From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail202.messagelabs.com (mail202.messagelabs.com [216.82.254.227]) by kanga.kvack.org (Postfix) with SMTP id E9B4F6B02A8 for ; Thu, 29 Jul 2010 05:48:36 -0400 (EDT) Received: by qwk4 with SMTP id 4so62525qwk.14 for ; Thu, 29 Jul 2010 02:48:35 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: References: <20100727200804.2F40.A69D9226@jp.fujitsu.com> <20100728135850.7A92.A69D9226@jp.fujitsu.com> From: dave b Date: Thu, 29 Jul 2010 19:48:14 +1000 Message-ID: Subject: Re: PROBLEM: oom killer and swap weirdness on 2.6.3* kernels Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Sender: owner-linux-mm@kvack.org To: KOSAKI Motohiro Cc: David Rientjes , Hugh Dickins , linux-kernel@vger.kernel.org, linux-mm@kvack.org List-ID: On 29 July 2010 19:47, dave b wrote: > On 28 July 2010 17:14, dave b wrote: >> On 28 July 2010 15:06, KOSAKI Motohiro = wrote: >>>> On 27 July 2010 21:14, KOSAKI Motohiro wrote: >>>> >> On 27 July 2010 18:09, dave b wrote: >>>> >> > On 27 July 2010 16:09, KOSAKI Motohiro wrote: >>>> >> >>> > Do you mean the issue will be gone if disabling intel graphic= s? >>>> >> >>> It may be a general issue or it could just be specific :) >>>> >> > >>>> >> > I will try with the latest ubuntu and report how that goes (that = will >>>> >> > be using fairly new xorg etc.) it is likely to be hidden issue ju= st >>>> >> > with the intel graphics driver. However, my concern is that it is= n't - >>>> >> > and it is about how shared graphics memory is handled :) >>>> >> >>>> >> >>>> >> Ok my desktop still stalled and no oom killer was invoked when I ad= ded >>>> >> swap to a live-cd of 10.04 amd64. >>>> >> >>>> >> *Without* *swap* *on* - the oom killer was invoked - here is a copy= of it. >>>> > >>>> > This stack seems similar following bug. can you please try to disabl= e intel graphics >>>> > driver? >>>> > >>>> > https://bugzilla.kernel.org/show_bug.cgi?id=3D14933 >>>> >>>> Ok I am not sure how to do that :) >>>> I could revert the patch and see if it 'fixes' this :) >>> >>> Oops, no, revert is not good action. the patch is correct. >>> probably my explanation was not clear. sorry. >>> >>> I did hope to disable 'driver' (i.e. using vga), not disable the patch. >> >> Oh you mean in xorg, I will also blacklist the module. Sure that patch >> might not it but in 2.6.26 the problem isn't there :) > > Ok I re-tested with 2.6.26 and 2.6.34.1 > So I will describe what happens below: > > 2.6.26 - with xorg running > "Given I have a test file called a.out > =C2=A0And I can see Xorg > =C2=A0And I am using 2.6.26 > =C2=A0And I have swap on > =C2=A0When I run it I run a.out > =C2=A0Then I see the system freeze up slightly > =C2=A0And the hard drive churns( and the cpu is doing something as the > large fan kicks) > =C2=A0And after a while the system unfreezes" > > 2.6.26 - from single mode - before xorg starts and i915 is *not* loaded. > "Given I have a test file called a.out > =C2=A0And I cannot see Xorg > =C2=A0And I am using 2.6.26 > =C2=A0And I have swap on > =C2=A0When I run it I run a.out > =C2=A0Then I see the system freeze up > =C2=A0And the system fan doesn't spin any faster > =C2=A0And the system just sits idle" > > 2.6.34.1 > With and without xorg - WITH spam on the same behaviour as in the > 2.6.26 kernel appears (when xorg is not loaded). > > OOM attached from the 2.6.26 kernel when I used magic keys to invoke > the oom killer :) (this was on the 2.6.26 kernel - before i915 had > loaded and in single mode). s/spam/same/ -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org