F12/Rawhide+nomodset vs. F11 radeon r700 support

[François Cami]

On Tue, 05 Jan 2010 19:56:54 +0100
"Dennis J." <dennisml at conversis.de> wrote:

> On 01/05/2010 06:56 PM, Adam Williamson wrote:
> > On Tue, 2010-01-05 at 16:48 +0000, "Jóhann B. Guðmundsson" wrote:
> >> this is and the
> >> output from lspci
> >
> > there's no need for this on X.org bugs, the Xorg.0.log file contains
> > sufficient information on the graphics hardware.
> 
> This bug seems to be both a KMS bug and a Xorg bug. I can get boot into 
> runlevel 3 but the display isn't initialized properly and the only way I 
> got some of the debug information attached to the bug was by blindly 
> logging in, setting up sshd and then logging in from another machine.
> 
> The second problem is that when I boot with nomodeset I get to the desktop 
> but it is so slow it cannot be used for regular work. I'm not sure if that 
> is a bug or because the radeon driver now requires KMS for any sort of 2D 
> acceleration.

On most radeons, both UMS and KMS are accelerated on Fedora 12.
That may not be the case for R700, and at any rate we'd rather have
the KMS codepath working than the older UMS.
If booting into KMS/runlevel 3 fails, it's a KMS problem, Xorg has no
way to recover from that.
On the other hand, your logs show both nouveau and radeon, as pointed
out by Jóhann. Could you try booting with the parameter:
nouveau.modeset=0
and report?

François