From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id <S129758AbRB0Svk>; Tue, 27 Feb 2001 13:51:40 -0500
Received: (majordomo@vger.kernel.org) by vger.kernel.org
	id <S129766AbRB0Svb>; Tue, 27 Feb 2001 13:51:31 -0500
Received: from e21.nc.us.ibm.com ([32.97.136.227]:4843 "EHLO e21.nc.us.ibm.com")
	by vger.kernel.org with ESMTP id <S129758AbRB0SvL>;
	Tue, 27 Feb 2001 13:51:11 -0500
Date: Tue, 27 Feb 2001 13:50:10 -0500 (EST)
From: Richard A Nelson <cowboy@vnet.ibm.com>
X-X-Sender: <cowboy@badlands.lexington.ibm.com>
To: Alan Cox <alan@lxorguk.ukuu.org.uk>
cc: <linux-kernel@vger.kernel.org>
Subject: olympic (tokenring) & IPv6 (was Re: Linux 2.4.1ac20)
In-Reply-To: <Pine.LNX.4.33.0102231005220.2402-100000@badlands.lexington.ibm.com>
Message-ID: <Pine.LNX.4.33.0102271345460.2280-100000@badlands.lexington.ibm.com>
X-No-Markup: yes
x-No-ProductLinks: yes
x-No-Archive: yes
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: linux-kernel-owner@vger.kernel.org
X-Mailing-List: linux-kernel@vger.kernel.org

Problem recap:
  * modprobe oopses with null pointer right after loading olympic module
  * problem still occurs upto 2.4.2ac5

I've not been able to capture the oops due to other hardware problems,
but have narrowed the problem down:

 1) modprobe ipv6
    modprobe olympic
    ---> Oops & reboot

 2) modprobe olympic
    modprobe ipv6
    ---> still running (ok, so its only been 20 minutes, but hey that
         is *significantly* better than Oops on bootup)

> On Wed, 21 Feb 2001, Richard A Nelson wrote:
>
> > On Wed, 21 Feb 2001, Alan Cox wrote:
> >
> > >
> > > Can you stick an if(dev!=NULL) in front of that and let me know if that
> > > fixes it - just to verify thats the problem spot
> >
> > Compiling now, will reboot in the am (if it aint done soon) -
> > don't want it rebooting all night if this isn't it ;-}
>
> Sigh, it appears that the problem is actually neither of the above
> lines.
>
> I even tried 2.4.2 and had the same problem, then I noticed that 2.4.2
> also contains the same update ;-}
>
> I put an if(dev == NULL)printk...else in both spots and rebooted.
>
> Then (I gotta start drinking more coffee, or increase Copenhagen
> intake), I noticed that the actual oops is from modprobe!!!!
>
> I've not been able to get a oops trace because the error either causes
> a reboot, or hang (either case, e2fsck is run at reboot).
>
> Heres some system info that might be useful:
> Kernel modules         2.4.2
> Gnu C                  2.95.3
> Gnu Make               3.79.1
> Binutils               2.10.91.0.2
> Linux C Library        2.2.2
> Dynamic linker         ldd (GNU libc) 2.2.2
> Procps                 2.0.7
> Mount                  2.10s
> Net-tools              2.05
> Console-tools          0.2.3
> Sh-utils               2.0.11
>
> I'll be happy to reboot and capture the oops, if someone can help me
> make sure it gets logged (serial console not a possibility, the other
> box is currently in pieces).
>
>

-- 
Rick Nelson
Life'll kill ya                         -- Warren Zevon
Then you'll be dead                     -- Life'll kill ya