diff options
| author | Ronald S. Bultje <rsbultje@gmail.com> | 2010-09-29 14:03:30 +0000 |
|---|---|---|
| committer | Ronald S. Bultje <rsbultje@gmail.com> | 2010-09-29 14:03:30 +0000 |
| commit | 0cc8a5d0885396bfec7e02f1aea9e5eee386855a (patch) | |
| tree | 6b0caadccb1d4f0c286021796dd824d280018a48 | |
| parent | c0673f2cf4e2aab240af4ed3d94a8dac27114fef (diff) | |
| download | ffmpeg-0cc8a5d0885396bfec7e02f1aea9e5eee386855a.tar.gz | |
Remove mv_mask variable. Replace the related pand -1/0 instructions by either
a pxor, or remove the instruction alltogether. Altogether, this saves 1
instruction.
Originally committed as revision 25255 to svn://svn.ffmpeg.org/ffmpeg/trunk
| -rw-r--r-- | libavcodec/x86/h264dsp_mmx.c | 13 |
1 files changed, 7 insertions, 6 deletions
diff --git a/libavcodec/x86/h264dsp_mmx.c b/libavcodec/x86/h264dsp_mmx.c index 8743456bc7..92d08d90a9 100644 --- a/libavcodec/x86/h264dsp_mmx.c +++ b/libavcodec/x86/h264dsp_mmx.c @@ -66,14 +66,15 @@ void ff_h264_idct_add8_sse2 (uint8_t **dest, const int *block_offset, DCTEL static av_always_inline void h264_loop_filter_strength_iteration_mmx2(int16_t bS[2][4][4], uint8_t nnz[40], int8_t ref[2][40], int16_t mv[2][40][2], int bidir, int edges, int step, - int mask_mv, int dir, const int d_idx) + int mask_mv, int dir, const int d_idx, + const uint64_t mask_dir) { - DECLARE_ALIGNED(8, const uint64_t, mask_dir) = dir ? 0 : 0xffffffffffffffffULL; int b_idx, edge; for( b_idx=12, edge=0; edge<edges; edge+=step, b_idx+=8*step ) { + if (!mask_dir) __asm__ volatile( - "pand %0, %%mm0 \n\t" - ::"m"(mask_dir) + "pxor %%mm0, %%mm0 \n\t" + :: ); if(!(mask_mv & edge)) { if(bidir) { @@ -193,8 +194,8 @@ static void h264_loop_filter_strength_mmx2( int16_t bS[2][4][4], uint8_t nnz[40] // could do a special case for dir==0 && edges==1, but it only reduces the // average filter time by 1.2% - h264_loop_filter_strength_iteration_mmx2(bS, nnz, ref, mv, bidir, edges, step, mask_mv1, 1, -8); - h264_loop_filter_strength_iteration_mmx2(bS, nnz, ref, mv, bidir, 4, 1, mask_mv0, 0, -1); + h264_loop_filter_strength_iteration_mmx2(bS, nnz, ref, mv, bidir, edges, step, mask_mv1, 1, -8, 0); + h264_loop_filter_strength_iteration_mmx2(bS, nnz, ref, mv, bidir, 4, 1, mask_mv0, 0, -1, -1); __asm__ volatile( "movq (%0), %%mm0 \n\t" |