From e3f2c4a3db1413bebfd502f7ac94fb55e3ba8c84 Mon Sep 17 00:00:00 2001 From: Soheil Hassas Yeganeh Date: Wed, 3 Jan 2018 21:47:10 -0500 Subject: ip: do not set RFS core on error queue reads We should only record RPS on normal reads and writes. In single threaded processes, all calls record the same state. In multi-threaded processes where a separate thread processes errors, the RFS table mispredicts. Note that, when CONFIG_RPS is disabled, sock_rps_record_flow is a noop and no branch is added as a result of this patch. Signed-off-by: Soheil Hassas Yeganeh Signed-off-by: Willem de Bruijn Signed-off-by: Eric Dumazet Signed-off-by: Neal Cardwell Signed-off-by: David S. Miller --- net/ipv4/af_inet.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index bab98a4fedad..54cccdd8b1e3 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -790,7 +790,8 @@ int inet_recvmsg(struct socket *sock, struct msghdr *msg, size_t size, int addr_len = 0; int err; - sock_rps_record_flow(sk); + if (likely(!(flags & MSG_ERRQUEUE))) + sock_rps_record_flow(sk); err = sk->sk_prot->recvmsg(sk, msg, size, flags & MSG_DONTWAIT, flags & ~MSG_DONTWAIT, &addr_len); -- cgit v1.2.3-59-g8ed1b