From 27f39c73e63833b4c081a0d681d88b4184a0491d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 19 May 2010 22:07:23 +0000 Subject: net: Use __this_cpu_inc() in fast path This patch saves 224 bytes of text on my machine. __this_cpu_inc() generates a single instruction, using no scratch registers : 65 ff 04 25 a8 30 01 00 incl %gs:0x130a8 instead of : 48 c7 c2 80 30 01 00 mov $0x13080,%rdx 65 48 8b 04 25 88 ea 00 00 mov %gs:0xea88,%rax 83 44 10 28 01 addl $0x1,0x28(%rax,%rdx,1) Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/dev.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 1845b08c624e..7d76b056aa3d 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2822,8 +2822,7 @@ static int __netif_receive_skb(struct sk_buff *skb) skb->dev = master; } - __get_cpu_var(softnet_data).processed++; - + __this_cpu_inc(softnet_data.processed); skb_reset_network_header(skb); skb_reset_transport_header(skb); skb->mac_len = skb->network_header - skb->mac_header; -- cgit v1.2.3 From 15e83ed78864d0625e87a85f09b297c0919a4797 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 19 May 2010 23:16:03 +0000 Subject: net: remove zap_completion_queue netpoll does an interesting work in zap_completion_queue(), but this was before we did skb orphaning before delivering packets to device. It now makes sense to add a test in dev_kfree_skb_irq() to not queue a skb if already orphaned, and to remove netpoll zap_completion_queue() as a bonus. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/dev.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 7d76b056aa3d..983a3c1d65c4 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1577,7 +1577,9 @@ EXPORT_SYMBOL(__netif_schedule); void dev_kfree_skb_irq(struct sk_buff *skb) { - if (atomic_dec_and_test(&skb->users)) { + if (!skb->destructor) + dev_kfree_skb(skb); + else if (atomic_dec_and_test(&skb->users)) { struct softnet_data *sd; unsigned long flags; -- cgit v1.2.3 From bc135b23d01acf7ee926aaf75b0020c86d3869f9 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 2 Jun 2010 03:23:51 -0700 Subject: net: Define accessors to manipulate QDISC_STATE_RUNNING Define three helpers to manipulate QDISC_STATE_RUNNIG flag, that a second patch will move on another location. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/dev.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 983a3c1d65c4..2733226d90b2 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2047,7 +2047,7 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, kfree_skb(skb); rc = NET_XMIT_DROP; } else if ((q->flags & TCQ_F_CAN_BYPASS) && !qdisc_qlen(q) && - !test_and_set_bit(__QDISC_STATE_RUNNING, &q->state)) { + qdisc_run_begin(q)) { /* * This is a work-conserving queue; there are no old skbs * waiting to be sent out; and the qdisc is not running - @@ -2059,7 +2059,7 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, if (sch_direct_xmit(skb, q, dev, txq, root_lock)) __qdisc_run(q); else - clear_bit(__QDISC_STATE_RUNNING, &q->state); + qdisc_run_end(q); rc = NET_XMIT_SUCCESS; } else { -- cgit v1.2.3 From 79640a4ca6955e3ebdb7038508fa7a0cd7fa5527 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 2 Jun 2010 05:09:29 -0700 Subject: net: add additional lock to qdisc to increase throughput When many cpus compete for sending frames on a given qdisc, the qdisc spinlock suffers from very high contention. The cpu owning __QDISC_STATE_RUNNING bit has same priority to acquire the lock, and cannot dequeue packets fast enough, since it must wait for this lock for each dequeued packet. One solution to this problem is to force all cpus spinning on a second lock before trying to get the main lock, when/if they see __QDISC_STATE_RUNNING already set. The owning cpu then compete with at most one other cpu for the main lock, allowing for higher dequeueing rate. Based on a previous patch from Alexander Duyck. I added the heuristic to avoid the atomic in fast path, and put the new lock far away from the cache line used by the dequeue worker. Also try to release the busylock lock as late as possible. Tests with following script gave a boost from ~50.000 pps to ~600.000 pps on a dual quad core machine (E5450 @3.00GHz), tg3 driver. (A single netperf flow can reach ~800.000 pps on this platform) for j in `seq 0 3`; do for i in `seq 0 7`; do netperf -H 192.168.0.1 -t UDP_STREAM -l 60 -N -T $i -- -m 6 & done done Signed-off-by: Eric Dumazet Acked-by: Alexander Duyck Signed-off-by: David S. Miller --- net/core/dev.c | 29 +++++++++++++++++++++++++---- 1 file changed, 25 insertions(+), 4 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 2733226d90b2..ffca5c1066fa 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2040,8 +2040,18 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, struct netdev_queue *txq) { spinlock_t *root_lock = qdisc_lock(q); + bool contended = qdisc_is_running(q); int rc; + /* + * Heuristic to force contended enqueues to serialize on a + * separate lock before trying to get qdisc main lock. + * This permits __QDISC_STATE_RUNNING owner to get the lock more often + * and dequeue packets faster. + */ + if (unlikely(contended)) + spin_lock(&q->busylock); + spin_lock(root_lock); if (unlikely(test_bit(__QDISC_STATE_DEACTIVATED, &q->state))) { kfree_skb(skb); @@ -2056,19 +2066,30 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, if (!(dev->priv_flags & IFF_XMIT_DST_RELEASE)) skb_dst_force(skb); __qdisc_update_bstats(q, skb->len); - if (sch_direct_xmit(skb, q, dev, txq, root_lock)) + if (sch_direct_xmit(skb, q, dev, txq, root_lock)) { + if (unlikely(contended)) { + spin_unlock(&q->busylock); + contended = false; + } __qdisc_run(q); - else + } else qdisc_run_end(q); rc = NET_XMIT_SUCCESS; } else { skb_dst_force(skb); rc = qdisc_enqueue_root(skb, q); - qdisc_run(q); + if (qdisc_run_begin(q)) { + if (unlikely(contended)) { + spin_unlock(&q->busylock); + contended = false; + } + __qdisc_run(q); + } } spin_unlock(root_lock); - + if (unlikely(contended)) + spin_unlock(&q->busylock); return rc; } -- cgit v1.2.3 From ab95bfe01f9872459c8678572ccadbf646badad0 Mon Sep 17 00:00:00 2001 From: Jiri Pirko Date: Tue, 1 Jun 2010 21:52:08 +0000 Subject: net: replace hooks in __netif_receive_skb V5 What this patch does is it removes two receive frame hooks (for bridge and for macvlan) from __netif_receive_skb. These are replaced them with a single hook for both. It only supports one hook per device because it makes no sense to do bridging and macvlan on the same device. Then a network driver (of virtual netdev like macvlan or bridge) can register an rx_handler for needed net device. Signed-off-by: Jiri Pirko Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- net/core/dev.c | 119 ++++++++++++++++++++++++++------------------------------- 1 file changed, 55 insertions(+), 64 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index ffca5c1066fa..ec01a5998d70 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2604,70 +2604,14 @@ static inline int deliver_skb(struct sk_buff *skb, return pt_prev->func(skb, skb->dev, pt_prev, orig_dev); } -#if defined(CONFIG_BRIDGE) || defined (CONFIG_BRIDGE_MODULE) - -#if defined(CONFIG_ATM_LANE) || defined(CONFIG_ATM_LANE_MODULE) +#if (defined(CONFIG_BRIDGE) || defined(CONFIG_BRIDGE_MODULE)) && \ + (defined(CONFIG_ATM_LANE) || defined(CONFIG_ATM_LANE_MODULE)) /* This hook is defined here for ATM LANE */ int (*br_fdb_test_addr_hook)(struct net_device *dev, unsigned char *addr) __read_mostly; EXPORT_SYMBOL_GPL(br_fdb_test_addr_hook); #endif -/* - * If bridge module is loaded call bridging hook. - * returns NULL if packet was consumed. - */ -struct sk_buff *(*br_handle_frame_hook)(struct net_bridge_port *p, - struct sk_buff *skb) __read_mostly; -EXPORT_SYMBOL_GPL(br_handle_frame_hook); - -static inline struct sk_buff *handle_bridge(struct sk_buff *skb, - struct packet_type **pt_prev, int *ret, - struct net_device *orig_dev) -{ - struct net_bridge_port *port; - - if (skb->pkt_type == PACKET_LOOPBACK || - (port = rcu_dereference(skb->dev->br_port)) == NULL) - return skb; - - if (*pt_prev) { - *ret = deliver_skb(skb, *pt_prev, orig_dev); - *pt_prev = NULL; - } - - return br_handle_frame_hook(port, skb); -} -#else -#define handle_bridge(skb, pt_prev, ret, orig_dev) (skb) -#endif - -#if defined(CONFIG_MACVLAN) || defined(CONFIG_MACVLAN_MODULE) -struct sk_buff *(*macvlan_handle_frame_hook)(struct macvlan_port *p, - struct sk_buff *skb) __read_mostly; -EXPORT_SYMBOL_GPL(macvlan_handle_frame_hook); - -static inline struct sk_buff *handle_macvlan(struct sk_buff *skb, - struct packet_type **pt_prev, - int *ret, - struct net_device *orig_dev) -{ - struct macvlan_port *port; - - port = rcu_dereference(skb->dev->macvlan_port); - if (!port) - return skb; - - if (*pt_prev) { - *ret = deliver_skb(skb, *pt_prev, orig_dev); - *pt_prev = NULL; - } - return macvlan_handle_frame_hook(port, skb); -} -#else -#define handle_macvlan(skb, pt_prev, ret, orig_dev) (skb) -#endif - #ifdef CONFIG_NET_CLS_ACT /* TODO: Maybe we should just force sch_ingress to be compiled in * when CONFIG_NET_CLS_ACT is? otherwise some useless instructions @@ -2763,6 +2707,47 @@ void netif_nit_deliver(struct sk_buff *skb) rcu_read_unlock(); } +/** + * netdev_rx_handler_register - register receive handler + * @dev: device to register a handler for + * @rx_handler: receive handler to register + * + * Register a receive hander for a device. This handler will then be + * called from __netif_receive_skb. A negative errno code is returned + * on a failure. + * + * The caller must hold the rtnl_mutex. + */ +int netdev_rx_handler_register(struct net_device *dev, + rx_handler_func_t *rx_handler) +{ + ASSERT_RTNL(); + + if (dev->rx_handler) + return -EBUSY; + + rcu_assign_pointer(dev->rx_handler, rx_handler); + + return 0; +} +EXPORT_SYMBOL_GPL(netdev_rx_handler_register); + +/** + * netdev_rx_handler_unregister - unregister receive handler + * @dev: device to unregister a handler from + * + * Unregister a receive hander from a device. + * + * The caller must hold the rtnl_mutex. + */ +void netdev_rx_handler_unregister(struct net_device *dev) +{ + + ASSERT_RTNL(); + rcu_assign_pointer(dev->rx_handler, NULL); +} +EXPORT_SYMBOL_GPL(netdev_rx_handler_unregister); + static inline void skb_bond_set_mac_by_master(struct sk_buff *skb, struct net_device *master) { @@ -2815,6 +2800,7 @@ EXPORT_SYMBOL(__skb_bond_should_drop); static int __netif_receive_skb(struct sk_buff *skb) { struct packet_type *ptype, *pt_prev; + rx_handler_func_t *rx_handler; struct net_device *orig_dev; struct net_device *master; struct net_device *null_or_orig; @@ -2877,12 +2863,17 @@ static int __netif_receive_skb(struct sk_buff *skb) ncls: #endif - skb = handle_bridge(skb, &pt_prev, &ret, orig_dev); - if (!skb) - goto out; - skb = handle_macvlan(skb, &pt_prev, &ret, orig_dev); - if (!skb) - goto out; + /* Handle special case of bridge or macvlan */ + rx_handler = rcu_dereference(skb->dev->rx_handler); + if (rx_handler) { + if (pt_prev) { + ret = deliver_skb(skb, pt_prev, orig_dev); + pt_prev = NULL; + } + skb = rx_handler(skb); + if (!skb) + goto out; + } /* * Make sure frames received on VLAN interfaces stacked on -- cgit v1.2.3 From b78462ebc6a4ef9074aa80abebcdd470dc5f0ce0 Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Wed, 2 Jun 2010 12:24:37 +0000 Subject: skbuff: add check for non-linear to warn_if_lro and needs_linearize We can avoid an unecessary cache miss by checking if the skb is non-linear before accessing gso_size/gso_type in skb_warn_if_lro, the same can also be done to avoid a cache miss on nr_frags if data_len is 0. Signed-off-by: Alexander Duyck Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- net/core/dev.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index ec01a5998d70..3abb3a6058be 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2103,9 +2103,10 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, static inline int skb_needs_linearize(struct sk_buff *skb, struct net_device *dev) { - return (skb_has_frags(skb) && !(dev->features & NETIF_F_FRAGLIST)) || - (skb_shinfo(skb)->nr_frags && (!(dev->features & NETIF_F_SG) || - illegal_highdma(dev, skb))); + return skb_is_nonlinear(skb) && + ((skb_has_frags(skb) && !(dev->features & NETIF_F_FRAGLIST)) || + (skb_shinfo(skb)->nr_frags && (!(dev->features & NETIF_F_SG) || + illegal_highdma(dev, skb)))); } /** -- cgit v1.2.3 From 271c1dfa61bc90a57648ff96f3eb92d4b4d4f11e Mon Sep 17 00:00:00 2001 From: jamal Date: Fri, 4 Jun 2010 02:06:22 +0000 Subject: net: Remove unnecessary net action assertion The extra assertion to allow packet munging only when there are no other ptypes listening which may have worked around an old bug is unnecessary. It is sufficient to check if the skb is cloned before trampling on it. Thanks to Herbert Xu for being persistent and patient in getting this across. [Note that cloning checks and assertions are the general rule used by tc actions (documentation/networking/tc-actions-env-rules.txt)]. Signed-off-by: Jamal Hadi Salim Signed-off-by: David S. Miller --- net/core/dev.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index b65347c2cf2a..c8d127718ff1 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2663,9 +2663,6 @@ static inline struct sk_buff *handle_ing(struct sk_buff *skb, if (*pt_prev) { *ret = deliver_skb(skb, *pt_prev, orig_dev); *pt_prev = NULL; - } else { - /* Huh? Why does turning on AF_PACKET affect this? */ - skb->tc_verd = SET_TC_OK2MUNGE(skb->tc_verd); } switch (ing_filter(skb)) { -- cgit v1.2.3 From bb69ae049fcc986fcd742eb90ca0d44a7a49c9f1 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 7 Jun 2010 11:42:13 +0000 Subject: anycast: Some RCU conversions - dev_get_by_flags() changed to dev_get_by_flags_rcu() - ipv6_sock_ac_join() dont touch dev & idev refcounts - ipv6_sock_ac_drop() dont touch dev & idev refcounts - ipv6_sock_ac_close() dont touch dev & idev refcounts - ipv6_dev_ac_dec() dount touch idev refcount - ipv6_chk_acast_addr() dont touch idev refcount Signed-off-by: Eric Dumazet CC: Hideaki YOSHIFUJI Signed-off-by: David S. Miller --- net/core/dev.c | 14 +++++--------- 1 file changed, 5 insertions(+), 9 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index c8d127718ff1..6f330cee79a6 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -803,35 +803,31 @@ struct net_device *dev_getfirstbyhwtype(struct net *net, unsigned short type) EXPORT_SYMBOL(dev_getfirstbyhwtype); /** - * dev_get_by_flags - find any device with given flags + * dev_get_by_flags_rcu - find any device with given flags * @net: the applicable net namespace * @if_flags: IFF_* values * @mask: bitmask of bits in if_flags to check * * Search for any interface with the given flags. Returns NULL if a device - * is not found or a pointer to the device. The device returned has - * had a reference added and the pointer is safe until the user calls - * dev_put to indicate they have finished with it. + * is not found or a pointer to the device. Must be called inside + * rcu_read_lock(), and result refcount is unchanged. */ -struct net_device *dev_get_by_flags(struct net *net, unsigned short if_flags, +struct net_device *dev_get_by_flags_rcu(struct net *net, unsigned short if_flags, unsigned short mask) { struct net_device *dev, *ret; ret = NULL; - rcu_read_lock(); for_each_netdev_rcu(net, dev) { if (((dev->flags ^ if_flags) & mask) == 0) { - dev_hold(dev); ret = dev; break; } } - rcu_read_unlock(); return ret; } -EXPORT_SYMBOL(dev_get_by_flags); +EXPORT_SYMBOL(dev_get_by_flags_rcu); /** * dev_valid_name - check if name is okay for network device -- cgit v1.2.3 From be1f3c2c027cc5ad735df6a45a542ed1db7ec48b Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Tue, 8 Jun 2010 07:19:54 +0000 Subject: net: Enable 64-bit net device statistics on 32-bit architectures Use struct rtnl_link_stats64 as the statistics structure. On 32-bit architectures, insert 32 bits of padding after/before each field of struct net_device_stats to make its layout compatible with struct rtnl_link_stats64. Add an anonymous union in net_device; move stats into the union and add struct rtnl_link_stats64 stats64. Add net_device_ops::ndo_get_stats64, implementations of which will return a pointer to struct rtnl_link_stats64. Drivers that implement this operation must not update the structure asynchronously. Change dev_get_stats() to call ndo_get_stats64 if available, and to return a pointer to struct rtnl_link_stats64. Change callers of dev_get_stats() accordingly. Signed-off-by: Ben Hutchings Signed-off-by: David S. Miller --- net/core/dev.c | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 277844901ce3..a1abc10db08a 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -3701,10 +3701,10 @@ void dev_seq_stop(struct seq_file *seq, void *v) static void dev_seq_printf_stats(struct seq_file *seq, struct net_device *dev) { - const struct net_device_stats *stats = dev_get_stats(dev); + const struct rtnl_link_stats64 *stats = dev_get_stats(dev); - seq_printf(seq, "%6s: %7lu %7lu %4lu %4lu %4lu %5lu %10lu %9lu " - "%8lu %7lu %4lu %4lu %4lu %5lu %7lu %10lu\n", + seq_printf(seq, "%6s: %7llu %7llu %4llu %4llu %4llu %5llu %10llu %9llu " + "%8llu %7llu %4llu %4llu %4llu %5llu %7llu %10llu\n", dev->name, stats->rx_bytes, stats->rx_packets, stats->rx_errors, stats->rx_dropped + stats->rx_missed_errors, @@ -5281,18 +5281,21 @@ EXPORT_SYMBOL(dev_txq_stats_fold); * @dev: device to get statistics from * * Get network statistics from device. The device driver may provide - * its own method by setting dev->netdev_ops->get_stats; otherwise - * the internal statistics structure is used. + * its own method by setting dev->netdev_ops->get_stats64 or + * dev->netdev_ops->get_stats; otherwise the internal statistics + * structure is used. */ -const struct net_device_stats *dev_get_stats(struct net_device *dev) +const struct rtnl_link_stats64 *dev_get_stats(struct net_device *dev) { const struct net_device_ops *ops = dev->netdev_ops; + if (ops->ndo_get_stats64) + return ops->ndo_get_stats64(dev); if (ops->ndo_get_stats) - return ops->ndo_get_stats(dev); + return (struct rtnl_link_stats64 *)ops->ndo_get_stats(dev); dev_txq_stats_fold(dev, &dev->stats); - return &dev->stats; + return &dev->stats64; } EXPORT_SYMBOL(dev_get_stats); -- cgit v1.2.3 From 93e2c32b5cb2ad92ceb1d7a4684f20a0d25bf530 Mon Sep 17 00:00:00 2001 From: Jiri Pirko Date: Thu, 10 Jun 2010 03:34:59 +0000 Subject: net: add rx_handler data pointer Add possibility to register rx_handler data pointer along with a rx_handler. Signed-off-by: Jiri Pirko Signed-off-by: David S. Miller --- net/core/dev.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index a1abc10db08a..abdb19e547a7 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2703,6 +2703,7 @@ void netif_nit_deliver(struct sk_buff *skb) * netdev_rx_handler_register - register receive handler * @dev: device to register a handler for * @rx_handler: receive handler to register + * @rx_handler_data: data pointer that is used by rx handler * * Register a receive hander for a device. This handler will then be * called from __netif_receive_skb. A negative errno code is returned @@ -2711,13 +2712,15 @@ void netif_nit_deliver(struct sk_buff *skb) * The caller must hold the rtnl_mutex. */ int netdev_rx_handler_register(struct net_device *dev, - rx_handler_func_t *rx_handler) + rx_handler_func_t *rx_handler, + void *rx_handler_data) { ASSERT_RTNL(); if (dev->rx_handler) return -EBUSY; + rcu_assign_pointer(dev->rx_handler_data, rx_handler_data); rcu_assign_pointer(dev->rx_handler, rx_handler); return 0; @@ -2737,6 +2740,7 @@ void netdev_rx_handler_unregister(struct net_device *dev) ASSERT_RTNL(); rcu_assign_pointer(dev->rx_handler, NULL); + rcu_assign_pointer(dev->rx_handler_data, NULL); } EXPORT_SYMBOL_GPL(netdev_rx_handler_unregister); -- cgit v1.2.3 From f350a0a87374418635689471606454abc7beaa3a Mon Sep 17 00:00:00 2001 From: Jiri Pirko Date: Tue, 15 Jun 2010 06:50:45 +0000 Subject: bridge: use rx_handler_data pointer to store net_bridge_port pointer Register net_bridge_port pointer as rx_handler data pointer. As br_port is removed from struct net_device, another netdev priv_flag is added to indicate the device serves as a bridge port. Also rcuized pointers are now correctly dereferenced in br_fdb.c and in netfilter parts. Signed-off-by: Jiri Pirko Signed-off-by: David S. Miller --- net/core/dev.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index abdb19e547a7..5902426ef585 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2765,7 +2765,8 @@ int __skb_bond_should_drop(struct sk_buff *skb, struct net_device *master) if (master->priv_flags & IFF_MASTER_ARPMON) dev->last_rx = jiffies; - if ((master->priv_flags & IFF_MASTER_ALB) && master->br_port) { + if ((master->priv_flags & IFF_MASTER_ALB) && + (master->priv_flags & IFF_BRIDGE_PORT)) { /* Do address unmangle. The local destination address * will be always the one master has. Provides the right * functionality in a bridge. -- cgit v1.2.3 From 6afff0caa721211e8c04bdc7627ee3bff95bcb95 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Wed, 16 Jun 2010 14:18:12 +0000 Subject: net: consolidate netif_needs_gso() checks netif_needs_gso() is checked twice in the TX path once, before submitting the skb to the qdisc and once after it is dequeued from the qdisc just before calling ndo_hard_start(). This opens a window for a user to change the gso/tso or tx checksum settings that can cause netif_needs_gso to be true in one check and false in the other. Specifically, changing TX checksum setting may cause the warning in skb_gso_segment() to be triggered if the checksum is calculated earlier. This consolidates the netif_needs_gso() calls so that the stack only checks if gso is needed in dev_hard_start_xmit(). Signed-off-by: John Fastabend Cc: Herbert Xu Signed-off-by: Jeff Kirsher Acked-by: Herbert Xu Signed-off-by: David S. Miller --- net/core/dev.c | 68 +++++++++++++++++++++++++++------------------------------- 1 file changed, 32 insertions(+), 36 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 5902426ef585..7f390b52caab 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1895,6 +1895,22 @@ static inline void skb_orphan_try(struct sk_buff *skb) skb_orphan(skb); } +/* + * Returns true if either: + * 1. skb has frag_list and the device doesn't support FRAGLIST, or + * 2. skb is fragmented and the device does not support SG, or if + * at least one of fragments is in highmem and device does not + * support DMA from it. + */ +static inline int skb_needs_linearize(struct sk_buff *skb, + struct net_device *dev) +{ + return skb_is_nonlinear(skb) && + ((skb_has_frags(skb) && !(dev->features & NETIF_F_FRAGLIST)) || + (skb_shinfo(skb)->nr_frags && (!(dev->features & NETIF_F_SG) || + illegal_highdma(dev, skb)))); +} + int dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev, struct netdev_queue *txq) { @@ -1919,6 +1935,22 @@ int dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev, goto out_kfree_skb; if (skb->next) goto gso; + } else { + if (skb_needs_linearize(skb, dev) && + __skb_linearize(skb)) + goto out_kfree_skb; + + /* If packet is not checksummed and device does not + * support checksumming for this protocol, complete + * checksumming here. + */ + if (skb->ip_summed == CHECKSUM_PARTIAL) { + skb_set_transport_header(skb, skb->csum_start - + skb_headroom(skb)); + if (!dev_can_checksum(dev, skb) && + skb_checksum_help(skb)) + goto out_kfree_skb; + } } rc = ops->ndo_start_xmit(skb, dev); @@ -2089,22 +2121,6 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, return rc; } -/* - * Returns true if either: - * 1. skb has frag_list and the device doesn't support FRAGLIST, or - * 2. skb is fragmented and the device does not support SG, or if - * at least one of fragments is in highmem and device does not - * support DMA from it. - */ -static inline int skb_needs_linearize(struct sk_buff *skb, - struct net_device *dev) -{ - return skb_is_nonlinear(skb) && - ((skb_has_frags(skb) && !(dev->features & NETIF_F_FRAGLIST)) || - (skb_shinfo(skb)->nr_frags && (!(dev->features & NETIF_F_SG) || - illegal_highdma(dev, skb)))); -} - /** * dev_queue_xmit - transmit a buffer * @skb: buffer to transmit @@ -2137,25 +2153,6 @@ int dev_queue_xmit(struct sk_buff *skb) struct Qdisc *q; int rc = -ENOMEM; - /* GSO will handle the following emulations directly. */ - if (netif_needs_gso(dev, skb)) - goto gso; - - /* Convert a paged skb to linear, if required */ - if (skb_needs_linearize(skb, dev) && __skb_linearize(skb)) - goto out_kfree_skb; - - /* If packet is not checksummed and device does not support - * checksumming for this protocol, complete checksumming here. - */ - if (skb->ip_summed == CHECKSUM_PARTIAL) { - skb_set_transport_header(skb, skb->csum_start - - skb_headroom(skb)); - if (!dev_can_checksum(dev, skb) && skb_checksum_help(skb)) - goto out_kfree_skb; - } - -gso: /* Disable soft irqs for various locks below. Also * stops preemption for RCU. */ @@ -2214,7 +2211,6 @@ gso: rc = -ENETDOWN; rcu_read_unlock_bh(); -out_kfree_skb: kfree_skb(skb); return rc; out: -- cgit v1.2.3 From 70777d03466e7a8a41b0d34677454c92f4e93d89 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Wed, 30 Jun 2010 10:39:19 -0700 Subject: net/core: use ntohs for skb->protocol This is only noticed by people that are not doing everything correct in the first place. Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: David S. Miller --- net/core/dev.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 7f390b52caab..e85cc5fa3c4e 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1537,7 +1537,8 @@ static void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev) if (net_ratelimit()) printk(KERN_CRIT "protocol %04x is " "buggy, dev %s\n", - skb2->protocol, dev->name); + ntohs(skb2->protocol), + dev->name); skb_reset_network_header(skb2); } -- cgit v1.2.3 From 256df2f3879efdb2e9808bdb1b54b16fbb11fa38 Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Sun, 27 Jun 2010 01:02:35 +0000 Subject: netdevice.h net/core/dev.c: Convert netdev_ logging macros to functions Reduces an x86 defconfig text and data ~2k. text is smaller, data is larger. $ size vmlinux* text data bss dec hex filename 7198862 720112 1366288 9285262 8dae8e vmlinux 7205273 716016 1366288 9287577 8db799 vmlinux.device_h Uses %pV and struct va_format Format arguments are verified before printk Signed-off-by: Joe Perches Acked-by: Greg Kroah-Hartman Signed-off-by: David S. Miller --- net/core/dev.c | 62 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 62 insertions(+) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index e85cc5fa3c4e..93b8929fa21d 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5802,6 +5802,68 @@ char *netdev_drivername(const struct net_device *dev, char *buffer, int len) return buffer; } +static int __netdev_printk(const char *level, const struct net_device *dev, + struct va_format *vaf) +{ + int r; + + if (dev && dev->dev.parent) + r = dev_printk(level, dev->dev.parent, "%s: %pV", + netdev_name(dev), vaf); + else if (dev) + r = printk("%s%s: %pV", level, netdev_name(dev), vaf); + else + r = printk("%s(NULL net_device): %pV", level, vaf); + + return r; +} + +int netdev_printk(const char *level, const struct net_device *dev, + const char *format, ...) +{ + struct va_format vaf; + va_list args; + int r; + + va_start(args, format); + + vaf.fmt = format; + vaf.va = &args; + + r = __netdev_printk(level, dev, &vaf); + va_end(args); + + return r; +} +EXPORT_SYMBOL(netdev_printk); + +#define define_netdev_printk_level(func, level) \ +int func(const struct net_device *dev, const char *fmt, ...) \ +{ \ + int r; \ + struct va_format vaf; \ + va_list args; \ + \ + va_start(args, fmt); \ + \ + vaf.fmt = fmt; \ + vaf.va = &args; \ + \ + r = __netdev_printk(level, dev, &vaf); \ + va_end(args); \ + \ + return r; \ +} \ +EXPORT_SYMBOL(func); + +define_netdev_printk_level(netdev_emerg, KERN_EMERG); +define_netdev_printk_level(netdev_alert, KERN_ALERT); +define_netdev_printk_level(netdev_crit, KERN_CRIT); +define_netdev_printk_level(netdev_err, KERN_ERR); +define_netdev_printk_level(netdev_warn, KERN_WARNING); +define_netdev_printk_level(netdev_notice, KERN_NOTICE); +define_netdev_printk_level(netdev_info, KERN_INFO); + static void __net_exit netdev_exit(struct net *net) { kfree(net->dev_name_head); -- cgit v1.2.3 From 28172739f0a276eb8d6ca917b3974c2edb036da3 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 7 Jul 2010 14:58:56 -0700 Subject: net: fix 64 bit counters on 32 bit arches There is a small possibility that a reader gets incorrect values on 32 bit arches. SNMP applications could catch incorrect counters when a 32bit high part is changed by another stats consumer/provider. One way to solve this is to add a rtnl_link_stats64 param to all ndo_get_stats64() methods, and also add such a parameter to dev_get_stats(). Rule is that we are not allowed to use dev->stats64 as a temporary storage for 64bit stats, but a caller provided area (usually on stack) Old drivers (only providing get_stats() method) need no changes. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/dev.c | 25 ++++++++++++++++--------- 1 file changed, 16 insertions(+), 9 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 93b8929fa21d..92482d7a87a9 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -3703,7 +3703,8 @@ void dev_seq_stop(struct seq_file *seq, void *v) static void dev_seq_printf_stats(struct seq_file *seq, struct net_device *dev) { - const struct rtnl_link_stats64 *stats = dev_get_stats(dev); + struct rtnl_link_stats64 temp; + const struct rtnl_link_stats64 *stats = dev_get_stats(dev, &temp); seq_printf(seq, "%6s: %7llu %7llu %4llu %4llu %4llu %5llu %10llu %9llu " "%8llu %7llu %4llu %4llu %4llu %5llu %7llu %10llu\n", @@ -5281,23 +5282,29 @@ EXPORT_SYMBOL(dev_txq_stats_fold); /** * dev_get_stats - get network device statistics * @dev: device to get statistics from + * @storage: place to store stats * * Get network statistics from device. The device driver may provide * its own method by setting dev->netdev_ops->get_stats64 or * dev->netdev_ops->get_stats; otherwise the internal statistics * structure is used. */ -const struct rtnl_link_stats64 *dev_get_stats(struct net_device *dev) +const struct rtnl_link_stats64 *dev_get_stats(struct net_device *dev, + struct rtnl_link_stats64 *storage) { const struct net_device_ops *ops = dev->netdev_ops; - if (ops->ndo_get_stats64) - return ops->ndo_get_stats64(dev); - if (ops->ndo_get_stats) - return (struct rtnl_link_stats64 *)ops->ndo_get_stats(dev); - - dev_txq_stats_fold(dev, &dev->stats); - return &dev->stats64; + if (ops->ndo_get_stats64) { + memset(storage, 0, sizeof(*storage)); + return ops->ndo_get_stats64(dev, storage); + } + if (ops->ndo_get_stats) { + memcpy(storage, ops->ndo_get_stats(dev), sizeof(*storage)); + return storage; + } + memcpy(storage, &dev->stats, sizeof(*storage)); + dev_txq_stats_fold(dev, (struct net_device_stats *)storage); + return storage; } EXPORT_SYMBOL(dev_get_stats); -- cgit v1.2.3 From 3cfde79c6c7c8002375c4a8e5be7f602fbb9675d Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Fri, 9 Jul 2010 09:11:52 +0000 Subject: net: Get rid of rtnl_link_stats64 / net_device_stats union In commit be1f3c2c027cc5ad735df6a45a542ed1db7ec48b "net: Enable 64-bit net device statistics on 32-bit architectures" I redefined struct net_device_stats so that it could be used in a union with struct rtnl_link_stats64, avoiding the need for explicit copying or conversion between the two. However, this is unsafe because there is no locking required and no lock consistently held around calls to dev_get_stats() and use of the statistics structure it returns. In commit 28172739f0a276eb8d6ca917b3974c2edb036da3 "net: fix 64 bit counters on 32 bit arches" Eric Dumazet dealt with that problem by requiring callers of dev_get_stats() to provide storage for the result. This means that the net_device::stats64 field and the padding in struct net_device_stats are now redundant, so remove them. Update the comment on net_device_ops::ndo_get_stats64 to reflect its new usage. Change dev_txq_stats_fold() to use struct rtnl_link_stats64, since that is what all its callers are really using and it is no longer going to be compatible with struct net_device_stats. Eric Dumazet suggested the separate function for the structure conversion. Signed-off-by: Ben Hutchings Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/dev.c | 31 ++++++++++++++++++++++++++----- 1 file changed, 26 insertions(+), 5 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index eb4201cf9c8c..79ee26ef5095 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5274,10 +5274,10 @@ void netdev_run_todo(void) /** * dev_txq_stats_fold - fold tx_queues stats * @dev: device to get statistics from - * @stats: struct net_device_stats to hold results + * @stats: struct rtnl_link_stats64 to hold results */ void dev_txq_stats_fold(const struct net_device *dev, - struct net_device_stats *stats) + struct rtnl_link_stats64 *stats) { unsigned long tx_bytes = 0, tx_packets = 0, tx_dropped = 0; unsigned int i; @@ -5297,6 +5297,27 @@ void dev_txq_stats_fold(const struct net_device *dev, } EXPORT_SYMBOL(dev_txq_stats_fold); +/* Convert net_device_stats to rtnl_link_stats64. They have the same + * fields in the same order, with only the type differing. + */ +static void netdev_stats_to_stats64(struct rtnl_link_stats64 *stats64, + const struct net_device_stats *netdev_stats) +{ +#if BITS_PER_LONG == 64 + BUILD_BUG_ON(sizeof(*stats64) != sizeof(*netdev_stats)); + memcpy(stats64, netdev_stats, sizeof(*stats64)); +#else + size_t i, n = sizeof(*stats64) / sizeof(u64); + const unsigned long *src = (const unsigned long *)netdev_stats; + u64 *dst = (u64 *)stats64; + + BUILD_BUG_ON(sizeof(*netdev_stats) / sizeof(unsigned long) != + sizeof(*stats64) / sizeof(u64)); + for (i = 0; i < n; i++) + dst[i] = src[i]; +#endif +} + /** * dev_get_stats - get network device statistics * @dev: device to get statistics from @@ -5317,11 +5338,11 @@ const struct rtnl_link_stats64 *dev_get_stats(struct net_device *dev, return ops->ndo_get_stats64(dev, storage); } if (ops->ndo_get_stats) { - memcpy(storage, ops->ndo_get_stats(dev), sizeof(*storage)); + netdev_stats_to_stats64(storage, ops->ndo_get_stats(dev)); return storage; } - memcpy(storage, &dev->stats, sizeof(*storage)); - dev_txq_stats_fold(dev, (struct net_device_stats *)storage); + netdev_stats_to_stats64(storage, &dev->stats); + dev_txq_stats_fold(dev, storage); return storage; } EXPORT_SYMBOL(dev_get_stats); -- cgit v1.2.3 From d77535162e736c47978d5c01469c56e1781dc91b Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Fri, 9 Jul 2010 09:12:41 +0000 Subject: net: Document that dev_get_stats() returns the given pointer Document that dev_get_stats() returns the same stats pointer it was given. Remove const qualification from the returned pointer since the caller may do what it likes with that structure. Signed-off-by: Ben Hutchings Signed-off-by: David S. Miller --- net/core/dev.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 79ee26ef5095..e2b9fa2c917e 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5323,13 +5323,13 @@ static void netdev_stats_to_stats64(struct rtnl_link_stats64 *stats64, * @dev: device to get statistics from * @storage: place to store stats * - * Get network statistics from device. The device driver may provide - * its own method by setting dev->netdev_ops->get_stats64 or - * dev->netdev_ops->get_stats; otherwise the internal statistics - * structure is used. + * Get network statistics from device. Return @storage. + * The device driver may provide its own method by setting + * dev->netdev_ops->get_stats64 or dev->netdev_ops->get_stats; + * otherwise the internal statistics structure is used. */ -const struct rtnl_link_stats64 *dev_get_stats(struct net_device *dev, - struct rtnl_link_stats64 *storage) +struct rtnl_link_stats64 *dev_get_stats(struct net_device *dev, + struct rtnl_link_stats64 *storage) { const struct net_device_ops *ops = dev->netdev_ops; -- cgit v1.2.3 From c1f19b51d1d87f3e3bb7e6648f43f7d57ed2da6b Mon Sep 17 00:00:00 2001 From: Richard Cochran Date: Sat, 17 Jul 2010 08:49:36 +0000 Subject: net: support time stamping in phy devices. This patch adds a new networking option to allow hardware time stamps from PHY devices. When enabled, likely candidates among incoming and outgoing network packets are offered to the PHY driver for possible time stamping. When accepted by the PHY driver, incoming packets are deferred for later delivery by the driver. The patch also adds phylib driver methods for the SIOCSHWTSTAMP ioctl and callbacks for transmit and receive time stamping. Drivers may optionally implement these functions. Signed-off-by: Richard Cochran Signed-off-by: David S. Miller --- net/core/dev.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index e2b9fa2c917e..1c002c7ef5d5 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2957,6 +2957,9 @@ int netif_receive_skb(struct sk_buff *skb) if (netdev_tstamp_prequeue) net_timestamp_check(skb); + if (skb_defer_rx_timestamp(skb)) + return NET_RX_SUCCESS; + #ifdef CONFIG_RPS { struct rps_dev_flow voidflow, *rflow = &voidflow; -- cgit v1.2.3 From bd27290a593f80cb99e95287cb29c72c0d57608b Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 19 Jul 2010 09:35:40 -0700 Subject: net: 64bit stats for netdev_queue Since struct netdev_queue tx_bytes/tx_packets/tx_dropped are already protected by _xmit_lock, its easy to convert these fields to u64 instead of unsigned long. This completes 64bit stats for devices using them (vlan, macvlan, ...) Strictly, we could avoid the locking in dev_txq_stats_fold() on 64bit arches, but its slow path and we prefer keep it simple. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/dev.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 1c002c7ef5d5..9de75cdade56 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5282,15 +5282,17 @@ void netdev_run_todo(void) void dev_txq_stats_fold(const struct net_device *dev, struct rtnl_link_stats64 *stats) { - unsigned long tx_bytes = 0, tx_packets = 0, tx_dropped = 0; + u64 tx_bytes = 0, tx_packets = 0, tx_dropped = 0; unsigned int i; struct netdev_queue *txq; for (i = 0; i < dev->num_tx_queues; i++) { txq = netdev_get_tx_queue(dev, i); + spin_lock_bh(&txq->_xmit_lock); tx_bytes += txq->tx_bytes; tx_packets += txq->tx_packets; tx_dropped += txq->tx_dropped; + spin_unlock_bh(&txq->_xmit_lock); } if (tx_bytes || tx_packets || tx_dropped) { stats->tx_bytes = tx_bytes; -- cgit v1.2.3 From de38483010bae523f533bb6bf9f7b7353772f6eb Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Sun, 1 Aug 2010 00:33:23 -0700 Subject: net: ingress filter message limit If user misconfigures ingress and causes a redirection loop, don't overwhelm the log. This is also a error case so make it unlikely. Found by inspection, luckily not in real system. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- net/core/dev.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index b74fcd3e9365..5d1282df2fe3 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2646,10 +2646,10 @@ static int ing_filter(struct sk_buff *skb) int result = TC_ACT_OK; struct Qdisc *q; - if (MAX_RED_LOOP < ttl++) { - printk(KERN_WARNING - "Redir loop detected Dropping packet (%d->%d)\n", - skb->skb_iif, dev->ifindex); + if (unlikely(MAX_RED_LOOP < ttl++)) { + if (net_ratelimit()) + pr_warning( "Redir loop detected Dropping packet (%d->%d)\n", + skb->skb_iif, dev->ifindex); return TC_ACT_SHOT; } -- cgit v1.2.3 From a427615e0420f179eab801b929111abaadea2ed3 Mon Sep 17 00:00:00 2001 From: Changli Gao Date: Mon, 2 Aug 2010 22:45:49 -0700 Subject: net: cleanup inclusion Commit ab95bfe01f9872459c8678572ccadbf646badad0 replaces bridge and macvlan hooks in __netif_receive_skb(), so dev.c doesn't need to include their headers. Signed-off-by: Changli Gao Signed-off-by: David S. Miller --- net/core/dev.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 5d1282df2fe3..8c663dbf1d77 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -101,8 +101,6 @@ #include #include #include -#include -#include #include #include #include -- cgit v1.2.3 From 3578b0c8abc7bdb4f02152ce5db7e09d484c6866 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 3 Aug 2010 00:24:04 -0700 Subject: Revert "net: remove zap_completion_queue" This reverts commit 15e83ed78864d0625e87a85f09b297c0919a4797. As explained by Johannes Berg, the optimization made here is invalid. Or, at best, incomplete. Not only destructor invocation, but conntract entry releasing must be executed outside of hw IRQ context. So just checking "skb->destructor" is insufficient. Signed-off-by: David S. Miller --- net/core/dev.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 8c663dbf1d77..e1c1cdcc2bb0 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1591,9 +1591,7 @@ EXPORT_SYMBOL(__netif_schedule); void dev_kfree_skb_irq(struct sk_buff *skb) { - if (!skb->destructor) - dev_kfree_skb(skb); - else if (atomic_dec_and_test(&skb->users)) { + if (atomic_dec_and_test(&skb->users)) { struct softnet_data *sd; unsigned long flags; -- cgit v1.2.3