2 * Copyright (C) 2001-2011 Michael Niedermayer <michaelni@gmx.at>
4 * This file is part of FFmpeg.
6 * FFmpeg is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
11 * FFmpeg 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 GNU
14 * Lesser General Public License for more details.
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with FFmpeg; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
26 #if COMPILE_TEMPLATE_MMXEXT
27 #define PREFETCH "prefetchnta"
29 #define PREFETCH " # nop"
32 #if COMPILE_TEMPLATE_MMXEXT
33 #define REAL_MOVNTQ(a,b) "movntq " #a ", " #b " \n\t"
34 #define MOVNTQ2 "movntq "
36 #define REAL_MOVNTQ(a,b) "movq " #a ", " #b " \n\t"
37 #define MOVNTQ2 "movq "
39 #define MOVNTQ(a,b) REAL_MOVNTQ(a,b)
41 #if !COMPILE_TEMPLATE_MMXEXT
42 static av_always_inline void
43 dither_8to16(const uint8_t *srcDither, int rot)
46 __asm__ volatile("pxor %%mm0, %%mm0\n\t"
47 "movq (%0), %%mm3\n\t"
48 "movq %%mm3, %%mm4\n\t"
49 "psrlq $24, %%mm3\n\t"
50 "psllq $40, %%mm4\n\t"
51 "por %%mm4, %%mm3\n\t"
52 "movq %%mm3, %%mm4\n\t"
53 "punpcklbw %%mm0, %%mm3\n\t"
54 "punpckhbw %%mm0, %%mm4\n\t"
58 __asm__ volatile("pxor %%mm0, %%mm0\n\t"
59 "movq (%0), %%mm3\n\t"
60 "movq %%mm3, %%mm4\n\t"
61 "punpcklbw %%mm0, %%mm3\n\t"
62 "punpckhbw %%mm0, %%mm4\n\t"
69 static void RENAME(yuv2yuvX)(const int16_t *filter, int filterSize,
70 const int16_t **src, uint8_t *dest, int dstW,
71 const uint8_t *dither, int offset)
73 dither_8to16(dither, offset);
77 "punpcklwd %%mm1, %%mm1\n\t"
78 "punpckldq %%mm1, %%mm1\n\t"
80 "paddw %%mm1, %%mm3\n\t"
81 "paddw %%mm1, %%mm4\n\t"
88 "movq %%mm3, %%mm6\n\t"
89 "movq %%mm4, %%mm7\n\t"
91 "mov %0, %%"REG_d" \n\t"\
92 "mov (%%"REG_d"), %%"REG_S" \n\t"\
93 ".p2align 4 \n\t" /* FIXME Unroll? */\
95 "movq 8(%%"REG_d"), %%mm0 \n\t" /* filterCoeff */\
96 "movq (%%"REG_S", %%"REG_c", 2), %%mm2 \n\t" /* srcData */\
97 "movq 8(%%"REG_S", %%"REG_c", 2), %%mm5 \n\t" /* srcData */\
98 "add $16, %%"REG_d" \n\t"\
99 "mov (%%"REG_d"), %%"REG_S" \n\t"\
100 "test %%"REG_S", %%"REG_S" \n\t"\
101 "pmulhw %%mm0, %%mm2 \n\t"\
102 "pmulhw %%mm0, %%mm5 \n\t"\
103 "paddw %%mm2, %%mm3 \n\t"\
104 "paddw %%mm5, %%mm4 \n\t"\
106 "psraw $3, %%mm3 \n\t"\
107 "psraw $3, %%mm4 \n\t"\
108 "packuswb %%mm4, %%mm3 \n\t"
109 MOVNTQ2 " %%mm3, (%1, %%"REG_c")\n\t"
110 "add $8, %%"REG_c" \n\t"\
111 "cmp %2, %%"REG_c" \n\t"\
112 "movq %%mm6, %%mm3\n\t"
113 "movq %%mm7, %%mm4\n\t"
114 "mov %0, %%"REG_d" \n\t"\
115 "mov (%%"REG_d"), %%"REG_S" \n\t"\
118 "r" (dest-offset), "g" ((x86_reg)(dstW+offset)), "m" (offset)
119 : "%"REG_d, "%"REG_S, "%"REG_c
123 #define YSCALEYUV2PACKEDX_UV \
125 "xor %%"REG_a", %%"REG_a" \n\t"\
129 "lea "CHR_MMX_FILTER_OFFSET"(%0), %%"REG_d" \n\t"\
130 "mov (%%"REG_d"), %%"REG_S" \n\t"\
131 "movq "VROUNDER_OFFSET"(%0), %%mm3 \n\t"\
132 "movq %%mm3, %%mm4 \n\t"\
135 "movq 8(%%"REG_d"), %%mm0 \n\t" /* filterCoeff */\
136 "movq (%%"REG_S", %%"REG_a"), %%mm2 \n\t" /* UsrcData */\
137 "add %6, %%"REG_S" \n\t" \
138 "movq (%%"REG_S", %%"REG_a"), %%mm5 \n\t" /* VsrcData */\
139 "add $16, %%"REG_d" \n\t"\
140 "mov (%%"REG_d"), %%"REG_S" \n\t"\
141 "pmulhw %%mm0, %%mm2 \n\t"\
142 "pmulhw %%mm0, %%mm5 \n\t"\
143 "paddw %%mm2, %%mm3 \n\t"\
144 "paddw %%mm5, %%mm4 \n\t"\
145 "test %%"REG_S", %%"REG_S" \n\t"\
148 #define YSCALEYUV2PACKEDX_YA(offset,coeff,src1,src2,dst1,dst2) \
149 "lea "offset"(%0), %%"REG_d" \n\t"\
150 "mov (%%"REG_d"), %%"REG_S" \n\t"\
151 "movq "VROUNDER_OFFSET"(%0), "#dst1" \n\t"\
152 "movq "#dst1", "#dst2" \n\t"\
155 "movq 8(%%"REG_d"), "#coeff" \n\t" /* filterCoeff */\
156 "movq (%%"REG_S", %%"REG_a", 2), "#src1" \n\t" /* Y1srcData */\
157 "movq 8(%%"REG_S", %%"REG_a", 2), "#src2" \n\t" /* Y2srcData */\
158 "add $16, %%"REG_d" \n\t"\
159 "mov (%%"REG_d"), %%"REG_S" \n\t"\
160 "pmulhw "#coeff", "#src1" \n\t"\
161 "pmulhw "#coeff", "#src2" \n\t"\
162 "paddw "#src1", "#dst1" \n\t"\
163 "paddw "#src2", "#dst2" \n\t"\
164 "test %%"REG_S", %%"REG_S" \n\t"\
167 #define YSCALEYUV2PACKEDX \
168 YSCALEYUV2PACKEDX_UV \
169 YSCALEYUV2PACKEDX_YA(LUM_MMX_FILTER_OFFSET,%%mm0,%%mm2,%%mm5,%%mm1,%%mm7) \
171 #define YSCALEYUV2PACKEDX_END \
172 :: "r" (&c->redDither), \
173 "m" (dummy), "m" (dummy), "m" (dummy),\
174 "r" (dest), "m" (dstW_reg), "m"(uv_off) \
175 : "%"REG_a, "%"REG_d, "%"REG_S \
178 #define YSCALEYUV2PACKEDX_ACCURATE_UV \
180 "xor %%"REG_a", %%"REG_a" \n\t"\
184 "lea "CHR_MMX_FILTER_OFFSET"(%0), %%"REG_d" \n\t"\
185 "mov (%%"REG_d"), %%"REG_S" \n\t"\
186 "pxor %%mm4, %%mm4 \n\t"\
187 "pxor %%mm5, %%mm5 \n\t"\
188 "pxor %%mm6, %%mm6 \n\t"\
189 "pxor %%mm7, %%mm7 \n\t"\
192 "movq (%%"REG_S", %%"REG_a"), %%mm0 \n\t" /* UsrcData */\
193 "add %6, %%"REG_S" \n\t" \
194 "movq (%%"REG_S", %%"REG_a"), %%mm2 \n\t" /* VsrcData */\
195 "mov "STR(APCK_PTR2)"(%%"REG_d"), %%"REG_S" \n\t"\
196 "movq (%%"REG_S", %%"REG_a"), %%mm1 \n\t" /* UsrcData */\
197 "movq %%mm0, %%mm3 \n\t"\
198 "punpcklwd %%mm1, %%mm0 \n\t"\
199 "punpckhwd %%mm1, %%mm3 \n\t"\
200 "movq "STR(APCK_COEF)"(%%"REG_d"),%%mm1 \n\t" /* filterCoeff */\
201 "pmaddwd %%mm1, %%mm0 \n\t"\
202 "pmaddwd %%mm1, %%mm3 \n\t"\
203 "paddd %%mm0, %%mm4 \n\t"\
204 "paddd %%mm3, %%mm5 \n\t"\
205 "add %6, %%"REG_S" \n\t" \
206 "movq (%%"REG_S", %%"REG_a"), %%mm3 \n\t" /* VsrcData */\
207 "mov "STR(APCK_SIZE)"(%%"REG_d"), %%"REG_S" \n\t"\
208 "add $"STR(APCK_SIZE)", %%"REG_d" \n\t"\
209 "test %%"REG_S", %%"REG_S" \n\t"\
210 "movq %%mm2, %%mm0 \n\t"\
211 "punpcklwd %%mm3, %%mm2 \n\t"\
212 "punpckhwd %%mm3, %%mm0 \n\t"\
213 "pmaddwd %%mm1, %%mm2 \n\t"\
214 "pmaddwd %%mm1, %%mm0 \n\t"\
215 "paddd %%mm2, %%mm6 \n\t"\
216 "paddd %%mm0, %%mm7 \n\t"\
218 "psrad $16, %%mm4 \n\t"\
219 "psrad $16, %%mm5 \n\t"\
220 "psrad $16, %%mm6 \n\t"\
221 "psrad $16, %%mm7 \n\t"\
222 "movq "VROUNDER_OFFSET"(%0), %%mm0 \n\t"\
223 "packssdw %%mm5, %%mm4 \n\t"\
224 "packssdw %%mm7, %%mm6 \n\t"\
225 "paddw %%mm0, %%mm4 \n\t"\
226 "paddw %%mm0, %%mm6 \n\t"\
227 "movq %%mm4, "U_TEMP"(%0) \n\t"\
228 "movq %%mm6, "V_TEMP"(%0) \n\t"\
230 #define YSCALEYUV2PACKEDX_ACCURATE_YA(offset) \
231 "lea "offset"(%0), %%"REG_d" \n\t"\
232 "mov (%%"REG_d"), %%"REG_S" \n\t"\
233 "pxor %%mm1, %%mm1 \n\t"\
234 "pxor %%mm5, %%mm5 \n\t"\
235 "pxor %%mm7, %%mm7 \n\t"\
236 "pxor %%mm6, %%mm6 \n\t"\
239 "movq (%%"REG_S", %%"REG_a", 2), %%mm0 \n\t" /* Y1srcData */\
240 "movq 8(%%"REG_S", %%"REG_a", 2), %%mm2 \n\t" /* Y2srcData */\
241 "mov "STR(APCK_PTR2)"(%%"REG_d"), %%"REG_S" \n\t"\
242 "movq (%%"REG_S", %%"REG_a", 2), %%mm4 \n\t" /* Y1srcData */\
243 "movq %%mm0, %%mm3 \n\t"\
244 "punpcklwd %%mm4, %%mm0 \n\t"\
245 "punpckhwd %%mm4, %%mm3 \n\t"\
246 "movq "STR(APCK_COEF)"(%%"REG_d"), %%mm4 \n\t" /* filterCoeff */\
247 "pmaddwd %%mm4, %%mm0 \n\t"\
248 "pmaddwd %%mm4, %%mm3 \n\t"\
249 "paddd %%mm0, %%mm1 \n\t"\
250 "paddd %%mm3, %%mm5 \n\t"\
251 "movq 8(%%"REG_S", %%"REG_a", 2), %%mm3 \n\t" /* Y2srcData */\
252 "mov "STR(APCK_SIZE)"(%%"REG_d"), %%"REG_S" \n\t"\
253 "add $"STR(APCK_SIZE)", %%"REG_d" \n\t"\
254 "test %%"REG_S", %%"REG_S" \n\t"\
255 "movq %%mm2, %%mm0 \n\t"\
256 "punpcklwd %%mm3, %%mm2 \n\t"\
257 "punpckhwd %%mm3, %%mm0 \n\t"\
258 "pmaddwd %%mm4, %%mm2 \n\t"\
259 "pmaddwd %%mm4, %%mm0 \n\t"\
260 "paddd %%mm2, %%mm7 \n\t"\
261 "paddd %%mm0, %%mm6 \n\t"\
263 "psrad $16, %%mm1 \n\t"\
264 "psrad $16, %%mm5 \n\t"\
265 "psrad $16, %%mm7 \n\t"\
266 "psrad $16, %%mm6 \n\t"\
267 "movq "VROUNDER_OFFSET"(%0), %%mm0 \n\t"\
268 "packssdw %%mm5, %%mm1 \n\t"\
269 "packssdw %%mm6, %%mm7 \n\t"\
270 "paddw %%mm0, %%mm1 \n\t"\
271 "paddw %%mm0, %%mm7 \n\t"\
272 "movq "U_TEMP"(%0), %%mm3 \n\t"\
273 "movq "V_TEMP"(%0), %%mm4 \n\t"\
275 #define YSCALEYUV2PACKEDX_ACCURATE \
276 YSCALEYUV2PACKEDX_ACCURATE_UV \
277 YSCALEYUV2PACKEDX_ACCURATE_YA(LUM_MMX_FILTER_OFFSET)
279 #define YSCALEYUV2RGBX \
280 "psubw "U_OFFSET"(%0), %%mm3 \n\t" /* (U-128)8*/\
281 "psubw "V_OFFSET"(%0), %%mm4 \n\t" /* (V-128)8*/\
282 "movq %%mm3, %%mm2 \n\t" /* (U-128)8*/\
283 "movq %%mm4, %%mm5 \n\t" /* (V-128)8*/\
284 "pmulhw "UG_COEFF"(%0), %%mm3 \n\t"\
285 "pmulhw "VG_COEFF"(%0), %%mm4 \n\t"\
286 /* mm2=(U-128)8, mm3=ug, mm4=vg mm5=(V-128)8 */\
287 "pmulhw "UB_COEFF"(%0), %%mm2 \n\t"\
288 "pmulhw "VR_COEFF"(%0), %%mm5 \n\t"\
289 "psubw "Y_OFFSET"(%0), %%mm1 \n\t" /* 8(Y-16)*/\
290 "psubw "Y_OFFSET"(%0), %%mm7 \n\t" /* 8(Y-16)*/\
291 "pmulhw "Y_COEFF"(%0), %%mm1 \n\t"\
292 "pmulhw "Y_COEFF"(%0), %%mm7 \n\t"\
293 /* mm1= Y1, mm2=ub, mm3=ug, mm4=vg mm5=vr, mm7=Y2 */\
294 "paddw %%mm3, %%mm4 \n\t"\
295 "movq %%mm2, %%mm0 \n\t"\
296 "movq %%mm5, %%mm6 \n\t"\
297 "movq %%mm4, %%mm3 \n\t"\
298 "punpcklwd %%mm2, %%mm2 \n\t"\
299 "punpcklwd %%mm5, %%mm5 \n\t"\
300 "punpcklwd %%mm4, %%mm4 \n\t"\
301 "paddw %%mm1, %%mm2 \n\t"\
302 "paddw %%mm1, %%mm5 \n\t"\
303 "paddw %%mm1, %%mm4 \n\t"\
304 "punpckhwd %%mm0, %%mm0 \n\t"\
305 "punpckhwd %%mm6, %%mm6 \n\t"\
306 "punpckhwd %%mm3, %%mm3 \n\t"\
307 "paddw %%mm7, %%mm0 \n\t"\
308 "paddw %%mm7, %%mm6 \n\t"\
309 "paddw %%mm7, %%mm3 \n\t"\
310 /* mm0=B1, mm2=B2, mm3=G2, mm4=G1, mm5=R1, mm6=R2 */\
311 "packuswb %%mm0, %%mm2 \n\t"\
312 "packuswb %%mm6, %%mm5 \n\t"\
313 "packuswb %%mm3, %%mm4 \n\t"\
315 #define REAL_WRITEBGR32(dst, dstw, index, b, g, r, a, q0, q2, q3, t) \
316 "movq "#b", "#q2" \n\t" /* B */\
317 "movq "#r", "#t" \n\t" /* R */\
318 "punpcklbw "#g", "#b" \n\t" /* GBGBGBGB 0 */\
319 "punpcklbw "#a", "#r" \n\t" /* ARARARAR 0 */\
320 "punpckhbw "#g", "#q2" \n\t" /* GBGBGBGB 2 */\
321 "punpckhbw "#a", "#t" \n\t" /* ARARARAR 2 */\
322 "movq "#b", "#q0" \n\t" /* GBGBGBGB 0 */\
323 "movq "#q2", "#q3" \n\t" /* GBGBGBGB 2 */\
324 "punpcklwd "#r", "#q0" \n\t" /* ARGBARGB 0 */\
325 "punpckhwd "#r", "#b" \n\t" /* ARGBARGB 1 */\
326 "punpcklwd "#t", "#q2" \n\t" /* ARGBARGB 2 */\
327 "punpckhwd "#t", "#q3" \n\t" /* ARGBARGB 3 */\
329 MOVNTQ( q0, (dst, index, 4))\
330 MOVNTQ( b, 8(dst, index, 4))\
331 MOVNTQ( q2, 16(dst, index, 4))\
332 MOVNTQ( q3, 24(dst, index, 4))\
334 "add $8, "#index" \n\t"\
335 "cmp "#dstw", "#index" \n\t"\
337 #define WRITEBGR32(dst, dstw, index, b, g, r, a, q0, q2, q3, t) REAL_WRITEBGR32(dst, dstw, index, b, g, r, a, q0, q2, q3, t)
339 static void RENAME(yuv2rgb32_X_ar)(SwsContext *c, const int16_t *lumFilter,
340 const int16_t **lumSrc, int lumFilterSize,
341 const int16_t *chrFilter, const int16_t **chrUSrc,
342 const int16_t **chrVSrc,
343 int chrFilterSize, const int16_t **alpSrc,
344 uint8_t *dest, int dstW, int dstY)
347 x86_reg dstW_reg = dstW;
348 x86_reg uv_off = c->uv_offx2;
350 if (CONFIG_SWSCALE_ALPHA && c->alpPixBuf) {
351 YSCALEYUV2PACKEDX_ACCURATE
353 "movq %%mm2, "U_TEMP"(%0) \n\t"
354 "movq %%mm4, "V_TEMP"(%0) \n\t"
355 "movq %%mm5, "Y_TEMP"(%0) \n\t"
356 YSCALEYUV2PACKEDX_ACCURATE_YA(ALP_MMX_FILTER_OFFSET)
357 "movq "Y_TEMP"(%0), %%mm5 \n\t"
358 "psraw $3, %%mm1 \n\t"
359 "psraw $3, %%mm7 \n\t"
360 "packuswb %%mm7, %%mm1 \n\t"
361 WRITEBGR32(%4, %5, %%REGa, %%mm3, %%mm4, %%mm5, %%mm1, %%mm0, %%mm7, %%mm2, %%mm6)
362 YSCALEYUV2PACKEDX_END
364 YSCALEYUV2PACKEDX_ACCURATE
366 "pcmpeqd %%mm7, %%mm7 \n\t"
367 WRITEBGR32(%4, %5, %%REGa, %%mm2, %%mm4, %%mm5, %%mm7, %%mm0, %%mm1, %%mm3, %%mm6)
368 YSCALEYUV2PACKEDX_END
372 static void RENAME(yuv2rgb32_X)(SwsContext *c, const int16_t *lumFilter,
373 const int16_t **lumSrc, int lumFilterSize,
374 const int16_t *chrFilter, const int16_t **chrUSrc,
375 const int16_t **chrVSrc,
376 int chrFilterSize, const int16_t **alpSrc,
377 uint8_t *dest, int dstW, int dstY)
380 x86_reg dstW_reg = dstW;
381 x86_reg uv_off = c->uv_offx2;
383 if (CONFIG_SWSCALE_ALPHA && c->alpPixBuf) {
386 YSCALEYUV2PACKEDX_YA(ALP_MMX_FILTER_OFFSET, %%mm0, %%mm3, %%mm6, %%mm1, %%mm7)
387 "psraw $3, %%mm1 \n\t"
388 "psraw $3, %%mm7 \n\t"
389 "packuswb %%mm7, %%mm1 \n\t"
390 WRITEBGR32(%4, %5, %%REGa, %%mm2, %%mm4, %%mm5, %%mm1, %%mm0, %%mm7, %%mm3, %%mm6)
391 YSCALEYUV2PACKEDX_END
395 "pcmpeqd %%mm7, %%mm7 \n\t"
396 WRITEBGR32(%4, %5, %%REGa, %%mm2, %%mm4, %%mm5, %%mm7, %%mm0, %%mm1, %%mm3, %%mm6)
397 YSCALEYUV2PACKEDX_END
401 #define REAL_WRITERGB16(dst, dstw, index) \
402 "pand "MANGLE(bF8)", %%mm2 \n\t" /* B */\
403 "pand "MANGLE(bFC)", %%mm4 \n\t" /* G */\
404 "pand "MANGLE(bF8)", %%mm5 \n\t" /* R */\
405 "psrlq $3, %%mm2 \n\t"\
407 "movq %%mm2, %%mm1 \n\t"\
408 "movq %%mm4, %%mm3 \n\t"\
410 "punpcklbw %%mm7, %%mm3 \n\t"\
411 "punpcklbw %%mm5, %%mm2 \n\t"\
412 "punpckhbw %%mm7, %%mm4 \n\t"\
413 "punpckhbw %%mm5, %%mm1 \n\t"\
415 "psllq $3, %%mm3 \n\t"\
416 "psllq $3, %%mm4 \n\t"\
418 "por %%mm3, %%mm2 \n\t"\
419 "por %%mm4, %%mm1 \n\t"\
421 MOVNTQ(%%mm2, (dst, index, 2))\
422 MOVNTQ(%%mm1, 8(dst, index, 2))\
424 "add $8, "#index" \n\t"\
425 "cmp "#dstw", "#index" \n\t"\
427 #define WRITERGB16(dst, dstw, index) REAL_WRITERGB16(dst, dstw, index)
429 static void RENAME(yuv2rgb565_X_ar)(SwsContext *c, const int16_t *lumFilter,
430 const int16_t **lumSrc, int lumFilterSize,
431 const int16_t *chrFilter, const int16_t **chrUSrc,
432 const int16_t **chrVSrc,
433 int chrFilterSize, const int16_t **alpSrc,
434 uint8_t *dest, int dstW, int dstY)
437 x86_reg dstW_reg = dstW;
438 x86_reg uv_off = c->uv_offx2;
440 YSCALEYUV2PACKEDX_ACCURATE
442 "pxor %%mm7, %%mm7 \n\t"
443 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */
445 "paddusb "BLUE_DITHER"(%0), %%mm2\n\t"
446 "paddusb "GREEN_DITHER"(%0), %%mm4\n\t"
447 "paddusb "RED_DITHER"(%0), %%mm5\n\t"
449 WRITERGB16(%4, %5, %%REGa)
450 YSCALEYUV2PACKEDX_END
453 static void RENAME(yuv2rgb565_X)(SwsContext *c, const int16_t *lumFilter,
454 const int16_t **lumSrc, int lumFilterSize,
455 const int16_t *chrFilter, const int16_t **chrUSrc,
456 const int16_t **chrVSrc,
457 int chrFilterSize, const int16_t **alpSrc,
458 uint8_t *dest, int dstW, int dstY)
461 x86_reg dstW_reg = dstW;
462 x86_reg uv_off = c->uv_offx2;
466 "pxor %%mm7, %%mm7 \n\t"
467 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */
469 "paddusb "BLUE_DITHER"(%0), %%mm2 \n\t"
470 "paddusb "GREEN_DITHER"(%0), %%mm4 \n\t"
471 "paddusb "RED_DITHER"(%0), %%mm5 \n\t"
473 WRITERGB16(%4, %5, %%REGa)
474 YSCALEYUV2PACKEDX_END
477 #define REAL_WRITERGB15(dst, dstw, index) \
478 "pand "MANGLE(bF8)", %%mm2 \n\t" /* B */\
479 "pand "MANGLE(bF8)", %%mm4 \n\t" /* G */\
480 "pand "MANGLE(bF8)", %%mm5 \n\t" /* R */\
481 "psrlq $3, %%mm2 \n\t"\
482 "psrlq $1, %%mm5 \n\t"\
484 "movq %%mm2, %%mm1 \n\t"\
485 "movq %%mm4, %%mm3 \n\t"\
487 "punpcklbw %%mm7, %%mm3 \n\t"\
488 "punpcklbw %%mm5, %%mm2 \n\t"\
489 "punpckhbw %%mm7, %%mm4 \n\t"\
490 "punpckhbw %%mm5, %%mm1 \n\t"\
492 "psllq $2, %%mm3 \n\t"\
493 "psllq $2, %%mm4 \n\t"\
495 "por %%mm3, %%mm2 \n\t"\
496 "por %%mm4, %%mm1 \n\t"\
498 MOVNTQ(%%mm2, (dst, index, 2))\
499 MOVNTQ(%%mm1, 8(dst, index, 2))\
501 "add $8, "#index" \n\t"\
502 "cmp "#dstw", "#index" \n\t"\
504 #define WRITERGB15(dst, dstw, index) REAL_WRITERGB15(dst, dstw, index)
506 static void RENAME(yuv2rgb555_X_ar)(SwsContext *c, const int16_t *lumFilter,
507 const int16_t **lumSrc, int lumFilterSize,
508 const int16_t *chrFilter, const int16_t **chrUSrc,
509 const int16_t **chrVSrc,
510 int chrFilterSize, const int16_t **alpSrc,
511 uint8_t *dest, int dstW, int dstY)
514 x86_reg dstW_reg = dstW;
515 x86_reg uv_off = c->uv_offx2;
517 YSCALEYUV2PACKEDX_ACCURATE
519 "pxor %%mm7, %%mm7 \n\t"
520 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */
522 "paddusb "BLUE_DITHER"(%0), %%mm2\n\t"
523 "paddusb "GREEN_DITHER"(%0), %%mm4\n\t"
524 "paddusb "RED_DITHER"(%0), %%mm5\n\t"
526 WRITERGB15(%4, %5, %%REGa)
527 YSCALEYUV2PACKEDX_END
530 static void RENAME(yuv2rgb555_X)(SwsContext *c, const int16_t *lumFilter,
531 const int16_t **lumSrc, int lumFilterSize,
532 const int16_t *chrFilter, const int16_t **chrUSrc,
533 const int16_t **chrVSrc,
534 int chrFilterSize, const int16_t **alpSrc,
535 uint8_t *dest, int dstW, int dstY)
538 x86_reg dstW_reg = dstW;
539 x86_reg uv_off = c->uv_offx2;
543 "pxor %%mm7, %%mm7 \n\t"
544 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */
546 "paddusb "BLUE_DITHER"(%0), %%mm2 \n\t"
547 "paddusb "GREEN_DITHER"(%0), %%mm4 \n\t"
548 "paddusb "RED_DITHER"(%0), %%mm5 \n\t"
550 WRITERGB15(%4, %5, %%REGa)
551 YSCALEYUV2PACKEDX_END
554 #define WRITEBGR24MMX(dst, dstw, index) \
555 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */\
556 "movq %%mm2, %%mm1 \n\t" /* B */\
557 "movq %%mm5, %%mm6 \n\t" /* R */\
558 "punpcklbw %%mm4, %%mm2 \n\t" /* GBGBGBGB 0 */\
559 "punpcklbw %%mm7, %%mm5 \n\t" /* 0R0R0R0R 0 */\
560 "punpckhbw %%mm4, %%mm1 \n\t" /* GBGBGBGB 2 */\
561 "punpckhbw %%mm7, %%mm6 \n\t" /* 0R0R0R0R 2 */\
562 "movq %%mm2, %%mm0 \n\t" /* GBGBGBGB 0 */\
563 "movq %%mm1, %%mm3 \n\t" /* GBGBGBGB 2 */\
564 "punpcklwd %%mm5, %%mm0 \n\t" /* 0RGB0RGB 0 */\
565 "punpckhwd %%mm5, %%mm2 \n\t" /* 0RGB0RGB 1 */\
566 "punpcklwd %%mm6, %%mm1 \n\t" /* 0RGB0RGB 2 */\
567 "punpckhwd %%mm6, %%mm3 \n\t" /* 0RGB0RGB 3 */\
569 "movq %%mm0, %%mm4 \n\t" /* 0RGB0RGB 0 */\
570 "movq %%mm2, %%mm6 \n\t" /* 0RGB0RGB 1 */\
571 "movq %%mm1, %%mm5 \n\t" /* 0RGB0RGB 2 */\
572 "movq %%mm3, %%mm7 \n\t" /* 0RGB0RGB 3 */\
574 "psllq $40, %%mm0 \n\t" /* RGB00000 0 */\
575 "psllq $40, %%mm2 \n\t" /* RGB00000 1 */\
576 "psllq $40, %%mm1 \n\t" /* RGB00000 2 */\
577 "psllq $40, %%mm3 \n\t" /* RGB00000 3 */\
579 "punpckhdq %%mm4, %%mm0 \n\t" /* 0RGBRGB0 0 */\
580 "punpckhdq %%mm6, %%mm2 \n\t" /* 0RGBRGB0 1 */\
581 "punpckhdq %%mm5, %%mm1 \n\t" /* 0RGBRGB0 2 */\
582 "punpckhdq %%mm7, %%mm3 \n\t" /* 0RGBRGB0 3 */\
584 "psrlq $8, %%mm0 \n\t" /* 00RGBRGB 0 */\
585 "movq %%mm2, %%mm6 \n\t" /* 0RGBRGB0 1 */\
586 "psllq $40, %%mm2 \n\t" /* GB000000 1 */\
587 "por %%mm2, %%mm0 \n\t" /* GBRGBRGB 0 */\
588 MOVNTQ(%%mm0, (dst))\
590 "psrlq $24, %%mm6 \n\t" /* 0000RGBR 1 */\
591 "movq %%mm1, %%mm5 \n\t" /* 0RGBRGB0 2 */\
592 "psllq $24, %%mm1 \n\t" /* BRGB0000 2 */\
593 "por %%mm1, %%mm6 \n\t" /* BRGBRGBR 1 */\
594 MOVNTQ(%%mm6, 8(dst))\
596 "psrlq $40, %%mm5 \n\t" /* 000000RG 2 */\
597 "psllq $8, %%mm3 \n\t" /* RGBRGB00 3 */\
598 "por %%mm3, %%mm5 \n\t" /* RGBRGBRG 2 */\
599 MOVNTQ(%%mm5, 16(dst))\
601 "add $24, "#dst" \n\t"\
603 "add $8, "#index" \n\t"\
604 "cmp "#dstw", "#index" \n\t"\
607 #define WRITEBGR24MMXEXT(dst, dstw, index) \
608 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */\
609 "movq "MANGLE(ff_M24A)", %%mm0 \n\t"\
610 "movq "MANGLE(ff_M24C)", %%mm7 \n\t"\
611 "pshufw $0x50, %%mm2, %%mm1 \n\t" /* B3 B2 B3 B2 B1 B0 B1 B0 */\
612 "pshufw $0x50, %%mm4, %%mm3 \n\t" /* G3 G2 G3 G2 G1 G0 G1 G0 */\
613 "pshufw $0x00, %%mm5, %%mm6 \n\t" /* R1 R0 R1 R0 R1 R0 R1 R0 */\
615 "pand %%mm0, %%mm1 \n\t" /* B2 B1 B0 */\
616 "pand %%mm0, %%mm3 \n\t" /* G2 G1 G0 */\
617 "pand %%mm7, %%mm6 \n\t" /* R1 R0 */\
619 "psllq $8, %%mm3 \n\t" /* G2 G1 G0 */\
620 "por %%mm1, %%mm6 \n\t"\
621 "por %%mm3, %%mm6 \n\t"\
622 MOVNTQ(%%mm6, (dst))\
624 "psrlq $8, %%mm4 \n\t" /* 00 G7 G6 G5 G4 G3 G2 G1 */\
625 "pshufw $0xA5, %%mm2, %%mm1 \n\t" /* B5 B4 B5 B4 B3 B2 B3 B2 */\
626 "pshufw $0x55, %%mm4, %%mm3 \n\t" /* G4 G3 G4 G3 G4 G3 G4 G3 */\
627 "pshufw $0xA5, %%mm5, %%mm6 \n\t" /* R5 R4 R5 R4 R3 R2 R3 R2 */\
629 "pand "MANGLE(ff_M24B)", %%mm1 \n\t" /* B5 B4 B3 */\
630 "pand %%mm7, %%mm3 \n\t" /* G4 G3 */\
631 "pand %%mm0, %%mm6 \n\t" /* R4 R3 R2 */\
633 "por %%mm1, %%mm3 \n\t" /* B5 G4 B4 G3 B3 */\
634 "por %%mm3, %%mm6 \n\t"\
635 MOVNTQ(%%mm6, 8(dst))\
637 "pshufw $0xFF, %%mm2, %%mm1 \n\t" /* B7 B6 B7 B6 B7 B6 B6 B7 */\
638 "pshufw $0xFA, %%mm4, %%mm3 \n\t" /* 00 G7 00 G7 G6 G5 G6 G5 */\
639 "pshufw $0xFA, %%mm5, %%mm6 \n\t" /* R7 R6 R7 R6 R5 R4 R5 R4 */\
641 "pand %%mm7, %%mm1 \n\t" /* B7 B6 */\
642 "pand %%mm0, %%mm3 \n\t" /* G7 G6 G5 */\
643 "pand "MANGLE(ff_M24B)", %%mm6 \n\t" /* R7 R6 R5 */\
645 "por %%mm1, %%mm3 \n\t"\
646 "por %%mm3, %%mm6 \n\t"\
647 MOVNTQ(%%mm6, 16(dst))\
649 "add $24, "#dst" \n\t"\
651 "add $8, "#index" \n\t"\
652 "cmp "#dstw", "#index" \n\t"\
655 #if COMPILE_TEMPLATE_MMXEXT
657 #define WRITEBGR24(dst, dstw, index) WRITEBGR24MMXEXT(dst, dstw, index)
660 #define WRITEBGR24(dst, dstw, index) WRITEBGR24MMX(dst, dstw, index)
663 static void RENAME(yuv2bgr24_X_ar)(SwsContext *c, const int16_t *lumFilter,
664 const int16_t **lumSrc, int lumFilterSize,
665 const int16_t *chrFilter, const int16_t **chrUSrc,
666 const int16_t **chrVSrc,
667 int chrFilterSize, const int16_t **alpSrc,
668 uint8_t *dest, int dstW, int dstY)
671 x86_reg dstW_reg = dstW;
672 x86_reg uv_off = c->uv_offx2;
674 YSCALEYUV2PACKEDX_ACCURATE
676 "pxor %%mm7, %%mm7 \n\t"
677 "lea (%%"REG_a", %%"REG_a", 2), %%"REG_c"\n\t" //FIXME optimize
678 "add %4, %%"REG_c" \n\t"
679 WRITEBGR24(%%REGc, %5, %%REGa)
680 :: "r" (&c->redDither),
681 "m" (dummy), "m" (dummy), "m" (dummy),
682 "r" (dest), "m" (dstW_reg), "m"(uv_off)
683 : "%"REG_a, "%"REG_c, "%"REG_d, "%"REG_S
687 static void RENAME(yuv2bgr24_X)(SwsContext *c, const int16_t *lumFilter,
688 const int16_t **lumSrc, int lumFilterSize,
689 const int16_t *chrFilter, const int16_t **chrUSrc,
690 const int16_t **chrVSrc,
691 int chrFilterSize, const int16_t **alpSrc,
692 uint8_t *dest, int dstW, int dstY)
695 x86_reg dstW_reg = dstW;
696 x86_reg uv_off = c->uv_offx2;
700 "pxor %%mm7, %%mm7 \n\t"
701 "lea (%%"REG_a", %%"REG_a", 2), %%"REG_c" \n\t" //FIXME optimize
702 "add %4, %%"REG_c" \n\t"
703 WRITEBGR24(%%REGc, %5, %%REGa)
704 :: "r" (&c->redDither),
705 "m" (dummy), "m" (dummy), "m" (dummy),
706 "r" (dest), "m" (dstW_reg), "m"(uv_off)
707 : "%"REG_a, "%"REG_c, "%"REG_d, "%"REG_S
711 #define REAL_WRITEYUY2(dst, dstw, index) \
712 "packuswb %%mm3, %%mm3 \n\t"\
713 "packuswb %%mm4, %%mm4 \n\t"\
714 "packuswb %%mm7, %%mm1 \n\t"\
715 "punpcklbw %%mm4, %%mm3 \n\t"\
716 "movq %%mm1, %%mm7 \n\t"\
717 "punpcklbw %%mm3, %%mm1 \n\t"\
718 "punpckhbw %%mm3, %%mm7 \n\t"\
720 MOVNTQ(%%mm1, (dst, index, 2))\
721 MOVNTQ(%%mm7, 8(dst, index, 2))\
723 "add $8, "#index" \n\t"\
724 "cmp "#dstw", "#index" \n\t"\
726 #define WRITEYUY2(dst, dstw, index) REAL_WRITEYUY2(dst, dstw, index)
728 static void RENAME(yuv2yuyv422_X_ar)(SwsContext *c, const int16_t *lumFilter,
729 const int16_t **lumSrc, int lumFilterSize,
730 const int16_t *chrFilter, const int16_t **chrUSrc,
731 const int16_t **chrVSrc,
732 int chrFilterSize, const int16_t **alpSrc,
733 uint8_t *dest, int dstW, int dstY)
736 x86_reg dstW_reg = dstW;
737 x86_reg uv_off = c->uv_offx2;
739 YSCALEYUV2PACKEDX_ACCURATE
740 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */
741 "psraw $3, %%mm3 \n\t"
742 "psraw $3, %%mm4 \n\t"
743 "psraw $3, %%mm1 \n\t"
744 "psraw $3, %%mm7 \n\t"
745 WRITEYUY2(%4, %5, %%REGa)
746 YSCALEYUV2PACKEDX_END
749 static void RENAME(yuv2yuyv422_X)(SwsContext *c, const int16_t *lumFilter,
750 const int16_t **lumSrc, int lumFilterSize,
751 const int16_t *chrFilter, const int16_t **chrUSrc,
752 const int16_t **chrVSrc,
753 int chrFilterSize, const int16_t **alpSrc,
754 uint8_t *dest, int dstW, int dstY)
757 x86_reg dstW_reg = dstW;
758 x86_reg uv_off = c->uv_offx2;
761 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */
762 "psraw $3, %%mm3 \n\t"
763 "psraw $3, %%mm4 \n\t"
764 "psraw $3, %%mm1 \n\t"
765 "psraw $3, %%mm7 \n\t"
766 WRITEYUY2(%4, %5, %%REGa)
767 YSCALEYUV2PACKEDX_END
770 #define REAL_YSCALEYUV2RGB_UV(index, c) \
771 "xor "#index", "#index" \n\t"\
774 "movq (%2, "#index"), %%mm2 \n\t" /* uvbuf0[eax]*/\
775 "movq (%3, "#index"), %%mm3 \n\t" /* uvbuf1[eax]*/\
776 "add "UV_OFF_BYTE"("#c"), "#index" \n\t" \
777 "movq (%2, "#index"), %%mm5 \n\t" /* uvbuf0[eax+2048]*/\
778 "movq (%3, "#index"), %%mm4 \n\t" /* uvbuf1[eax+2048]*/\
779 "sub "UV_OFF_BYTE"("#c"), "#index" \n\t" \
780 "psubw %%mm3, %%mm2 \n\t" /* uvbuf0[eax] - uvbuf1[eax]*/\
781 "psubw %%mm4, %%mm5 \n\t" /* uvbuf0[eax+2048] - uvbuf1[eax+2048]*/\
782 "movq "CHR_MMX_FILTER_OFFSET"+8("#c"), %%mm0 \n\t"\
783 "pmulhw %%mm0, %%mm2 \n\t" /* (uvbuf0[eax] - uvbuf1[eax])uvalpha1>>16*/\
784 "pmulhw %%mm0, %%mm5 \n\t" /* (uvbuf0[eax+2048] - uvbuf1[eax+2048])uvalpha1>>16*/\
785 "psraw $4, %%mm3 \n\t" /* uvbuf0[eax] - uvbuf1[eax] >>4*/\
786 "psraw $4, %%mm4 \n\t" /* uvbuf0[eax+2048] - uvbuf1[eax+2048] >>4*/\
787 "paddw %%mm2, %%mm3 \n\t" /* uvbuf0[eax]uvalpha1 - uvbuf1[eax](1-uvalpha1)*/\
788 "paddw %%mm5, %%mm4 \n\t" /* uvbuf0[eax+2048]uvalpha1 - uvbuf1[eax+2048](1-uvalpha1)*/\
789 "psubw "U_OFFSET"("#c"), %%mm3 \n\t" /* (U-128)8*/\
790 "psubw "V_OFFSET"("#c"), %%mm4 \n\t" /* (V-128)8*/\
791 "movq %%mm3, %%mm2 \n\t" /* (U-128)8*/\
792 "movq %%mm4, %%mm5 \n\t" /* (V-128)8*/\
793 "pmulhw "UG_COEFF"("#c"), %%mm3 \n\t"\
794 "pmulhw "VG_COEFF"("#c"), %%mm4 \n\t"\
795 /* mm2=(U-128)8, mm3=ug, mm4=vg mm5=(V-128)8 */\
797 #define REAL_YSCALEYUV2RGB_YA(index, c, b1, b2) \
798 "movq ("#b1", "#index", 2), %%mm0 \n\t" /*buf0[eax]*/\
799 "movq ("#b2", "#index", 2), %%mm1 \n\t" /*buf1[eax]*/\
800 "movq 8("#b1", "#index", 2), %%mm6 \n\t" /*buf0[eax]*/\
801 "movq 8("#b2", "#index", 2), %%mm7 \n\t" /*buf1[eax]*/\
802 "psubw %%mm1, %%mm0 \n\t" /* buf0[eax] - buf1[eax]*/\
803 "psubw %%mm7, %%mm6 \n\t" /* buf0[eax] - buf1[eax]*/\
804 "pmulhw "LUM_MMX_FILTER_OFFSET"+8("#c"), %%mm0 \n\t" /* (buf0[eax] - buf1[eax])yalpha1>>16*/\
805 "pmulhw "LUM_MMX_FILTER_OFFSET"+8("#c"), %%mm6 \n\t" /* (buf0[eax] - buf1[eax])yalpha1>>16*/\
806 "psraw $4, %%mm1 \n\t" /* buf0[eax] - buf1[eax] >>4*/\
807 "psraw $4, %%mm7 \n\t" /* buf0[eax] - buf1[eax] >>4*/\
808 "paddw %%mm0, %%mm1 \n\t" /* buf0[eax]yalpha1 + buf1[eax](1-yalpha1) >>16*/\
809 "paddw %%mm6, %%mm7 \n\t" /* buf0[eax]yalpha1 + buf1[eax](1-yalpha1) >>16*/\
811 #define REAL_YSCALEYUV2RGB_COEFF(c) \
812 "pmulhw "UB_COEFF"("#c"), %%mm2 \n\t"\
813 "pmulhw "VR_COEFF"("#c"), %%mm5 \n\t"\
814 "psubw "Y_OFFSET"("#c"), %%mm1 \n\t" /* 8(Y-16)*/\
815 "psubw "Y_OFFSET"("#c"), %%mm7 \n\t" /* 8(Y-16)*/\
816 "pmulhw "Y_COEFF"("#c"), %%mm1 \n\t"\
817 "pmulhw "Y_COEFF"("#c"), %%mm7 \n\t"\
818 /* mm1= Y1, mm2=ub, mm3=ug, mm4=vg mm5=vr, mm7=Y2 */\
819 "paddw %%mm3, %%mm4 \n\t"\
820 "movq %%mm2, %%mm0 \n\t"\
821 "movq %%mm5, %%mm6 \n\t"\
822 "movq %%mm4, %%mm3 \n\t"\
823 "punpcklwd %%mm2, %%mm2 \n\t"\
824 "punpcklwd %%mm5, %%mm5 \n\t"\
825 "punpcklwd %%mm4, %%mm4 \n\t"\
826 "paddw %%mm1, %%mm2 \n\t"\
827 "paddw %%mm1, %%mm5 \n\t"\
828 "paddw %%mm1, %%mm4 \n\t"\
829 "punpckhwd %%mm0, %%mm0 \n\t"\
830 "punpckhwd %%mm6, %%mm6 \n\t"\
831 "punpckhwd %%mm3, %%mm3 \n\t"\
832 "paddw %%mm7, %%mm0 \n\t"\
833 "paddw %%mm7, %%mm6 \n\t"\
834 "paddw %%mm7, %%mm3 \n\t"\
835 /* mm0=B1, mm2=B2, mm3=G2, mm4=G1, mm5=R1, mm6=R2 */\
836 "packuswb %%mm0, %%mm2 \n\t"\
837 "packuswb %%mm6, %%mm5 \n\t"\
838 "packuswb %%mm3, %%mm4 \n\t"\
840 #define YSCALEYUV2RGB_YA(index, c, b1, b2) REAL_YSCALEYUV2RGB_YA(index, c, b1, b2)
842 #define YSCALEYUV2RGB(index, c) \
843 REAL_YSCALEYUV2RGB_UV(index, c) \
844 REAL_YSCALEYUV2RGB_YA(index, c, %0, %1) \
845 REAL_YSCALEYUV2RGB_COEFF(c)
848 * vertical bilinear scale YV12 to RGB
850 static void RENAME(yuv2rgb32_2)(SwsContext *c, const int16_t *buf[2],
851 const int16_t *ubuf[2], const int16_t *vbuf[2],
852 const int16_t *abuf[2], uint8_t *dest,
853 int dstW, int yalpha, int uvalpha, int y)
855 const int16_t *buf0 = buf[0], *buf1 = buf[1],
856 *ubuf0 = ubuf[0], *ubuf1 = ubuf[1];
858 if (CONFIG_SWSCALE_ALPHA && c->alpPixBuf) {
859 const int16_t *abuf0 = abuf[0], *abuf1 = abuf[1];
862 YSCALEYUV2RGB(%%r8, %5)
863 YSCALEYUV2RGB_YA(%%r8, %5, %6, %7)
864 "psraw $3, %%mm1 \n\t" /* abuf0[eax] - abuf1[eax] >>7*/
865 "psraw $3, %%mm7 \n\t" /* abuf0[eax] - abuf1[eax] >>7*/
866 "packuswb %%mm7, %%mm1 \n\t"
867 WRITEBGR32(%4, 8280(%5), %%r8, %%mm2, %%mm4, %%mm5, %%mm1, %%mm0, %%mm7, %%mm3, %%mm6)
868 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "r" (dest),
870 "r" (abuf0), "r" (abuf1)
874 c->u_temp=(intptr_t)abuf0;
875 c->v_temp=(intptr_t)abuf1;
877 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
878 "mov %4, %%"REG_b" \n\t"
879 "push %%"REG_BP" \n\t"
880 YSCALEYUV2RGB(%%REGBP, %5)
883 "mov "U_TEMP"(%5), %0 \n\t"
884 "mov "V_TEMP"(%5), %1 \n\t"
885 YSCALEYUV2RGB_YA(%%REGBP, %5, %0, %1)
886 "psraw $3, %%mm1 \n\t" /* abuf0[eax] - abuf1[eax] >>7*/
887 "psraw $3, %%mm7 \n\t" /* abuf0[eax] - abuf1[eax] >>7*/
888 "packuswb %%mm7, %%mm1 \n\t"
891 WRITEBGR32(%%REGb, 8280(%5), %%REGBP, %%mm2, %%mm4, %%mm5, %%mm1, %%mm0, %%mm7, %%mm3, %%mm6)
892 "pop %%"REG_BP" \n\t"
893 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
894 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
900 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
901 "mov %4, %%"REG_b" \n\t"
902 "push %%"REG_BP" \n\t"
903 YSCALEYUV2RGB(%%REGBP, %5)
904 "pcmpeqd %%mm7, %%mm7 \n\t"
905 WRITEBGR32(%%REGb, 8280(%5), %%REGBP, %%mm2, %%mm4, %%mm5, %%mm7, %%mm0, %%mm1, %%mm3, %%mm6)
906 "pop %%"REG_BP" \n\t"
907 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
908 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
914 static void RENAME(yuv2bgr24_2)(SwsContext *c, const int16_t *buf[2],
915 const int16_t *ubuf[2], const int16_t *vbuf[2],
916 const int16_t *abuf[2], uint8_t *dest,
917 int dstW, int yalpha, int uvalpha, int y)
919 const int16_t *buf0 = buf[0], *buf1 = buf[1],
920 *ubuf0 = ubuf[0], *ubuf1 = ubuf[1];
922 //Note 8280 == DSTW_OFFSET but the preprocessor can't handle that there :(
924 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
925 "mov %4, %%"REG_b" \n\t"
926 "push %%"REG_BP" \n\t"
927 YSCALEYUV2RGB(%%REGBP, %5)
928 "pxor %%mm7, %%mm7 \n\t"
929 WRITEBGR24(%%REGb, 8280(%5), %%REGBP)
930 "pop %%"REG_BP" \n\t"
931 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
932 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
937 static void RENAME(yuv2rgb555_2)(SwsContext *c, const int16_t *buf[2],
938 const int16_t *ubuf[2], const int16_t *vbuf[2],
939 const int16_t *abuf[2], uint8_t *dest,
940 int dstW, int yalpha, int uvalpha, int y)
942 const int16_t *buf0 = buf[0], *buf1 = buf[1],
943 *ubuf0 = ubuf[0], *ubuf1 = ubuf[1];
945 //Note 8280 == DSTW_OFFSET but the preprocessor can't handle that there :(
947 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
948 "mov %4, %%"REG_b" \n\t"
949 "push %%"REG_BP" \n\t"
950 YSCALEYUV2RGB(%%REGBP, %5)
951 "pxor %%mm7, %%mm7 \n\t"
952 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */
954 "paddusb "BLUE_DITHER"(%5), %%mm2 \n\t"
955 "paddusb "GREEN_DITHER"(%5), %%mm4 \n\t"
956 "paddusb "RED_DITHER"(%5), %%mm5 \n\t"
958 WRITERGB15(%%REGb, 8280(%5), %%REGBP)
959 "pop %%"REG_BP" \n\t"
960 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
961 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
966 static void RENAME(yuv2rgb565_2)(SwsContext *c, const int16_t *buf[2],
967 const int16_t *ubuf[2], const int16_t *vbuf[2],
968 const int16_t *abuf[2], uint8_t *dest,
969 int dstW, int yalpha, int uvalpha, int y)
971 const int16_t *buf0 = buf[0], *buf1 = buf[1],
972 *ubuf0 = ubuf[0], *ubuf1 = ubuf[1];
974 //Note 8280 == DSTW_OFFSET but the preprocessor can't handle that there :(
976 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
977 "mov %4, %%"REG_b" \n\t"
978 "push %%"REG_BP" \n\t"
979 YSCALEYUV2RGB(%%REGBP, %5)
980 "pxor %%mm7, %%mm7 \n\t"
981 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */
983 "paddusb "BLUE_DITHER"(%5), %%mm2 \n\t"
984 "paddusb "GREEN_DITHER"(%5), %%mm4 \n\t"
985 "paddusb "RED_DITHER"(%5), %%mm5 \n\t"
987 WRITERGB16(%%REGb, 8280(%5), %%REGBP)
988 "pop %%"REG_BP" \n\t"
989 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
990 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
995 #define REAL_YSCALEYUV2PACKED(index, c) \
996 "movq "CHR_MMX_FILTER_OFFSET"+8("#c"), %%mm0 \n\t"\
997 "movq "LUM_MMX_FILTER_OFFSET"+8("#c"), %%mm1 \n\t"\
998 "psraw $3, %%mm0 \n\t"\
999 "psraw $3, %%mm1 \n\t"\
1000 "movq %%mm0, "CHR_MMX_FILTER_OFFSET"+8("#c") \n\t"\
1001 "movq %%mm1, "LUM_MMX_FILTER_OFFSET"+8("#c") \n\t"\
1002 "xor "#index", "#index" \n\t"\
1005 "movq (%2, "#index"), %%mm2 \n\t" /* uvbuf0[eax]*/\
1006 "movq (%3, "#index"), %%mm3 \n\t" /* uvbuf1[eax]*/\
1007 "add "UV_OFF_BYTE"("#c"), "#index" \n\t" \
1008 "movq (%2, "#index"), %%mm5 \n\t" /* uvbuf0[eax+2048]*/\
1009 "movq (%3, "#index"), %%mm4 \n\t" /* uvbuf1[eax+2048]*/\
1010 "sub "UV_OFF_BYTE"("#c"), "#index" \n\t" \
1011 "psubw %%mm3, %%mm2 \n\t" /* uvbuf0[eax] - uvbuf1[eax]*/\
1012 "psubw %%mm4, %%mm5 \n\t" /* uvbuf0[eax+2048] - uvbuf1[eax+2048]*/\
1013 "movq "CHR_MMX_FILTER_OFFSET"+8("#c"), %%mm0 \n\t"\
1014 "pmulhw %%mm0, %%mm2 \n\t" /* (uvbuf0[eax] - uvbuf1[eax])uvalpha1>>16*/\
1015 "pmulhw %%mm0, %%mm5 \n\t" /* (uvbuf0[eax+2048] - uvbuf1[eax+2048])uvalpha1>>16*/\
1016 "psraw $7, %%mm3 \n\t" /* uvbuf0[eax] - uvbuf1[eax] >>4*/\
1017 "psraw $7, %%mm4 \n\t" /* uvbuf0[eax+2048] - uvbuf1[eax+2048] >>4*/\
1018 "paddw %%mm2, %%mm3 \n\t" /* uvbuf0[eax]uvalpha1 - uvbuf1[eax](1-uvalpha1)*/\
1019 "paddw %%mm5, %%mm4 \n\t" /* uvbuf0[eax+2048]uvalpha1 - uvbuf1[eax+2048](1-uvalpha1)*/\
1020 "movq (%0, "#index", 2), %%mm0 \n\t" /*buf0[eax]*/\
1021 "movq (%1, "#index", 2), %%mm1 \n\t" /*buf1[eax]*/\
1022 "movq 8(%0, "#index", 2), %%mm6 \n\t" /*buf0[eax]*/\
1023 "movq 8(%1, "#index", 2), %%mm7 \n\t" /*buf1[eax]*/\
1024 "psubw %%mm1, %%mm0 \n\t" /* buf0[eax] - buf1[eax]*/\
1025 "psubw %%mm7, %%mm6 \n\t" /* buf0[eax] - buf1[eax]*/\
1026 "pmulhw "LUM_MMX_FILTER_OFFSET"+8("#c"), %%mm0 \n\t" /* (buf0[eax] - buf1[eax])yalpha1>>16*/\
1027 "pmulhw "LUM_MMX_FILTER_OFFSET"+8("#c"), %%mm6 \n\t" /* (buf0[eax] - buf1[eax])yalpha1>>16*/\
1028 "psraw $7, %%mm1 \n\t" /* buf0[eax] - buf1[eax] >>4*/\
1029 "psraw $7, %%mm7 \n\t" /* buf0[eax] - buf1[eax] >>4*/\
1030 "paddw %%mm0, %%mm1 \n\t" /* buf0[eax]yalpha1 + buf1[eax](1-yalpha1) >>16*/\
1031 "paddw %%mm6, %%mm7 \n\t" /* buf0[eax]yalpha1 + buf1[eax](1-yalpha1) >>16*/\
1033 #define YSCALEYUV2PACKED(index, c) REAL_YSCALEYUV2PACKED(index, c)
1035 static void RENAME(yuv2yuyv422_2)(SwsContext *c, const int16_t *buf[2],
1036 const int16_t *ubuf[2], const int16_t *vbuf[2],
1037 const int16_t *abuf[2], uint8_t *dest,
1038 int dstW, int yalpha, int uvalpha, int y)
1040 const int16_t *buf0 = buf[0], *buf1 = buf[1],
1041 *ubuf0 = ubuf[0], *ubuf1 = ubuf[1];
1043 //Note 8280 == DSTW_OFFSET but the preprocessor can't handle that there :(
1045 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1046 "mov %4, %%"REG_b" \n\t"
1047 "push %%"REG_BP" \n\t"
1048 YSCALEYUV2PACKED(%%REGBP, %5)
1049 WRITEYUY2(%%REGb, 8280(%5), %%REGBP)
1050 "pop %%"REG_BP" \n\t"
1051 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1052 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1057 #define REAL_YSCALEYUV2RGB1(index, c) \
1058 "xor "#index", "#index" \n\t"\
1061 "movq (%2, "#index"), %%mm3 \n\t" /* uvbuf0[eax]*/\
1062 "add "UV_OFF_BYTE"("#c"), "#index" \n\t" \
1063 "movq (%2, "#index"), %%mm4 \n\t" /* uvbuf0[eax+2048]*/\
1064 "sub "UV_OFF_BYTE"("#c"), "#index" \n\t" \
1065 "psraw $4, %%mm3 \n\t" /* uvbuf0[eax] - uvbuf1[eax] >>4*/\
1066 "psraw $4, %%mm4 \n\t" /* uvbuf0[eax+2048] - uvbuf1[eax+2048] >>4*/\
1067 "psubw "U_OFFSET"("#c"), %%mm3 \n\t" /* (U-128)8*/\
1068 "psubw "V_OFFSET"("#c"), %%mm4 \n\t" /* (V-128)8*/\
1069 "movq %%mm3, %%mm2 \n\t" /* (U-128)8*/\
1070 "movq %%mm4, %%mm5 \n\t" /* (V-128)8*/\
1071 "pmulhw "UG_COEFF"("#c"), %%mm3 \n\t"\
1072 "pmulhw "VG_COEFF"("#c"), %%mm4 \n\t"\
1073 /* mm2=(U-128)8, mm3=ug, mm4=vg mm5=(V-128)8 */\
1074 "movq (%0, "#index", 2), %%mm1 \n\t" /*buf0[eax]*/\
1075 "movq 8(%0, "#index", 2), %%mm7 \n\t" /*buf0[eax]*/\
1076 "psraw $4, %%mm1 \n\t" /* buf0[eax] - buf1[eax] >>4*/\
1077 "psraw $4, %%mm7 \n\t" /* buf0[eax] - buf1[eax] >>4*/\
1078 "pmulhw "UB_COEFF"("#c"), %%mm2 \n\t"\
1079 "pmulhw "VR_COEFF"("#c"), %%mm5 \n\t"\
1080 "psubw "Y_OFFSET"("#c"), %%mm1 \n\t" /* 8(Y-16)*/\
1081 "psubw "Y_OFFSET"("#c"), %%mm7 \n\t" /* 8(Y-16)*/\
1082 "pmulhw "Y_COEFF"("#c"), %%mm1 \n\t"\
1083 "pmulhw "Y_COEFF"("#c"), %%mm7 \n\t"\
1084 /* mm1= Y1, mm2=ub, mm3=ug, mm4=vg mm5=vr, mm7=Y2 */\
1085 "paddw %%mm3, %%mm4 \n\t"\
1086 "movq %%mm2, %%mm0 \n\t"\
1087 "movq %%mm5, %%mm6 \n\t"\
1088 "movq %%mm4, %%mm3 \n\t"\
1089 "punpcklwd %%mm2, %%mm2 \n\t"\
1090 "punpcklwd %%mm5, %%mm5 \n\t"\
1091 "punpcklwd %%mm4, %%mm4 \n\t"\
1092 "paddw %%mm1, %%mm2 \n\t"\
1093 "paddw %%mm1, %%mm5 \n\t"\
1094 "paddw %%mm1, %%mm4 \n\t"\
1095 "punpckhwd %%mm0, %%mm0 \n\t"\
1096 "punpckhwd %%mm6, %%mm6 \n\t"\
1097 "punpckhwd %%mm3, %%mm3 \n\t"\
1098 "paddw %%mm7, %%mm0 \n\t"\
1099 "paddw %%mm7, %%mm6 \n\t"\
1100 "paddw %%mm7, %%mm3 \n\t"\
1101 /* mm0=B1, mm2=B2, mm3=G2, mm4=G1, mm5=R1, mm6=R2 */\
1102 "packuswb %%mm0, %%mm2 \n\t"\
1103 "packuswb %%mm6, %%mm5 \n\t"\
1104 "packuswb %%mm3, %%mm4 \n\t"\
1106 #define YSCALEYUV2RGB1(index, c) REAL_YSCALEYUV2RGB1(index, c)
1108 // do vertical chrominance interpolation
1109 #define REAL_YSCALEYUV2RGB1b(index, c) \
1110 "xor "#index", "#index" \n\t"\
1113 "movq (%2, "#index"), %%mm2 \n\t" /* uvbuf0[eax]*/\
1114 "movq (%3, "#index"), %%mm3 \n\t" /* uvbuf1[eax]*/\
1115 "add "UV_OFF_BYTE"("#c"), "#index" \n\t" \
1116 "movq (%2, "#index"), %%mm5 \n\t" /* uvbuf0[eax+2048]*/\
1117 "movq (%3, "#index"), %%mm4 \n\t" /* uvbuf1[eax+2048]*/\
1118 "sub "UV_OFF_BYTE"("#c"), "#index" \n\t" \
1119 "paddw %%mm2, %%mm3 \n\t" /* uvbuf0[eax] + uvbuf1[eax]*/\
1120 "paddw %%mm5, %%mm4 \n\t" /* uvbuf0[eax+2048] + uvbuf1[eax+2048]*/\
1121 "psrlw $5, %%mm3 \n\t" /*FIXME might overflow*/\
1122 "psrlw $5, %%mm4 \n\t" /*FIXME might overflow*/\
1123 "psubw "U_OFFSET"("#c"), %%mm3 \n\t" /* (U-128)8*/\
1124 "psubw "V_OFFSET"("#c"), %%mm4 \n\t" /* (V-128)8*/\
1125 "movq %%mm3, %%mm2 \n\t" /* (U-128)8*/\
1126 "movq %%mm4, %%mm5 \n\t" /* (V-128)8*/\
1127 "pmulhw "UG_COEFF"("#c"), %%mm3 \n\t"\
1128 "pmulhw "VG_COEFF"("#c"), %%mm4 \n\t"\
1129 /* mm2=(U-128)8, mm3=ug, mm4=vg mm5=(V-128)8 */\
1130 "movq (%0, "#index", 2), %%mm1 \n\t" /*buf0[eax]*/\
1131 "movq 8(%0, "#index", 2), %%mm7 \n\t" /*buf0[eax]*/\
1132 "psraw $4, %%mm1 \n\t" /* buf0[eax] - buf1[eax] >>4*/\
1133 "psraw $4, %%mm7 \n\t" /* buf0[eax] - buf1[eax] >>4*/\
1134 "pmulhw "UB_COEFF"("#c"), %%mm2 \n\t"\
1135 "pmulhw "VR_COEFF"("#c"), %%mm5 \n\t"\
1136 "psubw "Y_OFFSET"("#c"), %%mm1 \n\t" /* 8(Y-16)*/\
1137 "psubw "Y_OFFSET"("#c"), %%mm7 \n\t" /* 8(Y-16)*/\
1138 "pmulhw "Y_COEFF"("#c"), %%mm1 \n\t"\
1139 "pmulhw "Y_COEFF"("#c"), %%mm7 \n\t"\
1140 /* mm1= Y1, mm2=ub, mm3=ug, mm4=vg mm5=vr, mm7=Y2 */\
1141 "paddw %%mm3, %%mm4 \n\t"\
1142 "movq %%mm2, %%mm0 \n\t"\
1143 "movq %%mm5, %%mm6 \n\t"\
1144 "movq %%mm4, %%mm3 \n\t"\
1145 "punpcklwd %%mm2, %%mm2 \n\t"\
1146 "punpcklwd %%mm5, %%mm5 \n\t"\
1147 "punpcklwd %%mm4, %%mm4 \n\t"\
1148 "paddw %%mm1, %%mm2 \n\t"\
1149 "paddw %%mm1, %%mm5 \n\t"\
1150 "paddw %%mm1, %%mm4 \n\t"\
1151 "punpckhwd %%mm0, %%mm0 \n\t"\
1152 "punpckhwd %%mm6, %%mm6 \n\t"\
1153 "punpckhwd %%mm3, %%mm3 \n\t"\
1154 "paddw %%mm7, %%mm0 \n\t"\
1155 "paddw %%mm7, %%mm6 \n\t"\
1156 "paddw %%mm7, %%mm3 \n\t"\
1157 /* mm0=B1, mm2=B2, mm3=G2, mm4=G1, mm5=R1, mm6=R2 */\
1158 "packuswb %%mm0, %%mm2 \n\t"\
1159 "packuswb %%mm6, %%mm5 \n\t"\
1160 "packuswb %%mm3, %%mm4 \n\t"\
1162 #define YSCALEYUV2RGB1b(index, c) REAL_YSCALEYUV2RGB1b(index, c)
1164 #define REAL_YSCALEYUV2RGB1_ALPHA(index) \
1165 "movq (%1, "#index", 2), %%mm7 \n\t" /* abuf0[index ] */\
1166 "movq 8(%1, "#index", 2), %%mm1 \n\t" /* abuf0[index+4] */\
1167 "psraw $7, %%mm7 \n\t" /* abuf0[index ] >>7 */\
1168 "psraw $7, %%mm1 \n\t" /* abuf0[index+4] >>7 */\
1169 "packuswb %%mm1, %%mm7 \n\t"
1170 #define YSCALEYUV2RGB1_ALPHA(index) REAL_YSCALEYUV2RGB1_ALPHA(index)
1173 * YV12 to RGB without scaling or interpolating
1175 static void RENAME(yuv2rgb32_1)(SwsContext *c, const int16_t *buf0,
1176 const int16_t *ubuf[2], const int16_t *vbuf[2],
1177 const int16_t *abuf0, uint8_t *dest,
1178 int dstW, int uvalpha, int y)
1180 const int16_t *ubuf0 = ubuf[0];
1181 const int16_t *buf1= buf0; //FIXME needed for RGB1/BGR1
1183 if (uvalpha < 2048) { // note this is not correct (shifts chrominance by 0.5 pixels) but it is a bit faster
1184 const int16_t *ubuf1 = ubuf[0];
1185 if (CONFIG_SWSCALE_ALPHA && c->alpPixBuf) {
1187 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1188 "mov %4, %%"REG_b" \n\t"
1189 "push %%"REG_BP" \n\t"
1190 YSCALEYUV2RGB1(%%REGBP, %5)
1191 YSCALEYUV2RGB1_ALPHA(%%REGBP)
1192 WRITEBGR32(%%REGb, 8280(%5), %%REGBP, %%mm2, %%mm4, %%mm5, %%mm7, %%mm0, %%mm1, %%mm3, %%mm6)
1193 "pop %%"REG_BP" \n\t"
1194 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1195 :: "c" (buf0), "d" (abuf0), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1200 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1201 "mov %4, %%"REG_b" \n\t"
1202 "push %%"REG_BP" \n\t"
1203 YSCALEYUV2RGB1(%%REGBP, %5)
1204 "pcmpeqd %%mm7, %%mm7 \n\t"
1205 WRITEBGR32(%%REGb, 8280(%5), %%REGBP, %%mm2, %%mm4, %%mm5, %%mm7, %%mm0, %%mm1, %%mm3, %%mm6)
1206 "pop %%"REG_BP" \n\t"
1207 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1208 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1213 const int16_t *ubuf1 = ubuf[1];
1214 if (CONFIG_SWSCALE_ALPHA && c->alpPixBuf) {
1216 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1217 "mov %4, %%"REG_b" \n\t"
1218 "push %%"REG_BP" \n\t"
1219 YSCALEYUV2RGB1b(%%REGBP, %5)
1220 YSCALEYUV2RGB1_ALPHA(%%REGBP)
1221 WRITEBGR32(%%REGb, 8280(%5), %%REGBP, %%mm2, %%mm4, %%mm5, %%mm7, %%mm0, %%mm1, %%mm3, %%mm6)
1222 "pop %%"REG_BP" \n\t"
1223 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1224 :: "c" (buf0), "d" (abuf0), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1229 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1230 "mov %4, %%"REG_b" \n\t"
1231 "push %%"REG_BP" \n\t"
1232 YSCALEYUV2RGB1b(%%REGBP, %5)
1233 "pcmpeqd %%mm7, %%mm7 \n\t"
1234 WRITEBGR32(%%REGb, 8280(%5), %%REGBP, %%mm2, %%mm4, %%mm5, %%mm7, %%mm0, %%mm1, %%mm3, %%mm6)
1235 "pop %%"REG_BP" \n\t"
1236 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1237 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1244 static void RENAME(yuv2bgr24_1)(SwsContext *c, const int16_t *buf0,
1245 const int16_t *ubuf[2], const int16_t *vbuf[2],
1246 const int16_t *abuf0, uint8_t *dest,
1247 int dstW, int uvalpha, int y)
1249 const int16_t *ubuf0 = ubuf[0];
1250 const int16_t *buf1= buf0; //FIXME needed for RGB1/BGR1
1252 if (uvalpha < 2048) { // note this is not correct (shifts chrominance by 0.5 pixels) but it is a bit faster
1253 const int16_t *ubuf1 = ubuf[0];
1255 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1256 "mov %4, %%"REG_b" \n\t"
1257 "push %%"REG_BP" \n\t"
1258 YSCALEYUV2RGB1(%%REGBP, %5)
1259 "pxor %%mm7, %%mm7 \n\t"
1260 WRITEBGR24(%%REGb, 8280(%5), %%REGBP)
1261 "pop %%"REG_BP" \n\t"
1262 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1263 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1267 const int16_t *ubuf1 = ubuf[1];
1269 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1270 "mov %4, %%"REG_b" \n\t"
1271 "push %%"REG_BP" \n\t"
1272 YSCALEYUV2RGB1b(%%REGBP, %5)
1273 "pxor %%mm7, %%mm7 \n\t"
1274 WRITEBGR24(%%REGb, 8280(%5), %%REGBP)
1275 "pop %%"REG_BP" \n\t"
1276 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1277 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1283 static void RENAME(yuv2rgb555_1)(SwsContext *c, const int16_t *buf0,
1284 const int16_t *ubuf[2], const int16_t *vbuf[2],
1285 const int16_t *abuf0, uint8_t *dest,
1286 int dstW, int uvalpha, int y)
1288 const int16_t *ubuf0 = ubuf[0];
1289 const int16_t *buf1= buf0; //FIXME needed for RGB1/BGR1
1291 if (uvalpha < 2048) { // note this is not correct (shifts chrominance by 0.5 pixels) but it is a bit faster
1292 const int16_t *ubuf1 = ubuf[0];
1294 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1295 "mov %4, %%"REG_b" \n\t"
1296 "push %%"REG_BP" \n\t"
1297 YSCALEYUV2RGB1(%%REGBP, %5)
1298 "pxor %%mm7, %%mm7 \n\t"
1299 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */
1301 "paddusb "BLUE_DITHER"(%5), %%mm2 \n\t"
1302 "paddusb "GREEN_DITHER"(%5), %%mm4 \n\t"
1303 "paddusb "RED_DITHER"(%5), %%mm5 \n\t"
1305 WRITERGB15(%%REGb, 8280(%5), %%REGBP)
1306 "pop %%"REG_BP" \n\t"
1307 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1308 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1312 const int16_t *ubuf1 = ubuf[1];
1314 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1315 "mov %4, %%"REG_b" \n\t"
1316 "push %%"REG_BP" \n\t"
1317 YSCALEYUV2RGB1b(%%REGBP, %5)
1318 "pxor %%mm7, %%mm7 \n\t"
1319 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */
1321 "paddusb "BLUE_DITHER"(%5), %%mm2 \n\t"
1322 "paddusb "GREEN_DITHER"(%5), %%mm4 \n\t"
1323 "paddusb "RED_DITHER"(%5), %%mm5 \n\t"
1325 WRITERGB15(%%REGb, 8280(%5), %%REGBP)
1326 "pop %%"REG_BP" \n\t"
1327 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1328 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1334 static void RENAME(yuv2rgb565_1)(SwsContext *c, const int16_t *buf0,
1335 const int16_t *ubuf[2], const int16_t *vbuf[2],
1336 const int16_t *abuf0, uint8_t *dest,
1337 int dstW, int uvalpha, int y)
1339 const int16_t *ubuf0 = ubuf[0];
1340 const int16_t *buf1= buf0; //FIXME needed for RGB1/BGR1
1342 if (uvalpha < 2048) { // note this is not correct (shifts chrominance by 0.5 pixels) but it is a bit faster
1343 const int16_t *ubuf1 = ubuf[0];
1345 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1346 "mov %4, %%"REG_b" \n\t"
1347 "push %%"REG_BP" \n\t"
1348 YSCALEYUV2RGB1(%%REGBP, %5)
1349 "pxor %%mm7, %%mm7 \n\t"
1350 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */
1352 "paddusb "BLUE_DITHER"(%5), %%mm2 \n\t"
1353 "paddusb "GREEN_DITHER"(%5), %%mm4 \n\t"
1354 "paddusb "RED_DITHER"(%5), %%mm5 \n\t"
1356 WRITERGB16(%%REGb, 8280(%5), %%REGBP)
1357 "pop %%"REG_BP" \n\t"
1358 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1359 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1363 const int16_t *ubuf1 = ubuf[1];
1365 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1366 "mov %4, %%"REG_b" \n\t"
1367 "push %%"REG_BP" \n\t"
1368 YSCALEYUV2RGB1b(%%REGBP, %5)
1369 "pxor %%mm7, %%mm7 \n\t"
1370 /* mm2=B, %%mm4=G, %%mm5=R, %%mm7=0 */
1372 "paddusb "BLUE_DITHER"(%5), %%mm2 \n\t"
1373 "paddusb "GREEN_DITHER"(%5), %%mm4 \n\t"
1374 "paddusb "RED_DITHER"(%5), %%mm5 \n\t"
1376 WRITERGB16(%%REGb, 8280(%5), %%REGBP)
1377 "pop %%"REG_BP" \n\t"
1378 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1379 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1385 #define REAL_YSCALEYUV2PACKED1(index, c) \
1386 "xor "#index", "#index" \n\t"\
1389 "movq (%2, "#index"), %%mm3 \n\t" /* uvbuf0[eax]*/\
1390 "add "UV_OFF_BYTE"("#c"), "#index" \n\t" \
1391 "movq (%2, "#index"), %%mm4 \n\t" /* uvbuf0[eax+2048]*/\
1392 "sub "UV_OFF_BYTE"("#c"), "#index" \n\t" \
1393 "psraw $7, %%mm3 \n\t" \
1394 "psraw $7, %%mm4 \n\t" \
1395 "movq (%0, "#index", 2), %%mm1 \n\t" /*buf0[eax]*/\
1396 "movq 8(%0, "#index", 2), %%mm7 \n\t" /*buf0[eax]*/\
1397 "psraw $7, %%mm1 \n\t" \
1398 "psraw $7, %%mm7 \n\t" \
1400 #define YSCALEYUV2PACKED1(index, c) REAL_YSCALEYUV2PACKED1(index, c)
1402 #define REAL_YSCALEYUV2PACKED1b(index, c) \
1403 "xor "#index", "#index" \n\t"\
1406 "movq (%2, "#index"), %%mm2 \n\t" /* uvbuf0[eax]*/\
1407 "movq (%3, "#index"), %%mm3 \n\t" /* uvbuf1[eax]*/\
1408 "add "UV_OFF_BYTE"("#c"), "#index" \n\t" \
1409 "movq (%2, "#index"), %%mm5 \n\t" /* uvbuf0[eax+2048]*/\
1410 "movq (%3, "#index"), %%mm4 \n\t" /* uvbuf1[eax+2048]*/\
1411 "sub "UV_OFF_BYTE"("#c"), "#index" \n\t" \
1412 "paddw %%mm2, %%mm3 \n\t" /* uvbuf0[eax] + uvbuf1[eax]*/\
1413 "paddw %%mm5, %%mm4 \n\t" /* uvbuf0[eax+2048] + uvbuf1[eax+2048]*/\
1414 "psrlw $8, %%mm3 \n\t" \
1415 "psrlw $8, %%mm4 \n\t" \
1416 "movq (%0, "#index", 2), %%mm1 \n\t" /*buf0[eax]*/\
1417 "movq 8(%0, "#index", 2), %%mm7 \n\t" /*buf0[eax]*/\
1418 "psraw $7, %%mm1 \n\t" \
1419 "psraw $7, %%mm7 \n\t"
1420 #define YSCALEYUV2PACKED1b(index, c) REAL_YSCALEYUV2PACKED1b(index, c)
1422 static void RENAME(yuv2yuyv422_1)(SwsContext *c, const int16_t *buf0,
1423 const int16_t *ubuf[2], const int16_t *vbuf[2],
1424 const int16_t *abuf0, uint8_t *dest,
1425 int dstW, int uvalpha, int y)
1427 const int16_t *ubuf0 = ubuf[0];
1428 const int16_t *buf1= buf0; //FIXME needed for RGB1/BGR1
1430 if (uvalpha < 2048) { // note this is not correct (shifts chrominance by 0.5 pixels) but it is a bit faster
1431 const int16_t *ubuf1 = ubuf[0];
1433 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1434 "mov %4, %%"REG_b" \n\t"
1435 "push %%"REG_BP" \n\t"
1436 YSCALEYUV2PACKED1(%%REGBP, %5)
1437 WRITEYUY2(%%REGb, 8280(%5), %%REGBP)
1438 "pop %%"REG_BP" \n\t"
1439 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1440 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1444 const int16_t *ubuf1 = ubuf[1];
1446 "mov %%"REG_b", "ESP_OFFSET"(%5) \n\t"
1447 "mov %4, %%"REG_b" \n\t"
1448 "push %%"REG_BP" \n\t"
1449 YSCALEYUV2PACKED1b(%%REGBP, %5)
1450 WRITEYUY2(%%REGb, 8280(%5), %%REGBP)
1451 "pop %%"REG_BP" \n\t"
1452 "mov "ESP_OFFSET"(%5), %%"REG_b" \n\t"
1453 :: "c" (buf0), "d" (buf1), "S" (ubuf0), "D" (ubuf1), "m" (dest),
1459 #if COMPILE_TEMPLATE_MMXEXT
1460 static void RENAME(hyscale_fast)(SwsContext *c, int16_t *dst,
1461 int dstWidth, const uint8_t *src,
1464 int32_t *filterPos = c->hLumFilterPos;
1465 int16_t *filter = c->hLumFilter;
1466 void *mmxextFilterCode = c->lumMmxextFilterCode;
1477 "mov %%"REG_b", %5 \n\t"
1479 "mov -8(%%rsp), %%"REG_a" \n\t"
1480 "mov %%"REG_a", %6 \n\t"
1484 "mov -8(%%rsp), %%"REG_a" \n\t"
1485 "mov %%"REG_a", %5 \n\t"
1488 "pxor %%mm7, %%mm7 \n\t"
1489 "mov %0, %%"REG_c" \n\t"
1490 "mov %1, %%"REG_D" \n\t"
1491 "mov %2, %%"REG_d" \n\t"
1492 "mov %3, %%"REG_b" \n\t"
1493 "xor %%"REG_a", %%"REG_a" \n\t" // i
1494 PREFETCH" (%%"REG_c") \n\t"
1495 PREFETCH" 32(%%"REG_c") \n\t"
1496 PREFETCH" 64(%%"REG_c") \n\t"
1499 #define CALL_MMXEXT_FILTER_CODE \
1500 "movl (%%"REG_b"), %%esi \n\t"\
1502 "movl (%%"REG_b", %%"REG_a"), %%esi \n\t"\
1503 "add %%"REG_S", %%"REG_c" \n\t"\
1504 "add %%"REG_a", %%"REG_D" \n\t"\
1505 "xor %%"REG_a", %%"REG_a" \n\t"\
1508 #define CALL_MMXEXT_FILTER_CODE \
1509 "movl (%%"REG_b"), %%esi \n\t"\
1511 "addl (%%"REG_b", %%"REG_a"), %%"REG_c" \n\t"\
1512 "add %%"REG_a", %%"REG_D" \n\t"\
1513 "xor %%"REG_a", %%"REG_a" \n\t"\
1515 #endif /* ARCH_X86_64 */
1517 CALL_MMXEXT_FILTER_CODE
1518 CALL_MMXEXT_FILTER_CODE
1519 CALL_MMXEXT_FILTER_CODE
1520 CALL_MMXEXT_FILTER_CODE
1521 CALL_MMXEXT_FILTER_CODE
1522 CALL_MMXEXT_FILTER_CODE
1523 CALL_MMXEXT_FILTER_CODE
1524 CALL_MMXEXT_FILTER_CODE
1527 "mov %5, %%"REG_b" \n\t"
1529 "mov %6, %%"REG_a" \n\t"
1530 "mov %%"REG_a", -8(%%rsp) \n\t"
1534 "mov %5, %%"REG_a" \n\t"
1535 "mov %%"REG_a", -8(%%rsp) \n\t"
1538 :: "m" (src), "m" (dst), "m" (filter), "m" (filterPos),
1539 "m" (mmxextFilterCode)
1546 : "%"REG_a, "%"REG_c, "%"REG_d, "%"REG_S, "%"REG_D
1552 for (i=dstWidth-1; (i*xInc)>>16 >=srcW-1; i--)
1553 dst[i] = src[srcW-1]*128;
1556 static void RENAME(hcscale_fast)(SwsContext *c, int16_t *dst1, int16_t *dst2,
1557 int dstWidth, const uint8_t *src1,
1558 const uint8_t *src2, int srcW, int xInc)
1560 int32_t *filterPos = c->hChrFilterPos;
1561 int16_t *filter = c->hChrFilter;
1562 void *mmxextFilterCode = c->chrMmxextFilterCode;
1565 DECLARE_ALIGNED(8, uint64_t, ebxsave);
1568 DECLARE_ALIGNED(8, uint64_t, retsave);
1573 "mov %%"REG_b", %7 \n\t"
1575 "mov -8(%%rsp), %%"REG_a" \n\t"
1576 "mov %%"REG_a", %8 \n\t"
1580 "mov -8(%%rsp), %%"REG_a" \n\t"
1581 "mov %%"REG_a", %7 \n\t"
1584 "pxor %%mm7, %%mm7 \n\t"
1585 "mov %0, %%"REG_c" \n\t"
1586 "mov %1, %%"REG_D" \n\t"
1587 "mov %2, %%"REG_d" \n\t"
1588 "mov %3, %%"REG_b" \n\t"
1589 "xor %%"REG_a", %%"REG_a" \n\t" // i
1590 PREFETCH" (%%"REG_c") \n\t"
1591 PREFETCH" 32(%%"REG_c") \n\t"
1592 PREFETCH" 64(%%"REG_c") \n\t"
1594 CALL_MMXEXT_FILTER_CODE
1595 CALL_MMXEXT_FILTER_CODE
1596 CALL_MMXEXT_FILTER_CODE
1597 CALL_MMXEXT_FILTER_CODE
1598 "xor %%"REG_a", %%"REG_a" \n\t" // i
1599 "mov %5, %%"REG_c" \n\t" // src
1600 "mov %6, %%"REG_D" \n\t" // buf2
1601 PREFETCH" (%%"REG_c") \n\t"
1602 PREFETCH" 32(%%"REG_c") \n\t"
1603 PREFETCH" 64(%%"REG_c") \n\t"
1605 CALL_MMXEXT_FILTER_CODE
1606 CALL_MMXEXT_FILTER_CODE
1607 CALL_MMXEXT_FILTER_CODE
1608 CALL_MMXEXT_FILTER_CODE
1611 "mov %7, %%"REG_b" \n\t"
1613 "mov %8, %%"REG_a" \n\t"
1614 "mov %%"REG_a", -8(%%rsp) \n\t"
1618 "mov %7, %%"REG_a" \n\t"
1619 "mov %%"REG_a", -8(%%rsp) \n\t"
1622 :: "m" (src1), "m" (dst1), "m" (filter), "m" (filterPos),
1623 "m" (mmxextFilterCode), "m" (src2), "m"(dst2)
1630 : "%"REG_a, "%"REG_c, "%"REG_d, "%"REG_S, "%"REG_D
1636 for (i=dstWidth-1; (i*xInc)>>16 >=srcW-1; i--) {
1637 dst1[i] = src1[srcW-1]*128;
1638 dst2[i] = src2[srcW-1]*128;
1641 #endif /* COMPILE_TEMPLATE_MMXEXT */
1643 static av_cold void RENAME(sws_init_swscale)(SwsContext *c)
1645 enum AVPixelFormat dstFormat = c->dstFormat;
1647 c->use_mmx_vfilter= 0;
1648 if (!is16BPS(dstFormat) && !is9_OR_10BPS(dstFormat) && dstFormat != AV_PIX_FMT_NV12
1649 && dstFormat != AV_PIX_FMT_NV21 && !(c->flags & SWS_BITEXACT)) {
1650 if (c->flags & SWS_ACCURATE_RND) {
1651 if (!(c->flags & SWS_FULL_CHR_H_INT)) {
1652 switch (c->dstFormat) {
1653 case AV_PIX_FMT_RGB32: c->yuv2packedX = RENAME(yuv2rgb32_X_ar); break;
1654 case AV_PIX_FMT_BGR24: c->yuv2packedX = RENAME(yuv2bgr24_X_ar); break;
1655 case AV_PIX_FMT_RGB555: c->yuv2packedX = RENAME(yuv2rgb555_X_ar); break;
1656 case AV_PIX_FMT_RGB565: c->yuv2packedX = RENAME(yuv2rgb565_X_ar); break;
1657 case AV_PIX_FMT_YUYV422: c->yuv2packedX = RENAME(yuv2yuyv422_X_ar); break;
1662 c->use_mmx_vfilter= 1;
1663 c->yuv2planeX = RENAME(yuv2yuvX );
1664 if (!(c->flags & SWS_FULL_CHR_H_INT)) {
1665 switch (c->dstFormat) {
1666 case AV_PIX_FMT_RGB32: c->yuv2packedX = RENAME(yuv2rgb32_X); break;
1667 case AV_PIX_FMT_BGR24: c->yuv2packedX = RENAME(yuv2bgr24_X); break;
1668 case AV_PIX_FMT_RGB555: c->yuv2packedX = RENAME(yuv2rgb555_X); break;
1669 case AV_PIX_FMT_RGB565: c->yuv2packedX = RENAME(yuv2rgb565_X); break;
1670 case AV_PIX_FMT_YUYV422: c->yuv2packedX = RENAME(yuv2yuyv422_X); break;
1675 if (!(c->flags & SWS_FULL_CHR_H_INT)) {
1676 switch (c->dstFormat) {
1677 case AV_PIX_FMT_RGB32:
1678 c->yuv2packed1 = RENAME(yuv2rgb32_1);
1679 c->yuv2packed2 = RENAME(yuv2rgb32_2);
1681 case AV_PIX_FMT_BGR24:
1682 c->yuv2packed1 = RENAME(yuv2bgr24_1);
1683 c->yuv2packed2 = RENAME(yuv2bgr24_2);
1685 case AV_PIX_FMT_RGB555:
1686 c->yuv2packed1 = RENAME(yuv2rgb555_1);
1687 c->yuv2packed2 = RENAME(yuv2rgb555_2);
1689 case AV_PIX_FMT_RGB565:
1690 c->yuv2packed1 = RENAME(yuv2rgb565_1);
1691 c->yuv2packed2 = RENAME(yuv2rgb565_2);
1693 case AV_PIX_FMT_YUYV422:
1694 c->yuv2packed1 = RENAME(yuv2yuyv422_1);
1695 c->yuv2packed2 = RENAME(yuv2yuyv422_2);
1703 if (c->srcBpc == 8 && c->dstBpc <= 14) {
1704 // Use the new MMX scaler if the MMXEXT one can't be used (it is faster than the x86 ASM one).
1705 #if COMPILE_TEMPLATE_MMXEXT
1706 if (c->flags & SWS_FAST_BILINEAR && c->canMMXEXTBeUsed) {
1707 c->hyscale_fast = RENAME(hyscale_fast);
1708 c->hcscale_fast = RENAME(hcscale_fast);
1710 #endif /* COMPILE_TEMPLATE_MMXEXT */
1711 c->hyscale_fast = NULL;
1712 c->hcscale_fast = NULL;
1713 #if COMPILE_TEMPLATE_MMXEXT
1715 #endif /* COMPILE_TEMPLATE_MMXEXT */