Bug 10240 - radeon driver locks computer with 9250 VIVO card when dri is enabled
Summary: radeon driver locks computer with 9250 VIVO card when dri is enabled
Status: RESOLVED FIXED
Alias: None
Product: xorg
Classification: Unclassified
Component: Driver/Radeon (show other bugs)
Version: 7.1 (2006.05)
Hardware: x86 (IA32) Linux (All)
: high major
Assignee: xf86-video-ati maintainers
QA Contact: Xorg Project Team
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2007-03-10 03:49 UTC by Nyyr :-)
Modified: 2007-08-09 02:58 UTC (History)
0 users

See Also:
i915 platform:
i915 features:


Attachments
/etc/X11/xorg.conf (1.48 KB, text/plain)
2007-03-10 03:49 UTC, Nyyr :-)
no flags Details
/etc/X11/xorg.conf.radeon (878 bytes, text/plain)
2007-03-10 03:50 UTC, Nyyr :-)
no flags Details
/var/log/Xorg.0.log (60.85 KB, text/plain)
2007-03-10 03:50 UTC, Nyyr :-)
no flags Details
/var/log/Xorg.0.log.save (54.75 KB, text/plain)
2007-03-10 03:51 UTC, Nyyr :-)
no flags Details

Description Nyyr :-) 2007-03-10 03:49:21 UTC
When configured as dual head (clone mode), radeon driver disables clone mode (because it does not find 2nd monitor (TV)) and DRI and then it is working.
But when using standard simple single head configuration, driver locks computer.
Attached files:
Xorg.conf - dual head cfg
Xorg.conf.radeon - single head cfg
Xorg.0.log - dual head log
Xorg.0.log.save - single head log
Comment 1 Nyyr :-) 2007-03-10 03:49:59 UTC
Created attachment 9067 [details]
/etc/X11/xorg.conf
Comment 2 Nyyr :-) 2007-03-10 03:50:24 UTC
Created attachment 9068 [details]
/etc/X11/xorg.conf.radeon
Comment 3 Nyyr :-) 2007-03-10 03:50:50 UTC
Created attachment 9069 [details]
/var/log/Xorg.0.log
Comment 4 Nyyr :-) 2007-03-10 03:51:10 UTC
Created attachment 9070 [details]
/var/log/Xorg.0.log.save
Comment 5 Roland Scheidegger 2007-03-10 15:49:50 UTC
Setup is very different if dri is enabled or not, so the reason it's crashing is almost certainly due to dri being enabled (you could disable it to confirm).
It could be the same as https://bugs.freedesktop.org/show_bug.cgi?id=9284, so try the workarounds there. If that doesn't help, and since you didn't say when exactly it locks up and it's not impossible your setup would use that at server startup, try disabling twopointlines as per Michel's comment in https://bugs.freedesktop.org/show_bug.cgi?id=5986.
In the future, please set the type of attachments correctly, as the content type for the four you provided really got messed up which just creates an unnecessary hassle for others to view them.
Comment 6 Nyyr :-) 2007-04-06 12:03:27 UTC
1.) I cannot find how to disable DRI. I don't have any *Load "DRI"* lines in my config and *Option "DRI" "false"* does not do the job. I can still see *Loading dri* line in Xorg.0.log

2.) Computer locks up during X boot.

3.) Adding those XaaNoSolidBresenhamLine and XaaNoSolidTwoPointLine statements helped only for the first time (first reboot after xorg.conf change). Since then  (next reboots) computer locks up during X boot again.

4.) In my BIOS AGP speed is set to 8 and FastWrites to ON. With these settings Windows XP work fine. I tried to put *Option "AGPMode" "8"* into xorg.conf, did not help. So I tried to add FastWrites ON option (as it is disabled by the X driver by default - according to Xorg.0.log), didn't help either.
Comment 7 Daniel Stone 2007-04-06 12:11:22 UTC
have you tried disabling fast writes in the bios?
Comment 8 Nyyr :-) 2007-04-06 13:14:07 UTC
Yes, I have disabled fast writes in bios and then tried all again. Still cannot start X.
Comment 9 Daniel Stone 2007-04-06 13:33:18 UTC
(didn't mean to click 'fixed', btw.)
Comment 10 Michel Dänzer 2007-04-07 04:39:08 UTC
(In reply to comment #6)
> 1.) I cannot find how to disable DRI.

With xf86-video-ati 6.6.191, Option "DRI" "off" should do the trick. With older versions you may need to prevent the X server dri module or the kernel radeon module from loading, e.g. by moving them away.


> 4.) In my BIOS AGP speed is set to 8 and FastWrites to ON. With these settings
> Windows XP work fine. I tried to put *Option "AGPMode" "8"* into xorg.conf, did
> not help. So I tried to add FastWrites ON option (as it is disabled by the X
> driver by default - according to Xorg.0.log), didn't help either.

Generally the settings should match between the BIOS setup and xorg.conf. The safest settings should be FastWrites off and mode 4x or 8x in both.
Comment 11 Nyyr :-) 2007-04-08 13:55:17 UTC
Option "DRI" "off" does NOT do the trick. (ati version 6.6.3).
I had to remove /usr/lib/xorg/modules/extensions/libdri.so to avoid loading it.

After removing DRI, X starts, even with FastWrites ON. But X now ignore mode 1280x960 and sets it to 1024x768 (I have TV attached to my VIVO card and even that X does not recognize it, it limits screen resolution, I guess....).

So it seems that it is a DRI bug....
Comment 12 Alex Deucher 2007-04-08 19:31:32 UTC
(In reply to comment #11)
> Option "DRI" "off" does NOT do the trick. (ati version 6.6.3).
> I had to remove /usr/lib/xorg/modules/extensions/libdri.so to avoid loading it.
> 
> After removing DRI, X starts, even with FastWrites ON. But X now ignore mode
> 1280x960 and sets it to 1024x768 (I have TV attached to my VIVO card and even
> that X does not recognize it, it limits screen resolution, I guess....).
> 
> So it seems that it is a DRI bug....
> 

6.6.3 still has the crtc2 size limiting bug and it does not have the dri option, so neither of those will work with 6.6.3.
Comment 13 Michel Dänzer 2007-06-09 04:00:32 UTC
Does it work with DRI enabled but with Option "BusType" "PCI"?
Comment 14 Nyyr :-) 2007-07-04 12:41:20 UTC
I have installed Fedora 7 (AMD_64) and now the open source radeon driver works fine. About 1580 FPS according to glxgears output.

So I'm not able to test the previous configuration anymore. (I have also another mainboard now, since my 32-bit Athlon XP CPU burned out :-( )
Comment 15 Timo Jyrinki 2007-08-09 02:58:31 UTC
If it works with Fedora 7, I guess this is fixed then. Please re-open if necessary.


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.