diff options
Diffstat (limited to 'net/core/dev.c')
-rw-r--r-- | net/core/dev.c | 514 |
1 files changed, 387 insertions, 127 deletions
diff --git a/net/core/dev.c b/net/core/dev.c index 343883f65ea7..ed4550fd9ece 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -126,6 +126,7 @@ #include <linux/in.h> #include <linux/jhash.h> #include <linux/random.h> +#include <trace/napi.h> #include "net-sysfs.h" @@ -1336,7 +1337,12 @@ static void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev) { struct packet_type *ptype; +#ifdef CONFIG_NET_CLS_ACT + if (!(skb->tstamp.tv64 && (G_TC_FROM(skb->tc_verd) & AT_INGRESS))) + net_timestamp(skb); +#else net_timestamp(skb); +#endif rcu_read_lock(); list_for_each_entry_rcu(ptype, &ptype_all, list) { @@ -1683,7 +1689,17 @@ int dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev, goto gso; } + /* + * If device doesnt need skb->dst, release it right now while + * its hot in this cpu cache + */ + if ((dev->priv_flags & IFF_XMIT_DST_RELEASE) && skb->dst) { + dst_release(skb->dst); + skb->dst = NULL; + } rc = ops->ndo_start_xmit(skb, dev); + if (rc == 0) + txq_trans_update(txq); /* * TODO: if skb_orphan() was called by * dev->hard_start_xmit() (for example, the unmodified @@ -1713,6 +1729,7 @@ gso: skb->next = nskb; return rc; } + txq_trans_update(txq); if (unlikely(netif_tx_queue_stopped(txq) && skb->next)) return NETDEV_TX_BUSY; } while (skb->next); @@ -1732,9 +1749,14 @@ u16 skb_tx_hash(const struct net_device *dev, const struct sk_buff *skb) if (skb_rx_queue_recorded(skb)) { hash = skb_get_rx_queue(skb); - } else if (skb->sk && skb->sk->sk_hash) { + while (unlikely (hash >= dev->real_num_tx_queues)) + hash -= dev->real_num_tx_queues; + return hash; + } + + if (skb->sk && skb->sk->sk_hash) hash = skb->sk->sk_hash; - } else + else hash = skb->protocol; hash = jhash_1word(hash, skb_tx_hashrnd); @@ -2368,26 +2390,6 @@ void napi_gro_flush(struct napi_struct *napi) } EXPORT_SYMBOL(napi_gro_flush); -void *skb_gro_header(struct sk_buff *skb, unsigned int hlen) -{ - unsigned int offset = skb_gro_offset(skb); - - hlen += offset; - if (hlen <= skb_headlen(skb)) - return skb->data + offset; - - if (unlikely(!skb_shinfo(skb)->nr_frags || - skb_shinfo(skb)->frags[0].size <= - hlen - skb_headlen(skb) || - PageHighMem(skb_shinfo(skb)->frags[0].page))) - return pskb_may_pull(skb, hlen) ? skb->data + offset : NULL; - - return page_address(skb_shinfo(skb)->frags[0].page) + - skb_shinfo(skb)->frags[0].page_offset + - offset - skb_headlen(skb); -} -EXPORT_SYMBOL(skb_gro_header); - int dev_gro_receive(struct napi_struct *napi, struct sk_buff *skb) { struct sk_buff **pp = NULL; @@ -2450,10 +2452,25 @@ int dev_gro_receive(struct napi_struct *napi, struct sk_buff *skb) ret = GRO_HELD; pull: - if (unlikely(!pskb_may_pull(skb, skb_gro_offset(skb)))) { - if (napi->gro_list == skb) - napi->gro_list = skb->next; - ret = GRO_DROP; + if (skb_headlen(skb) < skb_gro_offset(skb)) { + int grow = skb_gro_offset(skb) - skb_headlen(skb); + + BUG_ON(skb->end - skb->tail < grow); + + memcpy(skb_tail_pointer(skb), NAPI_GRO_CB(skb)->frag0, grow); + + skb->tail += grow; + skb->data_len -= grow; + + skb_shinfo(skb)->frags[0].page_offset += grow; + skb_shinfo(skb)->frags[0].size -= grow; + + if (unlikely(!skb_shinfo(skb)->frags[0].size)) { + put_page(skb_shinfo(skb)->frags[0].page); + memmove(skb_shinfo(skb)->frags, + skb_shinfo(skb)->frags + 1, + --skb_shinfo(skb)->nr_frags); + } } ok: @@ -2503,6 +2520,22 @@ int napi_skb_finish(int ret, struct sk_buff *skb) } EXPORT_SYMBOL(napi_skb_finish); +void skb_gro_reset_offset(struct sk_buff *skb) +{ + NAPI_GRO_CB(skb)->data_offset = 0; + NAPI_GRO_CB(skb)->frag0 = NULL; + NAPI_GRO_CB(skb)->frag0_len = 0; + + if (skb->mac_header == skb->tail && + !PageHighMem(skb_shinfo(skb)->frags[0].page)) { + NAPI_GRO_CB(skb)->frag0 = + page_address(skb_shinfo(skb)->frags[0].page) + + skb_shinfo(skb)->frags[0].page_offset; + NAPI_GRO_CB(skb)->frag0_len = skb_shinfo(skb)->frags[0].size; + } +} +EXPORT_SYMBOL(skb_gro_reset_offset); + int napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb) { skb_gro_reset_offset(skb); @@ -2520,16 +2553,10 @@ void napi_reuse_skb(struct napi_struct *napi, struct sk_buff *skb) } EXPORT_SYMBOL(napi_reuse_skb); -struct sk_buff *napi_fraginfo_skb(struct napi_struct *napi, - struct napi_gro_fraginfo *info) +struct sk_buff *napi_get_frags(struct napi_struct *napi) { struct net_device *dev = napi->dev; struct sk_buff *skb = napi->skb; - struct ethhdr *eth; - skb_frag_t *frag; - int i; - - napi->skb = NULL; if (!skb) { skb = netdev_alloc_skb(dev, GRO_MAX_HEAD + NET_IP_ALIGN); @@ -2537,47 +2564,14 @@ struct sk_buff *napi_fraginfo_skb(struct napi_struct *napi, goto out; skb_reserve(skb, NET_IP_ALIGN); - } - - BUG_ON(info->nr_frags > MAX_SKB_FRAGS); - frag = &info->frags[info->nr_frags - 1]; - - for (i = skb_shinfo(skb)->nr_frags; i < info->nr_frags; i++) { - skb_fill_page_desc(skb, i, frag->page, frag->page_offset, - frag->size); - frag++; - } - skb_shinfo(skb)->nr_frags = info->nr_frags; - - skb->data_len = info->len; - skb->len += info->len; - skb->truesize += info->len; - skb_reset_mac_header(skb); - skb_gro_reset_offset(skb); - - eth = skb_gro_header(skb, sizeof(*eth)); - if (!eth) { - napi_reuse_skb(napi, skb); - skb = NULL; - goto out; + napi->skb = skb; } - skb_gro_pull(skb, sizeof(*eth)); - - /* - * This works because the only protocols we care about don't require - * special handling. We'll fix it up properly at the end. - */ - skb->protocol = eth->h_proto; - - skb->ip_summed = info->ip_summed; - skb->csum = info->csum; - out: return skb; } -EXPORT_SYMBOL(napi_fraginfo_skb); +EXPORT_SYMBOL(napi_get_frags); int napi_frags_finish(struct napi_struct *napi, struct sk_buff *skb, int ret) { @@ -2607,9 +2601,46 @@ int napi_frags_finish(struct napi_struct *napi, struct sk_buff *skb, int ret) } EXPORT_SYMBOL(napi_frags_finish); -int napi_gro_frags(struct napi_struct *napi, struct napi_gro_fraginfo *info) +struct sk_buff *napi_frags_skb(struct napi_struct *napi) { - struct sk_buff *skb = napi_fraginfo_skb(napi, info); + struct sk_buff *skb = napi->skb; + struct ethhdr *eth; + unsigned int hlen; + unsigned int off; + + napi->skb = NULL; + + skb_reset_mac_header(skb); + skb_gro_reset_offset(skb); + + off = skb_gro_offset(skb); + hlen = off + sizeof(*eth); + eth = skb_gro_header_fast(skb, off); + if (skb_gro_header_hard(skb, hlen)) { + eth = skb_gro_header_slow(skb, hlen, off); + if (unlikely(!eth)) { + napi_reuse_skb(napi, skb); + skb = NULL; + goto out; + } + } + + skb_gro_pull(skb, sizeof(*eth)); + + /* + * This works because the only protocols we care about don't require + * special handling. We'll fix it up properly at the end. + */ + skb->protocol = eth->h_proto; + +out: + return skb; +} +EXPORT_SYMBOL(napi_frags_skb); + +int napi_gro_frags(struct napi_struct *napi) +{ + struct sk_buff *skb = napi_frags_skb(napi); if (!skb) return NET_RX_DROP; @@ -2713,7 +2744,7 @@ void netif_napi_del(struct napi_struct *napi) struct sk_buff *skb, *next; list_del_init(&napi->dev_list); - kfree_skb(napi->skb); + napi_free_frags(napi); for (skb = napi->gro_list; skb; skb = next) { next = skb->next; @@ -2767,8 +2798,10 @@ static void net_rx_action(struct softirq_action *h) * accidently calling ->poll() when NAPI is not scheduled. */ work = 0; - if (test_bit(NAPI_STATE_SCHED, &n->state)) + if (test_bit(NAPI_STATE_SCHED, &n->state)) { work = n->poll(n, weight); + trace_napi_poll(n); + } WARN_ON_ONCE(work > weight); @@ -3438,6 +3471,252 @@ void dev_set_rx_mode(struct net_device *dev) netif_addr_unlock_bh(dev); } +/* hw addresses list handling functions */ + +static int __hw_addr_add(struct list_head *list, unsigned char *addr, + int addr_len, unsigned char addr_type) +{ + struct netdev_hw_addr *ha; + int alloc_size; + + if (addr_len > MAX_ADDR_LEN) + return -EINVAL; + + alloc_size = sizeof(*ha); + if (alloc_size < L1_CACHE_BYTES) + alloc_size = L1_CACHE_BYTES; + ha = kmalloc(alloc_size, GFP_ATOMIC); + if (!ha) + return -ENOMEM; + memcpy(ha->addr, addr, addr_len); + ha->type = addr_type; + list_add_tail_rcu(&ha->list, list); + return 0; +} + +static void ha_rcu_free(struct rcu_head *head) +{ + struct netdev_hw_addr *ha; + + ha = container_of(head, struct netdev_hw_addr, rcu_head); + kfree(ha); +} + +static int __hw_addr_del_ii(struct list_head *list, unsigned char *addr, + int addr_len, unsigned char addr_type, + int ignore_index) +{ + struct netdev_hw_addr *ha; + int i = 0; + + list_for_each_entry(ha, list, list) { + if (i++ != ignore_index && + !memcmp(ha->addr, addr, addr_len) && + (ha->type == addr_type || !addr_type)) { + list_del_rcu(&ha->list); + call_rcu(&ha->rcu_head, ha_rcu_free); + return 0; + } + } + return -ENOENT; +} + +static int __hw_addr_add_multiple_ii(struct list_head *to_list, + struct list_head *from_list, + int addr_len, unsigned char addr_type, + int ignore_index) +{ + int err; + struct netdev_hw_addr *ha, *ha2; + unsigned char type; + + list_for_each_entry(ha, from_list, list) { + type = addr_type ? addr_type : ha->type; + err = __hw_addr_add(to_list, ha->addr, addr_len, type); + if (err) + goto unroll; + } + return 0; + +unroll: + list_for_each_entry(ha2, from_list, list) { + if (ha2 == ha) + break; + type = addr_type ? addr_type : ha2->type; + __hw_addr_del_ii(to_list, ha2->addr, addr_len, type, + ignore_index); + } + return err; +} + +static void __hw_addr_del_multiple_ii(struct list_head *to_list, + struct list_head *from_list, + int addr_len, unsigned char addr_type, + int ignore_index) +{ + struct netdev_hw_addr *ha; + unsigned char type; + + list_for_each_entry(ha, from_list, list) { + type = addr_type ? addr_type : ha->type; + __hw_addr_del_ii(to_list, ha->addr, addr_len, addr_type, + ignore_index); + } +} + +static void __hw_addr_flush(struct list_head *list) +{ + struct netdev_hw_addr *ha, *tmp; + + list_for_each_entry_safe(ha, tmp, list, list) { + list_del_rcu(&ha->list); + call_rcu(&ha->rcu_head, ha_rcu_free); + } +} + +/* Device addresses handling functions */ + +static void dev_addr_flush(struct net_device *dev) +{ + /* rtnl_mutex must be held here */ + + __hw_addr_flush(&dev->dev_addr_list); + dev->dev_addr = NULL; +} + +static int dev_addr_init(struct net_device *dev) +{ + unsigned char addr[MAX_ADDR_LEN]; + struct netdev_hw_addr *ha; + int err; + + /* rtnl_mutex must be held here */ + + INIT_LIST_HEAD(&dev->dev_addr_list); + memset(addr, 0, sizeof(*addr)); + err = __hw_addr_add(&dev->dev_addr_list, addr, sizeof(*addr), + NETDEV_HW_ADDR_T_LAN); + if (!err) { + /* + * Get the first (previously created) address from the list + * and set dev_addr pointer to this location. + */ + ha = list_first_entry(&dev->dev_addr_list, + struct netdev_hw_addr, list); + dev->dev_addr = ha->addr; + } + return err; +} + +/** + * dev_addr_add - Add a device address + * @dev: device + * @addr: address to add + * @addr_type: address type + * + * Add a device address to the device or increase the reference count if + * it already exists. + * + * The caller must hold the rtnl_mutex. + */ +int dev_addr_add(struct net_device *dev, unsigned char *addr, + unsigned char addr_type) +{ + int err; + + ASSERT_RTNL(); + + err = __hw_addr_add(&dev->dev_addr_list, addr, dev->addr_len, + addr_type); + if (!err) + call_netdevice_notifiers(NETDEV_CHANGEADDR, dev); + return err; +} +EXPORT_SYMBOL(dev_addr_add); + +/** + * dev_addr_del - Release a device address. + * @dev: device + * @addr: address to delete + * @addr_type: address type + * + * Release reference to a device address and remove it from the device + * if the reference count drops to zero. + * + * The caller must hold the rtnl_mutex. + */ +int dev_addr_del(struct net_device *dev, unsigned char *addr, + unsigned char addr_type) +{ + int err; + + ASSERT_RTNL(); + + err = __hw_addr_del_ii(&dev->dev_addr_list, addr, dev->addr_len, + addr_type, 0); + if (!err) + call_netdevice_notifiers(NETDEV_CHANGEADDR, dev); + return err; +} +EXPORT_SYMBOL(dev_addr_del); + +/** + * dev_addr_add_multiple - Add device addresses from another device + * @to_dev: device to which addresses will be added + * @from_dev: device from which addresses will be added + * @addr_type: address type - 0 means type will be used from from_dev + * + * Add device addresses of the one device to another. + ** + * The caller must hold the rtnl_mutex. + */ +int dev_addr_add_multiple(struct net_device *to_dev, + struct net_device *from_dev, + unsigned char addr_type) +{ + int err; + + ASSERT_RTNL(); + + if (from_dev->addr_len != to_dev->addr_len) + return -EINVAL; + err = __hw_addr_add_multiple_ii(&to_dev->dev_addr_list, + &from_dev->dev_addr_list, + to_dev->addr_len, addr_type, 0); + if (!err) + call_netdevice_notifiers(NETDEV_CHANGEADDR, to_dev); + return err; +} +EXPORT_SYMBOL(dev_addr_add_multiple); + +/** + * dev_addr_del_multiple - Delete device addresses by another device + * @to_dev: device where the addresses will be deleted + * @from_dev: device by which addresses the addresses will be deleted + * @addr_type: address type - 0 means type will used from from_dev + * + * Deletes addresses in to device by the list of addresses in from device. + * + * The caller must hold the rtnl_mutex. + */ +int dev_addr_del_multiple(struct net_device *to_dev, + struct net_device *from_dev, + unsigned char addr_type) +{ + ASSERT_RTNL(); + + if (from_dev->addr_len != to_dev->addr_len) + return -EINVAL; + __hw_addr_del_multiple_ii(&to_dev->dev_addr_list, + &from_dev->dev_addr_list, + to_dev->addr_len, addr_type, 0); + call_netdevice_notifiers(NETDEV_CHANGEADDR, to_dev); + return 0; +} +EXPORT_SYMBOL(dev_addr_del_multiple); + +/* unicast and multicast addresses handling functions */ + int __dev_addr_delete(struct dev_addr_list **list, int *count, void *addr, int alen, int glbl) { @@ -4327,39 +4606,6 @@ unsigned long netdev_fix_features(unsigned long features, const char *name) } EXPORT_SYMBOL(netdev_fix_features); -/* Some devices need to (re-)set their netdev_ops inside - * ->init() or similar. If that happens, we have to setup - * the compat pointers again. - */ -void netdev_resync_ops(struct net_device *dev) -{ -#ifdef CONFIG_COMPAT_NET_DEV_OPS - const struct net_device_ops *ops = dev->netdev_ops; - - dev->init = ops->ndo_init; - dev->uninit = ops->ndo_uninit; - dev->open = ops->ndo_open; - dev->change_rx_flags = ops->ndo_change_rx_flags; - dev->set_rx_mode = ops->ndo_set_rx_mode; - dev->set_multicast_list = ops->ndo_set_multicast_list; - dev->set_mac_address = ops->ndo_set_mac_address; - dev->validate_addr = ops->ndo_validate_addr; - dev->do_ioctl = ops->ndo_do_ioctl; - dev->set_config = ops->ndo_set_config; - dev->change_mtu = ops->ndo_change_mtu; - dev->neigh_setup = ops->ndo_neigh_setup; - dev->tx_timeout = ops->ndo_tx_timeout; - dev->get_stats = ops->ndo_get_stats; - dev->vlan_rx_register = ops->ndo_vlan_rx_register; - dev->vlan_rx_add_vid = ops->ndo_vlan_rx_add_vid; - dev->vlan_rx_kill_vid = ops->ndo_vlan_rx_kill_vid; -#ifdef CONFIG_NET_POLL_CONTROLLER - dev->poll_controller = ops->ndo_poll_controller; -#endif -#endif -} -EXPORT_SYMBOL(netdev_resync_ops); - /** * register_netdevice - register a network device * @dev: device to register @@ -4399,23 +4645,6 @@ int register_netdevice(struct net_device *dev) dev->iflink = -1; -#ifdef CONFIG_COMPAT_NET_DEV_OPS - /* Netdevice_ops API compatiability support. - * This is temporary until all network devices are converted. - */ - if (dev->netdev_ops) { - netdev_resync_ops(dev); - } else { - char drivername[64]; - pr_info("%s (%s): not using net_device_ops yet\n", - dev->name, netdev_drivername(dev, drivername, 64)); - - /* This works only because net_device_ops and the - compatiablity structure are the same. */ - dev->netdev_ops = (void *) &(dev->init); - } -#endif - /* Init, if this function is available */ if (dev->netdev_ops->ndo_init) { ret = dev->netdev_ops->ndo_init(dev); @@ -4701,13 +4930,30 @@ void netdev_run_todo(void) * the internal statistics structure is used. */ const struct net_device_stats *dev_get_stats(struct net_device *dev) - { +{ const struct net_device_ops *ops = dev->netdev_ops; if (ops->ndo_get_stats) return ops->ndo_get_stats(dev); - else - return &dev->stats; + else { + unsigned long tx_bytes = 0, tx_packets = 0, tx_dropped = 0; + struct net_device_stats *stats = &dev->stats; + unsigned int i; + struct netdev_queue *txq; + + for (i = 0; i < dev->num_tx_queues; i++) { + txq = netdev_get_tx_queue(dev, i); + tx_bytes += txq->tx_bytes; + tx_packets += txq->tx_packets; + tx_dropped += txq->tx_dropped; + } + if (tx_bytes || tx_packets || tx_dropped) { + stats->tx_bytes = tx_bytes; + stats->tx_packets = tx_packets; + stats->tx_dropped = tx_dropped; + } + return stats; + } } EXPORT_SYMBOL(dev_get_stats); @@ -4765,13 +5011,16 @@ struct net_device *alloc_netdev_mq(int sizeof_priv, const char *name, if (!tx) { printk(KERN_ERR "alloc_netdev: Unable to allocate " "tx qdiscs.\n"); - kfree(p); - return NULL; + goto free_p; } dev = (struct net_device *) (((long)p + NETDEV_ALIGN_CONST) & ~NETDEV_ALIGN_CONST); dev->padded = (char *)dev - (char *)p; + + if (dev_addr_init(dev)) + goto free_tx; + dev_net_set(dev, &init_net); dev->_tx = tx; @@ -4783,9 +5032,17 @@ struct net_device *alloc_netdev_mq(int sizeof_priv, const char *name, netdev_init_queues(dev); INIT_LIST_HEAD(&dev->napi_list); + dev->priv_flags = IFF_XMIT_DST_RELEASE; setup(dev); strcpy(dev->name, name); return dev; + +free_tx: + kfree(tx); + +free_p: + kfree(p); + return NULL; } EXPORT_SYMBOL(alloc_netdev_mq); @@ -4805,6 +5062,9 @@ void free_netdev(struct net_device *dev) kfree(dev->_tx); + /* Flush device addresses */ + dev_addr_flush(dev); + list_for_each_entry_safe(p, n, &dev->napi_list, dev_list) netif_napi_del(p); |