lk/linux-kernel-test
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

tcp: Combat per-cpu skew in orphan tests.

Browse Source 
As reported by Anton Blanchard when we use
percpu_counter_read_positive() to make our orphan socket limit checks,
the check can be off by up to num_cpus_online() * batch (which is 32
by default) which on a 128 cpu machine can be as large as the default
orphan limit itself.

Fix this by doing the full expensive sum check if the optimized check
triggers.

Reported-by: Anton Blanchard <anton@samba.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
This commit is contained in:
David S. Miller
2010-08-25 02:27:49 -07:00
parent b2bc85631e
commit ad1af0fedb
3 changed files with 19 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
net/ipv4/tcp.c
View File

@@ -2011,11 +2011,8 @@ adjudge_to_death:
}
}
if (sk->sk_state != TCP_CLOSE) {
int orphan_count = percpu_counter_read_positive(
sk->sk_prot->orphan_count);
sk_mem_reclaim(sk);
if (tcp_too_many_orphans(sk, orphan_count)) {
if (tcp_too_many_orphans(sk, 0)) {
if (net_ratelimit())
printk(KERN_INFO "TCP: too many of orphaned "
"sockets\n");