for connected socket, the incoming_cpu field in the sock struct
is not going to change frequently, but we are setting it
unconditionally for each packet.
Since sk_incoming_cpu and sk_flags share the same cacheline,
and the latter is access by udp_recvmsg(), this cause a cache
miss for each packet for UDP connected socket.
With this patch, we set the incoming cpu field only when the
ingress cpu really changes.
This gives a small but measurable performance improvement for
connected UDP socket.
Signed-off-by: Paolo Abeni <pab...@redhat.com>
---
include/net/sock.h | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/include/net/sock.h b/include/net/sock.h
index 858891c..00d0914 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -907,7 +907,10 @@ static inline int sk_backlog_rcv(struct sock *sk, struct
sk_buff *skb)
static inline void sk_incoming_cpu_update(struct sock *sk)
{
- sk->sk_incoming_cpu = raw_smp_processor_id();
+ int cpu = raw_smp_processor_id();
+
+ if (unlikely(sk->sk_incoming_cpu != cpu))
+ sk->sk_incoming_cpu = cpu;
}
static inline void sock_rps_record_flow_hash(__u32 hash)
--
2.9.4
