Re: gen6 (SNB) depthbuffer issue with OpenGL games

Re: gen6 (SNB) depthbuffer issue with OpenGL games

[Nicolas Kalkhof]

[-- Attachment #1: Type: text/html, Size: 2376 bytes --]

[-- Attachment #2: Type: text/plain, Size: 159 bytes --]

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: gen6 (SNB) depthbuffer issue with OpenGL games

[Safety0ff]

On 19/07/11 07:18 PM, Nicolas Kalkhof wrote:
> Hi,
> absolutely correct! I've experienced this issue the same way you
> described during the 2.6.39-rc series in random occurance but much
> less intensive. however it disappeared during the 3.0-rc series. now
> the depth/alpha issue occurs instantly after starting the OGL app.
> I'll try to nail down the bad commit.
>
> regards,
> nic
Now that I think of it, I believe my ubuntu 32bit installation running
2.6.38 from xorg-edgers was also affected.

Regards,
-Safety0ff

Re: gen6 (SNB) depthbuffer issue with OpenGL games

[Ian Romanick]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 07/20/2011 01:24 AM, Nicolas Kalkhof wrote:
> Hi Ian,
> 
> ok I've definately nailed the issue down to the "i915_enable_rc6" parameter in i915_drv.c
> Someone disabled this parameter and that causes the depthbuffer issue. I've enabled the switch by setting i915_enable_rc6=1; and the problem dissapeared.
> 
> glxinfo | grep "OpenGL renderer" shows:
> OpenGL renderer string: Mesa DRI Intel(R) Sandybridge Mobile
> 
> lspci -vn | grep VGA yields:
> 00:02.0 0300: 8086:0126 (rev 09) (prog-if 00 [VGA controller])
> 
> My CPU is a SNB I72620M with a 3000 graphics. No other GPU present.
> 
> Stupid question: Can I file a bug report on https://bugs.freedesktop.org or on  https://bugzilla.kernel.org/

bugzilla.kernel.org is the right place.  Please include all of the
information from this e-mail and the image showing the corruption.

> -----Ursprüngliche Nachricht-----
> Von: "Ian Romanick" <idr@freedesktop.org>
> Gesendet: Jul 19, 2011 11:08:14 PM
> An: "Nicolas Kalkhof" <nkalkhof@web.de>
> Betreff: Re: [Intel-gfx] gen6 (SNB) depthbuffer issue with OpenGL games
> 
> On 07/19/2011 01:21 PM, Nicolas Kalkhof wrote:
>>>> Hi all,
>>>>
>>>> ok I've nailed the issue down to 3.0.0-rc7 and 3.0.0-rc7-git1. I suspect that the changes made in
>>>> drivers/gpu/drm/i915/i915_dma.c are the cause of the problem.
>>>>
>>>> http://www.kernel.org/diff/diffview.cgi?file=%2Fpub%2Flinux%2Fkernel%2Fv3.0%2Fsnapshots%2Fpatch-3.0-rc7-git1.bz2;z=14
>>>>
>>>> Any Clues?
> 
> Okay, so that's the commit below, which changes some error clean-up
> paths. That is also odd. What *exact* GPU do you have? Specificially,
> what's the output of
> 
> glxinfo | grep "OpenGL renderer"
> 
> and
> 
> lspci -vn | grep VGA
> 
> Does this appear in all games or just certain games? If it's just in
> certain games, which ones?
> 
> commit a7b85d2aa63ed09cd5a4a640772b3272f5ac7caa
> Author: Keith Packard <keithp@keithp.com>
> Date: Sun Jul 10 13:12:17 2011 -0700
> 
> drm/i915: Clean up i915_driver_load failure path
> 
> i915_driver_load adds a write-combining MTRR region for the GTT
> aperture to improve memory speeds through the aperture. If
> i915_driver_load fails after this, it would not have cleaned up the
> MTRR. This shouldn't cause any problems, except for consuming an MTRR
> register. Still, it's best to clean up completely in the failure path,
> which is easily done by calling mtrr_del if the mtrr was successfully
> allocated.
> 
> i915_driver_load calls i915_gem_load which register
> i915_gem_inactive_shrink. If i915_driver_load fails after calling
> i915_gem_load, the shrinker will be left registered. When called, it
> will access freed memory and crash. The fix is to unregister the
> shrinker in the
> failure path using code duplicated from i915_driver_unload.
> 
> i915_driver_load also has some incorrect gotos in the error cleanup
> paths:
> 
> * After failing to initialize the GTT (which cannot happen, btw,
> intel_gtt_get returns a fixed (non-NULL) value), it tries to
> free the uninitialized WC IO mapping. Fixed this by changing the
> target from out_iomapfree to out_rmmap
> 
> Signed-off-by: Keith Packard <keithp@keithp.com>
> Tested-by: Lin Ming <ming.m.lin@intel.com>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/

iEYEARECAAYFAk4m6yAACgkQX1gOwKyEAw/6MQCgjs/YWI3rpwN8XsgHy/rwuq5P
c84AnjsjBjudlE9QZBuLFhuZgW+giw+/
=eJ0i
-----END PGP SIGNATURE-----
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: gen6 (SNB) depthbuffer issue with OpenGL games

[Nicolas Kalkhof]

Hi all,

ok I've nailed the issue down to 3.0.0-rc7 and 3.0.0-rc7-git1. I suspect that the changes made in
drivers/gpu/drm/i915/i915_dma.c are the cause of the problem.

http://www.kernel.org/diff/diffview.cgi?file=%2Fpub%2Flinux%2Fkernel%2Fv3.0%2Fsnapshots%2Fpatch-3.0-rc7-git1.bz2;z=14

Any Clues?

Regards
Nic


-----Ursprüngliche Nachricht-----
Von: "Nicolas Kalkhof" <nkalkhof@web.de>
Gesendet: Jul 19, 2011 9:08:26 PM
An: "Ian Romanick" <idr@freedesktop.org>
Betreff: Re: [Intel-gfx] gen6 (SNB) depthbuffer issue with OpenGL games

>Hi Ian,
>
>thx for your reply. ok I give it a try and will report my findings....
>
>regards
>nic
>
>-----Ursprüngliche Nachricht-----
>Von: "Ian Romanick" <idr@freedesktop.org>
>Gesendet: Jul 19, 2011 8:45:53 PM
>An: "Nicolas Kalkhof" <nkalkhof@web.de>
>Betreff: Re: [Intel-gfx] gen6 (SNB) depthbuffer issue with OpenGL games
>
>>-----BEGIN PGP SIGNED MESSAGE-----
>>Hash: SHA1
>>
>>On 07/19/2011 07:58 AM, Nicolas Kalkhof wrote:
>>> Hi there,
>>>
>>> I've experienced a strange depth buffer issue recently with OpenGL games
>>> (see attached screenshots). It seems that the depth buffer fails on some
>>> pixels. This problem was introduced somewhere between
>>> Kernel-3.0.0-rc6-git6 and kernel-3.0.0-rc7. Both xf-86-video-intel and
>>> mesa are latest git.
>>> Could someone please look into it?
>>
>>That's very odd. Any chance you could bisect to find the bad commit?
>>You should be able to just bisect the drivers/gpu/drm directory. I'm a
>>bit suspicious that this is a kernel issue. There were only 5 commits
>>in drivers/gpu/drm from rc6 to rc7, and none of them should affect any
>>Intel GPUs at all.
>>-----BEGIN PGP SIGNATURE-----
>>Version: GnuPG v1.4.11 (GNU/Linux)
>>Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/
>>
>>iEYEARECAAYFAk4l0OEACgkQX1gOwKyEAw/BWQCfeySgLPcT1bdNxPcTJr/S2aw0
>>vvkAnjS5OJF18pRwrRKaMyG2zoXFqwqB
>>=AiSG
>>-----END PGP SIGNATURE-----
>
>
>___________________________________________________________
>Schon gehört? WEB.DE hat einen genialen Phishing-Filter in die
>Toolbar eingebaut! http://produkte.web.de/go/toolbar
>_______________________________________________
>Intel-gfx mailing list
>Intel-gfx@lists.freedesktop.org
>http://lists.freedesktop.org/mailman/listinfo/intel-gfx


___________________________________________________________
Schon gehört? WEB.DE hat einen genialen Phishing-Filter in die
Toolbar eingebaut! http://produkte.web.de/go/toolbar
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: gen6 (SNB) depthbuffer issue with OpenGL games

[Ian Romanick]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 07/19/2011 01:21 PM, Nicolas Kalkhof wrote:
> Hi all,
> 
> ok I've nailed the issue down to 3.0.0-rc7 and 3.0.0-rc7-git1. I suspect that the changes made in
> drivers/gpu/drm/i915/i915_dma.c are the cause of the problem.
> 
> http://www.kernel.org/diff/diffview.cgi?file=%2Fpub%2Flinux%2Fkernel%2Fv3.0%2Fsnapshots%2Fpatch-3.0-rc7-git1.bz2;z=14
> 
> Any Clues?

Okay, so that's the commit below, which changes some error clean-up
paths.  That is also odd.  What *exact* GPU do you have?  Specificially,
what's the output of

    glxinfo | grep "OpenGL renderer"

and

    lspci -vn | grep VGA

Does this appear in all games or just certain games?  If it's just in
certain games, which ones?

commit a7b85d2aa63ed09cd5a4a640772b3272f5ac7caa
Author: Keith Packard <keithp@keithp.com>
Date:   Sun Jul 10 13:12:17 2011 -0700

    drm/i915: Clean up i915_driver_load failure path

    i915_driver_load adds a write-combining MTRR region for the GTT
    aperture to improve memory speeds through the aperture. If
    i915_driver_load fails after this, it would not have cleaned up the
    MTRR. This shouldn't cause any problems, except for consuming an MTRR
    register. Still, it's best to clean up completely in the failure path,
    which is easily done by calling mtrr_del if the mtrr was successfully
    allocated.

    i915_driver_load calls i915_gem_load which register
    i915_gem_inactive_shrink. If i915_driver_load fails after calling
    i915_gem_load, the shrinker will be left registered. When called, it
    will access freed memory and crash. The fix is to unregister the
shrinker in the
    failure path using code duplicated from i915_driver_unload.

    i915_driver_load also has some incorrect gotos in the error cleanup
    paths:

     * After failing to initialize the GTT (which cannot happen, btw,
       intel_gtt_get returns a fixed (non-NULL) value), it tries to
       free the uninitialized WC IO mapping. Fixed this by changing the
       target from out_iomapfree to out_rmmap

    Signed-off-by: Keith Packard <keithp@keithp.com>
    Tested-by: Lin Ming <ming.m.lin@intel.com>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/

iEYEARECAAYFAk4l8j4ACgkQX1gOwKyEAw+XRACaAzuKEihGpktWtA0UbTwx78NR
kBEAoJ7FxkXWEtxC5enkFwnFc19+sPGT
=abq+
-----END PGP SIGNATURE-----

Re: gen6 (SNB) depthbuffer issue with OpenGL games

[Nicolas Kalkhof]

Hi Ian,

thx for your reply. ok I give it a try and will report my findings....

regards
nic

-----Ursprüngliche Nachricht-----
Von: "Ian Romanick" <idr@freedesktop.org>
Gesendet: Jul 19, 2011 8:45:53 PM
An: "Nicolas Kalkhof" <nkalkhof@web.de>
Betreff: Re: [Intel-gfx] gen6 (SNB) depthbuffer issue with OpenGL games

>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>On 07/19/2011 07:58 AM, Nicolas Kalkhof wrote:
>> Hi there,
>>
>> I've experienced a strange depth buffer issue recently with OpenGL games
>> (see attached screenshots). It seems that the depth buffer fails on some
>> pixels. This problem was introduced somewhere between
>> Kernel-3.0.0-rc6-git6 and kernel-3.0.0-rc7. Both xf-86-video-intel and
>> mesa are latest git.
>> Could someone please look into it?
>
>That's very odd. Any chance you could bisect to find the bad commit?
>You should be able to just bisect the drivers/gpu/drm directory. I'm a
>bit suspicious that this is a kernel issue. There were only 5 commits
>in drivers/gpu/drm from rc6 to rc7, and none of them should affect any
>Intel GPUs at all.
>-----BEGIN PGP SIGNATURE-----
>Version: GnuPG v1.4.11 (GNU/Linux)
>Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/
>
>iEYEARECAAYFAk4l0OEACgkQX1gOwKyEAw/BWQCfeySgLPcT1bdNxPcTJr/S2aw0
>vvkAnjS5OJF18pRwrRKaMyG2zoXFqwqB
>=AiSG
>-----END PGP SIGNATURE-----


___________________________________________________________
Schon gehört? WEB.DE hat einen genialen Phishing-Filter in die
Toolbar eingebaut! http://produkte.web.de/go/toolbar
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: gen6 (SNB) depthbuffer issue with OpenGL games

[Safety0ff]

[-- Attachment #1.1: Type: text/plain, Size: 859 bytes --]

On 19/07/11 02:58 PM, Nicolas Kalkhof wrote:

> I've experienced a strange depth buffer issue recently with OpenGL games (see attached screenshots). It seems that the depth buffer fails on some pixels. This problem was introduced somewhere between Kernel-3.0.0-rc6-git6 and kernel-3.0.0-rc7. Both xf-86-video-intel and mesa are latest git. Could someone please look into it?
I've experienced this issue far earlier than the version stated above (i.e. with 2.39 rc's using mesa and xf86-video-intel from git from that time period).
It does not occur constantly in my case, just randomly, and not necessarily from application start (i.e. it sometimes started after it had been running a short period of time).

I would have mentioned this earlier if I wasn't having ram issues which make it difficult to have any certainty about problems which occur randomly.


[-- Attachment #1.2: Type: text/html, Size: 1415 bytes --]

[-- Attachment #2: Type: text/plain, Size: 159 bytes --]

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

gen6 (SNB) depthbuffer issue with OpenGL games

[Nicolas Kalkhof]

[-- Attachment #1: Type: text/html, Size: 1005 bytes --]

[-- Attachment #2: snapshot080418_1.png --]
[-- Type: image/png, Size: 245259 bytes --]

[-- Attachment #3: snapshot080419_1.png --]
[-- Type: image/png, Size: 251673 bytes --]

[-- Attachment #4: Type: text/plain, Size: 159 bytes --]

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: gen6 (SNB) depthbuffer issue with OpenGL games

[Ian Romanick]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 07/19/2011 07:58 AM, Nicolas Kalkhof wrote:
> Hi there,
> 
> I've experienced a strange depth buffer issue recently with OpenGL games
> (see attached screenshots). It seems that the depth buffer fails on some
> pixels. This problem was introduced somewhere between
> Kernel-3.0.0-rc6-git6 and kernel-3.0.0-rc7. Both xf-86-video-intel and
> mesa are latest git.
> Could someone please look into it?

That's very odd.  Any chance you could bisect to find the bad commit?
You should be able to just bisect the drivers/gpu/drm directory.  I'm a
bit suspicious that this is a kernel issue.  There were only 5 commits
in drivers/gpu/drm from rc6 to rc7, and none of them should affect any
Intel GPUs at all.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/

iEYEARECAAYFAk4l0OEACgkQX1gOwKyEAw/BWQCfeySgLPcT1bdNxPcTJr/S2aw0
vvkAnjS5OJF18pRwrRKaMyG2zoXFqwqB
=AiSG
-----END PGP SIGNATURE-----