diff options
| author |   Eric Dumazet <edumazet@google.com> | 2016-04-13 22:05:40 -0700 |
|---|---|---|
| committer |   David S. Miller <davem@davemloft.net> | 2016-04-15 16:45:44 -0400 |
| commit | 8804b2722dc5d6f9b7ba0a9e812eae9ee5ce95bc (patch) | |
| tree | fd1010fca7c152ca547f02d80a02dc9d1b9d7b20 /net/ipv4/tcp_input.c | |
| parent | tcp: do not mess with listener sk_wmem_alloc (diff) | |
| download | linux-dev-8804b2722dc5d6f9b7ba0a9e812eae9ee5ce95bc.tar.xz linux-dev-8804b2722dc5d6f9b7ba0a9e812eae9ee5ce95bc.zip | |
tcp: remove false sharing in tcp_rcv_state_process()
Last known hot point during SYNFLOOD attack is the clearing
of rx_opt.saw_tstamp in tcp_rcv_state_process()
It is not needed for a listener, so we move it where it matters.
Performance while a SYNFLOOD hits a single listener socket
went from 5 Mpps to 6 Mpps on my test server (24 cores, 8 NIC RX queues)
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/ipv4/tcp_input.c')
| -rw-r--r-- | net/ipv4/tcp_input.c | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 7ea7034af83f..90e0d9256b74 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -5796,8 +5796,6 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb) int queued = 0; bool acceptable; - tp->rx_opt.saw_tstamp = 0; - switch (sk->sk_state) { case TCP_CLOSE: goto discard; @@ -5838,6 +5836,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb) goto discard; case TCP_SYN_SENT: + tp->rx_opt.saw_tstamp = 0; queued = tcp_rcv_synsent_state_process(sk, skb, th); if (queued >= 0) return queued; @@ -5849,6 +5848,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb) return 0; } + tp->rx_opt.saw_tstamp = 0; req = tp->fastopen_rsk; if (req) { WARN_ON_ONCE(sk->sk_state != TCP_SYN_RECV && |