[PATCH] new radeon memory map fixes

[Benjamin Herrenschmidt]

On Sat, 2006-02-04 at 07:41 -0800, Jesse Barnes wrote:
> On Wednesday, January 25, 2006 3:43 pm, Benjamin Herrenschmidt wrote:
> > Ok, so finally here is a new version of the patch. This time, it's
> > against modular and it comes with a DRM patch. The X driver and the
> > DRM patch should both work with the unpatched counterpart though
> > you'll only get the full benefit of the fixes with both patches
> > applied.
> 
> Ah, modular is *so* much nicer to work with.  I was able to just rebuild 
> my ATI driver and test it out without too much pain.
> 
> > As I had to shuffle a lot of code around in the X driver, there may
> > still be bugs lurking around. Especially look for regressions around
> > Xinerama and MergedFB as I haven't yet had a chance to test with
> > those (especially Xinerama is doing a lot of very dodgy stuffs in the
> > radeon driver).
> >
> > Please, try to test all sort of combinations of color tiling on/off,
> > dri enabled/disabled, bit depth, hw/sw cursor etc...
> 
> You mentioned that this might help with the VT switch problem on my 
> pbook.  Unfortunately, it doesn't.  I have MergedFB off, but other than 
> that I'm using stock settings.  The server seems to run ok, but if I 
> switch to a console VT (ctl-alt-f1) then back to the X server 
> (ctl-alt-f7) the machine hangs in a bad way and display gets corrupted 
> (turns rainbow colors).
> 
> Is this a known bug?  Has anyone else seen this behavior?

Hrm.. it works on mine which happens to be the same model... Can you
send me your xorg.conf and your dmesg log ? Also, do you have DRI
enabled, and if yes, do you have the matching DRI patch ?

Ben