kernel: Uhhuh. NMI received for unknown reason...

[Dave Airlie]

On Fri, 2008-03-14 at 08:25 -0700, darrell pfeifer wrote:
> 2008/3/10 "Jóhann B. Guðmundsson" <johannbg at hi.is>:
> > This has been going on of depending on kernels shipped with FC6-8 I
> >  noticed it...
> >  ( HW: Dell inspiron 6000 ) ( yes with the latest kernel FC8 )
> >
> >  Uhhuh. NMI received for unknown reason bX on CPU 0.
> >  You have some hardware problem, likely on the PCI bus.
> >  Dazed and confused, but trying to continue....
> >
> >  Can we fix this messages somehow it's kinda misleading, it
> >  suggest that the hw is failing when in 99% it's a buggy driver..
> >
> >  When I read "You have some hardware problem" I think
> >  my hardware is failing and I need to backup my data
> >  and replace the part that is failing....
> >
> >  Better would be..
> >
> >  Uhhuh. NMI received for unknown reason bX on CPU 0.
> >  I have been feed [ driver X ] buggy driver ...
> >  Dazed and confused, but trying to continue....
> >
> 
> I have a Dell Inspiron 9300 with a radeon card, using latest rawhide.
> 
> As I mentioned yesterday, I had seen these errors a few months ago but
> they went away.
> 
> Yesterday I had 3 NMI's in the space of about 4 hours. In each case, I
> had just selected a menu or submenu and it had just finished being
> drawn to the screen.
> 
> I turned off compiz and haven't had an NMI all morning. If you're
> using compiz you might want to give that a try.
> 
> That does tend to point to a buggy driver.

Could still be heat related, running the 3D engine may cause more heat
to be generated.. I've been running compiz on my Inspiron 6000 for the
last week and not seen any NMIs, but I've also not being using it for
long periods of time, as that machine has been relegated to being the
beside the couch machine...

Dave.


> 
> darrell
>