libswscale/yuv2rgb_template.c

   1 /*
   2  * yuv2rgb_mmx.c, software YUV to RGB converter with Intel MMX "technology"
   3  *
   4  * Copyright (C) 2000, Silicon Integrated System Corp
   5  *
   6  * Author: Olie Lho <ollie@sis.com.tw>
   7  *
   8  * 15,24 bpp and dithering from Michael Niedermayer (michaelni@gmx.at)
   9  * MMX/MMX2 Template stuff from Michael Niedermayer (needed for fast movntq support)
  10  * context / deglobalize stuff by Michael Niedermayer
  11  *
  12  * This file is part of mpeg2dec, a free MPEG-2 video decoder
  13  *
  14  * mpeg2dec is free software; you can redistribute it and/or modify
  15  * it under the terms of the GNU General Public License as published by
  16  * the Free Software Foundation; either version 2, or (at your option)
  17  * any later version.
  18  *
  19  * mpeg2dec is distributed in the hope that it will be useful,
  20  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  21  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  22  * GNU General Public License for more details.
  23  *
  24  * You should have received a copy of the GNU General Public License
  25  * along with mpeg2dec; if not, write to the Free Software
  26  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  27  */
  28
  29 #undef MOVNTQ
  30 #undef EMMS
  31 #undef SFENCE
  32
  33 #if HAVE_AMD3DNOW
  34 /* On K6 femms is faster than emms. On K7 femms is directly mapped to emms. */
  35 #define EMMS     "femms"
  36 #else
  37 #define EMMS     "emms"
  38 #endif
  39
  40 #if HAVE_MMX2
  41 #define MOVNTQ "movntq"
  42 #define SFENCE "sfence"
  43 #else
  44 #define MOVNTQ "movq"
  45 #define SFENCE "/nop"
  46 #endif
  47
  48 #define YUV2RGB \
  49     /* Do the multiply part of the conversion for even and odd pixels,
  50        register usage:
  51        mm0 -> Cblue, mm1 -> Cred, mm2 -> Cgreen even pixels,
  52        mm3 -> Cblue, mm4 -> Cred, mm5 -> Cgreen odd pixels,
  53        mm6 -> Y even, mm7 -> Y odd */\
  54     /* convert the chroma part */\
  55     "punpcklbw %%mm4, %%mm0;" /* scatter 4 Cb 00 u3 00 u2 00 u1 00 u0 */ \
  56     "punpcklbw %%mm4, %%mm1;" /* scatter 4 Cr 00 v3 00 v2 00 v1 00 v0 */ \
  57 \
  58     "psllw $3, %%mm0;" /* Promote precision */ \
  59     "psllw $3, %%mm1;" /* Promote precision */ \
  60 \
  61     "psubsw "U_OFFSET"(%4), %%mm0;" /* Cb -= 128 */ \
  62     "psubsw "V_OFFSET"(%4), %%mm1;" /* Cr -= 128 */ \
  63 \
  64     "movq %%mm0, %%mm2;" /* Copy 4 Cb 00 u3 00 u2 00 u1 00 u0 */ \
  65     "movq %%mm1, %%mm3;" /* Copy 4 Cr 00 v3 00 v2 00 v1 00 v0 */ \
  66 \
  67     "pmulhw "UG_COEFF"(%4), %%mm2;" /* Mul Cb with green coeff -> Cb green */ \
  68     "pmulhw "VG_COEFF"(%4), %%mm3;" /* Mul Cr with green coeff -> Cr green */ \
  69 \
  70     "pmulhw "UB_COEFF"(%4), %%mm0;" /* Mul Cb -> Cblue 00 b3 00 b2 00 b1 00 b0 */\
  71     "pmulhw "VR_COEFF"(%4), %%mm1;" /* Mul Cr -> Cred 00 r3 00 r2 00 r1 00 r0 */\
  72 \
  73     "paddsw %%mm3, %%mm2;" /* Cb green + Cr green -> Cgreen */\
  74 \
  75     /* convert the luma part */\
  76     "movq %%mm6, %%mm7;" /* Copy 8 Y Y7 Y6 Y5 Y4 Y3 Y2 Y1 Y0 */\
  77     "pand "MANGLE(mmx_00ffw)", %%mm6;" /* get Y even 00 Y6 00 Y4 00 Y2 00 Y0 */\
  78 \
  79     "psrlw $8, %%mm7;" /* get Y odd 00 Y7 00 Y5 00 Y3 00 Y1 */\
  80 \
  81     "psllw $3, %%mm6;" /* Promote precision */\
  82     "psllw $3, %%mm7;" /* Promote precision */\
  83 \
  84     "psubw "Y_OFFSET"(%4), %%mm6;" /* Y -= 16 */\
  85     "psubw "Y_OFFSET"(%4), %%mm7;" /* Y -= 16 */\
  86 \
  87     "pmulhw "Y_COEFF"(%4), %%mm6;" /* Mul 4 Y even 00 y6 00 y4 00 y2 00 y0 */\
  88     "pmulhw "Y_COEFF"(%4), %%mm7;" /* Mul 4 Y odd 00 y7 00 y5 00 y3 00 y1 */\
  89 \
  90     /* Do the addition part of the conversion for even and odd pixels,
  91        register usage:
  92        mm0 -> Cblue, mm1 -> Cred, mm2 -> Cgreen even pixels,
  93        mm3 -> Cblue, mm4 -> Cred, mm5 -> Cgreen odd pixels,
  94        mm6 -> Y even, mm7 -> Y odd */\
  95     "movq %%mm0, %%mm3;" /* Copy Cblue */\
  96     "movq %%mm1, %%mm4;" /* Copy Cred */\
  97     "movq %%mm2, %%mm5;" /* Copy Cgreen */\
  98 \
  99     "paddsw %%mm6, %%mm0;" /* Y even + Cblue 00 B6 00 B4 00 B2 00 B0 */\
 100     "paddsw %%mm7, %%mm3;" /* Y odd + Cblue 00 B7 00 B5 00 B3 00 B1 */\
 101 \
 102     "paddsw %%mm6, %%mm1;" /* Y even + Cred 00 R6 00 R4 00 R2 00 R0 */\
 103     "paddsw %%mm7, %%mm4;" /* Y odd + Cred 00 R7 00 R5 00 R3 00 R1 */\
 104 \
 105     "paddsw %%mm6, %%mm2;" /* Y even + Cgreen 00 G6 00 G4 00 G2 00 G0 */\
 106     "paddsw %%mm7, %%mm5;" /* Y odd + Cgreen 00 G7 00 G5 00 G3 00 G1 */\
 107 \
 108     /* Limit RGB even to 0..255 */\
 109     "packuswb %%mm0, %%mm0;" /* B6 B4 B2 B0  B6 B4 B2 B0 */\
 110     "packuswb %%mm1, %%mm1;" /* R6 R4 R2 R0  R6 R4 R2 R0 */\
 111     "packuswb %%mm2, %%mm2;" /* G6 G4 G2 G0  G6 G4 G2 G0 */\
 112 \
 113     /* Limit RGB odd to 0..255 */\
 114     "packuswb %%mm3, %%mm3;" /* B7 B5 B3 B1  B7 B5 B3 B1 */\
 115     "packuswb %%mm4, %%mm4;" /* R7 R5 R3 R1  R7 R5 R3 R1 */\
 116     "packuswb %%mm5, %%mm5;" /* G7 G5 G3 G1  G7 G5 G3 G1 */\
 117 \
 118     /* Interleave RGB even and odd */\
 119     "punpcklbw %%mm3, %%mm0;" /* B7 B6 B5 B4 B3 B2 B1 B0 */\
 120     "punpcklbw %%mm4, %%mm1;" /* R7 R6 R5 R4 R3 R2 R1 R0 */\
 121     "punpcklbw %%mm5, %%mm2;" /* G7 G6 G5 G4 G3 G2 G1 G0 */\
 122
 123
 124 #define YUV422_UNSHIFT                   \
 125     if(c->srcFormat == PIX_FMT_YUV422P){ \
 126         srcStride[1] *= 2;               \
 127         srcStride[2] *= 2;               \
 128     }                                    \
 129
 130 #define YUV2RGB_LOOP(depth)                                   \
 131     h_size= (c->dstW+7)&~7;                                   \
 132     if(h_size*depth > FFABS(dstStride[0])) h_size-=8;         \
 133 \
 134     __asm__ volatile ("pxor %mm4, %mm4;" /* zero mm4 */ );    \
 135     for (y= 0; y<srcSliceH; y++ ) {                           \
 136         uint8_t *image = dst[0] + (y+srcSliceY)*dstStride[0]; \
 137         uint8_t *py = src[0] + y*srcStride[0];                \
 138         uint8_t *pu = src[1] + (y>>1)*srcStride[1];           \
 139         uint8_t *pv = src[2] + (y>>1)*srcStride[2];           \
 140         long index= -h_size/2;                                \
 141
 142 #define YUV2RGB_INIT                                                       \
 143         /* This MMX assembly code deals with a SINGLE scan line at a time, \
 144          * it converts 8 pixels in each iteration. */                      \
 145         __asm__ volatile (                                                 \
 146         /* load data for start of next scan line */                        \
 147         "movd    (%2, %0), %%mm0;" /* Load 4 Cb 00 00 00 00 u3 u2 u1 u0 */ \
 148         "movd    (%3, %0), %%mm1;" /* Load 4 Cr 00 00 00 00 v3 v2 v1 v0 */ \
 149         "movq (%5, %0, 2), %%mm6;" /* Load 8  Y Y7 Y6 Y5 Y4 Y3 Y2 Y1 Y0 */ \
 150         /*                                                                 \
 151         ".balign 16     \n\t"                                              \
 152         */                                                                 \
 153         "1:             \n\t"                                              \
 154         /* No speed difference on my p3@500 with prefetch,                 \
 155          * if it is faster for anyone with -benchmark then tell me.        \
 156         PREFETCH" 64(%0) \n\t"                                             \
 157         PREFETCH" 64(%1) \n\t"                                             \
 158         PREFETCH" 64(%2) \n\t"                                             \
 159         */                                                                 \
 160
 161 #define YUV2RGB_ENDLOOP(depth) \
 162         "add $"AV_STRINGIFY(depth*8)", %1    \n\t" \
 163         "add                       $4, %0    \n\t" \
 164         " js                       1b        \n\t" \
 165 \
 166         : "+r" (index), "+r" (image) \
 167         : "r" (pu - index), "r" (pv - index), "r"(&c->redDither), "r" (py - 2*index) \
 168         ); \
 169     } \
 170     __asm__ volatile (EMMS); \
 171     return srcSliceH; \
 172
 173 static inline int RENAME(yuv420_rgb16)(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
 174                                        int srcSliceH, uint8_t* dst[], int dstStride[]){
 175     int y, h_size;
 176
 177     YUV422_UNSHIFT
 178     YUV2RGB_LOOP(2)
 179
 180         c->blueDither= ff_dither8[y&1];
 181         c->greenDither= ff_dither4[y&1];
 182         c->redDither= ff_dither8[(y+1)&1];
 183
 184         YUV2RGB_INIT
 185         YUV2RGB
 186
 187 #ifdef DITHER1XBPP
 188         "paddusb "BLUE_DITHER"(%4), %%mm0;"
 189         "paddusb "GREEN_DITHER"(%4), %%mm2;"
 190         "paddusb "RED_DITHER"(%4), %%mm1;"
 191 #endif
 192         /* mask unneeded bits off */
 193         "pand "MANGLE(mmx_redmask)", %%mm0;" /* b7b6b5b4 b3_0_0_0 b7b6b5b4 b3_0_0_0 */
 194         "pand "MANGLE(mmx_grnmask)", %%mm2;" /* g7g6g5g4 g3g2_0_0 g7g6g5g4 g3g2_0_0 */
 195         "pand "MANGLE(mmx_redmask)", %%mm1;" /* r7r6r5r4 r3_0_0_0 r7r6r5r4 r3_0_0_0 */
 196
 197         "psrlw   $3, %%mm0;" /* 0_0_0_b7 b6b5b4b3 0_0_0_b7 b6b5b4b3 */
 198         "pxor %%mm4, %%mm4;" /* zero mm4 */
 199
 200         "movq %%mm0, %%mm5;" /* Copy B7-B0 */
 201         "movq %%mm2, %%mm7;" /* Copy G7-G0 */
 202
 203         /* convert RGB24 plane to RGB16 pack for pixel 0-3 */
 204         "punpcklbw %%mm4, %%mm2;" /* 0_0_0_0 0_0_0_0 g7g6g5g4 g3g2_0_0 */
 205         "punpcklbw %%mm1, %%mm0;" /* r7r6r5r4 r3_0_0_0 0_0_0_b7 b6b5b4b3 */
 206
 207         "psllw  $3, %%mm2;" /* 0_0_0_0 0_g7g6g5 g4g3g2_0 0_0_0_0 */
 208         "por %%mm2, %%mm0;" /* r7r6r5r4 r3g7g6g5 g4g3g2b7 b6b5b4b3 */
 209
 210         "movq 8 (%5, %0, 2), %%mm6;" /* Load 8 Y Y7 Y6 Y5 Y4 Y3 Y2 Y1 Y0 */
 211         MOVNTQ "      %%mm0, (%1);" /* store pixel 0-3 */
 212
 213         /* convert RGB24 plane to RGB16 pack for pixel 0-3 */
 214         "punpckhbw %%mm4, %%mm7;" /* 0_0_0_0 0_0_0_0 g7g6g5g4 g3g2_0_0 */
 215         "punpckhbw %%mm1, %%mm5;" /* r7r6r5r4 r3_0_0_0 0_0_0_b7 b6b5b4b3 */
 216
 217         "psllw        $3, %%mm7;" /* 0_0_0_0 0_g7g6g5 g4g3g2_0 0_0_0_0 */
 218         "movd 4 (%2, %0), %%mm0;" /* Load 4 Cb 00 00 00 00 u3 u2 u1 u0 */
 219
 220         "por       %%mm7, %%mm5;" /* r7r6r5r4 r3g7g6g5 g4g3g2b7 b6b5b4b3 */
 221         "movd 4 (%3, %0), %%mm1;" /* Load 4 Cr 00 00 00 00 v3 v2 v1 v0 */
 222
 223         MOVNTQ "   %%mm5, 8 (%1);" /* store pixel 4-7 */
 224
 225     YUV2RGB_ENDLOOP(2)
 226 }
 227
 228 static inline int RENAME(yuv420_rgb15)(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
 229                                        int srcSliceH, uint8_t* dst[], int dstStride[]){
 230     int y, h_size;
 231
 232     YUV422_UNSHIFT
 233     YUV2RGB_LOOP(2)
 234
 235         c->blueDither= ff_dither8[y&1];
 236         c->greenDither= ff_dither8[y&1];
 237         c->redDither= ff_dither8[(y+1)&1];
 238
 239         YUV2RGB_INIT
 240         YUV2RGB
 241
 242 #ifdef DITHER1XBPP
 243         "paddusb "BLUE_DITHER"(%4), %%mm0  \n\t"
 244         "paddusb "GREEN_DITHER"(%4), %%mm2  \n\t"
 245         "paddusb "RED_DITHER"(%4), %%mm1  \n\t"
 246 #endif
 247
 248         /* mask unneeded bits off */
 249         "pand "MANGLE(mmx_redmask)", %%mm0;" /* b7b6b5b4 b3_0_0_0 b7b6b5b4 b3_0_0_0 */
 250         "pand "MANGLE(mmx_redmask)", %%mm2;" /* g7g6g5g4 g3_0_0_0 g7g6g5g4 g3_0_0_0 */
 251         "pand "MANGLE(mmx_redmask)", %%mm1;" /* r7r6r5r4 r3_0_0_0 r7r6r5r4 r3_0_0_0 */
 252
 253         "psrlw   $3, %%mm0;" /* 0_0_0_b7 b6b5b4b3 0_0_0_b7 b6b5b4b3 */
 254         "psrlw   $1, %%mm1;" /* 0_r7r6r5  r4r3_0_0 0_r7r6r5 r4r3_0_0 */
 255         "pxor %%mm4, %%mm4;" /* zero mm4 */
 256
 257         "movq %%mm0, %%mm5;" /* Copy B7-B0 */
 258         "movq %%mm2, %%mm7;" /* Copy G7-G0 */
 259
 260         /* convert RGB24 plane to RGB16 pack for pixel 0-3 */
 261         "punpcklbw %%mm4, %%mm2;" /* 0_0_0_0 0_0_0_0 g7g6g5g4 g3_0_0_0 */
 262         "punpcklbw %%mm1, %%mm0;" /* r7r6r5r4 r3_0_0_0 0_0_0_b7 b6b5b4b3 */
 263
 264         "psllw  $2, %%mm2;" /* 0_0_0_0 0_0_g7g6 g5g4g3_0 0_0_0_0 */
 265         "por %%mm2, %%mm0;" /* 0_r7r6r5 r4r3g7g6 g5g4g3b7 b6b5b4b3 */
 266
 267         "movq 8 (%5, %0, 2), %%mm6;" /* Load 8 Y Y7 Y6 Y5 Y4 Y3 Y2 Y1 Y0 */
 268         MOVNTQ "      %%mm0, (%1);"  /* store pixel 0-3 */
 269
 270         /* convert RGB24 plane to RGB16 pack for pixel 0-3 */
 271         "punpckhbw %%mm4, %%mm7;" /* 0_0_0_0 0_0_0_0 0_g7g6g5 g4g3_0_0 */
 272         "punpckhbw %%mm1, %%mm5;" /* r7r6r5r4 r3_0_0_0 0_0_0_b7 b6b5b4b3 */
 273
 274         "psllw        $2, %%mm7;" /* 0_0_0_0 0_0_g7g6 g5g4g3_0 0_0_0_0 */
 275         "movd 4 (%2, %0), %%mm0;" /* Load 4 Cb 00 00 00 00 u3 u2 u1 u0 */
 276
 277         "por       %%mm7, %%mm5;" /* 0_r7r6r5 r4r3g7g6 g5g4g3b7 b6b5b4b3 */
 278         "movd 4 (%3, %0), %%mm1;" /* Load 4 Cr 00 00 00 00 v3 v2 v1 v0 */
 279
 280         MOVNTQ " %%mm5, 8 (%1);" /* store pixel 4-7 */
 281
 282     YUV2RGB_ENDLOOP(2)
 283 }
 284
 285 static inline int RENAME(yuv420_rgb24)(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
 286                                        int srcSliceH, uint8_t* dst[], int dstStride[]){
 287     int y, h_size;
 288
 289     YUV422_UNSHIFT
 290     YUV2RGB_LOOP(3)
 291
 292         YUV2RGB_INIT
 293         YUV2RGB
 294         /* mm0=B, %%mm2=G, %%mm1=R */
 295 #if HAVE_MMX2
 296         "movq "MANGLE(ff_M24A)", %%mm4     \n\t"
 297         "movq "MANGLE(ff_M24C)", %%mm7     \n\t"
 298         "pshufw $0x50, %%mm0, %%mm5     \n\t" /* B3 B2 B3 B2  B1 B0 B1 B0 */
 299         "pshufw $0x50, %%mm2, %%mm3     \n\t" /* G3 G2 G3 G2  G1 G0 G1 G0 */
 300         "pshufw $0x00, %%mm1, %%mm6     \n\t" /* R1 R0 R1 R0  R1 R0 R1 R0 */
 301
 302         "pand   %%mm4, %%mm5            \n\t" /*    B2        B1       B0 */
 303         "pand   %%mm4, %%mm3            \n\t" /*    G2        G1       G0 */
 304         "pand   %%mm7, %%mm6            \n\t" /*       R1        R0       */
 305
 306         "psllq     $8, %%mm3            \n\t" /* G2        G1       G0    */
 307         "por    %%mm5, %%mm6            \n\t"
 308         "por    %%mm3, %%mm6            \n\t"
 309         MOVNTQ" %%mm6, (%1)             \n\t"
 310
 311         "psrlq     $8, %%mm2            \n\t" /* 00 G7 G6 G5  G4 G3 G2 G1 */
 312         "pshufw $0xA5, %%mm0, %%mm5     \n\t" /* B5 B4 B5 B4  B3 B2 B3 B2 */
 313         "pshufw $0x55, %%mm2, %%mm3     \n\t" /* G4 G3 G4 G3  G4 G3 G4 G3 */
 314         "pshufw $0xA5, %%mm1, %%mm6     \n\t" /* R5 R4 R5 R4  R3 R2 R3 R2 */
 315
 316         "pand "MANGLE(ff_M24B)", %%mm5     \n\t" /* B5       B4        B3    */
 317         "pand          %%mm7, %%mm3     \n\t" /*       G4        G3       */
 318         "pand          %%mm4, %%mm6     \n\t" /*    R4        R3       R2 */
 319
 320         "por    %%mm5, %%mm3            \n\t" /* B5    G4 B4     G3 B3    */
 321         "por    %%mm3, %%mm6            \n\t"
 322         MOVNTQ" %%mm6, 8(%1)            \n\t"
 323
 324         "pshufw $0xFF, %%mm0, %%mm5     \n\t" /* B7 B6 B7 B6  B7 B6 B6 B7 */
 325         "pshufw $0xFA, %%mm2, %%mm3     \n\t" /* 00 G7 00 G7  G6 G5 G6 G5 */
 326         "pshufw $0xFA, %%mm1, %%mm6     \n\t" /* R7 R6 R7 R6  R5 R4 R5 R4 */
 327         "movd 4 (%2, %0), %%mm0;" /* Load 4 Cb 00 00 00 00 u3 u2 u1 u0 */
 328
 329         "pand          %%mm7, %%mm5     \n\t" /*       B7        B6       */
 330         "pand          %%mm4, %%mm3     \n\t" /*    G7        G6       G5 */
 331         "pand "MANGLE(ff_M24B)", %%mm6     \n\t" /* R7       R6        R5    */
 332         "movd 4 (%3, %0), %%mm1;" /* Load 4 Cr 00 00 00 00 v3 v2 v1 v0 */
 333 \
 334         "por          %%mm5, %%mm3      \n\t"
 335         "por          %%mm3, %%mm6      \n\t"
 336         MOVNTQ"       %%mm6, 16(%1)     \n\t"
 337         "movq 8 (%5, %0, 2), %%mm6;" /* Load 8 Y Y7 Y6 Y5 Y4 Y3 Y2 Y1 Y0 */
 338         "pxor         %%mm4, %%mm4      \n\t"
 339
 340 #else
 341
 342         "pxor      %%mm4, %%mm4     \n\t"
 343         "movq      %%mm0, %%mm5     \n\t" /* B */
 344         "movq      %%mm1, %%mm6     \n\t" /* R */
 345         "punpcklbw %%mm2, %%mm0     \n\t" /* GBGBGBGB 0 */
 346         "punpcklbw %%mm4, %%mm1     \n\t" /* 0R0R0R0R 0 */
 347         "punpckhbw %%mm2, %%mm5     \n\t" /* GBGBGBGB 2 */
 348         "punpckhbw %%mm4, %%mm6     \n\t" /* 0R0R0R0R 2 */
 349         "movq      %%mm0, %%mm7     \n\t" /* GBGBGBGB 0 */
 350         "movq      %%mm5, %%mm3     \n\t" /* GBGBGBGB 2 */
 351         "punpcklwd %%mm1, %%mm7     \n\t" /* 0RGB0RGB 0 */
 352         "punpckhwd %%mm1, %%mm0     \n\t" /* 0RGB0RGB 1 */
 353         "punpcklwd %%mm6, %%mm5     \n\t" /* 0RGB0RGB 2 */
 354         "punpckhwd %%mm6, %%mm3     \n\t" /* 0RGB0RGB 3 */
 355
 356         "movq      %%mm7, %%mm2     \n\t" /* 0RGB0RGB 0 */
 357         "movq      %%mm0, %%mm6     \n\t" /* 0RGB0RGB 1 */
 358         "movq      %%mm5, %%mm1     \n\t" /* 0RGB0RGB 2 */
 359         "movq      %%mm3, %%mm4     \n\t" /* 0RGB0RGB 3 */
 360
 361         "psllq       $40, %%mm7     \n\t" /* RGB00000 0 */
 362         "psllq       $40, %%mm0     \n\t" /* RGB00000 1 */
 363         "psllq       $40, %%mm5     \n\t" /* RGB00000 2 */
 364         "psllq       $40, %%mm3     \n\t" /* RGB00000 3 */
 365
 366         "punpckhdq %%mm2, %%mm7     \n\t" /* 0RGBRGB0 0 */
 367         "punpckhdq %%mm6, %%mm0     \n\t" /* 0RGBRGB0 1 */
 368         "punpckhdq %%mm1, %%mm5     \n\t" /* 0RGBRGB0 2 */
 369         "punpckhdq %%mm4, %%mm3     \n\t" /* 0RGBRGB0 3 */
 370
 371         "psrlq        $8, %%mm7     \n\t" /* 00RGBRGB 0 */
 372         "movq      %%mm0, %%mm6     \n\t" /* 0RGBRGB0 1 */
 373         "psllq       $40, %%mm0     \n\t" /* GB000000 1 */
 374         "por       %%mm0, %%mm7     \n\t" /* GBRGBRGB 0 */
 375         MOVNTQ"    %%mm7, (%1)      \n\t"
 376
 377         "movd 4 (%2, %0), %%mm0;" /* Load 4 Cb 00 00 00 00 u3 u2 u1 u0 */
 378
 379         "psrlq       $24, %%mm6     \n\t" /* 0000RGBR 1 */
 380         "movq      %%mm5, %%mm1     \n\t" /* 0RGBRGB0 2 */
 381         "psllq       $24, %%mm5     \n\t" /* BRGB0000 2 */
 382         "por       %%mm5, %%mm6     \n\t" /* BRGBRGBR 1 */
 383         MOVNTQ"    %%mm6, 8(%1)     \n\t"
 384
 385         "movq 8 (%5, %0, 2), %%mm6;" /* Load 8 Y Y7 Y6 Y5 Y4 Y3 Y2 Y1 Y0 */
 386
 387         "psrlq       $40, %%mm1     \n\t" /* 000000RG 2 */
 388         "psllq        $8, %%mm3     \n\t" /* RGBRGB00 3 */
 389         "por       %%mm3, %%mm1     \n\t" /* RGBRGBRG 2 */
 390         MOVNTQ"    %%mm1, 16(%1)    \n\t"
 391
 392         "movd 4 (%3, %0), %%mm1;" /* Load 4 Cr 00 00 00 00 v3 v2 v1 v0 */
 393         "pxor      %%mm4, %%mm4     \n\t"
 394 #endif
 395
 396     YUV2RGB_ENDLOOP(3)
 397 }
 398
 399 #define RGB_PLANAR2PACKED32                                             \
 400     /* convert RGB plane to RGB packed format,                          \
 401        mm0 ->  B, mm1 -> R, mm2 -> G, mm3 -> A,                         \
 402        mm4 -> GB, mm5 -> AR pixel 4-7,                                  \
 403        mm6 -> GB, mm7 -> AR pixel 0-3 */                                \
 404     "movq      %%mm0, %%mm6;"   /* B7 B6 B5 B4 B3 B2 B1 B0 */           \
 405     "movq      %%mm1, %%mm7;"   /* R7 R6 R5 R4 R3 R2 R1 R0 */           \
 406 \
 407     "movq      %%mm0, %%mm4;"   /* B7 B6 B5 B4 B3 B2 B1 B0 */           \
 408     "movq      %%mm1, %%mm5;"   /* R7 R6 R5 R4 R3 R2 R1 R0 */           \
 409 \
 410     "punpcklbw %%mm2, %%mm6;"   /* G3 B3 G2 B2 G1 B1 G0 B0 */           \
 411     "punpcklbw %%mm3, %%mm7;"   /* A3 R3 A2 R2 A1 R1 A0 R0 */           \
 412 \
 413     "punpcklwd %%mm7, %%mm6;"   /* A1 R1 B1 G1 A0 R0 B0 G0 */           \
 414     MOVNTQ "   %%mm6, (%1);"    /* Store ARGB1 ARGB0 */                 \
 415 \
 416     "movq      %%mm0, %%mm6;"   /* B7 B6 B5 B4 B3 B2 B1 B0 */           \
 417     "punpcklbw %%mm2, %%mm6;"   /* G3 B3 G2 B2 G1 B1 G0 B0 */           \
 418 \
 419     "punpckhwd %%mm7, %%mm6;"   /* A3 R3 G3 B3 A2 R2 B3 G2 */           \
 420     MOVNTQ "   %%mm6, 8 (%1);"  /* Store ARGB3 ARGB2 */                 \
 421 \
 422     "punpckhbw %%mm2, %%mm4;"   /* G7 B7 G6 B6 G5 B5 G4 B4 */           \
 423     "punpckhbw %%mm3, %%mm5;"   /* A7 R7 A6 R6 A5 R5 A4 R4 */           \
 424 \
 425     "punpcklwd %%mm5, %%mm4;"   /* A5 R5 B5 G5 A4 R4 B4 G4 */           \
 426     MOVNTQ "   %%mm4, 16 (%1);" /* Store ARGB5 ARGB4 */                 \
 427 \
 428     "movq      %%mm0, %%mm4;"   /* B7 B6 B5 B4 B3 B2 B1 B0 */           \
 429     "punpckhbw %%mm2, %%mm4;"   /* G7 B7 G6 B6 G5 B5 G4 B4 */           \
 430 \
 431     "punpckhwd %%mm5, %%mm4;"   /* A7 R7 G7 B7 A6 R6 B6 G6 */           \
 432     MOVNTQ "   %%mm4, 24 (%1);" /* Store ARGB7 ARGB6 */                 \
 433 \
 434     "movd 4 (%2, %0), %%mm0;"   /* Load 4 Cb 00 00 00 00 u3 u2 u1 u0 */ \
 435     "movd 4 (%3, %0), %%mm1;"   /* Load 4 Cr 00 00 00 00 v3 v2 v1 v0 */ \
 436 \
 437     "pxor         %%mm4, %%mm4;" /* zero mm4 */                         \
 438     "movq 8 (%5, %0, 2), %%mm6;" /* Load 8 Y Y7 Y6 Y5 Y4 Y3 Y2 Y1 Y0 */ \
 439
 440 static inline int RENAME(yuv420_rgb32)(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
 441                                        int srcSliceH, uint8_t* dst[], int dstStride[]){
 442     int y, h_size;
 443
 444     YUV422_UNSHIFT
 445     YUV2RGB_LOOP(4)
 446
 447         YUV2RGB_INIT
 448         YUV2RGB
 449         "pcmpeqd   %%mm3, %%mm3;"   /* fill mm3 */
 450         RGB_PLANAR2PACKED32
 451
 452     YUV2RGB_ENDLOOP(4)
 453 }