Bug 2018

Summary: 3D depth display issues after suspend/resume
Product: Mesa Reporter: Andrew J. Montalenti <pixelmonkey>
Component: Drivers/DRI/UnichromeAssignee: Default DRI bug account <dri-devel>
Status: RESOLVED WONTFIX QA Contact:
Severity: normal    
Priority: high CC: xavier
Version: 6.2   
Hardware: x86 (IA32)   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:

Description Andrew J. Montalenti 2004-12-05 17:05:31 UTC
After suspending/resuming my KM400-based laptop, I usually get strange depth issues in 3D 
applications.

To see what I mean, check out these screenshots.  Sometimes things seem "beneath" other things, as in 
these shots from one resume.

Tux beneath himself:

http://www.pixelmonkey.org/pub/dri/tux.png

Balls beneath foobillard table:

http://www.pixelmonkey.org/pub/dri/foo.png

Other times, things seem completely messed-up depth wise.  Here are some shots of tuxracer after 
such a resume:

http://www.pixelmonkey.org/pub/dri/tux2.png
http://www.pixelmonkey.org/pub/dri/tux3.png
http://www.pixelmonkey.org/pub/dri/tux4.png

Strangely enough, certain apps don't experience these issues.  GLMatrix, for example, seems to work 
okay.  Celestia seems to work okay.  3ddeskd works reasonably well, but when rotating the workspaces, 
workspaces that should appear "behind" another sometimes "intersect" with the one in the foreground.  
This confirms my hunch that this is depth (z-ordering?  I'm not a 3D programmer ;-)) related.

http://www.pixelmonkey.org/pub/dri/tuxmenu.png

This is a shot of tux menu, which, as you can see, is perfectly fine.

In the resume where tux's body is all messed up, foobillard appears but only half the table is actually 
visible.  In addition, glxgears spins, but the right-most gear has some sort of depth issue as well.

In terms of performance, nothing is hurt by a suspend/resume, and stability is fine.  It's just these 
weird display issues.

On the Unichrome mailing list, Thomas said it might be the AGP kernel module if what are getting 
messed up are textures, but otherwise it may be the unichrome_dri driver.  Just in case, I filed a bug 
report on the kernel bugzilla as well.  Can anyone give me insight as to "what driver is to blame" for 
these issues, so at least I can make sure the proper dev is handling it (or maybe even try to fix it 
myself)?  Thanks.
Comment 1 almos 2012-08-14 23:14:37 UTC
Can this be closed?
Comment 2 almos 2014-05-24 21:13:23 UTC
The Unichrome driver was removed in 8.0 more than 2 years ago. I close this ancient bug as wontfix, because I have no idea if it was ever fixed, but I'm sure nobody is interested in developing that driver anymore.

Sidenote: the motherboard (with VIA chipset) in my AMD AthlonXP machine died earlier this year.

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.