| Line | Branch | Exec | Source |
|---|---|---|---|
| 1 | /* | ||
| 2 | * H.26L/H.264/AVC/JVT/14496-10/... encoder/decoder | ||
| 3 | * Copyright (c) 2003-2010 Michael Niedermayer <michaelni@gmx.at> | ||
| 4 | * | ||
| 5 | * This file is part of FFmpeg. | ||
| 6 | * | ||
| 7 | * FFmpeg is free software; you can redistribute it and/or | ||
| 8 | * modify it under the terms of the GNU Lesser General Public | ||
| 9 | * License as published by the Free Software Foundation; either | ||
| 10 | * version 2.1 of the License, or (at your option) any later version. | ||
| 11 | * | ||
| 12 | * FFmpeg is distributed in the hope that it will be useful, | ||
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | ||
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | ||
| 15 | * Lesser General Public License for more details. | ||
| 16 | * | ||
| 17 | * You should have received a copy of the GNU Lesser General Public | ||
| 18 | * License along with FFmpeg; if not, write to the Free Software | ||
| 19 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA | ||
| 20 | */ | ||
| 21 | |||
| 22 | #include "libavutil/common.h" | ||
| 23 | #include "libavutil/intreadwrite.h" | ||
| 24 | |||
| 25 | #include "bit_depth_template.c" | ||
| 26 | #include "hpel_template.c" | ||
| 27 | #include "pel_template.c" | ||
| 28 | |||
| 29 | 65508 | static inline void FUNC(copy_block2)(uint8_t *dst, const uint8_t *restrict src, int dstStride, int srcStride, int h) | |
| 30 | { | ||
| 31 | int i; | ||
| 32 |
2/2✓ Branch 0 taken 458556 times.
✓ Branch 1 taken 65508 times.
|
524064 | for(i=0; i<h; i++) |
| 33 | { | ||
| 34 | 458556 | AV_WN2P(dst , AV_RN2P(src )); | |
| 35 | 458556 | dst+=dstStride; | |
| 36 | 458556 | src+=srcStride; | |
| 37 | } | ||
| 38 | 65508 | } | |
| 39 | |||
| 40 | 25376984 | static inline void FUNC(copy_block4)(uint8_t *dst, const uint8_t *restrict src, int dstStride, int srcStride, int h) | |
| 41 | { | ||
| 42 | int i; | ||
| 43 |
2/2✓ Branch 0 taken 114196428 times.
✓ Branch 1 taken 12688492 times.
|
253769840 | for(i=0; i<h; i++) |
| 44 | { | ||
| 45 | 228392856 | AV_WN4P(dst , AV_RN4P(src )); | |
| 46 | 228392856 | dst+=dstStride; | |
| 47 | 228392856 | src+=srcStride; | |
| 48 | } | ||
| 49 | 25376984 | } | |
| 50 | |||
| 51 | 13944600 | static inline void FUNC(copy_block8)(uint8_t *dst, const uint8_t *restrict src, int dstStride, int srcStride, int h) | |
| 52 | { | ||
| 53 | int i; | ||
| 54 |
2/2✓ Branch 0 taken 90639900 times.
✓ Branch 1 taken 6972300 times.
|
195224400 | for(i=0; i<h; i++) |
| 55 | { | ||
| 56 | 181279800 | AV_WN4P(dst , AV_RN4P(src )); | |
| 57 | 181279800 | AV_WN4P(dst+4*sizeof(pixel), AV_RN4P(src+4*sizeof(pixel))); | |
| 58 | 181279800 | dst+=dstStride; | |
| 59 | 181279800 | src+=srcStride; | |
| 60 | } | ||
| 61 | 13944600 | } | |
| 62 | |||
| 63 | 6735230 | static inline void FUNC(copy_block16)(uint8_t *dst, const uint8_t *restrict src, int dstStride, int srcStride, int h) | |
| 64 | { | ||
| 65 | int i; | ||
| 66 |
2/2✓ Branch 0 taken 70719915 times.
✓ Branch 1 taken 3367615 times.
|
148175060 | for(i=0; i<h; i++) |
| 67 | { | ||
| 68 | 141439830 | AV_WN4P(dst , AV_RN4P(src )); | |
| 69 | 141439830 | AV_WN4P(dst+ 4*sizeof(pixel), AV_RN4P(src+ 4*sizeof(pixel))); | |
| 70 | 141439830 | AV_WN4P(dst+ 8*sizeof(pixel), AV_RN4P(src+ 8*sizeof(pixel))); | |
| 71 | 141439830 | AV_WN4P(dst+12*sizeof(pixel), AV_RN4P(src+12*sizeof(pixel))); | |
| 72 | 141439830 | dst+=dstStride; | |
| 73 | 141439830 | src+=srcStride; | |
| 74 | } | ||
| 75 | 6735230 | } | |
| 76 | |||
| 77 | #define H264_LOWPASS(OPNAME, OP, OP2) \ | ||
| 78 | static void FUNC(OPNAME ## h264_qpel4_h_lowpass)(uint8_t *_dst, const uint8_t *restrict _src, int dstStride, int srcStride)\ | ||
| 79 | {\ | ||
| 80 | const int h=4;\ | ||
| 81 | int i;\ | ||
| 82 | pixel *dst = (pixel*)_dst;\ | ||
| 83 | const pixel *restrict src = (const pixel*)_src;\ | ||
| 84 | dstStride >>= sizeof(pixel)-1;\ | ||
| 85 | srcStride >>= sizeof(pixel)-1;\ | ||
| 86 | for(i=0; i<h; i++)\ | ||
| 87 | {\ | ||
| 88 | OP(dst[0], (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3]));\ | ||
| 89 | OP(dst[1], (src[1]+src[2])*20 - (src[0 ]+src[3])*5 + (src[-1]+src[4]));\ | ||
| 90 | OP(dst[2], (src[2]+src[3])*20 - (src[1 ]+src[4])*5 + (src[0 ]+src[5]));\ | ||
| 91 | OP(dst[3], (src[3]+src[4])*20 - (src[2 ]+src[5])*5 + (src[1 ]+src[6]));\ | ||
| 92 | dst+=dstStride;\ | ||
| 93 | src+=srcStride;\ | ||
| 94 | }\ | ||
| 95 | }\ | ||
| 96 | \ | ||
| 97 | static void FUNC(OPNAME ## h264_qpel4_v_lowpass)(uint8_t *_dst, const uint8_t *restrict _src, int dstStride, int srcStride)\ | ||
| 98 | {\ | ||
| 99 | const int w=4;\ | ||
| 100 | int i;\ | ||
| 101 | pixel *dst = (pixel*)_dst;\ | ||
| 102 | const pixel *restrict src = (const pixel*)_src;\ | ||
| 103 | dstStride >>= sizeof(pixel)-1;\ | ||
| 104 | srcStride >>= sizeof(pixel)-1;\ | ||
| 105 | for(i=0; i<w; i++)\ | ||
| 106 | {\ | ||
| 107 | const int srcB= src[-2*srcStride];\ | ||
| 108 | const int srcA= src[-1*srcStride];\ | ||
| 109 | const int src0= src[0 *srcStride];\ | ||
| 110 | const int src1= src[1 *srcStride];\ | ||
| 111 | const int src2= src[2 *srcStride];\ | ||
| 112 | const int src3= src[3 *srcStride];\ | ||
| 113 | const int src4= src[4 *srcStride];\ | ||
| 114 | const int src5= src[5 *srcStride];\ | ||
| 115 | const int src6= src[6 *srcStride];\ | ||
| 116 | OP(dst[0*dstStride], (src0+src1)*20 - (srcA+src2)*5 + (srcB+src3));\ | ||
| 117 | OP(dst[1*dstStride], (src1+src2)*20 - (src0+src3)*5 + (srcA+src4));\ | ||
| 118 | OP(dst[2*dstStride], (src2+src3)*20 - (src1+src4)*5 + (src0+src5));\ | ||
| 119 | OP(dst[3*dstStride], (src3+src4)*20 - (src2+src5)*5 + (src1+src6));\ | ||
| 120 | dst++;\ | ||
| 121 | src++;\ | ||
| 122 | }\ | ||
| 123 | }\ | ||
| 124 | \ | ||
| 125 | static void FUNC(OPNAME ## h264_qpel4_hv_lowpass)(uint8_t *_dst, pixeltmp *tmp, const uint8_t *restrict _src, int dstStride, int tmpStride, int srcStride)\ | ||
| 126 | {\ | ||
| 127 | const int h=4;\ | ||
| 128 | const int w=4;\ | ||
| 129 | const int pad = (BIT_DEPTH == 10) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\ | ||
| 130 | int i;\ | ||
| 131 | pixel *dst = (pixel*)_dst;\ | ||
| 132 | const pixel *restrict src = (const pixel*)_src;\ | ||
| 133 | dstStride >>= sizeof(pixel)-1;\ | ||
| 134 | srcStride >>= sizeof(pixel)-1;\ | ||
| 135 | src -= 2*srcStride;\ | ||
| 136 | for(i=0; i<h+5; i++)\ | ||
| 137 | {\ | ||
| 138 | tmp[0]= (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3]) + pad;\ | ||
| 139 | tmp[1]= (src[1]+src[2])*20 - (src[0 ]+src[3])*5 + (src[-1]+src[4]) + pad;\ | ||
| 140 | tmp[2]= (src[2]+src[3])*20 - (src[1 ]+src[4])*5 + (src[0 ]+src[5]) + pad;\ | ||
| 141 | tmp[3]= (src[3]+src[4])*20 - (src[2 ]+src[5])*5 + (src[1 ]+src[6]) + pad;\ | ||
| 142 | tmp+=tmpStride;\ | ||
| 143 | src+=srcStride;\ | ||
| 144 | }\ | ||
| 145 | tmp -= tmpStride*(h+5-2);\ | ||
| 146 | for(i=0; i<w; i++)\ | ||
| 147 | {\ | ||
| 148 | const int tmpB= tmp[-2*tmpStride] - pad;\ | ||
| 149 | const int tmpA= tmp[-1*tmpStride] - pad;\ | ||
| 150 | const int tmp0= tmp[0 *tmpStride] - pad;\ | ||
| 151 | const int tmp1= tmp[1 *tmpStride] - pad;\ | ||
| 152 | const int tmp2= tmp[2 *tmpStride] - pad;\ | ||
| 153 | const int tmp3= tmp[3 *tmpStride] - pad;\ | ||
| 154 | const int tmp4= tmp[4 *tmpStride] - pad;\ | ||
| 155 | const int tmp5= tmp[5 *tmpStride] - pad;\ | ||
| 156 | const int tmp6= tmp[6 *tmpStride] - pad;\ | ||
| 157 | OP2(dst[0*dstStride], (tmp0+tmp1)*20 - (tmpA+tmp2)*5 + (tmpB+tmp3));\ | ||
| 158 | OP2(dst[1*dstStride], (tmp1+tmp2)*20 - (tmp0+tmp3)*5 + (tmpA+tmp4));\ | ||
| 159 | OP2(dst[2*dstStride], (tmp2+tmp3)*20 - (tmp1+tmp4)*5 + (tmp0+tmp5));\ | ||
| 160 | OP2(dst[3*dstStride], (tmp3+tmp4)*20 - (tmp2+tmp5)*5 + (tmp1+tmp6));\ | ||
| 161 | dst++;\ | ||
| 162 | tmp++;\ | ||
| 163 | }\ | ||
| 164 | }\ | ||
| 165 | \ | ||
| 166 | static void FUNC(OPNAME ## h264_qpel8_h_lowpass)(uint8_t *_dst, const uint8_t *restrict _src, int dstStride, int srcStride)\ | ||
| 167 | {\ | ||
| 168 | const int h=8;\ | ||
| 169 | int i;\ | ||
| 170 | pixel *dst = (pixel*)_dst;\ | ||
| 171 | const pixel *restrict src = (const pixel*)_src;\ | ||
| 172 | dstStride >>= sizeof(pixel)-1;\ | ||
| 173 | srcStride >>= sizeof(pixel)-1;\ | ||
| 174 | for(i=0; i<h; i++)\ | ||
| 175 | {\ | ||
| 176 | OP(dst[0], (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3 ]));\ | ||
| 177 | OP(dst[1], (src[1]+src[2])*20 - (src[0 ]+src[3])*5 + (src[-1]+src[4 ]));\ | ||
| 178 | OP(dst[2], (src[2]+src[3])*20 - (src[1 ]+src[4])*5 + (src[0 ]+src[5 ]));\ | ||
| 179 | OP(dst[3], (src[3]+src[4])*20 - (src[2 ]+src[5])*5 + (src[1 ]+src[6 ]));\ | ||
| 180 | OP(dst[4], (src[4]+src[5])*20 - (src[3 ]+src[6])*5 + (src[2 ]+src[7 ]));\ | ||
| 181 | OP(dst[5], (src[5]+src[6])*20 - (src[4 ]+src[7])*5 + (src[3 ]+src[8 ]));\ | ||
| 182 | OP(dst[6], (src[6]+src[7])*20 - (src[5 ]+src[8])*5 + (src[4 ]+src[9 ]));\ | ||
| 183 | OP(dst[7], (src[7]+src[8])*20 - (src[6 ]+src[9])*5 + (src[5 ]+src[10]));\ | ||
| 184 | dst+=dstStride;\ | ||
| 185 | src+=srcStride;\ | ||
| 186 | }\ | ||
| 187 | }\ | ||
| 188 | \ | ||
| 189 | static void FUNC(OPNAME ## h264_qpel8_v_lowpass)(uint8_t *_dst, const uint8_t *restrict _src, int dstStride, int srcStride)\ | ||
| 190 | {\ | ||
| 191 | const int w=8;\ | ||
| 192 | int i;\ | ||
| 193 | pixel *dst = (pixel*)_dst;\ | ||
| 194 | const pixel *restrict src = (const pixel*)_src;\ | ||
| 195 | dstStride >>= sizeof(pixel)-1;\ | ||
| 196 | srcStride >>= sizeof(pixel)-1;\ | ||
| 197 | for(i=0; i<w; i++)\ | ||
| 198 | {\ | ||
| 199 | const int srcB= src[-2*srcStride];\ | ||
| 200 | const int srcA= src[-1*srcStride];\ | ||
| 201 | const int src0= src[0 *srcStride];\ | ||
| 202 | const int src1= src[1 *srcStride];\ | ||
| 203 | const int src2= src[2 *srcStride];\ | ||
| 204 | const int src3= src[3 *srcStride];\ | ||
| 205 | const int src4= src[4 *srcStride];\ | ||
| 206 | const int src5= src[5 *srcStride];\ | ||
| 207 | const int src6= src[6 *srcStride];\ | ||
| 208 | const int src7= src[7 *srcStride];\ | ||
| 209 | const int src8= src[8 *srcStride];\ | ||
| 210 | const int src9= src[9 *srcStride];\ | ||
| 211 | const int src10=src[10*srcStride];\ | ||
| 212 | OP(dst[0*dstStride], (src0+src1)*20 - (srcA+src2)*5 + (srcB+src3));\ | ||
| 213 | OP(dst[1*dstStride], (src1+src2)*20 - (src0+src3)*5 + (srcA+src4));\ | ||
| 214 | OP(dst[2*dstStride], (src2+src3)*20 - (src1+src4)*5 + (src0+src5));\ | ||
| 215 | OP(dst[3*dstStride], (src3+src4)*20 - (src2+src5)*5 + (src1+src6));\ | ||
| 216 | OP(dst[4*dstStride], (src4+src5)*20 - (src3+src6)*5 + (src2+src7));\ | ||
| 217 | OP(dst[5*dstStride], (src5+src6)*20 - (src4+src7)*5 + (src3+src8));\ | ||
| 218 | OP(dst[6*dstStride], (src6+src7)*20 - (src5+src8)*5 + (src4+src9));\ | ||
| 219 | OP(dst[7*dstStride], (src7+src8)*20 - (src6+src9)*5 + (src5+src10));\ | ||
| 220 | dst++;\ | ||
| 221 | src++;\ | ||
| 222 | }\ | ||
| 223 | }\ | ||
| 224 | \ | ||
| 225 | static void FUNC(OPNAME ## h264_qpel8_hv_lowpass)(uint8_t *_dst, pixeltmp *tmp, const uint8_t *restrict _src, int dstStride, int tmpStride, int srcStride)\ | ||
| 226 | {\ | ||
| 227 | const int h=8;\ | ||
| 228 | const int w=8;\ | ||
| 229 | const int pad = (BIT_DEPTH == 10) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\ | ||
| 230 | int i;\ | ||
| 231 | pixel *dst = (pixel*)_dst;\ | ||
| 232 | const pixel *restrict src = (const pixel*)_src;\ | ||
| 233 | dstStride >>= sizeof(pixel)-1;\ | ||
| 234 | srcStride >>= sizeof(pixel)-1;\ | ||
| 235 | src -= 2*srcStride;\ | ||
| 236 | for(i=0; i<h+5; i++)\ | ||
| 237 | {\ | ||
| 238 | tmp[0]= (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3 ]) + pad;\ | ||
| 239 | tmp[1]= (src[1]+src[2])*20 - (src[0 ]+src[3])*5 + (src[-1]+src[4 ]) + pad;\ | ||
| 240 | tmp[2]= (src[2]+src[3])*20 - (src[1 ]+src[4])*5 + (src[0 ]+src[5 ]) + pad;\ | ||
| 241 | tmp[3]= (src[3]+src[4])*20 - (src[2 ]+src[5])*5 + (src[1 ]+src[6 ]) + pad;\ | ||
| 242 | tmp[4]= (src[4]+src[5])*20 - (src[3 ]+src[6])*5 + (src[2 ]+src[7 ]) + pad;\ | ||
| 243 | tmp[5]= (src[5]+src[6])*20 - (src[4 ]+src[7])*5 + (src[3 ]+src[8 ]) + pad;\ | ||
| 244 | tmp[6]= (src[6]+src[7])*20 - (src[5 ]+src[8])*5 + (src[4 ]+src[9 ]) + pad;\ | ||
| 245 | tmp[7]= (src[7]+src[8])*20 - (src[6 ]+src[9])*5 + (src[5 ]+src[10]) + pad;\ | ||
| 246 | tmp+=tmpStride;\ | ||
| 247 | src+=srcStride;\ | ||
| 248 | }\ | ||
| 249 | tmp -= tmpStride*(h+5-2);\ | ||
| 250 | for(i=0; i<w; i++)\ | ||
| 251 | {\ | ||
| 252 | const int tmpB= tmp[-2*tmpStride] - pad;\ | ||
| 253 | const int tmpA= tmp[-1*tmpStride] - pad;\ | ||
| 254 | const int tmp0= tmp[0 *tmpStride] - pad;\ | ||
| 255 | const int tmp1= tmp[1 *tmpStride] - pad;\ | ||
| 256 | const int tmp2= tmp[2 *tmpStride] - pad;\ | ||
| 257 | const int tmp3= tmp[3 *tmpStride] - pad;\ | ||
| 258 | const int tmp4= tmp[4 *tmpStride] - pad;\ | ||
| 259 | const int tmp5= tmp[5 *tmpStride] - pad;\ | ||
| 260 | const int tmp6= tmp[6 *tmpStride] - pad;\ | ||
| 261 | const int tmp7= tmp[7 *tmpStride] - pad;\ | ||
| 262 | const int tmp8= tmp[8 *tmpStride] - pad;\ | ||
| 263 | const int tmp9= tmp[9 *tmpStride] - pad;\ | ||
| 264 | const int tmp10=tmp[10*tmpStride] - pad;\ | ||
| 265 | OP2(dst[0*dstStride], (tmp0+tmp1)*20 - (tmpA+tmp2)*5 + (tmpB+tmp3));\ | ||
| 266 | OP2(dst[1*dstStride], (tmp1+tmp2)*20 - (tmp0+tmp3)*5 + (tmpA+tmp4));\ | ||
| 267 | OP2(dst[2*dstStride], (tmp2+tmp3)*20 - (tmp1+tmp4)*5 + (tmp0+tmp5));\ | ||
| 268 | OP2(dst[3*dstStride], (tmp3+tmp4)*20 - (tmp2+tmp5)*5 + (tmp1+tmp6));\ | ||
| 269 | OP2(dst[4*dstStride], (tmp4+tmp5)*20 - (tmp3+tmp6)*5 + (tmp2+tmp7));\ | ||
| 270 | OP2(dst[5*dstStride], (tmp5+tmp6)*20 - (tmp4+tmp7)*5 + (tmp3+tmp8));\ | ||
| 271 | OP2(dst[6*dstStride], (tmp6+tmp7)*20 - (tmp5+tmp8)*5 + (tmp4+tmp9));\ | ||
| 272 | OP2(dst[7*dstStride], (tmp7+tmp8)*20 - (tmp6+tmp9)*5 + (tmp5+tmp10));\ | ||
| 273 | dst++;\ | ||
| 274 | tmp++;\ | ||
| 275 | }\ | ||
| 276 | }\ | ||
| 277 | \ | ||
| 278 | static void FUNC(OPNAME ## h264_qpel16_v_lowpass)(uint8_t *dst, const uint8_t *restrict src, int dstStride, int srcStride)\ | ||
| 279 | {\ | ||
| 280 | FUNC(OPNAME ## h264_qpel8_v_lowpass)(dst , src , dstStride, srcStride);\ | ||
| 281 | FUNC(OPNAME ## h264_qpel8_v_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\ | ||
| 282 | src += 8*srcStride;\ | ||
| 283 | dst += 8*dstStride;\ | ||
| 284 | FUNC(OPNAME ## h264_qpel8_v_lowpass)(dst , src , dstStride, srcStride);\ | ||
| 285 | FUNC(OPNAME ## h264_qpel8_v_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\ | ||
| 286 | }\ | ||
| 287 | \ | ||
| 288 | static void FUNC(OPNAME ## h264_qpel16_h_lowpass)(uint8_t *dst, const uint8_t *restrict src, int dstStride, int srcStride)\ | ||
| 289 | {\ | ||
| 290 | FUNC(OPNAME ## h264_qpel8_h_lowpass)(dst , src , dstStride, srcStride);\ | ||
| 291 | FUNC(OPNAME ## h264_qpel8_h_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\ | ||
| 292 | src += 8*srcStride;\ | ||
| 293 | dst += 8*dstStride;\ | ||
| 294 | FUNC(OPNAME ## h264_qpel8_h_lowpass)(dst , src , dstStride, srcStride);\ | ||
| 295 | FUNC(OPNAME ## h264_qpel8_h_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\ | ||
| 296 | }\ | ||
| 297 | \ | ||
| 298 | static void FUNC(OPNAME ## h264_qpel16_hv_lowpass)(uint8_t *dst, pixeltmp *tmp, const uint8_t *restrict src, int dstStride, int tmpStride, int srcStride){\ | ||
| 299 | FUNC(OPNAME ## h264_qpel8_hv_lowpass)(dst , tmp , src , dstStride, tmpStride, srcStride);\ | ||
| 300 | FUNC(OPNAME ## h264_qpel8_hv_lowpass)(dst+8*sizeof(pixel), tmp+8, src+8*sizeof(pixel), dstStride, tmpStride, srcStride);\ | ||
| 301 | src += 8*srcStride;\ | ||
| 302 | dst += 8*dstStride;\ | ||
| 303 | FUNC(OPNAME ## h264_qpel8_hv_lowpass)(dst , tmp , src , dstStride, tmpStride, srcStride);\ | ||
| 304 | FUNC(OPNAME ## h264_qpel8_hv_lowpass)(dst+8*sizeof(pixel), tmp+8, src+8*sizeof(pixel), dstStride, tmpStride, srcStride);\ | ||
| 305 | }\ | ||
| 306 | |||
| 307 | #define H264_MC(OPNAME, NAME, SIZE) \ | ||
| 308 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc00)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 309 | {\ | ||
| 310 | FUNCC(OPNAME ## pixels ## SIZE)(dst, src, stride, SIZE);\ | ||
| 311 | }\ | ||
| 312 | \ | ||
| 313 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc10)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 314 | {\ | ||
| 315 | uint8_t half[SIZE*SIZE*sizeof(pixel)];\ | ||
| 316 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _h_lowpass)(half, src, SIZE*sizeof(pixel), stride);\ | ||
| 317 | FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, src, half, stride, stride, SIZE*sizeof(pixel), SIZE);\ | ||
| 318 | }\ | ||
| 319 | \ | ||
| 320 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc20)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 321 | {\ | ||
| 322 | FUNC(OPNAME ## NAME ## _qpel ## SIZE ## _h_lowpass)(dst, src, stride, stride);\ | ||
| 323 | }\ | ||
| 324 | \ | ||
| 325 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc30)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 326 | {\ | ||
| 327 | uint8_t half[SIZE*SIZE*sizeof(pixel)];\ | ||
| 328 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _h_lowpass)(half, src, SIZE*sizeof(pixel), stride);\ | ||
| 329 | FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, src+sizeof(pixel), half, stride, stride, SIZE*sizeof(pixel), SIZE);\ | ||
| 330 | }\ | ||
| 331 | \ | ||
| 332 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc01)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 333 | {\ | ||
| 334 | uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 335 | uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\ | ||
| 336 | uint8_t half[SIZE*SIZE*sizeof(pixel)];\ | ||
| 337 | FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel), stride, SIZE + 5);\ | ||
| 338 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _v_lowpass)(half, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\ | ||
| 339 | FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, full_mid, half, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\ | ||
| 340 | }\ | ||
| 341 | \ | ||
| 342 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc02)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 343 | {\ | ||
| 344 | uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 345 | uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\ | ||
| 346 | FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel), stride, SIZE + 5);\ | ||
| 347 | FUNC(OPNAME ## NAME ## _qpel ## SIZE ## _v_lowpass)(dst, full_mid, stride, SIZE*sizeof(pixel));\ | ||
| 348 | }\ | ||
| 349 | \ | ||
| 350 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc03)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 351 | {\ | ||
| 352 | uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 353 | uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\ | ||
| 354 | uint8_t half[SIZE*SIZE*sizeof(pixel)];\ | ||
| 355 | FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel), stride, SIZE + 5);\ | ||
| 356 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _v_lowpass)(half, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\ | ||
| 357 | FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, full_mid+SIZE*sizeof(pixel), half, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\ | ||
| 358 | }\ | ||
| 359 | \ | ||
| 360 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc11)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 361 | {\ | ||
| 362 | uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 363 | uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\ | ||
| 364 | uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\ | ||
| 365 | uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\ | ||
| 366 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _h_lowpass)(halfH, src, SIZE*sizeof(pixel), stride);\ | ||
| 367 | FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel), stride, SIZE + 5);\ | ||
| 368 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _v_lowpass)(halfV, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\ | ||
| 369 | FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\ | ||
| 370 | }\ | ||
| 371 | \ | ||
| 372 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc31)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 373 | {\ | ||
| 374 | uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 375 | uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\ | ||
| 376 | uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\ | ||
| 377 | uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\ | ||
| 378 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _h_lowpass)(halfH, src, SIZE*sizeof(pixel), stride);\ | ||
| 379 | FUNC(copy_block ## SIZE )(full, src - stride*2 + sizeof(pixel), SIZE*sizeof(pixel), stride, SIZE + 5);\ | ||
| 380 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _v_lowpass)(halfV, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\ | ||
| 381 | FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\ | ||
| 382 | }\ | ||
| 383 | \ | ||
| 384 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc13)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 385 | {\ | ||
| 386 | uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 387 | uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\ | ||
| 388 | uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\ | ||
| 389 | uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\ | ||
| 390 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _h_lowpass)(halfH, src + stride, SIZE*sizeof(pixel), stride);\ | ||
| 391 | FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel), stride, SIZE + 5);\ | ||
| 392 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _v_lowpass)(halfV, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\ | ||
| 393 | FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\ | ||
| 394 | }\ | ||
| 395 | \ | ||
| 396 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc33)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 397 | {\ | ||
| 398 | uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 399 | uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\ | ||
| 400 | uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\ | ||
| 401 | uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\ | ||
| 402 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _h_lowpass)(halfH, src + stride, SIZE*sizeof(pixel), stride);\ | ||
| 403 | FUNC(copy_block ## SIZE )(full, src - stride*2 + sizeof(pixel), SIZE*sizeof(pixel), stride, SIZE + 5);\ | ||
| 404 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _v_lowpass)(halfV, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\ | ||
| 405 | FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\ | ||
| 406 | }\ | ||
| 407 | \ | ||
| 408 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc22)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 409 | {\ | ||
| 410 | pixeltmp tmp[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 411 | FUNC(OPNAME ## NAME ## _qpel ## SIZE ## _hv_lowpass)(dst, tmp, src, stride, SIZE*sizeof(pixel), stride);\ | ||
| 412 | }\ | ||
| 413 | \ | ||
| 414 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc21)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 415 | {\ | ||
| 416 | pixeltmp tmp[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 417 | uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\ | ||
| 418 | uint8_t halfHV[SIZE*SIZE*sizeof(pixel)];\ | ||
| 419 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _h_lowpass)(halfH, src, SIZE*sizeof(pixel), stride);\ | ||
| 420 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _hv_lowpass)(halfHV, tmp, src, SIZE*sizeof(pixel), SIZE*sizeof(pixel), stride);\ | ||
| 421 | FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfHV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\ | ||
| 422 | }\ | ||
| 423 | \ | ||
| 424 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc23)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 425 | {\ | ||
| 426 | pixeltmp tmp[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 427 | uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\ | ||
| 428 | uint8_t halfHV[SIZE*SIZE*sizeof(pixel)];\ | ||
| 429 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _h_lowpass)(halfH, src + stride, SIZE*sizeof(pixel), stride);\ | ||
| 430 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _hv_lowpass)(halfHV, tmp, src, SIZE*sizeof(pixel), SIZE*sizeof(pixel), stride);\ | ||
| 431 | FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfHV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\ | ||
| 432 | }\ | ||
| 433 | \ | ||
| 434 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc12)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 435 | {\ | ||
| 436 | uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 437 | uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\ | ||
| 438 | pixeltmp tmp[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 439 | uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\ | ||
| 440 | uint8_t halfHV[SIZE*SIZE*sizeof(pixel)];\ | ||
| 441 | FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel), stride, SIZE + 5);\ | ||
| 442 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _v_lowpass)(halfV, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\ | ||
| 443 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _hv_lowpass)(halfHV, tmp, src, SIZE*sizeof(pixel), SIZE*sizeof(pixel), stride);\ | ||
| 444 | FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfV, halfHV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\ | ||
| 445 | }\ | ||
| 446 | \ | ||
| 447 | static void FUNCC(OPNAME ## NAME ## _qpel ## SIZE ## _mc32)(uint8_t *dst, const uint8_t *restrict src, ptrdiff_t stride)\ | ||
| 448 | {\ | ||
| 449 | uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 450 | uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\ | ||
| 451 | pixeltmp tmp[SIZE*(SIZE+5)*sizeof(pixel)];\ | ||
| 452 | uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\ | ||
| 453 | uint8_t halfHV[SIZE*SIZE*sizeof(pixel)];\ | ||
| 454 | FUNC(copy_block ## SIZE )(full, src - stride*2 + sizeof(pixel), SIZE*sizeof(pixel), stride, SIZE + 5);\ | ||
| 455 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _v_lowpass)(halfV, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\ | ||
| 456 | FUNC(put_ ## NAME ## _qpel ## SIZE ## _hv_lowpass)(halfHV, tmp, src, SIZE*sizeof(pixel), SIZE*sizeof(pixel), stride);\ | ||
| 457 | FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfV, halfHV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\ | ||
| 458 | }\ | ||
| 459 | |||
| 460 | #define op_avg(a, b) a = (((a)+CLIP(((b) + 16)>>5)+1)>>1) | ||
| 461 | //#define op_avg2(a, b) a = (((a)*w1+cm[((b) + 16)>>5]*w2 + o + 64)>>7) | ||
| 462 | #define op_put(a, b) a = CLIP(((b) + 16)>>5) | ||
| 463 | #define op2_avg(a, b) a = (((a)+CLIP(((b) + 512)>>10)+1)>>1) | ||
| 464 | #define op2_put(a, b) a = CLIP(((b) + 512)>>10) | ||
| 465 | |||
| 466 | #ifndef SNOW | ||
| 467 |
4/4✓ Branch 0 taken 624172341 times.
✓ Branch 1 taken 84957460 times.
✓ Branch 2 taken 86775048 times.
✓ Branch 3 taken 11796717 times.
|
1610733792 | H264_LOWPASS(put_ , op_put, op2_put) |
| 468 |
4/4✓ Branch 0 taken 9289763 times.
✓ Branch 1 taken 1138358 times.
✓ Branch 2 taken 1435028 times.
✓ Branch 3 taken 192131 times.
|
24095310 | H264_LOWPASS(avg_ , op_avg, op2_avg) |
| 469 | 29207196 | H264_MC(put_, h264, 4) | |
| 470 | 27522890 | H264_MC(put_, h264, 8) | |
| 471 | 22532766 | H264_MC(put_, h264, 16) | |
| 472 | 7888064 | H264_MC(avg_, h264, 4) | |
| 473 | 2442308 | H264_MC(avg_, h264, 8) | |
| 474 | 4816358 | H264_MC(avg_, h264, 16) | |
| 475 | #endif | ||
| 476 | |||
| 477 | #undef op_avg | ||
| 478 | #undef op_put | ||
| 479 | #undef op2_avg | ||
| 480 | #undef op2_put | ||
| 481 |