| Message ID | 150591153693.113604.8604505743746410801.stgit@buzz |
|---|---|
| State | Accepted, archived |
| Delegated to: | David Miller |
| Headers | show |
| Series | net_sched: always reset qdisc backlog in qdisc_reset() | expand |
On Wed, Sep 20, 2017 at 5:45 AM, Konstantin Khlebnikov <khlebnikov@yandex-team.ru> wrote: > SKB stored in qdisc->gso_skb also counted into backlog. > > Some qdiscs don't reset backlog to zero in ->reset(), > for example sfq just dequeue and free all queued skb. > > Signed-off-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru> > Fixes: 2ccccf5fb43f ("net_sched: update hierarchical backlog too") Looks good. sch->qstats.backlog = 0 can be removed from each ->reset() after this patch.
From: Konstantin Khlebnikov <khlebnikov@yandex-team.ru> Date: Wed, 20 Sep 2017 15:45:36 +0300 > SKB stored in qdisc->gso_skb also counted into backlog. > > Some qdiscs don't reset backlog to zero in ->reset(), > for example sfq just dequeue and free all queued skb. > > Signed-off-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru> > Fixes: 2ccccf5fb43f ("net_sched: update hierarchical backlog too") Applied and queued up for -stable.
diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index 92237e75dbbc..bf8c81e07c70 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -685,6 +685,7 @@ void qdisc_reset(struct Qdisc *qdisc) qdisc->gso_skb = NULL; } qdisc->q.qlen = 0; + qdisc->qstats.backlog = 0; } EXPORT_SYMBOL(qdisc_reset);
SKB stored in qdisc->gso_skb also counted into backlog. Some qdiscs don't reset backlog to zero in ->reset(), for example sfq just dequeue and free all queued skb. Signed-off-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru> Fixes: 2ccccf5fb43f ("net_sched: update hierarchical backlog too") --- net/sched/sch_generic.c | 1 + 1 file changed, 1 insertion(+)