From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hermann Pitton Subject: Re: [2.6.31] ir-kbd-i2c oops. Date: Wed, 30 Sep 2009 01:26:23 +0200 Message-ID: <1254266783.2657.11.camel@localhost> References: <200909160300.28382.pluto@agmk.net> <20090916085701.6e883600@hyperion.delvare> <200909161003.33090.pluto@agmk.net> <20090929161629.2a5c8d30@hyperion.delvare> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: <20090929161629.2a5c8d30-ig7AzVSIIG7kN2dkZ6Wm7A@public.gmane.org> Sender: linux-i2c-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Jean Delvare Cc: =?UTF-8?Q?Pawe=C5=82?= Sikora , linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, LMML List-Id: linux-i2c@vger.kernel.org Hi Jean, Am Dienstag, den 29.09.2009, 16:16 +0200 schrieb Jean Delvare: > On Wed, 16 Sep 2009 10:03:32 +0200, Pawe=C5=82 Sikora wrote: > > On Wednesday 16 September 2009 08:57:01 Jean Delvare wrote: > > > Hi Pawel, > > >=20 > > > I think this would be fixed by the following patch: > > > http://patchwork.kernel.org/patch/45707/ > >=20 > > still oopses. this time i've attached full dmesg. >=20 > Any news on this? Do you have a refined list of kernels which have th= e > bug and kernels which do not? Tried 2.6.32-rc1? Tried the v4l-dvb > repository? >=20 > Anyone else seeing this bug? I can see you ask the other way round, but just in case, I don't have that bug neither on some self compiled 2.6.30 with recent mercurial v4l-dvb on some outdated Fedora nor on a=20 Linux localhost.localdomain 2.6.30.5-43.fc11.x86_64 #1 SMP Thu Aug 27 21:39:52 EDT 2009 x86_64 x86_64 x86_64 GNU/Linux with my recently purchased older Pinnacle 310i. Hm, there are different versions of that card, to have it mentioned, obviously also with different remotes, and I can't tell how to identify them. > Your kernel stack trace doesn't look terribly reliable and I am not > able to come to any conclusion. The crash is supposed to happen in > ir_input_init(), but the stack trace doesn't lead there. I am also > skeptical about the +0x64/0x1a52, ir_input_init() is a rather small > function and I fail to see how it could be 6738 bytes in binary size. > Might be that the bug caused a stack corruption. Building a debug > kernel may help. Cheers, Hermann