Return-Path: <linux-kernel-owner+w=401wt.eu-S1755797AbZCRVHv@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1755797AbZCRVHv (ORCPT <rfc822;w@1wt.eu>);
	Wed, 18 Mar 2009 17:07:51 -0400
Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752681AbZCRVHi
	(ORCPT <rfc822;linux-kernel-outgoing>);
	Wed, 18 Mar 2009 17:07:38 -0400
Received: from e35.co.us.ibm.com ([32.97.110.153]:39706 "EHLO
	e35.co.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1752242AbZCRVHh (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Wed, 18 Mar 2009 17:07:37 -0400
Message-ID: <49C16294.8050101@us.ibm.com>
Date: Wed, 18 Mar 2009 14:07:32 -0700
From: Vernon Mauery <vernux@us.ibm.com>
User-Agent: Thunderbird 2.0.0.19 (X11/20090105)
MIME-Version: 1.0
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
References: <49C12E64.1000301@us.ibm.com> <87prge1rhu.fsf@basil.nowhere.org>
In-Reply-To: <87prge1rhu.fsf@basil.nowhere.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org
Content-Length: 2259
Lines: 47

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
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/