diff options
| author | Ronald S. Bultje <rsbultje@gmail.com> | 2013-02-18 21:03:02 -0800 |
|---|---|---|
| committer | Martin Storsjö <martin@martin.st> | 2013-04-10 11:03:06 +0300 |
| commit | 62844c3fd66940c7747e9b2bb7804e265319f43f (patch) | |
| tree | b0e5a05644457aa5d7598d1fefa1a41f83550753 /libavcodec/h264dsp.c | |
| parent | e8cafd2773bc56455c8816593cbd9368f2d69a80 (diff) | |
| download | ffmpeg-62844c3fd66940c7747e9b2bb7804e265319f43f.tar.gz | |
h264: Integrate clear_blocks calls with IDCT
The non-intra-pcm branch in hl_decode_mb (simple, 8bpp) goes from 700
to 672 cycles, and the complete loop of decode_mb_cabac and hl_decode_mb
(in the decode_slice loop) goes from 1759 to 1733 cycles on the clip
tested (cathedral), i.e. almost 30 cycles per mb faster.
Signed-off-by: Martin Storsjö <martin@martin.st>
Diffstat (limited to 'libavcodec/h264dsp.c')
| -rw-r--r-- | libavcodec/h264dsp.c | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/libavcodec/h264dsp.c b/libavcodec/h264dsp.c index b739758e24..8a70aefedc 100644 --- a/libavcodec/h264dsp.c +++ b/libavcodec/h264dsp.c @@ -57,8 +57,8 @@ void ff_h264dsp_init(H264DSPContext *c, const int bit_depth, const int chroma_fo #define FUNC(a, depth) a ## _ ## depth ## _c #define ADDPX_DSP(depth) \ - c->h264_add_pixels4 = FUNC(ff_h264_add_pixels4, depth);\ - c->h264_add_pixels8 = FUNC(ff_h264_add_pixels8, depth) + c->h264_add_pixels4_clear = FUNC(ff_h264_add_pixels4, depth);\ + c->h264_add_pixels8_clear = FUNC(ff_h264_add_pixels8, depth) if (bit_depth > 8 && bit_depth <= 16) { ADDPX_DSP(16); |