Re: intel graphic card hanging (Hangcheck timer elapsed... GPU hung)

Re: intel graphic card hanging (Hangcheck timer elapsed... GPU hung)

[Leandro Lucarella]

> Dear all,
> (please Cc)
> 
> I am running 2.6.34-rc2 + git 01e7770 commit (yesterday) with KMS on
> Debian/sid:
> xserver-xorg: 7.5+5
> video-intel: 2.9.1-3
> 
> Running D2X-xl (self compiled) (descent2 open source code) X suddenly
> died,
> and after that gdm could not be restarted because after a few secs
> it went dead again.
> 
> In the logs I see:
> kernel: [ 2486.341227] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck
> timer elapsed... GPU hung
> kernel: [ 2486.341240] render error detected, EIR: 0x00000000
> kernel: [ 2486.341278] [drm:i915_do_wait_request] *ERROR*
> i915_do_wait_request returns -5 (awaiting 113671 at 113604)
> kernel: [ 2486.669008] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck
> timer elapsed... GPU hung
> kernel: [ 2486.669012] render error detected, EIR: 0x00000000
> kernel: [ 2486.669522] [drm:i915_do_wait_request] *ERROR*
> i915_do_wait_request returns -5 (awaiting 113672 at 113604)
> kernel: [ 2487.524230] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck
> timer elapsed... GPU hung
> kernel: [ 2487.524243] render error detected, EIR: 0x00000000
> kernel: [ 2487.524284] [drm:i915_do_wait_request] *ERROR*
> i915_do_wait_request returns -5 (awaiting 113673 at 113604)
> kernel: [ 2487.912237] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck
> timer elapsed... GPU hung
> kernel: [ 2487.916175] render error detected, EIR: 0x00000000
> kernel: [ 2487.920137] [drm:i915_do_wait_request] *ERROR*
> i915_do_wait_request returns -5 (awaiting 113674 at 113604)
> Stopping gdm and restarting didn't help to get the graphics card back to
> normal.
> 
> If you have any ideas how to fix/track/debug that please let me know,
> same if you need more information.

Hi. I just wanted to say that I was hitting this bug too. It seems to be
a problem associated with xscreensaver and/or OpenGL. Looking for
information on the web I found some bug report that said something about
OpenGL (I can't find it now) so I tried deactivating xscreensaver and
the system is working very stable, no GPU hungs in 10 days of uptime.
When I had xscreensaver active the X server crashed in a couple of days
and sometime I got the GPU hung error in dmesg but X didn't crashed.

Thanks.

(please Cc)

-- 
Leandro Lucarella (AKA luca)                     http://llucax.com.ar/
----------------------------------------------------------------------
GPG Key: 5F5A8D05 (F8CD F9A7 BF00 5431 4145  104C 949E BFB6 5F5A 8D05)
----------------------------------------------------------------------
This homeless guy asked me for some money the other day.
And I was gonna give it to him but then I thought you're
just gonna use it on drugs or alcohol.
And then I thought, that's what I'm gonna use it on.
Why am I judging this poor bastard.

Re: intel graphic card hanging (Hangcheck timer elapsed... GPU hung)

[Takashi Iwai]

At Sun, 30 May 2010 21:44:50 -0300,
Leandro Lucarella wrote:
> 
> > Dear all,
> > (please Cc)
> > 
> > I am running 2.6.34-rc2 + git 01e7770 commit (yesterday) with KMS on
> > Debian/sid:
> > xserver-xorg: 7.5+5
> > video-intel: 2.9.1-3
> > 
> > Running D2X-xl (self compiled) (descent2 open source code) X suddenly
> > died,
> > and after that gdm could not be restarted because after a few secs
> > it went dead again.
> > 
> > In the logs I see:
> > kernel: [ 2486.341227] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck
> > timer elapsed... GPU hung
> > kernel: [ 2486.341240] render error detected, EIR: 0x00000000
> > kernel: [ 2486.341278] [drm:i915_do_wait_request] *ERROR*
> > i915_do_wait_request returns -5 (awaiting 113671 at 113604)
> > kernel: [ 2486.669008] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck
> > timer elapsed... GPU hung
> > kernel: [ 2486.669012] render error detected, EIR: 0x00000000
> > kernel: [ 2486.669522] [drm:i915_do_wait_request] *ERROR*
> > i915_do_wait_request returns -5 (awaiting 113672 at 113604)
> > kernel: [ 2487.524230] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck
> > timer elapsed... GPU hung
> > kernel: [ 2487.524243] render error detected, EIR: 0x00000000
> > kernel: [ 2487.524284] [drm:i915_do_wait_request] *ERROR*
> > i915_do_wait_request returns -5 (awaiting 113673 at 113604)
> > kernel: [ 2487.912237] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck
> > timer elapsed... GPU hung
> > kernel: [ 2487.916175] render error detected, EIR: 0x00000000
> > kernel: [ 2487.920137] [drm:i915_do_wait_request] *ERROR*
> > i915_do_wait_request returns -5 (awaiting 113674 at 113604)
> > Stopping gdm and restarting didn't help to get the graphics card back to
> > normal.
> > 
> > If you have any ideas how to fix/track/debug that please let me know,
> > same if you need more information.
> 
> Hi. I just wanted to say that I was hitting this bug too. It seems to be
> a problem associated with xscreensaver and/or OpenGL. Looking for
> information on the web I found some bug report that said something about
> OpenGL (I can't find it now) so I tried deactivating xscreensaver and
> the system is working very stable, no GPU hungs in 10 days of uptime.
> When I had xscreensaver active the X server crashed in a couple of days
> and sometime I got the GPU hung error in dmesg but X didn't crashed.

We've seen a similar hang with PineView when GLknots screensaver mode
is running.  Run /usr/lib/xscreensaver/glknots, and the screen gets
frozen immediately (mouse works, so the rendering simply doesn't
work), and no way to recover until reboot.

Reported at:
	https://bugs.freedesktop.org/show_bug.cgi?id=28274


thanks,

Takashi

Re: intel graphic card hanging (Hangcheck timer elapsed... GPU hung)

[Leandro Lucarella]

Takashi Iwai, el 31 de mayo a las 08:46 me escribiste:
> > > In the logs I see:
> > > kernel: [ 2486.341227] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck
> > > timer elapsed... GPU hung
> > > kernel: [ 2486.341240] render error detected, EIR: 0x00000000
> > > kernel: [ 2486.341278] [drm:i915_do_wait_request] *ERROR*
> > > i915_do_wait_request returns -5 (awaiting 113671 at 113604)
> > > kernel: [ 2486.669008] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck
> > > timer elapsed... GPU hung
> > > kernel: [ 2486.669012] render error detected, EIR: 0x00000000
> > > 
> > > If you have any ideas how to fix/track/debug that please let me know,
> > > same if you need more information.
> > 
> > Hi. I just wanted to say that I was hitting this bug too. It seems to be
> > a problem associated with xscreensaver and/or OpenGL. Looking for
> > information on the web I found some bug report that said something about
> > OpenGL (I can't find it now) so I tried deactivating xscreensaver and
> > the system is working very stable, no GPU hungs in 10 days of uptime.
> > When I had xscreensaver active the X server crashed in a couple of days
> > and sometime I got the GPU hung error in dmesg but X didn't crashed.
> 
> We've seen a similar hang with PineView when GLknots screensaver mode
> is running.  Run /usr/lib/xscreensaver/glknots, and the screen gets
> frozen immediately (mouse works, so the rendering simply doesn't
> work), and no way to recover until reboot.
> 
> Reported at:
> 	https://bugs.freedesktop.org/show_bug.cgi?id=28274

GLknots works fine (I tried a couple of times and I never got the GPU
hung message in dmesg). BTW, X crashes when I have the error (or
sometimes it gets rendering corruption, but continue working besides
that).


-- 
Leandro Lucarella (AKA luca)                     http://llucax.com.ar/
----------------------------------------------------------------------
GPG Key: 5F5A8D05 (F8CD F9A7 BF00 5431 4145  104C 949E BFB6 5F5A 8D05)
----------------------------------------------------------------------
All evidence has been buried
All tapes have been erased
But your footsteps give you away
So you're backtracking

Re: intel graphic card hanging (Hangcheck timer elapsed... GPU hung)

[Oliver Neukum]

Am Montag, 31. Mai 2010 02:44:50 schrieb Leandro Lucarella:
> Hi. I just wanted to say that I was hitting this bug too. It seems to be
> a problem associated with xscreensaver and/or OpenGL. Looking for
> information on the web I found some bug report that said something about
> OpenGL (I can't find it now) so I tried deactivating xscreensaver and
> the system is working very stable, no GPU hungs in 10 days of uptime.
> When I had xscreensaver active the X server crashed in a couple of days
> and sometime I got the GPU hung error in dmesg but X didn't crashed.

Please try the 3DKnots screensaver.

	Regards
		Oliver

intel graphic card hanging (Hangcheck timer elapsed... GPU hung)

[Norbert Preining]

Dear all,

(please Cc)

I am running 2.6.34-rc2 + git 01e7770 commit (yesterday) with KMS on Debian/sid:
xserver-xorg: 7.5+5
video-intel: 2.9.1-3

Running D2X-xl (self compiled) (descent2 open source code) X suddenly died,
and after that gdm could not be restarted because after a few secs
it went dead again.

In the logs I see:
kernel: [ 2486.341227] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
kernel: [ 2486.341240] render error detected, EIR: 0x00000000
kernel: [ 2486.341278] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 113671 at 113604)
kernel: [ 2486.669008] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
kernel: [ 2486.669012] render error detected, EIR: 0x00000000
kernel: [ 2486.669522] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 113672 at 113604)
kernel: [ 2487.524230] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
kernel: [ 2487.524243] render error detected, EIR: 0x00000000
kernel: [ 2487.524284] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 113673 at 113604)
kernel: [ 2487.912237] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
kernel: [ 2487.916175] render error detected, EIR: 0x00000000
kernel: [ 2487.920137] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 113674 at 113604)

Stopping gdm and restarting didn't help to get the graphics card back to normal.

If you have any ideas how to fix/track/debug that please let me know,
same if you need more information.

Best wishes

Norbert
------------------------------------------------------------------------
Norbert Preining            preining@{jaist.ac.jp, logic.at, debian.org}
JAIST, Japan            TU Wien, Austria           Debian TeX Task Force
DSA: 0x09C5B094   fp: 14DF 2E6C 0307 BE6D AD76  A9C0 D2BF 4AA3 09C5 B094
------------------------------------------------------------------------
HUCKNALL (vb.)
To crouch upwards: as in the movement of a seated person's feet and
legs made in order to allow a cleaner's hoover to pass beneath them.
			--- Douglas Adams, The Meaning of Liff

Re: intel graphic card hanging (Hangcheck timer elapsed... GPU hung)

[Maciej Rutecki]

On sobota, 27 marca 2010 o 17:11:05 Norbert Preining wrote:
> Dear all,
> 
> (please Cc)

[Added more CCs]

Similar:
http://marc.info/?l=linux-kernel&m=126954749618319&w=2

> 
> I am running 2.6.34-rc2 + git 01e7770 commit (yesterday) with KMS on
>  Debian/sid: xserver-xorg: 7.5+5
> video-intel: 2.9.1-3
> 
> Running D2X-xl (self compiled) (descent2 open source code) X suddenly died,
> and after that gdm could not be restarted because after a few secs
> it went dead again.
> 
> In the logs I see:
> kernel: [ 2486.341227] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
>  elapsed... GPU hung kernel: [ 2486.341240] render error detected, EIR:
>  0x00000000
> kernel: [ 2486.341278] [drm:i915_do_wait_request] *ERROR*
>  i915_do_wait_request returns -5 (awaiting 113671 at 113604) kernel: [
>  2486.669008] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
>  elapsed... GPU hung kernel: [ 2486.669012] render error detected, EIR:
>  0x00000000
> kernel: [ 2486.669522] [drm:i915_do_wait_request] *ERROR*
>  i915_do_wait_request returns -5 (awaiting 113672 at 113604) kernel: [
>  2487.524230] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
>  elapsed... GPU hung kernel: [ 2487.524243] render error detected, EIR:
>  0x00000000
> kernel: [ 2487.524284] [drm:i915_do_wait_request] *ERROR*
>  i915_do_wait_request returns -5 (awaiting 113673 at 113604) kernel: [
>  2487.912237] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
>  elapsed... GPU hung kernel: [ 2487.916175] render error detected, EIR:
>  0x00000000
> kernel: [ 2487.920137] [drm:i915_do_wait_request] *ERROR*
>  i915_do_wait_request returns -5 (awaiting 113674 at 113604)
> 
> Stopping gdm and restarting didn't help to get the graphics card back to
>  normal.
> 
> If you have any ideas how to fix/track/debug that please let me know,
> same if you need more information.
> 
> Best wishes
> 
> Norbert
> ------------------------------------------------------------------------
> Norbert Preining            preining@{jaist.ac.jp, logic.at, debian.org}
> JAIST, Japan            TU Wien, Austria           Debian TeX Task Force
> DSA: 0x09C5B094   fp: 14DF 2E6C 0307 BE6D AD76  A9C0 D2BF 4AA3 09C5 B094
> ------------------------------------------------------------------------
> HUCKNALL (vb.)
> To crouch upwards: as in the movement of a seated person's feet and
> legs made in order to allow a cleaner's hoover to pass beneath them.
> 			--- Douglas Adams, The Meaning of Liff
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 

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

Re: intel graphic card hanging (Hangcheck timer elapsed... GPU hung)

[Norbert Preining]

On Sa, 27 Mär 2010, Maciej Rutecki wrote:
> http://marc.info/?l=linux-kernel&m=126954749618319&w=2

Yes, I am using modeset, too.

> > I am running 2.6.34-rc2 + git 01e7770 commit (yesterday) with KMS on
> >  Debian/sid: xserver-xorg: 7.5+5
> > video-intel: 2.9.1-3
> > 
> > Running D2X-xl (self compiled) (descent2 open source code) X suddenly died,
> > and after that gdm could not be restarted because after a few secs
> > it went dead again.
> > 
> > In the logs I see:
> > kernel: [ 2486.341227] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> >  elapsed... GPU hung kernel: [ 2486.341240] render error detected, EIR:
> >  0x00000000
> > kernel: [ 2486.341278] [drm:i915_do_wait_request] *ERROR*
> >  i915_do_wait_request returns -5 (awaiting 113671 at 113604) kernel: [
> >  2486.669008] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> >  elapsed... GPU hung kernel: [ 2486.669012] render error detected, EIR:
> >  0x00000000
> > kernel: [ 2486.669522] [drm:i915_do_wait_request] *ERROR*
> >  i915_do_wait_request returns -5 (awaiting 113672 at 113604) kernel: [
> >  2487.524230] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> >  elapsed... GPU hung kernel: [ 2487.524243] render error detected, EIR:
> >  0x00000000
> > kernel: [ 2487.524284] [drm:i915_do_wait_request] *ERROR*
> >  i915_do_wait_request returns -5 (awaiting 113673 at 113604) kernel: [
> >  2487.912237] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer
> >  elapsed... GPU hung kernel: [ 2487.916175] render error detected, EIR:
> >  0x00000000
> > kernel: [ 2487.920137] [drm:i915_do_wait_request] *ERROR*
> >  i915_do_wait_request returns -5 (awaiting 113674 at 113604)
> > 
> > Stopping gdm and restarting didn't help to get the graphics card back to
> >  normal.
> > 
> > If you have any ideas how to fix/track/debug that please let me know,
> > same if you need more information.

Best wishes

Norbert
------------------------------------------------------------------------
Norbert Preining            preining@{jaist.ac.jp, logic.at, debian.org}
JAIST, Japan            TU Wien, Austria           Debian TeX Task Force
DSA: 0x09C5B094   fp: 14DF 2E6C 0307 BE6D AD76  A9C0 D2BF 4AA3 09C5 B094
------------------------------------------------------------------------
`How do you feel?' he asked him.
bits of me keep
passing out.' ....
`We're safe,' he said.
`Oh good,' said Arthur.
in one of the
spaceships of the Vogon Constructor Fleet.'
this is obviously some strange usage of
the word "safe" that I wasn't previously aware of.'
                 --- Arthur after his first ever teleport ride.
                 --- Douglas Adams, The Hitchhikers Guide to the Galaxy

Re: intel graphic card hanging (Hangcheck timer elapsed... GPU hung)

[Jesse Barnes]

On Sun, 28 Mar 2010 01:11:05 +0900
Norbert Preining <preining@logic.at> wrote:

> Dear all,
> 
> (please Cc)
> 
> I am running 2.6.34-rc2 + git 01e7770 commit (yesterday) with KMS on Debian/sid:
> xserver-xorg: 7.5+5
> video-intel: 2.9.1-3
> 
> Running D2X-xl (self compiled) (descent2 open source code) X suddenly died,
> and after that gdm could not be restarted because after a few secs
> it went dead again.
> 
> In the logs I see:
> kernel: [ 2486.341227] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
> kernel: [ 2486.341240] render error detected, EIR: 0x00000000
> kernel: [ 2486.341278] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 113671 at 113604)
> kernel: [ 2486.669008] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
> kernel: [ 2486.669012] render error detected, EIR: 0x00000000
> kernel: [ 2486.669522] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 113672 at 113604)
> kernel: [ 2487.524230] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
> kernel: [ 2487.524243] render error detected, EIR: 0x00000000
> kernel: [ 2487.524284] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 113673 at 113604)
> kernel: [ 2487.912237] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
> kernel: [ 2487.916175] render error detected, EIR: 0x00000000
> kernel: [ 2487.920137] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 113674 at 113604)
> 
> Stopping gdm and restarting didn't help to get the graphics card back to normal.
> 
> If you have any ideas how to fix/track/debug that please let me know,
> same if you need more information.

I was seeing something similar on my 945, can you try this patchset?
It seems to make things much more stable for me.

http://lists.freedesktop.org/archives/intel-gfx/2010-March/006387.html

There are 7 patches starting with that one.

-- 
Jesse Barnes, Intel Open Source Technology Center

Re: intel graphic card hanging (Hangcheck timer elapsed... GPU hung)

[Norbert Preining]

On Mo, 29 Mär 2010, Jesse Barnes wrote:
> I was seeing something similar on my 945, can you try this patchset?

Tried it with git kernel commit b72c409 and your patches, but
didn't change anything. Starting d2x-xl killed the X server:
[ 1067.920246] render error detected, EIR: 0x00000000
[ 1067.920293] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 48294 at 48216)
[ 1068.412095] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
[ 1068.412107] render error detected, EIR: 0x00000000
[ 1068.412135] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 48299 at 48216)
[ 1068.776229] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung

Anything else I could try?

Best wishes

Norbert
------------------------------------------------------------------------
Norbert Preining            preining@{jaist.ac.jp, logic.at, debian.org}
JAIST, Japan            TU Wien, Austria           Debian TeX Task Force
DSA: 0x09C5B094   fp: 14DF 2E6C 0307 BE6D AD76  A9C0 D2BF 4AA3 09C5 B094
------------------------------------------------------------------------
PAPPLE (vb.)
To do what babies do to soup with their spoons.
			--- Douglas Adams, The Meaning of Liff

Re: intel graphic card hanging (Hangcheck timer elapsed... GPU hung)

[Jesse Barnes]

On Tue, 30 Mar 2010 22:57:06 +0900
Norbert Preining <preining@logic.at> wrote:

> On Mo, 29 Mär 2010, Jesse Barnes wrote:
> > I was seeing something similar on my 945, can you try this patchset?
> 
> Tried it with git kernel commit b72c409 and your patches, but
> didn't change anything. Starting d2x-xl killed the X server:
> [ 1067.920246] render error detected, EIR: 0x00000000
> [ 1067.920293] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 48294 at 48216)
> [ 1068.412095] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
> [ 1068.412107] render error detected, EIR: 0x00000000
> [ 1068.412135] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 48299 at 48216)
> [ 1068.776229] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
> 
> Anything else I could try?

One of my 945 machines became pretty stable after increasing the
hancheck timeout from 75 to 750 jiffies in i915_drv.h.  But I have
reports that it doesn't work on all machines, so it's probably just
covering up a bug somewhere else.

-- 
Jesse Barnes, Intel Open Source Technology Center