Re: repeatable slab corruption with LTP msgctl08

[Andrew Morton]

On Thu, 12 Jun 2008 10:35:55 +0200 Andi Kleen <andi@xxxxxxxxxxxxxx> wrote:

> Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> writes:
> >
> > Doing the same thing on 2.6.26-rc5-mm3, msgctl08 also runs to
> > completion, in 20.9 seconds.  So
> >
> > - it got slower
> 
> That is because it scales itself to the number of available msg queues.
> So with Nadia's patch there are more and it runs slower.
> 
> In fact it seems to start one process per message queue, so perhaps it's 
> just running out of processes or something. Ok it should not segfault.
> 
> BTW a great way to debug slab corruptions with LTP faster is to run with
> a slab thrasher stress module like http://firstfloor.org/~andi/crasher-26.diff
> 

Something like that might be needed.  I ran it again and it took 49
minutes to crash.

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