netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Vernon Mauery <vernux@us.ibm.com>
To: Andi Kleen <andi@firstfloor.org>
Cc: netdev <netdev@vger.kernel.org>,
	LKML <linux-kernel@vger.kernel.org>,
	rt-users <linux-rt-users@vger.kernel.org>
Subject: Re: High contention on the sk_buff_head.lock
Date: Wed, 18 Mar 2009 14:07:32 -0700	[thread overview]
Message-ID: <49C16294.8050101@us.ibm.com> (raw)
In-Reply-To: <87prge1rhu.fsf@basil.nowhere.org>

Andi Kleen wrote:
> Vernon Mauery <vernux@us.ibm.com> writes:
>> So while this issue really hits -rt kernels hard, it has a real effect on
>> mainline kernels as well.  The contention of the spinlocks is amplified
>> when they get turned into rt-mutexes, which causes a double context switch.
> 
> The new adaptive spin heuristics that have been discussed some time
> ago didn't help? Unsurprisingly making locks a lot more expensive
> has drawbacks as you discovered.

Yes.  Well, while the adaptive spinlocks did great things for the
network throughput last time I tested them, they also didn't quite
give the determinism in other areas.  It would be nice to be able to
target a handful of trouble locks with adaptive spinlocks.

Even so, though I saw dramatic throughput increases with adaptive
spinlocks, they would still be bound by this same lock contention
that I am seeing when the locks are true spinlocks.

>>    &list->lock#3:      24517307       24643791           0.71        1286.62      56516392.42       34834296       44904018           0.60        164.79    31314786.02
>>     -------------
>>    &list->lock#3       15596927    [<ffffffff812474da>] dev_queue_xmit+0x2ea/0x468
> 
> The real "fix" would be probably to use a multi queue capable NIC
> and a NIC driver that sets up multiple queues for TX (normally they
> only do for RX). Then cores or a set of cores (often the number
> of cores is larger than the number of NIC queues) could avoid this
> problem. Disadvantage: more memory use.

Hmmm.  So do either the netxen_nic or bnx2x drivers support multiple
queues?  (that is the HW that I have access to right now).  And do I
need to do anything to set them up?

> But then again I'm not sure it's  worth it if the problem only
> happens in out of tree RT.

The effects of the high contention are not quite so pronounced in the
vanilla kernel, but I think we are still limited by this lock.  In the
-rt kernel, it is obvious that the lock contention is causing lots of
trouble.

--Vernon

  parent reply	other threads:[~2009-03-18 21:07 UTC|newest]

Thread overview: 40+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-03-18 17:24 High contention on the sk_buff_head.lock Vernon Mauery
2009-03-18 19:07 ` Eric Dumazet
2009-03-18 20:17   ` Vernon Mauery
2009-03-20 23:29     ` Jarek Poplawski
2009-03-23  8:32       ` Eric Dumazet
2009-03-23  8:37         ` David Miller
2009-03-23  8:50           ` Jarek Poplawski
2009-04-02 14:13           ` Herbert Xu
2009-04-02 14:15             ` Herbert Xu
2009-03-18 20:54 ` Andi Kleen
2009-03-18 21:03   ` David Miller
2009-03-18 21:10     ` Vernon Mauery
2009-03-18 21:38       ` David Miller
2009-03-18 21:49         ` Vernon Mauery
2009-03-19  1:02           ` David Miller
2009-03-18 21:54         ` Gregory Haskins
2009-03-19  1:03           ` David Miller
2009-03-19  1:13             ` Sven-Thorsten Dietrich
2009-03-19  1:17               ` David Miller
2009-03-19  1:43                 ` Sven-Thorsten Dietrich
2009-03-19  1:54                   ` David Miller
2009-03-19  5:49                     ` Eric Dumazet
2009-03-19  5:58                       ` David Miller
2009-03-19 14:04                         ` [PATCH] net: reorder struct Qdisc for better SMP performance Eric Dumazet
2009-03-20  8:33                           ` David Miller
2009-03-19 13:45                   ` High contention on the sk_buff_head.lock Andi Kleen
2009-03-19  3:48             ` Gregory Haskins
2009-03-19  5:38               ` David Miller
2009-03-19 12:42                 ` Gregory Haskins
2009-03-19 20:52                   ` David Miller
2009-03-19 12:50             ` Peter W. Morreale
2009-03-19  7:15           ` Evgeniy Polyakov
2009-03-18 21:07   ` Vernon Mauery [this message]
2009-03-18 21:45     ` Eilon Greenstein
2009-03-18 21:51       ` Vernon Mauery
2009-03-18 21:59         ` Andi Kleen
2009-03-18 22:19           ` Rick Jones
2009-03-19 12:59   ` Peter W. Morreale
2009-03-19 13:36     ` Peter W. Morreale
2009-03-19 13:46     ` Andi Kleen

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=49C16294.8050101@us.ibm.com \
    --to=vernux@us.ibm.com \
    --cc=andi@firstfloor.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-rt-users@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).