From: Patrick McHardy <kaber-dcUjhNyLwpNeoWH0uzbU5w@public•gmane.org>
To: David Miller <davem-fT/PcQaiUtIeIZ0/mPfg9Q@public•gmane.org>
Cc: netdev-u79uwXL29TY76Z2rM5mHXA@public•gmane.org,
johannes-cdvu00un1VgdHxzADdlk8Q@public•gmane.org,
linux-wireless-u79uwXL29TY76Z2rM5mHXA@public•gmane.org
Subject: Re: [PATCH 20/31]: pkt_sched: Perform bulk of qdisc destruction in RCU.
Date: Fri, 18 Jul 2008 01:58:02 +0200 [thread overview]
Message-ID: <487FDC8A.9090101@trash.net> (raw)
In-Reply-To: <20080717.153609.142867331.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
David Miller wrote:
> From: Patrick McHardy <kaber-dcUjhNyLwpNeoWH0uzbU5w@public•gmane.org>
> Date: Thu, 17 Jul 2008 15:48:22 +0200
>
>> One thought that occured to me - we could avoid all the visiblity
>> issues wrt. dev->qdisc_list by simply getting rid of it :)
>>
>> If we move the qdisc list from the device to the root Qdisc itself,
>> it would become invisible automatically as soon as we assign a new
>> root qdisc to the netdev_queue. Iteration would become slightly
>> more complicated since we'd have to iterate over all netdev_queues,
>> but I think it should avoid most of the problems I mentioned
>> (besides the u32_list thing).
>
> What might make sense is to have a special Qdisc_root structure which
> is simply:
>
> struct Qdisc_root {
> struct Qdisc qd;
> struct list_head qdisc_list;
> };
>
> Everything about tree level synchronization would be type explicit.
Device level grafting is also explicit, so that looks like
a clean way.
> Yes, as you say, the qdisc iteration would get slightly ugly. But
> that doesn't seem to be a huge deal.
>
> But it seems a clean solution to the child qdisc visibility problem.
>
> About u32_list, that thing definitely needs some spinlock. The
> consultation of that list, and refcount mods, only occur during config
> operations. So it's not like we have to grab this lock in the data
> paths.
>
> If we really want to sweep this problem under the rug, there is another
> way. Have the qdisc_destroy() RCU handler kick off a workqueue, and
> grab the RTNL semaphore there during the final destruction calls. :-)
That would be the safe way. The RCU destruction used to cause us
bugs for at least two years, but I actually believe the Qdisc_root
thing will work :)
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public•gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
next prev parent reply other threads:[~2008-07-17 23:58 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-07-17 12:17 [PATCH 20/31]: pkt_sched: Perform bulk of qdisc destruction in RCU David Miller
[not found] ` <20080717.051726.226040470.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2008-07-17 13:03 ` Patrick McHardy
[not found] ` <487F4327.1000107-dcUjhNyLwpNeoWH0uzbU5w@public.gmane.org>
2008-07-17 13:12 ` David Miller
2008-07-17 13:48 ` Patrick McHardy
[not found] ` <487F4DA6.6010009-dcUjhNyLwpNeoWH0uzbU5w@public.gmane.org>
2008-07-17 22:36 ` David Miller
[not found] ` <20080717.153609.142867331.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2008-07-17 23:58 ` Patrick McHardy [this message]
2008-07-17 13:35 ` jamal
2008-07-17 14:02 ` Patrick McHardy
[not found] ` <487F50EC.80008-dcUjhNyLwpNeoWH0uzbU5w@public.gmane.org>
2008-07-17 22:24 ` David Miller
[not found] ` <20080717.152447.89672084.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2008-07-17 23:48 ` Patrick McHardy
2008-07-18 13:10 ` jamal
2008-07-18 13:27 ` jamal
2008-07-18 21:05 ` David Miller
2008-07-20 15:16 ` jamal
2008-07-20 17:25 ` David Miller
[not found] ` <20080720.102534.246150854.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2008-07-20 17:34 ` Tomas Winkler
[not found] ` <1ba2fa240807201034u1fbe3343tdcc858f7772d2b39-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2008-07-20 17:35 ` David Miller
2008-07-20 22:32 ` jamal
2008-07-20 23:59 ` David Miller
2008-07-21 2:20 ` jamal
2008-07-21 11:20 ` jamal
2008-07-21 16:45 ` David Miller
2008-07-21 11:58 ` Herbert Xu
[not found] ` <E1KKu2D-0002kd-00-XQvu0L+U/CjiRBuR/1fSEKKkPtS2pBon@public.gmane.org>
2008-07-21 13:08 ` jamal
2008-07-21 13:19 ` Herbert Xu
2008-07-21 13:56 ` jamal
2008-07-21 13:58 ` Herbert Xu
[not found] ` <20080721135846.GA11543-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2008-07-21 15:09 ` David Miller
[not found] ` <20080721.080901.214889979.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2008-07-21 15:22 ` Herbert Xu
2008-07-21 15:26 ` David Miller
[not found] ` <20080721.082627.128593651.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2008-07-21 16:16 ` Herbert Xu
[not found] ` <20080721161616.GA12938-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2008-07-21 16:25 ` David Miller
2008-07-21 16:43 ` Herbert Xu
2008-07-21 16:51 ` David Miller
[not found] ` <20080721.095124.89249903.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2008-07-21 17:02 ` Herbert Xu
[not found] ` <20080721170233.GA13417-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2008-07-21 17:08 ` David Miller
[not found] ` <20080721.100821.38432201.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2008-07-21 17:11 ` Herbert Xu
2008-08-22 6:56 ` Herbert Xu
2008-08-22 7:16 ` David Miller
[not found] ` <20080822.001620.141235108.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2008-08-22 7:41 ` Herbert Xu
[not found] ` <20080822074115.GB25615-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2008-08-22 10:42 ` David Miller
[not found] ` <20080822.034217.135152551.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2008-08-22 10:47 ` Herbert Xu
2008-08-22 13:52 ` jamal
[not found] ` <20080822065655.GA18471-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2008-08-22 13:43 ` jamal
2008-07-21 17:35 ` David Miller
2008-07-18 17:10 ` Roland Dreier
[not found] ` <adaljzz6ruq.fsf-FYB4Gu1CFyUAvxtiuMwx3w@public.gmane.org>
2008-07-20 14:58 ` jamal
2008-07-20 14:32 ` Herbert Xu
2008-07-20 17:20 ` David Miller
2008-07-20 14:20 ` Herbert Xu
[not found] ` <E1KKZmE-0000ty-00-XQvu0L+U/CjiRBuR/1fSEKKkPtS2pBon@public.gmane.org>
2008-07-20 15:35 ` jamal
2008-07-21 0:11 ` Herbert Xu
[not found] ` <20080721001119.GA6515-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2008-07-21 2:33 ` jamal
2008-07-21 3:17 ` Herbert Xu
2008-07-21 11:14 ` jamal
2008-07-21 11:36 ` Herbert Xu
[not found] ` <20080721113600.GA10322-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2008-07-21 11:39 ` jamal
2008-07-19 3:59 ` David Miller
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=487FDC8A.9090101@trash.net \
--to=kaber-dcujhnylwpneowh0uzbu5w@public$(echo .)gmane.org \
--cc=davem-fT/PcQaiUtIeIZ0/mPfg9Q@public$(echo .)gmane.org \
--cc=johannes-cdvu00un1VgdHxzADdlk8Q@public$(echo .)gmane.org \
--cc=linux-wireless-u79uwXL29TY76Z2rM5mHXA@public$(echo .)gmane.org \
--cc=netdev-u79uwXL29TY76Z2rM5mHXA@public$(echo .)gmane.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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