[FFmpeg-cvslog] lavfi/gradfun: remove rounding to match C and SSE code.
Clément Bœsch
git at videolan.org
Wed Dec 19 03:18:27 CET 2012
ffmpeg | branch: master | Clément Bœsch <ubitux at gmail.com> | Fri Dec 7 00:41:04 2012 +0100| [63e1fc25884fe7cf3f66a8b99602ca98a47fb763] | committer: Clément Bœsch
lavfi/gradfun: remove rounding to match C and SSE code.
There is no noticable benefit for such precision.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=63e1fc25884fe7cf3f66a8b99602ca98a47fb763
---
libavfilter/x86/gradfun.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/libavfilter/x86/gradfun.c b/libavfilter/x86/gradfun.c
index be3e6a9..214e764 100644
--- a/libavfilter/x86/gradfun.c
+++ b/libavfilter/x86/gradfun.c
@@ -136,9 +136,9 @@ static void gradfun_filter_line_ssse3(uint8_t *dst, const uint8_t *src, const ui
"psubw %%xmm6, %%xmm2 \n"
"pminsw %%xmm7, %%xmm2 \n" // m = -max(0, 127-m)
"pmullw %%xmm2, %%xmm2 \n"
- "psllw $1, %%xmm2 \n"
+ "psllw $2, %%xmm1 \n"
"paddw %%xmm4, %%xmm0 \n" // pix += dither
- "pmulhrsw %%xmm2, %%xmm1 \n" // m = m*m*delta >> 14
+ "pmulhw %%xmm2, %%xmm1 \n" // m = m*m*delta >> 14
"paddw %%xmm1, %%xmm0 \n" // pix += m
"psraw $7, %%xmm0 \n"
"packuswb %%xmm0, %%xmm0 \n"
More information about the ffmpeg-cvslog
mailing list