diff options
| author | Pavel Kubelun <be.dissent@gmail.com> | 2017-07-25 03:57:31 -0400 |
|---|---|---|
| committer | Stijn Tintel <stijn@linux-ipv6.be> | 2017-10-08 15:19:07 +0300 |
| commit | b2ea46fe236ae02d845f94cf984cbc1786870333 (patch) | |
| tree | 277ff9cd154420367c261324e69ba831e3c762b7 /target/linux/generic/backport-4.9/024-5-tcp-tsq-add-a-shortcut-in-tcp_small_queue_check.patch | |
| parent | 3a69ad3b2a0df578ca340ba95f35230ee9529c9c (diff) | |
| download | upstream-b2ea46fe236ae02d845f94cf984cbc1786870333.tar.gz upstream-b2ea46fe236ae02d845f94cf984cbc1786870333.tar.bz2 upstream-b2ea46fe236ae02d845f94cf984cbc1786870333.zip | |
generic: net: tcp: backport tcp tx performance patches
An overall throughput gain of 22 % for heavy TCP use over a single TX queue.
Original patchset comment
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v4.13&id=3f4888adae7c1619b990d98a9b967536f71822b8
Signed-off-by: Pavel Kubelun <be.dissent@gmail.com>
Diffstat (limited to 'target/linux/generic/backport-4.9/024-5-tcp-tsq-add-a-shortcut-in-tcp_small_queue_check.patch')
| -rw-r--r-- | target/linux/generic/backport-4.9/024-5-tcp-tsq-add-a-shortcut-in-tcp_small_queue_check.patch | 37 |
1 files changed, 37 insertions, 0 deletions
diff --git a/target/linux/generic/backport-4.9/024-5-tcp-tsq-add-a-shortcut-in-tcp_small_queue_check.patch b/target/linux/generic/backport-4.9/024-5-tcp-tsq-add-a-shortcut-in-tcp_small_queue_check.patch new file mode 100644 index 0000000000..dd1a37eb21 --- /dev/null +++ b/target/linux/generic/backport-4.9/024-5-tcp-tsq-add-a-shortcut-in-tcp_small_queue_check.patch @@ -0,0 +1,37 @@ +From 75eefc6c59fd2c5f1ab95a3a113c217237d12a31 Mon Sep 17 00:00:00 2001 +From: Eric Dumazet <edumazet@google.com> +Date: Sat, 3 Dec 2016 11:14:54 -0800 +Subject: [PATCH 05/10] tcp: tsq: add a shortcut in tcp_small_queue_check() + +Always allow the two first skbs in write queue to be sent, +regardless of sk_wmem_alloc/sk_pacing_rate values. + +This helps a lot in situations where TX completions are delayed either +because of driver latencies or softirq latencies. + +Test is done with no cache line misses. + +Signed-off-by: Eric Dumazet <edumazet@google.com> +Signed-off-by: David S. Miller <davem@davemloft.net> +--- + net/ipv4/tcp_output.c | 9 +++++++++ + 1 file changed, 9 insertions(+) + +--- a/net/ipv4/tcp_output.c ++++ b/net/ipv4/tcp_output.c +@@ -2084,6 +2084,15 @@ static bool tcp_small_queue_check(struct + limit <<= factor; + + if (atomic_read(&sk->sk_wmem_alloc) > limit) { ++ /* Always send the 1st or 2nd skb in write queue. ++ * No need to wait for TX completion to call us back, ++ * after softirq/tasklet schedule. ++ * This helps when TX completions are delayed too much. ++ */ ++ if (skb == sk->sk_write_queue.next || ++ skb->prev == sk->sk_write_queue.next) ++ return false; ++ + set_bit(TSQ_THROTTLED, &tcp_sk(sk)->tsq_flags); + /* It is possible TX completion already happened + * before we set TSQ_THROTTLED, so we must |