From: "Toke Høiland-Jørgensen" <toke@toke.dk>
To: Pete Heist <pete@heistp.net>
Cc: Cake List <cake@lists.bufferbloat.net>
Subject: Re: [Cake] cake infinite loop(?) with hfsc on one-armed router
Date: Sat, 05 Jan 2019 11:06:59 +0100 [thread overview]
Message-ID: <87wonjxvss.fsf@toke.dk> (raw)
In-Reply-To: <B1F84884-7150-41BB-9E39-640F560F99D5@heistp.net>
Pete Heist <pete@heistp.net> writes:
>> On Jan 4, 2019, at 11:34 PM, Toke Høiland-Jørgensen <toke@toke.dk> wrote:
>>
>> Pete Heist <pete@heistp.net> writes:
>>
>> This basically means that we can't use CAKE as a leaf qdisc with GSO
>> splitting as it stands currently. I *think* the solution is for CAKE to
>> notify its parents; could you try the patch below and see if it helps?
>
> Aha, good news. :)
>
> I’m probably not currently in a position to try it on my old kernels with the out of tree build:
>
> On 3.16.7:
Hmm, try this version for 3.16 - probably doesn't work on later kernels.
I'll look into a proper backport once you've confirmed that it works :)
-Toke
diff --git a/net/sched/sch_cake.c b/net/sched/sch_cake.c
index b910cd5c56f7..ef3acdbb8429 100644
--- a/net/sched/sch_cake.c
+++ b/net/sched/sch_cake.c
@@ -1617,6 +1617,44 @@ static u32 cake_classify(struct Qdisc *sch, struct cake_tin_data **t,
static void cake_reconfigure(struct Qdisc *sch);
+
+static struct Qdisc *qdisc_match_from_root(struct Qdisc *root, u32 handle)
+{
+ struct Qdisc *q;
+
+ if (!(root->flags & TCQ_F_BUILTIN) &&
+ root->handle == handle)
+ return root;
+
+ list_for_each_entry(q, &root->list, list) {
+ if (q->handle == handle)
+ return q;
+ }
+ return NULL;
+}
+
+void adjust_parent_qlen(struct Qdisc *sch, unsigned int n,
+ unsigned int len)
+{
+ u32 parentid;
+ if (n == 0 && len == 0)
+ return;
+ rcu_read_lock();
+ while ((parentid = sch->parent)) {
+ if (TC_H_MAJ(parentid) == TC_H_MAJ(TC_H_INGRESS))
+ break;
+
+ sch = qdisc_match_from_root(qdisc_dev(sch), TC_H_MAJ(parentid));
+ if (sch == NULL) {
+ WARN_ON_ONCE(parentid != TC_H_ROOT);
+ break;
+ }
+ sch->q.qlen += n;
+ sch->qstats.backlog += len;
+ }
+ rcu_read_unlock();
+}
+
static s32 cake_enqueue(struct sk_buff *skb, struct Qdisc *sch,
struct sk_buff **to_free)
{
@@ -1667,7 +1705,7 @@ static s32 cake_enqueue(struct sk_buff *skb, struct Qdisc *sch,
if (skb_is_gso(skb) && q->rate_flags & CAKE_FLAG_SPLIT_GSO) {
struct sk_buff *segs, *nskb;
netdev_features_t features = netif_skb_features(skb);
- unsigned int slen = 0;
+ unsigned int slen = 0, numsegs = 0;
segs = skb_gso_segment(skb, features & ~NETIF_F_GSO_MASK);
if (IS_ERR_OR_NULL(segs))
@@ -1684,6 +1722,7 @@ static s32 cake_enqueue(struct sk_buff *skb, struct Qdisc *sch,
sch->q.qlen++;
slen += segs->len;
+ numsegs++;
q->buffer_used += segs->truesize;
b->packets++;
segs = nskb;
@@ -1696,7 +1735,7 @@ static s32 cake_enqueue(struct sk_buff *skb, struct Qdisc *sch,
sch->qstats.backlog += slen;
q->avg_window_bytes += slen;
- qdisc_tree_reduce_backlog(sch, 1, len);
+ adjust_parent_qlen(sch, numsegs - 1, slen - len);
consume_skb(skb);
} else {
/* not splitting */
next prev parent reply other threads:[~2019-01-05 10:07 UTC|newest]
Thread overview: 47+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-12-27 23:30 Pete Heist
2018-12-28 12:58 ` Pete Heist
2018-12-28 21:22 ` Pete Heist
2018-12-28 22:07 ` Jonathan Morton
2018-12-28 22:42 ` Pete Heist
2019-01-04 21:34 ` Toke Høiland-Jørgensen
2019-01-04 22:10 ` Pete Heist
2019-01-04 22:12 ` Pete Heist
2019-01-04 22:34 ` Toke Høiland-Jørgensen
2019-01-05 5:58 ` Pete Heist
2019-01-05 10:06 ` Toke Høiland-Jørgensen [this message]
2019-01-05 10:59 ` Pete Heist
2019-01-05 11:06 ` Pete Heist
2019-01-05 11:18 ` Toke Høiland-Jørgensen
2019-01-05 11:26 ` Pete Heist
2019-01-05 11:35 ` Pete Heist
2019-01-05 12:38 ` Toke Høiland-Jørgensen
2019-01-05 12:51 ` Pete Heist
2019-01-05 13:10 ` Toke Høiland-Jørgensen
2019-01-05 13:20 ` Pete Heist
2019-01-05 13:35 ` Toke Høiland-Jørgensen
2019-01-05 15:34 ` Pete Heist
2019-01-05 15:52 ` Jonathan Morton
2019-01-05 16:32 ` Toke Høiland-Jørgensen
2019-01-05 19:27 ` Sebastian Moeller
2019-01-05 20:01 ` Pete Heist
2019-01-05 20:10 ` Toke Høiland-Jørgensen
2019-01-05 20:31 ` Pete Heist
2019-01-05 22:27 ` Toke Høiland-Jørgensen
2019-01-05 22:41 ` Pete Heist
2019-01-06 9:37 ` Pete Heist
2019-01-06 20:56 ` Toke Høiland-Jørgensen
2019-01-07 0:30 ` Pete Heist
2019-01-07 2:11 ` Dave Taht
2019-01-07 11:30 ` Toke Høiland-Jørgensen
2019-01-07 15:07 ` Pete Heist
2019-01-08 20:03 ` Pete Heist
2019-01-08 20:44 ` Dave Taht
2019-01-08 22:01 ` Pete Heist
2019-01-08 22:33 ` Dave Taht
2019-01-09 6:13 ` Pete Heist
2019-01-08 22:27 ` Toke Høiland-Jørgensen
2019-01-09 5:29 ` Pete Heist
2019-01-09 8:36 ` Toke Høiland-Jørgensen
2019-01-06 20:55 ` Toke Høiland-Jørgensen
2019-01-05 10:44 ` Jonathan Morton
2019-01-05 11:17 ` Toke Høiland-Jørgensen
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
List information: https://lists.bufferbloat.net/postorius/lists/cake.lists.bufferbloat.net/
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=87wonjxvss.fsf@toke.dk \
--to=toke@toke.dk \
--cc=cake@lists.bufferbloat.net \
--cc=pete@heistp.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox