yading@10: /*
yading@10:  * aligned/packed access motion
yading@10:  *
yading@10:  * Copyright (c) 2001-2003 BERO <bero@geocities.co.jp>
yading@10:  *
yading@10:  * This file is part of FFmpeg.
yading@10:  *
yading@10:  * FFmpeg is free software; you can redistribute it and/or
yading@10:  * modify it under the terms of the GNU Lesser General Public
yading@10:  * License as published by the Free Software Foundation; either
yading@10:  * version 2.1 of the License, or (at your option) any later version.
yading@10:  *
yading@10:  * FFmpeg is distributed in the hope that it will be useful,
yading@10:  * but WITHOUT ANY WARRANTY; without even the implied warranty of
yading@10:  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
yading@10:  * Lesser General Public License for more details.
yading@10:  *
yading@10:  * You should have received a copy of the GNU Lesser General Public
yading@10:  * License along with FFmpeg; if not, write to the Free Software
yading@10:  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
yading@10:  */
yading@10: 
yading@10: #include <assert.h>
yading@10: #include <stdint.h>
yading@10: 
yading@10: #include "libavutil/attributes.h"
yading@10: #include "libavcodec/h264chroma.h"
yading@10: 
yading@10: #define H264_CHROMA_MC(OPNAME, OP)\
yading@10: static void OPNAME ## h264_chroma_mc2_sh4(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){\
yading@10:     const int A=(8-x)*(8-y);\
yading@10:     const int B=(  x)*(8-y);\
yading@10:     const int C=(8-x)*(  y);\
yading@10:     const int D=(  x)*(  y);\
yading@10:     \
yading@10:     assert(x<8 && y<8 && x>=0 && y>=0);\
yading@10: \
yading@10:     do {\
yading@10:         int t0,t1,t2,t3; \
yading@10:         uint8_t *s0 = src; \
yading@10:         uint8_t *s1 = src+stride; \
yading@10:         t0 = *s0++; t2 = *s1++; \
yading@10:         t1 = *s0++; t3 = *s1++; \
yading@10:         OP(dst[0], (A*t0 + B*t1 + C*t2 + D*t3));\
yading@10:         t0 = *s0++; t2 = *s1++; \
yading@10:         OP(dst[1], (A*t1 + B*t0 + C*t3 + D*t2));\
yading@10:         dst+= stride;\
yading@10:         src+= stride;\
yading@10:     }while(--h);\
yading@10: }\
yading@10: \
yading@10: static void OPNAME ## h264_chroma_mc4_sh4(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){\
yading@10:     const int A=(8-x)*(8-y);\
yading@10:     const int B=(  x)*(8-y);\
yading@10:     const int C=(8-x)*(  y);\
yading@10:     const int D=(  x)*(  y);\
yading@10:     \
yading@10:     assert(x<8 && y<8 && x>=0 && y>=0);\
yading@10: \
yading@10:     do {\
yading@10:         int t0,t1,t2,t3; \
yading@10:         uint8_t *s0 = src; \
yading@10:         uint8_t *s1 = src+stride; \
yading@10:         t0 = *s0++; t2 = *s1++; \
yading@10:         t1 = *s0++; t3 = *s1++; \
yading@10:         OP(dst[0], (A*t0 + B*t1 + C*t2 + D*t3));\
yading@10:         t0 = *s0++; t2 = *s1++; \
yading@10:         OP(dst[1], (A*t1 + B*t0 + C*t3 + D*t2));\
yading@10:         t1 = *s0++; t3 = *s1++; \
yading@10:         OP(dst[2], (A*t0 + B*t1 + C*t2 + D*t3));\
yading@10:         t0 = *s0++; t2 = *s1++; \
yading@10:         OP(dst[3], (A*t1 + B*t0 + C*t3 + D*t2));\
yading@10:         dst+= stride;\
yading@10:         src+= stride;\
yading@10:     }while(--h);\
yading@10: }\
yading@10: \
yading@10: static void OPNAME ## h264_chroma_mc8_sh4(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){\
yading@10:     const int A=(8-x)*(8-y);\
yading@10:     const int B=(  x)*(8-y);\
yading@10:     const int C=(8-x)*(  y);\
yading@10:     const int D=(  x)*(  y);\
yading@10:     \
yading@10:     assert(x<8 && y<8 && x>=0 && y>=0);\
yading@10: \
yading@10:     do {\
yading@10:         int t0,t1,t2,t3; \
yading@10:         uint8_t *s0 = src; \
yading@10:         uint8_t *s1 = src+stride; \
yading@10:         t0 = *s0++; t2 = *s1++; \
yading@10:         t1 = *s0++; t3 = *s1++; \
yading@10:         OP(dst[0], (A*t0 + B*t1 + C*t2 + D*t3));\
yading@10:         t0 = *s0++; t2 = *s1++; \
yading@10:         OP(dst[1], (A*t1 + B*t0 + C*t3 + D*t2));\
yading@10:         t1 = *s0++; t3 = *s1++; \
yading@10:         OP(dst[2], (A*t0 + B*t1 + C*t2 + D*t3));\
yading@10:         t0 = *s0++; t2 = *s1++; \
yading@10:         OP(dst[3], (A*t1 + B*t0 + C*t3 + D*t2));\
yading@10:         t1 = *s0++; t3 = *s1++; \
yading@10:         OP(dst[4], (A*t0 + B*t1 + C*t2 + D*t3));\
yading@10:         t0 = *s0++; t2 = *s1++; \
yading@10:         OP(dst[5], (A*t1 + B*t0 + C*t3 + D*t2));\
yading@10:         t1 = *s0++; t3 = *s1++; \
yading@10:         OP(dst[6], (A*t0 + B*t1 + C*t2 + D*t3));\
yading@10:         t0 = *s0++; t2 = *s1++; \
yading@10:         OP(dst[7], (A*t1 + B*t0 + C*t3 + D*t2));\
yading@10:         dst+= stride;\
yading@10:         src+= stride;\
yading@10:     }while(--h);\
yading@10: }
yading@10: 
yading@10: #define op_avg(a, b) a = (((a)+(((b) + 32)>>6)+1)>>1)
yading@10: #define op_put(a, b) a = (((b) + 32)>>6)
yading@10: 
yading@10: H264_CHROMA_MC(put_       , op_put)
yading@10: H264_CHROMA_MC(avg_       , op_avg)
yading@10: #undef op_avg
yading@10: #undef op_put
yading@10: 
yading@10: av_cold void ff_h264chroma_init_sh4(H264ChromaContext *c, int bit_depth)
yading@10: {
yading@10:     const int high_bit_depth = bit_depth > 8;
yading@10: 
yading@10:     if (!high_bit_depth) {
yading@10:     c->put_h264_chroma_pixels_tab[0]= put_h264_chroma_mc8_sh4;
yading@10:     c->put_h264_chroma_pixels_tab[1]= put_h264_chroma_mc4_sh4;
yading@10:     c->put_h264_chroma_pixels_tab[2]= put_h264_chroma_mc2_sh4;
yading@10:     c->avg_h264_chroma_pixels_tab[0]= avg_h264_chroma_mc8_sh4;
yading@10:     c->avg_h264_chroma_pixels_tab[1]= avg_h264_chroma_mc4_sh4;
yading@10:     c->avg_h264_chroma_pixels_tab[2]= avg_h264_chroma_mc2_sh4;
yading@10:     }
yading@10: }