diff options
| author | Janne Grunau <janne-libav@jannau.net> | 2014-01-10 14:07:24 +0100 |
|---|---|---|
| committer | Janne Grunau <janne-libav@jannau.net> | 2014-01-15 12:31:07 +0100 |
| commit | f896bca03fc63b93851c1c14c9321c20b3cd44a6 (patch) | |
| tree | 082492458722a4699eca8d2b795299a703d3df34 /libavcodec/aarch64/h264dsp_init_aarch64.c | |
| parent | 36e3b1f2fd262028834a9d7b1eb533c1218ee6c2 (diff) | |
| download | ffmpeg-f896bca03fc63b93851c1c14c9321c20b3cd44a6.tar.gz | |
aarch64: h264 (bi)weight NEON optimizations
Ported from ARMv7 NEON.
Diffstat (limited to 'libavcodec/aarch64/h264dsp_init_aarch64.c')
| -rw-r--r-- | libavcodec/aarch64/h264dsp_init_aarch64.c | 25 |
1 files changed, 25 insertions, 0 deletions
diff --git a/libavcodec/aarch64/h264dsp_init_aarch64.c b/libavcodec/aarch64/h264dsp_init_aarch64.c index 307b30cd38..b106f11134 100644 --- a/libavcodec/aarch64/h264dsp_init_aarch64.c +++ b/libavcodec/aarch64/h264dsp_init_aarch64.c @@ -34,6 +34,23 @@ void ff_h264_v_loop_filter_chroma_neon(uint8_t *pix, int stride, int alpha, void ff_h264_h_loop_filter_chroma_neon(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0); +void ff_weight_h264_pixels_16_neon(uint8_t *dst, int stride, int height, + int log2_den, int weight, int offset); +void ff_weight_h264_pixels_8_neon(uint8_t *dst, int stride, int height, + int log2_den, int weight, int offset); +void ff_weight_h264_pixels_4_neon(uint8_t *dst, int stride, int height, + int log2_den, int weight, int offset); + +void ff_biweight_h264_pixels_16_neon(uint8_t *dst, uint8_t *src, int stride, + int height, int log2_den, int weightd, + int weights, int offset); +void ff_biweight_h264_pixels_8_neon(uint8_t *dst, uint8_t *src, int stride, + int height, int log2_den, int weightd, + int weights, int offset); +void ff_biweight_h264_pixels_4_neon(uint8_t *dst, uint8_t *src, int stride, + int height, int log2_den, int weightd, + int weights, int offset); + void ff_h264_idct_add_neon(uint8_t *dst, int16_t *block, int stride); void ff_h264_idct_dc_add_neon(uint8_t *dst, int16_t *block, int stride); void ff_h264_idct_add16_neon(uint8_t *dst, const int *block_offset, @@ -63,6 +80,14 @@ av_cold void ff_h264dsp_init_aarch64(H264DSPContext *c, const int bit_depth, c->h264_v_loop_filter_chroma = ff_h264_v_loop_filter_chroma_neon; c->h264_h_loop_filter_chroma = ff_h264_h_loop_filter_chroma_neon; + c->weight_h264_pixels_tab[0] = ff_weight_h264_pixels_16_neon; + c->weight_h264_pixels_tab[1] = ff_weight_h264_pixels_8_neon; + c->weight_h264_pixels_tab[2] = ff_weight_h264_pixels_4_neon; + + c->biweight_h264_pixels_tab[0] = ff_biweight_h264_pixels_16_neon; + c->biweight_h264_pixels_tab[1] = ff_biweight_h264_pixels_8_neon; + c->biweight_h264_pixels_tab[2] = ff_biweight_h264_pixels_4_neon; + c->h264_idct_add = ff_h264_idct_add_neon; c->h264_idct_dc_add = ff_h264_idct_dc_add_neon; c->h264_idct_add16 = ff_h264_idct_add16_neon; |