Bug#369167: DRI probably broken: where to reportbug

[Michel Dänzer]

On Fri, 2006-10-13 at 18:21 -0400, David Nusinow wrote:
> On Tue, Jun 13, 2006 at 08:39:34AM +0200, Michel Dänzer wrote:
> > On Mon, 2006-06-12 at 20:19 -0400, David Nusinow wrote:
> > > On Mon, Jun 12, 2006 at 08:26:53AM +0200, Michel Dänzer wrote:
> > > > On Sun, 2006-06-11 at 17:45 -0400, David Nusinow wrote:
> > > > > 
> > > > > Should we ship our driver with this patch as well?
> > > > 
> > > > Eventually, yes, but I'm not sure it's correct for integrated chipsets
> > > > as is.
> > > 
> > > Ok, let me know. I actually have an IGP 340M here that I can test on if
> > > that's sufficient to make you feel more confident in it.
> > 
> > Oh, yes, would be awesome if you could test whether this patch has any
> > impact wrt DRI on that. At any rate though, please provide a full
> > logfile from running with this patch.
> 
> Ok, so I finally got my notebook up and running again, and tested the patch
> with the latest driver. It worked fine both with fb enabled and without.
> I've attached the server logs for both. 

Thanks, but it looks like the patch wasn't applied correctly; there
should be a line like

(II) RADEON(0): Detected total video RAM=131072K, accessible=131072K (PCI BAR=131072K)

before

(--) RADEON(0): Mapped VideoRAM: 32768 kByte (64 bit DDR SDRAM)

in both logs, but it's not in either.


> It's worth noting that using the framebuffer was insanely slow in comparison to 
> not using, although I don't know if this is normal or not. 

Looks like it doesn't enable write-combining for some reason with it
enabled.

> I'm committing the patch to our svn repo now, let me know if 
> you want me to upload it to unstable.

It should actually be tested first. :) Although AFAIK Fabio applied it
to the Ubuntu package a while ago and hasn't reported any regressions.


-- 
Earthling Michel Dänzer           |          http://tungstengraphics.com
Libre software enthusiast         |          Debian, X and DRI developer