diff options
| author | Eric Dumazet <edumazet@google.com> | 2013-07-30 17:55:08 -0700 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2013-07-31 14:59:49 -0700 |
| commit | f2f872f9272a79a1048877ea14c15576f46c225e (patch) | |
| tree | 22af2e425ac96068c2d383c052e1d287db867211 /net/core/sock.c | |
| parent | ca4c3fc24e293719fe7410c4e63da9b6bc633b83 (diff) | |
| download | linux-stable-f2f872f9272a79a1048877ea14c15576f46c225e.tar.gz linux-stable-f2f872f9272a79a1048877ea14c15576f46c225e.tar.bz2 linux-stable-f2f872f9272a79a1048877ea14c15576f46c225e.zip | |
netem: Introduce skb_orphan_partial() helper
Commit 547669d483e578 ("tcp: xps: fix reordering issues") added
unexpected reorders in case netem is used in a MQ setup for high
performance test bed.
ETH=eth0
tc qd del dev $ETH root 2>/dev/null
tc qd add dev $ETH root handle 1: mq
for i in `seq 1 32`
do
tc qd add dev $ETH parent 1:$i netem delay 100ms
done
As all tcp packets are orphaned by netem, TCP stack believes it can
set skb->ooo_okay on all packets.
In order to allow producers to send more packets, we want to
keep sk_wmem_alloc from reaching sk_sndbuf limit.
We can do that by accounting one byte per skb in netem queues,
so that TCP stack is not fooled too much.
Tested:
With above MQ/netem setup, scaling number of concurrent flows gives
linear results and no reorders/retransmits
lpq83:~# for n in 1 10 20 30 40 50 60 70 80 90 100
do echo -n "n:$n " ; ./super_netperf $n -H 10.7.7.84; done
n:1 198.46
n:10 2002.69
n:20 4000.98
n:30 6006.35
n:40 8020.93
n:50 10032.3
n:60 12081.9
n:70 13971.3
n:80 16009.7
n:90 17117.3
n:100 17425.5
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/core/sock.c')
| -rw-r--r-- | net/core/sock.c | 19 |
1 files changed, 19 insertions, 0 deletions
diff --git a/net/core/sock.c b/net/core/sock.c index 85e8de1bc7fd..a753d97434dc 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1576,6 +1576,25 @@ void sock_wfree(struct sk_buff *skb) } EXPORT_SYMBOL(sock_wfree); +void skb_orphan_partial(struct sk_buff *skb) +{ + /* TCP stack sets skb->ooo_okay based on sk_wmem_alloc, + * so we do not completely orphan skb, but transfert all + * accounted bytes but one, to avoid unexpected reorders. + */ + if (skb->destructor == sock_wfree +#ifdef CONFIG_INET + || skb->destructor == tcp_wfree +#endif + ) { + atomic_sub(skb->truesize - 1, &skb->sk->sk_wmem_alloc); + skb->truesize = 1; + } else { + skb_orphan(skb); + } +} +EXPORT_SYMBOL(skb_orphan_partial); + /* * Read buffer destructor automatically called from kfree_skb. */ |