[Pixman] [PATCH] sse2: Implement simple bilinear scaling for x8r8g8b8 to a8r8g8b8

[Chris Wilson]

On Wed, Jan 23, 2013 at 11:51:37AM -0800, Matt Turner wrote:
> On Wed, Jan 23, 2013 at 6:37 AM, Chris Wilson <chris at chris-wilson.co.uk> wrote:
> > Improves firefon-tron on a IVB i7-3720qm: 68.6s to 45.2s.
> >
> > Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> > ---
> >  pixman/pixman-sse2.c |   63 ++++++++++++++++++++++++++++++++++++++++++++++++++
> >  1 file changed, 63 insertions(+)
> >
> > diff --git a/pixman/pixman-sse2.c b/pixman/pixman-sse2.c
> > index fc873cc..bc3e2f1 100644
> > --- a/pixman/pixman-sse2.c
> > +++ b/pixman/pixman-sse2.c
> > @@ -5679,6 +5679,67 @@ FAST_BILINEAR_MAINLOOP_COMMON (sse2_8888_8888_normal_SRC,
> >                                NORMAL, FLAG_NONE)
> >
> >  static force_inline void
> > +scaled_bilinear_scanline_sse2_0888_8888_SRC (uint32_t *       dst,
> 
> Maybe some funny whitespace before dst? Or maybe just a spaces vs tabs issue.

Historical accident, this was cut'n'paste from 
scaled_bilinear_scanline_sse2_8888_8888_SRC, which at least means that
the function is consistent with all the other
scaled_bilinear_scanline_sse2_*() :)
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre