From: Eric Dumazet <eric.dumazet@gmail.com>
To: Stephen Hemminger <shemminger@vyatta.com>,
David Miller <davem@davemloft.net>
Cc: alexander.h.duyck@intel.com, netdev@vger.kernel.org
Subject: Re: [RFC PATCH] net: add additional lock to qdisc to increase enqueue/dequeue fairness
Date: Wed, 02 Jun 2010 11:48:28 +0200 [thread overview]
Message-ID: <1275472108.2725.142.camel@edumazet-laptop> (raw)
In-Reply-To: <1274463643.2439.473.camel@edumazet-laptop>
Le vendredi 21 mai 2010 à 19:40 +0200, Eric Dumazet a écrit :
> Le vendredi 21 mai 2010 à 18:44 +0200, Eric Dumazet a écrit :
>
> > We could use cmpxchg() and manipulate several bits at once in fast path.
> > ( __QDISC_STATE_RUNNING, __QDISC_STATE_LOCKED ... ) but making the crowd
> > of cpus spin on the same bits/cacheline than dequeue worker would
> > definitely slowdown the worker.
> >
> >
>
> Maybe I am missing something, but __QDISC_STATE_RUNNING is always
> manipulated with the lock held...
>
> We might avoid two atomic ops when changing this state (if moved to a
> separate container) in fast path (when a cpu sends only one packet and
> returns)
>
>
Here are two patches to implement this idea.
First patch to abstract QDISC_STATE_RUNNING access.
Second patch to add a __qstate container and remove the atomic ops.
next prev parent reply other threads:[~2010-06-02 9:48 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-03-23 20:25 [RFC PATCH] net: add additional lock to qdisc to increase enqueue/dequeue fairness Alexander Duyck
2010-03-23 20:31 ` David Miller
2010-03-24 2:12 ` Andi Kleen
2010-03-23 20:40 ` Rick Jones
2010-03-23 20:54 ` Eric Dumazet
2010-03-23 21:18 ` Eric Dumazet
2010-03-23 21:45 ` David Miller
2010-03-23 22:08 ` Duyck, Alexander H
2010-03-24 2:58 ` David Miller
2010-03-24 5:42 ` Eric Dumazet
2010-03-24 6:10 ` David Miller
2010-03-23 22:13 ` Eric Dumazet
2010-05-21 15:43 ` Stephen Hemminger
2010-05-21 16:44 ` Eric Dumazet
2010-05-21 17:38 ` Stephen Hemminger
2010-05-21 17:40 ` Eric Dumazet
2010-06-02 9:48 ` Eric Dumazet [this message]
2010-06-02 9:49 ` [PATCH 1/2 net-next-2.6] net: Define accessors to manipulate QDISC_STATE_RUNNING Eric Dumazet
2010-06-02 10:24 ` David Miller
2010-06-02 9:50 ` [PATCH 2/2 net-next-2.6] net: QDISC_STATE_RUNNING dont need atomic bit ops Eric Dumazet
2010-06-02 10:25 ` David Miller
2010-03-24 14:24 ` [RFC PATCH] net: add additional lock to qdisc to increase enqueue/dequeue fairness Eric Dumazet
2010-05-21 15:08 ` [PATCH] net: add additional lock to qdisc to increase throughput Eric Dumazet
2010-05-21 20:04 ` Duyck, Alexander H
2010-06-02 12:10 ` David Miller
2010-06-02 14:52 ` Eric Dumazet
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=1275472108.2725.142.camel@edumazet-laptop \
--to=eric.dumazet@gmail.com \
--cc=alexander.h.duyck@intel.com \
--cc=davem@davemloft.net \
--cc=netdev@vger.kernel.org \
--cc=shemminger@vyatta.com \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.