linux-ext4 - Re: EXT4: kernel BUG at fs/jbd2/commit.c:533!

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Wed, 02 Sep 2009 15:07:37 -0600
From:	Andreas Dilger <adilger@....com>
To:	Theodore Tso <tytso@....edu>
Cc:	Sachin Sant <sachinp@...ibm.com>, linux-ext4@...r.kernel.org
Subject: Re: EXT4: kernel BUG at fs/jbd2/commit.c:533!

On Sep 02, 2009  08:09 -0400, Theodore Ts'o wrote:
> On Wed, Sep 02, 2009 at 04:48:46PM +0530, Sachin Sant wrote:
> > While executing fsstress against ext4 with 2.6.31-rc8 on a
> > powerpc box, ran into the following Bug.
> >
> > ------------[ cut here ]------------
> > cpu 0x0: Vector: 700 (Program Check) at [c0000000f3ae3970]
> >    pc: c000000000280f28: .jbd2_journal_commit_transaction+0x8ac/0x1bfc
> >    lr: c000000000280e58: .jbd2_journal_commit_transaction+0x7dc/0x1bfc
> >    sp: c0000000f3ae3bf0
> >   msr: 8000000000029032
> >  current = 0xc0000000f315c7d0
> >  paca    = 0xc000000000b62600
> >    pid   = 23139, comm = kjournald2
> > kernel BUG at fs/jbd2/commit.c:533!
> 
> Hmm, that's this ASSERT:
> 
>      J_ASSERT(commit_transaction->t_nr_buffers <=
>               commit_transaction->t_outstanding_credits);
> 
> in jbd2_journal_commit_transaction().  That's not something I've ever
> seen before, nor have we made any changes in the jbd2 logic in quite
> some time that would could be related to that code.
> 
> Is it possible to extract out the contents of the commit_transaction
> structure?  In particular, the t_nr_buffers and t_outstanding_credits value?
> 
> At this point, I can't rule out a wild pointer corrupting the data
> structure, however improbable that sounds... 

Given the nature of the test being run (fsstress) it seems entirely
possible that there is some transaction which doesn't reserve enough
credits for the operation in some rare corner case?

Cheers, Andreas
--
Andreas Dilger
Sr. Staff Engineer, Lustre Group
Sun Microsystems of Canada, Inc.

--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html