diff options
| author | Daniel Kang <daniel.d.kang@gmail.com> | 2011-07-03 14:56:09 -0400 |
|---|---|---|
| committer | Ronald S. Bultje <rsbultje@gmail.com> | 2011-07-08 15:59:29 -0700 |
| commit | c0483d0c7a8f12564f615dcd73890bcc2f7c2796 (patch) | |
| tree | 6e3235406090fbfa18849d5eb5a4f4a723629163 /libavcodec/x86/h264_intrapred_init.c | |
| parent | 5c391a161a1f71dcdafd3f59b45d80220845f501 (diff) | |
| download | ffmpeg-streaming-c0483d0c7a8f12564f615dcd73890bcc2f7c2796.zip ffmpeg-streaming-c0483d0c7a8f12564f615dcd73890bcc2f7c2796.tar.gz | |
H.264: Add x86 assembly for 10-bit H.264 predict functions
Mainly ported from 8-bit H.264 predict.
Some code ported from x264. LGPL ok by author.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
Diffstat (limited to 'libavcodec/x86/h264_intrapred_init.c')
| -rw-r--r-- | libavcodec/x86/h264_intrapred_init.c | 58 |
1 files changed, 58 insertions, 0 deletions
diff --git a/libavcodec/x86/h264_intrapred_init.c b/libavcodec/x86/h264_intrapred_init.c index 9d6726c..a0c5164 100644 --- a/libavcodec/x86/h264_intrapred_init.c +++ b/libavcodec/x86/h264_intrapred_init.c @@ -43,9 +43,41 @@ PRED4x4(horizontal_down, 10, avx) #define PRED8x8(TYPE, DEPTH, OPT) \ void ff_pred8x8_ ## TYPE ## _ ## DEPTH ## _ ## OPT (uint8_t *src, int stride); +PRED8x8(dc, 10, mmxext) +PRED8x8(dc, 10, sse2) +PRED8x8(top_dc, 10, mmxext) +PRED8x8(top_dc, 10, sse2) PRED8x8(vertical, 10, sse2) PRED8x8(horizontal, 10, sse2) +#define PRED8x8L(TYPE, DEPTH, OPT)\ +void ff_pred8x8l_ ## TYPE ## _ ## DEPTH ## _ ## OPT (uint8_t *src, int has_topleft, int has_topright, int stride); + +PRED8x8L(dc, 10, sse2) +PRED8x8L(dc, 10, ssse3) +PRED8x8L(top_dc, 10, sse2) +PRED8x8L(top_dc, 10, ssse3) +PRED8x8L(vertical, 10, sse2) +PRED8x8L(vertical, 10, ssse3) +PRED8x8L(horizontal, 10, sse2) +PRED8x8L(horizontal, 10, ssse3) +PRED8x8L(down_left, 10, sse2) +PRED8x8L(down_left, 10, ssse3) +PRED8x8L(down_right, 10, sse2) +PRED8x8L(down_right, 10, ssse3) +PRED8x8L(vertical_right, 10, sse2) +PRED8x8L(vertical_right, 10, ssse3) +PRED8x8L(horizontal_up, 10, sse2) +PRED8x8L(horizontal_up, 10, ssse3) + +#define PRED16x16(TYPE, DEPTH, OPT)\ +void ff_pred16x16_ ## TYPE ## _ ## DEPTH ## _ ## OPT (uint8_t *src, int stride); + +PRED16x16(vertical, 10, mmxext) +PRED16x16(vertical, 10, sse2) +PRED16x16(horizontal, 10, mmxext) +PRED16x16(horizontal, 10, sse2) + void ff_pred16x16_vertical_mmx (uint8_t *src, int stride); void ff_pred16x16_vertical_sse (uint8_t *src, int stride); void ff_pred16x16_horizontal_mmx (uint8_t *src, int stride); @@ -253,6 +285,12 @@ void ff_h264_pred_init_x86(H264PredContext *h, int codec_id, const int bit_depth if (mm_flags & AV_CPU_FLAG_MMX2) { h->pred4x4[DC_PRED ] = ff_pred4x4_dc_10_mmxext; h->pred4x4[HOR_UP_PRED ] = ff_pred4x4_horizontal_up_10_mmxext; + + h->pred8x8[DC_PRED8x8 ] = ff_pred8x8_dc_10_mmxext; + h->pred8x8[TOP_DC_PRED8x8 ] = ff_pred8x8_top_dc_10_mmxext; + + h->pred16x16[VERT_PRED8x8 ] = ff_pred16x16_vertical_10_mmxext; + h->pred16x16[HOR_PRED8x8 ] = ff_pred16x16_horizontal_10_mmxext; } if (mm_flags & AV_CPU_FLAG_SSE2) { h->pred4x4[DIAG_DOWN_LEFT_PRED ] = ff_pred4x4_down_left_10_sse2; @@ -261,13 +299,33 @@ void ff_h264_pred_init_x86(H264PredContext *h, int codec_id, const int bit_depth h->pred4x4[VERT_RIGHT_PRED ] = ff_pred4x4_vertical_right_10_sse2; h->pred4x4[HOR_DOWN_PRED ] = ff_pred4x4_horizontal_down_10_sse2; + h->pred8x8[DC_PRED8x8 ] = ff_pred8x8_dc_10_sse2; + h->pred8x8[TOP_DC_PRED8x8 ] = ff_pred8x8_top_dc_10_sse2; h->pred8x8[VERT_PRED8x8 ] = ff_pred8x8_vertical_10_sse2; h->pred8x8[HOR_PRED8x8 ] = ff_pred8x8_horizontal_10_sse2; + + h->pred8x8l[VERT_PRED ] = ff_pred8x8l_vertical_10_sse2; + h->pred8x8l[HOR_PRED ] = ff_pred8x8l_horizontal_10_sse2; + h->pred8x8l[DC_PRED ] = ff_pred8x8l_dc_10_sse2; + h->pred8x8l[TOP_DC_PRED ] = ff_pred8x8l_top_dc_10_sse2; + h->pred8x8l[DIAG_DOWN_LEFT_PRED ] = ff_pred8x8l_down_left_10_sse2; + h->pred8x8l[DIAG_DOWN_RIGHT_PRED] = ff_pred8x8l_down_right_10_sse2; + h->pred8x8l[VERT_RIGHT_PRED ] = ff_pred8x8l_vertical_right_10_sse2; + h->pred8x8l[HOR_UP_PRED ] = ff_pred8x8l_horizontal_up_10_sse2; + + h->pred16x16[VERT_PRED8x8 ] = ff_pred16x16_vertical_10_sse2; + h->pred16x16[HOR_PRED8x8 ] = ff_pred16x16_horizontal_10_sse2; } if (mm_flags & AV_CPU_FLAG_SSSE3) { h->pred4x4[DIAG_DOWN_RIGHT_PRED] = ff_pred4x4_down_right_10_ssse3; h->pred4x4[VERT_RIGHT_PRED ] = ff_pred4x4_vertical_right_10_ssse3; h->pred4x4[HOR_DOWN_PRED ] = ff_pred4x4_horizontal_down_10_ssse3; + + h->pred8x8l[VERT_PRED ] = ff_pred8x8l_vertical_10_ssse3; + h->pred8x8l[HOR_PRED ] = ff_pred8x8l_horizontal_10_ssse3; + h->pred8x8l[DC_PRED ] = ff_pred8x8l_dc_10_ssse3; + h->pred8x8l[TOP_DC_PRED ] = ff_pred8x8l_top_dc_10_ssse3; + h->pred8x8l[DIAG_DOWN_LEFT_PRED ] = ff_pred8x8l_down_left_10_ssse3; } #if HAVE_AVX if (mm_flags & AV_CPU_FLAG_AVX) { |
