net: sched: do not acquire qdisc spinlock in qdisc/class stats dump
Large tc dumps (tc -s {qdisc|class} sh dev ethX) done by Google BwE host
agent [1] are problematic at scale :
For each qdisc/class found in the dump, we currently lock the root qdisc
spinlock in order to get stats. Sampling stats every 5 seconds from
thousands of HTB classes is a challenge when the root qdisc spinlock is
under high pressure. Not only the dumps take time, they also slow
down the fast path (queue/dequeue packets) by 10 % to 20 % in some cases.
An audit of existing qdiscs showed that sch_fq_codel is the only qdisc
that might need the qdisc lock in fq_codel_dump_stats() and
fq_codel_dump_class_stats()
In v2 of this patch, I now use the Qdisc running seqcount to provide
consistent reads of packets/bytes counters, regardless of 32/64 bit arches.
I also changed rate estimators to use the same infrastructure
so that they no longer need to lock root qdisc lock.
[1]
http://static.googleusercontent.com/media/research.google.com/en//pubs/archive/43838.pdf
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Kevin Athey <kda@google.com>
Cc: Xiaotian Pei <xiaotian@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c
index ddf047d..d4a8bbf 100644
--- a/net/sched/sch_api.c
+++ b/net/sched/sch_api.c
@@ -982,7 +982,7 @@
rcu_assign_pointer(sch->stab, stab);
}
if (tca[TCA_RATE]) {
- spinlock_t *root_lock;
+ seqcount_t *running;
err = -EOPNOTSUPP;
if (sch->flags & TCQ_F_MQROOT)
@@ -991,14 +991,15 @@
if ((sch->parent != TC_H_ROOT) &&
!(sch->flags & TCQ_F_INGRESS) &&
(!p || !(p->flags & TCQ_F_MQROOT)))
- root_lock = qdisc_root_sleeping_lock(sch);
+ running = qdisc_root_sleeping_running(sch);
else
- root_lock = qdisc_lock(sch);
+ running = &sch->running;
err = gen_new_estimator(&sch->bstats,
sch->cpu_bstats,
&sch->rate_est,
- root_lock,
+ NULL,
+ running,
tca[TCA_RATE]);
if (err)
goto err_out4;
@@ -1061,7 +1062,8 @@
gen_replace_estimator(&sch->bstats,
sch->cpu_bstats,
&sch->rate_est,
- qdisc_root_sleeping_lock(sch),
+ NULL,
+ qdisc_root_sleeping_running(sch),
tca[TCA_RATE]);
}
out:
@@ -1369,8 +1371,7 @@
goto nla_put_failure;
if (gnet_stats_start_copy_compat(skb, TCA_STATS2, TCA_STATS, TCA_XSTATS,
- qdisc_root_sleeping_lock(q), &d,
- TCA_PAD) < 0)
+ NULL, &d, TCA_PAD) < 0)
goto nla_put_failure;
if (q->ops->dump_stats && q->ops->dump_stats(q, &d) < 0)
@@ -1381,7 +1382,8 @@
cpu_qstats = q->cpu_qstats;
}
- if (gnet_stats_copy_basic(&d, cpu_bstats, &q->bstats) < 0 ||
+ if (gnet_stats_copy_basic(qdisc_root_sleeping_running(q),
+ &d, cpu_bstats, &q->bstats) < 0 ||
gnet_stats_copy_rate_est(&d, &q->bstats, &q->rate_est) < 0 ||
gnet_stats_copy_queue(&d, cpu_qstats, &q->qstats, qlen) < 0)
goto nla_put_failure;
@@ -1684,8 +1686,7 @@
goto nla_put_failure;
if (gnet_stats_start_copy_compat(skb, TCA_STATS2, TCA_STATS, TCA_XSTATS,
- qdisc_root_sleeping_lock(q), &d,
- TCA_PAD) < 0)
+ NULL, &d, TCA_PAD) < 0)
goto nla_put_failure;
if (cl_ops->dump_stats && cl_ops->dump_stats(q, cl, &d) < 0)
