tcp: make challenge acks faster

When backporting upstream commit 75ff39ccc1bd ("tcp: make challenge acks
less predictable") I negelected to use the correct ACCESS* type macros.
This fixes that up to hopefully speed things up a bit more.

Thanks to Chas Wiliams for the 3.10 backport which reminded me of this.

Cc: Yue Cao <ycao009@ucr.edu>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Chas Williams <ciwillia@brocade.com>
Cc: Willy Tarreau <w@1wt.eu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 90f9d00a3fbc1e9971323fd862373004ff23a1cf..963b7f74677753d2dc5dbac8dce30144fc357750 100644 (file)
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -3299,12 +3299,12 @@ static void tcp_send_challenge_ack(struct sock *sk)
                u32 half = (sysctl_tcp_challenge_ack_limit + 1) >> 1;
 
                challenge_timestamp = now;
-               challenge_count = half +
+               ACCESS_ONCE(challenge_count) = half +
                                  prandom_u32_max(sysctl_tcp_challenge_ack_limit);
        }
-       count = challenge_count;
+       count = ACCESS_ONCE(challenge_count);
        if (count > 0) {
-               challenge_count = count - 1;
+               ACCESS_ONCE(challenge_count) = count - 1;
                NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPCHALLENGEACK);
                tcp_send_ack(sk);
        }