Re: troubles with radeon driver on PowerBook G4
On Fre, 2010-10-29 at 17:50 -0400, Joseph Jezak wrote:
> On 10/29/10 09:04, Michel Dänzer wrote:
> > On Fre, 2010-10-29 at 14:17 +0200, Gerfried Fuchs wrote:
> >> * Michel Dänzer <firstname.lastname@example.org> [2010-10-27 15:08:48 CEST]:
> >>> On Mit, 2010-10-27 at 13:55 +0200, Gerfried Fuchs wrote:
> >>>> When using 3d applications like with warzone2100 or neverball the
> >>>> graphics display is totally unusable, even in windowed mode. There are
> >>>> extremely large bit planes all over the screen making anything else
> >>>> unrecognizable.
> >>> The full Xorg.0.log file would be a starting point.
> >>> This is more likely an issue with Mesa or the kernel though, so the
> >>> output of dmesg and
> >>> LIBGL_DEBUG=verbose glxinfo 2>&1
> >>> would also be interesting.
> >> Find attached all three outputs. Hope they help to track down the
> >> issue.
> > As Joseph pointed out, it's probably
> > https://bugs.freedesktop.org/show_bug.cgi?id=26496 . Someone will need
> > to track down what's the problem there. I don't have time for it, though
> > I can try to answer questions.
> If you have suggestions for tracking it down further than the bisect,
> I'd be happy to take a look.
See e.g. https://bugs.freedesktop.org/show_bug.cgi?id=26496#c6 , i.e.
find out if it's possible for the fake BO manager to end up putting VBOs
into VRAM, and if that actually happens here.
Or more generally, given that the problem doesn't occur with KMS, find
out what's different between the two cases. E.g., do they take different
paths affected by this change, ...
Earthling Michel Dänzer | http://www.vmware.com
Libre software enthusiast | Debian, X and DRI developer