radeon EXA & fallbacks

[Benjamin Herrenschmidt]

On Thu, 2005-09-22 at 13:54 +0200, Roland Scheidegger wrote:
> Eric Anholt wrote:
> > Yep, you've got subpixel antialiasing on your fonts, right?  This is 
> > really hard to accelerate.  Basically you've got *two* ARGB8888
> > values for source that need to get passed in to the blending along
> > with the dest, which the card's BLEND_CNTL-type blending doesn't
> > support.
> Sounds pretty terrible. I don't think there is ANY hardware which could
> support this directly (in the alpha blend stage), blending is just not 
> defined that way. Though in this example the source seems to be 1x1 
> sized so maybe it could be used as the fixed blend color instead (only 
> on r200, not r100, however)? Not sure though, I've no idea what sort of 
> blend operation exa wants to do there, if it's a dot product it still 
> won't work :-).

Looking at the log, there's indeed a lot of those 1x1 source, so a
special case for this would make sense...

Ben.