diff options
| author | Michael Niedermayer <michaelni@gmx.at> | 2007-12-21 10:16:22 +0000 |
|---|---|---|
| committer | Michael Niedermayer <michaelni@gmx.at> | 2007-12-21 10:16:22 +0000 |
| commit | febdd0b9b414b2adfe4c3ea2fa06ded2e32e510b (patch) | |
| tree | 8d2aa283115c22ad6462897ae110b627b531ce81 | |
| parent | a06a18c55dfade76ba5f560d9d8be71951638792 (diff) | |
| download | ffmpeg-febdd0b9b414b2adfe4c3ea2fa06ded2e32e510b.tar.gz | |
~15% faster h264_chroma_mc2/4_c() these also prevent some possible out
of array reads.
Originally committed as revision 11290 to svn://svn.ffmpeg.org/ffmpeg/trunk
| -rw-r--r-- | libavcodec/dsputil.c | 26 |
1 files changed, 26 insertions, 0 deletions
diff --git a/libavcodec/dsputil.c b/libavcodec/dsputil.c index e3b5db7479..523921ec17 100644 --- a/libavcodec/dsputil.c +++ b/libavcodec/dsputil.c @@ -1440,6 +1440,7 @@ static void OPNAME ## h264_chroma_mc2_c(uint8_t *dst/*align 8*/, uint8_t *src/*a \ assert(x<8 && y<8 && x>=0 && y>=0);\ \ + if(D){\ for(i=0; i<h; i++)\ {\ OP(dst[0], (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1]));\ @@ -1447,6 +1448,17 @@ static void OPNAME ## h264_chroma_mc2_c(uint8_t *dst/*align 8*/, uint8_t *src/*a dst+= stride;\ src+= stride;\ }\ + }else{\ + const int E= B+C;\ + const int step= C ? stride : 1;\ + for(i=0; i<h; i++)\ + {\ + OP(dst[0], (A*src[0] + E*src[step+0]));\ + OP(dst[1], (A*src[1] + E*src[step+1]));\ + dst+= stride;\ + src+= stride;\ + }\ + }\ }\ \ static void OPNAME ## h264_chroma_mc4_c(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){\ @@ -1458,6 +1470,7 @@ static void OPNAME ## h264_chroma_mc4_c(uint8_t *dst/*align 8*/, uint8_t *src/*a \ assert(x<8 && y<8 && x>=0 && y>=0);\ \ + if(D){\ for(i=0; i<h; i++)\ {\ OP(dst[0], (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1]));\ @@ -1467,6 +1480,19 @@ static void OPNAME ## h264_chroma_mc4_c(uint8_t *dst/*align 8*/, uint8_t *src/*a dst+= stride;\ src+= stride;\ }\ + }else{\ + const int E= B+C;\ + const int step= C ? stride : 1;\ + for(i=0; i<h; i++)\ + {\ + OP(dst[0], (A*src[0] + E*src[step+0]));\ + OP(dst[1], (A*src[1] + E*src[step+1]));\ + OP(dst[2], (A*src[2] + E*src[step+2]));\ + OP(dst[3], (A*src[3] + E*src[step+3]));\ + dst+= stride;\ + src+= stride;\ + }\ + }\ }\ \ static void OPNAME ## h264_chroma_mc8_c(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){\ |