lkml.org 
[lkml]   [2011]   [Aug]   [30]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: Linux 3.1-rc4
On Tue, Aug 30, 2011 at 11:14:30AM +0200, Peter Zijlstra wrote:
> On Tue, 2011-08-30 at 09:07 +0200, Borislav Petkov wrote:
> > On Sun, Aug 28, 2011 at 09:52:43PM -0700, Linus Torvalds wrote:
> > > We're back on track on weekly -rc releases.
> >
> > Peter,
> >
> > I get the following lockdep splat on -rc4, it looks like load balancing
> > something.
> >
> > [ 5183.478132] ------------[ cut here ]------------
> > [ 5183.478155] WARNING: at kernel/lockdep.c:690 __lock_acquire+0x189/0x916()
>
> that's:
>
> look_up_lock_class():
> WARN_ON_ONCE(class->name != lock->name);
>
> That's something that really shouldn't happen and my first guess is that
> you're suffering some memory corruption.

Hmm, strange. That was one of Linus' suggestions last time I triggered
something similar in look_up_lock_class(). (Thanks Yong, I'd completely
forgotten about it). The old issue was a list corruption at the same
spot and this one looks similar except the list element is there to not
trap on the memory access but it triggers on the warning because it is
obviously a different element.

And the fact that I'd completely forgotten about it is that it hasn't
reappeared since then (v2.6.37-rc5-62-g6313e3c) which fits with the
pattern of a memory corruption.

Anyway, this is just FYI in case someone else triggers it too, I'll be
on the lookout for other strange things cooking up while the machine is
running.

Thanks.

--
Regards/Gruss,
Boris.



\
 
 \ /
  Last update: 2011-08-30 12:05    [W:2.913 / U:0.388 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site