Bug 21682 - White screen with compiz/AIGLX on X.org server 1.5.2 when running on 16bpp (intel 945GM)
Summary: White screen with compiz/AIGLX on X.org server 1.5.2 when running on 16bpp (i...
Status: RESOLVED WONTFIX
Alias: None
Product: Mesa
Classification: Unclassified
Component: Drivers/DRI/i915 (show other bugs)
Version: unspecified
Hardware: Other All
: medium normal
Assignee: Default DRI bug account
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2009-05-11 09:46 UTC by Tamás Németh
Modified: 2018-07-17 09:51 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
16bit.jpg (52.61 KB, image/jpeg)
2009-05-11 09:47 UTC, Tamás Németh
Details
My desktop in tis normal state (122.59 KB, image/jpeg)
2009-05-11 09:48 UTC, Tamás Németh
Details
Xorg.0.log (34.02 KB, text/plain)
2009-05-11 09:50 UTC, Tamás Németh
Details
xorg.conf (7.10 KB, text/plain)
2009-05-11 09:52 UTC, Tamás Németh
Details
hwinfo --gfxcard (1.73 KB, text/plain)
2009-05-11 09:53 UTC, Tamás Németh
Details
My normal openSUSE 11.3 desktop (24bpp) (403.64 KB, image/png)
2011-03-30 06:48 UTC, Tamás Németh
Details
My openSUSE 11.3 desktop distorted at 16bpp (308.19 KB, image/png)
2011-03-30 06:50 UTC, Tamás Németh
Details

Description Tamás Németh 2009-05-11 09:46:58 UTC
I'm using 64 bit openSUSE 11.1's X.org 7.4 packages and compiz 0.7.8. It works fine with 1400x1050 24bpp, when I switch to 16bpp (in order to be able to use krfb) then compiz draws almost everything totally white, except application windows (without decoration), the cube caps, the cube background and cube gears for me (see the attached 16bit.jpg). When I switch back to 24bpp, but disable the the dri module (and also unload the i915 and drm kernel modules), thus using sofware openGL, the sitution becomes even worse: application windows also be come white, so I was unable to take a screenshot of this scenario.
Comment 1 Tamás Németh 2009-05-11 09:47:41 UTC
Created attachment 25735 [details]
16bit.jpg

Hardware accelerated compiz with 16 bit colour depth
Comment 2 Tamás Németh 2009-05-11 09:48:07 UTC
Created attachment 25736 [details]
My desktop in tis normal state
Comment 3 Tamás Németh 2009-05-11 09:50:38 UTC
Created attachment 25737 [details]
Xorg.0.log
Comment 4 Tamás Németh 2009-05-11 09:52:06 UTC
Created attachment 25738 [details]
xorg.conf
Comment 5 Tamás Németh 2009-05-11 09:53:35 UTC
Created attachment 25739 [details]
hwinfo --gfxcard
Comment 6 Stefan Dirsch 2009-05-11 12:00:41 UTC
You can't use compiz in 16bpp or without hardware acceleration. 24bpp and hardware acceleration are requirements.
Comment 7 Tamás Németh 2009-05-11 13:18:20 UTC
(In reply to comment #6)
> You can't use compiz in 16bpp or without hardware acceleration. 24bpp and
> hardware acceleration are requirements.
> 

Understood, but it's not an acceptable behaviour that it starts wihout a single warning message (and it's even working hard to do something :), rendering the whole desktop unusable. It shouldn't start at all.
Comment 8 Michel Dänzer 2009-05-12 03:24:59 UTC
(In reply to comment #6)
> You can't use compiz in 16bpp or without hardware acceleration. 24bpp and
> hardware acceleration are requirements.

Huh? The white screen of death with software rendering is just a known longstanding bug which is pretty low priority to fix, as compiz likely wouldn't be usable anyway like that, but there is certainly no requirement for depth 24.
Comment 9 Michel Dänzer 2009-05-12 03:28:13 UTC
Likely a driver bug, most likely Mesa.
Comment 10 Eric Anholt 2009-08-07 18:44:21 UTC
Does this also occur with DRI2?
Comment 11 Tamás Németh 2009-08-08 00:00:34 UTC
(In reply to comment #10)
> Does this also occur with DRI2?
> 

Yes, but only if I use 16bpp instead of 24bpp (read #1). (This is compiz 0.7.8, I haven't tried 0.8.2 yet.)
Comment 12 Eric Anholt 2009-08-11 12:54:14 UTC
I read #1, and #1-#5 all show DRI1 in use, not DRI2.  Could you attach a Xorg.0.log from reproducing the problem with DRI2?
Comment 13 Tamás Németh 2009-10-15 05:10:27 UTC
(In reply to comment #12)
> I read #1, and #1-#5 all show DRI1 in use, not DRI2.  Could you attach a
> Xorg.0.log from reproducing the problem with DRI2?
> 

OK, I installed openSUSE 11.2 MI8, now I'm able to try to reproduce it with DRI2, but I don't know how to change to 16bpp, because there is no /etc/X11/xorg.conf. Is there a way to create /etc/X11/xorg.conf with the current running parameters of X.org (i.e. extracting the "running config" from the running X server in xorg.conf format)? Or do I have to alter bpp in another way?
Comment 14 Stefan Dirsch 2010-09-11 09:53:00 UTC
(In reply to comment #13)
> (In reply to comment #12)
> > I read #1, and #1-#5 all show DRI1 in use, not DRI2.  Could you attach a
> > Xorg.0.log from reproducing the problem with DRI2?
> > 
> 
> OK, I installed openSUSE 11.2 MI8, now I'm able to try to reproduce it with
> DRI2, but I don't know how to change to 16bpp, because there is no
> /etc/X11/xorg.conf. Is there a way to create /etc/X11/xorg.conf with the
> current running parameters of X.org (i.e. extracting the "running config" from
> the running X server in xorg.conf format)? Or do I have to alter bpp in another
> way?

Assuming you're using xdm (/etc/sysconfig/displaymanager:DISPLAYMANAGER=xdm) the easiest would be to add 'depth 16' to the Xserver options list in /etc/X11/xdm/Xservers. Not sure whether intel is already a DRI2 capable driver on 
openSUSE 11.2.
Comment 15 Stefan Dirsch 2010-09-17 23:36:59 UTC
Tamás is already using openSUSE 11.3, which ships xorg-server 1.8 and therefore includes DRI2.
Comment 16 Tamás Németh 2010-09-27 08:35:56 UTC
(In reply to comment #15)
> Tamás is already using openSUSE 11.3, which ships xorg-server 1.8 and therefore
> includes DRI2.

In fact I still have a netbook with openSUSE 11.2 but compiz ceased to work on it. I assume I will have to update it to opesSUSE 11.3. Can I do anything for you given that I only have 11.3 (also including a disfunctional compiz: https://bugzilla.novell.com/show_bug.cgi?id=641596) now?
Comment 17 Stefan Dirsch 2010-10-09 10:22:20 UTC
Tamás, looks like you resolved your issues with compiz on openSUSE 11.3. So can you finally verify whether the issue with 16bpp color depth still exists with a DRI2 capable driver?
Comment 18 Stefan Dirsch 2011-03-29 03:48:23 UTC
Tamas, any news on that one?
Comment 19 Tamás Németh 2011-03-30 02:19:50 UTC
(In reply to comment #18)
> Tamas, any news on that one?

I just tried it on openSUSE 11.4. Compiz 0.9.x is practically unusable (it crashes in half a minute after starting), however I was able to recognize that the white screen problem is gone in 16bpp but now the color palette of windows totally went wrong and the windows are displayed scaled down four times and every window is composed four of its scaled down image ad then tiled together. Totally unusable. Rendering is correct at higher color depth but as I mentioned earlier, compiz always crashes in half a minute after starting.

I still have an openSUSE 11.3 machine. Should I try that too?
Comment 20 Stefan Dirsch 2011-03-30 02:49:02 UTC
(In reply to comment #19)
> I still have an openSUSE 11.3 machine. Should I try that too?

That would be great, yes.
Comment 21 Tamás Németh 2011-03-30 06:48:31 UTC
Created attachment 45041 [details]
My normal openSUSE 11.3 desktop (24bpp)
Comment 22 Tamás Németh 2011-03-30 06:50:47 UTC
Created attachment 45042 [details]
My openSUSE 11.3 desktop distorted at 16bpp
Comment 23 Tamás Németh 2011-03-30 07:00:01 UTC
(In reply to comment #20)
> (In reply to comment #19)
> > I still have an openSUSE 11.3 machine. Should I try that too?
> 
> That would be great, yes.

Sadly, openSUSE 11.3 virtually behaves the same way as 11.4.

(See my two attachments.)
Comment 24 Elizabeth 2017-10-27 19:22:05 UTC
Too old, still valid for latest SW?? Thank you.
Comment 25 Stefan Dirsch 2018-07-17 09:51:11 UTC
compiz is no longer been shipped by any still supported Linux distribution I guess. Also 945GM were still 32 bit machines. Probably no longer been used by anyone. Let's better close this one as WONTFIX.


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.