AF_UNIX stream socket performance on P4 CPUs tends to suffer due to a lot of pipeline flushes from atomic operations. The patch below removes the sock_hold() and sock_put() in unix_stream_sendmsg(). This should be safe as the socket still holds a reference to its peer which is only released after the file descriptor's final user invokes unix_release_sock(). The only consideration is that we must add a memory barrier before setting the peer initially.
-ben Signed-off-by: Benjamin LaHaise <[EMAIL PROTECTED]> diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index acc73ba..fc444a8 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -1063,10 +1063,12 @@ restart: /* Set credentials */ sk->sk_peercred = other->sk_peercred; - sock_hold(newsk); - unix_peer(sk) = newsk; sock->state = SS_CONNECTED; sk->sk_state = TCP_ESTABLISHED; + sock_hold(newsk); + + smp_mb__after_atomic_inc(); /* sock_hold() does an atomic_inc() */ + unix_peer(sk) = newsk; unix_state_wunlock(sk); @@ -1414,7 +1416,7 @@ static int unix_stream_sendmsg(struct ki } else { sunaddr = NULL; err = -ENOTCONN; - other = unix_peer_get(sk); + other = unix_peer(sk); if (!other) goto out_err; } @@ -1476,7 +1478,6 @@ static int unix_stream_sendmsg(struct ki other->sk_data_ready(other, size); sent+=size; } - sock_put(other); scm_destroy(siocb->scm); siocb->scm = NULL; -- "You know, I've seen some crystals do some pretty trippy shit, man." Don't Email: <[EMAIL PROTECTED]>. - To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html