Re: [PATCH net-next 2/3] tcp: implement coalescing on backlog queue

[Eric Dumazet]

On Thu, Nov 22, 2018 at 10:16 AM Eric Dumazet <eduma...@google.com> wrote:

> Yes, I was considering properly filtering SACK as a refinement later [1]
> but you raise a valid point for alien stacks that are not yet using SACK :/
>
> [1] This version of the patch will not aggregate sacks since the
> memcmp() on tcp options would fail.
>
> Neal can you double check if cake_ack_filter() does not have the issue
> you just mentioned ?

Note that aggregated pure acks will have a gso_segs set to the number
of aggregated acks,
we might simply use this value later in the stack, instead of forcing
having X pure acks in the backlog
and increase memory needs and cpu costs.

Then I guess I need this fix :

diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 
36c9d715bf2aa7eb7bf58b045bfeb85a2ec1a696..736f7f24cdb4fe61769faaa1644c8bff01c746c4
100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1669,7 +1669,8 @@ bool tcp_add_backlog(struct sock *sk, struct sk_buff *skb)
                __skb_pull(skb, hdrlen);
                if (skb_try_coalesce(tail, skb, &fragstolen, &delta)) {
                        TCP_SKB_CB(tail)->end_seq = TCP_SKB_CB(skb)->end_seq;
-                       TCP_SKB_CB(tail)->ack_seq = TCP_SKB_CB(skb)->ack_seq;
+                       if (after(TCP_SKB_CB(skb)->ack_seq,
TCP_SKB_CB(tail)->ack_seq))
+                               TCP_SKB_CB(tail)->ack_seq =
TCP_SKB_CB(skb)->ack_seq;
                        TCP_SKB_CB(tail)->tcp_flags |=
TCP_SKB_CB(skb)->tcp_flags;

                        if (TCP_SKB_CB(skb)->has_rxtstamp) {