Lock-ups with Radeon and xorg 6.9

[Erwin Rol]

Hey all,

just a "me too" for the radeon problems. My setup is a fedora rawhide
and since i updated to the new modular X the X server locks up my
machine hard (only power/reset button helps). The GDM login works fine,
and i can even login, but very shortly after login in it hangs my
machine. The VESA driver caused problems with wrong colors and shifted
icons, so that didn't work either (see also:
https://bugs.freedesktop.org/attachment.cgi?id=3851 )
The radeonfb kernel module plus the fbdev X-server seem to be working
the best at the moment but sometimes i also get a lockup with that
(about one time per day, for a machine that is on 24/7) The hardware
worked without lockups before the switch to the modular X Server.  

The software is;
X Window System Version 6.99.99.902 (7.0.0 RC 2)
Release Date: 09 November 2005
X Protocol Version 11, Revision 0, Release 6.99.99.902
Build Operating System:Linux 2.6.9-22.ELsmp x86_64Red Hat, Inc.
Current Operating System: Linux xpc.home.erwinrol.com
2.6.14-1.1696_FC5#1 SMP S at Nov 19 19:00:28 EST 2005 x86_64
Build Date: 17 November 2005

The hardware is;
01:00.0 VGA compatible controller: ATI Technologies Inc RV370 5B60
[Radeon X300 (PCIE)]
01:00.1 Display controller: ATI Technologies Inc RV370 [Radeon X300SE]
(which is a el'cheapo PCIe card)


- Erwin

On Wed, 2005-11-23 at 14:38 +0100, Matthias Hopf wrote:
> On Nov 16, 05 20:30:06 +0100, Alexej Davidov wrote:
> > since I've upgraded to xorg 6.9, I experience frequent lockups. With
> > DRI enabled, the system locks up when starting xorg (reproducable),
> > with DRI disabled it will lock up at some random time later. The lockup
> > is complete: no net, no sysrq, NMI watchdog doesn't "bite" either. The
> > logs don't schow anything unusual. Nothing in syslog. X log shows that
> > DRI could be enabled (if enabled in xorg.conf).
> > 
> > Computer: IBM Thinkpad T42p
> > Video card: ATI FireGL Mobility T2 (M10) NT
> > Xorg: 6.8.99.901
> > Kernel: 2.6.14
> 
> Just to asure you that you are not alone. We have multiple bug reports
> regarding radeon with exactly your symptoms. I'm trying to track down
> the change, and it seems that is not only the radeon driver that's
> involved, as a back-ported CVS driver (to 6.8.2) seems to work fine.
> 
> For details:
> 
> https://bugzilla.novell.com/show_bug.cgi?id=127757
> 
> However, I do not really have much knowledge about the inner workings of
> the radeon driver, so help would be appreciated.
> 
> Matthias