[Xorg-driver-geode] Geode LX video testing results

Jordan Crouse jordan.crouse at amd.com
Fri Dec 14 11:36:43 PST 2007


On 14/12/07 14:16 -0500, Bart Trojanowski wrote:
> * Bart Trojanowski <bart at jukie.net> [071214 12:00]:
> > * Jordan Crouse <jordan.crouse at amd.com> [071214 11:53]:
> > > So at this point, its clear we're walking into a trap somewhere in the
> > > VBE code and not coming back.  At this point, I'm afraid the course of
> > > action is to slog forward and figure out what part of the VBE code is
> > > breaking.  Then we might get a better idea of what our options are.  I
> > > would help, but unfortunately, I don't have any hardware handy that
> > > breaks (unless the db61 would suffice?).
> > 
> > I am perfectly happy to find where we go wrong, I may need your
> > expertise to figure out how to fix it when I get there.
> 
> It dies in vbeDoEDID() called by GeodeDoDDC().

http://gitweb.freedesktop.org/?p=xorg/xserver.git;a=blob;h=06559bbd5c41bef06bbfc5935b62da0c240302d2;hb=86730337001ba4db6d77fe42406695e32784b157;f=hw/xfree86/vbe/vbe.c#l310

There are a lot of INT10 calls in that code path.  WE're getting pretty close
to needing somebody with a FS2 and knowledge of the BIOS to jump in and
watch what happens on the other side.

Jordan

-- 
Jordan Crouse
Systems Software Development Engineer 
Advanced Micro Devices, Inc.




More information about the Xorg-driver-geode mailing list