diff options
| author | Paolo Abeni <pabeni@redhat.com> | 2023-09-12 19:10:03 +0200 |
|---|---|---|
| committer | Paolo Abeni <pabeni@redhat.com> | 2023-09-12 19:10:04 +0200 |
| commit | 8fc8911b66962c6ff4345e7000930a4bcc54ae5a (patch) | |
| tree | 987f1ac9958a489af0b4bfd4726fb91ce94f028d /net/ipv4/tcp_output.c | |
| parent | cd8bae85815416d19f46e3828d457442f77de292 (diff) | |
| parent | 133c4c0d37175f510a10fa9bed51e223936073fc (diff) | |
Merge branch 'tcp-backlog-processing-optims'
Eric Dumazet says:
====================
tcp: backlog processing optims
First patches are mostly preparing the ground for the last one.
Last patch of the series implements sort of ACK reduction
only for the cases a TCP receiver is under high stress,
which happens for high throughput flows.
This gives us a ~20% increase of single TCP flow (100Gbit -> 120Gbit)
====================
Link: https://lore.kernel.org/r/20230911170531.828100-1-edumazet@google.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Diffstat (limited to 'net/ipv4/tcp_output.c')
| -rw-r--r-- | net/ipv4/tcp_output.c | 15 |
1 files changed, 4 insertions, 11 deletions
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index ccfc8bbf74558..1fc1f879cfd6c 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -1077,7 +1077,8 @@ static void tcp_tasklet_func(struct tasklet_struct *t) #define TCP_DEFERRED_ALL (TCPF_TSQ_DEFERRED | \ TCPF_WRITE_TIMER_DEFERRED | \ TCPF_DELACK_TIMER_DEFERRED | \ - TCPF_MTU_REDUCED_DEFERRED) + TCPF_MTU_REDUCED_DEFERRED | \ + TCPF_ACK_DEFERRED) /** * tcp_release_cb - tcp release_sock() callback * @sk: socket @@ -1101,16 +1102,6 @@ void tcp_release_cb(struct sock *sk) tcp_tsq_write(sk); __sock_put(sk); } - /* Here begins the tricky part : - * We are called from release_sock() with : - * 1) BH disabled - * 2) sk_lock.slock spinlock held - * 3) socket owned by us (sk->sk_lock.owned == 1) - * - * But following code is meant to be called from BH handlers, - * so we should keep BH disabled, but early release socket ownership - */ - sock_release_ownership(sk); if (flags & TCPF_WRITE_TIMER_DEFERRED) { tcp_write_timer_handler(sk); @@ -1124,6 +1115,8 @@ void tcp_release_cb(struct sock *sk) inet_csk(sk)->icsk_af_ops->mtu_reduced(sk); __sock_put(sk); } + if ((flags & TCPF_ACK_DEFERRED) && inet_csk_ack_scheduled(sk)) + tcp_send_ack(sk); } EXPORT_SYMBOL(tcp_release_cb); |