summaryrefslogtreecommitdiffstats
path: root/net/core/dev.c
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2012-10-10 11:12:54 +0900
committerLinus Torvalds <torvalds@linux-foundation.org>2012-10-10 11:12:54 +0900
commitaac2b1f5747ea34696d0da5bdc4d8247aa6437af (patch)
tree8fc8499aad6a28b044c9bdab3f920f64a98460c1 /net/core/dev.c
parent23d5385f382a7c7d8b6bf19b0c2cfb3acbb12d31 (diff)
parent5175a5e76bbdf20a614fb47ce7a38f0f39e70226 (diff)
Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
Pull networking updates from David Miller: 1) UAPI changes for networking from David Howells 2) A netlink dump is an operation we can sleep within, and therefore we need to make sure the dump provider module doesn't disappear on us meanwhile. Fix from Gao Feng. 3) Now that tunnels support GRO, we have to be more careful in skb_gro_reset_offset() otherwise we OOPS, from Eric Dumazet. 4) We can end up processing packets for VLANs we aren't actually configured to be on, fix from Florian Zumbiehl. 5) Fix routing cache removal regression in redirects and IPVS. The core issue on the IPVS side is that it wants to rewrite who the nexthop is and we have to explicitly accomodate that case. From Julian Anastasov. 6) Error code return fixes all over the networking drivers from Peter Senna Tschudin. 7) Fix routing cache removal regressions in IPSEC, from Steffen Klassert. 8) Fix deadlock in RDS during pings, from Jeff Liu. 9) Neighbour packet queue can trigger skb_under_panic() because we do not reset the network header of the SKB in the right spot. From Ramesh Nagappa. * git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (61 commits) RDS: fix rds-ping spinlock recursion netdev/phy: Prototype of_mdio_find_bus() farsync: fix support for over 30 cards be2net: Remove code that stops further access to BE NIC based on UE bits pch_gbe: Fix build error by selecting all the possible dependencies. e1000e: add device IDs for i218 ixgbe/ixgbevf: Limit maximum jumbo frame size to 9.5K to avoid Tx hangs ixgbevf: Set the netdev number of Tx queues UAPI: (Scripted) Disintegrate include/linux/tc_ematch UAPI: (Scripted) Disintegrate include/linux/tc_act UAPI: (Scripted) Disintegrate include/linux/netfilter_ipv6 UAPI: (Scripted) Disintegrate include/linux/netfilter_ipv4 UAPI: (Scripted) Disintegrate include/linux/netfilter_bridge UAPI: (Scripted) Disintegrate include/linux/netfilter_arp UAPI: (Scripted) Disintegrate include/linux/netfilter/ipset UAPI: (Scripted) Disintegrate include/linux/netfilter UAPI: (Scripted) Disintegrate include/linux/isdn UAPI: (Scripted) Disintegrate include/linux/caif net: fix typo in freescale/ucc_geth.c vxlan: fix more sparse warnings ...
Diffstat (limited to 'net/core/dev.c')
-rw-r--r--net/core/dev.c59
1 files changed, 44 insertions, 15 deletions
diff --git a/net/core/dev.c b/net/core/dev.c
index 1e0a1847c3b..09cb3f6dc40 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3300,18 +3300,18 @@ ncls:
&& !skb_pfmemalloc_protocol(skb))
goto drop;
- rx_handler = rcu_dereference(skb->dev->rx_handler);
if (vlan_tx_tag_present(skb)) {
if (pt_prev) {
ret = deliver_skb(skb, pt_prev, orig_dev);
pt_prev = NULL;
}
- if (vlan_do_receive(&skb, !rx_handler))
+ if (vlan_do_receive(&skb))
goto another_round;
else if (unlikely(!skb))
goto unlock;
}
+ rx_handler = rcu_dereference(skb->dev->rx_handler);
if (rx_handler) {
if (pt_prev) {
ret = deliver_skb(skb, pt_prev, orig_dev);
@@ -3331,6 +3331,9 @@ ncls:
}
}
+ if (vlan_tx_nonzero_tag_present(skb))
+ skb->pkt_type = PACKET_OTHERHOST;
+
/* deliver only exact match when indicated */
null_or_dev = deliver_exact ? skb->dev : NULL;
@@ -3471,17 +3474,31 @@ out:
return netif_receive_skb(skb);
}
-inline void napi_gro_flush(struct napi_struct *napi)
+/* napi->gro_list contains packets ordered by age.
+ * youngest packets at the head of it.
+ * Complete skbs in reverse order to reduce latencies.
+ */
+void napi_gro_flush(struct napi_struct *napi, bool flush_old)
{
- struct sk_buff *skb, *next;
+ struct sk_buff *skb, *prev = NULL;
- for (skb = napi->gro_list; skb; skb = next) {
- next = skb->next;
+ /* scan list and build reverse chain */
+ for (skb = napi->gro_list; skb != NULL; skb = skb->next) {
+ skb->prev = prev;
+ prev = skb;
+ }
+
+ for (skb = prev; skb; skb = prev) {
skb->next = NULL;
+
+ if (flush_old && NAPI_GRO_CB(skb)->age == jiffies)
+ return;
+
+ prev = skb->prev;
napi_gro_complete(skb);
+ napi->gro_count--;
}
- napi->gro_count = 0;
napi->gro_list = NULL;
}
EXPORT_SYMBOL(napi_gro_flush);
@@ -3542,6 +3559,7 @@ enum gro_result dev_gro_receive(struct napi_struct *napi, struct sk_buff *skb)
napi->gro_count++;
NAPI_GRO_CB(skb)->count = 1;
+ NAPI_GRO_CB(skb)->age = jiffies;
skb_shinfo(skb)->gso_size = skb_gro_len(skb);
skb->next = napi->gro_list;
napi->gro_list = skb;
@@ -3631,20 +3649,22 @@ gro_result_t napi_skb_finish(gro_result_t ret, struct sk_buff *skb)
}
EXPORT_SYMBOL(napi_skb_finish);
-void skb_gro_reset_offset(struct sk_buff *skb)
+static void skb_gro_reset_offset(struct sk_buff *skb)
{
+ const struct skb_shared_info *pinfo = skb_shinfo(skb);
+ const skb_frag_t *frag0 = &pinfo->frags[0];
+
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_frag_page(&skb_shinfo(skb)->frags[0]))) {
- NAPI_GRO_CB(skb)->frag0 =
- skb_frag_address(&skb_shinfo(skb)->frags[0]);
- NAPI_GRO_CB(skb)->frag0_len = skb_frag_size(&skb_shinfo(skb)->frags[0]);
+ pinfo->nr_frags &&
+ !PageHighMem(skb_frag_page(frag0))) {
+ NAPI_GRO_CB(skb)->frag0 = skb_frag_address(frag0);
+ NAPI_GRO_CB(skb)->frag0_len = skb_frag_size(frag0);
}
}
-EXPORT_SYMBOL(skb_gro_reset_offset);
gro_result_t napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb)
{
@@ -3876,7 +3896,7 @@ void napi_complete(struct napi_struct *n)
if (unlikely(test_bit(NAPI_STATE_NPSVC, &n->state)))
return;
- napi_gro_flush(n);
+ napi_gro_flush(n, false);
local_irq_save(flags);
__napi_complete(n);
local_irq_restore(flags);
@@ -3981,8 +4001,17 @@ static void net_rx_action(struct softirq_action *h)
local_irq_enable();
napi_complete(n);
local_irq_disable();
- } else
+ } else {
+ if (n->gro_list) {
+ /* flush too old packets
+ * If HZ < 1000, flush all packets.
+ */
+ local_irq_enable();
+ napi_gro_flush(n, HZ >= 1000);
+ local_irq_disable();
+ }
list_move_tail(&n->poll_list, &sd->poll_list);
+ }
}
netpoll_poll_unlock(have);