ffmpeg | branch: master | Peter Cordes <[email protected]> | Thu Jan 29 17:20:27 2015 -0400| [9e5687adf20aff27cd4922d0f68f54c9622418bd] | committer: Michael Niedermayer
pixelutils: Comment on (lack of) sad_8x8_sse2 Signed-off-by: Peter Cordes <[email protected]> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=9e5687adf20aff27cd4922d0f68f54c9622418bd --- libavutil/x86/pixelutils_init.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/libavutil/x86/pixelutils_init.c b/libavutil/x86/pixelutils_init.c index d600510..c24a533 100644 --- a/libavutil/x86/pixelutils_init.c +++ b/libavutil/x86/pixelutils_init.c @@ -43,6 +43,12 @@ void ff_pixelutils_sad_init_x86(av_pixelutils_sad_fn *sad, int aligned) sad[2] = ff_pixelutils_sad_8x8_mmx; } + // The best way to use SSE2 would be to do 2 SADs in parallel, + // but we'd have to modify the pixelutils API to return SIMD functions. + + // It's probably not faster to shuffle data around + // to get two lines of 8 pixels into a single 16byte register, + // so just use the MMX 8x8 version even when SSE2 is available. if (EXTERNAL_MMXEXT(cpu_flags)) { sad[2] = ff_pixelutils_sad_8x8_mmxext; sad[3] = ff_pixelutils_sad_16x16_mmxext; _______________________________________________ ffmpeg-cvslog mailing list [email protected] http://ffmpeg.org/mailman/listinfo/ffmpeg-cvslog
