linux-kernel - Re: repeatable slab corruption with LTP msgctl08

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]

Message-Id: <20080612011537.6146c41d.akpm@linux-foundation.org>
Date:	Thu, 12 Jun 2008 01:15:37 -0700
From:	Andrew Morton <akpm@...ux-foundation.org>
To:	Nadia Derbey <Nadia.Derbey@...l.net>,
	Manfred Spraul <manfred@...orfullife.com>,
	linux-kernel@...r.kernel.org, linux-mm@...ck.org,
	Pekka Enberg <penberg@...helsinki.fi>,
	Christoph Lameter <clameter@....com>
Subject: Re: repeatable slab corruption with LTP msgctl08

On Thu, 12 Jun 2008 01:02:00 -0700 Andrew Morton <akpm@...ux-foundation.org> wrote:

> On Wed, 11 Jun 2008 23:34:49 -0700 Andrew Morton <akpm@...ux-foundation.org> wrote:
> 
> > On Wed, 11 Jun 2008 22:13:24 -0700 Andrew Morton <akpm@...ux-foundation.org> wrote:
> > 
> > > Running current mainline on my old 2-way PIII.  Distro is RH FC1.  LTP
> > > version is ltp-full-20070228 (lots of retro-computing there).
> > > 
> > > Config is at http://userweb.kernel.org/~akpm/config-vmm.txt
> > > 
> > > 
> > > ./testcases/bin/msgctl08 crashes after ten minutes or so:
> > 
> > ah, it runs to completion in about ten seconds on 2.6.25, so it'll be
> > easy for someone to bisect it.
> > 
> > What's that?  Sigh.  OK.  I wasn't doing anything much anyway.
> 
> Oh drat.  git-bisect tells me that this one-year-old msgctl08's
> execution time vastly increased when we added
> 
> commit f7bf3df8be72d98afa84f5ff183e14c1ba1e560d
> Author: Nadia Derbey <Nadia.Derbey@...l.net>
> Date:   Tue Apr 29 01:00:39 2008 -0700
> 
>     ipc: scale msgmni to the amount of lowmem
>     
> 
> But we already knew that, and LTP got changed to fix it.
> 
> So I was wrong in assuming that the long-execution-time correlates with
> the slab-corruption bug.
> 
> And the slab corruption bug takes half an hour to reproduce and an
> unknown amount of time to not-reproduce.  I don't think I'll be able to
> complete this before I disappear for over a week.
> 

Doing

	echo 16 > /proc/sys/kernel/msgmni

on the tree at the above bisection point
(f7bf3df8be72d98afa84f5ff183e14c1ba1e560d is most-recently-applied
patch), msgctl08 runs to completion in 15.7 seconds.

Doing the same thing on 2.6.26-rc5-mm3, msgctl08 also runs to
completion, in 20.9 seconds.  So

- it got slower

- the crash requires the huge value of /proc/sys/kernel/msgmni (1746)
  to reproduce.

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