[PATCH net-next] tcp: defer sacked assignment

[PATCH net-next] tcp: defer sacked assignment

[Eric Dumazet]

From: Eric Dumazet <edumazet@google.com>

While chasing tcp_xmit_retransmit_queue() kasan issue, I found
that we could avoid reading sacked field of skb that we wont send,
possibly removing one cache line miss.

Very minor change in slow path, but why not ? ;) 

Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 net/ipv4/tcp_output.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index bdaef7fd6e47701b2d2280e1f40f816e1920c432..8b45794eb6b2600c914e504b8031f61dbbe164a3 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -2776,7 +2776,7 @@ void tcp_xmit_retransmit_queue(struct sock *sk)
 
 	max_segs = tcp_tso_autosize(sk, tcp_current_mss(sk));
 	tcp_for_write_queue_from(skb, sk) {
-		__u8 sacked = TCP_SKB_CB(skb)->sacked;
+		__u8 sacked;
 		int segs;
 
 		if (skb == tcp_send_head(sk))
@@ -2788,6 +2788,7 @@ void tcp_xmit_retransmit_queue(struct sock *sk)
 		segs = tp->snd_cwnd - tcp_packets_in_flight(tp);
 		if (segs <= 0)
 			return;
+		sacked = TCP_SKB_CB(skb)->sacked;
 		/* In case tcp_shift_skb_data() have aggregated large skbs,
 		 * we need to make sure not sending too bigs TSO packets
 		 */

Re: [PATCH net-next] tcp: defer sacked assignment

[Neal Cardwell]

On Wed, Aug 17, 2016 at 10:48 AM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
>
> From: Eric Dumazet <edumazet@google.com>
>
> While chasing tcp_xmit_retransmit_queue() kasan issue, I found
> that we could avoid reading sacked field of skb that we wont send,
> possibly removing one cache line miss.
>
> Very minor change in slow path, but why not ? ;)
>
> Signed-off-by: Eric Dumazet <edumazet@google.com>
> ---

Acked-by: Neal Cardwell <ncardwell@google.com>

neal

Re: [PATCH net-next] tcp: defer sacked assignment

[David Miller]

From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Wed, 17 Aug 2016 07:48:36 -0700

> From: Eric Dumazet <edumazet@google.com>
> 
> While chasing tcp_xmit_retransmit_queue() kasan issue, I found
> that we could avoid reading sacked field of skb that we wont send,
> possibly removing one cache line miss.
> 
> Very minor change in slow path, but why not ? ;) 
> 
> Signed-off-by: Eric Dumazet <edumazet@google.com>

Applied.