Re: ATI on FreeBSD! FireGL Linux Driver Port

On Tue 27 Dec 05 00:39, Chris Gilbert <Chris_at_lainos.org> wrote:
> Sounds good!
>
> Also, make sure to add a couple of xorg.conf lines in regards to this
> as explained in the README.
>
> By default, it will try to involve Xv with its GL bits and fail,
> therefore not exposing proper Xv acceleration. (xvinfo will report no
> available adapters)
>
> By adding the following two lines, the correct "GL-less" accelerated
> behavior will be restored: (These go in the "Device" section)
>
>         Option "VideoOverlay"   "on"
>         Option "OpenGLOverlay"  "off"

Yeah, I got that from the README. Even so, thanks for the heads up.

- jt

> On Tuesday 27 December 2005 08:33, Joshua Tinnin wrote:
> > On Tue 27 Dec 05 00:22, Eric Anholt <eta_at_lclark.edu> wrote:
> > > On Tue, 2005-12-27 at 00:10 -0700, Joshua Tinnin wrote:
> > > > On Mon 26 Dec 05 22:30, Chris Gilbert <Chris_at_lainos.org> wrote:
> > > > > Sure!
> > > > >
> > > > > Below is a support list based on what ATI reports the driver
> > > > > supports.
> > > > >
> > > > > Everything here should work as advertised, but to be sure, if
> > > > > you have success with a particular card then let me know and
> > > > > I'll mark the support as confirmed.
> > > > >
> > > > > It supports: (Put into a sane order, as opposed to what is on
> > > > > the ATI site)
> > > >
> > > > <snip>
> > > >
> > > > > Radeon® 9600 series
> > > >
> > > > I have an ATI Radeon 9600XT, and I'm proud to report it's been
> > > > working for the last hour with no problems. I can live without
> > > > the 3D acceleration in FreeBSD, but 2D acceleration helps
> > > > significantly. Finally using my card's memory instead of the
> > > > system's.
> > >
> > > Hmm.  X.Org should be supporting the same 2d acceleration for
> > > 9600s as fglrx.  Was that not the case?
> >
> > Well, rather, it seems to be working much better now. The radeon
> > X.Org driver never did seem to do 2D accel very well, at least in
> > my experience, though the logs tell me it was working. It seems
> > like it's actually working now, though I guess what I said
> > previously wasn't precise.
> >
> > - jt