Does DRI/DRM still work on non-coherent DMA platforms?

Does DRI/DRM still work on non-coherent DMA platforms?

[Gerhard Pircher]

Hi,

I heard rumors that DRI doesn't work on SAM440EP (onboard Radeon GPU)
boards. I had DRI halfway working on my AmigaOne (Radeon 9250) with an
early release candidate of the 2.6.25 kernel and IIRC this patch here
(manually applied):
http://kerneltrap.org/index.php?q=mailarchive/git-commits-head/2008/3/30/1301044

I just tested DRI again on my A1 (I don't have a SAM board) with
v2.6.29-rc6 and the machine looks up while starting the X server.
Curiously DRI seems to work, if I specify "udbg-immortal" on the
kernel command line and enable DRM debugging.
Did anyone try out DRI on a non-coherent DMA PPC platform and can
confirm that it still works?

best regards,

Gerhard

-- 
Psssst! Schon vom neuen GMX MultiMessenger gehört? Der kann`s mit allen: http://www.gmx.net/de/go/multimessenger01

Re: Does DRI/DRM still work on non-coherent DMA platforms?

[Benjamin Herrenschmidt]

On Mon, 2009-03-09 at 22:30 +0100, Gerhard Pircher wrote:
> Hi,
> 
> I heard rumors that DRI doesn't work on SAM440EP (onboard Radeon GPU)
> boards. I had DRI halfway working on my AmigaOne (Radeon 9250) with an
> early release candidate of the 2.6.25 kernel and IIRC this patch here
> (manually applied):
> http://kerneltrap.org/index.php?q=mailarchive/git-commits-head/2008/3/30/1301044
> 
> I just tested DRI again on my A1 (I don't have a SAM board) with
> v2.6.29-rc6 and the machine looks up while starting the X server.
> Curiously DRI seems to work, if I specify "udbg-immortal" on the
> kernel command line and enable DRM debugging.
> Did anyone try out DRI on a non-coherent DMA PPC platform and can
> confirm that it still works?

It should work... try using dri-next branch from airlied tree, I had
that working on a Canyonlands after DaveM and I fixed a whole bunch of
issues. Also make sure you use the latest X side driver too.

Cheers,
Ben.