Parent Directory
| 
Revision Log
Revision 262 - (view) (download)
| 1 : | Isibaar | 262 | ;/************************************************************************** |
| 2 : | ; * | ||
| 3 : | ; * XVID MPEG-4 VIDEO CODEC | ||
| 4 : | ; * xmm sum of absolute difference | ||
| 5 : | ; * | ||
| 6 : | ; * This program is free software; you can redistribute it and/or modify | ||
| 7 : | ; * it under the terms of the GNU General Public License as published by | ||
| 8 : | ; * the Free Software Foundation; either version 2 of the License, or | ||
| 9 : | ; * (at your option) any later version. | ||
| 10 : | ; * | ||
| 11 : | ; * This program is distributed in the hope that it will be useful, | ||
| 12 : | ; * but WITHOUT ANY WARRANTY; without even the implied warranty of | ||
| 13 : | ; * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | ||
| 14 : | ; * GNU General Public License for more details. | ||
| 15 : | ; * | ||
| 16 : | ; * You should have received a copy of the GNU General Public License | ||
| 17 : | ; * along with this program; if not, write to the Free Software | ||
| 18 : | ; * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. | ||
| 19 : | ; * | ||
| 20 : | ; *************************************************************************/ | ||
| 21 : | |||
| 22 : | ;/************************************************************************** | ||
| 23 : | ; * | ||
| 24 : | ; * History: | ||
| 25 : | ; * | ||
| 26 : | ; * 23.07.2002 sad8bi_xmm; <pross@xvid.org> | ||
| 27 : | ; * 04.06.2002 rewrote some funcs (XMM mainly) -Skal- | ||
| 28 : | ; * 17.11.2001 bugfix and small improvement for dev16_xmm, | ||
| 29 : | ; * removed terminate early in sad16_xmm (Isibaar) | ||
| 30 : | ; * 12.11.2001 inital version; (c)2001 peter ross <pross@cs.rmit.edu.au> | ||
| 31 : | ; * | ||
| 32 : | ; *************************************************************************/ | ||
| 33 : | |||
| 34 : | bits 32 | ||
| 35 : | %macro cglobal 1 | ||
| 36 : | %ifdef PREFIX | ||
| 37 : | global _%1 | ||
| 38 : | %define %1 _%1 | ||
| 39 : | %else | ||
| 40 : | global %1 | ||
| 41 : | %endif | ||
| 42 : | %endmacro | ||
| 43 : | |||
| 44 : | section .data | ||
| 45 : | |||
| 46 : | align 16 | ||
| 47 : | mmx_one times 4 dw 1 | ||
| 48 : | |||
| 49 : | section .text | ||
| 50 : | |||
| 51 : | cglobal sad16_xmm | ||
| 52 : | cglobal sad8_xmm | ||
| 53 : | cglobal sad16bi_xmm | ||
| 54 : | cglobal sad8bi_xmm | ||
| 55 : | cglobal dev16_xmm | ||
| 56 : | |||
| 57 : | ;=========================================================================== | ||
| 58 : | ; | ||
| 59 : | ; uint32_t sad16_xmm(const uint8_t * const cur, | ||
| 60 : | ; const uint8_t * const ref, | ||
| 61 : | ; const uint32_t best_sad); | ||
| 62 : | cglobal sad8_xmm | ||
| 63 : | ;=========================================================================== | ||
| 64 : | |||
| 65 : | %macro SAD_16x16_SSE 0 | ||
| 66 : | movq mm0, [eax] | ||
| 67 : | psadbw mm0, [edx] | ||
| 68 : | movq mm1, [eax+8] | ||
| 69 : | add eax, ecx | ||
| 70 : | psadbw mm1, [edx+8] | ||
| 71 : | paddusw mm5,mm0 | ||
| 72 : | add edx, ecx | ||
| 73 : | paddusw mm6,mm1 | ||
| 74 : | %endmacro | ||
| 75 : | |||
| 76 : | align 16 | ||
| 77 : | sad16_xmm: | ||
| 78 : | |||
| 79 : | mov eax, [esp+ 4] ; Src1 | ||
| 80 : | mov edx, [esp+ 8] ; Src2 | ||
| 81 : | mov ecx, [esp+12] ; Stride | ||
| 82 : | |||
| 83 : | pxor mm5, mm5 ; accum1 | ||
| 84 : | pxor mm6, mm6 ; accum2 | ||
| 85 : | |||
| 86 : | SAD_16x16_SSE | ||
| 87 : | SAD_16x16_SSE | ||
| 88 : | SAD_16x16_SSE | ||
| 89 : | SAD_16x16_SSE | ||
| 90 : | SAD_16x16_SSE | ||
| 91 : | SAD_16x16_SSE | ||
| 92 : | SAD_16x16_SSE | ||
| 93 : | SAD_16x16_SSE | ||
| 94 : | |||
| 95 : | SAD_16x16_SSE | ||
| 96 : | SAD_16x16_SSE | ||
| 97 : | SAD_16x16_SSE | ||
| 98 : | SAD_16x16_SSE | ||
| 99 : | SAD_16x16_SSE | ||
| 100 : | SAD_16x16_SSE | ||
| 101 : | SAD_16x16_SSE | ||
| 102 : | SAD_16x16_SSE | ||
| 103 : | |||
| 104 : | paddusw mm6,mm5 | ||
| 105 : | movd eax, mm6 | ||
| 106 : | ret | ||
| 107 : | |||
| 108 : | |||
| 109 : | ;=========================================================================== | ||
| 110 : | ; | ||
| 111 : | ; uint32_t sad8_xmm(const uint8_t * const cur, | ||
| 112 : | ; const uint8_t * const ref, | ||
| 113 : | ; const uint32_t stride); | ||
| 114 : | ; | ||
| 115 : | ;=========================================================================== | ||
| 116 : | |||
| 117 : | %macro SADBI_16x16_SSE 0 | ||
| 118 : | movq mm0, [eax] | ||
| 119 : | movq mm1, [eax+8] | ||
| 120 : | |||
| 121 : | movq mm2, [edx] | ||
| 122 : | movq mm3, [edx+8] | ||
| 123 : | |||
| 124 : | pavgb mm2, [ebx] | ||
| 125 : | add edx, ecx | ||
| 126 : | |||
| 127 : | pavgb mm3, [ebx+8] | ||
| 128 : | add ebx, ecx | ||
| 129 : | |||
| 130 : | psadbw mm0, mm2 | ||
| 131 : | add eax, ecx | ||
| 132 : | |||
| 133 : | psadbw mm1, mm3 | ||
| 134 : | paddusw mm5,mm0 | ||
| 135 : | |||
| 136 : | paddusw mm6,mm1 | ||
| 137 : | %endmacro | ||
| 138 : | |||
| 139 : | align 16 | ||
| 140 : | sad16bi_xmm: | ||
| 141 : | push ebx | ||
| 142 : | mov eax, [esp+4+ 4] ; Src | ||
| 143 : | mov edx, [esp+4+ 8] ; Ref1 | ||
| 144 : | mov ebx, [esp+4+12] ; Ref2 | ||
| 145 : | mov ecx, [esp+4+16] ; Stride | ||
| 146 : | |||
| 147 : | pxor mm5, mm5 ; accum1 | ||
| 148 : | pxor mm6, mm6 ; accum2 | ||
| 149 : | |||
| 150 : | SADBI_16x16_SSE | ||
| 151 : | SADBI_16x16_SSE | ||
| 152 : | SADBI_16x16_SSE | ||
| 153 : | SADBI_16x16_SSE | ||
| 154 : | SADBI_16x16_SSE | ||
| 155 : | SADBI_16x16_SSE | ||
| 156 : | SADBI_16x16_SSE | ||
| 157 : | SADBI_16x16_SSE | ||
| 158 : | |||
| 159 : | SADBI_16x16_SSE | ||
| 160 : | SADBI_16x16_SSE | ||
| 161 : | SADBI_16x16_SSE | ||
| 162 : | SADBI_16x16_SSE | ||
| 163 : | SADBI_16x16_SSE | ||
| 164 : | SADBI_16x16_SSE | ||
| 165 : | SADBI_16x16_SSE | ||
| 166 : | SADBI_16x16_SSE | ||
| 167 : | |||
| 168 : | paddusw mm6,mm5 | ||
| 169 : | movd eax, mm6 | ||
| 170 : | pop ebx | ||
| 171 : | ret | ||
| 172 : | |||
| 173 : | ;=========================================================================== | ||
| 174 : | ; | ||
| 175 : | ; uint32_t sad8bi_xmm(const uint8_t * const cur, | ||
| 176 : | ; const uint8_t * const ref1, | ||
| 177 : | ; const uint8_t * const ref2, | ||
| 178 : | ; const uint32_t stride); | ||
| 179 : | ; | ||
| 180 : | ;=========================================================================== | ||
| 181 : | |||
| 182 : | ;=========================================================================== | ||
| 183 : | ; | ||
| 184 : | ; uint32_t sad8_xmm(const uint8_t * const cur, | ||
| 185 : | ; const uint8_t * const ref, | ||
| 186 : | ; const uint32_t stride); | ||
| 187 : | ; | ||
| 188 : | ;=========================================================================== | ||
| 189 : | |||
| 190 : | %macro SAD_8x8_SSE 0 | ||
| 191 : | movq mm0, [eax] | ||
| 192 : | movq mm1, [eax+ecx] | ||
| 193 : | %macro MEAN_16x16_SSE 0 | ||
| 194 : | psadbw mm0, [edx] | ||
| 195 : | psadbw mm1, [edx+ecx] | ||
| 196 : | add eax, ebx | ||
| 197 : | add edx, ebx | ||
| 198 : | |||
| 199 : | paddusw mm5,mm0 | ||
| 200 : | paddusw mm6,mm1 | ||
| 201 : | %endmacro | ||
| 202 : | |||
| 203 : | align 16 | ||
| 204 : | sad8_xmm: | ||
| 205 : | |||
| 206 : | mov eax, [esp+ 4] ; Src1 | ||
| 207 : | mov edx, [esp+ 8] ; Src2 | ||
| 208 : | mov ecx, [esp+12] ; Stride | ||
| 209 : | push ebx | ||
| 210 : | lea ebx, [ecx+ecx] | ||
| 211 : | |||
| 212 : | pxor mm5, mm5 ; accum1 | ||
| 213 : | pxor mm6, mm6 ; accum2 | ||
| 214 : | |||
| 215 : | SAD_8x8_SSE | ||
| 216 : | SAD_8x8_SSE | ||
| 217 : | SAD_8x8_SSE | ||
| 218 : | |||
| 219 : | movq mm0, [eax] | ||
| 220 : | movq mm1, [eax+ecx] | ||
| 221 : | psadbw mm0, [edx] | ||
| 222 : | psadbw mm1, [edx+ecx] | ||
| 223 : | |||
| 224 : | pop ebx | ||
| 225 : | |||
| 226 : | paddusw mm5,mm0 | ||
| 227 : | paddusw mm6,mm1 | ||
| 228 : | |||
| 229 : | paddusw mm6,mm5 | ||
| 230 : | movd eax, mm6 | ||
| 231 : | |||
| 232 : | ret | ||
| 233 : | movq mm0, [eax] | ||
| 234 : | movq mm1, [eax+8] | ||
| 235 : | psadbw mm0, mm7 | ||
| 236 : | psadbw mm1, mm7 | ||
| 237 : | add eax, ecx | ||
| 238 : | paddw mm5, mm0 | ||
| 239 : | paddw mm6, mm1 | ||
| 240 : | %endmacro | ||
| 241 : | |||
| 242 : | %macro ABS_16x16_SSE 0 | ||
| 243 : | movq mm0, [eax] | ||
| 244 : | movq mm1, [eax+8] | ||
| 245 : | psadbw mm0, mm4 | ||
| 246 : | psadbw mm1, mm4 | ||
| 247 : | lea eax,[eax+ecx] | ||
| 248 : | paddw mm5, mm0 | ||
| 249 : | paddw mm6, mm1 | ||
| 250 : | %endmacro | ||
| 251 : | |||
| 252 : | align 16 | ||
| 253 : | dev16_xmm: | ||
| 254 : | |||
| 255 : | mov eax, [esp+ 4] ; Src | ||
| 256 : | mov ecx, [esp+ 8] ; Stride | ||
| 257 : | |||
| 258 : | pxor mm7, mm7 ; zero | ||
| 259 : | pxor mm5, mm5 ; mean accums | ||
| 260 : | pxor mm6, mm6 | ||
| 261 : | |||
| 262 : | MEAN_16x16_SSE | ||
| 263 : | MEAN_16x16_SSE | ||
| 264 : | MEAN_16x16_SSE | ||
| 265 : | MEAN_16x16_SSE | ||
| 266 : | MEAN_16x16_SSE | ||
| 267 : | MEAN_16x16_SSE | ||
| 268 : | MEAN_16x16_SSE | ||
| 269 : | MEAN_16x16_SSE | ||
| 270 : | |||
| 271 : | MEAN_16x16_SSE | ||
| 272 : | MEAN_16x16_SSE | ||
| 273 : | MEAN_16x16_SSE | ||
| 274 : | MEAN_16x16_SSE | ||
| 275 : | MEAN_16x16_SSE | ||
| 276 : | MEAN_16x16_SSE | ||
| 277 : | MEAN_16x16_SSE | ||
| 278 : | MEAN_16x16_SSE | ||
| 279 : | |||
| 280 : | paddusw mm6, mm5 | ||
| 281 : | |||
| 282 : | movq mm4, mm6 | ||
| 283 : | psllq mm4, 32 | ||
| 284 : | paddd mm4, mm6 | ||
| 285 : | psrld mm4, 8 ; /= (16*16) | ||
| 286 : | |||
| 287 : | packssdw mm4, mm4 | ||
| 288 : | packuswb mm4, mm4 | ||
| 289 : | |||
| 290 : | ; mm4 contains the mean | ||
| 291 : | |||
| 292 : | mov eax, [esp+ 4] ; Src | ||
| 293 : | |||
| 294 : | pxor mm5, mm5 ; sums | ||
| 295 : | pxor mm6, mm6 | ||
| 296 : | |||
| 297 : | ABS_16x16_SSE | ||
| 298 : | ABS_16x16_SSE | ||
| 299 : | ABS_16x16_SSE | ||
| 300 : | ABS_16x16_SSE | ||
| 301 : | ABS_16x16_SSE | ||
| 302 : | ABS_16x16_SSE | ||
| 303 : | ABS_16x16_SSE | ||
| 304 : | ABS_16x16_SSE | ||
| 305 : | |||
| 306 : | ABS_16x16_SSE | ||
| 307 : | ABS_16x16_SSE | ||
| 308 : | ABS_16x16_SSE | ||
| 309 : | ABS_16x16_SSE | ||
| 310 : | ABS_16x16_SSE | ||
| 311 : | ABS_16x16_SSE | ||
| 312 : | ABS_16x16_SSE | ||
| 313 : | ABS_16x16_SSE | ||
| 314 : | |||
| 315 : | paddusw mm6, mm5 | ||
| 316 : | movq mm7, mm6 | ||
| 317 : | psllq mm7, 32 | ||
| 318 : | paddd mm6, mm7 | ||
| 319 : | |||
| 320 : | movd eax, mm6 | ||
| 321 : | ret |
| No admin address has been configured | ViewVC Help |
| Powered by ViewVC 1.0.4 |