aboutsummaryrefslogtreecommitdiffstats
path: root/net/ipv4
diff options
context:
space:
mode:
authorEric Dumazet <eric.dumazet@gmail.com>2010-04-27 18:13:20 -0400
committerDavid S. Miller <davem@davemloft.net>2010-04-27 18:13:20 -0400
commitc377411f2494a931ff7facdbb3a6839b1266bcf6 (patch)
tree6846cdcec913f50839e3916856f78f7e059ff5fb /net/ipv4
parent6e7676c1a76aed6e957611d8d7a9e5592e23aeba (diff)
net: sk_add_backlog() take rmem_alloc into account
Current socket backlog limit is not enough to really stop DDOS attacks, because user thread spend many time to process a full backlog each round, and user might crazy spin on socket lock. We should add backlog size and receive_queue size (aka rmem_alloc) to pace writers, and let user run without being slow down too much. Introduce a sk_rcvqueues_full() helper, to avoid taking socket lock in stress situations. Under huge stress from a multiqueue/RPS enabled NIC, a single flow udp receiver can now process ~200.000 pps (instead of ~100 pps before the patch) on a 8 core machine. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/ipv4')
-rw-r--r--net/ipv4/udp.c4
1 files changed, 4 insertions, 0 deletions
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index fa3d2874db4..63eb56b2d87 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -1372,6 +1372,10 @@ int udp_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
1372 goto drop; 1372 goto drop;
1373 } 1373 }
1374 1374
1375
1376 if (sk_rcvqueues_full(sk, skb))
1377 goto drop;
1378
1375 rc = 0; 1379 rc = 0;
1376 1380
1377 bh_lock_sock(sk); 1381 bh_lock_sock(sk);