Bug 21243

Summary: [radeon-rewrite] shadowtex: warnings and darker image
Product: Mesa Reporter: Fabio Pedretti <pedretti.fabio>
Component: Drivers/DRI/r300Assignee: Default DRI bug account <dri-devel>
Status: RESOLVED FIXED QA Contact:
Severity: normal    
Priority: medium    
Version: git   
Hardware: x86 (IA32)   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:

Description Fabio Pedretti 2009-04-17 04:14:50 UTC 
Running progs/demos/shadowtex under Ubuntu 9.04 32 bit with current git of the radeon-rewrite branch (up to commit 9028335371af465a73c13282a3608f76168e4d63, compiled without libdrm-radeon and using standard Ubuntu kernel) with a RV530  I get thousand of this warnings:

Mesa 7.5-devel implementation error: unexpected texture format in setup_hardware_state
Please report at bugzilla.freedesktop.org

Also, the image output is darker than that of r300 master, where the warning is also not shown (however both don't show the base shadow like the sw render does).
Comment 1 Maciej Cencora 2009-05-07 18:01:21 UTC 
The warning is gone with a7bb95bf142ac9727a4223639742a8524f9c724b. Also that patch almost makes it work (try pressing 'p' when it didn't work from start).
I think there is a race condition in context initialization because I can get it to work from start once per 10 attempts, and it works every time when started under valgrind.
Comment 2 Fabio Pedretti 2009-05-08 00:27:53 UTC 
Nice work!

I confirm the warnings are gone, I get the darker image 1/10 of the times, while 9/10 works fine.
Comment 3 Fabio Pedretti 2009-05-12 03:14:21 UTC 
A maybe related issue affects progs/tests/rubberband , sometimes it works as expected, sometimes it shows an empty window (the window shows the content under the window itself).
Comment 4 Jerome Glisse 2009-05-18 07:40:52 UTC 
Please open a different bugs if you have q different issue than the one reported here in the first place.
Comment 5 Fabio Pedretti 2009-06-12 01:58:08 UTC 
(In reply to comment #1)
> The warning is gone with a7bb95bf142ac9727a4223639742a8524f9c724b. Also that
> patch almost makes it work (try pressing 'p' when it didn't work from start).
> I think there is a race condition in context initialization because I can get
> it to work from start once per 10 attempts, and it works every time when
> started under valgrind.
> 

Apparently this is fixed with current mesa master with radeon-rewrite merged.
Comment 6 Adam Jackson 2009-08-24 12:32:04 UTC 
Mass version move, cvs -> git

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.