Re: ATI M9 DRM issues,
On Thu, 2003-09-04 at 17:54, Aaron Van Couwenberghe wrote:
> Hi all!
>
> I've managed to get my ATI Mobility Radeon 9000 to work great under X
> finally - after a good deal of tweaking, of course. the 2.4.22 radeonfb
> driver likes my card, the ati driver from X 4.3.0 works - I installed of
> course the radeon DRM driver (radeon-20030901-linux.i386.tar -- DRI
> project CVS snap).
>
> I'm using Daniel Stone's XF 4.3.0 packages -- these are of course built
> on woody and I'm running Sarge, but I can't imagine this causing any
> problems. I will rebuild locally to be sure but I honestly doubt this
> will solve anything.
If you are using X 4.3.0 then the standard drm kernel drivers won't work
and you need to get updated drivers. look in www.dri.org iirc for
updated drivers.
Standard kernel drivers are only for X 4.2 (And maybe 4.1 don't know).
>
> Now on to the issue: When I startx, any time I have Option "UseFBDev"
> "true" in my device section, I get this wierd error message about not
> being able to probe the VGA bios in PCI space. odd. And the server dies.
>
> If I turn off the use of the framebuffer, my server starts great, and
> it's fully 2d accelerated. fairly easy to tell by running aviplay - it
> renders fast and uses almost no CPU time. However, glxinfo reports
> "direct rendering: No" and any 3d applications run slow as molasses.
>
> I don't know a whole lot about X, but is turning off the FBDev nuking my
> 3d accel, or something else? And if that's it, how do I fix this probing
> problem? XF86Config-4, startx output, and the error logfile are attached
> below.
>
> Also I have this wierd thing happening - when i exit X my framebuffer
> resolution doesn't change -- and I get a whole bunch of tiled (640x480?)
> consoles on a 1600x1200 display. Bit difficult to see :P. How do i
> change a fbdev's resolution?
>
> Thanks in advance!
> -Aaron Van Couwenberghe
>
> PS note of course the only line different in my alternate XF86Config is
> to replace "false" with "true" in the "UseFBDev" line of my device
> section.
>
--
Micha Feigin
michf@math.tau.ac.il
Reply to: