Skip to contentSkip to conversation

Graphical corruption post 2.6.38

DISCUSSION Colin Guthrie 4 participants 12 messages Mar 23, 2011 – May 12, 2011

lore
Share & tools
lore
#0 lore raw

Hi,

Running latest drivers from git + recent drm+xserver.

After upgrading to 2.6.38 I seem to get occasional graphical corruption. This manifests itself mostly obviously in e.g gnome-terminal when getting git status and such like (although there are many other triggers
- have seen it in various apps).

Here is an example of what it looks like, but it can be much worse.
http://colin.guthr.ie/intel-2.6.38.png

Any idea of what could cause this before I go digging too deeply?

HW is:
unknown : Intel Corporation|Mobile 945GM/GMS/GME, 943/940GML
Express Integrated Graphics Controller [DISPLAY_OTHER] (vendor:8086 device:27a6 subv:1028 subd:01bd) (rev: 03)

Card:Intel 810 and later: Intel Corporation|Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller [DISPLAY_VGA] (vendor:8086 device:27a2 subv:1028 subd:01bd) (rev: 03)

Col

Signature · 5 lines
+------------------------+
|     Colin Guthrie      |
+------------------------+
| http://colin.guthr.ie/ |
+------------------------+
Reviewed-by Daniel Vetter Tested-by Dave Airlie
#1 lore raw
Quoted 9 lines · Colin Guthrie

Hi,

Running latest drivers from git + recent drm+xserver.

After upgrading to 2.6.38 I seem to get occasional graphical corruption. This manifests itself mostly obviously in e.g gnome-terminal when getting git status and such like (although there are many other triggers
- have seen it in various apps).

Quoted 4 lines

Here is an example of what it looks like, but it can be much worse.
http://colin.guthr.ie/intel-2.6.38.png

Any idea of what could cause this before I go digging too deeply?

First try:

commit 29c5a587284195278e233eec5c2234c24fb2c204

Author: Chris Wilson <[email protected]>
Date:   Thu Mar 17 15:23:22 2011 +0000

drm/i915: Fix tiling corruption from pipelined fencing

... even though it was disabled. A mistake in the handling of fence reuse caused us to skip the vital delay of waiting for the object to finish rendering before changing the register. This resulted in us changing the fence register whilst the bo was active and so causing the blits to complete using the wrong stride or even the wrong tiling. (Visually the effect is that small blocks of the screen look like they have been interlaced). The fix is to wait for the GPU to finish using the memory region pointed to by the fence before changing it.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=34584

Cc Andy Whitcroft <[email protected]>
Cc Daniel Vetter <[email protected]>
[Note for 2.6.38-stable, we need to reintroduce the interruptible passing]

Thanks,
-Chris

Signature · 1 line
Chris Wilson, Intel Open Source Technology Centre
#2 lore raw

'Twas brillig, and Chris Wilson at 23/03/11 14:08 did gyre and gimble:

Quoted 15 lines · Colin Guthrie
Quoted 9 lines

Hi,

Running latest drivers from git + recent drm+xserver.

After upgrading to 2.6.38 I seem to get occasional graphical corruption. This manifests itself mostly obviously in e.g gnome-terminal when getting git status and such like (although there are many other triggers
- have seen it in various apps).

Quoted 4 lines

Here is an example of what it looks like, but it can be much worse.
http://colin.guthr.ie/intel-2.6.38.png

Any idea of what could cause this before I go digging too deeply?

Quoted 7 lines

First try:

commit 29c5a587284195278e233eec5c2234c24fb2c204

Author: Chris Wilson <[email protected]>
Date:   Thu Mar 17 15:23:22 2011 +0000

drm/i915: Fix tiling corruption from pipelined fencing

Looks like a good bet. I'll try and get this tested sometime soon :)

Cheers

Col

Signature · 5 lines
+------------------------+
|     Colin Guthrie      |
+------------------------+
| http://colin.guthr.ie/ |
+------------------------+
#3 lore raw
In reply to Chris Wilson
Quoted 15 lines · Chris Wilson
Quoted 9 lines · Colin Guthrie

Hi,

Running latest drivers from git + recent drm+xserver.

After upgrading to 2.6.38 I seem to get occasional graphical corruption. This manifests itself mostly obviously in e.g gnome-terminal when getting git status and such like (although there are many other triggers
- have seen it in various apps).

Quoted 4 lines

Here is an example of what it looks like, but it can be much worse.
http://colin.guthr.ie/intel-2.6.38.png

Any idea of what could cause this before I go digging too deeply?

Quoted 7 lines

First try:

commit 29c5a587284195278e233eec5c2234c24fb2c204

Author: Chris Wilson <[email protected]>
Date:   Thu Mar 17 15:23:22 2011 +0000

drm/i915: Fix tiling corruption from pipelined fencing

Hi,

where exactly can I find a version of the patch that applies against 2.6.38?

I tried several attachements of the bugzilla entry, without success.

Regards,
Tino

#4 lore raw
Quoted 10 lines · Tino Keitel
Quoted 5 lines · Chris Wilson

commit 29c5a587284195278e233eec5c2234c24fb2c204

Author: Chris Wilson <[email protected]>
Date:   Thu Mar 17 15:23:22 2011 +0000

drm/i915: Fix tiling corruption from pipelined fencing

Hi,

where exactly can I find a version of the patch that applies against 2.6.38?

It's supposed to be available from kernel.org, but that appears to be lagging severely, so try:

http://cgit.freedesktop.org/~ickle/linux-2.6/commit/?h=2.6.38&id=3b936797f5e4623ae1b80a5f0c4df30a17360bf5
-Chris

Signature · 1 line
Chris Wilson, Intel Open Source Technology Centre
#5 lore raw

'Twas brillig, and Chris Wilson at 23/03/11 18:51 did gyre and gimble:

Quoted 8 lines · Tino Keitel
Quoted 7 lines · Chris Wilson
Quoted 5 lines

commit 29c5a587284195278e233eec5c2234c24fb2c204

Author: Chris Wilson <[email protected]>
Date:   Thu Mar 17 15:23:22 2011 +0000

drm/i915: Fix tiling corruption from pipelined fencing

Quoted 8 lines
Quoted 3 lines

Hi,

where exactly can I find a version of the patch that applies against 2.6.38?

It's supposed to be available from kernel.org, but that appears to be lagging severely, so try:

http://cgit.freedesktop.org/~ickle/linux-2.6/commit/?h=2.6.38&id=3b936797f5e4623ae1b80a5f0c4df30a17360bf5

Thanks Chris, that seems to have done the trick.

Still see a couple weird bits occasionally but I've only seen then in Firefox on font glyphs so I think that's a FF problem - it certainly doesn't seem to be generic like the one that patch fixes. Hopefully it'll go into 2.6.38.2 :)

Cheers

Col

Signature · 5 lines
+------------------------+
|     Colin Guthrie      |
+------------------------+
| http://colin.guthr.ie/ |
+------------------------+
#6 lore raw
Quoted 4 lines · Colin Guthrie

Still see a couple weird bits occasionally but I've only seen then in Firefox on font glyphs so I think that's a FF problem - it certainly doesn't seem to be generic like the one that patch fixes. Hopefully it'll go into 2.6.38.2 :)

I've an unusual bug https://bugs.freedesktop.org/show_bug.cgi?id=35557
that I've not begun to dig into. Similar?
-Chris

Signature · 1 line
Chris Wilson, Intel Open Source Technology Centre
#7 lore raw

'Twas brillig, and Chris Wilson at 26/03/11 17:26 did gyre and gimble:

Quoted 8 lines · Colin Guthrie
Quoted 4 lines

Still see a couple weird bits occasionally but I've only seen then in Firefox on font glyphs so I think that's a FF problem - it certainly doesn't seem to be generic like the one that patch fixes. Hopefully it'll go into 2.6.38.2 :)

I've an unusual bug https://bugs.freedesktop.org/show_bug.cgi?id=35557
that I've not begun to dig into. Similar?

Seems kinda similar, but I've not seen it quite as holistic as that before. I tend to see a particular word messed up, rather than an individual glyph messed up consistently throughout the page, but perhaps I've just not noticed (it is quite subtle in that screenshot).

Col

Signature · 5 lines
+------------------------+
|     Colin Guthrie      |
+------------------------+
| http://colin.guthr.ie/ |
+------------------------+
#8 lore raw
In reply to Colin Guthrie

Dear Chris,

Quoted 14 lines · an unknown sender

'Twas brillig, and Chris Wilson at 23/03/11 18:51 did gyre and gimble:

Quoted 13 lines · Tino Keitel
Quoted 8 lines · Chris Wilson
Quoted 5 lines

commit 29c5a587284195278e233eec5c2234c24fb2c204

Author: Chris Wilson <chris at chris-wilson.co.uk>
Date:   Thu Mar 17 15:23:22 2011 +0000

drm/i915: Fix tiling corruption from pipelined fencing

where exactly can I find a version of the patch that applies against 2.6.38?

It's supposed to be available from kernel.org, but that appears to be lagging severely, so try:

Quoted 8 lines
Quoted 1 line

http://cgit.freedesktop.org/~ickle/linux-2.6/commit/?h=2.6.38&id=3b936797f5e4623ae1b80a5f0c4df30a17360bf5

Thanks Chris, that seems to have done the trick.

Still see a couple weird bits occasionally but I've only seen then in Firefox on font glyphs so I think that's a FF problem - it certainly doesn't seem to be generic like the one that patch fixes. Hopefully it'll go into 2.6.38.2 :)

I did not see the above commit in 2.6.38.x. Could it please be send to <[email protected]> for inclusion.

Thanks,

Paul

PS: I hope the message is threaded correctly. I have delivery problems with lists hosted at Freedesktop.org and using the compressed archives does not work so easily since Mailman seems to mangle the messages before putting them into the archive [1].

[1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=624111

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#9 lore raw
Quoted 9 lines · Paul Menzel
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines · Chris Wilson
Quoted 5 lines

commit 29c5a587284195278e233eec5c2234c24fb2c204

Author: Chris Wilson <chris at chris-wilson.co.uk>
Date:   Thu Mar 17 15:23:22 2011 +0000

drm/i915: Fix tiling corruption from pipelined fencing

I did not see the above commit in 2.6.38.x. Could it please be send to <[email protected]> for inclusion.

I have a backported fix for 2.6.38.x.

Thanks for the reminder,
-Chris

Signature · 1 line
Chris Wilson, Intel Open Source Technology Centre
#10 lore raw

Am Montag, den 25.04.2011, 22:13 +0100 schrieb Chris Wilson:

Quoted 6 lines · Paul Menzel
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines · Chris Wilson
Quoted 4 lines

commit 29c5a587284195278e233eec5c2234c24fb2c204

Author: Chris Wilson <chris at chris-wilson.co.uk>
Date:   Thu Mar 17 15:23:22 2011 +0000
Quoted 6 lines
Quoted 4 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

drm/i915: Fix tiling corruption from pipelined fencing

I did not see the above commit in 2.6.38.x. Could it please be send to <[email protected]> for inclusion.

I have a backported fix for 2.6.38.x.

Great. I did not see it in 2.6.38.5-stable review [1] though. Is the backported fix included there?

Thanks,

Paul

[1] https://lkml.org/lkml/2011/4/29/259

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#11 lore raw

Am Freitag, den 29.04.2011, 22:16 +0200 schrieb Paul Menzel:

Quoted 9 lines

Am Montag, den 25.04.2011, 22:13 +0100 schrieb Chris Wilson:

Quoted 8 lines · Paul Menzel
Quoted 7 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines · Chris Wilson
Quoted 5 lines

commit 29c5a587284195278e233eec5c2234c24fb2c204

Author: Chris Wilson <chris at chris-wilson.co.uk>
Date:   Thu Mar 17 15:23:22 2011 +0000

drm/i915: Fix tiling corruption from pipelined fencing

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

I did not see the above commit in 2.6.38.x. Could it please be send to <[email protected]> for inclusion.

I have a backported fix for 2.6.38.x.

Great. I did not see it in 2.6.38.5-stable review [1] though. Is the backported fix included there?

I just want to report that using Fedora 15 Beta Live stick using Linux kernel 2.6.38.x also exhibits this problem.

Thanks,

Paul

Quoted 1 line

[1] https://lkml.org/lkml/2011/4/29/259

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx