[1868] in athena10
[grand.central.org #124531] OpenAFS 1.4.7 hangs in a kernel oops on reboot.
daemon@ATHENA.MIT.EDU (Derrick Brashear via RT)
Wed Mar 25 15:18:14 2009
Cc: debathena@mit.edu
MIME-Version: 1.0
In-Reply-To: <rt-124531@grand.central.org>
Content-Type: text/plain; charset="utf-8"
Reply-To: openafs-bugs@openafs.org
Message-ID: <rt-3.2.2-124531-413635-44.15.9865028417151@grand.central.org>
To: wdc@mit.edu
From: "Derrick Brashear via RT" <openafs-bugs@openafs.org>
Date: Wed, 25 Mar 2009 15:18:02 -0400 (EDT)
<URL: http://rt.central.org/rt/Ticket/Display.html?id=124531 >
> [guest - Tue Mar 24 19:12:22 2009]:
Try
STABLE14-shutdown-vcache-avoid-null-deref-20090324
i'd be curious to know if it has any affect without the other patch.
> This issue was logged under Ubuntu, and additional details
> can be found in:
> https://bugs.launchpad.net/bugs/333197
>
> We're deploying a new version of Athena based on Ubuntu,
> and we have a clever logout cleanup. Unfortunately when that
> clever mechanism fails, it reboots the system. So we're getting
> a lot more reboots than is normal. AFS hangs in an oops here.
>
> We have been running for a couple weeks the patch:
> /afs/andrew/usr/cg2v/cbr-only-free-what-you-alloc.diff
>
> Apparently that patch does not remedy the problem for the original
> reporter, but it DEFINITELY DOES remedy the problem we're having that
> looks like the same thing.
>
> I am opening this RT so that we can explicitly track the issue, and
> fully understand what the root cause is, what the most appropriate patch
> is, and to make sure that that patch becomes part of OpenAFS in the most
> efficient manner possible.
>
> I've been unable to ocr the images I took of the kernel oops trace,
> but I did use gimp to coalesce the images in the Ubuntu bug into one,
> easy to read image. I've attached that image to this case.
>
> How can I help move this issue forward?
>
> -Bill Cattey
>
>
