + memcpy(&data[0], &flow.addrs.v4addrs.src, 4); + memcpy(&data[4], &flow.addrs.v4addrs.dst, 4); + + hash = xenvif_find_hash(vif, data, sizeof(data)); + type = PKT_HASH_TYPE_L3; + } + + break; + + case htons(ETH_P_IPV6): + if (has_tcp_hdr && + (flags & XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP)) { + u8 data[36]; + + memcpy(&data[0], &flow.addrs.v6addrs.src, 16); + memcpy(&data[16], &flow.addrs.v6addrs.dst, 16); + memcpy(&data[32], &flow.ports.src, 2); + memcpy(&data[34], &flow.ports.dst, 2); + + hash = xenvif_find_hash(vif, data, sizeof(data)); + type = PKT_HASH_TYPE_L4; + } else if (flags & XEN_NETIF_CTRL_HASH_TYPE_IPV6) { + u8 data[32]; + + memcpy(&data[0], &flow.addrs.v6addrs.src, 16); + memcpy(&data[16], &flow.addrs.v6addrs.dst, 16); + + hash = xenvif_find_hash(vif, data, sizeof(data)); + type = PKT_HASH_TYPE_L3; + } + + break; + } + +done: + if (type == PKT_HASH_TYPE_NONE) + skb_clear_hash(skb); + else + __skb_set_sw_hash(skb, hash, type == PKT_HASH_TYPE_L4); +} + +u32 xenvif_set_hash_alg(struct xenvif *vif, u32 alg) +{ + switch (alg) { + case XEN_NETIF_CTRL_HASH_ALGORITHM_NONE: + case XEN_NETIF_CTRL_HASH_ALGORITHM_TOEPLITZ: + break; + + default: + return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER; + } + + vif->hash.alg = alg; + + return XEN_NETIF_CTRL_STATUS_SUCCESS; +} + +u32 xenvif_get_hash_flags(struct xenvif *vif, u32 *flags) +{ + if (vif->hash.alg == XEN_NETIF_CTRL_HASH_ALGORITHM_NONE) + return XEN_NETIF_CTRL_STATUS_NOT_SUPPORTED; + + *flags = XEN_NETIF_CTRL_HASH_TYPE_IPV4 | + XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP | + XEN_NETIF_CTRL_HASH_TYPE_IPV6 | + XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP; + + return XEN_NETIF_CTRL_STATUS_SUCCESS; +} + +u32 xenvif_set_hash_flags(struct xenvif *vif, u32 flags) +{ + if (flags & ~(XEN_NETIF_CTRL_HASH_TYPE_IPV4 | + XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP | + XEN_NETIF_CTRL_HASH_TYPE_IPV6 | + XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP)) + return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER; + + if (vif->hash.alg == XEN_NETIF_CTRL_HASH_ALGORITHM_NONE) + return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER; + + vif->hash.flags = flags; + + return XEN_NETIF_CTRL_STATUS_SUCCESS; +} + +u32 xenvif_set_hash_key(struct xenvif *vif, u32 gref, u32 len) +{ + u8 *key = vif->hash.key; + struct gnttab_copy copy_op = { + .source.u.ref = gref, + .source.domid = vif->domid, + .dest.u.gmfn = virt_to_gfn(key), + .dest.domid = DOMID_SELF, + .dest.offset = xen_offset_in_page(key), + .len = len, + .flags = GNTCOPY_source_gref + }; + + if (len > XEN_NETBK_MAX_HASH_KEY_SIZE) + return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER; + + if (copy_op.len != 0) { + gnttab_batch_copy(©_op, 1); + + if (copy_op.status != GNTST_okay) + return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER; + } + + /* Clear any remaining key octets */ + if (len < XEN_NETBK_MAX_HASH_KEY_SIZE) + memset(key + len, 0, XEN_NETBK_MAX_HASH_KEY_SIZE - len); + + xenvif_flush_hash(vif); + + return XEN_NETIF_CTRL_STATUS_SUCCESS; +} + +u32 xenvif_set_hash_mapping_size(struct xenvif *vif, u32 size) +{ + if (size > XEN_NETBK_MAX_HASH_MAPPING_SIZE) + return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER; + + vif->hash.size = size; + memset(vif->hash.mapping[vif->hash.mapping_sel], 0, + sizeof(u32) * size); + + return XEN_NETIF_CTRL_STATUS_SUCCESS; +} + +u32 xenvif_set_hash_mapping(struct xenvif *vif, u32 gref, u32 len, + u32 off) +{ + u32 *mapping = vif->hash.mapping[!vif->hash.mapping_sel]; + unsigned int nr = 1; + struct gnttab_copy copy_op[2] = {{ + .source.u.ref = gref, + .source.domid = vif->domid, + .dest.domid = DOMID_SELF, + .len = len * sizeof(*mapping), + .flags = GNTCOPY_source_gref + }}; + + if ((off + len < off) || (off + len > vif->hash.size) || + len > XEN_PAGE_SIZE / sizeof(*mapping)) + return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER; + + copy_op[0].dest.u.gmfn = virt_to_gfn(mapping + off); + copy_op[0].dest.offset = xen_offset_in_page(mapping + off); + if (copy_op[0].dest.offset + copy_op[0].len > XEN_PAGE_SIZE) { + copy_op[1] = copy_op[0]; + copy_op[1].source.offset = XEN_PAGE_SIZE - copy_op[0].dest.offset; + copy_op[1].dest.u.gmfn = virt_to_gfn(mapping + off + len); + copy_op[1].dest.offset = 0; + copy_op[1].len = copy_op[0].len - copy_op[1].source.offset; + copy_op[0].len = copy_op[1].source.offset; + nr = 2; + } + + memcpy(mapping, vif->hash.mapping[vif->hash.mapping_sel], + vif->hash.size * sizeof(*mapping)); + + if (copy_op[0].len != 0) { + gnttab_batch_copy(copy_op, nr); + + if (copy_op[0].status != GNTST_okay || + copy_op[nr - 1].status != GNTST_okay) + return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER; + } + + while (len-- != 0) + if (mapping[off++] >= vif->num_queues) + return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER; + + vif->hash.mapping_sel = !vif->hash.mapping_sel; + + return XEN_NETIF_CTRL_STATUS_SUCCESS; +} + +#ifdef CONFIG_DEBUG_FS +void xenvif_dump_hash_info(struct xenvif *vif, struct seq_file *m) +{ + unsigned int i; + + switch (vif->hash.alg) { + case XEN_NETIF_CTRL_HASH_ALGORITHM_TOEPLITZ: + seq_puts(m, "Hash Algorithm: TOEPLITZ\n"); + break; + + case XEN_NETIF_CTRL_HASH_ALGORITHM_NONE: + seq_puts(m, "Hash Algorithm: NONE\n"); + fallthrough; + default: + return; + } + + if (vif->hash.flags) { + seq_puts(m, "\nHash Flags:\n"); + + if (vif->hash.flags & XEN_NETIF_CTRL_HASH_TYPE_IPV4) + seq_puts(m, "- IPv4\n"); + if (vif->hash.flags & XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP) + seq_puts(m, "- IPv4 + TCP\n"); + if (vif->hash.flags & XEN_NETIF_CTRL_HASH_TYPE_IPV6) + seq_puts(m, "- IPv6\n"); + if (vif->hash.flags & XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP) + seq_puts(m, "- IPv6 + TCP\n"); + } + + seq_puts(m, "\nHash Key:\n"); + + for (i = 0; i < XEN_NETBK_MAX_HASH_KEY_SIZE; ) { + unsigned int j, n; + + n = 8; + if (i + n >= XEN_NETBK_MAX_HASH_KEY_SIZE) + n = XEN_NETBK_MAX_HASH_KEY_SIZE - i; + + seq_printf(m, "[%2u - %2u]: ", i, i + n - 1); + + for (j = 0; j < n; j++, i++) + seq_printf(m, "%02x ", vif->hash.key[i]); + + seq_puts(m, "\n"); + } + + if (vif->hash.size != 0) { + const u32 *mapping = vif->hash.mapping[vif->hash.mapping_sel]; + + seq_puts(m, "\nHash Mapping:\n"); + + for (i = 0; i < vif->hash.size; ) { + unsigned int j, n; + + n = 8; + if (i + n >= vif->hash.size) + n = vif->hash.size - i; + + seq_printf(m, "[%4u - %4u]: ", i, i + n - 1); + + for (j = 0; j < n; j++, i++) + seq_printf(m, "%4u ", mapping[i]); + + seq_puts(m, "\n"); + } + } +} +#endif /* CONFIG_DEBUG_FS */ + +void xenvif_init_hash(struct xenvif *vif) +{ + if (xenvif_hash_cache_size == 0) + return; + + BUG_ON(vif->hash.cache.count); + + spin_lock_init(&vif->hash.cache.lock); + INIT_LIST_HEAD(&vif->hash.cache.list); +} + +void xenvif_deinit_hash(struct xenvif *vif) +{ + xenvif_flush_hash(vif); +} diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c new file mode 100644 index 000000000..09fc4bc06 --- /dev/null +++ b/drivers/net/xen-netback/interface.c @@ -0,0 +1,856 @@ +/* + * Network-device interface management. + * + * Copyright (c) 2004-2005, Keir Fraser + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * 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 "common.h" + +#include <linux/kthread.h> +#include <linux/sched/task.h> +#include <linux/ethtool.h> +#include <linux/rtnetlink.h> +#include <linux/if_vlan.h> +#include <linux/vmalloc.h> + +#include <xen/events.h> +#include <asm/xen/hypercall.h> +#include <xen/balloon.h> + +#define XENVIF_NAPI_WEIGHT 64 + +/* Number of bytes allowed on the internal guest Rx queue. */ +#define XENVIF_RX_QUEUE_BYTES (XEN_NETIF_RX_RING_SIZE/2 * PAGE_SIZE) + +/* This function is used to set SKBTX_DEV_ZEROCOPY as well as + * increasing the inflight counter. We need to increase the inflight + * counter because core driver calls into xenvif_zerocopy_callback + * which calls xenvif_skb_zerocopy_complete. + */ +void xenvif_skb_zerocopy_prepare(struct xenvif_queue *queue, + struct sk_buff *skb) +{ + skb_shinfo(skb)->tx_flags |= SKBTX_DEV_ZEROCOPY; + atomic_inc(&queue->inflight_packets); +} + +void xenvif_skb_zerocopy_complete(struct xenvif_queue *queue) +{ + atomic_dec(&queue->inflight_packets); + + /* Wake the dealloc thread _after_ decrementing inflight_packets so + * that if kthread_stop() has already been called, the dealloc thread + * does not wait forever with nothing to wake it. + */ + wake_up(&queue->dealloc_wq); +} + +static int xenvif_schedulable(struct xenvif *vif) +{ + return netif_running(vif->dev) && + test_bit(VIF_STATUS_CONNECTED, &vif->status) && + !vif->disabled; +} + +static bool xenvif_handle_tx_interrupt(struct xenvif_queue *queue) +{ + bool rc; + + rc = RING_HAS_UNCONSUMED_REQUESTS(&queue->tx); + if (rc) + napi_schedule(&queue->napi); + return rc; +} + +static irqreturn_t xenvif_tx_interrupt(int irq, void *dev_id) +{ + struct xenvif_queue *queue = dev_id; + int old; + + old = atomic_fetch_or(NETBK_TX_EOI, &queue->eoi_pending); + WARN(old & NETBK_TX_EOI, "Interrupt while EOI pending\n"); + + if (!xenvif_handle_tx_interrupt(queue)) { + atomic_andnot(NETBK_TX_EOI, &queue->eoi_pending); + xen_irq_lateeoi(irq, XEN_EOI_FLAG_SPURIOUS); + } + + return IRQ_HANDLED; +} + +static int xenvif_poll(struct napi_struct *napi, int budget) +{ + struct xenvif_queue *queue = + container_of(napi, struct xenvif_queue, napi); + int work_done; + + /* This vif is rogue, we pretend we've there is nothing to do + * for this vif to deschedule it from NAPI. But this interface + * will be turned off in thread context later. + */ + if (unlikely(queue->vif->disabled)) { + napi_complete(napi); + return 0; + } + + work_done = xenvif_tx_action(queue, budget); + + if (work_done < budget) { + napi_complete_done(napi, work_done); + /* If the queue is rate-limited, it shall be + * rescheduled in the timer callback. + */ + if (likely(!queue->rate_limited)) + xenvif_napi_schedule_or_enable_events(queue); + } + + return work_done; +} + +static bool xenvif_handle_rx_interrupt(struct xenvif_queue *queue) +{ + bool rc; + + rc = xenvif_have_rx_work(queue, false); + if (rc) + xenvif_kick_thread(queue); + return rc; +} + +static irqreturn_t xenvif_rx_interrupt(int irq, void *dev_id) +{ + struct xenvif_queue *queue = dev_id; + int old; + + old = atomic_fetch_or(NETBK_RX_EOI, &queue->eoi_pending); + WARN(old & NETBK_RX_EOI, "Interrupt while EOI pending\n"); + + if (!xenvif_handle_rx_interrupt(queue)) { + atomic_andnot(NETBK_RX_EOI, &queue->eoi_pending); + xen_irq_lateeoi(irq, XEN_EOI_FLAG_SPURIOUS); + } + + return IRQ_HANDLED; +} + +irqreturn_t xenvif_interrupt(int irq, void *dev_id) +{ + struct xenvif_queue *queue = dev_id; + int old; + bool has_rx, has_tx; + + old = atomic_fetch_or(NETBK_COMMON_EOI, &queue->eoi_pending); + WARN(old, "Interrupt while EOI pending\n"); + + has_tx = xenvif_handle_tx_interrupt(queue); + has_rx = xenvif_handle_rx_interrupt(queue); + + if (!has_rx && !has_tx) { + atomic_andnot(NETBK_COMMON_EOI, &queue->eoi_pending); + xen_irq_lateeoi(irq, XEN_EOI_FLAG_SPURIOUS); + } + + return IRQ_HANDLED; +} + +static u16 xenvif_select_queue(struct net_device *dev, struct sk_buff *skb, + struct net_device *sb_dev) +{ + struct xenvif *vif = netdev_priv(dev); + unsigned int size = vif->hash.size; + unsigned int num_queues; + + /* If queues are not set up internally - always return 0 + * as the packet going to be dropped anyway */ + num_queues = READ_ONCE(vif->num_queues); + if (num_queues < 1) + return 0; + + if (vif->hash.alg == XEN_NETIF_CTRL_HASH_ALGORITHM_NONE) + return netdev_pick_tx(dev, skb, NULL) % + dev->real_num_tx_queues; + + xenvif_set_skb_hash(vif, skb); + + if (size == 0) + return skb_get_hash_raw(skb) % dev->real_num_tx_queues; + + return vif->hash.mapping[vif->hash.mapping_sel] + [skb_get_hash_raw(skb) % size]; +} + +static netdev_tx_t +xenvif_start_xmit(struct sk_buff *skb, struct net_device *dev) +{ + struct xenvif *vif = netdev_priv(dev); + struct xenvif_queue *queue = NULL; + unsigned int num_queues; + u16 index; + struct xenvif_rx_cb *cb; + + BUG_ON(skb->dev != dev); + + /* Drop the packet if queues are not set up. + * This handler should be called inside an RCU read section + * so we don't need to enter it here explicitly. + */ + num_queues = READ_ONCE(vif->num_queues); + if (num_queues < 1) + goto drop; + + /* Obtain the queue to be used to transmit this packet */ + index = skb_get_queue_mapping(skb); + if (index >= num_queues) { + pr_warn_ratelimited("Invalid queue %hu for packet on interface %s\n", + index, vif->dev->name); + index %= num_queues; + } + queue = &vif->queues[index]; + + /* Drop the packet if queue is not ready */ + if (queue->task == NULL || + queue->dealloc_task == NULL || + !xenvif_schedulable(vif)) + goto drop; + + if (vif->multicast_control && skb->pkt_type == PACKET_MULTICAST) { + struct ethhdr *eth = (struct ethhdr *)skb->data; + + if (!xenvif_mcast_match(vif, eth->h_dest)) + goto drop; + } + + cb = XENVIF_RX_CB(skb); + cb->expires = jiffies + vif->drain_timeout; + + /* If there is no hash algorithm configured then make sure there + * is no hash information in the socket buffer otherwise it + * would be incorrectly forwarded to the frontend. + */ + if (vif->hash.alg == XEN_NETIF_CTRL_HASH_ALGORITHM_NONE) + skb_clear_hash(skb); + + if (!xenvif_rx_queue_tail(queue, skb)) + goto drop; + + xenvif_kick_thread(queue); + + return NETDEV_TX_OK; + + drop: + vif->dev->stats.tx_dropped++; + dev_kfree_skb_any(skb); + return NETDEV_TX_OK; +} + +static struct net_device_stats *xenvif_get_stats(struct net_device *dev) +{ + struct xenvif *vif = netdev_priv(dev); + struct xenvif_queue *queue = NULL; + unsigned int num_queues; + u64 rx_bytes = 0; + u64 rx_packets = 0; + u64 tx_bytes = 0; + u64 tx_packets = 0; + unsigned int index; + + rcu_read_lock(); + num_queues = READ_ONCE(vif->num_queues); + + /* Aggregate tx and rx stats from each queue */ + for (index = 0; index < num_queues; ++index) { + queue = &vif->queues[index]; + rx_bytes += queue->stats.rx_bytes; + rx_packets += queue->stats.rx_packets; + tx_bytes += queue->stats.tx_bytes; + tx_packets += queue->stats.tx_packets; + } + + rcu_read_unlock(); + + vif->dev->stats.rx_bytes = rx_bytes; + vif->dev->stats.rx_packets = rx_packets; + vif->dev->stats.tx_bytes = tx_bytes; + vif->dev->stats.tx_packets = tx_packets; + + return &vif->dev->stats; +} + +static void xenvif_up(struct xenvif *vif) +{ + struct xenvif_queue *queue = NULL; + unsigned int num_queues = vif->num_queues; + unsigned int queue_index; + + for (queue_index = 0; queue_index < num_queues; ++queue_index) { + queue = &vif->queues[queue_index]; + napi_enable(&queue->napi); + enable_irq(queue->tx_irq); + if (queue->tx_irq != queue->rx_irq) + enable_irq(queue->rx_irq); + xenvif_napi_schedule_or_enable_events(queue); + } +} + +static void xenvif_down(struct xenvif *vif) +{ + struct xenvif_queue *queue = NULL; + unsigned int num_queues = vif->num_queues; + unsigned int queue_index; + + for (queue_index = 0; queue_index < num_queues; ++queue_index) { + queue = &vif->queues[queue_index]; + disable_irq(queue->tx_irq); + if (queue->tx_irq != queue->rx_irq) + disable_irq(queue->rx_irq); + napi_disable(&queue->napi); + del_timer_sync(&queue->credit_timeout); + } +} + +static int xenvif_open(struct net_device *dev) +{ + struct xenvif *vif = netdev_priv(dev); + if (test_bit(VIF_STATUS_CONNECTED, &vif->status)) + xenvif_up(vif); + netif_tx_start_all_queues(dev); + return 0; +} + +static int xenvif_close(struct net_device *dev) +{ + struct xenvif *vif = netdev_priv(dev); + if (test_bit(VIF_STATUS_CONNECTED, &vif->status)) + xenvif_down(vif); + netif_tx_stop_all_queues(dev); + return 0; +} + +static int xenvif_change_mtu(struct net_device *dev, int mtu) +{ + struct xenvif *vif = netdev_priv(dev); + int max = vif->can_sg ? ETH_MAX_MTU - VLAN_ETH_HLEN : ETH_DATA_LEN; + + if (mtu > max) + return -EINVAL; + dev->mtu = mtu; + return 0; +} + +static netdev_features_t xenvif_fix_features(struct net_device *dev, + netdev_features_t features) +{ + struct xenvif *vif = netdev_priv(dev); + + if (!vif->can_sg) + features &= ~NETIF_F_SG; + if (~(vif->gso_mask) & GSO_BIT(TCPV4)) + features &= ~NETIF_F_TSO; + if (~(vif->gso_mask) & GSO_BIT(TCPV6)) + features &= ~NETIF_F_TSO6; + if (!vif->ip_csum) + features &= ~NETIF_F_IP_CSUM; + if (!vif->ipv6_csum) + features &= ~NETIF_F_IPV6_CSUM; + + return features; +} + +static const struct xenvif_stat { + char name[ETH_GSTRING_LEN]; + u16 offset; +} xenvif_stats[] = { + { + "rx_gso_checksum_fixup", + offsetof(struct xenvif_stats, rx_gso_checksum_fixup) + }, + /* If (sent != success + fail), there are probably packets never + * freed up properly! + */ + { + "tx_zerocopy_sent", + offsetof(struct xenvif_stats, tx_zerocopy_sent), + }, + { + "tx_zerocopy_success", + offsetof(struct xenvif_stats, tx_zerocopy_success), + }, + { + "tx_zerocopy_fail", + offsetof(struct xenvif_stats, tx_zerocopy_fail) + }, + /* Number of packets exceeding MAX_SKB_FRAG slots. You should use + * a guest with the same MAX_SKB_FRAG + */ + { + "tx_frag_overflow", + offsetof(struct xenvif_stats, tx_frag_overflow) + }, +}; + +static int xenvif_get_sset_count(struct net_device *dev, int string_set) +{ + switch (string_set) { + case ETH_SS_STATS: + return ARRAY_SIZE(xenvif_stats); + default: + return -EINVAL; + } +} + +static void xenvif_get_ethtool_stats(struct net_device *dev, + struct ethtool_stats *stats, u64 * data) +{ + struct xenvif *vif = netdev_priv(dev); + unsigned int num_queues; + int i; + unsigned int queue_index; + + rcu_read_lock(); + num_queues = READ_ONCE(vif->num_queues); + + for (i = 0; i < ARRAY_SIZE(xenvif_stats); i++) { + unsigned long accum = 0; + for (queue_index = 0; queue_index < num_queues; ++queue_index) { + void *vif_stats = &vif->queues[queue_index].stats; + accum += *(unsigned long *)(vif_stats + xenvif_stats[i].offset); + } + data[i] = accum; + } + + rcu_read_unlock(); +} + +static void xenvif_get_strings(struct net_device *dev, u32 stringset, u8 * data) +{ + int i; + + switch (stringset) { + case ETH_SS_STATS: + for (i = 0; i < ARRAY_SIZE(xenvif_stats); i++) + memcpy(data + i * ETH_GSTRING_LEN, + xenvif_stats[i].name, ETH_GSTRING_LEN); + break; + } +} + +static const struct ethtool_ops xenvif_ethtool_ops = { + .get_link = ethtool_op_get_link, + + .get_sset_count = xenvif_get_sset_count, + .get_ethtool_stats = xenvif_get_ethtool_stats, + .get_strings = xenvif_get_strings, +}; + +static const struct net_device_ops xenvif_netdev_ops = { + .ndo_select_queue = xenvif_select_queue, + .ndo_start_xmit = xenvif_start_xmit, + .ndo_get_stats = xenvif_get_stats, + .ndo_open = xenvif_open, + .ndo_stop = xenvif_close, + .ndo_change_mtu = xenvif_change_mtu, + .ndo_fix_features = xenvif_fix_features, + .ndo_set_mac_address = eth_mac_addr, + .ndo_validate_addr = eth_validate_addr, +}; + +struct xenvif *xenvif_alloc(struct device *parent, domid_t domid, + unsigned int handle) +{ + int err; + struct net_device *dev; + struct xenvif *vif; + char name[IFNAMSIZ] = {}; + + snprintf(name, IFNAMSIZ - 1, "vif%u.%u", domid, handle); + /* Allocate a netdev with the max. supported number of queues. + * When the guest selects the desired number, it will be updated + * via netif_set_real_num_*_queues(). + */ + dev = alloc_netdev_mq(sizeof(struct xenvif), name, NET_NAME_UNKNOWN, + ether_setup, xenvif_max_queues); + if (dev == NULL) { + pr_warn("Could not allocate netdev for %s\n", name); + return ERR_PTR(-ENOMEM); + } + + SET_NETDEV_DEV(dev, parent); + + vif = netdev_priv(dev); + + vif->domid = domid; + vif->handle = handle; + vif->can_sg = 1; + vif->ip_csum = 1; + vif->dev = dev; + vif->disabled = false; + vif->drain_timeout = msecs_to_jiffies(rx_drain_timeout_msecs); + vif->stall_timeout = msecs_to_jiffies(rx_stall_timeout_msecs); + + /* Start out with no queues. */ + vif->queues = NULL; + vif->num_queues = 0; + + vif->xdp_headroom = 0; + + spin_lock_init(&vif->lock); + INIT_LIST_HEAD(&vif->fe_mcast_addr); + + dev->netdev_ops = &xenvif_netdev_ops; + dev->hw_features = NETIF_F_SG | + NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM | + NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_FRAGLIST; + dev->features = dev->hw_features | NETIF_F_RXCSUM; + dev->ethtool_ops = &xenvif_ethtool_ops; + + dev->min_mtu = ETH_MIN_MTU; + dev->max_mtu = ETH_MAX_MTU - VLAN_ETH_HLEN; + + /* + * Initialise a dummy MAC address. We choose the numerically + * largest non-broadcast address to prevent the address getting + * stolen by an Ethernet bridge for STP purposes. + * (FE:FF:FF:FF:FF:FF) + */ + eth_broadcast_addr(dev->dev_addr); + dev->dev_addr[0] &= ~0x01; + + netif_carrier_off(dev); + + err = register_netdev(dev); + if (err) { + netdev_warn(dev, "Could not register device: err=%d\n", err); + free_netdev(dev); + return ERR_PTR(err); + } + + netdev_dbg(dev, "Successfully created xenvif\n"); + + __module_get(THIS_MODULE); + + return vif; +} + +int xenvif_init_queue(struct xenvif_queue *queue) +{ + int err, i; + + queue->credit_bytes = queue->remaining_credit = ~0UL; + queue->credit_usec = 0UL; + timer_setup(&queue->credit_timeout, xenvif_tx_credit_callback, 0); + queue->credit_window_start = get_jiffies_64(); + + queue->rx_queue_max = XENVIF_RX_QUEUE_BYTES; + + skb_queue_head_init(&queue->rx_queue); + skb_queue_head_init(&queue->tx_queue); + + queue->pending_cons = 0; + queue->pending_prod = MAX_PENDING_REQS; + for (i = 0; i < MAX_PENDING_REQS; ++i) + queue->pending_ring[i] = i; + + spin_lock_init(&queue->callback_lock); + spin_lock_init(&queue->response_lock); + + /* If ballooning is disabled, this will consume real memory, so you + * better enable it. The long term solution would be to use just a + * bunch of valid page descriptors, without dependency on ballooning + */ + err = gnttab_alloc_pages(MAX_PENDING_REQS, + queue->mmap_pages); + if (err) { + netdev_err(queue->vif->dev, "Could not reserve mmap_pages\n"); + return -ENOMEM; + } + + for (i = 0; i < MAX_PENDING_REQS; i++) { + queue->pending_tx_info[i].callback_struct = (struct ubuf_info) + { .callback = xenvif_zerocopy_callback, + { { .ctx = NULL, + .desc = i } } }; + queue->grant_tx_handle[i] = NETBACK_INVALID_HANDLE; + } + + return 0; +} + +void xenvif_carrier_on(struct xenvif *vif) +{ + rtnl_lock(); + if (!vif->can_sg && vif->dev->mtu > ETH_DATA_LEN) + dev_set_mtu(vif->dev, ETH_DATA_LEN); + netdev_update_features(vif->dev); + set_bit(VIF_STATUS_CONNECTED, &vif->status); + if (netif_running(vif->dev)) + xenvif_up(vif); + rtnl_unlock(); +} + +int xenvif_connect_ctrl(struct xenvif *vif, grant_ref_t ring_ref, + unsigned int evtchn) +{ + struct net_device *dev = vif->dev; + void *addr; + struct xen_netif_ctrl_sring *shared; + RING_IDX rsp_prod, req_prod; + int err; + + err = xenbus_map_ring_valloc(xenvif_to_xenbus_device(vif), + &ring_ref, 1, &addr); + if (err) + goto err; + + shared = (struct xen_netif_ctrl_sring *)addr; + rsp_prod = READ_ONCE(shared->rsp_prod); + req_prod = READ_ONCE(shared->req_prod); + + BACK_RING_ATTACH(&vif->ctrl, shared, rsp_prod, XEN_PAGE_SIZE); + + err = -EIO; + if (req_prod - rsp_prod > RING_SIZE(&vif->ctrl)) + goto err_unmap; + + err = bind_interdomain_evtchn_to_irq_lateeoi(vif->domid, evtchn); + if (err < 0) + goto err_unmap; + + vif->ctrl_irq = err; + + xenvif_init_hash(vif); + + err = request_threaded_irq(vif->ctrl_irq, NULL, xenvif_ctrl_irq_fn, + IRQF_ONESHOT, "xen-netback-ctrl", vif); + if (err) { + pr_warn("Could not setup irq handler for %s\n", dev->name); + goto err_deinit; + } + + return 0; + +err_deinit: + xenvif_deinit_hash(vif); + unbind_from_irqhandler(vif->ctrl_irq, vif); + vif->ctrl_irq = 0; + +err_unmap: + xenbus_unmap_ring_vfree(xenvif_to_xenbus_device(vif), + vif->ctrl.sring); + vif->ctrl.sring = NULL; + +err: + return err; +} + +static void xenvif_disconnect_queue(struct xenvif_queue *queue) +{ + if (queue->task) { + kthread_stop(queue->task); + put_task_struct(queue->task); + queue->task = NULL; + } + + if (queue->dealloc_task) { + kthread_stop(queue->dealloc_task); + queue->dealloc_task = NULL; + } + + if (queue->napi.poll) { + netif_napi_del(&queue->napi); + queue->napi.poll = NULL; + } + + if (queue->tx_irq) { + unbind_from_irqhandler(queue->tx_irq, queue); + if (queue->tx_irq == queue->rx_irq) + queue->rx_irq = 0; + queue->tx_irq = 0; + } + + if (queue->rx_irq) { + unbind_from_irqhandler(queue->rx_irq, queue); + queue->rx_irq = 0; + } + + xenvif_unmap_frontend_data_rings(queue); +} + +int xenvif_connect_data(struct xenvif_queue *queue, + unsigned long tx_ring_ref, + unsigned long rx_ring_ref, + unsigned int tx_evtchn, + unsigned int rx_evtchn) +{ + struct task_struct *task; + int err; + + BUG_ON(queue->tx_irq); + BUG_ON(queue->task); + BUG_ON(queue->dealloc_task); + + err = xenvif_map_frontend_data_rings(queue, tx_ring_ref, + rx_ring_ref); + if (err < 0) + goto err; + + init_waitqueue_head(&queue->wq); + init_waitqueue_head(&queue->dealloc_wq); + atomic_set(&queue->inflight_packets, 0); + + netif_napi_add(queue->vif->dev, &queue->napi, xenvif_poll, + XENVIF_NAPI_WEIGHT); + + queue->stalled = true; + + task = kthread_run(xenvif_kthread_guest_rx, queue, + "%s-guest-rx", queue->name); + if (IS_ERR(task)) + goto kthread_err; + queue->task = task; + /* + * Take a reference to the task in order to prevent it from being freed + * if the thread function returns before kthread_stop is called. + */ + get_task_struct(task); + + task = kthread_run(xenvif_dealloc_kthread, queue, + "%s-dealloc", queue->name); + if (IS_ERR(task)) + goto kthread_err; + queue->dealloc_task = task; + + if (tx_evtchn == rx_evtchn) { + /* feature-split-event-channels == 0 */ + err = bind_interdomain_evtchn_to_irqhandler_lateeoi( + queue->vif->domid, tx_evtchn, xenvif_interrupt, 0, + queue->name, queue); + if (err < 0) + goto err; + queue->tx_irq = queue->rx_irq = err; + disable_irq(queue->tx_irq); + } else { + /* feature-split-event-channels == 1 */ + snprintf(queue->tx_irq_name, sizeof(queue->tx_irq_name), + "%s-tx", queue->name); + err = bind_interdomain_evtchn_to_irqhandler_lateeoi( + queue->vif->domid, tx_evtchn, xenvif_tx_interrupt, 0, + queue->tx_irq_name, queue); + if (err < 0) + goto err; + queue->tx_irq = err; + disable_irq(queue->tx_irq); + + snprintf(queue->rx_irq_name, sizeof(queue->rx_irq_name), + "%s-rx", queue->name); + err = bind_interdomain_evtchn_to_irqhandler_lateeoi( + queue->vif->domid, rx_evtchn, xenvif_rx_interrupt, 0, + queue->rx_irq_name, queue); + if (err < 0) + goto err; + queue->rx_irq = err; + disable_irq(queue->rx_irq); + } + + return 0; + +kthread_err: + pr_warn("Could not allocate kthread for %s\n", queue->name); + err = PTR_ERR(task); +err: + xenvif_disconnect_queue(queue); + return err; +} + +void xenvif_carrier_off(struct xenvif *vif) +{ + struct net_device *dev = vif->dev; + + rtnl_lock(); + if (test_and_clear_bit(VIF_STATUS_CONNECTED, &vif->status)) { + netif_carrier_off(dev); /* discard queued packets */ + if (netif_running(dev)) + xenvif_down(vif); + } + rtnl_unlock(); +} + +void xenvif_disconnect_data(struct xenvif *vif) +{ + struct xenvif_queue *queue = NULL; + unsigned int num_queues = vif->num_queues; + unsigned int queue_index; + + xenvif_carrier_off(vif); + + for (queue_index = 0; queue_index < num_queues; ++queue_index) { + queue = &vif->queues[queue_index]; + + xenvif_disconnect_queue(queue); + } + + xenvif_mcast_addr_list_free(vif); +} + +void xenvif_disconnect_ctrl(struct xenvif *vif) +{ + if (vif->ctrl_irq) { + xenvif_deinit_hash(vif); + unbind_from_irqhandler(vif->ctrl_irq, vif); + vif->ctrl_irq = 0; + } + + if (vif->ctrl.sring) { + xenbus_unmap_ring_vfree(xenvif_to_xenbus_device(vif), + vif->ctrl.sring); + vif->ctrl.sring = NULL; + } +} + +/* Reverse the relevant parts of xenvif_init_queue(). + * Used for queue teardown from xenvif_free(), and on the + * error handling paths in xenbus.c:connect(). + */ +void xenvif_deinit_queue(struct xenvif_queue *queue) +{ + gnttab_free_pages(MAX_PENDING_REQS, queue->mmap_pages); +} + +void xenvif_free(struct xenvif *vif) +{ + struct xenvif_queue *queues = vif->queues; + unsigned int num_queues = vif->num_queues; + unsigned int queue_index; + + unregister_netdev(vif->dev); + free_netdev(vif->dev); + + for (queue_index = 0; queue_index < num_queues; ++queue_index) + xenvif_deinit_queue(&queues[queue_index]); + vfree(queues); + + module_put(THIS_MODULE); +} diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-netback/netback.c new file mode 100644 index 000000000..d25bb5b9a --- /dev/null +++ b/drivers/net/xen-netback/netback.c @@ -0,0 +1,1797 @@ +/* + * Back-end of the driver for virtual network devices. This portion of the + * driver exports a 'unified' network-device interface that can be accessed + * by any operating system that implements a compatible front end. A + * reference front-end implementation can be found in: + * drivers/net/xen-netfront.c + * + * Copyright (c) 2002-2005, K A Fraser + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * 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 "common.h" + +#include <linux/kthread.h> +#include <linux/if_vlan.h> +#include <linux/udp.h> +#include <linux/highmem.h> + +#include <net/tcp.h> + +#include <xen/xen.h> +#include <xen/events.h> +#include <xen/interface/memory.h> +#include <xen/page.h> + +#include <asm/xen/hypercall.h> + +/* Provide an option to disable split event channels at load time as + * event channels are limited resource. Split event channels are + * enabled by default. + */ +bool separate_tx_rx_irq = true; +module_param(separate_tx_rx_irq, bool, 0644); + +/* The time that packets can stay on the guest Rx internal queue + * before they are dropped. + */ +unsigned int rx_drain_timeout_msecs = 10000; +module_param(rx_drain_timeout_msecs, uint, 0444); + +/* The length of time before the frontend is considered unresponsive + * because it isn't providing Rx slots. + */ +unsigned int rx_stall_timeout_msecs = 60000; +module_param(rx_stall_timeout_msecs, uint, 0444); + +#define MAX_QUEUES_DEFAULT 8 +unsigned int xenvif_max_queues; +module_param_named(max_queues, xenvif_max_queues, uint, 0644); +MODULE_PARM_DESC(max_queues, + "Maximum number of queues per virtual interface"); + +/* + * This is the maximum slots a skb can have. If a guest sends a skb + * which exceeds this limit it is considered malicious. + */ +#define FATAL_SKB_SLOTS_DEFAULT 20 +static unsigned int fatal_skb_slots = FATAL_SKB_SLOTS_DEFAULT; +module_param(fatal_skb_slots, uint, 0444); + +/* The amount to copy out of the first guest Tx slot into the skb's + * linear area. If the first slot has more data, it will be mapped + * and put into the first frag. + * + * This is sized to avoid pulling headers from the frags for most + * TCP/IP packets. + */ +#define XEN_NETBACK_TX_COPY_LEN 128 + +/* This is the maximum number of flows in the hash cache. */ +#define XENVIF_HASH_CACHE_SIZE_DEFAULT 64 +unsigned int xenvif_hash_cache_size = XENVIF_HASH_CACHE_SIZE_DEFAULT; +module_param_named(hash_cache_size, xenvif_hash_cache_size, uint, 0644); +MODULE_PARM_DESC(hash_cache_size, "Number of flows in the hash cache"); + +/* The module parameter tells that we have to put data + * for xen-netfront with the XDP_PACKET_HEADROOM offset + * needed for XDP processing + */ +bool provides_xdp_headroom = true; +module_param(provides_xdp_headroom, bool, 0644); + +static void xenvif_idx_release(struct xenvif_queue *queue, u16 pending_idx, + u8 status); + +static void make_tx_response(struct xenvif_queue *queue, + struct xen_netif_tx_request *txp, + unsigned int extra_count, + s8 st); +static void push_tx_responses(struct xenvif_queue *queue); + +static void xenvif_idx_unmap(struct xenvif_queue *queue, u16 pending_idx); + +static inline int tx_work_todo(struct xenvif_queue *queue); + +static inline unsigned long idx_to_pfn(struct xenvif_queue *queue, + u16 idx) +{ + return page_to_pfn(queue->mmap_pages[idx]); +} + +static inline unsigned long idx_to_kaddr(struct xenvif_queue *queue, + u16 idx) +{ + return (unsigned long)pfn_to_kaddr(idx_to_pfn(queue, idx)); +} + +#define callback_param(vif, pending_idx) \ + (vif->pending_tx_info[pending_idx].callback_struct) + +/* Find the containing VIF's structure from a pointer in pending_tx_info array + */ +static inline struct xenvif_queue *ubuf_to_queue(const struct ubuf_info *ubuf) +{ + u16 pending_idx = ubuf->desc; + struct pending_tx_info *temp = + container_of(ubuf, struct pending_tx_info, callback_struct); + return container_of(temp - pending_idx, + struct xenvif_queue, + pending_tx_info[0]); +} + +static u16 frag_get_pending_idx(skb_frag_t *frag) +{ + return (u16)skb_frag_off(frag); +} + +static void frag_set_pending_idx(skb_frag_t *frag, u16 pending_idx) +{ + skb_frag_off_set(frag, pending_idx); +} + +static inline pending_ring_idx_t pending_index(unsigned i) +{ + return i & (MAX_PENDING_REQS-1); +} + +void xenvif_kick_thread(struct xenvif_queue *queue) +{ + wake_up(&queue->wq); +} + +void xenvif_napi_schedule_or_enable_events(struct xenvif_queue *queue) +{ + int more_to_do; + + RING_FINAL_CHECK_FOR_REQUESTS(&queue->tx, more_to_do); + + if (more_to_do) + napi_schedule(&queue->napi); + else if (atomic_fetch_andnot(NETBK_TX_EOI | NETBK_COMMON_EOI, + &queue->eoi_pending) & + (NETBK_TX_EOI | NETBK_COMMON_EOI)) + xen_irq_lateeoi(queue->tx_irq, 0); +} + +static void tx_add_credit(struct xenvif_queue *queue) +{ + unsigned long max_burst, max_credit; + + /* + * Allow a burst big enough to transmit a jumbo packet of up to 128kB. + * Otherwise the interface can seize up due to insufficient credit. + */ + max_burst = max(131072UL, queue->credit_bytes); + + /* Take care that adding a new chunk of credit doesn't wrap to zero. */ + max_credit = queue->remaining_credit + queue->credit_bytes; + if (max_credit < queue->remaining_credit) + max_credit = ULONG_MAX; /* wrapped: clamp to ULONG_MAX */ + + queue->remaining_credit = min(max_credit, max_burst); + queue->rate_limited = false; +} + +void xenvif_tx_credit_callback(struct timer_list *t) +{ + struct xenvif_queue *queue = from_timer(queue, t, credit_timeout); + tx_add_credit(queue); + xenvif_napi_schedule_or_enable_events(queue); +} + +static void xenvif_tx_err(struct xenvif_queue *queue, + struct xen_netif_tx_request *txp, + unsigned int extra_count, RING_IDX end) +{ + RING_IDX cons = queue->tx.req_cons; + unsigned long flags; + + do { + spin_lock_irqsave(&queue->response_lock, flags); + make_tx_response(queue, txp, extra_count, XEN_NETIF_RSP_ERROR); + push_tx_responses(queue); + spin_unlock_irqrestore(&queue->response_lock, flags); + if (cons == end) + break; + RING_COPY_REQUEST(&queue->tx, cons++, txp); + extra_count = 0; /* only the first frag can have extras */ + } while (1); + queue->tx.req_cons = cons; +} + +static void xenvif_fatal_tx_err(struct xenvif *vif) +{ + netdev_err(vif->dev, "fatal error; disabling device\n"); + vif->disabled = true; + /* Disable the vif from queue 0's kthread */ + if (vif->num_queues) + xenvif_kick_thread(&vif->queues[0]); +} + +static int xenvif_count_requests(struct xenvif_queue *queue, + struct xen_netif_tx_request *first, + unsigned int extra_count, + struct xen_netif_tx_request *txp, + int work_to_do) +{ + RING_IDX cons = queue->tx.req_cons; + int slots = 0; + int drop_err = 0; + int more_data; + + if (!(first->flags & XEN_NETTXF_more_data)) + return 0; + + do { + struct xen_netif_tx_request dropped_tx = { 0 }; + + if (slots >= work_to_do) { + netdev_err(queue->vif->dev, + "Asked for %d slots but exceeds this limit\n", + work_to_do); + xenvif_fatal_tx_err(queue->vif); + return -ENODATA; + } + + /* This guest is really using too many slots and + * considered malicious. + */ + if (unlikely(slots >= fatal_skb_slots)) { + netdev_err(queue->vif->dev, + "Malicious frontend using %d slots, threshold %u\n", + slots, fatal_skb_slots); + xenvif_fatal_tx_err(queue->vif); + return -E2BIG; + } + + /* Xen network protocol had implicit dependency on + * MAX_SKB_FRAGS. XEN_NETBK_LEGACY_SLOTS_MAX is set to + * the historical MAX_SKB_FRAGS value 18 to honor the + * same behavior as before. Any packet using more than + * 18 slots but less than fatal_skb_slots slots is + * dropped + */ + if (!drop_err && slots >= XEN_NETBK_LEGACY_SLOTS_MAX) { + if (net_ratelimit()) + netdev_dbg(queue->vif->dev, + "Too many slots (%d) exceeding limit (%d), dropping packet\n", + slots, XEN_NETBK_LEGACY_SLOTS_MAX); + drop_err = -E2BIG; + } + + if (drop_err) + txp = &dropped_tx; + + RING_COPY_REQUEST(&queue->tx, cons + slots, txp); + + /* If the guest submitted a frame >= 64 KiB then + * first->size overflowed and following slots will + * appear to be larger than the frame. + * + * This cannot be fatal error as there are buggy + * frontends that do this. + * + * Consume all slots and drop the packet. + */ + if (!drop_err && txp->size > first->size) { + if (net_ratelimit()) + netdev_dbg(queue->vif->dev, + "Invalid tx request, slot size %u > remaining size %u\n", + txp->size, first->size); + drop_err = -EIO; + } + + first->size -= txp->size; + slots++; + + if (unlikely((txp->offset + txp->size) > XEN_PAGE_SIZE)) { + netdev_err(queue->vif->dev, "Cross page boundary, txp->offset: %u, size: %u\n", + txp->offset, txp->size); + xenvif_fatal_tx_err(queue->vif); + return -EINVAL; + } + + more_data = txp->flags & XEN_NETTXF_more_data; + + if (!drop_err) + txp++; + + } while (more_data); + + if (drop_err) { + xenvif_tx_err(queue, first, extra_count, cons + slots); + return drop_err; + } + + return slots; +} + + +struct xenvif_tx_cb { + u16 copy_pending_idx[XEN_NETBK_LEGACY_SLOTS_MAX + 1]; + u8 copy_count; + u32 split_mask; +}; + +#define XENVIF_TX_CB(skb) ((struct xenvif_tx_cb *)(skb)->cb) +#define copy_pending_idx(skb, i) (XENVIF_TX_CB(skb)->copy_pending_idx[i]) +#define copy_count(skb) (XENVIF_TX_CB(skb)->copy_count) + +static inline void xenvif_tx_create_map_op(struct xenvif_queue *queue, + u16 pending_idx, + struct xen_netif_tx_request *txp, + unsigned int extra_count, + struct gnttab_map_grant_ref *mop) +{ + queue->pages_to_map[mop-queue->tx_map_ops] = queue->mmap_pages[pending_idx]; + gnttab_set_map_op(mop, idx_to_kaddr(queue, pending_idx), + GNTMAP_host_map | GNTMAP_readonly, + txp->gref, queue->vif->domid); + + memcpy(&queue->pending_tx_info[pending_idx].req, txp, + sizeof(*txp)); + queue->pending_tx_info[pending_idx].extra_count = extra_count; +} + +static inline struct sk_buff *xenvif_alloc_skb(unsigned int size) +{ + struct sk_buff *skb = + alloc_skb(size + NET_SKB_PAD + NET_IP_ALIGN, + GFP_ATOMIC | __GFP_NOWARN); + + BUILD_BUG_ON(sizeof(*XENVIF_TX_CB(skb)) > sizeof(skb->cb)); + if (unlikely(skb == NULL)) + return NULL; + + /* Packets passed to netif_rx() must have some headroom. */ + skb_reserve(skb, NET_SKB_PAD + NET_IP_ALIGN); + + /* Initialize it here to avoid later surprises */ + skb_shinfo(skb)->destructor_arg = NULL; + + return skb; +} + +static void xenvif_get_requests(struct xenvif_queue *queue, + struct sk_buff *skb, + struct xen_netif_tx_request *first, + struct xen_netif_tx_request *txfrags, + unsigned *copy_ops, + unsigned *map_ops, + unsigned int frag_overflow, + struct sk_buff *nskb, + unsigned int extra_count, + unsigned int data_len) +{ + struct skb_shared_info *shinfo = skb_shinfo(skb); + skb_frag_t *frags = shinfo->frags; + u16 pending_idx; + pending_ring_idx_t index; + unsigned int nr_slots; + struct gnttab_copy *cop = queue->tx_copy_ops + *copy_ops; + struct gnttab_map_grant_ref *gop = queue->tx_map_ops + *map_ops; + struct xen_netif_tx_request *txp = first; + + nr_slots = shinfo->nr_frags + frag_overflow + 1; + + copy_count(skb) = 0; + XENVIF_TX_CB(skb)->split_mask = 0; + + /* Create copy ops for exactly data_len bytes into the skb head. */ + __skb_put(skb, data_len); + while (data_len > 0) { + int amount = data_len > txp->size ? txp->size : data_len; + bool split = false; + + cop->source.u.ref = txp->gref; + cop->source.domid = queue->vif->domid; + cop->source.offset = txp->offset; + + cop->dest.domid = DOMID_SELF; + cop->dest.offset = (offset_in_page(skb->data + + skb_headlen(skb) - + data_len)) & ~XEN_PAGE_MASK; + cop->dest.u.gmfn = virt_to_gfn(skb->data + skb_headlen(skb) + - data_len); + + /* Don't cross local page boundary! */ + if (cop->dest.offset + amount > XEN_PAGE_SIZE) { + amount = XEN_PAGE_SIZE - cop->dest.offset; + XENVIF_TX_CB(skb)->split_mask |= 1U << copy_count(skb); + split = true; + } + + cop->len = amount; + cop->flags = GNTCOPY_source_gref; + + index = pending_index(queue->pending_cons); + pending_idx = queue->pending_ring[index]; + callback_param(queue, pending_idx).ctx = NULL; + copy_pending_idx(skb, copy_count(skb)) = pending_idx; + if (!split) + copy_count(skb)++; + + cop++; + data_len -= amount; + + if (amount == txp->size) { + /* The copy op covered the full tx_request */ + + memcpy(&queue->pending_tx_info[pending_idx].req, + txp, sizeof(*txp)); + queue->pending_tx_info[pending_idx].extra_count = + (txp == first) ? extra_count : 0; + + if (txp == first) + txp = txfrags; + else + txp++; + queue->pending_cons++; + nr_slots--; + } else { + /* The copy op partially covered the tx_request. + * The remainder will be mapped or copied in the next + * iteration. + */ + txp->offset += amount; + txp->size -= amount; + } + } + + for (shinfo->nr_frags = 0; nr_slots > 0 && shinfo->nr_frags < MAX_SKB_FRAGS; + nr_slots--) { + if (unlikely(!txp->size)) { + unsigned long flags; + + spin_lock_irqsave(&queue->response_lock, flags); + make_tx_response(queue, txp, 0, XEN_NETIF_RSP_OKAY); + push_tx_responses(queue); + spin_unlock_irqrestore(&queue->response_lock, flags); + ++txp; + continue; + } + + index = pending_index(queue->pending_cons++); + pending_idx = queue->pending_ring[index]; + xenvif_tx_create_map_op(queue, pending_idx, txp, + txp == first ? extra_count : 0, gop); + frag_set_pending_idx(&frags[shinfo->nr_frags], pending_idx); + ++shinfo->nr_frags; + ++gop; + + if (txp == first) + txp = txfrags; + else + txp++; + } + + if (nr_slots > 0) { + + shinfo = skb_shinfo(nskb); + frags = shinfo->frags; + + for (shinfo->nr_frags = 0; shinfo->nr_frags < nr_slots; ++txp) { + if (unlikely(!txp->size)) { + unsigned long flags; + + spin_lock_irqsave(&queue->response_lock, flags); + make_tx_response(queue, txp, 0, + XEN_NETIF_RSP_OKAY); + push_tx_responses(queue); + spin_unlock_irqrestore(&queue->response_lock, + flags); + continue; + } + + index = pending_index(queue->pending_cons++); + pending_idx = queue->pending_ring[index]; + xenvif_tx_create_map_op(queue, pending_idx, txp, 0, + gop); + frag_set_pending_idx(&frags[shinfo->nr_frags], + pending_idx); + ++shinfo->nr_frags; + ++gop; + } + + if (shinfo->nr_frags) { + skb_shinfo(skb)->frag_list = nskb; + nskb = NULL; + } + } + + if (nskb) { + /* A frag_list skb was allocated but it is no longer needed + * because enough slots were converted to copy ops above or some + * were empty. + */ + kfree_skb(nskb); + } + + (*copy_ops) = cop - queue->tx_copy_ops; + (*map_ops) = gop - queue->tx_map_ops; +} + +static inline void xenvif_grant_handle_set(struct xenvif_queue *queue, + u16 pending_idx, + grant_handle_t handle) +{ + if (unlikely(queue->grant_tx_handle[pending_idx] != + NETBACK_INVALID_HANDLE)) { + netdev_err(queue->vif->dev, + "Trying to overwrite active handle! pending_idx: 0x%x\n", + pending_idx); + BUG(); + } + queue->grant_tx_handle[pending_idx] = handle; +} + +static inline void xenvif_grant_handle_reset(struct xenvif_queue *queue, + u16 pending_idx) +{ + if (unlikely(queue->grant_tx_handle[pending_idx] == + NETBACK_INVALID_HANDLE)) { + netdev_err(queue->vif->dev, + "Trying to unmap invalid handle! pending_idx: 0x%x\n", + pending_idx); + BUG(); + } + queue->grant_tx_handle[pending_idx] = NETBACK_INVALID_HANDLE; +} + +static int xenvif_tx_check_gop(struct xenvif_queue *queue, + struct sk_buff *skb, + struct gnttab_map_grant_ref **gopp_map, + struct gnttab_copy **gopp_copy) +{ + struct gnttab_map_grant_ref *gop_map = *gopp_map; + u16 pending_idx; + /* This always points to the shinfo of the skb being checked, which + * could be either the first or the one on the frag_list + */ + struct skb_shared_info *shinfo = skb_shinfo(skb); + /* If this is non-NULL, we are currently checking the frag_list skb, and + * this points to the shinfo of the first one + */ + struct skb_shared_info *first_shinfo = NULL; + int nr_frags = shinfo->nr_frags; + const bool sharedslot = nr_frags && + frag_get_pending_idx(&shinfo->frags[0]) == + copy_pending_idx(skb, copy_count(skb) - 1); + int i, err = 0; + + for (i = 0; i < copy_count(skb); i++) { + int newerr; + + /* Check status of header. */ + pending_idx = copy_pending_idx(skb, i); + + newerr = (*gopp_copy)->status; + + /* Split copies need to be handled together. */ + if (XENVIF_TX_CB(skb)->split_mask & (1U << i)) { + (*gopp_copy)++; + if (!newerr) + newerr = (*gopp_copy)->status; + } + if (likely(!newerr)) { + /* The first frag might still have this slot mapped */ + if (i < copy_count(skb) - 1 || !sharedslot) + xenvif_idx_release(queue, pending_idx, + XEN_NETIF_RSP_OKAY); + } else { + err = newerr; + if (net_ratelimit()) + netdev_dbg(queue->vif->dev, + "Grant copy of header failed! status: %d pending_idx: %u ref: %u\n", + (*gopp_copy)->status, + pending_idx, + (*gopp_copy)->source.u.ref); + /* The first frag might still have this slot mapped */ + if (i < copy_count(skb) - 1 || !sharedslot) + xenvif_idx_release(queue, pending_idx, + XEN_NETIF_RSP_ERROR); + } + (*gopp_copy)++; + } + +check_frags: + for (i = 0; i < nr_frags; i++, gop_map++) { + int j, newerr; + + pending_idx = frag_get_pending_idx(&shinfo->frags[i]); + + /* Check error status: if okay then remember grant handle. */ + newerr = gop_map->status; + + if (likely(!newerr)) { + xenvif_grant_handle_set(queue, + pending_idx, + gop_map->handle); + /* Had a previous error? Invalidate this fragment. */ + if (unlikely(err)) { + xenvif_idx_unmap(queue, pending_idx); + /* If the mapping of the first frag was OK, but + * the header's copy failed, and they are + * sharing a slot, send an error + */ + if (i == 0 && !first_shinfo && sharedslot) + xenvif_idx_release(queue, pending_idx, + XEN_NETIF_RSP_ERROR); + else + xenvif_idx_release(queue, pending_idx, + XEN_NETIF_RSP_OKAY); + } + continue; + } + + /* Error on this fragment: respond to client with an error. */ + if (net_ratelimit()) + netdev_dbg(queue->vif->dev, + "Grant map of %d. frag failed! status: %d pending_idx: %u ref: %u\n", + i, + gop_map->status, + pending_idx, + gop_map->ref); + + xenvif_idx_release(queue, pending_idx, XEN_NETIF_RSP_ERROR); + + /* Not the first error? Preceding frags already invalidated. */ + if (err) + continue; + + /* Invalidate preceding fragments of this skb. */ + for (j = 0; j < i; j++) { + pending_idx = frag_get_pending_idx(&shinfo->frags[j]); + xenvif_idx_unmap(queue, pending_idx); + xenvif_idx_release(queue, pending_idx, + XEN_NETIF_RSP_OKAY); + } + + /* And if we found the error while checking the frag_list, unmap + * the first skb's frags + */ + if (first_shinfo) { + for (j = 0; j < first_shinfo->nr_frags; j++) { + pending_idx = frag_get_pending_idx(&first_shinfo->frags[j]); + xenvif_idx_unmap(queue, pending_idx); + xenvif_idx_release(queue, pending_idx, + XEN_NETIF_RSP_OKAY); + } + } + + /* Remember the error: invalidate all subsequent fragments. */ + err = newerr; + } + + if (skb_has_frag_list(skb) && !first_shinfo) { + first_shinfo = skb_shinfo(skb); + shinfo = skb_shinfo(skb_shinfo(skb)->frag_list); + nr_frags = shinfo->nr_frags; + + goto check_frags; + } + + *gopp_map = gop_map; + return err; +} + +static void xenvif_fill_frags(struct xenvif_queue *queue, struct sk_buff *skb) +{ + struct skb_shared_info *shinfo = skb_shinfo(skb); + int nr_frags = shinfo->nr_frags; + int i; + u16 prev_pending_idx = INVALID_PENDING_IDX; + + for (i = 0; i < nr_frags; i++) { + skb_frag_t *frag = shinfo->frags + i; + struct xen_netif_tx_request *txp; + struct page *page; + u16 pending_idx; + + pending_idx = frag_get_pending_idx(frag); + + /* If this is not the first frag, chain it to the previous*/ + if (prev_pending_idx == INVALID_PENDING_IDX) + skb_shinfo(skb)->destructor_arg = + &callback_param(queue, pending_idx); + else + callback_param(queue, prev_pending_idx).ctx = + &callback_param(queue, pending_idx); + + callback_param(queue, pending_idx).ctx = NULL; + prev_pending_idx = pending_idx; + + txp = &queue->pending_tx_info[pending_idx].req; + page = virt_to_page(idx_to_kaddr(queue, pending_idx)); + __skb_fill_page_desc(skb, i, page, txp->offset, txp->size); + skb->len += txp->size; + skb->data_len += txp->size; + skb->truesize += txp->size; + + /* Take an extra reference to offset network stack's put_page */ + get_page(queue->mmap_pages[pending_idx]); + } +} + +static int xenvif_get_extras(struct xenvif_queue *queue, + struct xen_netif_extra_info *extras, + unsigned int *extra_count, + int work_to_do) +{ + struct xen_netif_extra_info extra; + RING_IDX cons = queue->tx.req_cons; + + do { + if (unlikely(work_to_do-- <= 0)) { + netdev_err(queue->vif->dev, "Missing extra info\n"); + xenvif_fatal_tx_err(queue->vif); + return -EBADR; + } + + RING_COPY_REQUEST(&queue->tx, cons, &extra); + + queue->tx.req_cons = ++cons; + (*extra_count)++; + + if (unlikely(!extra.type || + extra.type >= XEN_NETIF_EXTRA_TYPE_MAX)) { + netdev_err(queue->vif->dev, + "Invalid extra type: %d\n", extra.type); + xenvif_fatal_tx_err(queue->vif); + return -EINVAL; + } + + memcpy(&extras[extra.type - 1], &extra, sizeof(extra)); + } while (extra.flags & XEN_NETIF_EXTRA_FLAG_MORE); + + return work_to_do; +} + +static int xenvif_set_skb_gso(struct xenvif *vif, + struct sk_buff *skb, + struct xen_netif_extra_info *gso) +{ + if (!gso->u.gso.size) { + netdev_err(vif->dev, "GSO size must not be zero.\n"); + xenvif_fatal_tx_err(vif); + return -EINVAL; + } + + switch (gso->u.gso.type) { + case XEN_NETIF_GSO_TYPE_TCPV4: + skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4; + break; + case XEN_NETIF_GSO_TYPE_TCPV6: + skb_shinfo(skb)->gso_type = SKB_GSO_TCPV6; + break; + default: + netdev_err(vif->dev, "Bad GSO type %d.\n", gso->u.gso.type); + xenvif_fatal_tx_err(vif); + return -EINVAL; + } + + skb_shinfo(skb)->gso_size = gso->u.gso.size; + /* gso_segs will be calculated later */ + + return 0; +} + +static int checksum_setup(struct xenvif_queue *queue, struct sk_buff *skb) +{ + bool recalculate_partial_csum = false; + + /* A GSO SKB must be CHECKSUM_PARTIAL. However some buggy + * peers can fail to set NETRXF_csum_blank when sending a GSO + * frame. In this case force the SKB to CHECKSUM_PARTIAL and + * recalculate the partial checksum. + */ + if (skb->ip_summed != CHECKSUM_PARTIAL && skb_is_gso(skb)) { + queue->stats.rx_gso_checksum_fixup++; + skb->ip_summed = CHECKSUM_PARTIAL; + recalculate_partial_csum = true; + } + + /* A non-CHECKSUM_PARTIAL SKB does not require setup. */ + if (skb->ip_summed != CHECKSUM_PARTIAL) + return 0; + + return skb_checksum_setup(skb, recalculate_partial_csum); +} + +static bool tx_credit_exceeded(struct xenvif_queue *queue, unsigned size) +{ + u64 now = get_jiffies_64(); + u64 next_credit = queue->credit_window_start + + msecs_to_jiffies(queue->credit_usec / 1000); + + /* Timer could already be pending in rare cases. */ + if (timer_pending(&queue->credit_timeout)) { + queue->rate_limited = true; + return true; + } + + /* Passed the point where we can replenish credit? */ + if (time_after_eq64(now, next_credit)) { + queue->credit_window_start = now; + tx_add_credit(queue); + } + + /* Still too big to send right now? Set a callback. */ + if (size > queue->remaining_credit) { + mod_timer(&queue->credit_timeout, + next_credit); + queue->credit_window_start = next_credit; + queue->rate_limited = true; + + return true; + } + + return false; +} + +/* No locking is required in xenvif_mcast_add/del() as they are + * only ever invoked from NAPI poll. An RCU list is used because + * xenvif_mcast_match() is called asynchronously, during start_xmit. + */ + +static int xenvif_mcast_add(struct xenvif *vif, const u8 *addr) +{ + struct xenvif_mcast_addr *mcast; + + if (vif->fe_mcast_count == XEN_NETBK_MCAST_MAX) { + if (net_ratelimit()) + netdev_err(vif->dev, + "Too many multicast addresses\n"); + return -ENOSPC; + } + + mcast = kzalloc(sizeof(*mcast), GFP_ATOMIC); + if (!mcast) + return -ENOMEM; + + ether_addr_copy(mcast->addr, addr); + list_add_tail_rcu(&mcast->entry, &vif->fe_mcast_addr); + vif->fe_mcast_count++; + + return 0; +} + +static void xenvif_mcast_del(struct xenvif *vif, const u8 *addr) +{ + struct xenvif_mcast_addr *mcast; + + list_for_each_entry_rcu(mcast, &vif->fe_mcast_addr, entry) { + if (ether_addr_equal(addr, mcast->addr)) { + --vif->fe_mcast_count; + list_del_rcu(&mcast->entry); + kfree_rcu(mcast, rcu); + break; + } + } +} + +bool xenvif_mcast_match(struct xenvif *vif, const u8 *addr) +{ + struct xenvif_mcast_addr *mcast; + + rcu_read_lock(); + list_for_each_entry_rcu(mcast, &vif->fe_mcast_addr, entry) { + if (ether_addr_equal(addr, mcast->addr)) { + rcu_read_unlock(); + return true; + } + } + rcu_read_unlock(); + + return false; +} + +void xenvif_mcast_addr_list_free(struct xenvif *vif) +{ + /* No need for locking or RCU here. NAPI poll and TX queue + * are stopped. + */ + while (!list_empty(&vif->fe_mcast_addr)) { + struct xenvif_mcast_addr *mcast; + + mcast = list_first_entry(&vif->fe_mcast_addr, + struct xenvif_mcast_addr, + entry); + --vif->fe_mcast_count; + list_del(&mcast->entry); + kfree(mcast); + } +} + +static void xenvif_tx_build_gops(struct xenvif_queue *queue, + int budget, + unsigned *copy_ops, + unsigned *map_ops) +{ + struct sk_buff *skb, *nskb; + int ret; + unsigned int frag_overflow; + + while (skb_queue_len(&queue->tx_queue) < budget) { + struct xen_netif_tx_request txreq; + struct xen_netif_tx_request txfrags[XEN_NETBK_LEGACY_SLOTS_MAX]; + struct xen_netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX-1]; + unsigned int extra_count; + u16 pending_idx; + RING_IDX idx; + int work_to_do; + unsigned int data_len; + pending_ring_idx_t index; + + if (queue->tx.sring->req_prod - queue->tx.req_cons > + XEN_NETIF_TX_RING_SIZE) { + netdev_err(queue->vif->dev, + "Impossible number of requests. " + "req_prod %d, req_cons %d, size %ld\n", + queue->tx.sring->req_prod, queue->tx.req_cons, + XEN_NETIF_TX_RING_SIZE); + xenvif_fatal_tx_err(queue->vif); + break; + } + + work_to_do = RING_HAS_UNCONSUMED_REQUESTS(&queue->tx); + if (!work_to_do) + break; + + idx = queue->tx.req_cons; + rmb(); /* Ensure that we see the request before we copy it. */ + RING_COPY_REQUEST(&queue->tx, idx, &txreq); + + /* Credit-based scheduling. */ + if (txreq.size > queue->remaining_credit && + tx_credit_exceeded(queue, txreq.size)) + break; + + queue->remaining_credit -= txreq.size; + + work_to_do--; + queue->tx.req_cons = ++idx; + + memset(extras, 0, sizeof(extras)); + extra_count = 0; + if (txreq.flags & XEN_NETTXF_extra_info) { + work_to_do = xenvif_get_extras(queue, extras, + &extra_count, + work_to_do); + idx = queue->tx.req_cons; + if (unlikely(work_to_do < 0)) + break; + } + + if (extras[XEN_NETIF_EXTRA_TYPE_MCAST_ADD - 1].type) { + struct xen_netif_extra_info *extra; + + extra = &extras[XEN_NETIF_EXTRA_TYPE_MCAST_ADD - 1]; + ret = xenvif_mcast_add(queue->vif, extra->u.mcast.addr); + + make_tx_response(queue, &txreq, extra_count, + (ret == 0) ? + XEN_NETIF_RSP_OKAY : + XEN_NETIF_RSP_ERROR); + push_tx_responses(queue); + continue; + } + + if (extras[XEN_NETIF_EXTRA_TYPE_MCAST_DEL - 1].type) { + struct xen_netif_extra_info *extra; + + extra = &extras[XEN_NETIF_EXTRA_TYPE_MCAST_DEL - 1]; + xenvif_mcast_del(queue->vif, extra->u.mcast.addr); + + make_tx_response(queue, &txreq, extra_count, + XEN_NETIF_RSP_OKAY); + push_tx_responses(queue); + continue; + } + + data_len = (txreq.size > XEN_NETBACK_TX_COPY_LEN) ? + XEN_NETBACK_TX_COPY_LEN : txreq.size; + + ret = xenvif_count_requests(queue, &txreq, extra_count, + txfrags, work_to_do); + + if (unlikely(ret < 0)) + break; + + idx += ret; + + if (unlikely(txreq.size < ETH_HLEN)) { + netdev_dbg(queue->vif->dev, + "Bad packet size: %d\n", txreq.size); + xenvif_tx_err(queue, &txreq, extra_count, idx); + break; + } + + /* No crossing a page as the payload mustn't fragment. */ + if (unlikely((txreq.offset + txreq.size) > XEN_PAGE_SIZE)) { + netdev_err(queue->vif->dev, "Cross page boundary, txreq.offset: %u, size: %u\n", + txreq.offset, txreq.size); + xenvif_fatal_tx_err(queue->vif); + break; + } + + index = pending_index(queue->pending_cons); + pending_idx = queue->pending_ring[index]; + + if (ret >= XEN_NETBK_LEGACY_SLOTS_MAX - 1 && data_len < txreq.size) + data_len = txreq.size; + + skb = xenvif_alloc_skb(data_len); + if (unlikely(skb == NULL)) { + netdev_dbg(queue->vif->dev, + "Can't allocate a skb in start_xmit.\n"); + xenvif_tx_err(queue, &txreq, extra_count, idx); + break; + } + + skb_shinfo(skb)->nr_frags = ret; + /* At this point shinfo->nr_frags is in fact the number of + * slots, which can be as large as XEN_NETBK_LEGACY_SLOTS_MAX. + */ + frag_overflow = 0; + nskb = NULL; + if (skb_shinfo(skb)->nr_frags > MAX_SKB_FRAGS) { + frag_overflow = skb_shinfo(skb)->nr_frags - MAX_SKB_FRAGS; + BUG_ON(frag_overflow > MAX_SKB_FRAGS); + skb_shinfo(skb)->nr_frags = MAX_SKB_FRAGS; + nskb = xenvif_alloc_skb(0); + if (unlikely(nskb == NULL)) { + skb_shinfo(skb)->nr_frags = 0; + kfree_skb(skb); + xenvif_tx_err(queue, &txreq, extra_count, idx); + if (net_ratelimit()) + netdev_err(queue->vif->dev, + "Can't allocate the frag_list skb.\n"); + break; + } + } + + if (extras[XEN_NETIF_EXTRA_TYPE_GSO - 1].type) { + struct xen_netif_extra_info *gso; + gso = &extras[XEN_NETIF_EXTRA_TYPE_GSO - 1]; + + if (xenvif_set_skb_gso(queue->vif, skb, gso)) { + /* Failure in xenvif_set_skb_gso is fatal. */ + skb_shinfo(skb)->nr_frags = 0; + kfree_skb(skb); + kfree_skb(nskb); + break; + } + } + + if (extras[XEN_NETIF_EXTRA_TYPE_HASH - 1].type) { + struct xen_netif_extra_info *extra; + enum pkt_hash_types type = PKT_HASH_TYPE_NONE; + + extra = &extras[XEN_NETIF_EXTRA_TYPE_HASH - 1]; + + switch (extra->u.hash.type) { + case _XEN_NETIF_CTRL_HASH_TYPE_IPV4: + case _XEN_NETIF_CTRL_HASH_TYPE_IPV6: + type = PKT_HASH_TYPE_L3; + break; + + case _XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP: + case _XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP: + type = PKT_HASH_TYPE_L4; + break; + + default: + break; + } + + if (type != PKT_HASH_TYPE_NONE) + skb_set_hash(skb, + *(u32 *)extra->u.hash.value, + type); + } + + xenvif_get_requests(queue, skb, &txreq, txfrags, copy_ops, + map_ops, frag_overflow, nskb, extra_count, + data_len); + + __skb_queue_tail(&queue->tx_queue, skb); + + queue->tx.req_cons = idx; + + if ((*map_ops >= ARRAY_SIZE(queue->tx_map_ops)) || + (*copy_ops >= ARRAY_SIZE(queue->tx_copy_ops))) + break; + } + + return; +} + +/* Consolidate skb with a frag_list into a brand new one with local pages on + * frags. Returns 0 or -ENOMEM if can't allocate new pages. + */ +static int xenvif_handle_frag_list(struct xenvif_queue *queue, struct sk_buff *skb) +{ + unsigned int offset = skb_headlen(skb); + skb_frag_t frags[MAX_SKB_FRAGS]; + int i, f; + struct ubuf_info *uarg; + struct sk_buff *nskb = skb_shinfo(skb)->frag_list; + + queue->stats.tx_zerocopy_sent += 2; + queue->stats.tx_frag_overflow++; + + xenvif_fill_frags(queue, nskb); + /* Subtract frags size, we will correct it later */ + skb->truesize -= skb->data_len; + skb->len += nskb->len; + skb->data_len += nskb->len; + + /* create a brand new frags array and coalesce there */ + for (i = 0; offset < skb->len; i++) { + struct page *page; + unsigned int len; + + BUG_ON(i >= MAX_SKB_FRAGS); + page = alloc_page(GFP_ATOMIC); + if (!page) { + int j; + skb->truesize += skb->data_len; + for (j = 0; j < i; j++) + put_page(skb_frag_page(&frags[j])); + return -ENOMEM; + } + + if (offset + PAGE_SIZE < skb->len) + len = PAGE_SIZE; + else + len = skb->len - offset; + if (skb_copy_bits(skb, offset, page_address(page), len)) + BUG(); + + offset += len; + __skb_frag_set_page(&frags[i], page); + skb_frag_off_set(&frags[i], 0); + skb_frag_size_set(&frags[i], len); + } + + /* Release all the original (foreign) frags. */ + for (f = 0; f < skb_shinfo(skb)->nr_frags; f++) + skb_frag_unref(skb, f); + uarg = skb_shinfo(skb)->destructor_arg; + /* increase inflight counter to offset decrement in callback */ + atomic_inc(&queue->inflight_packets); + uarg->callback(uarg, true); + skb_shinfo(skb)->destructor_arg = NULL; + + /* Fill the skb with the new (local) frags. */ + memcpy(skb_shinfo(skb)->frags, frags, i * sizeof(skb_frag_t)); + skb_shinfo(skb)->nr_frags = i; + skb->truesize += i * PAGE_SIZE; + + return 0; +} + +static int xenvif_tx_submit(struct xenvif_queue *queue) +{ + struct gnttab_map_grant_ref *gop_map = queue->tx_map_ops; + struct gnttab_copy *gop_copy = queue->tx_copy_ops; + struct sk_buff *skb; + int work_done = 0; + + while ((skb = __skb_dequeue(&queue->tx_queue)) != NULL) { + struct xen_netif_tx_request *txp; + u16 pending_idx; + + pending_idx = copy_pending_idx(skb, 0); + txp = &queue->pending_tx_info[pending_idx].req; + + /* Check the remap error code. */ + if (unlikely(xenvif_tx_check_gop(queue, skb, &gop_map, &gop_copy))) { + /* If there was an error, xenvif_tx_check_gop is + * expected to release all the frags which were mapped, + * so kfree_skb shouldn't do it again + */ + skb_shinfo(skb)->nr_frags = 0; + if (skb_has_frag_list(skb)) { + struct sk_buff *nskb = + skb_shinfo(skb)->frag_list; + skb_shinfo(nskb)->nr_frags = 0; + } + kfree_skb(skb); + continue; + } + + if (txp->flags & XEN_NETTXF_csum_blank) + skb->ip_summed = CHECKSUM_PARTIAL; + else if (txp->flags & XEN_NETTXF_data_validated) + skb->ip_summed = CHECKSUM_UNNECESSARY; + + xenvif_fill_frags(queue, skb); + + if (unlikely(skb_has_frag_list(skb))) { + struct sk_buff *nskb = skb_shinfo(skb)->frag_list; + xenvif_skb_zerocopy_prepare(queue, nskb); + if (xenvif_handle_frag_list(queue, skb)) { + if (net_ratelimit()) + netdev_err(queue->vif->dev, + "Not enough memory to consolidate frag_list!\n"); + xenvif_skb_zerocopy_prepare(queue, skb); + kfree_skb(skb); + continue; + } + /* Copied all the bits from the frag list -- free it. */ + skb_frag_list_init(skb); + kfree_skb(nskb); + } + + skb->dev = queue->vif->dev; + skb->protocol = eth_type_trans(skb, skb->dev); + skb_reset_network_header(skb); + + if (checksum_setup(queue, skb)) { + netdev_dbg(queue->vif->dev, + "Can't setup checksum in net_tx_action\n"); + /* We have to set this flag to trigger the callback */ + if (skb_shinfo(skb)->destructor_arg) + xenvif_skb_zerocopy_prepare(queue, skb); + kfree_skb(skb); + continue; + } + + skb_probe_transport_header(skb); + + /* If the packet is GSO then we will have just set up the + * transport header offset in checksum_setup so it's now + * straightforward to calculate gso_segs. + */ + if (skb_is_gso(skb)) { + int mss, hdrlen; + + /* GSO implies having the L4 header. */ + WARN_ON_ONCE(!skb_transport_header_was_set(skb)); + if (unlikely(!skb_transport_header_was_set(skb))) { + kfree_skb(skb); + continue; + } + + mss = skb_shinfo(skb)->gso_size; + hdrlen = skb_transport_header(skb) - + skb_mac_header(skb) + + tcp_hdrlen(skb); + + skb_shinfo(skb)->gso_segs = + DIV_ROUND_UP(skb->len - hdrlen, mss); + } + + queue->stats.rx_bytes += skb->len; + queue->stats.rx_packets++; + + work_done++; + + /* Set this flag right before netif_receive_skb, otherwise + * someone might think this packet already left netback, and + * do a skb_copy_ubufs while we are still in control of the + * skb. E.g. the __pskb_pull_tail earlier can do such thing. + */ + if (skb_shinfo(skb)->destructor_arg) { + xenvif_skb_zerocopy_prepare(queue, skb); + queue->stats.tx_zerocopy_sent++; + } + + netif_receive_skb(skb); + } + + return work_done; +} + +void xenvif_zerocopy_callback(struct ubuf_info *ubuf, bool zerocopy_success) +{ + unsigned long flags; + pending_ring_idx_t index; + struct xenvif_queue *queue = ubuf_to_queue(ubuf); + + /* This is the only place where we grab this lock, to protect callbacks + * from each other. + */ + spin_lock_irqsave(&queue->callback_lock, flags); + do { + u16 pending_idx = ubuf->desc; + ubuf = (struct ubuf_info *) ubuf->ctx; + BUG_ON(queue->dealloc_prod - queue->dealloc_cons >= + MAX_PENDING_REQS); + index = pending_index(queue->dealloc_prod); + queue->dealloc_ring[index] = pending_idx; + /* Sync with xenvif_tx_dealloc_action: + * insert idx then incr producer. + */ + smp_wmb(); + queue->dealloc_prod++; + } while (ubuf); + spin_unlock_irqrestore(&queue->callback_lock, flags); + + if (likely(zerocopy_success)) + queue->stats.tx_zerocopy_success++; + else + queue->stats.tx_zerocopy_fail++; + xenvif_skb_zerocopy_complete(queue); +} + +static inline void xenvif_tx_dealloc_action(struct xenvif_queue *queue) +{ + struct gnttab_unmap_grant_ref *gop; + pending_ring_idx_t dc, dp; + u16 pending_idx, pending_idx_release[MAX_PENDING_REQS]; + unsigned int i = 0; + + dc = queue->dealloc_cons; + gop = queue->tx_unmap_ops; + + /* Free up any grants we have finished using */ + do { + dp = queue->dealloc_prod; + + /* Ensure we see all indices enqueued by all + * xenvif_zerocopy_callback(). + */ + smp_rmb(); + + while (dc != dp) { + BUG_ON(gop - queue->tx_unmap_ops >= MAX_PENDING_REQS); + pending_idx = + queue->dealloc_ring[pending_index(dc++)]; + + pending_idx_release[gop - queue->tx_unmap_ops] = + pending_idx; + queue->pages_to_unmap[gop - queue->tx_unmap_ops] = + queue->mmap_pages[pending_idx]; + gnttab_set_unmap_op(gop, + idx_to_kaddr(queue, pending_idx), + GNTMAP_host_map, + queue->grant_tx_handle[pending_idx]); + xenvif_grant_handle_reset(queue, pending_idx); + ++gop; + } + + } while (dp != queue->dealloc_prod); + + queue->dealloc_cons = dc; + + if (gop - queue->tx_unmap_ops > 0) { + int ret; + ret = gnttab_unmap_refs(queue->tx_unmap_ops, + NULL, + queue->pages_to_unmap, + gop - queue->tx_unmap_ops); + if (ret) { + netdev_err(queue->vif->dev, "Unmap fail: nr_ops %tu ret %d\n", + gop - queue->tx_unmap_ops, ret); + for (i = 0; i < gop - queue->tx_unmap_ops; ++i) { + if (gop[i].status != GNTST_okay) + netdev_err(queue->vif->dev, + " host_addr: 0x%llx handle: 0x%x status: %d\n", + gop[i].host_addr, + gop[i].handle, + gop[i].status); + } + BUG(); + } + } + + for (i = 0; i < gop - queue->tx_unmap_ops; ++i) + xenvif_idx_release(queue, pending_idx_release[i], + XEN_NETIF_RSP_OKAY); +} + + +/* Called after netfront has transmitted */ +int xenvif_tx_action(struct xenvif_queue *queue, int budget) +{ + unsigned nr_mops = 0, nr_cops = 0; + int work_done, ret; + + if (unlikely(!tx_work_todo(queue))) + return 0; + + xenvif_tx_build_gops(queue, budget, &nr_cops, &nr_mops); + + if (nr_cops == 0) + return 0; + + gnttab_batch_copy(queue->tx_copy_ops, nr_cops); + if (nr_mops != 0) { + ret = gnttab_map_refs(queue->tx_map_ops, + NULL, + queue->pages_to_map, + nr_mops); + if (ret) { + unsigned int i; + + netdev_err(queue->vif->dev, "Map fail: nr %u ret %d\n", + nr_mops, ret); + for (i = 0; i < nr_mops; ++i) + WARN_ON_ONCE(queue->tx_map_ops[i].status == + GNTST_okay); + } + } + + work_done = xenvif_tx_submit(queue); + + return work_done; +} + +static void xenvif_idx_release(struct xenvif_queue *queue, u16 pending_idx, + u8 status) +{ + struct pending_tx_info *pending_tx_info; + pending_ring_idx_t index; + unsigned long flags; + + pending_tx_info = &queue->pending_tx_info[pending_idx]; + + spin_lock_irqsave(&queue->response_lock, flags); + + make_tx_response(queue, &pending_tx_info->req, + pending_tx_info->extra_count, status); + + /* Release the pending index before pusing the Tx response so + * its available before a new Tx request is pushed by the + * frontend. + */ + index = pending_index(queue->pending_prod++); + queue->pending_ring[index] = pending_idx; + + push_tx_responses(queue); + + spin_unlock_irqrestore(&queue->response_lock, flags); +} + + +static void make_tx_response(struct xenvif_queue *queue, + struct xen_netif_tx_request *txp, + unsigned int extra_count, + s8 st) +{ + RING_IDX i = queue->tx.rsp_prod_pvt; + struct xen_netif_tx_response *resp; + + resp = RING_GET_RESPONSE(&queue->tx, i); + resp->id = txp->id; + resp->status = st; + + while (extra_count-- != 0) + RING_GET_RESPONSE(&queue->tx, ++i)->status = XEN_NETIF_RSP_NULL; + + queue->tx.rsp_prod_pvt = ++i; +} + +static void push_tx_responses(struct xenvif_queue *queue) +{ + int notify; + + RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&queue->tx, notify); + if (notify) + notify_remote_via_irq(queue->tx_irq); +} + +static void xenvif_idx_unmap(struct xenvif_queue *queue, u16 pending_idx) +{ + int ret; + struct gnttab_unmap_grant_ref tx_unmap_op; + + gnttab_set_unmap_op(&tx_unmap_op, + idx_to_kaddr(queue, pending_idx), + GNTMAP_host_map, + queue->grant_tx_handle[pending_idx]); + xenvif_grant_handle_reset(queue, pending_idx); + + ret = gnttab_unmap_refs(&tx_unmap_op, NULL, + &queue->mmap_pages[pending_idx], 1); + if (ret) { + netdev_err(queue->vif->dev, + "Unmap fail: ret: %d pending_idx: %d host_addr: %llx handle: 0x%x status: %d\n", + ret, + pending_idx, + tx_unmap_op.host_addr, + tx_unmap_op.handle, + tx_unmap_op.status); + BUG(); + } +} + +static inline int tx_work_todo(struct xenvif_queue *queue) +{ + if (likely(RING_HAS_UNCONSUMED_REQUESTS(&queue->tx))) + return 1; + + return 0; +} + +static inline bool tx_dealloc_work_todo(struct xenvif_queue *queue) +{ + return queue->dealloc_cons != queue->dealloc_prod; +} + +void xenvif_unmap_frontend_data_rings(struct xenvif_queue *queue) +{ + if (queue->tx.sring) + xenbus_unmap_ring_vfree(xenvif_to_xenbus_device(queue->vif), + queue->tx.sring); + if (queue->rx.sring) + xenbus_unmap_ring_vfree(xenvif_to_xenbus_device(queue->vif), + queue->rx.sring); +} + +int xenvif_map_frontend_data_rings(struct xenvif_queue *queue, + grant_ref_t tx_ring_ref, + grant_ref_t rx_ring_ref) +{ + void *addr; + struct xen_netif_tx_sring *txs; + struct xen_netif_rx_sring *rxs; + RING_IDX rsp_prod, req_prod; + int err = -ENOMEM; + + err = xenbus_map_ring_valloc(xenvif_to_xenbus_device(queue->vif), + &tx_ring_ref, 1, &addr); + if (err) + goto err; + + txs = (struct xen_netif_tx_sring *)addr; + rsp_prod = READ_ONCE(txs->rsp_prod); + req_prod = READ_ONCE(txs->req_prod); + + BACK_RING_ATTACH(&queue->tx, txs, rsp_prod, XEN_PAGE_SIZE); + + err = -EIO; + if (req_prod - rsp_prod > RING_SIZE(&queue->tx)) + goto err; + + err = xenbus_map_ring_valloc(xenvif_to_xenbus_device(queue->vif), + &rx_ring_ref, 1, &addr); + if (err) + goto err; + + rxs = (struct xen_netif_rx_sring *)addr; + rsp_prod = READ_ONCE(rxs->rsp_prod); + req_prod = READ_ONCE(rxs->req_prod); + + BACK_RING_ATTACH(&queue->rx, rxs, rsp_prod, XEN_PAGE_SIZE); + + err = -EIO; + if (req_prod - rsp_prod > RING_SIZE(&queue->rx)) + goto err; + + return 0; + +err: + xenvif_unmap_frontend_data_rings(queue); + return err; +} + +static bool xenvif_dealloc_kthread_should_stop(struct xenvif_queue *queue) +{ + /* Dealloc thread must remain running until all inflight + * packets complete. + */ + return kthread_should_stop() && + !atomic_read(&queue->inflight_packets); +} + +int xenvif_dealloc_kthread(void *data) +{ + struct xenvif_queue *queue = data; + + for (;;) { + wait_event_interruptible(queue->dealloc_wq, + tx_dealloc_work_todo(queue) || + xenvif_dealloc_kthread_should_stop(queue)); + if (xenvif_dealloc_kthread_should_stop(queue)) + break; + + xenvif_tx_dealloc_action(queue); + cond_resched(); + } + + /* Unmap anything remaining*/ + if (tx_dealloc_work_todo(queue)) + xenvif_tx_dealloc_action(queue); + + return 0; +} + +static void make_ctrl_response(struct xenvif *vif, + const struct xen_netif_ctrl_request *req, + u32 status, u32 data) +{ + RING_IDX idx = vif->ctrl.rsp_prod_pvt; + struct xen_netif_ctrl_response rsp = { + .id = req->id, + .type = req->type, + .status = status, + .data = data, + }; + + *RING_GET_RESPONSE(&vif->ctrl, idx) = rsp; + vif->ctrl.rsp_prod_pvt = ++idx; +} + +static void push_ctrl_response(struct xenvif *vif) +{ + int notify; + + RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&vif->ctrl, notify); + if (notify) + notify_remote_via_irq(vif->ctrl_irq); +} + +static void process_ctrl_request(struct xenvif *vif, + const struct xen_netif_ctrl_request *req) +{ + u32 status = XEN_NETIF_CTRL_STATUS_NOT_SUPPORTED; + u32 data = 0; + + switch (req->type) { + case XEN_NETIF_CTRL_TYPE_SET_HASH_ALGORITHM: + status = xenvif_set_hash_alg(vif, req->data[0]); + break; + + case XEN_NETIF_CTRL_TYPE_GET_HASH_FLAGS: + status = xenvif_get_hash_flags(vif, &data); + break; + + case XEN_NETIF_CTRL_TYPE_SET_HASH_FLAGS: + status = xenvif_set_hash_flags(vif, req->data[0]); + break; + + case XEN_NETIF_CTRL_TYPE_SET_HASH_KEY: + status = xenvif_set_hash_key(vif, req->data[0], + req->data[1]); + break; + + case XEN_NETIF_CTRL_TYPE_GET_HASH_MAPPING_SIZE: + status = XEN_NETIF_CTRL_STATUS_SUCCESS; + data = XEN_NETBK_MAX_HASH_MAPPING_SIZE; + break; + + case XEN_NETIF_CTRL_TYPE_SET_HASH_MAPPING_SIZE: + status = xenvif_set_hash_mapping_size(vif, + req->data[0]); + break; + + case XEN_NETIF_CTRL_TYPE_SET_HASH_MAPPING: + status = xenvif_set_hash_mapping(vif, req->data[0], + req->data[1], + req->data[2]); + break; + + default: + break; + } + + make_ctrl_response(vif, req, status, data); + push_ctrl_response(vif); +} + +static void xenvif_ctrl_action(struct xenvif *vif) +{ + for (;;) { + RING_IDX req_prod, req_cons; + + req_prod = vif->ctrl.sring->req_prod; + req_cons = vif->ctrl.req_cons; + + /* Make sure we can see requests before we process them. */ + rmb(); + + if (req_cons == req_prod) + break; + + while (req_cons != req_prod) { + struct xen_netif_ctrl_request req; + + RING_COPY_REQUEST(&vif->ctrl, req_cons, &req); + req_cons++; + + process_ctrl_request(vif, &req); + } + + vif->ctrl.req_cons = req_cons; + vif->ctrl.sring->req_event = req_cons + 1; + } +} + +static bool xenvif_ctrl_work_todo(struct xenvif *vif) +{ + if (likely(RING_HAS_UNCONSUMED_REQUESTS(&vif->ctrl))) + return true; + + return false; +} + +irqreturn_t xenvif_ctrl_irq_fn(int irq, void *data) +{ + struct xenvif *vif = data; + unsigned int eoi_flag = XEN_EOI_FLAG_SPURIOUS; + + while (xenvif_ctrl_work_todo(vif)) { + xenvif_ctrl_action(vif); + eoi_flag = 0; + } + + xen_irq_lateeoi(irq, eoi_flag); + + return IRQ_HANDLED; +} + +static int __init netback_init(void) +{ + int rc = 0; + + if (!xen_domain()) + return -ENODEV; + + /* Allow as many queues as there are CPUs but max. 8 if user has not + * specified a value. + */ + if (xenvif_max_queues == 0) + xenvif_max_queues = min_t(unsigned int, MAX_QUEUES_DEFAULT, + num_online_cpus()); + + if (fatal_skb_slots < XEN_NETBK_LEGACY_SLOTS_MAX) { + pr_info("fatal_skb_slots too small (%d), bump it to XEN_NETBK_LEGACY_SLOTS_MAX (%d)\n", + fatal_skb_slots, XEN_NETBK_LEGACY_SLOTS_MAX); + fatal_skb_slots = XEN_NETBK_LEGACY_SLOTS_MAX; + } + + rc = xenvif_xenbus_init(); + if (rc) + goto failed_init; + +#ifdef CONFIG_DEBUG_FS + xen_netback_dbg_root = debugfs_create_dir("xen-netback", NULL); +#endif /* CONFIG_DEBUG_FS */ + + return 0; + +failed_init: + return rc; +} + +module_init(netback_init); + +static void __exit netback_fini(void) +{ +#ifdef CONFIG_DEBUG_FS + debugfs_remove_recursive(xen_netback_dbg_root); +#endif /* CONFIG_DEBUG_FS */ + xenvif_xenbus_fini(); +} +module_exit(netback_fini); + +MODULE_LICENSE("Dual BSD/GPL"); +MODULE_ALIAS("xen-backend:vif"); diff --git a/drivers/net/xen-netback/rx.c b/drivers/net/xen-netback/rx.c new file mode 100644 index 000000000..0ba754ebc --- /dev/null +++ b/drivers/net/xen-netback/rx.c @@ -0,0 +1,680 @@ +/* + * Copyright (c) 2016 Citrix Systems Inc. + * Copyright (c) 2002-2005, K A Fraser + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * 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 "common.h" + +#include <linux/kthread.h> + +#include <xen/xen.h> +#include <xen/events.h> + +/* + * Update the needed ring page slots for the first SKB queued. + * Note that any call sequence outside the RX thread calling this function + * needs to wake up the RX thread via a call of xenvif_kick_thread() + * afterwards in order to avoid a race with putting the thread to sleep. + */ +static void xenvif_update_needed_slots(struct xenvif_queue *queue, + const struct sk_buff *skb) +{ + unsigned int needed = 0; + + if (skb) { + needed = DIV_ROUND_UP(skb->len, XEN_PAGE_SIZE); + if (skb_is_gso(skb)) + needed++; + if (skb->sw_hash) + needed++; + } + + WRITE_ONCE(queue->rx_slots_needed, needed); +} + +static bool xenvif_rx_ring_slots_available(struct xenvif_queue *queue) +{ + RING_IDX prod, cons; + unsigned int needed; + + needed = READ_ONCE(queue->rx_slots_needed); + if (!needed) + return false; + + do { + prod = queue->rx.sring->req_prod; + cons = queue->rx.req_cons; + + if (prod - cons >= needed) + return true; + + queue->rx.sring->req_event = prod + 1; + + /* Make sure event is visible before we check prod + * again. + */ + mb(); + } while (queue->rx.sring->req_prod != prod); + + return false; +} + +bool xenvif_rx_queue_tail(struct xenvif_queue *queue, struct sk_buff *skb) +{ + unsigned long flags; + bool ret = true; + + spin_lock_irqsave(&queue->rx_queue.lock, flags); + + if (queue->rx_queue_len >= queue->rx_queue_max) { + struct net_device *dev = queue->vif->dev; + + netif_tx_stop_queue(netdev_get_tx_queue(dev, queue->id)); + ret = false; + } else { + if (skb_queue_empty(&queue->rx_queue)) + xenvif_update_needed_slots(queue, skb); + + __skb_queue_tail(&queue->rx_queue, skb); + + queue->rx_queue_len += skb->len; + } + + spin_unlock_irqrestore(&queue->rx_queue.lock, flags); + + return ret; +} + +static struct sk_buff *xenvif_rx_dequeue(struct xenvif_queue *queue) +{ + struct sk_buff *skb; + + spin_lock_irq(&queue->rx_queue.lock); + + skb = __skb_dequeue(&queue->rx_queue); + if (skb) { + xenvif_update_needed_slots(queue, skb_peek(&queue->rx_queue)); + + queue->rx_queue_len -= skb->len; + if (queue->rx_queue_len < queue->rx_queue_max) { + struct netdev_queue *txq; + + txq = netdev_get_tx_queue(queue->vif->dev, queue->id); + netif_tx_wake_queue(txq); + } + } + + spin_unlock_irq(&queue->rx_queue.lock); + + return skb; +} + +static void xenvif_rx_queue_purge(struct xenvif_queue *queue) +{ + struct sk_buff *skb; + + while ((skb = xenvif_rx_dequeue(queue)) != NULL) + kfree_skb(skb); +} + +static void xenvif_rx_queue_drop_expired(struct xenvif_queue *queue) +{ + struct sk_buff *skb; + + for (;;) { + skb = skb_peek(&queue->rx_queue); + if (!skb) + break; + if (time_before(jiffies, XENVIF_RX_CB(skb)->expires)) + break; + xenvif_rx_dequeue(queue); + kfree_skb(skb); + queue->vif->dev->stats.rx_dropped++; + } +} + +static void xenvif_rx_copy_flush(struct xenvif_queue *queue) +{ + unsigned int i; + int notify; + + gnttab_batch_copy(queue->rx_copy.op, queue->rx_copy.num); + + for (i = 0; i < queue->rx_copy.num; i++) { + struct gnttab_copy *op; + + op = &queue->rx_copy.op[i]; + + /* If the copy failed, overwrite the status field in + * the corresponding response. + */ + if (unlikely(op->status != GNTST_okay)) { + struct xen_netif_rx_response *rsp; + + rsp = RING_GET_RESPONSE(&queue->rx, + queue->rx_copy.idx[i]); + rsp->status = op->status; + } + } + + queue->rx_copy.num = 0; + + /* Push responses for all completed packets. */ + RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&queue->rx, notify); + if (notify) + notify_remote_via_irq(queue->rx_irq); + + __skb_queue_purge(queue->rx_copy.completed); +} + +static void xenvif_rx_copy_add(struct xenvif_queue *queue, + struct xen_netif_rx_request *req, + unsigned int offset, void *data, size_t len) +{ + struct gnttab_copy *op; + struct page *page; + struct xen_page_foreign *foreign; + + if (queue->rx_copy.num == COPY_BATCH_SIZE) + xenvif_rx_copy_flush(queue); + + op = &queue->rx_copy.op[queue->rx_copy.num]; + + page = virt_to_page(data); + + op->flags = GNTCOPY_dest_gref; + + foreign = xen_page_foreign(page); + if (foreign) { + op->source.domid = foreign->domid; + op->source.u.ref = foreign->gref; + op->flags |= GNTCOPY_source_gref; + } else { + op->source.u.gmfn = virt_to_gfn(data); + op->source.domid = DOMID_SELF; + } + + op->source.offset = xen_offset_in_page(data); + op->dest.u.ref = req->gref; + op->dest.domid = queue->vif->domid; + op->dest.offset = offset; + op->len = len; + + queue->rx_copy.idx[queue->rx_copy.num] = queue->rx.req_cons; + queue->rx_copy.num++; +} + +static unsigned int xenvif_gso_type(struct sk_buff *skb) +{ + if (skb_is_gso(skb)) { + if (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV4) + return XEN_NETIF_GSO_TYPE_TCPV4; + else + return XEN_NETIF_GSO_TYPE_TCPV6; + } + return XEN_NETIF_GSO_TYPE_NONE; +} + +struct xenvif_pkt_state { + struct sk_buff *skb; + size_t remaining_len; + struct sk_buff *frag_iter; + int frag; /* frag == -1 => frag_iter->head */ + unsigned int frag_offset; + struct xen_netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX - 1]; + unsigned int extra_count; + unsigned int slot; +}; + +static void xenvif_rx_next_skb(struct xenvif_queue *queue, + struct xenvif_pkt_state *pkt) +{ + struct sk_buff *skb; + unsigned int gso_type; + + skb = xenvif_rx_dequeue(queue); + + queue->stats.tx_bytes += skb->len; + queue->stats.tx_packets++; + + /* Reset packet state. */ + memset(pkt, 0, sizeof(struct xenvif_pkt_state)); + + pkt->skb = skb; + pkt->frag_iter = skb; + pkt->remaining_len = skb->len; + pkt->frag = -1; + + gso_type = xenvif_gso_type(skb); + if ((1 << gso_type) & queue->vif->gso_mask) { + struct xen_netif_extra_info *extra; + + extra = &pkt->extras[XEN_NETIF_EXTRA_TYPE_GSO - 1]; + + extra->u.gso.type = gso_type; + extra->u.gso.size = skb_shinfo(skb)->gso_size; + extra->u.gso.pad = 0; + extra->u.gso.features = 0; + extra->type = XEN_NETIF_EXTRA_TYPE_GSO; + extra->flags = 0; + + pkt->extra_count++; + } + + if (queue->vif->xdp_headroom) { + struct xen_netif_extra_info *extra; + + extra = &pkt->extras[XEN_NETIF_EXTRA_TYPE_XDP - 1]; + + memset(extra, 0, sizeof(struct xen_netif_extra_info)); + extra->u.xdp.headroom = queue->vif->xdp_headroom; + extra->type = XEN_NETIF_EXTRA_TYPE_XDP; + extra->flags = 0; + + pkt->extra_count++; + } + + if (skb->sw_hash) { + struct xen_netif_extra_info *extra; + + extra = &pkt->extras[XEN_NETIF_EXTRA_TYPE_HASH - 1]; + + extra->u.hash.algorithm = + XEN_NETIF_CTRL_HASH_ALGORITHM_TOEPLITZ; + + if (skb->l4_hash) + extra->u.hash.type = + skb->protocol == htons(ETH_P_IP) ? + _XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP : + _XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP; + else + extra->u.hash.type = + skb->protocol == htons(ETH_P_IP) ? + _XEN_NETIF_CTRL_HASH_TYPE_IPV4 : + _XEN_NETIF_CTRL_HASH_TYPE_IPV6; + + *(uint32_t *)extra->u.hash.value = skb_get_hash_raw(skb); + + extra->type = XEN_NETIF_EXTRA_TYPE_HASH; + extra->flags = 0; + + pkt->extra_count++; + } +} + +static void xenvif_rx_complete(struct xenvif_queue *queue, + struct xenvif_pkt_state *pkt) +{ + /* All responses are ready to be pushed. */ + queue->rx.rsp_prod_pvt = queue->rx.req_cons; + + __skb_queue_tail(queue->rx_copy.completed, pkt->skb); +} + +static void xenvif_rx_next_frag(struct xenvif_pkt_state *pkt) +{ + struct sk_buff *frag_iter = pkt->frag_iter; + unsigned int nr_frags = skb_shinfo(frag_iter)->nr_frags; + + pkt->frag++; + pkt->frag_offset = 0; + + if (pkt->frag >= nr_frags) { + if (frag_iter == pkt->skb) + pkt->frag_iter = skb_shinfo(frag_iter)->frag_list; + else + pkt->frag_iter = frag_iter->next; + + pkt->frag = -1; + } +} + +static void xenvif_rx_next_chunk(struct xenvif_queue *queue, + struct xenvif_pkt_state *pkt, + unsigned int offset, void **data, + size_t *len) +{ + struct sk_buff *frag_iter = pkt->frag_iter; + void *frag_data; + size_t frag_len, chunk_len; + + BUG_ON(!frag_iter); + + if (pkt->frag == -1) { + frag_data = frag_iter->data; + frag_len = skb_headlen(frag_iter); + } else { + skb_frag_t *frag = &skb_shinfo(frag_iter)->frags[pkt->frag]; + + frag_data = skb_frag_address(frag); + frag_len = skb_frag_size(frag); + } + + frag_data += pkt->frag_offset; + frag_len -= pkt->frag_offset; + + chunk_len = min_t(size_t, frag_len, XEN_PAGE_SIZE - offset); + chunk_len = min_t(size_t, chunk_len, XEN_PAGE_SIZE - + xen_offset_in_page(frag_data)); + + pkt->frag_offset += chunk_len; + + /* Advance to next frag? */ + if (frag_len == chunk_len) + xenvif_rx_next_frag(pkt); + + *data = frag_data; + *len = chunk_len; +} + +static void xenvif_rx_data_slot(struct xenvif_queue *queue, + struct xenvif_pkt_state *pkt, + struct xen_netif_rx_request *req, + struct xen_netif_rx_response *rsp) +{ + unsigned int offset = queue->vif->xdp_headroom; + unsigned int flags; + + do { + size_t len; + void *data; + + xenvif_rx_next_chunk(queue, pkt, offset, &data, &len); + xenvif_rx_copy_add(queue, req, offset, data, len); + + offset += len; + pkt->remaining_len -= len; + + } while (offset < XEN_PAGE_SIZE && pkt->remaining_len > 0); + + if (pkt->remaining_len > 0) + flags = XEN_NETRXF_more_data; + else + flags = 0; + + if (pkt->slot == 0) { + struct sk_buff *skb = pkt->skb; + + if (skb->ip_summed == CHECKSUM_PARTIAL) + flags |= XEN_NETRXF_csum_blank | + XEN_NETRXF_data_validated; + else if (skb->ip_summed == CHECKSUM_UNNECESSARY) + flags |= XEN_NETRXF_data_validated; + + if (pkt->extra_count != 0) + flags |= XEN_NETRXF_extra_info; + } + + rsp->offset = 0; + rsp->flags = flags; + rsp->id = req->id; + rsp->status = (s16)offset; +} + +static void xenvif_rx_extra_slot(struct xenvif_queue *queue, + struct xenvif_pkt_state *pkt, + struct xen_netif_rx_request *req, + struct xen_netif_rx_response *rsp) +{ + struct xen_netif_extra_info *extra = (void *)rsp; + unsigned int i; + + pkt->extra_count--; + + for (i = 0; i < ARRAY_SIZE(pkt->extras); i++) { + if (pkt->extras[i].type) { + *extra = pkt->extras[i]; + + if (pkt->extra_count != 0) + extra->flags |= XEN_NETIF_EXTRA_FLAG_MORE; + + pkt->extras[i].type = 0; + return; + } + } + BUG(); +} + +static void xenvif_rx_skb(struct xenvif_queue *queue) +{ + struct xenvif_pkt_state pkt; + + xenvif_rx_next_skb(queue, &pkt); + + queue->last_rx_time = jiffies; + + do { + struct xen_netif_rx_request *req; + struct xen_netif_rx_response *rsp; + + req = RING_GET_REQUEST(&queue->rx, queue->rx.req_cons); + rsp = RING_GET_RESPONSE(&queue->rx, queue->rx.req_cons); + + /* Extras must go after the first data slot */ + if (pkt.slot != 0 && pkt.extra_count != 0) + xenvif_rx_extra_slot(queue, &pkt, req, rsp); + else + xenvif_rx_data_slot(queue, &pkt, req, rsp); + + queue->rx.req_cons++; + pkt.slot++; + } while (pkt.remaining_len > 0 || pkt.extra_count != 0); + + xenvif_rx_complete(queue, &pkt); +} + +#define RX_BATCH_SIZE 64 + +static void xenvif_rx_action(struct xenvif_queue *queue) +{ + struct sk_buff_head completed_skbs; + unsigned int work_done = 0; + + __skb_queue_head_init(&completed_skbs); + queue->rx_copy.completed = &completed_skbs; + + while (xenvif_rx_ring_slots_available(queue) && + !skb_queue_empty(&queue->rx_queue) && + work_done < RX_BATCH_SIZE) { + xenvif_rx_skb(queue); + work_done++; + } + + /* Flush any pending copies and complete all skbs. */ + xenvif_rx_copy_flush(queue); +} + +static RING_IDX xenvif_rx_queue_slots(const struct xenvif_queue *queue) +{ + RING_IDX prod, cons; + + prod = queue->rx.sring->req_prod; + cons = queue->rx.req_cons; + + return prod - cons; +} + +static bool xenvif_rx_queue_stalled(const struct xenvif_queue *queue) +{ + unsigned int needed = READ_ONCE(queue->rx_slots_needed); + + return !queue->stalled && + xenvif_rx_queue_slots(queue) < needed && + time_after(jiffies, + queue->last_rx_time + queue->vif->stall_timeout); +} + +static bool xenvif_rx_queue_ready(struct xenvif_queue *queue) +{ + unsigned int needed = READ_ONCE(queue->rx_slots_needed); + + return queue->stalled && xenvif_rx_queue_slots(queue) >= needed; +} + +bool xenvif_have_rx_work(struct xenvif_queue *queue, bool test_kthread) +{ + return xenvif_rx_ring_slots_available(queue) || + (queue->vif->stall_timeout && + (xenvif_rx_queue_stalled(queue) || + xenvif_rx_queue_ready(queue))) || + (test_kthread && kthread_should_stop()) || + queue->vif->disabled; +} + +static long xenvif_rx_queue_timeout(struct xenvif_queue *queue) +{ + struct sk_buff *skb; + long timeout; + + skb = skb_peek(&queue->rx_queue); + if (!skb) + return MAX_SCHEDULE_TIMEOUT; + + timeout = XENVIF_RX_CB(skb)->expires - jiffies; + return timeout < 0 ? 0 : timeout; +} + +/* Wait until the guest Rx thread has work. + * + * The timeout needs to be adjusted based on the current head of the + * queue (and not just the head at the beginning). In particular, if + * the queue is initially empty an infinite timeout is used and this + * needs to be reduced when a skb is queued. + * + * This cannot be done with wait_event_timeout() because it only + * calculates the timeout once. + */ +static void xenvif_wait_for_rx_work(struct xenvif_queue *queue) +{ + DEFINE_WAIT(wait); + + if (xenvif_have_rx_work(queue, true)) + return; + + for (;;) { + long ret; + + prepare_to_wait(&queue->wq, &wait, TASK_INTERRUPTIBLE); + if (xenvif_have_rx_work(queue, true)) + break; + if (atomic_fetch_andnot(NETBK_RX_EOI | NETBK_COMMON_EOI, + &queue->eoi_pending) & + (NETBK_RX_EOI | NETBK_COMMON_EOI)) + xen_irq_lateeoi(queue->rx_irq, 0); + + ret = schedule_timeout(xenvif_rx_queue_timeout(queue)); + if (!ret) + break; + } + finish_wait(&queue->wq, &wait); +} + +static void xenvif_queue_carrier_off(struct xenvif_queue *queue) +{ + struct xenvif *vif = queue->vif; + + queue->stalled = true; + + /* At least one queue has stalled? Disable the carrier. */ + spin_lock(&vif->lock); + if (vif->stalled_queues++ == 0) { + netdev_info(vif->dev, "Guest Rx stalled"); + netif_carrier_off(vif->dev); + } + spin_unlock(&vif->lock); +} + +static void xenvif_queue_carrier_on(struct xenvif_queue *queue) +{ + struct xenvif *vif = queue->vif; + + queue->last_rx_time = jiffies; /* Reset Rx stall detection. */ + queue->stalled = false; + + /* All queues are ready? Enable the carrier. */ + spin_lock(&vif->lock); + if (--vif->stalled_queues == 0) { + netdev_info(vif->dev, "Guest Rx ready"); + netif_carrier_on(vif->dev); + } + spin_unlock(&vif->lock); +} + +int xenvif_kthread_guest_rx(void *data) +{ + struct xenvif_queue *queue = data; + struct xenvif *vif = queue->vif; + + if (!vif->stall_timeout) + xenvif_queue_carrier_on(queue); + + for (;;) { + xenvif_wait_for_rx_work(queue); + + if (kthread_should_stop()) + break; + + /* This frontend is found to be rogue, disable it in + * kthread context. Currently this is only set when + * netback finds out frontend sends malformed packet, + * but we cannot disable the interface in softirq + * context so we defer it here, if this thread is + * associated with queue 0. + */ + if (unlikely(vif->disabled && queue->id == 0)) { + xenvif_carrier_off(vif); + break; + } + + if (!skb_queue_empty(&queue->rx_queue)) + xenvif_rx_action(queue); + + /* If the guest hasn't provided any Rx slots for a + * while it's probably not responsive, drop the + * carrier so packets are dropped earlier. + */ + if (vif->stall_timeout) { + if (xenvif_rx_queue_stalled(queue)) + xenvif_queue_carrier_off(queue); + else if (xenvif_rx_queue_ready(queue)) + xenvif_queue_carrier_on(queue); + } + + /* Queued packets may have foreign pages from other + * domains. These cannot be queued indefinitely as + * this would starve guests of grant refs and transmit + * slots. + */ + xenvif_rx_queue_drop_expired(queue); + + cond_resched(); + } + + /* Bin any remaining skbs */ + xenvif_rx_queue_purge(queue); + + return 0; +} diff --git a/drivers/net/xen-netback/xenbus.c b/drivers/net/xen-netback/xenbus.c new file mode 100644 index 000000000..9ee9ce049 --- /dev/null +++ b/drivers/net/xen-netback/xenbus.c @@ -0,0 +1,1181 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Xenbus code for netif backend + * + * Copyright (C) 2005 Rusty Russell <rusty@rustcorp.com.au> + * Copyright (C) 2005 XenSource Ltd +*/ + +#include "common.h" +#include <linux/vmalloc.h> +#include <linux/rtnetlink.h> + +static int connect_data_rings(struct backend_info *be, + struct xenvif_queue *queue); +static void connect(struct backend_info *be); +static int read_xenbus_vif_flags(struct backend_info *be); +static int backend_create_xenvif(struct backend_info *be); +static void unregister_hotplug_status_watch(struct backend_info *be); +static void xen_unregister_watchers(struct xenvif *vif); +static void set_backend_state(struct backend_info *be, + enum xenbus_state state); + +#ifdef CONFIG_DEBUG_FS +struct dentry *xen_netback_dbg_root = NULL; + +static int xenvif_read_io_ring(struct seq_file *m, void *v) +{ + struct xenvif_queue *queue = m->private; + struct xen_netif_tx_back_ring *tx_ring = &queue->tx; + struct xen_netif_rx_back_ring *rx_ring = &queue->rx; + struct netdev_queue *dev_queue; + + if (tx_ring->sring) { + struct xen_netif_tx_sring *sring = tx_ring->sring; + + seq_printf(m, "Queue %d\nTX: nr_ents %u\n", queue->id, + tx_ring->nr_ents); + seq_printf(m, "req prod %u (%d) cons %u (%d) event %u (%d)\n", + sring->req_prod, + sring->req_prod - sring->rsp_prod, + tx_ring->req_cons, + tx_ring->req_cons - sring->rsp_prod, + sring->req_event, + sring->req_event - sring->rsp_prod); + seq_printf(m, "rsp prod %u (base) pvt %u (%d) event %u (%d)\n", + sring->rsp_prod, + tx_ring->rsp_prod_pvt, + tx_ring->rsp_prod_pvt - sring->rsp_prod, + sring->rsp_event, + sring->rsp_event - sring->rsp_prod); + seq_printf(m, "pending prod %u pending cons %u nr_pending_reqs %u\n", + queue->pending_prod, + queue->pending_cons, + nr_pending_reqs(queue)); + seq_printf(m, "dealloc prod %u dealloc cons %u dealloc_queue %u\n\n", + queue->dealloc_prod, + queue->dealloc_cons, + queue->dealloc_prod - queue->dealloc_cons); + } + + if (rx_ring->sring) { + struct xen_netif_rx_sring *sring = rx_ring->sring; + + seq_printf(m, "RX: nr_ents %u\n", rx_ring->nr_ents); + seq_printf(m, "req prod %u (%d) cons %u (%d) event %u (%d)\n", + sring->req_prod, + sring->req_prod - sring->rsp_prod, + rx_ring->req_cons, + rx_ring->req_cons - sring->rsp_prod, + sring->req_event, + sring->req_event - sring->rsp_prod); + seq_printf(m, "rsp prod %u (base) pvt %u (%d) event %u (%d)\n\n", + sring->rsp_prod, + rx_ring->rsp_prod_pvt, + rx_ring->rsp_prod_pvt - sring->rsp_prod, + sring->rsp_event, + sring->rsp_event - sring->rsp_prod); + } + + seq_printf(m, "NAPI state: %lx NAPI weight: %d TX queue len %u\n" + "Credit timer_pending: %d, credit: %lu, usec: %lu\n" + "remaining: %lu, expires: %lu, now: %lu\n", + queue->napi.state, queue->napi.weight, + skb_queue_len(&queue->tx_queue), + timer_pending(&queue->credit_timeout), + queue->credit_bytes, + queue->credit_usec, + queue->remaining_credit, + queue->credit_timeout.expires, + jiffies); + + dev_queue = netdev_get_tx_queue(queue->vif->dev, queue->id); + + seq_printf(m, "\nRx internal queue: len %u max %u pkts %u %s\n", + queue->rx_queue_len, queue->rx_queue_max, + skb_queue_len(&queue->rx_queue), + netif_tx_queue_stopped(dev_queue) ? "stopped" : "running"); + + return 0; +} + +#define XENVIF_KICK_STR "kick" +#define BUFFER_SIZE 32 + +static ssize_t +xenvif_write_io_ring(struct file *filp, const char __user *buf, size_t count, + loff_t *ppos) +{ + struct xenvif_queue *queue = + ((struct seq_file *)filp->private_data)->private; + int len; + char write[BUFFER_SIZE]; + + /* don't allow partial writes and check the length */ + if (*ppos != 0) + return 0; + if (count >= sizeof(write)) + return -ENOSPC; + + len = simple_write_to_buffer(write, + sizeof(write) - 1, + ppos, + buf, + count); + if (len < 0) + return len; + + write[len] = '\0'; + + if (!strncmp(write, XENVIF_KICK_STR, sizeof(XENVIF_KICK_STR) - 1)) + xenvif_interrupt(0, (void *)queue); + else { + pr_warn("Unknown command to io_ring_q%d. Available: kick\n", + queue->id); + count = -EINVAL; + } + return count; +} + +static int xenvif_io_ring_open(struct inode *inode, struct file *filp) +{ + int ret; + void *queue = NULL; + + if (inode->i_private) + queue = inode->i_private; + ret = single_open(filp, xenvif_read_io_ring, queue); + filp->f_mode |= FMODE_PWRITE; + return ret; +} + +static const struct file_operations xenvif_dbg_io_ring_ops_fops = { + .owner = THIS_MODULE, + .open = xenvif_io_ring_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, + .write = xenvif_write_io_ring, +}; + +static int xenvif_ctrl_show(struct seq_file *m, void *v) +{ + struct xenvif *vif = m->private; + + xenvif_dump_hash_info(vif, m); + + return 0; +} +DEFINE_SHOW_ATTRIBUTE(xenvif_ctrl); + +static void xenvif_debugfs_addif(struct xenvif *vif) +{ + int i; + + vif->xenvif_dbg_root = debugfs_create_dir(vif->dev->name, + xen_netback_dbg_root); + for (i = 0; i < vif->num_queues; ++i) { + char filename[sizeof("io_ring_q") + 4]; + + snprintf(filename, sizeof(filename), "io_ring_q%d", i); + debugfs_create_file(filename, 0600, vif->xenvif_dbg_root, + &vif->queues[i], + &xenvif_dbg_io_ring_ops_fops); + } + + if (vif->ctrl_irq) + debugfs_create_file("ctrl", 0400, vif->xenvif_dbg_root, vif, + &xenvif_ctrl_fops); +} + +static void xenvif_debugfs_delif(struct xenvif *vif) +{ + debugfs_remove_recursive(vif->xenvif_dbg_root); + vif->xenvif_dbg_root = NULL; +} +#endif /* CONFIG_DEBUG_FS */ + +/* + * Handle the creation of the hotplug script environment. We add the script + * and vif variables to the environment, for the benefit of the vif-* hotplug + * scripts. + */ +static int netback_uevent(struct xenbus_device *xdev, + struct kobj_uevent_env *env) +{ + struct backend_info *be = dev_get_drvdata(&xdev->dev); + + if (!be) + return 0; + + if (add_uevent_var(env, "script=%s", be->hotplug_script)) + return -ENOMEM; + + if (!be->vif) + return 0; + + return add_uevent_var(env, "vif=%s", be->vif->dev->name); +} + + +static int backend_create_xenvif(struct backend_info *be) +{ + int err; + long handle; + struct xenbus_device *dev = be->dev; + struct xenvif *vif; + + if (be->vif != NULL) + return 0; + + err = xenbus_scanf(XBT_NIL, dev->nodename, "handle", "%li", &handle); + if (err != 1) { + xenbus_dev_fatal(dev, err, "reading handle"); + return (err < 0) ? err : -EINVAL; + } + + vif = xenvif_alloc(&dev->dev, dev->otherend_id, handle); + if (IS_ERR(vif)) { + err = PTR_ERR(vif); + xenbus_dev_fatal(dev, err, "creating interface"); + return err; + } + be->vif = vif; + vif->be = be; + + kobject_uevent(&dev->dev.kobj, KOBJ_ONLINE); + return 0; +} + +static void backend_disconnect(struct backend_info *be) +{ + struct xenvif *vif = be->vif; + + if (vif) { + unsigned int num_queues = vif->num_queues; + unsigned int queue_index; + + xen_unregister_watchers(vif); +#ifdef CONFIG_DEBUG_FS + xenvif_debugfs_delif(vif); +#endif /* CONFIG_DEBUG_FS */ + xenvif_disconnect_data(vif); + + /* At this point some of the handlers may still be active + * so we need to have additional synchronization here. + */ + vif->num_queues = 0; + synchronize_net(); + + for (queue_index = 0; queue_index < num_queues; ++queue_index) + xenvif_deinit_queue(&vif->queues[queue_index]); + + vfree(vif->queues); + vif->queues = NULL; + + xenvif_disconnect_ctrl(vif); + } +} + +static void backend_connect(struct backend_info *be) +{ + if (be->vif) + connect(be); +} + +static inline void backend_switch_state(struct backend_info *be, + enum xenbus_state state) +{ + struct xenbus_device *dev = be->dev; + + pr_debug("%s -> %s\n", dev->nodename, xenbus_strstate(state)); + be->state = state; + + /* If we are waiting for a hotplug script then defer the + * actual xenbus state change. + */ + if (!be->have_hotplug_status_watch) + xenbus_switch_state(dev, state); +} + +/* Handle backend state transitions: + * + * The backend state starts in Initialising and the following transitions are + * allowed. + * + * Initialising -> InitWait -> Connected + * \ + * \ ^ \ | + * \ | \ | + * \ | \ | + * \ | \ | + * \ | \ | + * \ | \ | + * V | V V + * + * Closed <-> Closing + * + * The state argument specifies the eventual state of the backend and the + * function transitions to that state via the shortest path. + */ +static void set_backend_state(struct backend_info *be, + enum xenbus_state state) +{ + while (be->state != state) { + switch (be->state) { + case XenbusStateInitialising: + switch (state) { + case XenbusStateInitWait: + case XenbusStateConnected: + case XenbusStateClosing: + backend_switch_state(be, XenbusStateInitWait); + break; + case XenbusStateClosed: + backend_switch_state(be, XenbusStateClosed); + break; + default: + BUG(); + } + break; + case XenbusStateClosed: + switch (state) { + case XenbusStateInitWait: + case XenbusStateConnected: + backend_switch_state(be, XenbusStateInitWait); + break; + case XenbusStateClosing: + backend_switch_state(be, XenbusStateClosing); + break; + default: + BUG(); + } + break; + case XenbusStateInitWait: + switch (state) { + case XenbusStateConnected: + backend_connect(be); + backend_switch_state(be, XenbusStateConnected); + break; + case XenbusStateClosing: + case XenbusStateClosed: + backend_switch_state(be, XenbusStateClosing); + break; + default: + BUG(); + } + break; + case XenbusStateConnected: + switch (state) { + case XenbusStateInitWait: + case XenbusStateClosing: + case XenbusStateClosed: + backend_disconnect(be); + backend_switch_state(be, XenbusStateClosing); + break; + default: + BUG(); + } + break; + case XenbusStateClosing: + switch (state) { + case XenbusStateInitWait: + case XenbusStateConnected: + case XenbusStateClosed: + backend_switch_state(be, XenbusStateClosed); + break; + default: + BUG(); + } + break; + default: + BUG(); + } + } +} + +static void read_xenbus_frontend_xdp(struct backend_info *be, + struct xenbus_device *dev) +{ + struct xenvif *vif = be->vif; + u16 headroom; + int err; + + err = xenbus_scanf(XBT_NIL, dev->otherend, + "xdp-headroom", "%hu", &headroom); + if (err != 1) { + vif->xdp_headroom = 0; + return; + } + if (headroom > XEN_NETIF_MAX_XDP_HEADROOM) + headroom = XEN_NETIF_MAX_XDP_HEADROOM; + vif->xdp_headroom = headroom; +} + +/** + * Callback received when the frontend's state changes. + */ +static void frontend_changed(struct xenbus_device *dev, + enum xenbus_state frontend_state) +{ + struct backend_info *be = dev_get_drvdata(&dev->dev); + + pr_debug("%s -> %s\n", dev->otherend, xenbus_strstate(frontend_state)); + + be->frontend_state = frontend_state; + + switch (frontend_state) { + case XenbusStateInitialising: + set_backend_state(be, XenbusStateInitWait); + break; + + case XenbusStateInitialised: + break; + + case XenbusStateConnected: + set_backend_state(be, XenbusStateConnected); + break; + + case XenbusStateReconfiguring: + read_xenbus_frontend_xdp(be, dev); + xenbus_switch_state(dev, XenbusStateReconfigured); + break; + + case XenbusStateClosing: + set_backend_state(be, XenbusStateClosing); + break; + + case XenbusStateClosed: + set_backend_state(be, XenbusStateClosed); + if (xenbus_dev_is_online(dev)) + break; + fallthrough; /* if not online */ + case XenbusStateUnknown: + set_backend_state(be, XenbusStateClosed); + device_unregister(&dev->dev); + break; + + default: + xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend", + frontend_state); + break; + } +} + + +static void xen_net_read_rate(struct xenbus_device *dev, + unsigned long *bytes, unsigned long *usec) +{ + char *s, *e; + unsigned long b, u; + char *ratestr; + + /* Default to unlimited bandwidth. */ + *bytes = ~0UL; + *usec = 0; + + ratestr = xenbus_read(XBT_NIL, dev->nodename, "rate", NULL); + if (IS_ERR(ratestr)) + return; + + s = ratestr; + b = simple_strtoul(s, &e, 10); + if ((s == e) || (*e != ',')) + goto fail; + + s = e + 1; + u = simple_strtoul(s, &e, 10); + if ((s == e) || (*e != '\0')) + goto fail; + + *bytes = b; + *usec = u; + + kfree(ratestr); + return; + + fail: + pr_warn("Failed to parse network rate limit. Traffic unlimited.\n"); + kfree(ratestr); +} + +static int xen_net_read_mac(struct xenbus_device *dev, u8 mac[]) +{ + char *s, *e, *macstr; + int i; + + macstr = s = xenbus_read(XBT_NIL, dev->nodename, "mac", NULL); + if (IS_ERR(macstr)) + return PTR_ERR(macstr); + + for (i = 0; i < ETH_ALEN; i++) { + mac[i] = simple_strtoul(s, &e, 16); + if ((s == e) || (*e != ((i == ETH_ALEN-1) ? '\0' : ':'))) { + kfree(macstr); + return -ENOENT; + } + s = e+1; + } + + kfree(macstr); + return 0; +} + +static void xen_net_rate_changed(struct xenbus_watch *watch, + const char *path, const char *token) +{ + struct xenvif *vif = container_of(watch, struct xenvif, credit_watch); + struct xenbus_device *dev = xenvif_to_xenbus_device(vif); + unsigned long credit_bytes; + unsigned long credit_usec; + unsigned int queue_index; + + xen_net_read_rate(dev, &credit_bytes, &credit_usec); + for (queue_index = 0; queue_index < vif->num_queues; queue_index++) { + struct xenvif_queue *queue = &vif->queues[queue_index]; + + queue->credit_bytes = credit_bytes; + queue->credit_usec = credit_usec; + if (!mod_timer_pending(&queue->credit_timeout, jiffies) && + queue->remaining_credit > queue->credit_bytes) { + queue->remaining_credit = queue->credit_bytes; + } + } +} + +static int xen_register_credit_watch(struct xenbus_device *dev, + struct xenvif *vif) +{ + int err = 0; + char *node; + unsigned maxlen = strlen(dev->nodename) + sizeof("/rate"); + + if (vif->credit_watch.node) + return -EADDRINUSE; + + node = kmalloc(maxlen, GFP_KERNEL); + if (!node) + return -ENOMEM; + snprintf(node, maxlen, "%s/rate", dev->nodename); + vif->credit_watch.node = node; + vif->credit_watch.will_handle = NULL; + vif->credit_watch.callback = xen_net_rate_changed; + err = register_xenbus_watch(&vif->credit_watch); + if (err) { + pr_err("Failed to set watcher %s\n", vif->credit_watch.node); + kfree(node); + vif->credit_watch.node = NULL; + vif->credit_watch.will_handle = NULL; + vif->credit_watch.callback = NULL; + } + return err; +} + +static void xen_unregister_credit_watch(struct xenvif *vif) +{ + if (vif->credit_watch.node) { + unregister_xenbus_watch(&vif->credit_watch); + kfree(vif->credit_watch.node); + vif->credit_watch.node = NULL; + } +} + +static void xen_mcast_ctrl_changed(struct xenbus_watch *watch, + const char *path, const char *token) +{ + struct xenvif *vif = container_of(watch, struct xenvif, + mcast_ctrl_watch); + struct xenbus_device *dev = xenvif_to_xenbus_device(vif); + + vif->multicast_control = !!xenbus_read_unsigned(dev->otherend, + "request-multicast-control", 0); +} + +static int xen_register_mcast_ctrl_watch(struct xenbus_device *dev, + struct xenvif *vif) +{ + int err = 0; + char *node; + unsigned maxlen = strlen(dev->otherend) + + sizeof("/request-multicast-control"); + + if (vif->mcast_ctrl_watch.node) { + pr_err_ratelimited("Watch is already registered\n"); + return -EADDRINUSE; + } + + node = kmalloc(maxlen, GFP_KERNEL); + if (!node) { + pr_err("Failed to allocate memory for watch\n"); + return -ENOMEM; + } + snprintf(node, maxlen, "%s/request-multicast-control", + dev->otherend); + vif->mcast_ctrl_watch.node = node; + vif->mcast_ctrl_watch.will_handle = NULL; + vif->mcast_ctrl_watch.callback = xen_mcast_ctrl_changed; + err = register_xenbus_watch(&vif->mcast_ctrl_watch); + if (err) { + pr_err("Failed to set watcher %s\n", + vif->mcast_ctrl_watch.node); + kfree(node); + vif->mcast_ctrl_watch.node = NULL; + vif->mcast_ctrl_watch.will_handle = NULL; + vif->mcast_ctrl_watch.callback = NULL; + } + return err; +} + +static void xen_unregister_mcast_ctrl_watch(struct xenvif *vif) +{ + if (vif->mcast_ctrl_watch.node) { + unregister_xenbus_watch(&vif->mcast_ctrl_watch); + kfree(vif->mcast_ctrl_watch.node); + vif->mcast_ctrl_watch.node = NULL; + } +} + +static void xen_register_watchers(struct xenbus_device *dev, + struct xenvif *vif) +{ + xen_register_credit_watch(dev, vif); + xen_register_mcast_ctrl_watch(dev, vif); +} + +static void xen_unregister_watchers(struct xenvif *vif) +{ + xen_unregister_mcast_ctrl_watch(vif); + xen_unregister_credit_watch(vif); +} + +static void unregister_hotplug_status_watch(struct backend_info *be) +{ + if (be->have_hotplug_status_watch) { + unregister_xenbus_watch(&be->hotplug_status_watch); + kfree(be->hotplug_status_watch.node); + } + be->have_hotplug_status_watch = 0; +} + +static void hotplug_status_changed(struct xenbus_watch *watch, + const char *path, + const char *token) +{ + struct backend_info *be = container_of(watch, + struct backend_info, + hotplug_status_watch); + char *str; + unsigned int len; + + str = xenbus_read(XBT_NIL, be->dev->nodename, "hotplug-status", &len); + if (IS_ERR(str)) + return; + if (len == sizeof("connected")-1 && !memcmp(str, "connected", len)) { + /* Complete any pending state change */ + xenbus_switch_state(be->dev, be->state); + + /* Not interested in this watch anymore. */ + unregister_hotplug_status_watch(be); + } + kfree(str); +} + +static int connect_ctrl_ring(struct backend_info *be) +{ + struct xenbus_device *dev = be->dev; + struct xenvif *vif = be->vif; + unsigned int val; + grant_ref_t ring_ref; + unsigned int evtchn; + int err; + + err = xenbus_scanf(XBT_NIL, dev->otherend, + "ctrl-ring-ref", "%u", &val); + if (err < 0) + goto done; /* The frontend does not have a control ring */ + + ring_ref = val; + + err = xenbus_scanf(XBT_NIL, dev->otherend, + "event-channel-ctrl", "%u", &val); + if (err < 0) { + xenbus_dev_fatal(dev, err, + "reading %s/event-channel-ctrl", + dev->otherend); + goto fail; + } + + evtchn = val; + + err = xenvif_connect_ctrl(vif, ring_ref, evtchn); + if (err) { + xenbus_dev_fatal(dev, err, + "mapping shared-frame %u port %u", + ring_ref, evtchn); + goto fail; + } + +done: + return 0; + +fail: + return err; +} + +static void connect(struct backend_info *be) +{ + int err; + struct xenbus_device *dev = be->dev; + unsigned long credit_bytes, credit_usec; + unsigned int queue_index; + unsigned int requested_num_queues; + struct xenvif_queue *queue; + + /* Check whether the frontend requested multiple queues + * and read the number requested. + */ + requested_num_queues = xenbus_read_unsigned(dev->otherend, + "multi-queue-num-queues", 1); + if (requested_num_queues > xenvif_max_queues) { + /* buggy or malicious guest */ + xenbus_dev_fatal(dev, -EINVAL, + "guest requested %u queues, exceeding the maximum of %u.", + requested_num_queues, xenvif_max_queues); + return; + } + + err = xen_net_read_mac(dev, be->vif->fe_dev_addr); + if (err) { + xenbus_dev_fatal(dev, err, "parsing %s/mac", dev->nodename); + return; + } + + xen_net_read_rate(dev, &credit_bytes, &credit_usec); + xen_unregister_watchers(be->vif); + xen_register_watchers(dev, be->vif); + read_xenbus_vif_flags(be); + + err = connect_ctrl_ring(be); + if (err) { + xenbus_dev_fatal(dev, err, "connecting control ring"); + return; + } + + /* Use the number of queues requested by the frontend */ + be->vif->queues = vzalloc(array_size(requested_num_queues, + sizeof(struct xenvif_queue))); + if (!be->vif->queues) { + xenbus_dev_fatal(dev, -ENOMEM, + "allocating queues"); + return; + } + + be->vif->num_queues = requested_num_queues; + be->vif->stalled_queues = requested_num_queues; + + for (queue_index = 0; queue_index < requested_num_queues; ++queue_index) { + queue = &be->vif->queues[queue_index]; + queue->vif = be->vif; + queue->id = queue_index; + snprintf(queue->name, sizeof(queue->name), "%s-q%u", + be->vif->dev->name, queue->id); + + err = xenvif_init_queue(queue); + if (err) { + /* xenvif_init_queue() cleans up after itself on + * failure, but we need to clean up any previously + * initialised queues. Set num_queues to i so that + * earlier queues can be destroyed using the regular + * disconnect logic. + */ + be->vif->num_queues = queue_index; + goto err; + } + + queue->credit_bytes = credit_bytes; + queue->remaining_credit = credit_bytes; + queue->credit_usec = credit_usec; + + err = connect_data_rings(be, queue); + if (err) { + /* connect_data_rings() cleans up after itself on + * failure, but we need to clean up after + * xenvif_init_queue() here, and also clean up any + * previously initialised queues. + */ + xenvif_deinit_queue(queue); + be->vif->num_queues = queue_index; + goto err; + } + } + +#ifdef CONFIG_DEBUG_FS + xenvif_debugfs_addif(be->vif); +#endif /* CONFIG_DEBUG_FS */ + + /* Initialisation completed, tell core driver the number of + * active queues. + */ + rtnl_lock(); + netif_set_real_num_tx_queues(be->vif->dev, requested_num_queues); + netif_set_real_num_rx_queues(be->vif->dev, requested_num_queues); + rtnl_unlock(); + + xenvif_carrier_on(be->vif); + + unregister_hotplug_status_watch(be); + err = xenbus_watch_pathfmt(dev, &be->hotplug_status_watch, NULL, + hotplug_status_changed, + "%s/%s", dev->nodename, "hotplug-status"); + if (!err) + be->have_hotplug_status_watch = 1; + + netif_tx_wake_all_queues(be->vif->dev); + + return; + +err: + if (be->vif->num_queues > 0) + xenvif_disconnect_data(be->vif); /* Clean up existing queues */ + for (queue_index = 0; queue_index < be->vif->num_queues; ++queue_index) + xenvif_deinit_queue(&be->vif->queues[queue_index]); + vfree(be->vif->queues); + be->vif->queues = NULL; + be->vif->num_queues = 0; + xenvif_disconnect_ctrl(be->vif); + return; +} + + +static int connect_data_rings(struct backend_info *be, + struct xenvif_queue *queue) +{ + struct xenbus_device *dev = be->dev; + unsigned int num_queues = queue->vif->num_queues; + unsigned long tx_ring_ref, rx_ring_ref; + unsigned int tx_evtchn, rx_evtchn; + int err; + char *xspath; + size_t xspathsize; + const size_t xenstore_path_ext_size = 11; /* sufficient for "/queue-NNN" */ + + /* If the frontend requested 1 queue, or we have fallen back + * to single queue due to lack of frontend support for multi- + * queue, expect the remaining XenStore keys in the toplevel + * directory. Otherwise, expect them in a subdirectory called + * queue-N. + */ + if (num_queues == 1) { + xspath = kzalloc(strlen(dev->otherend) + 1, GFP_KERNEL); + if (!xspath) { + xenbus_dev_fatal(dev, -ENOMEM, + "reading ring references"); + return -ENOMEM; + } + strcpy(xspath, dev->otherend); + } else { + xspathsize = strlen(dev->otherend) + xenstore_path_ext_size; + xspath = kzalloc(xspathsize, GFP_KERNEL); + if (!xspath) { + xenbus_dev_fatal(dev, -ENOMEM, + "reading ring references"); + return -ENOMEM; + } + snprintf(xspath, xspathsize, "%s/queue-%u", dev->otherend, + queue->id); + } + + err = xenbus_gather(XBT_NIL, xspath, + "tx-ring-ref", "%lu", &tx_ring_ref, + "rx-ring-ref", "%lu", &rx_ring_ref, NULL); + if (err) { + xenbus_dev_fatal(dev, err, + "reading %s/ring-ref", + xspath); + goto err; + } + + /* Try split event channels first, then single event channel. */ + err = xenbus_gather(XBT_NIL, xspath, + "event-channel-tx", "%u", &tx_evtchn, + "event-channel-rx", "%u", &rx_evtchn, NULL); + if (err < 0) { + err = xenbus_scanf(XBT_NIL, xspath, + "event-channel", "%u", &tx_evtchn); + if (err < 0) { + xenbus_dev_fatal(dev, err, + "reading %s/event-channel(-tx/rx)", + xspath); + goto err; + } + rx_evtchn = tx_evtchn; + } + + /* Map the shared frame, irq etc. */ + err = xenvif_connect_data(queue, tx_ring_ref, rx_ring_ref, + tx_evtchn, rx_evtchn); + if (err) { + xenbus_dev_fatal(dev, err, + "mapping shared-frames %lu/%lu port tx %u rx %u", + tx_ring_ref, rx_ring_ref, + tx_evtchn, rx_evtchn); + goto err; + } + + err = 0; +err: /* Regular return falls through with err == 0 */ + kfree(xspath); + return err; +} + +static int read_xenbus_vif_flags(struct backend_info *be) +{ + struct xenvif *vif = be->vif; + struct xenbus_device *dev = be->dev; + unsigned int rx_copy; + int err; + + err = xenbus_scanf(XBT_NIL, dev->otherend, "request-rx-copy", "%u", + &rx_copy); + if (err == -ENOENT) { + err = 0; + rx_copy = 0; + } + if (err < 0) { + xenbus_dev_fatal(dev, err, "reading %s/request-rx-copy", + dev->otherend); + return err; + } + if (!rx_copy) + return -EOPNOTSUPP; + + if (!xenbus_read_unsigned(dev->otherend, "feature-rx-notify", 0)) { + /* - Reduce drain timeout to poll more frequently for + * Rx requests. + * - Disable Rx stall detection. + */ + be->vif->drain_timeout = msecs_to_jiffies(30); + be->vif->stall_timeout = 0; + } + + vif->can_sg = !!xenbus_read_unsigned(dev->otherend, "feature-sg", 0); + + vif->gso_mask = 0; + + if (xenbus_read_unsigned(dev->otherend, "feature-gso-tcpv4", 0)) + vif->gso_mask |= GSO_BIT(TCPV4); + + if (xenbus_read_unsigned(dev->otherend, "feature-gso-tcpv6", 0)) + vif->gso_mask |= GSO_BIT(TCPV6); + + vif->ip_csum = !xenbus_read_unsigned(dev->otherend, + "feature-no-csum-offload", 0); + + vif->ipv6_csum = !!xenbus_read_unsigned(dev->otherend, + "feature-ipv6-csum-offload", 0); + + read_xenbus_frontend_xdp(be, dev); + + return 0; +} + +static int netback_remove(struct xenbus_device *dev) +{ + struct backend_info *be = dev_get_drvdata(&dev->dev); + + unregister_hotplug_status_watch(be); + xenbus_rm(XBT_NIL, dev->nodename, "hotplug-status"); + if (be->vif) { + kobject_uevent(&dev->dev.kobj, KOBJ_OFFLINE); + backend_disconnect(be); + xenvif_free(be->vif); + be->vif = NULL; + } + kfree(be->hotplug_script); + kfree(be); + dev_set_drvdata(&dev->dev, NULL); + return 0; +} + +/** + * Entry point to this code when a new device is created. Allocate the basic + * structures and switch to InitWait. + */ +static int netback_probe(struct xenbus_device *dev, + const struct xenbus_device_id *id) +{ + const char *message; + struct xenbus_transaction xbt; + int err; + int sg; + const char *script; + struct backend_info *be = kzalloc(sizeof(*be), GFP_KERNEL); + + if (!be) { + xenbus_dev_fatal(dev, -ENOMEM, + "allocating backend structure"); + return -ENOMEM; + } + + be->dev = dev; + dev_set_drvdata(&dev->dev, be); + + sg = 1; + + do { + err = xenbus_transaction_start(&xbt); + if (err) { + xenbus_dev_fatal(dev, err, "starting transaction"); + goto fail; + } + + err = xenbus_printf(xbt, dev->nodename, "feature-sg", "%d", sg); + if (err) { + message = "writing feature-sg"; + goto abort_transaction; + } + + err = xenbus_printf(xbt, dev->nodename, "feature-gso-tcpv4", + "%d", sg); + if (err) { + message = "writing feature-gso-tcpv4"; + goto abort_transaction; + } + + err = xenbus_printf(xbt, dev->nodename, "feature-gso-tcpv6", + "%d", sg); + if (err) { + message = "writing feature-gso-tcpv6"; + goto abort_transaction; + } + + /* We support partial checksum setup for IPv6 packets */ + err = xenbus_printf(xbt, dev->nodename, + "feature-ipv6-csum-offload", + "%d", 1); + if (err) { + message = "writing feature-ipv6-csum-offload"; + goto abort_transaction; + } + + /* We support rx-copy path. */ + err = xenbus_printf(xbt, dev->nodename, + "feature-rx-copy", "%d", 1); + if (err) { + message = "writing feature-rx-copy"; + goto abort_transaction; + } + + /* we can adjust a headroom for netfront XDP processing */ + err = xenbus_printf(xbt, dev->nodename, + "feature-xdp-headroom", "%d", + provides_xdp_headroom); + if (err) { + message = "writing feature-xdp-headroom"; + goto abort_transaction; + } + + /* We don't support rx-flip path (except old guests who + * don't grok this feature flag). + */ + err = xenbus_printf(xbt, dev->nodename, + "feature-rx-flip", "%d", 0); + if (err) { + message = "writing feature-rx-flip"; + goto abort_transaction; + } + + /* We support dynamic multicast-control. */ + err = xenbus_printf(xbt, dev->nodename, + "feature-multicast-control", "%d", 1); + if (err) { + message = "writing feature-multicast-control"; + goto abort_transaction; + } + + err = xenbus_printf(xbt, dev->nodename, + "feature-dynamic-multicast-control", + "%d", 1); + if (err) { + message = "writing feature-dynamic-multicast-control"; + goto abort_transaction; + } + + err = xenbus_transaction_end(xbt, 0); + } while (err == -EAGAIN); + + if (err) { + xenbus_dev_fatal(dev, err, "completing transaction"); + goto fail; + } + + /* Split event channels support, this is optional so it is not + * put inside the above loop. + */ + err = xenbus_printf(XBT_NIL, dev->nodename, + "feature-split-event-channels", + "%u", separate_tx_rx_irq); + if (err) + pr_debug("Error writing feature-split-event-channels\n"); + + /* Multi-queue support: This is an optional feature. */ + err = xenbus_printf(XBT_NIL, dev->nodename, + "multi-queue-max-queues", "%u", xenvif_max_queues); + if (err) + pr_debug("Error writing multi-queue-max-queues\n"); + + err = xenbus_printf(XBT_NIL, dev->nodename, + "feature-ctrl-ring", + "%u", true); + if (err) + pr_debug("Error writing feature-ctrl-ring\n"); + + backend_switch_state(be, XenbusStateInitWait); + + script = xenbus_read(XBT_NIL, dev->nodename, "script", NULL); + if (IS_ERR(script)) { + err = PTR_ERR(script); + xenbus_dev_fatal(dev, err, "reading script"); + goto fail; + } + + be->hotplug_script = script; + + /* This kicks hotplug scripts, so do it immediately. */ + err = backend_create_xenvif(be); + if (err) + goto fail; + + return 0; + +abort_transaction: + xenbus_transaction_end(xbt, 1); + xenbus_dev_fatal(dev, err, "%s", message); +fail: + pr_debug("failed\n"); + netback_remove(dev); + return err; +} + +static const struct xenbus_device_id netback_ids[] = { + { "vif" }, + { "" } +}; + +static struct xenbus_driver netback_driver = { + .ids = netback_ids, + .probe = netback_probe, + .remove = netback_remove, + .uevent = netback_uevent, + .otherend_changed = frontend_changed, + .allow_rebind = true, +}; + +int xenvif_xenbus_init(void) +{ + return xenbus_register_backend(&netback_driver); +} + +void xenvif_xenbus_fini(void) +{ + return xenbus_unregister_driver(&netback_driver); +} |