Regression with xf86-video-ati and ati-1-0-branch

[Benjamin Herrenschmidt]

On Mon, 2006-03-13 at 00:38 +0100, Hanno Böck wrote:
> Am Montag, 13. März 2006 00:28 schrieb Benjamin Herrenschmidt:
> > Note that the new "stable" branch has been barely tested at this point,
> > I haven't quite "finished" with it :) I'n happy to know it builds at all
> > in fact :) But heh, that specific issue sounds like it might bite HEAD
> > as well, so please open that report.
> 
> If you open a new "stable" branch of the ati-driver and plan a release, would 
> you mind having a look at this issue, which has been unfixed for a long time 
> and will probably become more important as this HD-video-stuff is becoming 
> more widespread these days:
> https://bugs.freedesktop.org/show_bug.cgi?id=1462
> 

Thanks for pointing me to it, I posted a couple of ideas of things to
try out to find where the problem really is and asked Hui from ATI just
in case he happens to know. There are a couple of dodgy things in
radeon_video.c regarding testing of the chip type that might explain it.

Ben.