Re: corruption happening sometimes when I do a shutdown -r now

[Eric Sandeen <sandeen redhat com>]

Christian Kujau wrote:
> On Sun, 21 Dec 2008, shirish wrote:
>> $ uname -a
>> Linux shirish-desktop 2.6.27-11-generic #1 SMP Fri Dec 19 16:29:52 UTC
>> 2008 i686 GNU/Linux
>>
>>> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git&a=commitdiff&h=72ed3d035855841ad611ee48b20909e9619d4a79
>>> fixed one cause of this.  Unless you can reproduce on an upstream kernel
>>> you might bug Ubuntu about it.
> 
> Hm, this commit is from 02/2007 and his Ubuntu kernel says 2.6.27 where 
> this particular patch has been included already. Are there any other 
> commits which may be able to fix this?

Hmm good point, it probably is in there.

>> [ 158.989493] EXT3-fs error (device sdb7): ext3_free_blocks_sb: bit 
>> already cleared for block 493569
> 
> ...could bad RAM be the reason for such a "hickup" too?

It's always worth testing ram I suppose :)

If Ubuntu supplies packaged unpatched upstream kernels for testing
that'd be worth it too, or hand-compiling an upstream kernel perhaps.  I
just don't know what's in the Ubuntu kernel, and I've not seen similar
reports upstream or in Fedora; the problem could be unique to his
hardware, his usecase, kernel, or who knows.  Reproducing a problem on a
stock, known kernel will get more people interested.  :)

-Eric (idly wondering if we could ask Arjan to track things like ext3
shutdowns on kerneloops.org in addition to bona-fide oopses, to look for
the prevalence of such things on various kernels...)