Bug#580208: drm:radeon_fence_wait and drm:radeon_ib_schedule ERROR

[Michel Dänzer]

On Mit, 2010-05-05 at 08:18 +0200, Mader, Alexander (N-MSR) wrote: 
> Am 04.05.2010 13:14, schrieb Mader, Alexander (N-MSR):
> > Am 04.05.2010 12:34, schrieb Cyril Brulebois:
> >> Alexander Mader<Alexander.Mader at niles.de> (04/05/2010):
> >>> There was an udev update yesterday: "[AKTUALISIERUNG] udev 151-3 ->
> >>> 153-1", perhaps I try a rollback.
> >> Did you upgrade to/install the 2.6.32-5-amd64 kernel at the same time?
> >> If you could try with 2.6.32-4-amd64, that might help figuring out
> >> whether that's a regression on the kernel side.
> >
> > The first start with failure was with 2.6.32-4-amd64, see kern.log from
> > line 1 :-( After upgrade to 2.6.32-5-amd64 today the behaviour did not
> > change, see kern.log from line 1553. I did not try 2.6.32-4-amd64
> > together with firmware-linux-free 2.6.32-12, just with the previous
> > 2.6.32-11. To sum it up:
> >
> > 2.6.32-4-amd64 with udev 153-1 and firmware-linux-free 2.6.32-11: Failure
> > 2.6.32-4-amd64 with udev 153-1 and firmware-linux-free 2.6.32-12: Not
> > tested
> > 2.6.32-5-amd64 with udev 153-1 and firmware-linux-free 2.6.32-12: Failure
> >
> > In fact I do not know whether:
> >
> > 2.6.32-4-amd64 with udev 151-3 and firmware-linux-free 2.6.32-11: Success
> >
> > is true any more, it definitely was on Monday :-/ see attached excerpt
> > from kern.log.1.
> 
> A short interim report: KMS works fine with 2.6.32-4-amd64 after I 
> downgraded the following packages, see attached dmesg output:

[...]

> [DEAKTUALISIERUNG] xserver-xorg-video-radeon 1:6.13.0-1 -> 1:6.12.6-1

The 6.12 driver doesn't support KMS, so you didn't in fact test KMS but
a completely unsupported configuration where the kernel and X drivers
are fighting for hardware access.


-- 
Earthling Michel Dänzer           |                http://www.vmware.com
Libre software enthusiast         |          Debian, X and DRI developer