From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Paris Subject: Re: [Bug #14011] Kernel paging request failed in kmem_cache_alloc Date: Wed, 26 Aug 2009 11:08:21 -0400 Message-ID: <1251299301.2308.31.camel@dhcp231-106.rdu.redhat.com> References: <_yaHeGjHEzG.A.2bH.CtGlKB@chimera> <84144f020908252317j202aa459u27c40eed34debd6@mail.gmail.com> <200908261601.05734.ml_kernel@mortal-soul.de> <84144f020908260759q6554971dva6cc89a0ad9821b@mail.gmail.com> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <84144f020908260759q6554971dva6cc89a0ad9821b-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org> Sender: kernel-testers-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-ID: Content-Type: text/plain; charset="us-ascii" To: Pekka Enberg Cc: Matthias Dahl , "Rafael J. Wysocki" , Linux Kernel Mailing List , Kernel Testers List On Wed, 2009-08-26 at 17:59 +0300, Pekka Enberg wrote: > On Wed, Aug 26, 2009 at 5:01 PM, Matthias Dahl wrote: > >> Can you reproduce the bug without the proprietary nvidia module that > >> seems to be loaded? > > > > I am sorry but I forgot to test that and right now I am not very keen on > > trying again since this is my primary machine and I had quite some fs > > corruption (ext4 on md raid5 -> no barriers) the last times. :-( But this also > > happened w/o Xorg ever being run during that session (though naturally the > > nvidia kernel module was still loaded). > > Sure, I can understand that. The bug looks like regular slab > corruption which could have been caused the nvidia blob. So I think > the issue should be closed unless someone can reproduce it without the > blob. > > That said, sys_inotify_add_watch() also appears in the trace and > there's been quite a few bug fixes in that area recently so I guess we > should CC Eric Paris just in case the oops rings a bell to him. Nope, no bells here :( That slab cache is declared globally and allocated at __init time. I haven't seen any reports of writes off the ends of marks which might mess up a chache... sorry..... -Eric