Re: [Bug #15004] i915: *ERROR* Execbuf while wedged

[Jesse Barnes]

On Mon, 8 Feb 2010 15:00:44 -0500
Chris Mason <chris.mason@xxxxxxxxxx> wrote:

> On Mon, Feb 08, 2010 at 09:35:51AM -0800, Jesse Barnes wrote:
> > > 
> > > Ok, updating to rc7 and updating my xf86 driver to 2.10 seems to
> > > have fixed up my suspend/resume problems.  So, I should be able
> > > to trigger the execbuf problem again.
> > 
> > I've heard some reports that the 2D driver introduces and fixes
> > hangs, so it's possible 2.10 will fix both issues for you.
> > 
> > > Well, the rebooting isn't a huge deal, but if there's something I
> > > can track/kick or force to core dump, would it help?
> > 
> > We do have a test in intel-gpu-tools that will instigate a hang,
> > but it would be best to figure out what's causing it in your
> > environment.
> > 
> 
> [63516.632060] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> elapsed... GPU hung [63516.632069] render error detected, EIR:
> 0x00000000 [63516.632092] [drm:i915_do_wait_request] *ERROR*
> i915_do_wait_request returns -5 (awaiting 2590581 at 2590579)
> 
> So, this popped up while surfing in firefox, which is usually where I
> hit the execbuf errors.  X was totally stuck afterwards, but I could
> switch to a vc and get the dmesg.
> 
> If there are specific procs that I can try to get traces of, just let
> me know for next time.

Hm, EIR is clear so this may be a failure of our hangcheck timer.

Chris Wilson saw these recently too; hoping he has ideas.

-- 
Jesse Barnes, Intel Open Source Technology Center
--
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/