diff options
| author | Eric Dumazet <edumazet@google.com> | 2018-11-11 07:34:28 -0800 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2018-11-11 13:57:03 -0800 |
| commit | c73e5807e4f6fc6d373a5db55b45f639f8bb6328 (patch) | |
| tree | 09c332f796e245e7ec15cc31905e61f793da451b | |
| parent | 83afb36a7097bdf6776e261fa3a2bf2dc5a4a8c2 (diff) | |
| download | linux-c73e5807e4f6fc6d373a5db55b45f639f8bb6328.tar.gz linux-c73e5807e4f6fc6d373a5db55b45f639f8bb6328.tar.bz2 linux-c73e5807e4f6fc6d373a5db55b45f639f8bb6328.zip | |
tcp: tsq: no longer use limit_output_bytes for paced flows
FQ pacing guarantees that paced packets queued by one flow do not
add head-of-line blocking for other flows.
After TCP GSO conversion, increasing limit_output_bytes to 1 MB is safe,
since this maps to 16 skbs at most in qdisc or device queues.
(or slightly more if some drivers lower {gso_max_segs|size})
We still can queue at most 1 ms worth of traffic (this can be scaled
by wifi drivers if they need to)
Tested:
# ethtool -c eth0 | egrep "tx-usecs:|tx-frames:" # 40 Gbit mlx4 NIC
tx-usecs: 16
tx-frames: 16
# tc qdisc replace dev eth0 root fq
# for f in {1..10};do netperf -P0 -H lpaa24,6 -o THROUGHPUT;done
Before patch:
27711
26118
27107
27377
27712
27388
27340
27117
27278
27509
After patch:
37434
36949
36658
36998
37711
37291
37605
36659
36544
37349
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
| -rw-r--r-- | Documentation/networking/ip-sysctl.txt | 2 | ||||
| -rw-r--r-- | net/ipv4/tcp_ipv4.c | 4 | ||||
| -rw-r--r-- | net/ipv4/tcp_output.c | 5 |
3 files changed, 6 insertions, 5 deletions
diff --git a/Documentation/networking/ip-sysctl.txt b/Documentation/networking/ip-sysctl.txt index aa9e6a331679..af2a69439b93 100644 --- a/Documentation/networking/ip-sysctl.txt +++ b/Documentation/networking/ip-sysctl.txt @@ -759,7 +759,7 @@ tcp_limit_output_bytes - INTEGER flows, for typical pfifo_fast qdiscs. tcp_limit_output_bytes limits the number of bytes on qdisc or device to reduce artificial RTT/cwnd and reduce bufferbloat. - Default: 262144 + Default: 1048576 (16 * 65536) tcp_challenge_ack_limit - INTEGER Limits number of Challenge ACK sent per second, as recommended diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 5424a4077c27..0952d4b772e7 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -2574,8 +2574,8 @@ static int __net_init tcp_sk_init(struct net *net) * which are too large can cause TCP streams to be bursty. */ net->ipv4.sysctl_tcp_tso_win_divisor = 3; - /* Default TSQ limit of four TSO segments */ - net->ipv4.sysctl_tcp_limit_output_bytes = 262144; + /* Default TSQ limit of 16 TSO segments */ + net->ipv4.sysctl_tcp_limit_output_bytes = 16 * 65536; /* rfc5961 challenge ack rate limiting */ net->ipv4.sysctl_tcp_challenge_ack_limit = 1000; net->ipv4.sysctl_tcp_min_tso_segs = 2; diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 75dcf4daca72..d40d4cc53319 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -2220,8 +2220,9 @@ static bool tcp_small_queue_check(struct sock *sk, const struct sk_buff *skb, limit = max_t(unsigned long, 2 * skb->truesize, sk->sk_pacing_rate >> sk->sk_pacing_shift); - limit = min_t(unsigned long, limit, - sock_net(sk)->ipv4.sysctl_tcp_limit_output_bytes); + if (sk->sk_pacing_status == SK_PACING_NONE) + limit = min_t(unsigned long, limit, + sock_net(sk)->ipv4.sysctl_tcp_limit_output_bytes); limit <<= factor; if (refcount_read(&sk->sk_wmem_alloc) > limit) { |