af_unix: optimize unix_dgram_poll()

unix_dgram_poll() is pretty expensive to check POLLOUT status, because
it has to lock the socket to get its peer, take a reference on the peer
to check its receive queue status, and queue another poll_wait on
peer_wait. This all can be avoided if the process calling
unix_dgram_poll() is not interested in POLLOUT status. It makes
unix_dgram_recvmsg() faster by not queueing irrelevant pollers in
peer_wait.

On a test program provided by Alan Crequy :

Before:

real 0m0.211s
user 0m0.000s
sys 0m0.208s

After:

real 0m0.044s
user 0m0.000s
sys 0m0.040s

Suggested-by: Davide Libenzi <[e-mail]>
Reported-by: Alban Crequy <[e-mail]>
Acked-by: Davide Libenzi <[e-mail]>
Signed-off-by: Eric Dumazet <[e-mail]>
Signed-off-by: David S. Miller <[e-mail]>

net/unix/af_unix.c

···
 			return mask;
 	}
 
+	/* No write status requested, avoid expensive OUT tests. */
+	if (wait && !(wait->key & (POLLWRBAND | POLLWRNORM | POLLOUT)))
+		return mask;
+
 	writable = unix_writable(sk);
 	other = unix_peer_get(sk);
 	if (other) {