On Tue, 06 Apr 2004 20:27:39 -0700
Ian Romanick <[EMAIL PROTECTED]> wrote:

> Dieter N�tzel wrote:
> 
> > Anyone working on the "t_vertex" conversion for radeon (r200)?
> > Then Andreas could do his TMU3 work, again. Which could help with Viewperf.
> 
> I started looking at this today.  I decided to start with just the swtcl 
> path.  I'm a bit confused (already).
> 
> All of the drivers that have previously been converted use t_dd_tritmp.h 
> as their template.  The r200 (and radeon) drivers use t_dd_vbtmp.h.  It 
> looks like a lot of the mechanism used in r200_swtcl.c and t_dd_vbtmp.h 
> is, basically, an earlier version of t_vertex.  Is that fairly accurate? 
>   If that's the case, should I convert the r200 driver to use t_dd_tritmp.h?

The driver is already using t_dd_tritmp.h. Pre t_vertex drivers use
t_dd_tritmp.h for submitting triangles (and other "simple" primitives)
from the tnl render stage and t_dd_vbtmp.h for vertex setup. I'm not
sure in which cases t_dd_tritmp is actually used, that is, the fallback
cases of the r200's own swtcl render stage.

The t_vertex stuff replaces t_dd_vbtmp stuff. I've done this for the
savage driver and it was pretty straight forward. t_dd_tritmp.h will
still be needed for the tnl render stage. I don't know how t_vertex is
going to be integrated in the r200's swtcl render stage.

> 
> I think I need to dig through the code more to get a better 
> understanding of what's really going on...
> 

Felix


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id70&alloc_id638&op=click
--
_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel

Reply via email to