From: Yunsheng Lin <linyunsheng@huawei.com>
To: Hillf Danton <hdanton@sina.com>
Cc: Juergen Gross <jgross@suse.com>, <netdev@vger.kernel.org>,
<linux-kernel@vger.kernel.org>, Jiri Kosina <JKosina@suse.com>
Subject: Re: [PATCH net v3] net: sched: fix packet stuck problem for lockless qdisc
Date: Tue, 13 Apr 2021 10:56:42 +0800 [thread overview]
Message-ID: <57ace28a-93bb-6581-bbba-18d77a9871f7@huawei.com> (raw)
In-Reply-To: <20210413022129.2203-1-hdanton@sina.com>
On 2021/4/13 10:21, Hillf Danton wrote:
> On Mon, 12 Apr 2021 20:00:43 Yunsheng Lin wrote:
>>
>> Yes, the below patch seems to fix the data race described in
>> the commit log.
>> Then what is the difference between my patch and your patch below:)
>
> Hehe, this is one of the tough questions over a bounch of weeks.
>
> If a seqcount can detect the race between skb enqueue and dequeue then we
> cant see any excuse for not rolling back to the point without NOLOCK.
I am not sure I understood what you meant above.
As my understanding, the below patch is essentially the same as
your previous patch, the only difference I see is it uses qdisc->pad
instead of __QDISC_STATE_NEED_RESCHEDULE.
So instead of proposing another patch, it would be better if you
comment on my patch, and make improvement upon that.
>
> --- a/net/sched/sch_generic.c
> +++ b/net/sched/sch_generic.c
> @@ -632,6 +632,7 @@ static int pfifo_fast_enqueue(struct sk_
> return qdisc_drop(skb, qdisc, to_free);
> }
>
> + qdisc->pad++;
As has been mentioned:
Doing updating in pfifo_fast_enqueue() unconditionally does not
seems to be performance friendly, which is something my patch
tries to avoid as much as possible.
> qdisc_update_stats_at_enqueue(qdisc, pkt_len);
> return NET_XMIT_SUCCESS;
> }
> @@ -642,6 +643,7 @@ static struct sk_buff *pfifo_fast_dequeu
> struct sk_buff *skb = NULL;
> int band;
>
> + qdisc->pad = 0;
> for (band = 0; band < PFIFO_FAST_BANDS && !skb; band++) {
> struct skb_array *q = band2list(priv, band);
>
> --- a/include/net/sch_generic.h
> +++ b/include/net/sch_generic.h
> @@ -176,8 +176,12 @@ static inline bool qdisc_run_begin(struc
> static inline void qdisc_run_end(struct Qdisc *qdisc)
> {
> write_seqcount_end(&qdisc->running);
> - if (qdisc->flags & TCQ_F_NOLOCK)
> + if (qdisc->flags & TCQ_F_NOLOCK) {
> spin_unlock(&qdisc->seqlock);
> +
> + if (qdisc->pad != 0)
> + __netif_schedule(qdisc);
> + }
> }
>
> static inline bool qdisc_may_bulk(const struct Qdisc *qdisc)
>
> .
>
next prev parent reply other threads:[~2021-04-13 2:56 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-03-25 3:13 [PATCH net v3] net: sched: fix packet stuck problem for lockless qdisc Yunsheng Lin
2021-04-09 5:31 ` Juergen Gross
2021-04-12 1:04 ` Yunsheng Lin
2021-04-12 8:21 ` Juergen Gross
[not found] ` <20210409090909.1767-1-hdanton@sina.com>
2021-04-12 1:24 ` Yunsheng Lin
[not found] ` <20210412032111.1887-1-hdanton@sina.com>
2021-04-12 3:37 ` Yunsheng Lin
[not found] ` <20210412072856.2046-1-hdanton@sina.com>
2021-04-12 12:00 ` Yunsheng Lin
[not found] ` <20210413022129.2203-1-hdanton@sina.com>
2021-04-13 2:56 ` Yunsheng Lin [this message]
[not found] ` <20210413032620.2259-1-hdanton@sina.com>
2021-04-13 3:34 ` Yunsheng Lin
[not found] ` <20210413071241.2325-1-hdanton@sina.com>
2021-04-13 7:57 ` Yunsheng Lin
[not found] ` <20210413083352.2424-1-hdanton@sina.com>
2021-04-13 9:03 ` Yunsheng Lin
2021-04-13 9:15 ` Juergen Gross
2021-04-16 11:46 ` Jiri Kosina
2021-04-18 22:59 ` Michal Kubecek
2021-04-19 2:04 ` Yunsheng Lin
2021-04-19 12:21 ` [Linuxarm] " Yunsheng Lin
2021-04-19 15:25 ` Michal Kubecek
2021-04-19 14:57 ` Michal Kubecek
2021-04-20 1:45 ` Yunsheng Lin
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
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=57ace28a-93bb-6581-bbba-18d77a9871f7@huawei.com \
--to=linyunsheng@huawei.com \
--cc=JKosina@suse.com \
--cc=hdanton@sina.com \
--cc=jgross@suse.com \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
/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;
as well as URLs for NNTP newsgroup(s).