summaryrefslogtreecommitdiffstats
path: root/net/sched/sch_fq_codel.c
diff options
context:
space:
mode:
authorEric Dumazet <edumazet@google.com>2016-06-06 18:37:16 +0200
committerDavid S. Miller <davem@davemloft.net>2016-06-08 01:37:14 +0200
commitedb09eb17ed89eaa82a52dd306beac93e292b485 (patch)
tree1f241506d6b781b65d1033925f1c1ce6a39c3394 /net/sched/sch_fq_codel.c
parentnet_sched: transform qdisc running bit into a seqcount (diff)
downloadlinux-edb09eb17ed89eaa82a52dd306beac93e292b485.tar.xz
linux-edb09eb17ed89eaa82a52dd306beac93e292b485.zip
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>
Diffstat (limited to 'net/sched/sch_fq_codel.c')
-rw-r--r--net/sched/sch_fq_codel.c15
1 files changed, 11 insertions, 4 deletions
diff --git a/net/sched/sch_fq_codel.c b/net/sched/sch_fq_codel.c
index 6883a8971562..1daa54237f4e 100644
--- a/net/sched/sch_fq_codel.c
+++ b/net/sched/sch_fq_codel.c
@@ -566,11 +566,13 @@ static int fq_codel_dump_stats(struct Qdisc *sch, struct gnet_dump *d)
st.qdisc_stats.memory_usage = q->memory_usage;
st.qdisc_stats.drop_overmemory = q->drop_overmemory;
+ sch_tree_lock(sch);
list_for_each(pos, &q->new_flows)
st.qdisc_stats.new_flows_len++;
list_for_each(pos, &q->old_flows)
st.qdisc_stats.old_flows_len++;
+ sch_tree_unlock(sch);
return gnet_stats_copy_app(d, &st, sizeof(st));
}
@@ -624,7 +626,7 @@ static int fq_codel_dump_class_stats(struct Qdisc *sch, unsigned long cl,
if (idx < q->flows_cnt) {
const struct fq_codel_flow *flow = &q->flows[idx];
- const struct sk_buff *skb = flow->head;
+ const struct sk_buff *skb;
memset(&xstats, 0, sizeof(xstats));
xstats.type = TCA_FQ_CODEL_XSTATS_CLASS;
@@ -642,9 +644,14 @@ static int fq_codel_dump_class_stats(struct Qdisc *sch, unsigned long cl,
codel_time_to_us(delta) :
-codel_time_to_us(-delta);
}
- while (skb) {
- qs.qlen++;
- skb = skb->next;
+ if (flow->head) {
+ sch_tree_lock(sch);
+ skb = flow->head;
+ while (skb) {
+ qs.qlen++;
+ skb = skb->next;
+ }
+ sch_tree_unlock(sch);
}
qs.backlog = q->backlogs[idx];
qs.drops = flow->dropped;