On Mon, Jun 16, 2003 at 02:31:45PM +0100, José Fonseca wrote: > On Mon, Jun 16, 2003 at 01:48:35PM +0200, Martin Spott wrote:
> > (II) RADEON(0): Direct rendering disabled > > This can't be. This log can't possible match the other one you gave me. > In the /var/log/messages it showed the agp buffer being added and so on. > > The only explanantion I can think of is that you're running _two_ X > servers. The first one acquires the AGP device but the second fails to > acquire, since the AGP device is already acquired by the first. The explanation sounds reasonable, but I have to disappoint you: There is definitely no second X server running, there also was only one XFree86.0.log because - to make shure not to mix up things - I removed every log that existred previously. How would you explain that the whole scene changes simply by loading a different 'radeon' kernel module ? Let me see, I'll strip down the kernel to only the necessary things and then I'll see if the 'radeon' module collides with anything else. I've attached the DRM patch I'm talking about the whole time. You might compare it with yours to see, if my CVS tree is missing some part. The patch applies against stock 2.4.20 kernel DRM. I can understand that you're somewhat confused. I'll do anything to resolve this issue as my time permits - if you like, Martin. -- Unix _IS_ user friendly - it's just selective about who its friends are ! --------------------------------------------------------------------------
DRM.gz
Description: application/gunzip
