xfs
[Top] [All Lists]

Re: Crash with 3.8.3 and TuxOnIce

To: Pedro Ribeiro <pedrib@xxxxxxxxx>
Subject: Re: Crash with 3.8.3 and TuxOnIce
From: Dave Chinner <david@xxxxxxxxxxxxx>
Date: Wed, 20 Mar 2013 11:44:05 +1100
Cc: xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <CAEDdjHeo3Giv3Y8pLfYHSbJi07_dv7ywzinCSSkxUwBQe-1fsw@xxxxxxxxxxxxxx>
References: <CAEDdjHeo3Giv3Y8pLfYHSbJi07_dv7ywzinCSSkxUwBQe-1fsw@xxxxxxxxxxxxxx>
User-agent: Mutt/1.5.21 (2010-09-15)
On Tue, Mar 19, 2013 at 11:31:16PM +0000, Pedro Ribeiro wrote:
> Hi,
> 
> I'm using a TuxOnIce enabled kernel, and recently moved from 3.7.1 to
> 3.8.3. The former used to work nicely, but now I get a hard crash when I
> try to hibernate.
> This is solely related to TuxOnIce as it does not happen in the default
> hibernation.
> 
> I know this is an unsupported out of tree patch, but can you please help me
> debug it or point in the right direction?
> 
> Unfortunately as this was a kernel crash all I have are crappy photos, but
> they do show a readable stack trace:
> img18[dot]imageshack[dot]us/img18/395/imag0228hp[dot]jpg
> img13[dot]imageshack[dot]us/img13/4375/imag0231p[dot]jpg
> 
> (I removed the HTTP part and the dots off the links else it would bounce
> back from the list)

The list doesn't bounce URLs. Please make then clicky in future.

As it is, your kernel has oopsed in submit_bio() when writing a log
buffer writing a dummy log record.

The xfs_log_worker is trying to cover the log, and it seems like the
IO subsystem below it is doing something wrong. Perhaps Tux-on-Ice
is killing the IO subsystem without having first stopped all the
async filesystem processing first?

Looks like a Tux-on-ICE problem to me...

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx

<Prev in Thread] Current Thread [Next in Thread>