On Mon, Sep 12, 2011 at 05:08:46PM +0200, Mark Kettenis wrote:
> > Date: Mon, 12 Sep 2011 17:51:14 +0300
> > From: Paul Irofti <[email protected]>
> >
> > This breaks X on my sandybridge workstation. (This plus the kernel
patch
> > you sent before). I mailed you the dmesg in private, but I'll attach
it
> > here as well for whoever else might be interested. (Watchout, big
dmesg
> > cache.)
>
> It doesn't really break Sandybridge, because the current driver
> doesn't work on sandybridge at all. I guess Owain left the
> snadbybridge PCI Ids in the driver, to facilitate development.
Mark is 100% correct here.
>
> The Sandybridge support isn't quite there yet. But getting this stuff
> in will help us finish support for it. So it is not neccessary to
> test this on Sandybridge hardware (Intel Core i3/5/7-2xxx CPUs), but
> please do test on anything that's older than that (including the first
> generation Intel Core i3/5/7)!
In more detail, this code has the full sandybridge rendering code from
upstream as of a couple of months ago (before another acceleration
architecture experement started), as well as the renderer improvements
for other chipsets. What it currently lacks is the modesetting code for
sandybridge which is not quite done but getting closer.
>
> Cheers,
>
> Mark
>
Ditto,
-0-
--
A Law of Computer Programming:
Make it possible for programmers to write in English and you
will find the programmers cannot write in English.
