Bug 104090

Summary:	VIA X-server dies on Maxdata M-1200X
Product:	[openSUSE] SUSE LINUX 10.0	Reporter:	Stefan Behlert <behlert>
Component:	X.Org	Assignee:	Stefan Dirsch <sndirsch>
Status:	RESOLVED FIXED	QA Contact:	Stefan Dirsch <sndirsch>
Severity:	Blocker
Priority:	P5 - None
Version:	Beta 3
Target Milestone:	---
Hardware:	Other
OS:	All
Whiteboard:
Found By:	Development	Services Priority:
Business Priority:		Blocker:	---
Marketing QA Status:	---	IT Deployment:	---
Attachments:	Xorg.0.log.old lspci-outputs New Xext module logfile xf86debug log

Description Stefan Behlert 2005-08-11 12:01:55 UTC

There are two possible ways to kill the X server on a Maxdata 1200X: 
 
a) Switch from X to virtual terminal and back - X Server dies 
b) Suspend and resume - X server dies 
 
I'll append the Xorg.log - the machine is available from us.

Comment 1 Stefan Behlert 2005-08-11 12:02:58 UTC

Created attachment 45720 [details]
Xorg.0.log.old

Comment 2 Forgotten User ZhJd0F0L3x 2005-08-11 12:04:25 UTC

bug #98907 may be related

Comment 3 Stefan Behlert 2005-08-11 12:06:17 UTC

Created attachment 45721 [details]
lspci-outputs

Comment 4 Stefan Dirsch 2005-08-11 12:11:57 UTC

> Symbol xf86XvMCRegisterDRInfo from module 
> /usr/X11R6/lib/modules/drivers/via_drv.o is unresolved!
> Symbol xf86XvMCRegisterDRInfo from module 
> /usr/X11R6/lib/modules/drivers/via_drv.o is unresolved!

Looks suspicious to me.

Comment 5 Stefan Behlert 2005-08-11 12:27:44 UTC

It's a plain beta1-installation, only resmgr and bluetooth-packages have been 
replaced by newer versions. 
The machine is available online as g72, root-pw is the usual 
mobile-devices-laptop-password.

Comment 6 Stefan Dirsch 2005-08-11 12:32:50 UTC

My fault. I forgot to package

  programs/Xserver/Xext/xvmc.c
  programs/Xserver/Xext/xvmcext.h

in via-CVS-2005-07-07.tar.bz2 (used in xorg-x11 build). I'll fix this.

Comment 7 Stefan Dirsch 2005-08-11 13:35:07 UTC

Created attachment 45752 [details]
New Xext module

Looks like this fixes the problem. Copy the file
to /usr/X11R6/lib/modules/extensions/libextmod.a.

Comment 8 Stefan Dirsch 2005-08-11 13:36:28 UTC

Please give it a try. Also on the machine of bug #98907.

Comment 9 Forgotten User ZhJd0F0L3x 2005-08-11 15:16:21 UTC

Created attachment 45784 [details]
logfile

Log from the bug #98907 machine.
Still dies at switch back to X.

Comment 10 Stefan Behlert 2005-08-11 15:41:00 UTC

 Ok, I tried it, and switching to virtual console does now no longer kill the  
machine. I'll check suspend/resume later. (Had a midair-collision and was in a 
meeting meanwhile, sorry)

Comment 11 Forgotten User ZhJd0F0L3x 2005-08-11 21:35:52 UTC

Created attachment 45830 [details]
xf86debug log

Comment 12 Stefan Dirsch 2005-08-11 21:39:07 UTC

This is exactly what I got on g72 without the new libextmod.a when doing a 
"chvt 1; chvt 8" after starting xf86debug.

Comment 13 Stefan Behlert 2005-08-12 12:24:07 UTC

X is surviving suspend  on the maxdata machine now, too.

Comment 14 Stefan Dirsch 2005-08-12 12:28:21 UTC

Ok. Then let's close this one as FIXED. I've commited a new xorg-x11 package for
STABLE.

Comment 15 Stefan Behlert 2005-08-25 11:34:53 UTC

switching to console and back kills X-Server on beta3 on that machine again. 
It's currently available as g228

Comment 16 Stefan Dirsch 2005-08-25 13:05:20 UTC

Adding

  Load "dri"

to Modules section seems to help. I'll add this to x11DB, so it will be fixed
with Beta4.