Bug 21448

Summary: [GM965] Artefacts while rendering 3D in crystalspace
Product: Mesa Reporter: Mateusz Kaduk <mateusz.kaduk>
Component: Drivers/DRI/i965Assignee: Ian Romanick <idr>
Status: RESOLVED FIXED QA Contact:
Severity: normal    
Priority: medium Keywords: NEEDINFO
Version: unspecified   
Hardware: Other   
OS: All   
Whiteboard:
i915 platform: i915 features:
Attachments: Screenshot showing artefacts.
Walktest log from Crystalspace 1.2.1-2

Description Mateusz Kaduk 2009-04-27 16:08:18 UTC
Created attachment 25208 [details]
Screenshot showing artefacts.

Hi,
I installed package crystalspace and on other systems it works fine, while on GM965 with latest mesa master branch (28-04-09) walktest produce 3D rendering with artifacts as on attached screenshot.
Comment 1 Eric Anholt 2009-05-12 16:46:43 UTC
Works fine for me with Mesa master or 36edcebb9012f530a9288956a7fd3b21f9bb8aed on GM45.
Comment 2 Mateusz Kaduk 2009-05-13 23:47:42 UTC
(In reply to comment #1)
> Works fine for me with Mesa master or 36edcebb9012f530a9288956a7fd3b21f9bb8aed
> on GM45.
> 

It does not on GM965. Run walktest-1.2 and move around with arrows, you will get what I showed on screenshot. If you try walktest-1.4 or walktest-1.9 the window will be black, showing from time to time very dark artefacts when you push arrow buttons.

walktest-1.2 is from debian package of crystalspace 1.2.1
walktest-1.4 and walktest-1.9 are custom compiled from sources (cs 1.4 and 1.9-dev), they don't overlap.

my master is d3912e301fd707738b0952cd11e19f34b87765b8
Comment 3 Mateusz Kaduk 2009-07-20 02:24:13 UTC
Tested with debian packaged mesa 7.5 bug still present.

Configuration is as follow
00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 0c)
Xorg 1.6.1
mesa 7.5
xf86-video-intel 2.7.1
libdrm 2.4.11
linux 2.6.30 (kms enabled)
Comment 4 Ian Romanick 2009-09-02 23:26:40 UTC
(In reply to comment #2)
> (In reply to comment #1)
> > Works fine for me with Mesa master or 36edcebb9012f530a9288956a7fd3b21f9bb8aed
> > on GM45.
> > 
> 
> It does not on GM965. Run walktest-1.2 and move around with arrows, you will
> get what I showed on screenshot.

I tried this with walktest-1.2 from debian and Mesa master 06bf6628aabd6744ebf4268319eba4a1d377ffe2.  I was not able to reproduce this error.  The date on the SHA1 you quote is from May, which is quite old at this point.  Are you still able to reproduce this bug?
Comment 5 Mateusz Kaduk 2009-09-03 01:00:01 UTC
(In reply to comment #4)
> (In reply to comment #2)
> > (In reply to comment #1)
> > > Works fine for me with Mesa master or 36edcebb9012f530a9288956a7fd3b21f9bb8aed
> > > on GM45.
> > > 
> > 
> > It does not on GM965. Run walktest-1.2 and move around with arrows, you will
> > get what I showed on screenshot.
> 
> I tried this with walktest-1.2 from debian and Mesa master
> 06bf6628aabd6744ebf4268319eba4a1d377ffe2.  I was not able to reproduce this
> error.  The date on the SHA1 you quote is from May, which is quite old at this
> point.  Are you still able to reproduce this bug?
> 

I tried with 
mesa master (c19b52f10fe11282f0a3466eda2eeb1d422b7aa3) ---> 7.6-devel
drm master (73b59c894380995a2889b98e79acadd2da0bb237)  ---> libdrm 2.4.13

Reinstalled crystalspace 1.2.1-2 from sid
xserver-xorg-core 2:1.6.3-1
linux 2.6.30.5

Same problem with 
OpenGL renderer string: Mesa DRI Intel(R) 965GM GEM 20090712 2009Q2 RC3 
OpenGL version string: 2.1 Mesa 7.6-devel

However Software Rasterizer which I hited at begging since I did not know that mesa needed newer libdrm, worked fine.

This is probably some kind of specific case that I hit, I tried removing
libtxc-dxtn for S3TC but it did not help.

I am going to attach log file from walktest.
Comment 6 Mateusz Kaduk 2009-09-03 01:02:43 UTC
Created attachment 29165 [details]
Walktest log from Crystalspace 1.2.1-2

Attaching /usr/bin/walktest > walktest.txt 2> walktest.txt
Crystalspace from Debian (sid)
ii  crystalspace                                         1.2.1-2

Problem still exist
Comment 7 Eugeni Dodonov 2011-08-23 12:54:27 UTC
I could not reproduce it anymore either, so I think it was fixed meanwhile. But if it still happens with latest Intel graphics stack, please reopen the bug so we could investigate it further.
Comment 8 Mateusz Kaduk 2012-01-25 02:27:17 UTC
Cannot reproduce it. Seems to be fixed. Reopen if needed.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.