modular: Changes to 'master'

[Luc Verhaegen]

On Tue, Oct 21, 2008 at 05:16:18PM +0100, Daniel Stone wrote:
> Hi,
> 
> On Tue, Oct 21, 2008 at 06:08:49PM +0200, Luc Verhaegen wrote:
> > 
> > Once again, a very unbiased opinion by Mr Stone.
> 
> I'm not sure what you're trying to say, but thanks.

This isn't exactly a first. Especially in this whole story, but also in 
others.
 
> On Tue, Oct 21, 2008 at 06:00:44PM +0200, Erik Andrén wrote:
> > Could someone enlighten me why there are two radeon drivers in the first place?
> 
> Originally, the key differentiator was the lack of ATOMBIOS support in
> radeonhd.  Then radeonhd had ATOM support forced into it.

For newer hardware only. Note also that no register level information 
has been made available for such hardware.

> It's got its
> own internal infrastructure that isn't RandR 1.2 because RandR 1.2 sucks
> and will kill us all, or something, but the only thing it has mapped on
> to it is ... RandR 1.2.

Yeah, hrm... modern hw really maps to randr 1.2... DCE 3.2 is fun.

> It now has EXA, DRI and Xv code copy and pasted
> from Radeon.

Try again.

> There is the CS (command submission) infrastructure, so if you
> desperately want 3D support without a DRM, radeonhd is the market leader.

You clearly haven't been watching this code at all. 
 
> 
> On Tue, Oct 21, 2008 at 06:15:58PM +0200, Luc Verhaegen wrote:
> > Is there a single technical reason why shipping both is a problem?
> 
> If you're asking whether or not annarchy will blow up if we ship both,
> whether or not the server will explode in the face of two drivers with
> an identical prefix, etc, then the answer is no.  But I don't think
> that's what you were trying to ask.

So what stops it from being shipped as well?

Nothing. Just you.

Luc Verhaegen.
SUSE X Driver Developer.