2009-12-14 12:50:40

by Tarkan Erimer

Subject: [BUG] Linux-2.6.32-git10 drm_i915 error

Frequently, I hit this bug on Linux-2.6.32-git10. It's always
re-producible for me on my i945 video device equipped HP machine. When
hit this bug, it crashes the X server and screen goes black. But,
possible to switch to console. Here is the error log, I had in dmesg:

[ 2596.328132] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
elapsed... GPU hung
[ 2596.328146] render error detected, EIR: 0x00000000
[ 2596.328167] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request
returns -5 (awaiting 694588 at 694584)
[ 2596.439006] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
elapsed... GPU hung
[ 2596.439013] render error detected, EIR: 0x00000000

Tarkan

2010-01-10 14:21:04

by Maciej Rutecki

Subject: Re: [BUG] Linux-2.6.32-git10 drm_i915 error

2009/12/14 Tarkan Erimer <[email protected]>:
> Frequently, I hit this bug on Linux-2.6.32-git10. It's always
> re-producible for me on my i945 video device equipped HP machine. When
> hit this bug, it crashes the X server and screen goes black. But,
> possible to switch to console. Here is the error log, I had in dmesg:
>
> [ 2596.328132] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> elapsed... GPU hung
> [ 2596.328146] render error detected, EIR: 0x00000000
> [ 2596.328167] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request
> returns -5 (awaiting 694588 at 694584)
> [ 2596.439006] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> elapsed... GPU hung
> [ 2596.439013] render error detected, EIR: 0x00000000
>
> Tarkan

I confirm this. A observed this on 2.6.33-rc3-git3 after resume from suspend:

Jan 10 08:51:31 gumis kernel: [ 486.880045]
[drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU
hung
Jan 10 08:51:31 gumis kernel: [ 486.880063] render error detected,
EIR: 0x00000000
Jan 10 08:51:31 gumis kernel: [ 486.880107]
[drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (
awaiting 60107 at 60105)
Jan 10 08:51:32 gumis kernel: [ 487.276302] [drm:i915_gem_execbuffer]
*ERROR* i915_gem_do_execbuffer returns -5
Jan 10 08:51:32 gumis kernel: [ 487.276799] [drm:i915_gem_execbuffer]
*ERROR* i915_gem_do_execbuffer returns -5
Jan 10 08:51:32 gumis kernel: [ 487.277230] [drm:i915_gem_execbuffer]
*ERROR* i915_gem_do_execbuffer returns -5
Jan 10 08:51:32 gumis kernel: [ 487.278308] [drm:i915_gem_execbuffer]
*ERROR* i915_gem_do_execbuffer returns -5
Jan 10 08:51:32 gumis kernel: [ 487.278533] [drm:i915_gem_execbuffer]
*ERROR* i915_gem_do_execbuffer returns -5

2.6.33-rc2 works fine.
--
Maciej Rutecki
http://www.maciek.unixy.pl

2010-01-10 18:33:18

by Rafael J. Wysocki

Subject: Re: [BUG] Linux-2.6.32-git10 drm_i915 error

On Sunday 10 January 2010, Maciej Rutecki wrote:
> 2009/12/14 Tarkan Erimer <[email protected]>:
> > Frequently, I hit this bug on Linux-2.6.32-git10. It's always
> > re-producible for me on my i945 video device equipped HP machine. When
> > hit this bug, it crashes the X server and screen goes black. But,
> > possible to switch to console. Here is the error log, I had in dmesg:
> >
> > [ 2596.328132] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> > elapsed... GPU hung
> > [ 2596.328146] render error detected, EIR: 0x00000000
> > [ 2596.328167] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request
> > returns -5 (awaiting 694588 at 694584)
> > [ 2596.439006] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> > elapsed... GPU hung
> > [ 2596.439013] render error detected, EIR: 0x00000000
> >
> > Tarkan
>
> I confirm this. A observed this on 2.6.33-rc3-git3 after resume from suspend:
>
> Jan 10 08:51:31 gumis kernel: [ 486.880045]
> [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU
> hung
> Jan 10 08:51:31 gumis kernel: [ 486.880063] render error detected,
> EIR: 0x00000000
> Jan 10 08:51:31 gumis kernel: [ 486.880107]
> [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (
> awaiting 60107 at 60105)
> Jan 10 08:51:32 gumis kernel: [ 487.276302] [drm:i915_gem_execbuffer]
> *ERROR* i915_gem_do_execbuffer returns -5
> Jan 10 08:51:32 gumis kernel: [ 487.276799] [drm:i915_gem_execbuffer]
> *ERROR* i915_gem_do_execbuffer returns -5
> Jan 10 08:51:32 gumis kernel: [ 487.277230] [drm:i915_gem_execbuffer]
> *ERROR* i915_gem_do_execbuffer returns -5
> Jan 10 08:51:32 gumis kernel: [ 487.278308] [drm:i915_gem_execbuffer]
> *ERROR* i915_gem_do_execbuffer returns -5
> Jan 10 08:51:32 gumis kernel: [ 487.278533] [drm:i915_gem_execbuffer]
> *ERROR* i915_gem_do_execbuffer returns -5
>
> 2.6.33-rc2 works fine.

http://bugzilla.kernel.org/show_bug.cgi?id=15026 perhaps?

Rafael


>

2010-01-10 19:37:11

by Constantin Baranov

Subject: Re: [BUG] Linux-2.6.32-git10 drm_i915 error

On Sun, 10 Jan 2010 19:33:39 +0100 "Rafael J. Wysocki" <[email protected]> wrote:
> On Sunday 10 January 2010, Maciej Rutecki wrote:
> > 2009/12/14 Tarkan Erimer <[email protected]>:
> > > Frequently, I hit this bug on Linux-2.6.32-git10. It's always
> > > re-producible for me on my i945 video device equipped HP machine. When
> > > hit this bug, it crashes the X server and screen goes black. But,
> > > possible to switch to console. Here is the error log, I had in dmesg:
> > >
> > > [ 2596.328132] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> > > elapsed... GPU hung
> > > [ 2596.328146] render error detected, EIR: 0x00000000
> > > [ 2596.328167] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request
> > > returns -5 (awaiting 694588 at 694584)
> > > [ 2596.439006] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> > > elapsed... GPU hung
> > > [ 2596.439013] render error detected, EIR: 0x00000000
> > >
> > > Tarkan
> >
> > I confirm this. A observed this on 2.6.33-rc3-git3 after resume from suspend:
> >
> > Jan 10 08:51:31 gumis kernel: [ 486.880045]
> > [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU
> > hung
> > Jan 10 08:51:31 gumis kernel: [ 486.880063] render error detected,
> > EIR: 0x00000000
> > Jan 10 08:51:31 gumis kernel: [ 486.880107]
> > [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (
> > awaiting 60107 at 60105)
> > Jan 10 08:51:32 gumis kernel: [ 487.276302] [drm:i915_gem_execbuffer]
> > *ERROR* i915_gem_do_execbuffer returns -5
> > Jan 10 08:51:32 gumis kernel: [ 487.276799] [drm:i915_gem_execbuffer]
> > *ERROR* i915_gem_do_execbuffer returns -5
> > Jan 10 08:51:32 gumis kernel: [ 487.277230] [drm:i915_gem_execbuffer]
> > *ERROR* i915_gem_do_execbuffer returns -5
> > Jan 10 08:51:32 gumis kernel: [ 487.278308] [drm:i915_gem_execbuffer]
> > *ERROR* i915_gem_do_execbuffer returns -5
> > Jan 10 08:51:32 gumis kernel: [ 487.278533] [drm:i915_gem_execbuffer]
> > *ERROR* i915_gem_do_execbuffer returns -5
> >
> > 2.6.33-rc2 works fine.
>
> http://bugzilla.kernel.org/show_bug.cgi?id=15026 perhaps?

I have caught this bug in 2.6.31.9, 2.6.32.2, 2.6.33-rc3 and 2.6.33-rc3-git3
on my Acer Aspire 1410T with Intel GM45 chipset. It is not needed to perform
suspend and resume to reproduce the bug. The bug appears for me at startup
of the Stellarium (trunk version).

2010-01-10 20:13:26

by Maciej Rutecki

Subject: Re: [BUG] Linux-2.6.32-git10 drm_i915 error

2010/1/10 Rafael J. Wysocki <[email protected]>:

>
> http://bugzilla.kernel.org/show_bug.cgi?id=15026 perhaps?

Yes, at now Linux works fine. No problems after resume from suspend (disk/ram).

Regards
--
Maciej Rutecki
http://www.maciek.unixy.pl

2010-01-10 20:41:23

by Rafael J. Wysocki

Subject: Re: [BUG] Linux-2.6.32-git10 drm_i915 error

On Sunday 10 January 2010, Constantin Baranov wrote:
> On Sun, 10 Jan 2010 19:33:39 +0100 "Rafael J. Wysocki" <[email protected]> wrote:
> > On Sunday 10 January 2010, Maciej Rutecki wrote:
> > > 2009/12/14 Tarkan Erimer <[email protected]>:
> > > > Frequently, I hit this bug on Linux-2.6.32-git10. It's always
> > > > re-producible for me on my i945 video device equipped HP machine. When
> > > > hit this bug, it crashes the X server and screen goes black. But,
> > > > possible to switch to console. Here is the error log, I had in dmesg:
> > > >
> > > > [ 2596.328132] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> > > > elapsed... GPU hung
> > > > [ 2596.328146] render error detected, EIR: 0x00000000
> > > > [ 2596.328167] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request
> > > > returns -5 (awaiting 694588 at 694584)
> > > > [ 2596.439006] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> > > > elapsed... GPU hung
> > > > [ 2596.439013] render error detected, EIR: 0x00000000
> > > >
> > > > Tarkan
> > >
> > > I confirm this. A observed this on 2.6.33-rc3-git3 after resume from suspend:
> > >
> > > Jan 10 08:51:31 gumis kernel: [ 486.880045]
> > > [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU
> > > hung
> > > Jan 10 08:51:31 gumis kernel: [ 486.880063] render error detected,
> > > EIR: 0x00000000
> > > Jan 10 08:51:31 gumis kernel: [ 486.880107]
> > > [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (
> > > awaiting 60107 at 60105)
> > > Jan 10 08:51:32 gumis kernel: [ 487.276302] [drm:i915_gem_execbuffer]
> > > *ERROR* i915_gem_do_execbuffer returns -5
> > > Jan 10 08:51:32 gumis kernel: [ 487.276799] [drm:i915_gem_execbuffer]
> > > *ERROR* i915_gem_do_execbuffer returns -5
> > > Jan 10 08:51:32 gumis kernel: [ 487.277230] [drm:i915_gem_execbuffer]
> > > *ERROR* i915_gem_do_execbuffer returns -5
> > > Jan 10 08:51:32 gumis kernel: [ 487.278308] [drm:i915_gem_execbuffer]
> > > *ERROR* i915_gem_do_execbuffer returns -5
> > > Jan 10 08:51:32 gumis kernel: [ 487.278533] [drm:i915_gem_execbuffer]
> > > *ERROR* i915_gem_do_execbuffer returns -5
> > >
> > > 2.6.33-rc2 works fine.
> >
> > http://bugzilla.kernel.org/show_bug.cgi?id=15026 perhaps?
>
> I have caught this bug in 2.6.31.9, 2.6.32.2, 2.6.33-rc3 and 2.6.33-rc3-git3
> on my Acer Aspire 1410T with Intel GM45 chipset. It is not needed to perform
> suspend and resume to reproduce the bug. The bug appears for me at startup
> of the Stellarium (trunk version).

The message may be the same, but the reason is different.

Rafael