FFmpeg: libavfilter/libmpcodecs/vf_fspp.c Source File

FFmpeg

[フレーム]

vf_fspp.c

Go to the documentation of this file.

1 /*

4 *

5 * This file is part of MPlayer.

6 *

7 * MPlayer is free software; you can redistribute it and/or modify

8 * it under the terms of the GNU General Public License as published by

9 * the Free Software Foundation; either version 2 of the License, or

10 * (at your option) any later version.

11 *

12 * MPlayer 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

15 * GNU General Public License for more details.

16 *

17 * You should have received a copy of the GNU General Public License along

18 * with MPlayer; if not, write to the Free Software Foundation, Inc.,

19 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.

20 */

22 /*

23 * This implementation is based on an algorithm described in

24 * "Aria Nosratinia Embedded Post-Processing for

25 * Enhancement of Compressed Images (1999)"

26 * (http://citeseer.nj.nec.com/nosratinia99embedded.html)

27 * Futher, with splitting (i)dct into hor/ver passes, one of them can be

28 * performed once per block, not pixel. This allows for much better speed.

29 */

31 /*

32 Heavily optimized version of SPP filter by Nikolaj

33 */

35 #include <stdio.h>

36 #include <stdlib.h>

37 #include <string.h>

38 #include <inttypes.h>

39 #include <math.h>

41 #include "config.h"

43 #include "mp_msg.h"

44 #include "cpudetect.h"

45 #include "img_format.h"

46 #include "mp_image.h"

47 #include "vf.h"

48 #include "av_helpers.h"

49 #include "libvo/fastmemcpy.h"

51 #include "libavutil/internal.h"

52 #include "libavutil/intreadwrite.h"

53 #include "libavutil/mem.h"

54 #include "libavutil/x86/asm.h"

55 #include "libavcodec/avcodec.h"

56 #include "libavcodec/dsputil.h"

58 #undef free

59 #undef malloc

61 //===========================================================================//

62 #define BLOCKSZ 12

64 static const short custom_threshold[64]=

65 // values (296) can't be too high

66 // -it causes too big quant dependence

67 // or maybe overflow(check), which results in some flashing

68 { 71, 296, 295, 237, 71, 40, 38, 19,

69 245, 193, 185, 121, 102, 73, 53, 27,

70 158, 129, 141, 107, 97, 73, 50, 26,

71 102, 116, 109, 98, 82, 66, 45, 23,

72 71, 94, 95, 81, 70, 56, 38, 20,

73 56, 77, 74, 66, 56, 44, 30, 15,

74 38, 53, 50, 45, 38, 30, 21, 11,

75 20, 27, 26, 23, 20, 15, 11, 5

76 };

78 static const uint8_t __attribute__((aligned(32))) dither[8][8]={

79 { 0, 48, 12, 60, 3, 51, 15, 63, },

80 { 32, 16, 44, 28, 35, 19, 47, 31, },

81 { 8, 56, 4, 52, 11, 59, 7, 55, },

82 { 40, 24, 36, 20, 43, 27, 39, 23, },

83 { 2, 50, 14, 62, 1, 49, 13, 61, },

84 { 34, 18, 46, 30, 33, 17, 45, 29, },

85 { 10, 58, 6, 54, 9, 57, 5, 53, },

86 { 42, 26, 38, 22, 41, 25, 37, 21, },

87 };

89 struct vf_priv_s { //align 16 !

90 uint64_t threshold_mtx_noq[8*2];

91 uint64_t threshold_mtx[8*2];//used in both C & MMX (& later SSE2) versions

93 int log2_count;

94 int temp_stride;

95 int qp;

96 int mpeg2;

97 int prev_q;

98 uint8_t *src;

99 int16_t *temp;

100 int bframes;

101 char *non_b_qp;

102 };

103

104

105 #if !HAVE_MMX

106

107 //This func reads from 1 slice, 1 and clears 0 & 1

108 static void store_slice_c(uint8_t *dst, int16_t *src, int dst_stride, int src_stride, int width, int height, int log2_scale)

109 {int y, x;

110 #define STORE(pos) \

111 temp= (src[x + pos] + (d[pos]>>log2_scale))>>(6-log2_scale); \

112 src[x + pos]=src[x + pos - 8*src_stride]=0; \

113 if(temp & 0x100) temp= ~(temp>>31); \

114 dst[x + pos]= temp;

115

116 for(y=0; y<height; y++){

117 const uint8_t *d= dither[y];

118 for(x=0; x<width; x+=8){

119 int temp;

120 STORE(0);

121 STORE(1);

122 STORE(2);

123 STORE(3);

124 STORE(4);

125 STORE(5);

126 STORE(6);

127 STORE(7);

128 }

129 src+=src_stride;

130 dst+=dst_stride;

131 }

132 }

133

134 //This func reads from 2 slices, 0 & 2 and clears 2-nd

135 static void store_slice2_c(uint8_t *dst, int16_t *src, int dst_stride, int src_stride, int width, int height, int log2_scale)

136 {int y, x;

137 #define STORE2(pos) \

138 temp= (src[x + pos] + src[x + pos + 16*src_stride] + (d[pos]>>log2_scale))>>(6-log2_scale); \

139 src[x + pos + 16*src_stride]=0; \

140 if(temp & 0x100) temp= ~(temp>>31); \

141 dst[x + pos]= temp;

142

143 for(y=0; y<height; y++){

144 const uint8_t *d= dither[y];

145 for(x=0; x<width; x+=8){

146 int temp;

147 STORE2(0);

148 STORE2(1);

149 STORE2(2);

150 STORE2(3);

151 STORE2(4);

152 STORE2(5);

153 STORE2(6);

154 STORE2(7);

155 }

156 src+=src_stride;

157 dst+=dst_stride;

158 }

159 }

160

161 static void mul_thrmat_c(struct vf_priv_s *p,int q)

162 {

163 int a;

164 for(a=0;a<64;a++)

165 ((short*)p->threshold_mtx)[a]=q * ((short*)p->threshold_mtx_noq)[a];//ints faster in C

166 }

167

168 static void column_fidct_c(int16_t* thr_adr, int16_t *data, int16_t *output, int cnt);

169 static void row_idct_c(int16_t* workspace,

170 int16_t* output_adr, int output_stride, int cnt);

171 static void row_fdct_c(int16_t *data, const uint8_t *pixels, int line_size, int cnt);

172

173 //this is rather ugly, but there is no need for function pointers

174 #define store_slice_s store_slice_c

175 #define store_slice2_s store_slice2_c

176 #define mul_thrmat_s mul_thrmat_c

177 #define column_fidct_s column_fidct_c

178 #define row_idct_s row_idct_c

179 #define row_fdct_s row_fdct_c

180

181 #else /* HAVE_MMX */

182

183 //This func reads from 1 slice, 1 and clears 0 & 1

184 static void store_slice_mmx(uint8_t *dst, int16_t *src, long dst_stride, long src_stride, long width, long height, long log2_scale)

185 {

186 const uint8_t *od=&dither[0][0];

187 const uint8_t *end=&dither[height][0];

188 width = (width+7)&~7;

189 dst_stride-=width;

190 //src_stride=(src_stride-width)*2;

191 __asm__ volatile(

192 "mov %5, %%"REG_d" \n\t"

193 "mov %6, %%"REG_S" \n\t"

194 "mov %7, %%"REG_D" \n\t"

195 "mov %1, %%"REG_a" \n\t"

196 "movd %%"REG_d", %%mm5 \n\t"

197 "xor $-1, %%"REG_d" \n\t"

198 "mov %%"REG_a", %%"REG_c" \n\t"

199 "add 7,ドル %%"REG_d" \n\t"

200 "neg %%"REG_a" \n\t"

201 "sub %0, %%"REG_c" \n\t"

202 "add %%"REG_c", %%"REG_c" \n\t"

203 "movd %%"REG_d", %%mm2 \n\t"

204 "mov %%"REG_c", %1 \n\t"

205 "mov %2, %%"REG_d" \n\t"

206 "shl 4,ドル %%"REG_a" \n\t"

207

208 "2: \n\t"

209 "movq (%%"REG_d"), %%mm3 \n\t"

210 "movq %%mm3, %%mm4 \n\t"

211 "pxor %%mm7, %%mm7 \n\t"

212 "punpcklbw %%mm7, %%mm3 \n\t"

213 "punpckhbw %%mm7, %%mm4 \n\t"

214 "mov %0, %%"REG_c" \n\t"

215 "psraw %%mm5, %%mm3 \n\t"

216 "psraw %%mm5, %%mm4 \n\t"

217 "1: \n\t"

218 "movq %%mm7, (%%"REG_S",%%"REG_a",) \n\t"

219 "movq (%%"REG_S"), %%mm0 \n\t"

220 "movq 8(%%"REG_S"), %%mm1 \n\t"

221

222 "movq %%mm7, 8(%%"REG_S",%%"REG_a",) \n\t"

223 "paddw %%mm3, %%mm0 \n\t"

224 "paddw %%mm4, %%mm1 \n\t"

225

226 "movq %%mm7, (%%"REG_S") \n\t"

227 "psraw %%mm2, %%mm0 \n\t"

228 "psraw %%mm2, %%mm1 \n\t"

229

230 "movq %%mm7, 8(%%"REG_S") \n\t"

231 "packuswb %%mm1, %%mm0 \n\t"

232 "add 16,ドル %%"REG_S" \n\t"

233

234 "movq %%mm0, (%%"REG_D") \n\t"

235 "add 8,ドル %%"REG_D" \n\t"

236 "sub 8,ドル %%"REG_c" \n\t"

237 "jg 1b \n\t"

238 "add %1, %%"REG_S" \n\t"

239 "add 8,ドル %%"REG_d" \n\t"

240 "add %3, %%"REG_D" \n\t"

241 "cmp %4, %%"REG_d" \n\t"

242 "jl 2b \n\t"

243

244 :

245 : "m" (width), "m" (src_stride), "erm" (od), "m" (dst_stride), "erm" (end),

246 "m" (log2_scale), "m" (src), "m" (dst) //input

247 : "%"REG_a, "%"REG_c, "%"REG_d, "%"REG_S, "%"REG_D

248 );

249 }

250

251 //This func reads from 2 slices, 0 & 2 and clears 2-nd

252 static void store_slice2_mmx(uint8_t *dst, int16_t *src, long dst_stride, long src_stride, long width, long height, long log2_scale)

253 {

254 const uint8_t *od=&dither[0][0];

255 const uint8_t *end=&dither[height][0];

256 width = (width+7)&~7;

257 dst_stride-=width;

258 //src_stride=(src_stride-width)*2;

259 __asm__ volatile(

260 "mov %5, %%"REG_d" \n\t"

261 "mov %6, %%"REG_S" \n\t"

262 "mov %7, %%"REG_D" \n\t"

263 "mov %1, %%"REG_a" \n\t"

264 "movd %%"REG_d", %%mm5 \n\t"

265 "xor $-1, %%"REG_d" \n\t"

266 "mov %%"REG_a", %%"REG_c" \n\t"

267 "add 7,ドル %%"REG_d" \n\t"

268 "sub %0, %%"REG_c" \n\t"

269 "add %%"REG_c", %%"REG_c" \n\t"

270 "movd %%"REG_d", %%mm2 \n\t"

271 "mov %%"REG_c", %1 \n\t"

272 "mov %2, %%"REG_d" \n\t"

273 "shl 5,ドル %%"REG_a" \n\t"

274

275 "2: \n\t"

276 "movq (%%"REG_d"), %%mm3 \n\t"

277 "movq %%mm3, %%mm4 \n\t"

278 "pxor %%mm7, %%mm7 \n\t"

279 "punpcklbw %%mm7, %%mm3 \n\t"

280 "punpckhbw %%mm7, %%mm4 \n\t"

281 "mov %0, %%"REG_c" \n\t"

282 "psraw %%mm5, %%mm3 \n\t"

283 "psraw %%mm5, %%mm4 \n\t"

284 "1: \n\t"

285 "movq (%%"REG_S"), %%mm0 \n\t"

286 "movq 8(%%"REG_S"), %%mm1 \n\t"

287 "paddw %%mm3, %%mm0 \n\t"

288

289 "paddw (%%"REG_S",%%"REG_a",), %%mm0 \n\t"

290 "paddw %%mm4, %%mm1 \n\t"

291 "movq 8(%%"REG_S",%%"REG_a",), %%mm6 \n\t"

292

293 "movq %%mm7, (%%"REG_S",%%"REG_a",) \n\t"

294 "psraw %%mm2, %%mm0 \n\t"

295 "paddw %%mm6, %%mm1 \n\t"

296

297 "movq %%mm7, 8(%%"REG_S",%%"REG_a",) \n\t"

298 "psraw %%mm2, %%mm1 \n\t"

299 "packuswb %%mm1, %%mm0 \n\t"

300

301 "movq %%mm0, (%%"REG_D") \n\t"

302 "add 16,ドル %%"REG_S" \n\t"

303 "add 8,ドル %%"REG_D" \n\t"

304 "sub 8,ドル %%"REG_c" \n\t"

305 "jg 1b \n\t"

306 "add %1, %%"REG_S" \n\t"

307 "add 8,ドル %%"REG_d" \n\t"

308 "add %3, %%"REG_D" \n\t"

309 "cmp %4, %%"REG_d" \n\t"

310 "jl 2b \n\t"

311

312 :

313 : "m" (width), "m" (src_stride), "erm" (od), "m" (dst_stride), "erm" (end),

314 "m" (log2_scale), "m" (src), "m" (dst) //input

315 : "%"REG_a, "%"REG_c, "%"REG_d, "%"REG_D, "%"REG_S

316 );

317 }

318

319 static void mul_thrmat_mmx(struct vf_priv_s *p, int q)

320 {

321 uint64_t *adr=&p->threshold_mtx_noq[0];

322 __asm__ volatile(

323 "movd %0, %%mm7 \n\t"

324 "add 8ドル*8*2, %%"REG_D" \n\t"

325 "movq 0*8(%%"REG_S"), %%mm0 \n\t"

326 "punpcklwd %%mm7, %%mm7 \n\t"

327 "movq 1*8(%%"REG_S"), %%mm1 \n\t"

328 "punpckldq %%mm7, %%mm7 \n\t"

329 "pmullw %%mm7, %%mm0 \n\t"

330

331 "movq 2*8(%%"REG_S"), %%mm2 \n\t"

332 "pmullw %%mm7, %%mm1 \n\t"

333

334 "movq 3*8(%%"REG_S"), %%mm3 \n\t"

335 "pmullw %%mm7, %%mm2 \n\t"

336

337 "movq %%mm0, 0*8(%%"REG_D") \n\t"

338 "movq 4*8(%%"REG_S"), %%mm4 \n\t"

339 "pmullw %%mm7, %%mm3 \n\t"

340

341 "movq %%mm1, 1*8(%%"REG_D") \n\t"

342 "movq 5*8(%%"REG_S"), %%mm5 \n\t"

343 "pmullw %%mm7, %%mm4 \n\t"

344

345 "movq %%mm2, 2*8(%%"REG_D") \n\t"

346 "movq 6*8(%%"REG_S"), %%mm6 \n\t"

347 "pmullw %%mm7, %%mm5 \n\t"

348

349 "movq %%mm3, 3*8(%%"REG_D") \n\t"

350 "movq 7*8+0*8(%%"REG_S"), %%mm0 \n\t"

351 "pmullw %%mm7, %%mm6 \n\t"

352

353 "movq %%mm4, 4*8(%%"REG_D") \n\t"

354 "movq 7*8+1*8(%%"REG_S"), %%mm1 \n\t"

355 "pmullw %%mm7, %%mm0 \n\t"

356

357 "movq %%mm5, 5*8(%%"REG_D") \n\t"

358 "movq 7*8+2*8(%%"REG_S"), %%mm2 \n\t"

359 "pmullw %%mm7, %%mm1 \n\t"

360

361 "movq %%mm6, 6*8(%%"REG_D") \n\t"

362 "movq 7*8+3*8(%%"REG_S"), %%mm3 \n\t"

363 "pmullw %%mm7, %%mm2 \n\t"

364

365 "movq %%mm0, 7*8+0*8(%%"REG_D") \n\t"

366 "movq 7*8+4*8(%%"REG_S"), %%mm4 \n\t"

367 "pmullw %%mm7, %%mm3 \n\t"

368

369 "movq %%mm1, 7*8+1*8(%%"REG_D") \n\t"

370 "movq 7*8+5*8(%%"REG_S"), %%mm5 \n\t"

371 "pmullw %%mm7, %%mm4 \n\t"

372

373 "movq %%mm2, 7*8+2*8(%%"REG_D") \n\t"

374 "movq 7*8+6*8(%%"REG_S"), %%mm6 \n\t"

375 "pmullw %%mm7, %%mm5 \n\t"

376

377 "movq %%mm3, 7*8+3*8(%%"REG_D") \n\t"

378 "movq 14*8+0*8(%%"REG_S"), %%mm0 \n\t"

379 "pmullw %%mm7, %%mm6 \n\t"

380

381 "movq %%mm4, 7*8+4*8(%%"REG_D") \n\t"

382 "movq 14*8+1*8(%%"REG_S"), %%mm1 \n\t"

383 "pmullw %%mm7, %%mm0 \n\t"

384

385 "movq %%mm5, 7*8+5*8(%%"REG_D") \n\t"

386 "pmullw %%mm7, %%mm1 \n\t"

387

388 "movq %%mm6, 7*8+6*8(%%"REG_D") \n\t"

389 "movq %%mm0, 14*8+0*8(%%"REG_D") \n\t"

390 "movq %%mm1, 14*8+1*8(%%"REG_D") \n\t"

391

392 : "+g" (q), "+S" (adr), "+D" (adr)

393 :

394 );

395 }

396

397 static void column_fidct_mmx(int16_t* thr_adr, int16_t *data, int16_t *output, int cnt);

398 static void row_idct_mmx(int16_t* workspace,

399 int16_t* output_adr, int output_stride, int cnt);

400 static void row_fdct_mmx(int16_t *data, const uint8_t *pixels, int line_size, int cnt);

401

402 #define store_slice_s store_slice_mmx

403 #define store_slice2_s store_slice2_mmx

404 #define mul_thrmat_s mul_thrmat_mmx

405 #define column_fidct_s column_fidct_mmx

406 #define row_idct_s row_idct_mmx

407 #define row_fdct_s row_fdct_mmx

408 #endif // HAVE_MMX

409

410 static void filter(struct vf_priv_s *p, uint8_t *dst, uint8_t *src,

411 int dst_stride, int src_stride,

412 int width, int height,

413 uint8_t *qp_store, int qp_stride, int is_luma)

414 {

415 int x, x0, y, es, qy, t;

416 const int stride= is_luma ? p->temp_stride : (width+16);//((width+16+15)&(~15))

417 const int step=6-p->log2_count;

418 const int qps= 3 + is_luma;

419 int32_t __attribute__((aligned(32))) block_align[4*8*BLOCKSZ+ 4*8*BLOCKSZ];

420 int16_t *block= (int16_t *)block_align;

421 int16_t *block3=(int16_t *)(block_align+4*8*BLOCKSZ);

422

423 memset(block3, 0, 4*8*BLOCKSZ);

424

425 //p->src=src-src_stride*8-8;//!

426 if (!src || !dst) return; // HACK avoid crash for Y8 colourspace

427 for(y=0; y<height; y++){

428 int index= 8 + 8*stride + y*stride;

429 fast_memcpy(p->src + index, src + y*src_stride, width);//this line can be avoided by using DR & user fr.buffers

430 for(x=0; x<8; x++){

431 p->src[index - x - 1]= p->src[index + x ];

432 p->src[index + width + x ]= p->src[index + width - x - 1];

433 }

434 }

435 for(y=0; y<8; y++){

436 fast_memcpy(p->src + ( 7-y)*stride, p->src + ( y+8)*stride, stride);

437 fast_memcpy(p->src + (height+8+y)*stride, p->src + (height-y+7)*stride, stride);

438 }

439 //FIXME (try edge emu)

440

441 for(y=8; y<24; y++)

442 memset(p->temp+ 8 +y*stride, 0,width*sizeof(int16_t));

443

444 for(y=step; y<height+8; y+=step){ //step= 1,2

445 qy=y-4;

446 if (qy>height-1) qy=height-1;

447 if (qy<0) qy=0;

448 qy=(qy>>qps)*qp_stride;

449 row_fdct_s(block, p->src + y*stride +2-(y&1), stride, 2);

450 for(x0=0; x0<width+8-8*(BLOCKSZ-1); x0+=8*(BLOCKSZ-1)){

451 row_fdct_s(block+8*8, p->src + y*stride+8+x0 +2-(y&1), stride, 2*(BLOCKSZ-1));

452 if(p->qp)

453 column_fidct_s((int16_t*)(&p->threshold_mtx[0]), block+0*8, block3+0*8, 8*(BLOCKSZ-1)); //yes, this is a HOTSPOT

454 else

455 for (x=0; x<8*(BLOCKSZ-1); x+=8) {

456 t=x+x0-2; //correct t=x+x0-2-(y&1), but its the same

457 if (t<0) t=0;//t always < width-2

458 t=qp_store[qy+(t>>qps)];

459 t=norm_qscale(t, p->mpeg2);

460 if (t!=p->prev_q) p->prev_q=t, mul_thrmat_s(p, t);

461 column_fidct_s((int16_t*)(&p->threshold_mtx[0]), block+x*8, block3+x*8, 8); //yes, this is a HOTSPOT

462 }

463 row_idct_s(block3+0*8, p->temp + (y&15)*stride+x0+2-(y&1), stride, 2*(BLOCKSZ-1));

464 memmove(block, block+(BLOCKSZ-1)*64, 8*8*sizeof(int16_t)); //cycling

465 memmove(block3, block3+(BLOCKSZ-1)*64, 6*8*sizeof(int16_t));

466 }

467 //

468 es=width+8-x0; // 8, ...

469 if (es>8)

470 row_fdct_s(block+8*8, p->src + y*stride+8+x0 +2-(y&1), stride, (es-4)>>2);

471 column_fidct_s((int16_t*)(&p->threshold_mtx[0]), block, block3, es&(~1));

472 row_idct_s(block3+0*8, p->temp + (y&15)*stride+x0+2-(y&1), stride, es>>2);

473 {const int y1=y-8+step;//l5-7 l4-6

474 if (!(y1&7) && y1) {

475 if (y1&8) store_slice_s(dst + (y1-8)*dst_stride, p->temp+ 8 +8*stride,

476 dst_stride, stride, width, 8, 5-p->log2_count);

477 else store_slice2_s(dst + (y1-8)*dst_stride, p->temp+ 8 +0*stride,

478 dst_stride, stride, width, 8, 5-p->log2_count);

479 } }

480 }

481

482 if (y&7) { // == height & 7

483 if (y&8) store_slice_s(dst + ((y-8)&~7)*dst_stride, p->temp+ 8 +8*stride,

484 dst_stride, stride, width, y&7, 5-p->log2_count);

485 else store_slice2_s(dst + ((y-8)&~7)*dst_stride, p->temp+ 8 +0*stride,

486 dst_stride, stride, width, y&7, 5-p->log2_count);

487 }

488 }

489

490 static int config(struct vf_instance *vf,

491 int width, int height, int d_width, int d_height,

492 unsigned int flags, unsigned int outfmt)

493 {

494 int h= (height+16+15)&(~15);

495

496 vf->priv->temp_stride= (width+16+15)&(~15);

497 vf->priv->temp= (int16_t*)av_mallocz(vf->priv->temp_stride*3*8*sizeof(int16_t));

498 //this can also be avoided, see above

499 vf->priv->src = (uint8_t*)av_malloc(vf->priv->temp_stride*h*sizeof(uint8_t));

500

501 return ff_vf_next_config(vf,width,height,d_width,d_height,flags,outfmt);

502 }

503

504 static void get_image(struct vf_instance *vf, mp_image_t *mpi)

505 {

506 if(mpi->flags&MP_IMGFLAG_PRESERVE) return; // don't change

507 // ok, we can do pp in-place (or pp disabled):

508 vf->dmpi=ff_vf_get_image(vf->next,mpi->imgfmt,

509 mpi->type, mpi->flags, mpi->width, mpi->height);

510 mpi->planes[0]=vf->dmpi->planes[0];

511 mpi->stride[0]=vf->dmpi->stride[0];

512 mpi->width=vf->dmpi->width;

513 if(mpi->flags&MP_IMGFLAG_PLANAR){

514 mpi->planes[1]=vf->dmpi->planes[1];

515 mpi->planes[2]=vf->dmpi->planes[2];

516 mpi->stride[1]=vf->dmpi->stride[1];

517 mpi->stride[2]=vf->dmpi->stride[2];

518 }

519 mpi->flags|=MP_IMGFLAG_DIRECT;

520 }

521

522 static int put_image(struct vf_instance *vf, mp_image_t *mpi, double pts)

523 {

524 mp_image_t *dmpi;

525 if(!(mpi->flags&MP_IMGFLAG_DIRECT)){

526 // no DR, so get a new image! hope we'll get DR buffer:

527 dmpi=ff_vf_get_image(vf->next,mpi->imgfmt,

528 MP_IMGTYPE_TEMP,

529 MP_IMGFLAG_ACCEPT_STRIDE|MP_IMGFLAG_PREFER_ALIGNED_STRIDE,

530 mpi->width,mpi->height);

531 ff_vf_clone_mpi_attributes(dmpi, mpi);

532 }else{

533 dmpi=vf->dmpi;

534 }

535

536 vf->priv->mpeg2= mpi->qscale_type;

537 if(mpi->pict_type != 3 && mpi->qscale && !vf->priv->qp){

538 int w = mpi->qstride;

539 int h = (mpi->h + 15) >> 4;

540 if (!w) {

541 w = (mpi->w + 15) >> 4;

542 h = 1;

543 }

544 if(!vf->priv->non_b_qp)

545 vf->priv->non_b_qp= malloc(w*h);

546 fast_memcpy(vf->priv->non_b_qp, mpi->qscale, w*h);

547 }

548 if(vf->priv->log2_count || !(mpi->flags&MP_IMGFLAG_DIRECT)){

549 char *qp_tab= vf->priv->non_b_qp;

550 if(vf->priv->bframes || !qp_tab)

551 qp_tab= mpi->qscale;

552

553 if(qp_tab || vf->priv->qp){

554 filter(vf->priv, dmpi->planes[0], mpi->planes[0], dmpi->stride[0], mpi->stride[0],

555 mpi->w, mpi->h, qp_tab, mpi->qstride, 1);

556 filter(vf->priv, dmpi->planes[1], mpi->planes[1], dmpi->stride[1], mpi->stride[1],

557 mpi->w>>mpi->chroma_x_shift, mpi->h>>mpi->chroma_y_shift, qp_tab, mpi->qstride, 0);

558 filter(vf->priv, dmpi->planes[2], mpi->planes[2], dmpi->stride[2], mpi->stride[2],

559 mpi->w>>mpi->chroma_x_shift, mpi->h>>mpi->chroma_y_shift, qp_tab, mpi->qstride, 0);

560 }else{

561 memcpy_pic(dmpi->planes[0], mpi->planes[0], mpi->w, mpi->h, dmpi->stride[0], mpi->stride[0]);

562 memcpy_pic(dmpi->planes[1], mpi->planes[1], mpi->w>>mpi->chroma_x_shift, mpi->h>>mpi->chroma_y_shift, dmpi->stride[1], mpi->stride[1]);

563 memcpy_pic(dmpi->planes[2], mpi->planes[2], mpi->w>>mpi->chroma_x_shift, mpi->h>>mpi->chroma_y_shift, dmpi->stride[2], mpi->stride[2]);

564 }

565 }

566

567 #if HAVE_MMX

568 if(ff_gCpuCaps.hasMMX) __asm__ volatile ("emms\n\t");

569 #endif

570 #if HAVE_MMX2

571 if(ff_gCpuCaps.hasMMX2) __asm__ volatile ("sfence\n\t");

572 #endif

573 return ff_vf_next_put_image(vf,dmpi, pts);

574 }

575

576 static void uninit(struct vf_instance *vf)

577 {

578 if(!vf->priv) return;

579

580 av_free(vf->priv->temp);

581 vf->priv->temp= NULL;

582 av_free(vf->priv->src);

583 vf->priv->src= NULL;

584 //free(vf->priv->avctx);

585 //vf->priv->avctx= NULL;

586 free(vf->priv->non_b_qp);

587 vf->priv->non_b_qp= NULL;

588

589 av_free(vf->priv);

590 vf->priv=NULL;

591 }

592

593 //===========================================================================//

594

595 static int query_format(struct vf_instance *vf, unsigned int fmt)

596 {

597 switch(fmt){

598 case IMGFMT_YVU9:

599 case IMGFMT_IF09:

600 case IMGFMT_YV12:

601 case IMGFMT_I420:

602 case IMGFMT_IYUV:

603 case IMGFMT_CLPL:

604 case IMGFMT_Y800:

605 case IMGFMT_Y8:

606 case IMGFMT_444P:

607 case IMGFMT_422P:

608 case IMGFMT_411P:

609 return ff_vf_next_query_format(vf,fmt);

610 }

611 return 0;

612 }

613

614 static int control(struct vf_instance *vf, int request, void* data)

615 {

616 switch(request){

617 case VFCTRL_QUERY_MAX_PP_LEVEL:

618 return 5;

619 case VFCTRL_SET_PP_LEVEL:

620 vf->priv->log2_count= *((unsigned int*)data);

621 if (vf->priv->log2_count < 4) vf->priv->log2_count=4;

622 return CONTROL_TRUE;

623 }

624 return ff_vf_next_control(vf,request,data);

625 }

626

627 static int vf_open(vf_instance_t *vf, char *args)

628 {

629 int i=0, bias;

630 int custom_threshold_m[64];

631 int log2c=-1;

632

633 vf->config=config;

634 vf->put_image=put_image;

635 vf->get_image=get_image;

636 vf->query_format=query_format;

637 vf->uninit=uninit;

638 vf->control= control;

639 vf->priv=av_mallocz(sizeof(struct vf_priv_s));//assumes align 16 !

640

641 ff_init_avcodec();

642

643 //vf->priv->avctx= avcodec_alloc_context();

644 //dsputil_init(&vf->priv->dsp, vf->priv->avctx);

645

646 vf->priv->log2_count= 4;

647 vf->priv->bframes = 0;

648

649 if (args) sscanf(args, "%d:%d:%d:%d", &log2c, &vf->priv->qp, &i, &vf->priv->bframes);

650

651 if( log2c >=4 && log2c <=5 )

652 vf->priv->log2_count = log2c;

653 else if( log2c >= 6 )

654 vf->priv->log2_count = 5;

655

656 if(vf->priv->qp < 0)

657 vf->priv->qp = 0;

658

659 if (i < -15) i = -15;

660 if (i > 32) i = 32;

661

662 bias= (1<<4)+i; //regulable

663 vf->priv->prev_q=0;

664 //

665 for(i=0;i<64;i++) //FIXME: tune custom_threshold[] and remove this !

666 custom_threshold_m[i]=(int)(custom_threshold[i]*(bias/71.)+ 0.5);

667 for(i=0;i<8;i++){

668 vf->priv->threshold_mtx_noq[2*i]=(uint64_t)custom_threshold_m[i*8+2]

669 |(((uint64_t)custom_threshold_m[i*8+6])<<16)

670 |(((uint64_t)custom_threshold_m[i*8+0])<<32)

671 |(((uint64_t)custom_threshold_m[i*8+4])<<48);

672 vf->priv->threshold_mtx_noq[2*i+1]=(uint64_t)custom_threshold_m[i*8+5]

673 |(((uint64_t)custom_threshold_m[i*8+3])<<16)

674 |(((uint64_t)custom_threshold_m[i*8+1])<<32)

675 |(((uint64_t)custom_threshold_m[i*8+7])<<48);

676 }

677

678 if (vf->priv->qp) vf->priv->prev_q=vf->priv->qp, mul_thrmat_s(vf->priv, vf->priv->qp);

679

680 return 1;

681 }

682

683 const vf_info_t ff_vf_info_fspp = {

684 "fast simple postprocess",

685 "fspp",

686 "Michael Niedermayer, Nikolaj Poroshin",

687 "",

688 vf_open,

689 NULL

690 };

691

692 //====================================================================

693 //Specific spp's dct, idct and threshold functions

694 //I'd prefer to have them in the separate file.

695

696 //#define MANGLE(a) #a

697

698 //typedef int16_t int16_t; //! only int16_t

699

700 #define DCTSIZE 8

701 #define DCTSIZE_S "8"

702

703 #define FIX(x,s) ((int) ((x) * (1<<s) + 0.5)&0xffff)

704 #define C64(x) ((uint64_t)((x)|(x)<<16))<<32 | (uint64_t)(x) | (uint64_t)(x)<<16

705 #define FIX64(x,s) C64(FIX(x,s))

706

707 #define MULTIPLY16H(x,k) (((x)*(k))>>16)

708 #define THRESHOLD(r,x,t) if(((unsigned)((x)+t))>t*2) r=(x);else r=0;

709 #define DESCALE(x,n) (((x) + (1 << ((n)-1))) >> n)

710

711 #if HAVE_MMX

712

713 DECLARE_ASM_CONST(8, uint64_t, MM_FIX_0_382683433)=FIX64(0.382683433, 14);

714 DECLARE_ALIGNED(8, uint64_t, ff_MM_FIX_0_541196100)=FIX64(0.541196100, 14);

715 DECLARE_ALIGNED(8, uint64_t, ff_MM_FIX_0_707106781)=FIX64(0.707106781, 14);

716 DECLARE_ASM_CONST(8, uint64_t, MM_FIX_1_306562965)=FIX64(1.306562965, 14);

717

718 DECLARE_ASM_CONST(8, uint64_t, MM_FIX_1_414213562_A)=FIX64(1.414213562, 14);

719

720 DECLARE_ASM_CONST(8, uint64_t, MM_FIX_1_847759065)=FIX64(1.847759065, 13);

721 DECLARE_ASM_CONST(8, uint64_t, MM_FIX_2_613125930)=FIX64(-2.613125930, 13); //-

722 DECLARE_ASM_CONST(8, uint64_t, MM_FIX_1_414213562)=FIX64(1.414213562, 13);

723 DECLARE_ASM_CONST(8, uint64_t, MM_FIX_1_082392200)=FIX64(1.082392200, 13);

724 //for t3,t5,t7 == 0 shortcut

725 DECLARE_ASM_CONST(8, uint64_t, MM_FIX_0_847759065)=FIX64(0.847759065, 14);

726 DECLARE_ASM_CONST(8, uint64_t, MM_FIX_0_566454497)=FIX64(0.566454497, 14);

727 DECLARE_ASM_CONST(8, uint64_t, MM_FIX_0_198912367)=FIX64(0.198912367, 14);

728

729 DECLARE_ASM_CONST(8, uint64_t, MM_DESCALE_RND)=C64(4);

730 DECLARE_ASM_CONST(8, uint64_t, MM_2)=C64(2);

731

732 #else /* !HAVE_MMX */

733

734 typedef int32_t int_simd16_t;

735 static const int16_t FIX_0_382683433=FIX(0.382683433, 14);

736 static const int16_t FIX_0_541196100=FIX(0.541196100, 14);

737 static const int16_t FIX_0_707106781=FIX(0.707106781, 14);

738 static const int16_t FIX_1_306562965=FIX(1.306562965, 14);

739 static const int16_t FIX_1_414213562_A=FIX(1.414213562, 14);

740 static const int16_t FIX_1_847759065=FIX(1.847759065, 13);

741 static const int16_t FIX_2_613125930=FIX(-2.613125930, 13); //-

742 static const int16_t FIX_1_414213562=FIX(1.414213562, 13);

743 static const int16_t FIX_1_082392200=FIX(1.082392200, 13);

744

745 #endif

746

747 #if !HAVE_MMX

748

749 static void column_fidct_c(int16_t* thr_adr, int16_t *data, int16_t *output, int cnt)

750 {

751 int_simd16_t tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;

752 int_simd16_t tmp10, tmp11, tmp12, tmp13;

753 int_simd16_t z1,z2,z3,z4,z5, z10, z11, z12, z13;

754 int_simd16_t d0, d1, d2, d3, d4, d5, d6, d7;

755

756 int16_t* dataptr;

757 int16_t* wsptr;

758 int16_t *threshold;

759 int ctr;

760

761 dataptr = data;

762 wsptr = output;

763

764 for (; cnt > 0; cnt-=2) { //start positions

765 threshold=(int16_t*)thr_adr;//threshold_mtx

766 for (ctr = DCTSIZE; ctr > 0; ctr--) {

767 // Process columns from input, add to output.

768 tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*7];

769 tmp7 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*7];

770

771 tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*6];

772 tmp6 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*6];

773

774 tmp2 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*5];

775 tmp5 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*5];

776

777 tmp3 = dataptr[DCTSIZE*3] + dataptr[DCTSIZE*4];

778 tmp4 = dataptr[DCTSIZE*3] - dataptr[DCTSIZE*4];

779

780 // Even part of FDCT

781

782 tmp10 = tmp0 + tmp3;

783 tmp13 = tmp0 - tmp3;

784 tmp11 = tmp1 + tmp2;

785 tmp12 = tmp1 - tmp2;

786

787 d0 = tmp10 + tmp11;

788 d4 = tmp10 - tmp11;

789

790 z1 = MULTIPLY16H((tmp12 + tmp13) <<2, FIX_0_707106781);

791 d2 = tmp13 + z1;

792 d6 = tmp13 - z1;

793

794 // Even part of IDCT

795

796 THRESHOLD(tmp0, d0, threshold[0*8]);

797 THRESHOLD(tmp1, d2, threshold[2*8]);

798 THRESHOLD(tmp2, d4, threshold[4*8]);

799 THRESHOLD(tmp3, d6, threshold[6*8]);

800 tmp0+=2;

801 tmp10 = (tmp0 + tmp2)>>2;

802 tmp11 = (tmp0 - tmp2)>>2;

803

804 tmp13 = (tmp1 + tmp3)>>2; //+2 ! (psnr decides)

805 tmp12 = MULTIPLY16H((tmp1 - tmp3), FIX_1_414213562_A) - tmp13; //<<2

806

807 tmp0 = tmp10 + tmp13; //->temps

808 tmp3 = tmp10 - tmp13; //->temps

809 tmp1 = tmp11 + tmp12; //->temps

810 tmp2 = tmp11 - tmp12; //->temps

811

812 // Odd part of FDCT

813

814 tmp10 = tmp4 + tmp5;

815 tmp11 = tmp5 + tmp6;

816 tmp12 = tmp6 + tmp7;

817

818 z5 = MULTIPLY16H((tmp10 - tmp12)<<2, FIX_0_382683433);

819 z2 = MULTIPLY16H(tmp10 <<2, FIX_0_541196100) + z5;

820 z4 = MULTIPLY16H(tmp12 <<2, FIX_1_306562965) + z5;

821 z3 = MULTIPLY16H(tmp11 <<2, FIX_0_707106781);

822

823 z11 = tmp7 + z3;

824 z13 = tmp7 - z3;

825

826 d5 = z13 + z2;

827 d3 = z13 - z2;

828 d1 = z11 + z4;

829 d7 = z11 - z4;

830

831 // Odd part of IDCT

832

833 THRESHOLD(tmp4, d1, threshold[1*8]);

834 THRESHOLD(tmp5, d3, threshold[3*8]);

835 THRESHOLD(tmp6, d5, threshold[5*8]);

836 THRESHOLD(tmp7, d7, threshold[7*8]);

837

838 //Simd version uses here a shortcut for the tmp5,tmp6,tmp7 == 0

839 z13 = tmp6 + tmp5;

840 z10 = (tmp6 - tmp5)<<1;

841 z11 = tmp4 + tmp7;

842 z12 = (tmp4 - tmp7)<<1;

843

844 tmp7 = (z11 + z13)>>2; //+2 !

845 tmp11 = MULTIPLY16H((z11 - z13)<<1, FIX_1_414213562);

846 z5 = MULTIPLY16H(z10 + z12, FIX_1_847759065);

847 tmp10 = MULTIPLY16H(z12, FIX_1_082392200) - z5;

848 tmp12 = MULTIPLY16H(z10, FIX_2_613125930) + z5; // - !!

849

850 tmp6 = tmp12 - tmp7;

851 tmp5 = tmp11 - tmp6;

852 tmp4 = tmp10 + tmp5;

853

854 wsptr[DCTSIZE*0]+= (tmp0 + tmp7);

855 wsptr[DCTSIZE*1]+= (tmp1 + tmp6);

856 wsptr[DCTSIZE*2]+= (tmp2 + tmp5);

857 wsptr[DCTSIZE*3]+= (tmp3 - tmp4);

858 wsptr[DCTSIZE*4]+= (tmp3 + tmp4);

859 wsptr[DCTSIZE*5]+= (tmp2 - tmp5);

860 wsptr[DCTSIZE*6]= (tmp1 - tmp6);

861 wsptr[DCTSIZE*7]= (tmp0 - tmp7);

862 //

863 dataptr++; //next column

864 wsptr++;

865 threshold++;

866 }

867 dataptr+=8; //skip each second start pos

868 wsptr +=8;

869 }

870 }

871

872 #else /* HAVE_MMX */

873

874 static void column_fidct_mmx(int16_t* thr_adr, int16_t *data, int16_t *output, int cnt)

875 {

876 uint64_t __attribute__((aligned(8))) temps[4];

877 __asm__ volatile(

878 ASMALIGN(4)

879 "1: \n\t"

880 "movq "DCTSIZE_S "*0*2(%%"REG_S"), %%mm1 \n\t"

881 //

882 "movq "DCTSIZE_S "*3*2(%%"REG_S"), %%mm7 \n\t"

883 "movq %%mm1, %%mm0 \n\t"

884

885 "paddw "DCTSIZE_S "*7*2(%%"REG_S"), %%mm1 \n\t" //t0

886 "movq %%mm7, %%mm3 \n\t"

887

888 "paddw "DCTSIZE_S "*4*2(%%"REG_S"), %%mm7 \n\t" //t3

889 "movq %%mm1, %%mm5 \n\t"

890

891 "movq "DCTSIZE_S "*1*2(%%"REG_S"), %%mm6 \n\t"

892 "psubw %%mm7, %%mm1 \n\t" //t13

893

894 "movq "DCTSIZE_S "*2*2(%%"REG_S"), %%mm2 \n\t"

895 "movq %%mm6, %%mm4 \n\t"

896

897 "paddw "DCTSIZE_S "*6*2(%%"REG_S"), %%mm6 \n\t" //t1

898 "paddw %%mm7, %%mm5 \n\t" //t10

899

900 "paddw "DCTSIZE_S "*5*2(%%"REG_S"), %%mm2 \n\t" //t2

901 "movq %%mm6, %%mm7 \n\t"

902

903 "paddw %%mm2, %%mm6 \n\t" //t11

904 "psubw %%mm2, %%mm7 \n\t" //t12

905

906 "movq %%mm5, %%mm2 \n\t"

907 "paddw %%mm6, %%mm5 \n\t" //d0

908 // i0 t13 t12 i3 i1 d0 - d4

909 "psubw %%mm6, %%mm2 \n\t" //d4

910 "paddw %%mm1, %%mm7 \n\t"

911

912 "movq 4*16(%%"REG_d"), %%mm6 \n\t"

913 "psllw 2,ドル %%mm7 \n\t"

914

915 "psubw 0*16(%%"REG_d"), %%mm5 \n\t"

916 "psubw %%mm6, %%mm2 \n\t"

917

918 "paddusw 0*16(%%"REG_d"), %%mm5 \n\t"

919 "paddusw %%mm6, %%mm2 \n\t"

920

921 "pmulhw "MANGLE(ff_MM_FIX_0_707106781)", %%mm7 \n\t"

922 //

923 "paddw 0*16(%%"REG_d"), %%mm5 \n\t"

924 "paddw %%mm6, %%mm2 \n\t"

925

926 "psubusw 0*16(%%"REG_d"), %%mm5 \n\t"

927 "psubusw %%mm6, %%mm2 \n\t"

928

929 //This func is totally compute-bound, operates at huge speed. So, DC shortcut

930 // at this place isn't worthwhile due to BTB miss penalty (checked on Pent. 3).

931 //However, typical numbers: nondc - 29%%, dc - 46%%, zero - 25%%. All <> 0 case is very rare.

932 "paddw "MANGLE(MM_2)", %%mm5 \n\t"

933 "movq %%mm2, %%mm6 \n\t"

934

935 "paddw %%mm5, %%mm2 \n\t"

936 "psubw %%mm6, %%mm5 \n\t"

937

938 "movq %%mm1, %%mm6 \n\t"

939 "paddw %%mm7, %%mm1 \n\t" //d2

940

941 "psubw 2*16(%%"REG_d"), %%mm1 \n\t"

942 "psubw %%mm7, %%mm6 \n\t" //d6

943

944 "movq 6*16(%%"REG_d"), %%mm7 \n\t"

945 "psraw 2,ドル %%mm5 \n\t"

946

947 "paddusw 2*16(%%"REG_d"), %%mm1 \n\t"

948 "psubw %%mm7, %%mm6 \n\t"

949 // t7 d2 /t11 t4 t6 - d6 /t10

950

951 "paddw 2*16(%%"REG_d"), %%mm1 \n\t"

952 "paddusw %%mm7, %%mm6 \n\t"

953

954 "psubusw 2*16(%%"REG_d"), %%mm1 \n\t"

955 "paddw %%mm7, %%mm6 \n\t"

956

957 "psubw "DCTSIZE_S "*4*2(%%"REG_S"), %%mm3 \n\t"

958 "psubusw %%mm7, %%mm6 \n\t"

959

960 //movq [edi+"DCTSIZE_S"*2*2], mm1

961 //movq [edi+"DCTSIZE_S"*6*2], mm6

962 "movq %%mm1, %%mm7 \n\t"

963 "psraw 2,ドル %%mm2 \n\t"

964

965 "psubw "DCTSIZE_S "*6*2(%%"REG_S"), %%mm4 \n\t"

966 "psubw %%mm6, %%mm1 \n\t"

967

968 "psubw "DCTSIZE_S "*7*2(%%"REG_S"), %%mm0 \n\t"

969 "paddw %%mm7, %%mm6 \n\t" //'t13

970

971 "psraw 2,ドル %%mm6 \n\t" //paddw mm6, MM_2 !! ---

972 "movq %%mm2, %%mm7 \n\t"

973

974 "pmulhw "MANGLE(MM_FIX_1_414213562_A)", %%mm1 \n\t"

975 "paddw %%mm6, %%mm2 \n\t" //'t0

976

977 "movq %%mm2, 0*8+%3 \n\t" //!

978 "psubw %%mm6, %%mm7 \n\t" //'t3

979

980 "movq "DCTSIZE_S "*2*2(%%"REG_S"), %%mm2 \n\t"

981 "psubw %%mm6, %%mm1 \n\t" //'t12

982

983 "psubw "DCTSIZE_S "*5*2(%%"REG_S"), %%mm2 \n\t" //t5

984 "movq %%mm5, %%mm6 \n\t"

985

986 "movq %%mm7, 3*8+%3 \n\t"

987 "paddw %%mm2, %%mm3 \n\t" //t10

988

989 "paddw %%mm4, %%mm2 \n\t" //t11

990 "paddw %%mm0, %%mm4 \n\t" //t12

991

992 "movq %%mm3, %%mm7 \n\t"

993 "psubw %%mm4, %%mm3 \n\t"

994

995 "psllw 2,ドル %%mm3 \n\t"

996 "psllw 2,ドル %%mm7 \n\t" //opt for P6

997

998 "pmulhw "MANGLE(MM_FIX_0_382683433)", %%mm3 \n\t"

999 "psllw 2,ドル %%mm4 \n\t"

1000

1001 "pmulhw "MANGLE(ff_MM_FIX_0_541196100)", %%mm7 \n\t"

1002 "psllw 2,ドル %%mm2 \n\t"

1003

1004 "pmulhw "MANGLE(MM_FIX_1_306562965)", %%mm4 \n\t"

1005 "paddw %%mm1, %%mm5 \n\t" //'t1

1006

1007 "pmulhw "MANGLE(ff_MM_FIX_0_707106781)", %%mm2 \n\t"

1008 "psubw %%mm1, %%mm6 \n\t" //'t2

1009 // t7 't12 't11 t4 t6 - 't13 't10 ---

1010

1011 "paddw %%mm3, %%mm7 \n\t" //z2

1012

1013 "movq %%mm5, 1*8+%3 \n\t"

1014 "paddw %%mm3, %%mm4 \n\t" //z4

1015

1016 "movq 3*16(%%"REG_d"), %%mm3 \n\t"

1017 "movq %%mm0, %%mm1 \n\t"

1018

1019 "movq %%mm6, 2*8+%3 \n\t"

1020 "psubw %%mm2, %%mm1 \n\t" //z13

1021

1022 //===

1023 "paddw %%mm2, %%mm0 \n\t" //z11

1024 "movq %%mm1, %%mm5 \n\t"

1025

1026 "movq 5*16(%%"REG_d"), %%mm2 \n\t"

1027 "psubw %%mm7, %%mm1 \n\t" //d3

1028

1029 "paddw %%mm7, %%mm5 \n\t" //d5

1030 "psubw %%mm3, %%mm1 \n\t"

1031

1032 "movq 1*16(%%"REG_d"), %%mm7 \n\t"

1033 "psubw %%mm2, %%mm5 \n\t"

1034

1035 "movq %%mm0, %%mm6 \n\t"

1036 "paddw %%mm4, %%mm0 \n\t" //d1

1037

1038 "paddusw %%mm3, %%mm1 \n\t"

1039 "psubw %%mm4, %%mm6 \n\t" //d7

1040

1041 // d1 d3 - - - d5 d7 -

1042 "movq 7*16(%%"REG_d"), %%mm4 \n\t"

1043 "psubw %%mm7, %%mm0 \n\t"

1044

1045 "psubw %%mm4, %%mm6 \n\t"

1046 "paddusw %%mm2, %%mm5 \n\t"

1047

1048 "paddusw %%mm4, %%mm6 \n\t"

1049 "paddw %%mm3, %%mm1 \n\t"

1050

1051 "paddw %%mm2, %%mm5 \n\t"

1052 "paddw %%mm4, %%mm6 \n\t"

1053

1054 "psubusw %%mm3, %%mm1 \n\t"

1055 "psubusw %%mm2, %%mm5 \n\t"

1056

1057 "psubusw %%mm4, %%mm6 \n\t"

1058 "movq %%mm1, %%mm4 \n\t"

1059

1060 "por %%mm5, %%mm4 \n\t"

1061 "paddusw %%mm7, %%mm0 \n\t"

1062

1063 "por %%mm6, %%mm4 \n\t"

1064 "paddw %%mm7, %%mm0 \n\t"

1065

1066 "packssdw %%mm4, %%mm4 \n\t"

1067 "psubusw %%mm7, %%mm0 \n\t"

1068

1069 "movd %%mm4, %%"REG_a" \n\t"

1070 "or %%"REG_a", %%"REG_a" \n\t"

1071 "jnz 2f \n\t"

1072 //movq [edi+"DCTSIZE_S"*3*2], mm1

1073 //movq [edi+"DCTSIZE_S"*5*2], mm5

1074 //movq [edi+"DCTSIZE_S"*1*2], mm0

1075 //movq [edi+"DCTSIZE_S"*7*2], mm6

1076 // t4 t5 - - - t6 t7 -

1077 //--- t4 (mm0) may be <>0; mm1, mm5, mm6 == 0

1078 //Typical numbers: nondc - 19%%, dc - 26%%, zero - 55%%. zero case alone isn't worthwhile

1079 "movq 0*8+%3, %%mm4 \n\t"

1080 "movq %%mm0, %%mm1 \n\t"

1081

1082 "pmulhw "MANGLE(MM_FIX_0_847759065)", %%mm0 \n\t" //tmp6

1083 "movq %%mm1, %%mm2 \n\t"

1084

1085 "movq "DCTSIZE_S "*0*2(%%"REG_D"), %%mm5 \n\t"

1086 "movq %%mm2, %%mm3 \n\t"

1087

1088 "pmulhw "MANGLE(MM_FIX_0_566454497)", %%mm1 \n\t" //tmp5

1089 "paddw %%mm4, %%mm5 \n\t"

1090

1091 "movq 1*8+%3, %%mm6 \n\t"

1092 //paddw mm3, MM_2

1093 "psraw 2,ドル %%mm3 \n\t" //tmp7

1094

1095 "pmulhw "MANGLE(MM_FIX_0_198912367)", %%mm2 \n\t" //-tmp4

1096 "psubw %%mm3, %%mm4 \n\t"

1097

1098 "movq "DCTSIZE_S "*1*2(%%"REG_D"), %%mm7 \n\t"

1099 "paddw %%mm3, %%mm5 \n\t"

1100

1101 "movq %%mm4, "DCTSIZE_S "*7*2(%%"REG_D") \n\t"

1102 "paddw %%mm6, %%mm7 \n\t"

1103

1104 "movq 2*8+%3, %%mm3 \n\t"

1105 "psubw %%mm0, %%mm6 \n\t"

1106

1107 "movq "DCTSIZE_S "*2*2(%%"REG_D"), %%mm4 \n\t"

1108 "paddw %%mm0, %%mm7 \n\t"

1109

1110 "movq %%mm5, "DCTSIZE_S "*0*2(%%"REG_D") \n\t"

1111 "paddw %%mm3, %%mm4 \n\t"

1112

1113 "movq %%mm6, "DCTSIZE_S "*6*2(%%"REG_D") \n\t"

1114 "psubw %%mm1, %%mm3 \n\t"

1115

1116 "movq "DCTSIZE_S "*5*2(%%"REG_D"), %%mm5 \n\t"

1117 "paddw %%mm1, %%mm4 \n\t"

1118

1119 "movq "DCTSIZE_S "*3*2(%%"REG_D"), %%mm6 \n\t"

1120 "paddw %%mm3, %%mm5 \n\t"

1121

1122 "movq 3*8+%3, %%mm0 \n\t"

1123 "add 8,ドル %%"REG_S" \n\t"

1124

1125 "movq %%mm7, "DCTSIZE_S "*1*2(%%"REG_D") \n\t"

1126 "paddw %%mm0, %%mm6 \n\t"

1127

1128 "movq %%mm4, "DCTSIZE_S "*2*2(%%"REG_D") \n\t"

1129 "psubw %%mm2, %%mm0 \n\t"

1130

1131 "movq "DCTSIZE_S "*4*2(%%"REG_D"), %%mm7 \n\t"

1132 "paddw %%mm2, %%mm6 \n\t"

1133

1134 "movq %%mm5, "DCTSIZE_S "*5*2(%%"REG_D") \n\t"

1135 "paddw %%mm0, %%mm7 \n\t"

1136

1137 "movq %%mm6, "DCTSIZE_S "*3*2(%%"REG_D") \n\t"

1138

1139 "movq %%mm7, "DCTSIZE_S "*4*2(%%"REG_D") \n\t"

1140 "add 8,ドル %%"REG_D" \n\t"

1141 "jmp 4f \n\t"

1142

1143 "2: \n\t"

1144 //--- non DC2

1145 //psraw mm1, 2 w/o it -> offset. thr1, thr1, thr1 (actually thr1, thr1, thr1-1)

1146 //psraw mm5, 2

1147 //psraw mm0, 2

1148 //psraw mm6, 2

1149 "movq %%mm5, %%mm3 \n\t"

1150 "psubw %%mm1, %%mm5 \n\t"

1151

1152 "psllw 1,ドル %%mm5 \n\t" //'z10

1153 "paddw %%mm1, %%mm3 \n\t" //'z13

1154

1155 "movq %%mm0, %%mm2 \n\t"

1156 "psubw %%mm6, %%mm0 \n\t"

1157

1158 "movq %%mm5, %%mm1 \n\t"

1159 "psllw 1,ドル %%mm0 \n\t" //'z12

1160

1161 "pmulhw "MANGLE(MM_FIX_2_613125930)", %%mm1 \n\t" //-

1162 "paddw %%mm0, %%mm5 \n\t"

1163

1164 "pmulhw "MANGLE(MM_FIX_1_847759065)", %%mm5 \n\t" //'z5

1165 "paddw %%mm6, %%mm2 \n\t" //'z11

1166

1167 "pmulhw "MANGLE(MM_FIX_1_082392200)", %%mm0 \n\t"

1168 "movq %%mm2, %%mm7 \n\t"

1169

1170 //---

1171 "movq 0*8+%3, %%mm4 \n\t"

1172 "psubw %%mm3, %%mm2 \n\t"

1173

1174 "psllw 1,ドル %%mm2 \n\t"

1175 "paddw %%mm3, %%mm7 \n\t" //'t7

1176

1177 "pmulhw "MANGLE(MM_FIX_1_414213562)", %%mm2 \n\t" //'t11

1178 "movq %%mm4, %%mm6 \n\t"

1179 //paddw mm7, MM_2

1180 "psraw 2,ドル %%mm7 \n\t"

1181

1182 "paddw "DCTSIZE_S "*0*2(%%"REG_D"), %%mm4 \n\t"

1183 "psubw %%mm7, %%mm6 \n\t"

1184

1185 "movq 1*8+%3, %%mm3 \n\t"

1186 "paddw %%mm7, %%mm4 \n\t"

1187

1188 "movq %%mm6, "DCTSIZE_S "*7*2(%%"REG_D") \n\t"

1189 "paddw %%mm5, %%mm1 \n\t" //'t12

1190

1191 "movq %%mm4, "DCTSIZE_S "*0*2(%%"REG_D") \n\t"

1192 "psubw %%mm7, %%mm1 \n\t" //'t6

1193

1194 "movq 2*8+%3, %%mm7 \n\t"

1195 "psubw %%mm5, %%mm0 \n\t" //'t10

1196

1197 "movq 3*8+%3, %%mm6 \n\t"

1198 "movq %%mm3, %%mm5 \n\t"

1199

1200 "paddw "DCTSIZE_S "*1*2(%%"REG_D"), %%mm3 \n\t"

1201 "psubw %%mm1, %%mm5 \n\t"

1202

1203 "psubw %%mm1, %%mm2 \n\t" //'t5

1204 "paddw %%mm1, %%mm3 \n\t"

1205

1206 "movq %%mm5, "DCTSIZE_S "*6*2(%%"REG_D") \n\t"

1207 "movq %%mm7, %%mm4 \n\t"

1208

1209 "paddw "DCTSIZE_S "*2*2(%%"REG_D"), %%mm7 \n\t"

1210 "psubw %%mm2, %%mm4 \n\t"

1211

1212 "paddw "DCTSIZE_S "*5*2(%%"REG_D"), %%mm4 \n\t"

1213 "paddw %%mm2, %%mm7 \n\t"

1214

1215 "movq %%mm3, "DCTSIZE_S "*1*2(%%"REG_D") \n\t"

1216 "paddw %%mm2, %%mm0 \n\t" //'t4

1217

1218 // 't4 't6 't5 - - - - 't7

1219 "movq %%mm7, "DCTSIZE_S "*2*2(%%"REG_D") \n\t"

1220 "movq %%mm6, %%mm1 \n\t"

1221

1222 "paddw "DCTSIZE_S "*4*2(%%"REG_D"), %%mm6 \n\t"

1223 "psubw %%mm0, %%mm1 \n\t"

1224

1225 "paddw "DCTSIZE_S "*3*2(%%"REG_D"), %%mm1 \n\t"

1226 "paddw %%mm0, %%mm6 \n\t"

1227

1228 "movq %%mm4, "DCTSIZE_S "*5*2(%%"REG_D") \n\t"

1229 "add 8,ドル %%"REG_S" \n\t"

1230

1231 "movq %%mm6, "DCTSIZE_S "*4*2(%%"REG_D") \n\t"

1232

1233 "movq %%mm1, "DCTSIZE_S "*3*2(%%"REG_D") \n\t"

1234 "add 8,ドル %%"REG_D" \n\t"

1235

1236 "4: \n\t"

1237 //=part 2 (the same)===========================================================

1238 "movq "DCTSIZE_S "*0*2(%%"REG_S"), %%mm1 \n\t"

1239 //

1240 "movq "DCTSIZE_S "*3*2(%%"REG_S"), %%mm7 \n\t"

1241 "movq %%mm1, %%mm0 \n\t"

1242

1243 "paddw "DCTSIZE_S "*7*2(%%"REG_S"), %%mm1 \n\t" //t0

1244 "movq %%mm7, %%mm3 \n\t"

1245

1246 "paddw "DCTSIZE_S "*4*2(%%"REG_S"), %%mm7 \n\t" //t3

1247 "movq %%mm1, %%mm5 \n\t"

1248

1249 "movq "DCTSIZE_S "*1*2(%%"REG_S"), %%mm6 \n\t"

1250 "psubw %%mm7, %%mm1 \n\t" //t13

1251

1252 "movq "DCTSIZE_S "*2*2(%%"REG_S"), %%mm2 \n\t"

1253 "movq %%mm6, %%mm4 \n\t"

1254

1255 "paddw "DCTSIZE_S "*6*2(%%"REG_S"), %%mm6 \n\t" //t1

1256 "paddw %%mm7, %%mm5 \n\t" //t10

1257

1258 "paddw "DCTSIZE_S "*5*2(%%"REG_S"), %%mm2 \n\t" //t2

1259 "movq %%mm6, %%mm7 \n\t"

1260

1261 "paddw %%mm2, %%mm6 \n\t" //t11

1262 "psubw %%mm2, %%mm7 \n\t" //t12

1263

1264 "movq %%mm5, %%mm2 \n\t"

1265 "paddw %%mm6, %%mm5 \n\t" //d0

1266 // i0 t13 t12 i3 i1 d0 - d4

1267 "psubw %%mm6, %%mm2 \n\t" //d4

1268 "paddw %%mm1, %%mm7 \n\t"

1269

1270 "movq 1*8+4*16(%%"REG_d"), %%mm6 \n\t"

1271 "psllw 2,ドル %%mm7 \n\t"

1272

1273 "psubw 1*8+0*16(%%"REG_d"), %%mm5 \n\t"

1274 "psubw %%mm6, %%mm2 \n\t"

1275

1276 "paddusw 1*8+0*16(%%"REG_d"), %%mm5 \n\t"

1277 "paddusw %%mm6, %%mm2 \n\t"

1278

1279 "pmulhw "MANGLE(ff_MM_FIX_0_707106781)", %%mm7 \n\t"

1280 //

1281 "paddw 1*8+0*16(%%"REG_d"), %%mm5 \n\t"

1282 "paddw %%mm6, %%mm2 \n\t"

1283

1284 "psubusw 1*8+0*16(%%"REG_d"), %%mm5 \n\t"

1285 "psubusw %%mm6, %%mm2 \n\t"

1286

1287 //This func is totally compute-bound, operates at huge speed. So, DC shortcut

1288 // at this place isn't worthwhile due to BTB miss penalty (checked on Pent. 3).

1289 //However, typical numbers: nondc - 29%%, dc - 46%%, zero - 25%%. All <> 0 case is very rare.

1290 "paddw "MANGLE(MM_2)", %%mm5 \n\t"

1291 "movq %%mm2, %%mm6 \n\t"

1292

1293 "paddw %%mm5, %%mm2 \n\t"

1294 "psubw %%mm6, %%mm5 \n\t"

1295

1296 "movq %%mm1, %%mm6 \n\t"

1297 "paddw %%mm7, %%mm1 \n\t" //d2

1298

1299 "psubw 1*8+2*16(%%"REG_d"), %%mm1 \n\t"

1300 "psubw %%mm7, %%mm6 \n\t" //d6

1301

1302 "movq 1*8+6*16(%%"REG_d"), %%mm7 \n\t"

1303 "psraw 2,ドル %%mm5 \n\t"

1304

1305 "paddusw 1*8+2*16(%%"REG_d"), %%mm1 \n\t"

1306 "psubw %%mm7, %%mm6 \n\t"

1307 // t7 d2 /t11 t4 t6 - d6 /t10

1308

1309 "paddw 1*8+2*16(%%"REG_d"), %%mm1 \n\t"

1310 "paddusw %%mm7, %%mm6 \n\t"

1311

1312 "psubusw 1*8+2*16(%%"REG_d"), %%mm1 \n\t"

1313 "paddw %%mm7, %%mm6 \n\t"

1314

1315 "psubw "DCTSIZE_S "*4*2(%%"REG_S"), %%mm3 \n\t"

1316 "psubusw %%mm7, %%mm6 \n\t"

1317

1318 //movq [edi+"DCTSIZE_S"*2*2], mm1

1319 //movq [edi+"DCTSIZE_S"*6*2], mm6

1320 "movq %%mm1, %%mm7 \n\t"

1321 "psraw 2,ドル %%mm2 \n\t"

1322

1323 "psubw "DCTSIZE_S "*6*2(%%"REG_S"), %%mm4 \n\t"

1324 "psubw %%mm6, %%mm1 \n\t"

1325

1326 "psubw "DCTSIZE_S "*7*2(%%"REG_S"), %%mm0 \n\t"

1327 "paddw %%mm7, %%mm6 \n\t" //'t13

1328

1329 "psraw 2,ドル %%mm6 \n\t" //paddw mm6, MM_2 !! ---

1330 "movq %%mm2, %%mm7 \n\t"

1331

1332 "pmulhw "MANGLE(MM_FIX_1_414213562_A)", %%mm1 \n\t"

1333 "paddw %%mm6, %%mm2 \n\t" //'t0

1334

1335 "movq %%mm2, 0*8+%3 \n\t" //!

1336 "psubw %%mm6, %%mm7 \n\t" //'t3

1337

1338 "movq "DCTSIZE_S "*2*2(%%"REG_S"), %%mm2 \n\t"

1339 "psubw %%mm6, %%mm1 \n\t" //'t12

1340

1341 "psubw "DCTSIZE_S "*5*2(%%"REG_S"), %%mm2 \n\t" //t5

1342 "movq %%mm5, %%mm6 \n\t"

1343

1344 "movq %%mm7, 3*8+%3 \n\t"

1345 "paddw %%mm2, %%mm3 \n\t" //t10

1346

1347 "paddw %%mm4, %%mm2 \n\t" //t11

1348 "paddw %%mm0, %%mm4 \n\t" //t12

1349

1350 "movq %%mm3, %%mm7 \n\t"

1351 "psubw %%mm4, %%mm3 \n\t"

1352

1353 "psllw 2,ドル %%mm3 \n\t"

1354 "psllw 2,ドル %%mm7 \n\t" //opt for P6

1355

1356 "pmulhw "MANGLE(MM_FIX_0_382683433)", %%mm3 \n\t"

1357 "psllw 2,ドル %%mm4 \n\t"

1358

1359 "pmulhw "MANGLE(ff_MM_FIX_0_541196100)", %%mm7 \n\t"

1360 "psllw 2,ドル %%mm2 \n\t"

1361

1362 "pmulhw "MANGLE(MM_FIX_1_306562965)", %%mm4 \n\t"

1363 "paddw %%mm1, %%mm5 \n\t" //'t1

1364

1365 "pmulhw "MANGLE(ff_MM_FIX_0_707106781)", %%mm2 \n\t"

1366 "psubw %%mm1, %%mm6 \n\t" //'t2

1367 // t7 't12 't11 t4 t6 - 't13 't10 ---

1368

1369 "paddw %%mm3, %%mm7 \n\t" //z2

1370

1371 "movq %%mm5, 1*8+%3 \n\t"

1372 "paddw %%mm3, %%mm4 \n\t" //z4

1373

1374 "movq 1*8+3*16(%%"REG_d"), %%mm3 \n\t"

1375 "movq %%mm0, %%mm1 \n\t"

1376

1377 "movq %%mm6, 2*8+%3 \n\t"

1378 "psubw %%mm2, %%mm1 \n\t" //z13

1379

1380 //===

1381 "paddw %%mm2, %%mm0 \n\t" //z11

1382 "movq %%mm1, %%mm5 \n\t"

1383

1384 "movq 1*8+5*16(%%"REG_d"), %%mm2 \n\t"

1385 "psubw %%mm7, %%mm1 \n\t" //d3

1386

1387 "paddw %%mm7, %%mm5 \n\t" //d5

1388 "psubw %%mm3, %%mm1 \n\t"

1389

1390 "movq 1*8+1*16(%%"REG_d"), %%mm7 \n\t"

1391 "psubw %%mm2, %%mm5 \n\t"

1392

1393 "movq %%mm0, %%mm6 \n\t"

1394 "paddw %%mm4, %%mm0 \n\t" //d1

1395

1396 "paddusw %%mm3, %%mm1 \n\t"

1397 "psubw %%mm4, %%mm6 \n\t" //d7

1398

1399 // d1 d3 - - - d5 d7 -

1400 "movq 1*8+7*16(%%"REG_d"), %%mm4 \n\t"

1401 "psubw %%mm7, %%mm0 \n\t"

1402

1403 "psubw %%mm4, %%mm6 \n\t"

1404 "paddusw %%mm2, %%mm5 \n\t"

1405

1406 "paddusw %%mm4, %%mm6 \n\t"

1407 "paddw %%mm3, %%mm1 \n\t"

1408

1409 "paddw %%mm2, %%mm5 \n\t"

1410 "paddw %%mm4, %%mm6 \n\t"

1411

1412 "psubusw %%mm3, %%mm1 \n\t"

1413 "psubusw %%mm2, %%mm5 \n\t"

1414

1415 "psubusw %%mm4, %%mm6 \n\t"

1416 "movq %%mm1, %%mm4 \n\t"

1417

1418 "por %%mm5, %%mm4 \n\t"

1419 "paddusw %%mm7, %%mm0 \n\t"

1420

1421 "por %%mm6, %%mm4 \n\t"

1422 "paddw %%mm7, %%mm0 \n\t"

1423

1424 "packssdw %%mm4, %%mm4 \n\t"

1425 "psubusw %%mm7, %%mm0 \n\t"

1426

1427 "movd %%mm4, %%"REG_a" \n\t"

1428 "or %%"REG_a", %%"REG_a" \n\t"

1429 "jnz 3f \n\t"

1430 //movq [edi+"DCTSIZE_S"*3*2], mm1

1431 //movq [edi+"DCTSIZE_S"*5*2], mm5

1432 //movq [edi+"DCTSIZE_S"*1*2], mm0

1433 //movq [edi+"DCTSIZE_S"*7*2], mm6

1434 // t4 t5 - - - t6 t7 -

1435 //--- t4 (mm0) may be <>0; mm1, mm5, mm6 == 0

1436 //Typical numbers: nondc - 19%%, dc - 26%%, zero - 55%%. zero case alone isn't worthwhile

1437 "movq 0*8+%3, %%mm4 \n\t"

1438 "movq %%mm0, %%mm1 \n\t"

1439

1440 "pmulhw "MANGLE(MM_FIX_0_847759065)", %%mm0 \n\t" //tmp6

1441 "movq %%mm1, %%mm2 \n\t"

1442

1443 "movq "DCTSIZE_S "*0*2(%%"REG_D"), %%mm5 \n\t"

1444 "movq %%mm2, %%mm3 \n\t"

1445

1446 "pmulhw "MANGLE(MM_FIX_0_566454497)", %%mm1 \n\t" //tmp5

1447 "paddw %%mm4, %%mm5 \n\t"

1448

1449 "movq 1*8+%3, %%mm6 \n\t"

1450 //paddw mm3, MM_2

1451 "psraw 2,ドル %%mm3 \n\t" //tmp7

1452

1453 "pmulhw "MANGLE(MM_FIX_0_198912367)", %%mm2 \n\t" //-tmp4

1454 "psubw %%mm3, %%mm4 \n\t"

1455

1456 "movq "DCTSIZE_S "*1*2(%%"REG_D"), %%mm7 \n\t"

1457 "paddw %%mm3, %%mm5 \n\t"

1458

1459 "movq %%mm4, "DCTSIZE_S "*7*2(%%"REG_D") \n\t"

1460 "paddw %%mm6, %%mm7 \n\t"

1461

1462 "movq 2*8+%3, %%mm3 \n\t"

1463 "psubw %%mm0, %%mm6 \n\t"

1464

1465 "movq "DCTSIZE_S "*2*2(%%"REG_D"), %%mm4 \n\t"

1466 "paddw %%mm0, %%mm7 \n\t"

1467

1468 "movq %%mm5, "DCTSIZE_S "*0*2(%%"REG_D") \n\t"

1469 "paddw %%mm3, %%mm4 \n\t"

1470

1471 "movq %%mm6, "DCTSIZE_S "*6*2(%%"REG_D") \n\t"

1472 "psubw %%mm1, %%mm3 \n\t"

1473

1474 "movq "DCTSIZE_S "*5*2(%%"REG_D"), %%mm5 \n\t"

1475 "paddw %%mm1, %%mm4 \n\t"

1476

1477 "movq "DCTSIZE_S "*3*2(%%"REG_D"), %%mm6 \n\t"

1478 "paddw %%mm3, %%mm5 \n\t"

1479

1480 "movq 3*8+%3, %%mm0 \n\t"

1481 "add 24,ドル %%"REG_S" \n\t"

1482

1483 "movq %%mm7, "DCTSIZE_S "*1*2(%%"REG_D") \n\t"

1484 "paddw %%mm0, %%mm6 \n\t"

1485

1486 "movq %%mm4, "DCTSIZE_S "*2*2(%%"REG_D") \n\t"

1487 "psubw %%mm2, %%mm0 \n\t"

1488

1489 "movq "DCTSIZE_S "*4*2(%%"REG_D"), %%mm7 \n\t"

1490 "paddw %%mm2, %%mm6 \n\t"

1491

1492 "movq %%mm5, "DCTSIZE_S "*5*2(%%"REG_D") \n\t"

1493 "paddw %%mm0, %%mm7 \n\t"

1494

1495 "movq %%mm6, "DCTSIZE_S "*3*2(%%"REG_D") \n\t"

1496

1497 "movq %%mm7, "DCTSIZE_S "*4*2(%%"REG_D") \n\t"

1498 "add 24,ドル %%"REG_D" \n\t"

1499 "sub 2,ドル %%"REG_c" \n\t"

1500 "jnz 1b \n\t"

1501 "jmp 5f \n\t"

1502

1503 "3: \n\t"

1504 //--- non DC2

1505 //psraw mm1, 2 w/o it -> offset. thr1, thr1, thr1 (actually thr1, thr1, thr1-1)

1506 //psraw mm5, 2

1507 //psraw mm0, 2

1508 //psraw mm6, 2

1509 "movq %%mm5, %%mm3 \n\t"

1510 "psubw %%mm1, %%mm5 \n\t"

1511

1512 "psllw 1,ドル %%mm5 \n\t" //'z10

1513 "paddw %%mm1, %%mm3 \n\t" //'z13

1514

1515 "movq %%mm0, %%mm2 \n\t"

1516 "psubw %%mm6, %%mm0 \n\t"

1517

1518 "movq %%mm5, %%mm1 \n\t"

1519 "psllw 1,ドル %%mm0 \n\t" //'z12

1520

1521 "pmulhw "MANGLE(MM_FIX_2_613125930)", %%mm1 \n\t" //-

1522 "paddw %%mm0, %%mm5 \n\t"

1523

1524 "pmulhw "MANGLE(MM_FIX_1_847759065)", %%mm5 \n\t" //'z5

1525 "paddw %%mm6, %%mm2 \n\t" //'z11

1526

1527 "pmulhw "MANGLE(MM_FIX_1_082392200)", %%mm0 \n\t"

1528 "movq %%mm2, %%mm7 \n\t"

1529

1530 //---

1531 "movq 0*8+%3, %%mm4 \n\t"

1532 "psubw %%mm3, %%mm2 \n\t"

1533

1534 "psllw 1,ドル %%mm2 \n\t"

1535 "paddw %%mm3, %%mm7 \n\t" //'t7

1536

1537 "pmulhw "MANGLE(MM_FIX_1_414213562)", %%mm2 \n\t" //'t11

1538 "movq %%mm4, %%mm6 \n\t"

1539 //paddw mm7, MM_2

1540 "psraw 2,ドル %%mm7 \n\t"

1541

1542 "paddw "DCTSIZE_S "*0*2(%%"REG_D"), %%mm4 \n\t"

1543 "psubw %%mm7, %%mm6 \n\t"

1544

1545 "movq 1*8+%3, %%mm3 \n\t"

1546 "paddw %%mm7, %%mm4 \n\t"

1547

1548 "movq %%mm6, "DCTSIZE_S "*7*2(%%"REG_D") \n\t"

1549 "paddw %%mm5, %%mm1 \n\t" //'t12

1550

1551 "movq %%mm4, "DCTSIZE_S "*0*2(%%"REG_D") \n\t"

1552 "psubw %%mm7, %%mm1 \n\t" //'t6

1553

1554 "movq 2*8+%3, %%mm7 \n\t"

1555 "psubw %%mm5, %%mm0 \n\t" //'t10

1556

1557 "movq 3*8+%3, %%mm6 \n\t"

1558 "movq %%mm3, %%mm5 \n\t"

1559

1560 "paddw "DCTSIZE_S "*1*2(%%"REG_D"), %%mm3 \n\t"

1561 "psubw %%mm1, %%mm5 \n\t"

1562

1563 "psubw %%mm1, %%mm2 \n\t" //'t5

1564 "paddw %%mm1, %%mm3 \n\t"

1565

1566 "movq %%mm5, "DCTSIZE_S "*6*2(%%"REG_D") \n\t"

1567 "movq %%mm7, %%mm4 \n\t"

1568

1569 "paddw "DCTSIZE_S "*2*2(%%"REG_D"), %%mm7 \n\t"

1570 "psubw %%mm2, %%mm4 \n\t"

1571

1572 "paddw "DCTSIZE_S "*5*2(%%"REG_D"), %%mm4 \n\t"

1573 "paddw %%mm2, %%mm7 \n\t"

1574

1575 "movq %%mm3, "DCTSIZE_S "*1*2(%%"REG_D") \n\t"

1576 "paddw %%mm2, %%mm0 \n\t" //'t4

1577

1578 // 't4 't6 't5 - - - - 't7

1579 "movq %%mm7, "DCTSIZE_S "*2*2(%%"REG_D") \n\t"

1580 "movq %%mm6, %%mm1 \n\t"

1581

1582 "paddw "DCTSIZE_S "*4*2(%%"REG_D"), %%mm6 \n\t"

1583 "psubw %%mm0, %%mm1 \n\t"

1584

1585 "paddw "DCTSIZE_S "*3*2(%%"REG_D"), %%mm1 \n\t"

1586 "paddw %%mm0, %%mm6 \n\t"

1587

1588 "movq %%mm4, "DCTSIZE_S "*5*2(%%"REG_D") \n\t"

1589 "add 24,ドル %%"REG_S" \n\t"

1590

1591 "movq %%mm6, "DCTSIZE_S "*4*2(%%"REG_D") \n\t"

1592

1593 "movq %%mm1, "DCTSIZE_S "*3*2(%%"REG_D") \n\t"

1594 "add 24,ドル %%"REG_D" \n\t"

1595 "sub 2,ドル %%"REG_c" \n\t"

1596 "jnz 1b \n\t"

1597 "5: \n\t"

1598

1599 : "+S"(data), "+D"(output), "+c"(cnt), "=o"(temps)

1600 : "d"(thr_adr)

1601 : "%"REG_a

1602 );

1603 }

1604

1605 #endif // HAVE_MMX

1606

1607 #if !HAVE_MMX

1608

1609 static void row_idct_c(int16_t* workspace,

1610 int16_t* output_adr, int output_stride, int cnt)

1611 {

1612 int_simd16_t tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;

1613 int_simd16_t tmp10, tmp11, tmp12, tmp13;

1614 int_simd16_t z5, z10, z11, z12, z13;

1615 int16_t* outptr;

1616 int16_t* wsptr;

1617

1618 cnt*=4;

1619 wsptr = workspace;

1620 outptr = output_adr;

1621 for (; cnt > 0; cnt--) {

1622 // Even part

1623 //Simd version reads 4x4 block and transposes it

1624 tmp10 = ( wsptr[2] + wsptr[3]);

1625 tmp11 = ( wsptr[2] - wsptr[3]);

1626

1627 tmp13 = ( wsptr[0] + wsptr[1]);

1628 tmp12 = (MULTIPLY16H( wsptr[0] - wsptr[1], FIX_1_414213562_A)<<2) - tmp13;//this shift order to avoid overflow

1629

1630 tmp0 = tmp10 + tmp13; //->temps

1631 tmp3 = tmp10 - tmp13; //->temps

1632 tmp1 = tmp11 + tmp12;

1633 tmp2 = tmp11 - tmp12;

1634

1635 // Odd part

1636 //Also transpose, with previous:

1637 // ---- ---- ||||

1638 // ---- ---- idct ||||

1639 // ---- ---- ---> ||||

1640 // ---- ---- ||||

1641 z13 = wsptr[4] + wsptr[5];

1642 z10 = wsptr[4] - wsptr[5];

1643 z11 = wsptr[6] + wsptr[7];

1644 z12 = wsptr[6] - wsptr[7];

1645

1646 tmp7 = z11 + z13;

1647 tmp11 = MULTIPLY16H(z11 - z13, FIX_1_414213562);

1648

1649 z5 = MULTIPLY16H(z10 + z12, FIX_1_847759065);

1650 tmp10 = MULTIPLY16H(z12, FIX_1_082392200) - z5;

1651 tmp12 = MULTIPLY16H(z10, FIX_2_613125930) + z5; // - FIX_

1652

1653 tmp6 = (tmp12<<3) - tmp7;

1654 tmp5 = (tmp11<<3) - tmp6;

1655 tmp4 = (tmp10<<3) + tmp5;

1656

1657 // Final output stage: descale and write column

1658 outptr[0*output_stride]+= DESCALE(tmp0 + tmp7, 3);

1659 outptr[1*output_stride]+= DESCALE(tmp1 + tmp6, 3);

1660 outptr[2*output_stride]+= DESCALE(tmp2 + tmp5, 3);

1661 outptr[3*output_stride]+= DESCALE(tmp3 - tmp4, 3);

1662 outptr[4*output_stride]+= DESCALE(tmp3 + tmp4, 3);

1663 outptr[5*output_stride]+= DESCALE(tmp2 - tmp5, 3);

1664 outptr[6*output_stride]+= DESCALE(tmp1 - tmp6, 3); //no += ?

1665 outptr[7*output_stride]+= DESCALE(tmp0 - tmp7, 3); //no += ?

1666 outptr++;

1667

1668 wsptr += DCTSIZE; // advance pointer to next row

1669 }

1670 }

1671

1672 #else /* HAVE_MMX */

1673

1674 static void row_idct_mmx (int16_t* workspace,

1675 int16_t* output_adr, int output_stride, int cnt)

1676 {

1677 uint64_t __attribute__((aligned(8))) temps[4];

1678 __asm__ volatile(

1679 "lea (%%"REG_a",%%"REG_a",2), %%"REG_d" \n\t"

1680 "1: \n\t"

1681 "movq "DCTSIZE_S "*0*2(%%"REG_S"), %%mm0 \n\t"

1682 //

1683

1684 "movq "DCTSIZE_S "*1*2(%%"REG_S"), %%mm1 \n\t"

1685 "movq %%mm0, %%mm4 \n\t"

1686

1687 "movq "DCTSIZE_S "*2*2(%%"REG_S"), %%mm2 \n\t"

1688 "punpcklwd %%mm1, %%mm0 \n\t"

1689

1690 "movq "DCTSIZE_S "*3*2(%%"REG_S"), %%mm3 \n\t"

1691 "punpckhwd %%mm1, %%mm4 \n\t"

1692

1693 //transpose 4x4

1694 "movq %%mm2, %%mm7 \n\t"

1695 "punpcklwd %%mm3, %%mm2 \n\t"

1696

1697 "movq %%mm0, %%mm6 \n\t"

1698 "punpckldq %%mm2, %%mm0 \n\t" //0

1699

1700 "punpckhdq %%mm2, %%mm6 \n\t" //1

1701 "movq %%mm0, %%mm5 \n\t"

1702

1703 "punpckhwd %%mm3, %%mm7 \n\t"

1704 "psubw %%mm6, %%mm0 \n\t"

1705

1706 "pmulhw "MANGLE(MM_FIX_1_414213562_A)", %%mm0 \n\t"

1707 "movq %%mm4, %%mm2 \n\t"

1708

1709 "punpckldq %%mm7, %%mm4 \n\t" //2

1710 "paddw %%mm6, %%mm5 \n\t"

1711

1712 "punpckhdq %%mm7, %%mm2 \n\t" //3

1713 "movq %%mm4, %%mm1 \n\t"

1714

1715 "psllw 2,ドル %%mm0 \n\t"

1716 "paddw %%mm2, %%mm4 \n\t" //t10

1717

1718 "movq "DCTSIZE_S "*0*2+"DCTSIZE_S "(%%"REG_S"), %%mm3 \n\t"

1719 "psubw %%mm2, %%mm1 \n\t" //t11

1720

1721 "movq "DCTSIZE_S "*1*2+"DCTSIZE_S "(%%"REG_S"), %%mm2 \n\t"

1722 "psubw %%mm5, %%mm0 \n\t"

1723

1724 "movq %%mm4, %%mm6 \n\t"

1725 "paddw %%mm5, %%mm4 \n\t" //t0

1726

1727 "psubw %%mm5, %%mm6 \n\t" //t3

1728 "movq %%mm1, %%mm7 \n\t"

1729

1730 "movq "DCTSIZE_S "*2*2+"DCTSIZE_S "(%%"REG_S"), %%mm5 \n\t"

1731 "paddw %%mm0, %%mm1 \n\t" //t1

1732

1733 "movq %%mm4, 0*8+%3 \n\t" //t0

1734 "movq %%mm3, %%mm4 \n\t"

1735

1736 "movq %%mm6, 1*8+%3 \n\t" //t3

1737 "punpcklwd %%mm2, %%mm3 \n\t"

1738

1739 //transpose 4x4

1740 "movq "DCTSIZE_S "*3*2+"DCTSIZE_S "(%%"REG_S"), %%mm6 \n\t"

1741 "punpckhwd %%mm2, %%mm4 \n\t"

1742

1743 "movq %%mm5, %%mm2 \n\t"

1744 "punpcklwd %%mm6, %%mm5 \n\t"

1745

1746 "psubw %%mm0, %%mm7 \n\t" //t2

1747 "punpckhwd %%mm6, %%mm2 \n\t"

1748

1749 "movq %%mm3, %%mm0 \n\t"

1750 "punpckldq %%mm5, %%mm3 \n\t" //4

1751

1752 "punpckhdq %%mm5, %%mm0 \n\t" //5

1753 "movq %%mm4, %%mm5 \n\t"

1754

1755 //

1756 "movq %%mm3, %%mm6 \n\t"

1757 "punpckldq %%mm2, %%mm4 \n\t" //6

1758

1759 "psubw %%mm0, %%mm3 \n\t" //z10

1760 "punpckhdq %%mm2, %%mm5 \n\t" //7

1761

1762 "paddw %%mm0, %%mm6 \n\t" //z13

1763 "movq %%mm4, %%mm2 \n\t"

1764

1765 "movq %%mm3, %%mm0 \n\t"

1766 "psubw %%mm5, %%mm4 \n\t" //z12

1767

1768 "pmulhw "MANGLE(MM_FIX_2_613125930)", %%mm0 \n\t" //-

1769 "paddw %%mm4, %%mm3 \n\t"

1770

1771 "pmulhw "MANGLE(MM_FIX_1_847759065)", %%mm3 \n\t" //z5

1772 "paddw %%mm5, %%mm2 \n\t" //z11 >

1773

1774 "pmulhw "MANGLE(MM_FIX_1_082392200)", %%mm4 \n\t"

1775 "movq %%mm2, %%mm5 \n\t"

1776

1777 "psubw %%mm6, %%mm2 \n\t"

1778 "paddw %%mm6, %%mm5 \n\t" //t7

1779

1780 "pmulhw "MANGLE(MM_FIX_1_414213562)", %%mm2 \n\t" //t11

1781 "paddw %%mm3, %%mm0 \n\t" //t12

1782

1783 "psllw 3,ドル %%mm0 \n\t"

1784 "psubw %%mm3, %%mm4 \n\t" //t10

1785

1786 "movq 0*8+%3, %%mm6 \n\t"

1787 "movq %%mm1, %%mm3 \n\t"

1788

1789 "psllw 3,ドル %%mm4 \n\t"

1790 "psubw %%mm5, %%mm0 \n\t" //t6

1791

1792 "psllw 3,ドル %%mm2 \n\t"

1793 "paddw %%mm0, %%mm1 \n\t" //d1

1794

1795 "psubw %%mm0, %%mm2 \n\t" //t5

1796 "psubw %%mm0, %%mm3 \n\t" //d6

1797

1798 "paddw %%mm2, %%mm4 \n\t" //t4

1799 "movq %%mm7, %%mm0 \n\t"

1800

1801 "paddw %%mm2, %%mm7 \n\t" //d2

1802 "psubw %%mm2, %%mm0 \n\t" //d5

1803

1804 "movq "MANGLE(MM_DESCALE_RND)", %%mm2 \n\t" //4

1805 "psubw %%mm5, %%mm6 \n\t" //d7

1806

1807 "paddw 0*8+%3, %%mm5 \n\t" //d0

1808 "paddw %%mm2, %%mm1 \n\t"

1809

1810 "paddw %%mm2, %%mm5 \n\t"

1811 "psraw 3,ドル %%mm1 \n\t"

1812

1813 "paddw %%mm2, %%mm7 \n\t"

1814 "psraw 3,ドル %%mm5 \n\t"

1815

1816 "paddw (%%"REG_D"), %%mm5 \n\t"

1817 "psraw 3,ドル %%mm7 \n\t"

1818

1819 "paddw (%%"REG_D",%%"REG_a",), %%mm1 \n\t"

1820 "paddw %%mm2, %%mm0 \n\t"

1821

1822 "paddw (%%"REG_D",%%"REG_a",2), %%mm7 \n\t"

1823 "paddw %%mm2, %%mm3 \n\t"

1824

1825 "movq %%mm5, (%%"REG_D") \n\t"

1826 "paddw %%mm2, %%mm6 \n\t"

1827

1828 "movq %%mm1, (%%"REG_D",%%"REG_a",) \n\t"

1829 "psraw 3,ドル %%mm0 \n\t"

1830

1831 "movq %%mm7, (%%"REG_D",%%"REG_a",2) \n\t"

1832 "add %%"REG_d", %%"REG_D" \n\t" //3*ls

1833

1834 "movq 1*8+%3, %%mm5 \n\t" //t3

1835 "psraw 3,ドル %%mm3 \n\t"

1836

1837 "paddw (%%"REG_D",%%"REG_a",2), %%mm0 \n\t"

1838 "psubw %%mm4, %%mm5 \n\t" //d3

1839

1840 "paddw (%%"REG_D",%%"REG_d",), %%mm3 \n\t"

1841 "psraw 3,ドル %%mm6 \n\t"

1842

1843 "paddw 1*8+%3, %%mm4 \n\t" //d4

1844 "paddw %%mm2, %%mm5 \n\t"

1845

1846 "paddw (%%"REG_D",%%"REG_a",4), %%mm6 \n\t"

1847 "paddw %%mm2, %%mm4 \n\t"

1848

1849 "movq %%mm0, (%%"REG_D",%%"REG_a",2) \n\t"

1850 "psraw 3,ドル %%mm5 \n\t"

1851

1852 "paddw (%%"REG_D"), %%mm5 \n\t"

1853 "psraw 3,ドル %%mm4 \n\t"

1854

1855 "paddw (%%"REG_D",%%"REG_a",), %%mm4 \n\t"

1856 "add $"DCTSIZE_S "*2*4, %%"REG_S" \n\t" //4 rows

1857

1858 "movq %%mm3, (%%"REG_D",%%"REG_d",) \n\t"

1859 "movq %%mm6, (%%"REG_D",%%"REG_a",4) \n\t"

1860 "movq %%mm5, (%%"REG_D") \n\t"

1861 "movq %%mm4, (%%"REG_D",%%"REG_a",) \n\t"

1862

1863 "sub %%"REG_d", %%"REG_D" \n\t"

1864 "add 8,ドル %%"REG_D" \n\t"

1865 "dec %%"REG_c" \n\t"

1866 "jnz 1b \n\t"

1867

1868 : "+S"(workspace), "+D"(output_adr), "+c"(cnt), "=o"(temps)

1869 : "a"(output_stride*sizeof(short))

1870 : "%"REG_d

1871 );

1872 }

1873

1874 #endif // HAVE_MMX

1875

1876 #if !HAVE_MMX

1877

1878 static void row_fdct_c(int16_t *data, const uint8_t *pixels, int line_size, int cnt)

1879 {

1880 int_simd16_t tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;

1881 int_simd16_t tmp10, tmp11, tmp12, tmp13;

1882 int_simd16_t z1, z2, z3, z4, z5, z11, z13;

1883 int16_t *dataptr;

1884

1885 cnt*=4;

1886 // Pass 1: process rows.

1887

1888 dataptr = data;

1889 for (; cnt > 0; cnt--) {

1890 tmp0 = pixels[line_size*0] + pixels[line_size*7];

1891 tmp7 = pixels[line_size*0] - pixels[line_size*7];

1892 tmp1 = pixels[line_size*1] + pixels[line_size*6];

1893 tmp6 = pixels[line_size*1] - pixels[line_size*6];

1894 tmp2 = pixels[line_size*2] + pixels[line_size*5];

1895 tmp5 = pixels[line_size*2] - pixels[line_size*5];

1896 tmp3 = pixels[line_size*3] + pixels[line_size*4];

1897 tmp4 = pixels[line_size*3] - pixels[line_size*4];

1898

1899 // Even part

1900

1901 tmp10 = tmp0 + tmp3;

1902 tmp13 = tmp0 - tmp3;

1903 tmp11 = tmp1 + tmp2;

1904 tmp12 = tmp1 - tmp2;

1905 //Even columns are written first, this leads to different order of columns

1906 //in column_fidct(), but they are processed independently, so all ok.

1907 //Later in the row_idct() columns readed at the same order.

1908 dataptr[2] = tmp10 + tmp11;

1909 dataptr[3] = tmp10 - tmp11;

1910

1911 z1 = MULTIPLY16H((tmp12 + tmp13)<<2, FIX_0_707106781);

1912 dataptr[0] = tmp13 + z1;

1913 dataptr[1] = tmp13 - z1;

1914

1915 // Odd part

1916

1917 tmp10 = (tmp4 + tmp5) <<2;

1918 tmp11 = (tmp5 + tmp6) <<2;

1919 tmp12 = (tmp6 + tmp7) <<2;

1920

1921 z5 = MULTIPLY16H(tmp10 - tmp12, FIX_0_382683433);

1922 z2 = MULTIPLY16H(tmp10, FIX_0_541196100) + z5;

1923 z4 = MULTIPLY16H(tmp12, FIX_1_306562965) + z5;

1924 z3 = MULTIPLY16H(tmp11, FIX_0_707106781);

1925

1926 z11 = tmp7 + z3;

1927 z13 = tmp7 - z3;

1928

1929 dataptr[4] = z13 + z2;

1930 dataptr[5] = z13 - z2;

1931 dataptr[6] = z11 + z4;

1932 dataptr[7] = z11 - z4;

1933

1934 pixels++; // advance pointer to next column

1935 dataptr += DCTSIZE;

1936 }

1937 }

1938

1939 #else /* HAVE_MMX */

1940

1941 static void row_fdct_mmx(int16_t *data, const uint8_t *pixels, int line_size, int cnt)

1942 {

1943 uint64_t __attribute__((aligned(8))) temps[4];

1944 __asm__ volatile(

1945 "lea (%%"REG_a",%%"REG_a",2), %%"REG_d" \n\t"

1946 "6: \n\t"

1947 "movd (%%"REG_S"), %%mm0 \n\t"

1948 "pxor %%mm7, %%mm7 \n\t"

1949

1950 "movd (%%"REG_S",%%"REG_a",), %%mm1 \n\t"

1951 "punpcklbw %%mm7, %%mm0 \n\t"

1952

1953 "movd (%%"REG_S",%%"REG_a",2), %%mm2 \n\t"

1954 "punpcklbw %%mm7, %%mm1 \n\t"

1955

1956 "punpcklbw %%mm7, %%mm2 \n\t"

1957 "add %%"REG_d", %%"REG_S" \n\t"

1958

1959 "movq %%mm0, %%mm5 \n\t"

1960 //

1961

1962 "movd (%%"REG_S",%%"REG_a",4), %%mm3 \n\t" //7 ;prefetch!

1963 "movq %%mm1, %%mm6 \n\t"

1964

1965 "movd (%%"REG_S",%%"REG_d",), %%mm4 \n\t" //6

1966 "punpcklbw %%mm7, %%mm3 \n\t"

1967

1968 "psubw %%mm3, %%mm5 \n\t"

1969 "punpcklbw %%mm7, %%mm4 \n\t"

1970

1971 "paddw %%mm3, %%mm0 \n\t"

1972 "psubw %%mm4, %%mm6 \n\t"

1973

1974 "movd (%%"REG_S",%%"REG_a",2), %%mm3 \n\t" //5

1975 "paddw %%mm4, %%mm1 \n\t"

1976

1977 "movq %%mm5, 0*8+%3 \n\t" //t7

1978 "punpcklbw %%mm7, %%mm3 \n\t"

1979

1980 "movq %%mm6, 1*8+%3 \n\t" //t6

1981 "movq %%mm2, %%mm4 \n\t"

1982

1983 "movd (%%"REG_S"), %%mm5 \n\t" //3

1984 "paddw %%mm3, %%mm2 \n\t"

1985

1986 "movd (%%"REG_S",%%"REG_a",), %%mm6 \n\t" //4

1987 "punpcklbw %%mm7, %%mm5 \n\t"

1988

1989 "psubw %%mm3, %%mm4 \n\t"

1990 "punpcklbw %%mm7, %%mm6 \n\t"

1991

1992 "movq %%mm5, %%mm3 \n\t"

1993 "paddw %%mm6, %%mm5 \n\t" //t3

1994

1995 "psubw %%mm6, %%mm3 \n\t" //t4 ; t0 t1 t2 t4 t5 t3 - -

1996 "movq %%mm0, %%mm6 \n\t"

1997

1998 "movq %%mm1, %%mm7 \n\t"

1999 "psubw %%mm5, %%mm0 \n\t" //t13

2000

2001 "psubw %%mm2, %%mm1 \n\t"

2002 "paddw %%mm2, %%mm7 \n\t" //t11

2003

2004 "paddw %%mm0, %%mm1 \n\t"

2005 "movq %%mm7, %%mm2 \n\t"

2006

2007 "psllw 2,ドル %%mm1 \n\t"

2008 "paddw %%mm5, %%mm6 \n\t" //t10

2009

2010 "pmulhw "MANGLE(ff_MM_FIX_0_707106781)", %%mm1 \n\t"

2011 "paddw %%mm6, %%mm7 \n\t" //d2

2012

2013 "psubw %%mm2, %%mm6 \n\t" //d3

2014 "movq %%mm0, %%mm5 \n\t"

2015

2016 //transpose 4x4

2017 "movq %%mm7, %%mm2 \n\t"

2018 "punpcklwd %%mm6, %%mm7 \n\t"

2019

2020 "paddw %%mm1, %%mm0 \n\t" //d0

2021 "punpckhwd %%mm6, %%mm2 \n\t"

2022

2023 "psubw %%mm1, %%mm5 \n\t" //d1

2024 "movq %%mm0, %%mm6 \n\t"

2025

2026 "movq 1*8+%3, %%mm1 \n\t"

2027 "punpcklwd %%mm5, %%mm0 \n\t"

2028

2029 "punpckhwd %%mm5, %%mm6 \n\t"

2030 "movq %%mm0, %%mm5 \n\t"

2031

2032 "punpckldq %%mm7, %%mm0 \n\t" //0

2033 "paddw %%mm4, %%mm3 \n\t"

2034

2035 "punpckhdq %%mm7, %%mm5 \n\t" //1

2036 "movq %%mm6, %%mm7 \n\t"

2037

2038 "movq %%mm0, "DCTSIZE_S "*0*2(%%"REG_D") \n\t"

2039 "punpckldq %%mm2, %%mm6 \n\t" //2

2040

2041 "movq %%mm5, "DCTSIZE_S "*1*2(%%"REG_D") \n\t"

2042 "punpckhdq %%mm2, %%mm7 \n\t" //3

2043

2044 "movq %%mm6, "DCTSIZE_S "*2*2(%%"REG_D") \n\t"

2045 "paddw %%mm1, %%mm4 \n\t"

2046

2047 "movq %%mm7, "DCTSIZE_S "*3*2(%%"REG_D") \n\t"

2048 "psllw 2,ドル %%mm3 \n\t" //t10

2049

2050 "movq 0*8+%3, %%mm2 \n\t"

2051 "psllw 2,ドル %%mm4 \n\t" //t11

2052

2053 "pmulhw "MANGLE(ff_MM_FIX_0_707106781)", %%mm4 \n\t" //z3

2054 "paddw %%mm2, %%mm1 \n\t"

2055

2056 "psllw 2,ドル %%mm1 \n\t" //t12

2057 "movq %%mm3, %%mm0 \n\t"

2058

2059 "pmulhw "MANGLE(ff_MM_FIX_0_541196100)", %%mm0 \n\t"

2060 "psubw %%mm1, %%mm3 \n\t"

2061

2062 "pmulhw "MANGLE(MM_FIX_0_382683433)", %%mm3 \n\t" //z5

2063 "movq %%mm2, %%mm5 \n\t"

2064

2065 "pmulhw "MANGLE(MM_FIX_1_306562965)", %%mm1 \n\t"

2066 "psubw %%mm4, %%mm2 \n\t" //z13

2067

2068 "paddw %%mm4, %%mm5 \n\t" //z11

2069 "movq %%mm2, %%mm6 \n\t"

2070

2071 "paddw %%mm3, %%mm0 \n\t" //z2

2072 "movq %%mm5, %%mm7 \n\t"

2073

2074 "paddw %%mm0, %%mm2 \n\t" //d4

2075 "psubw %%mm0, %%mm6 \n\t" //d5

2076

2077 "movq %%mm2, %%mm4 \n\t"

2078 "paddw %%mm3, %%mm1 \n\t" //z4

2079

2080 //transpose 4x4

2081 "punpcklwd %%mm6, %%mm2 \n\t"

2082 "paddw %%mm1, %%mm5 \n\t" //d6

2083

2084 "punpckhwd %%mm6, %%mm4 \n\t"

2085 "psubw %%mm1, %%mm7 \n\t" //d7

2086

2087 "movq %%mm5, %%mm6 \n\t"

2088 "punpcklwd %%mm7, %%mm5 \n\t"

2089

2090 "punpckhwd %%mm7, %%mm6 \n\t"

2091 "movq %%mm2, %%mm7 \n\t"

2092

2093 "punpckldq %%mm5, %%mm2 \n\t" //4

2094 "sub %%"REG_d", %%"REG_S" \n\t"

2095

2096 "punpckhdq %%mm5, %%mm7 \n\t" //5

2097 "movq %%mm4, %%mm5 \n\t"

2098

2099 "movq %%mm2, "DCTSIZE_S "*0*2+"DCTSIZE_S "(%%"REG_D") \n\t"

2100 "punpckldq %%mm6, %%mm4 \n\t" //6

2101

2102 "movq %%mm7, "DCTSIZE_S "*1*2+"DCTSIZE_S "(%%"REG_D") \n\t"

2103 "punpckhdq %%mm6, %%mm5 \n\t" //7

2104

2105 "movq %%mm4, "DCTSIZE_S "*2*2+"DCTSIZE_S "(%%"REG_D") \n\t"

2106 "add 4,ドル %%"REG_S" \n\t"

2107

2108 "movq %%mm5, "DCTSIZE_S "*3*2+"DCTSIZE_S "(%%"REG_D") \n\t"

2109 "add $"DCTSIZE_S "*2*4, %%"REG_D" \n\t" //4 rows

2110 "dec %%"REG_c" \n\t"

2111 "jnz 6b \n\t"

2112

2113 : "+S"(pixels), "+D"(data), "+c"(cnt), "=o"(temps)

2114 : "a"(line_size)

2115 : "%"REG_d);

2116 }

2117

2118 #endif // HAVE_MMX

Generated on Wed Jul 10 2013 23:48:09 for FFmpeg by doxygen 1.8.2