net: introduce __skb_header_release()
authorEric Dumazet <edumazet@google.com>
Mon, 22 Sep 2014 23:29:32 +0000 (16:29 -0700)
committerDavid S. Miller <davem@davemloft.net>
Fri, 26 Sep 2014 19:40:06 +0000 (15:40 -0400)
While profiling TCP stack, I noticed one useless atomic operation
in tcp_sendmsg(), caused by skb_header_release().

It turns out all current skb_header_release() users have a fresh skb,
that no other user can see, so we can avoid one atomic operation.

Introduce __skb_header_release() to clearly document this.

This gave me a 1.5 % improvement on TCP_RR workload.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

No differences found