diff options
| author | Eric Dumazet <edumazet@google.com> | 2014-10-04 10:11:31 -0700 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2014-10-06 00:55:10 -0400 |
| commit | f2600cf02b5b59aaee082c3485b7f01fc7f7b70c (patch) | |
| tree | f38d692c3f706ead3d0cfd77b216fbd629a22964 /net/sched/sch_tbf.c | |
| parent | 681d2421e1135b95f5cd9d16fe10eac7f570a9f2 (diff) | |
| download | linux-f2600cf02b5b59aaee082c3485b7f01fc7f7b70c.tar.bz2 | |
net: sched: avoid costly atomic operation in fq_dequeue()
Standard qdisc API to setup a timer implies an atomic operation on every
packet dequeue : qdisc_unthrottled()
It turns out this is not really needed for FQ, as FQ has no concept of
global qdisc throttling, being a qdisc handling many different flows,
some of them can be throttled, while others are not.
Fix is straightforward : add a 'bool throttle' to
qdisc_watchdog_schedule_ns(), and remove calls to qdisc_unthrottled()
in sch_fq.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/sched/sch_tbf.c')
| -rw-r--r-- | net/sched/sch_tbf.c | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/net/sched/sch_tbf.c b/net/sched/sch_tbf.c index 77edffe329c4..a4afde14e865 100644 --- a/net/sched/sch_tbf.c +++ b/net/sched/sch_tbf.c @@ -268,7 +268,8 @@ static struct sk_buff *tbf_dequeue(struct Qdisc *sch) } qdisc_watchdog_schedule_ns(&q->watchdog, - now + max_t(long, -toks, -ptoks)); + now + max_t(long, -toks, -ptoks), + true); /* Maybe we have a shorter packet in the queue, which can be sent now. It sounds cool, |