Re: [AD] blender operations SRC and DST

[Peter Wang <novalazy@xxxxxxxxxx>]

On 2011-06-08, allefant <info@xxxxxxxxxx> wrote:
> On Tue, Jun 7, 2011 at 8:18 PM, Jon Rafkind <workmin@xxxxxxxxxx> wrote:
> 
> >  I changed the patch further:
> >  * altered internal/aintern_blend.h to deal with the new factors
> >  * al_blend_memory just calls _al_blend_inline now instead of having its
> > own implementation of blending
> >
> > Here is ex_blend_bench before and after the patch (summary is 2-5%
> > slowdown). I have a 1.8ghz amd processor.
> > no change:
> >  Plain blit: 213.454 FPS
> >  Scaled blit: 213.883 FPS
> >  Rotated blit: 223.992 FPS
> >
> > my change:
> >  Plain blit: 201.183 FPS
> >  Scaled blit: 208.2 FPS
> >  Rotated blit: 214.747 FPS
> >
> >
> 2-5% sounds harmless enough. But maybe wait for Peter, he spent a lot of
> effort into optimizing this and might thing differently :)

It's much worse on my machine, 10-12%.  I'd want to see what we can do
to avoid that kind of regression.

Peter

before patch:
Plain blit: 659.283 FPS
Scaled blit: 663.27 FPS
Rotated blit: 678.819 FPS

after patch:
Plain blit: 576.974 FPS
Scaled blit: 592.111 FPS
Rotated blit: 610.166 FPS