Rename y400a to gray8a.
[ffmpeg.git] / libswscale / swscale_template.c
index 0a02437..c269281 100644 (file)
@@ -3,39 +3,32 @@
  *
  * This file is part of FFmpeg.
  *
- * FFmpeg is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
  *
  * FFmpeg is distributed in the hope that it will be useful,
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
  *
- * You should have received a copy of the GNU General Public License
- * along with FFmpeg; if not, write to the Free Software
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
- *
- * The C code (not assembly, MMX, ...) of this file can be used
- * under the LGPL license.
  */
 
 #undef REAL_MOVNTQ
 #undef MOVNTQ
 #undef PAVGB
 #undef PREFETCH
-#undef PREFETCHW
 
 #if COMPILE_TEMPLATE_AMD3DNOW
 #define PREFETCH  "prefetch"
-#define PREFETCHW "prefetchw"
 #elif COMPILE_TEMPLATE_MMX2
 #define PREFETCH "prefetchnta"
-#define PREFETCHW "prefetcht0"
 #else
 #define PREFETCH  " # nop"
-#define PREFETCHW " # nop"
 #endif
 
 #if COMPILE_TEMPLATE_MMX2
@@ -62,7 +55,7 @@
         "movq                             %%mm3, %%mm4      \n\t"\
         "lea                     " offset "(%0), %%"REG_d"  \n\t"\
         "mov                        (%%"REG_d"), %%"REG_S"  \n\t"\
-        ASMALIGN(4) /* FIXME Unroll? */\
+        ".p2align                             4             \n\t" /* FIXME Unroll? */\
         "1:                                                 \n\t"\
         "movq                      8(%%"REG_d"), %%mm0      \n\t" /* filterCoeff */\
         "movq   "  x "(%%"REG_S", %%"REG_a", 2), %%mm2      \n\t" /* srcData */\
@@ -87,7 +80,7 @@
         "mov                        (%%"REG_d"), %%"REG_S"  \n\t"\
         "jb                                  1b             \n\t"\
         :: "r" (&c->redDither),\
-        "r" (dest), "g" (width)\
+        "r" (dest), "g" ((x86_reg)width)\
         : "%"REG_a, "%"REG_d, "%"REG_S\
     );
 
         "pxor                             %%mm6, %%mm6      \n\t"\
         "pxor                             %%mm7, %%mm7      \n\t"\
         "mov                        (%%"REG_d"), %%"REG_S"  \n\t"\
-        ASMALIGN(4) \
+        ".p2align                             4             \n\t"\
         "1:                                                 \n\t"\
         "movq   "  x "(%%"REG_S", %%"REG_a", 2), %%mm0      \n\t" /* srcData */\
         "movq 8+"  x "(%%"REG_S", %%"REG_a", 2), %%mm2      \n\t" /* srcData */\
         "mov                        (%%"REG_d"), %%"REG_S"  \n\t"\
         "jb                                  1b             \n\t"\
         :: "r" (&c->redDither),\
-        "r" (dest), "g" (width)\
+        "r" (dest), "g" ((x86_reg)width)\
         : "%"REG_a, "%"REG_d, "%"REG_S\
     );
 
 #define YSCALEYUV2YV121 \
     "mov %2, %%"REG_a"                    \n\t"\
-    ASMALIGN(4) /* FIXME Unroll? */\
+    ".p2align               4             \n\t" /* FIXME Unroll? */\
     "1:                                   \n\t"\
     "movq  (%0, %%"REG_a", 2), %%mm0      \n\t"\
     "movq 8(%0, %%"REG_a", 2), %%mm1      \n\t"\
     "pcmpeqw %%mm7, %%mm7                 \n\t"\
     "psrlw                 $15, %%mm7     \n\t"\
     "psllw                  $6, %%mm7     \n\t"\
-    ASMALIGN(4) /* FIXME Unroll? */\
+    ".p2align                4            \n\t" /* FIXME Unroll? */\
     "1:                                   \n\t"\
     "movq  (%0, %%"REG_a", 2), %%mm0      \n\t"\
     "movq 8(%0, %%"REG_a", 2), %%mm1      \n\t"\
 /*
     :: "m" (-lumFilterSize), "m" (-chrFilterSize),
        "m" (lumMmxFilter+lumFilterSize*4), "m" (chrMmxFilter+chrFilterSize*4),
-       "r" (dest), "m" (dstW),
+       "r" (dest), "m" (dstW_reg),
        "m" (lumSrc+lumFilterSize), "m" (chrSrc+chrFilterSize)
     : "%eax", "%ebx", "%ecx", "%edx", "%esi"
 */
 #define YSCALEYUV2PACKEDX_UV \
     __asm__ volatile(\
         "xor                   %%"REG_a", %%"REG_a"     \n\t"\
-        ASMALIGN(4)\
+        ".p2align                      4                \n\t"\
         "nop                                            \n\t"\
         "1:                                             \n\t"\
         "lea "CHR_MMX_FILTER_OFFSET"(%0), %%"REG_d"     \n\t"\
         "mov                 (%%"REG_d"), %%"REG_S"     \n\t"\
         "movq      "VROUNDER_OFFSET"(%0), %%mm3         \n\t"\
         "movq                      %%mm3, %%mm4         \n\t"\
-        ASMALIGN(4)\
+        ".p2align                      4                \n\t"\
         "2:                                             \n\t"\
         "movq               8(%%"REG_d"), %%mm0         \n\t" /* filterCoeff */\
         "movq     (%%"REG_S", %%"REG_a"), %%mm2         \n\t" /* UsrcData */\
     "mov                 (%%"REG_d"), %%"REG_S"     \n\t"\
     "movq      "VROUNDER_OFFSET"(%0), "#dst1"       \n\t"\
     "movq                    "#dst1", "#dst2"       \n\t"\
-    ASMALIGN(4)\
+    ".p2align                      4                \n\t"\
     "2:                                             \n\t"\
     "movq               8(%%"REG_d"), "#coeff"      \n\t" /* filterCoeff */\
     "movq  (%%"REG_S", %%"REG_a", 2), "#src1"       \n\t" /* Y1srcData */\
 #define YSCALEYUV2PACKEDX_END                     \
         :: "r" (&c->redDither),                   \
             "m" (dummy), "m" (dummy), "m" (dummy),\
-            "r" (dest), "m" (dstW)                \
+            "r" (dest), "m" (dstW_reg)            \
         : "%"REG_a, "%"REG_d, "%"REG_S            \
     );
 
 #define YSCALEYUV2PACKEDX_ACCURATE_UV \
     __asm__ volatile(\
         "xor %%"REG_a", %%"REG_a"                       \n\t"\
-        ASMALIGN(4)\
+        ".p2align                      4                \n\t"\
         "nop                                            \n\t"\
         "1:                                             \n\t"\
         "lea "CHR_MMX_FILTER_OFFSET"(%0), %%"REG_d"     \n\t"\
         "pxor                      %%mm5, %%mm5         \n\t"\
         "pxor                      %%mm6, %%mm6         \n\t"\
         "pxor                      %%mm7, %%mm7         \n\t"\
-        ASMALIGN(4)\
+        ".p2align                      4                \n\t"\
         "2:                                             \n\t"\
         "movq     (%%"REG_S", %%"REG_a"), %%mm0         \n\t" /* UsrcData */\
         "movq "AV_STRINGIFY(VOF)"(%%"REG_S", %%"REG_a"), %%mm2         \n\t" /* VsrcData */\
     "pxor                      %%mm5, %%mm5         \n\t"\
     "pxor                      %%mm7, %%mm7         \n\t"\
     "pxor                      %%mm6, %%mm6         \n\t"\
-    ASMALIGN(4)\
+    ".p2align                      4                \n\t"\
     "2:                                             \n\t"\
     "movq  (%%"REG_S", %%"REG_a", 2), %%mm0         \n\t" /* Y1srcData */\
     "movq 8(%%"REG_S", %%"REG_a", 2), %%mm2         \n\t" /* Y2srcData */\
     "movq              %%mm0, "CHR_MMX_FILTER_OFFSET"+8("#c") \n\t"\
     "movq              %%mm1, "LUM_MMX_FILTER_OFFSET"+8("#c") \n\t"\
     "xor            "#index", "#index"                        \n\t"\
-    ASMALIGN(4)\
+    ".p2align              4            \n\t"\
     "1:                                 \n\t"\
     "movq     (%2, "#index"), %%mm2     \n\t" /* uvbuf0[eax]*/\
     "movq     (%3, "#index"), %%mm3     \n\t" /* uvbuf1[eax]*/\
 
 #define REAL_YSCALEYUV2RGB_UV(index, c) \
     "xor            "#index", "#index"  \n\t"\
-    ASMALIGN(4)\
+    ".p2align              4            \n\t"\
     "1:                                 \n\t"\
     "movq     (%2, "#index"), %%mm2     \n\t" /* uvbuf0[eax]*/\
     "movq     (%3, "#index"), %%mm3     \n\t" /* uvbuf1[eax]*/\
 
 #define REAL_YSCALEYUV2PACKED1(index, c) \
     "xor            "#index", "#index"  \n\t"\
-    ASMALIGN(4)\
+    ".p2align              4            \n\t"\
     "1:                                 \n\t"\
     "movq     (%2, "#index"), %%mm3     \n\t" /* uvbuf0[eax]*/\
     "movq "AV_STRINGIFY(VOF)"(%2, "#index"), %%mm4     \n\t" /* uvbuf0[eax+2048]*/\
 
 #define REAL_YSCALEYUV2RGB1(index, c) \
     "xor            "#index", "#index"  \n\t"\
-    ASMALIGN(4)\
+    ".p2align              4            \n\t"\
     "1:                                 \n\t"\
     "movq     (%2, "#index"), %%mm3     \n\t" /* uvbuf0[eax]*/\
     "movq "AV_STRINGIFY(VOF)"(%2, "#index"), %%mm4     \n\t" /* uvbuf0[eax+2048]*/\
 
 #define REAL_YSCALEYUV2PACKED1b(index, c) \
     "xor "#index", "#index"             \n\t"\
-    ASMALIGN(4)\
+    ".p2align              4            \n\t"\
     "1:                                 \n\t"\
     "movq     (%2, "#index"), %%mm2     \n\t" /* uvbuf0[eax]*/\
     "movq     (%3, "#index"), %%mm3     \n\t" /* uvbuf1[eax]*/\
 // do vertical chrominance interpolation
 #define REAL_YSCALEYUV2RGB1b(index, c) \
     "xor            "#index", "#index"  \n\t"\
-    ASMALIGN(4)\
+    ".p2align              4            \n\t"\
     "1:                                 \n\t"\
     "movq     (%2, "#index"), %%mm2     \n\t" /* uvbuf0[eax]*/\
     "movq     (%3, "#index"), %%mm3     \n\t" /* uvbuf1[eax]*/\
@@ -958,7 +951,7 @@ static inline void RENAME(yuv2yuv1)(SwsContext *c, const int16_t *lumSrc, const
 #if COMPILE_TEMPLATE_MMX
     if(!(c->flags & SWS_BITEXACT)) {
         long p= 4;
-        uint8_t *src[4]= {alpSrc + dstW, lumSrc + dstW, chrSrc + chrDstW, chrSrc + VOFW + chrDstW};
+        const uint8_t *src[4]= {alpSrc + dstW, lumSrc + dstW, chrSrc + chrDstW, chrSrc + VOFW + chrDstW};
         uint8_t *dst[4]= {aDest, dest, uDest, vDest};
         x86_reg counter[4]= {dstW, dstW, chrDstW, chrDstW};
 
@@ -1032,6 +1025,7 @@ static inline void RENAME(yuv2packedX)(SwsContext *c, const int16_t *lumFilter,
 {
 #if COMPILE_TEMPLATE_MMX
     x86_reg dummy=0;
+    x86_reg dstW_reg = dstW;
     if(!(c->flags & SWS_BITEXACT)) {
         if (c->flags & SWS_ACCURATE_RND) {
             switch(c->dstFormat) {
@@ -1070,7 +1064,7 @@ static inline void RENAME(yuv2packedX)(SwsContext *c, const int16_t *lumFilter,
 
                 :: "r" (&c->redDither),
                 "m" (dummy), "m" (dummy), "m" (dummy),
-                "r" (dest), "m" (dstW)
+                "r" (dest), "m" (dstW_reg)
                 : "%"REG_a, "%"REG_c, "%"REG_d, "%"REG_S
                 );
                 return;
@@ -1144,7 +1138,7 @@ static inline void RENAME(yuv2packedX)(SwsContext *c, const int16_t *lumFilter,
 
                 :: "r" (&c->redDither),
                 "m" (dummy), "m" (dummy), "m" (dummy),
-                "r" (dest),  "m" (dstW)
+                "r" (dest),  "m" (dstW_reg)
                 : "%"REG_a, "%"REG_c, "%"REG_d, "%"REG_S
                 );
                 return;
@@ -1226,21 +1220,21 @@ static inline void RENAME(yuv2packed2)(SwsContext *c, const uint16_t *buf0, cons
             if (CONFIG_SWSCALE_ALPHA && c->alpPixBuf) {
 #if ARCH_X86_64
                 __asm__ volatile(
-                    YSCALEYUV2RGB(%%REGBP, %5)
-                    YSCALEYUV2RGB_YA(%%REGBP, %5, %6, %7)
+                    YSCALEYUV2RGB(%%r8, %5)
+                    YSCALEYUV2RGB_YA(%%r8, %5, %6, %7)
                     "psraw                  $3, %%mm1       \n\t" /* abuf0[eax] - abuf1[eax] >>7*/
                     "psraw                  $3, %%mm7       \n\t" /* abuf0[eax] - abuf1[eax] >>7*/
                     "packuswb            %%mm7, %%mm1       \n\t"
-                    WRITEBGR32(%4, 8280(%5), %%REGBP, %%mm2, %%mm4, %%mm5, %%mm1, %%mm0, %%mm7, %%mm3, %%mm6)
+                    WRITEBGR32(%4, 8280(%5), %%r8, %%mm2, %%mm4, %%mm5, %%mm1, %%mm0, %%mm7, %%mm3, %%mm6)
 
                     :: "c" (buf0), "d" (buf1), "S" (uvbuf0), "D" (uvbuf1), "r" (dest),
                     "a" (&c->redDither)
                     ,"r" (abuf0), "r" (abuf1)
-                    : "%"REG_BP
+                    : "%r8"
                 );
 #else
-                *(uint16_t **)(&c->u_temp)=abuf0;
-                *(uint16_t **)(&c->v_temp)=abuf1;
+                *(const uint16_t **)(&c->u_temp)=abuf0;
+                *(const uint16_t **)(&c->v_temp)=abuf1;
                 __asm__ volatile(
                     "mov %%"REG_b", "ESP_OFFSET"(%5)        \n\t"
                     "mov        %4, %%"REG_b"               \n\t"
@@ -1832,6 +1826,29 @@ static inline void RENAME(nv21ToUV)(uint8_t *dstU, uint8_t *dstV,
     RENAME(nvXXtoUV)(dstV, dstU, src1, width);
 }
 
+// FIXME Maybe dither instead.
+#define YUV_NBPS(depth) \
+static inline void RENAME(yuv ## depth ## ToUV)(uint8_t *dstU, uint8_t *dstV, \
+                                     const uint16_t *srcU, const uint16_t *srcV, \
+                                     long width, uint32_t *unused) \
+{ \
+    int i; \
+    for (i = 0; i < width; i++) { \
+        dstU[i] = srcU[i]>>(depth-8); \
+        dstV[i] = srcV[i]>>(depth-8); \
+    } \
+} \
+\
+static inline void RENAME(yuv ## depth ## ToY)(uint8_t *dstY, const uint16_t *srcY, long width, uint32_t *unused) \
+{ \
+    int i; \
+    for (i = 0; i < width; i++) \
+        dstY[i] = srcY[i]>>(depth-8); \
+} \
+
+YUV_NBPS( 9)
+YUV_NBPS(10)
+
 #if COMPILE_TEMPLATE_MMX
 static inline void RENAME(bgr24ToY_mmx)(uint8_t *dst, const uint8_t *src, long width, enum PixelFormat srcFormat)
 {
@@ -1889,7 +1906,7 @@ static inline void RENAME(bgr24ToY_mmx)(uint8_t *dst, const uint8_t *src, long w
 static inline void RENAME(bgr24ToUV_mmx)(uint8_t *dstU, uint8_t *dstV, const uint8_t *src, long width, enum PixelFormat srcFormat)
 {
     __asm__ volatile(
-        "movq                    24+%4, %%mm6       \n\t"
+        "movq                   24(%4), %%mm6       \n\t"
         "mov                        %3, %%"REG_a"   \n\t"
         "pxor                    %%mm7, %%mm7       \n\t"
         "1:                                         \n\t"
@@ -1900,9 +1917,9 @@ static inline void RENAME(bgr24ToUV_mmx)(uint8_t *dstU, uint8_t *dstV, const uin
         "punpcklbw               %%mm7, %%mm1       \n\t"
         "movq                    %%mm0, %%mm2       \n\t"
         "movq                    %%mm1, %%mm3       \n\t"
-        "pmaddwd                    %4, %%mm0       \n\t"
-        "pmaddwd                  8+%4, %%mm1       \n\t"
-        "pmaddwd                 16+%4, %%mm2       \n\t"
+        "pmaddwd                  (%4), %%mm0       \n\t"
+        "pmaddwd                 8(%4), %%mm1       \n\t"
+        "pmaddwd                16(%4), %%mm2       \n\t"
         "pmaddwd                 %%mm6, %%mm3       \n\t"
         "paddd                   %%mm1, %%mm0       \n\t"
         "paddd                   %%mm3, %%mm2       \n\t"
@@ -1914,9 +1931,9 @@ static inline void RENAME(bgr24ToUV_mmx)(uint8_t *dstU, uint8_t *dstV, const uin
         "punpcklbw               %%mm7, %%mm3       \n\t"
         "movq                    %%mm1, %%mm4       \n\t"
         "movq                    %%mm3, %%mm5       \n\t"
-        "pmaddwd                    %4, %%mm1       \n\t"
-        "pmaddwd                  8+%4, %%mm3       \n\t"
-        "pmaddwd                 16+%4, %%mm4       \n\t"
+        "pmaddwd                  (%4), %%mm1       \n\t"
+        "pmaddwd                 8(%4), %%mm3       \n\t"
+        "pmaddwd                16(%4), %%mm4       \n\t"
         "pmaddwd                 %%mm6, %%mm5       \n\t"
         "paddd                   %%mm3, %%mm1       \n\t"
         "paddd                   %%mm5, %%mm4       \n\t"
@@ -1939,7 +1956,7 @@ static inline void RENAME(bgr24ToUV_mmx)(uint8_t *dstU, uint8_t *dstV, const uin
         "add                        $4, %%"REG_a"   \n\t"
         " js                        1b              \n\t"
     : "+r" (src)
-    : "r" (dstU+width), "r" (dstV+width), "g" ((x86_reg)-width), "m"(ff_bgr24toUV[srcFormat == PIX_FMT_RGB24][0])
+    : "r" (dstU+width), "r" (dstV+width), "g" ((x86_reg)-width), "r"(ff_bgr24toUV[srcFormat == PIX_FMT_RGB24])
     : "%"REG_a
     );
 }
@@ -2061,7 +2078,7 @@ static inline void RENAME(hScale)(int16_t *dst, int dstW, const uint8_t *src, in
             "pxor                %%mm7, %%mm7       \n\t"
             "push           %%"REG_BP"              \n\t" // we use 7 regs here ...
             "mov             %%"REG_a", %%"REG_BP"  \n\t"
-            ASMALIGN(4)
+            ".p2align                4              \n\t"
             "1:                                     \n\t"
             "movzwl   (%2, %%"REG_BP"), %%eax       \n\t"
             "movzwl  2(%2, %%"REG_BP"), %%ebx       \n\t"
@@ -2105,7 +2122,7 @@ static inline void RENAME(hScale)(int16_t *dst, int dstW, const uint8_t *src, in
             "pxor                 %%mm7, %%mm7      \n\t"
             "push            %%"REG_BP"             \n\t" // we use 7 regs here ...
             "mov              %%"REG_a", %%"REG_BP" \n\t"
-            ASMALIGN(4)
+            ".p2align                 4             \n\t"
             "1:                                     \n\t"
             "movzwl    (%2, %%"REG_BP"), %%eax      \n\t"
             "movzwl   2(%2, %%"REG_BP"), %%ebx      \n\t"
@@ -2149,14 +2166,14 @@ static inline void RENAME(hScale)(int16_t *dst, int dstW, const uint8_t *src, in
 #endif
         );
     } else {
-        uint8_t *offset = src+filterSize;
+        const uint8_t *offset = src+filterSize;
         x86_reg counter= -2*dstW;
         //filter-= counter*filterSize/2;
         filterPos-= counter/2;
         dst-= counter/2;
         __asm__ volatile(
             "pxor                  %%mm7, %%mm7     \n\t"
-            ASMALIGN(4)
+            ".p2align                  4            \n\t"
             "1:                                     \n\t"
             "mov                      %2, %%"REG_c" \n\t"
             "movzwl      (%%"REG_c", %0), %%eax     \n\t"
@@ -2215,10 +2232,41 @@ static inline void RENAME(hScale)(int16_t *dst, int dstW, const uint8_t *src, in
         dst[i] = FFMIN(val>>7, (1<<15)-1); // the cubic equation does overflow ...
         //dst[i] = val>>7;
     }
-#endif /* COMPILE_ALTIVEC */
+#endif /* COMPILE_TEMPLATE_ALTIVEC */
 #endif /* COMPILE_MMX */
 }
 
+//FIXME all pal and rgb srcFormats could do this convertion as well
+//FIXME all scalers more complex than bilinear could do half of this transform
+static void RENAME(chrRangeToJpeg)(int16_t *dst, int width)
+{
+    int i;
+    for (i = 0; i < width; i++) {
+        dst[i     ] = (FFMIN(dst[i     ],30775)*4663 - 9289992)>>12; //-264
+        dst[i+VOFW] = (FFMIN(dst[i+VOFW],30775)*4663 - 9289992)>>12; //-264
+    }
+}
+static void RENAME(chrRangeFromJpeg)(int16_t *dst, int width)
+{
+    int i;
+    for (i = 0; i < width; i++) {
+        dst[i     ] = (dst[i     ]*1799 + 4081085)>>11; //1469
+        dst[i+VOFW] = (dst[i+VOFW]*1799 + 4081085)>>11; //1469
+    }
+}
+static void RENAME(lumRangeToJpeg)(int16_t *dst, int width)
+{
+    int i;
+    for (i = 0; i < width; i++)
+        dst[i] = (FFMIN(dst[i],30189)*19077 - 39057361)>>14;
+}
+static void RENAME(lumRangeFromJpeg)(int16_t *dst, int width)
+{
+    int i;
+    for (i = 0; i < width; i++)
+        dst[i] = (dst[i]*14071 + 33561947)>>14;
+}
+
 #define FAST_BILINEAR_X86 \
     "subl    %%edi, %%esi    \n\t" /*  src[xx+1] - src[xx] */                   \
     "imull   %%ecx, %%esi    \n\t" /* (src[xx+1] - src[xx])*xalpha */           \
@@ -2228,9 +2276,117 @@ static inline void RENAME(hScale)(int16_t *dst, int dstW, const uint8_t *src, in
     "shrl       $9, %%esi    \n\t"                                              \
 
 static inline void RENAME(hyscale_fast)(SwsContext *c, int16_t *dst,
-                                        int dstWidth, const uint8_t *src, int srcW,
+                                        long dstWidth, const uint8_t *src, int srcW,
                                         int xInc)
 {
+#if ARCH_X86
+#if COMPILE_TEMPLATE_MMX2
+    int32_t *filterPos = c->hLumFilterPos;
+    int16_t *filter    = c->hLumFilter;
+    int     canMMX2BeUsed  = c->canMMX2BeUsed;
+    void    *mmx2FilterCode= c->lumMmx2FilterCode;
+    int i;
+#if defined(PIC)
+    DECLARE_ALIGNED(8, uint64_t, ebxsave);
+#endif
+    if (canMMX2BeUsed) {
+        __asm__ volatile(
+#if defined(PIC)
+            "mov               %%"REG_b", %5        \n\t"
+#endif
+            "pxor                  %%mm7, %%mm7     \n\t"
+            "mov                      %0, %%"REG_c" \n\t"
+            "mov                      %1, %%"REG_D" \n\t"
+            "mov                      %2, %%"REG_d" \n\t"
+            "mov                      %3, %%"REG_b" \n\t"
+            "xor               %%"REG_a", %%"REG_a" \n\t" // i
+            PREFETCH"        (%%"REG_c")            \n\t"
+            PREFETCH"      32(%%"REG_c")            \n\t"
+            PREFETCH"      64(%%"REG_c")            \n\t"
+
+#if ARCH_X86_64
+
+#define CALL_MMX2_FILTER_CODE \
+            "movl            (%%"REG_b"), %%esi     \n\t"\
+            "call                    *%4            \n\t"\
+            "movl (%%"REG_b", %%"REG_a"), %%esi     \n\t"\
+            "add               %%"REG_S", %%"REG_c" \n\t"\
+            "add               %%"REG_a", %%"REG_D" \n\t"\
+            "xor               %%"REG_a", %%"REG_a" \n\t"\
+
+#else
+
+#define CALL_MMX2_FILTER_CODE \
+            "movl (%%"REG_b"), %%esi        \n\t"\
+            "call         *%4                       \n\t"\
+            "addl (%%"REG_b", %%"REG_a"), %%"REG_c" \n\t"\
+            "add               %%"REG_a", %%"REG_D" \n\t"\
+            "xor               %%"REG_a", %%"REG_a" \n\t"\
+
+#endif /* ARCH_X86_64 */
+
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
+
+#if defined(PIC)
+            "mov                      %5, %%"REG_b" \n\t"
+#endif
+            :: "m" (src), "m" (dst), "m" (filter), "m" (filterPos),
+            "m" (mmx2FilterCode)
+#if defined(PIC)
+            ,"m" (ebxsave)
+#endif
+            : "%"REG_a, "%"REG_c, "%"REG_d, "%"REG_S, "%"REG_D
+#if !defined(PIC)
+            ,"%"REG_b
+#endif
+        );
+        for (i=dstWidth-1; (i*xInc)>>16 >=srcW-1; i--) dst[i] = src[srcW-1]*128;
+    } else {
+#endif /* COMPILE_TEMPLATE_MMX2 */
+    x86_reg xInc_shr16 = xInc >> 16;
+    uint16_t xInc_mask = xInc & 0xffff;
+    x86_reg dstWidth_reg = dstWidth;
+    //NO MMX just normal asm ...
+    __asm__ volatile(
+        "xor %%"REG_a", %%"REG_a"            \n\t" // i
+        "xor %%"REG_d", %%"REG_d"            \n\t" // xx
+        "xorl    %%ecx, %%ecx                \n\t" // xalpha
+        ".p2align    4                       \n\t"
+        "1:                                  \n\t"
+        "movzbl    (%0, %%"REG_d"), %%edi    \n\t" //src[xx]
+        "movzbl   1(%0, %%"REG_d"), %%esi    \n\t" //src[xx+1]
+        FAST_BILINEAR_X86
+        "movw     %%si, (%%"REG_D", %%"REG_a", 2)   \n\t"
+        "addw       %4, %%cx                 \n\t" //xalpha += xInc&0xFFFF
+        "adc        %3, %%"REG_d"            \n\t" //xx+= xInc>>16 + carry
+
+        "movzbl    (%0, %%"REG_d"), %%edi    \n\t" //src[xx]
+        "movzbl   1(%0, %%"REG_d"), %%esi    \n\t" //src[xx+1]
+        FAST_BILINEAR_X86
+        "movw     %%si, 2(%%"REG_D", %%"REG_a", 2)  \n\t"
+        "addw       %4, %%cx                 \n\t" //xalpha += xInc&0xFFFF
+        "adc        %3, %%"REG_d"            \n\t" //xx+= xInc>>16 + carry
+
+
+        "add        $2, %%"REG_a"            \n\t"
+        "cmp        %2, %%"REG_a"            \n\t"
+        " jb        1b                       \n\t"
+
+
+        :: "r" (src), "m" (dst), "m" (dstWidth_reg), "m" (xInc_shr16), "m" (xInc_mask)
+        : "%"REG_a, "%"REG_d, "%ecx", "%"REG_D, "%esi"
+    );
+#if COMPILE_TEMPLATE_MMX2
+    } //if MMX2 can't be used
+#endif
+#else
     int i;
     unsigned int xpos=0;
     for (i=0;i<dstWidth;i++) {
@@ -2239,162 +2395,142 @@ static inline void RENAME(hyscale_fast)(SwsContext *c, int16_t *dst,
         dst[i]= (src[xx]<<7) + (src[xx+1] - src[xx])*xalpha;
         xpos+=xInc;
     }
+#endif /* ARCH_X86 */
 }
 
       // *** horizontal scale Y line to temp buffer
 static inline void RENAME(hyscale)(SwsContext *c, uint16_t *dst, long dstWidth, const uint8_t *src, int srcW, int xInc,
-                                   int flags, const int16_t *hLumFilter,
+                                   const int16_t *hLumFilter,
                                    const int16_t *hLumFilterPos, int hLumFilterSize,
-                                   enum PixelFormat srcFormat, uint8_t *formatConvBuffer,
+                                   uint8_t *formatConvBuffer,
                                    uint32_t *pal, int isAlpha)
 {
-    int32_t av_unused *mmx2FilterPos = c->lumMmx2FilterPos;
-    int16_t av_unused *mmx2Filter    = c->lumMmx2Filter;
-    int     av_unused canMMX2BeUsed  = c->canMMX2BeUsed;
-    void    av_unused *mmx2FilterCode= c->lumMmx2FilterCode;
-    void (*internal_func)(uint8_t *, const uint8_t *, long, uint32_t *) = isAlpha ? c->hascale_internal : c->hyscale_internal;
+    void (*toYV12)(uint8_t *, const uint8_t *, long, uint32_t *) = isAlpha ? c->alpToYV12 : c->lumToYV12;
+    void (*convertRange)(uint16_t *, int) = isAlpha ? NULL : c->lumConvertRange;
 
     src += isAlpha ? c->alpSrcOffset : c->lumSrcOffset;
 
-    if (internal_func) {
-        internal_func(formatConvBuffer, src, srcW, pal);
+    if (toYV12) {
+        toYV12(formatConvBuffer, src, srcW, pal);
         src= formatConvBuffer;
     }
 
-#if COMPILE_TEMPLATE_MMX
-    // Use the new MMX scaler if the MMX2 one can't be used (it is faster than the x86 ASM one).
-    if (!(flags&SWS_FAST_BILINEAR) || (!canMMX2BeUsed))
-#else
-    if (!(flags&SWS_FAST_BILINEAR))
-#endif
-    {
+    if (!c->hyscale_fast) {
         c->hScale(dst, dstWidth, src, srcW, xInc, hLumFilter, hLumFilterPos, hLumFilterSize);
     } else { // fast bilinear upscale / crap downscale
-#if ARCH_X86 && CONFIG_GPL
+        c->hyscale_fast(c, dst, dstWidth, src, srcW, xInc);
+    }
+
+    if (convertRange)
+        convertRange(dst, dstWidth);
+}
+
+static inline void RENAME(hcscale_fast)(SwsContext *c, int16_t *dst,
+                                        long dstWidth, const uint8_t *src1,
+                                        const uint8_t *src2, int srcW, int xInc)
+{
+#if ARCH_X86
 #if COMPILE_TEMPLATE_MMX2
-        int i;
+    int32_t *filterPos = c->hChrFilterPos;
+    int16_t *filter    = c->hChrFilter;
+    int     canMMX2BeUsed  = c->canMMX2BeUsed;
+    void    *mmx2FilterCode= c->chrMmx2FilterCode;
+    int i;
 #if defined(PIC)
-        DECLARE_ALIGNED(8, uint64_t, ebxsave);
+    DECLARE_ALIGNED(8, uint64_t, ebxsave);
 #endif
-        if (canMMX2BeUsed) {
-            __asm__ volatile(
+    if (canMMX2BeUsed) {
+        __asm__ volatile(
 #if defined(PIC)
-                "mov               %%"REG_b", %5        \n\t"
+            "mov          %%"REG_b", %6         \n\t"
 #endif
-                "pxor                  %%mm7, %%mm7     \n\t"
-                "mov                      %0, %%"REG_c" \n\t"
-                "mov                      %1, %%"REG_D" \n\t"
-                "mov                      %2, %%"REG_d" \n\t"
-                "mov                      %3, %%"REG_b" \n\t"
-                "xor               %%"REG_a", %%"REG_a" \n\t" // i
-                PREFETCH"        (%%"REG_c")            \n\t"
-                PREFETCH"      32(%%"REG_c")            \n\t"
-                PREFETCH"      64(%%"REG_c")            \n\t"
-
-#if ARCH_X86_64
-
-#define CALL_MMX2_FILTER_CODE \
-                "movl            (%%"REG_b"), %%esi     \n\t"\
-                "call                    *%4            \n\t"\
-                "movl (%%"REG_b", %%"REG_a"), %%esi     \n\t"\
-                "add               %%"REG_S", %%"REG_c" \n\t"\
-                "add               %%"REG_a", %%"REG_D" \n\t"\
-                "xor               %%"REG_a", %%"REG_a" \n\t"\
-
-#else
-
-#define CALL_MMX2_FILTER_CODE \
-                "movl (%%"REG_b"), %%esi        \n\t"\
-                "call         *%4                       \n\t"\
-                "addl (%%"REG_b", %%"REG_a"), %%"REG_c" \n\t"\
-                "add               %%"REG_a", %%"REG_D" \n\t"\
-                "xor               %%"REG_a", %%"REG_a" \n\t"\
-
-#endif /* ARCH_X86_64 */
-
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
+            "pxor             %%mm7, %%mm7      \n\t"
+            "mov                 %0, %%"REG_c"  \n\t"
+            "mov                 %1, %%"REG_D"  \n\t"
+            "mov                 %2, %%"REG_d"  \n\t"
+            "mov                 %3, %%"REG_b"  \n\t"
+            "xor          %%"REG_a", %%"REG_a"  \n\t" // i
+            PREFETCH"   (%%"REG_c")             \n\t"
+            PREFETCH" 32(%%"REG_c")             \n\t"
+            PREFETCH" 64(%%"REG_c")             \n\t"
+
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
+            "xor          %%"REG_a", %%"REG_a"  \n\t" // i
+            "mov                 %5, %%"REG_c"  \n\t" // src
+            "mov                 %1, %%"REG_D"  \n\t" // buf1
+            "add              $"AV_STRINGIFY(VOF)", %%"REG_D"  \n\t"
+            PREFETCH"   (%%"REG_c")             \n\t"
+            PREFETCH" 32(%%"REG_c")             \n\t"
+            PREFETCH" 64(%%"REG_c")             \n\t"
+
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
+            CALL_MMX2_FILTER_CODE
 
 #if defined(PIC)
-                "mov                      %5, %%"REG_b" \n\t"
+            "mov %6, %%"REG_b"    \n\t"
 #endif
-                :: "m" (src), "m" (dst), "m" (mmx2Filter), "m" (mmx2FilterPos),
-                "m" (mmx2FilterCode)
+            :: "m" (src1), "m" (dst), "m" (filter), "m" (filterPos),
+            "m" (mmx2FilterCode), "m" (src2)
 #if defined(PIC)
-                ,"m" (ebxsave)
+            ,"m" (ebxsave)
 #endif
-                : "%"REG_a, "%"REG_c, "%"REG_d, "%"REG_S, "%"REG_D
+            : "%"REG_a, "%"REG_c, "%"REG_d, "%"REG_S, "%"REG_D
 #if !defined(PIC)
-                ,"%"REG_b
+            ,"%"REG_b
 #endif
-            );
-            for (i=dstWidth-1; (i*xInc)>>16 >=srcW-1; i--) dst[i] = src[srcW-1]*128;
-        } else {
+        );
+        for (i=dstWidth-1; (i*xInc)>>16 >=srcW-1; i--) {
+            //printf("%d %d %d\n", dstWidth, i, srcW);
+            dst[i] = src1[srcW-1]*128;
+            dst[i+VOFW] = src2[srcW-1]*128;
+        }
+    } else {
 #endif /* COMPILE_TEMPLATE_MMX2 */
-        x86_reg xInc_shr16 = xInc >> 16;
+        x86_reg xInc_shr16 = (x86_reg) (xInc >> 16);
         uint16_t xInc_mask = xInc & 0xffff;
-        //NO MMX just normal asm ...
+        x86_reg dstWidth_reg = dstWidth;
         __asm__ volatile(
-            "xor %%"REG_a", %%"REG_a"            \n\t" // i
-            "xor %%"REG_d", %%"REG_d"            \n\t" // xx
-            "xorl    %%ecx, %%ecx                \n\t" // xalpha
-            ASMALIGN(4)
-            "1:                                  \n\t"
-            "movzbl    (%0, %%"REG_d"), %%edi    \n\t" //src[xx]
-            "movzbl   1(%0, %%"REG_d"), %%esi    \n\t" //src[xx+1]
+            "xor %%"REG_a", %%"REG_a"               \n\t" // i
+            "xor %%"REG_d", %%"REG_d"               \n\t" // xx
+            "xorl    %%ecx, %%ecx                   \n\t" // xalpha
+            ".p2align    4                          \n\t"
+            "1:                                     \n\t"
+            "mov        %0, %%"REG_S"               \n\t"
+            "movzbl  (%%"REG_S", %%"REG_d"), %%edi  \n\t" //src[xx]
+            "movzbl 1(%%"REG_S", %%"REG_d"), %%esi  \n\t" //src[xx+1]
             FAST_BILINEAR_X86
             "movw     %%si, (%%"REG_D", %%"REG_a", 2)   \n\t"
-            "addw       %4, %%cx                 \n\t" //xalpha += xInc&0xFFFF
-            "adc        %3, %%"REG_d"            \n\t" //xx+= xInc>>16 + carry
 
-            "movzbl    (%0, %%"REG_d"), %%edi    \n\t" //src[xx]
-            "movzbl   1(%0, %%"REG_d"), %%esi    \n\t" //src[xx+1]
+            "movzbl    (%5, %%"REG_d"), %%edi       \n\t" //src[xx]
+            "movzbl   1(%5, %%"REG_d"), %%esi       \n\t" //src[xx+1]
             FAST_BILINEAR_X86
-            "movw     %%si, 2(%%"REG_D", %%"REG_a", 2)  \n\t"
-            "addw       %4, %%cx                 \n\t" //xalpha += xInc&0xFFFF
-            "adc        %3, %%"REG_d"            \n\t" //xx+= xInc>>16 + carry
+            "movw     %%si, "AV_STRINGIFY(VOF)"(%%"REG_D", %%"REG_a", 2)   \n\t"
 
+            "addw       %4, %%cx                    \n\t" //xalpha += xInc&0xFFFF
+            "adc        %3, %%"REG_d"               \n\t" //xx+= xInc>>16 + carry
+            "add        $1, %%"REG_a"               \n\t"
+            "cmp        %2, %%"REG_a"               \n\t"
+            " jb        1b                          \n\t"
 
-            "add        $2, %%"REG_a"            \n\t"
-            "cmp        %2, %%"REG_a"            \n\t"
-            " jb        1b                       \n\t"
-
-
-            :: "r" (src), "m" (dst), "m" (dstWidth), "m" (xInc_shr16), "m" (xInc_mask)
+/* GCC 3.3 makes MPlayer crash on IA-32 machines when using "g" operand here,
+which is needed to support GCC 4.0. */
+#if ARCH_X86_64 && AV_GCC_VERSION_AT_LEAST(3,4)
+            :: "m" (src1), "m" (dst), "g" (dstWidth_reg), "m" (xInc_shr16), "m" (xInc_mask),
+#else
+            :: "m" (src1), "m" (dst), "m" (dstWidth_reg), "m" (xInc_shr16), "m" (xInc_mask),
+#endif
+            "r" (src2)
             : "%"REG_a, "%"REG_d, "%ecx", "%"REG_D, "%esi"
         );
 #if COMPILE_TEMPLATE_MMX2
-        } //if MMX2 can't be used
+    } //if MMX2 can't be used
 #endif
 #else
-        c->hyscale_fast(c, dst, dstWidth, src, srcW, xInc);
-#endif /* ARCH_X86 */
-    }
-
-    if(!isAlpha && c->srcRange != c->dstRange && !(isRGB(c->dstFormat) || isBGR(c->dstFormat))) {
-        int i;
-        //FIXME all pal and rgb srcFormats could do this convertion as well
-        //FIXME all scalers more complex than bilinear could do half of this transform
-        if(c->srcRange) {
-            for (i=0; i<dstWidth; i++)
-                dst[i]= (dst[i]*14071 + 33561947)>>14;
-        } else {
-            for (i=0; i<dstWidth; i++)
-                dst[i]= (FFMIN(dst[i],30189)*19077 - 39057361)>>14;
-        }
-    }
-}
-
-static inline void RENAME(hcscale_fast)(SwsContext *c, int16_t *dst,
-                                        int dstWidth, const uint8_t *src1,
-                                        const uint8_t *src2, int srcW, int xInc)
-{
     int i;
     unsigned int xpos=0;
     for (i=0;i<dstWidth;i++) {
@@ -2408,163 +2544,40 @@ static inline void RENAME(hcscale_fast)(SwsContext *c, int16_t *dst,
         */
         xpos+=xInc;
     }
+#endif /* ARCH_X86 */
 }
 
 inline static void RENAME(hcscale)(SwsContext *c, uint16_t *dst, long dstWidth, const uint8_t *src1, const uint8_t *src2,
-                                   int srcW, int xInc, int flags, const int16_t *hChrFilter,
+                                   int srcW, int xInc, const int16_t *hChrFilter,
                                    const int16_t *hChrFilterPos, int hChrFilterSize,
-                                   enum PixelFormat srcFormat, uint8_t *formatConvBuffer,
+                                   uint8_t *formatConvBuffer,
                                    uint32_t *pal)
 {
-    int32_t av_unused *mmx2FilterPos = c->chrMmx2FilterPos;
-    int16_t av_unused *mmx2Filter    = c->chrMmx2Filter;
-    int     av_unused canMMX2BeUsed  = c->canMMX2BeUsed;
-    void    av_unused *mmx2FilterCode= c->chrMmx2FilterCode;
-
-    if (isGray(srcFormat) || srcFormat==PIX_FMT_MONOBLACK || srcFormat==PIX_FMT_MONOWHITE)
-        return;
 
     src1 += c->chrSrcOffset;
     src2 += c->chrSrcOffset;
 
-    if (c->hcscale_internal) {
-        c->hcscale_internal(formatConvBuffer, formatConvBuffer+VOFW, src1, src2, srcW, pal);
+    if (c->chrToYV12) {
+        c->chrToYV12(formatConvBuffer, formatConvBuffer+VOFW, src1, src2, srcW, pal);
         src1= formatConvBuffer;
         src2= formatConvBuffer+VOFW;
     }
 
-#if COMPILE_TEMPLATE_MMX
-    // Use the new MMX scaler if the MMX2 one can't be used (it is faster than the x86 ASM one).
-    if (!(flags&SWS_FAST_BILINEAR) || (!canMMX2BeUsed))
-#else
-    if (!(flags&SWS_FAST_BILINEAR))
-#endif
-    {
+    if (!c->hcscale_fast) {
         c->hScale(dst     , dstWidth, src1, srcW, xInc, hChrFilter, hChrFilterPos, hChrFilterSize);
         c->hScale(dst+VOFW, dstWidth, src2, srcW, xInc, hChrFilter, hChrFilterPos, hChrFilterSize);
     } else { // fast bilinear upscale / crap downscale
-#if ARCH_X86 && CONFIG_GPL
-#if COMPILE_TEMPLATE_MMX2
-        int i;
-#if defined(PIC)
-        DECLARE_ALIGNED(8, uint64_t, ebxsave);
-#endif
-        if (canMMX2BeUsed) {
-            __asm__ volatile(
-#if defined(PIC)
-                "mov          %%"REG_b", %6         \n\t"
-#endif
-                "pxor             %%mm7, %%mm7      \n\t"
-                "mov                 %0, %%"REG_c"  \n\t"
-                "mov                 %1, %%"REG_D"  \n\t"
-                "mov                 %2, %%"REG_d"  \n\t"
-                "mov                 %3, %%"REG_b"  \n\t"
-                "xor          %%"REG_a", %%"REG_a"  \n\t" // i
-                PREFETCH"   (%%"REG_c")             \n\t"
-                PREFETCH" 32(%%"REG_c")             \n\t"
-                PREFETCH" 64(%%"REG_c")             \n\t"
-
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
-                "xor          %%"REG_a", %%"REG_a"  \n\t" // i
-                "mov                 %5, %%"REG_c"  \n\t" // src
-                "mov                 %1, %%"REG_D"  \n\t" // buf1
-                "add              $"AV_STRINGIFY(VOF)", %%"REG_D"  \n\t"
-                PREFETCH"   (%%"REG_c")             \n\t"
-                PREFETCH" 32(%%"REG_c")             \n\t"
-                PREFETCH" 64(%%"REG_c")             \n\t"
-
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
-                CALL_MMX2_FILTER_CODE
-
-#if defined(PIC)
-                "mov %6, %%"REG_b"    \n\t"
-#endif
-                :: "m" (src1), "m" (dst), "m" (mmx2Filter), "m" (mmx2FilterPos),
-                "m" (mmx2FilterCode), "m" (src2)
-#if defined(PIC)
-                ,"m" (ebxsave)
-#endif
-                : "%"REG_a, "%"REG_c, "%"REG_d, "%"REG_S, "%"REG_D
-#if !defined(PIC)
-                ,"%"REG_b
-#endif
-            );
-            for (i=dstWidth-1; (i*xInc)>>16 >=srcW-1; i--) {
-                //printf("%d %d %d\n", dstWidth, i, srcW);
-                dst[i] = src1[srcW-1]*128;
-                dst[i+VOFW] = src2[srcW-1]*128;
-            }
-        } else {
-#endif /* COMPILE_TEMPLATE_MMX2 */
-            x86_reg xInc_shr16 = (x86_reg) (xInc >> 16);
-            uint16_t xInc_mask = xInc & 0xffff;
-            __asm__ volatile(
-                "xor %%"REG_a", %%"REG_a"               \n\t" // i
-                "xor %%"REG_d", %%"REG_d"               \n\t" // xx
-                "xorl    %%ecx, %%ecx                   \n\t" // xalpha
-                ASMALIGN(4)
-                "1:                                     \n\t"
-                "mov        %0, %%"REG_S"               \n\t"
-                "movzbl  (%%"REG_S", %%"REG_d"), %%edi  \n\t" //src[xx]
-                "movzbl 1(%%"REG_S", %%"REG_d"), %%esi  \n\t" //src[xx+1]
-                FAST_BILINEAR_X86
-                "movw     %%si, (%%"REG_D", %%"REG_a", 2)   \n\t"
-
-                "movzbl    (%5, %%"REG_d"), %%edi       \n\t" //src[xx]
-                "movzbl   1(%5, %%"REG_d"), %%esi       \n\t" //src[xx+1]
-                FAST_BILINEAR_X86
-                "movw     %%si, "AV_STRINGIFY(VOF)"(%%"REG_D", %%"REG_a", 2)   \n\t"
-
-                "addw       %4, %%cx                    \n\t" //xalpha += xInc&0xFFFF
-                "adc        %3, %%"REG_d"               \n\t" //xx+= xInc>>16 + carry
-                "add        $1, %%"REG_a"               \n\t"
-                "cmp        %2, %%"REG_a"               \n\t"
-                " jb        1b                          \n\t"
-
-/* GCC 3.3 makes MPlayer crash on IA-32 machines when using "g" operand here,
-   which is needed to support GCC 4.0. */
-#if ARCH_X86_64 && AV_GCC_VERSION_AT_LEAST(3,4)
-                :: "m" (src1), "m" (dst), "g" (dstWidth), "m" (xInc_shr16), "m" (xInc_mask),
-#else
-                :: "m" (src1), "m" (dst), "m" (dstWidth), "m" (xInc_shr16), "m" (xInc_mask),
-#endif
-                "r" (src2)
-                : "%"REG_a, "%"REG_d, "%ecx", "%"REG_D, "%esi"
-            );
-#if COMPILE_TEMPLATE_MMX2
-        } //if MMX2 can't be used
-#endif
-#else
         c->hcscale_fast(c, dst, dstWidth, src1, src2, srcW, xInc);
-#endif /* ARCH_X86 */
-    }
-    if(c->srcRange != c->dstRange && !(isRGB(c->dstFormat) || isBGR(c->dstFormat))) {
-        int i;
-        //FIXME all pal and rgb srcFormats could do this convertion as well
-        //FIXME all scalers more complex than bilinear could do half of this transform
-        if(c->srcRange) {
-            for (i=0; i<dstWidth; i++) {
-                dst[i     ]= (dst[i     ]*1799 + 4081085)>>11; //1469
-                dst[i+VOFW]= (dst[i+VOFW]*1799 + 4081085)>>11; //1469
-            }
-        } else {
-            for (i=0; i<dstWidth; i++) {
-                dst[i     ]= (FFMIN(dst[i     ],30775)*4663 - 9289992)>>12; //-264
-                dst[i+VOFW]= (FFMIN(dst[i+VOFW],30775)*4663 - 9289992)>>12; //-264
-            }
-        }
     }
+
+    if (c->chrConvertRange)
+        c->chrConvertRange(dst, dstWidth);
 }
 
 #define DEBUG_SWSCALE_BUFFERS 0
 #define DEBUG_BUFFERS(...) if (DEBUG_SWSCALE_BUFFERS) av_log(c, AV_LOG_DEBUG, __VA_ARGS__)
 
-static int RENAME(swScale)(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
+static int RENAME(swScale)(SwsContext *c, const uint8_t* src[], int srcStride[], int srcSliceY,
                            int srcSliceH, uint8_t* dst[], int dstStride[])
 {
     /* load a few things into local vars to make the code more readable? and faster */
@@ -2576,7 +2589,6 @@ static int RENAME(swScale)(SwsContext *c, uint8_t* src[], int srcStride[], int s
     const int lumXInc= c->lumXInc;
     const int chrXInc= c->chrXInc;
     const enum PixelFormat dstFormat= c->dstFormat;
-    const enum PixelFormat srcFormat= c->srcFormat;
     const int flags= c->flags;
     int16_t *vLumFilterPos= c->vLumFilterPos;
     int16_t *vChrFilterPos= c->vChrFilterPos;
@@ -2588,7 +2600,7 @@ static int RENAME(swScale)(SwsContext *c, uint8_t* src[], int srcStride[], int s
     int16_t *hChrFilter= c->hChrFilter;
     int32_t *lumMmxFilter= c->lumMmxFilter;
     int32_t *chrMmxFilter= c->chrMmxFilter;
-    int32_t *alpMmxFilter= c->alpMmxFilter;
+    int32_t av_unused *alpMmxFilter= c->alpMmxFilter;
     const int vLumFilterSize= c->vLumFilterSize;
     const int vChrFilterSize= c->vChrFilterSize;
     const int hLumFilterSize= c->hLumFilterSize;
@@ -2662,8 +2674,10 @@ static int RENAME(swScale)(SwsContext *c, uint8_t* src[], int srcStride[], int s
         unsigned char *aDest=(CONFIG_SWSCALE_ALPHA && alpPixBuf) ? dst[3]+dstStride[3]*dstY : NULL;
 
         const int firstLumSrcY= vLumFilterPos[dstY]; //First line needed as input
+        const int firstLumSrcY2= vLumFilterPos[FFMIN(dstY | ((1<<c->chrDstVSubSample) - 1), dstH-1)];
         const int firstChrSrcY= vChrFilterPos[chrDstY]; //First line needed as input
         int lastLumSrcY= firstLumSrcY + vLumFilterSize -1; // Last line needed as input
+        int lastLumSrcY2=firstLumSrcY2+ vLumFilterSize -1; // Last line needed as input
         int lastChrSrcY= firstChrSrcY + vChrFilterSize -1; // Last line needed as input
         int enough_lines;
 
@@ -2673,57 +2687,60 @@ static int RENAME(swScale)(SwsContext *c, uint8_t* src[], int srcStride[], int s
         assert(firstLumSrcY >= lastInLumBuf - vLumBufSize + 1);
         assert(firstChrSrcY >= lastInChrBuf - vChrBufSize + 1);
 
-        // Do we have enough lines in this slice to output the dstY line
-        enough_lines = lastLumSrcY < srcSliceY + srcSliceH && lastChrSrcY < -((-srcSliceY - srcSliceH)>>c->chrSrcVSubSample);
-        if (!enough_lines) {
-            lastLumSrcY = srcSliceY + srcSliceH - 1;
-            lastChrSrcY = chrSrcSliceY + chrSrcSliceH - 1;
-        }
-
         DEBUG_BUFFERS("dstY: %d\n", dstY);
         DEBUG_BUFFERS("\tfirstLumSrcY: %d lastLumSrcY: %d lastInLumBuf: %d\n",
                          firstLumSrcY,    lastLumSrcY,    lastInLumBuf);
         DEBUG_BUFFERS("\tfirstChrSrcY: %d lastChrSrcY: %d lastInChrBuf: %d\n",
                          firstChrSrcY,    lastChrSrcY,    lastInChrBuf);
 
+        // Do we have enough lines in this slice to output the dstY line
+        enough_lines = lastLumSrcY2 < srcSliceY + srcSliceH && lastChrSrcY < -((-srcSliceY - srcSliceH)>>c->chrSrcVSubSample);
+
+        if (!enough_lines) {
+            lastLumSrcY = srcSliceY + srcSliceH - 1;
+            lastChrSrcY = chrSrcSliceY + chrSrcSliceH - 1;
+            DEBUG_BUFFERS("buffering slice: lastLumSrcY %d lastChrSrcY %d\n",
+                                            lastLumSrcY, lastChrSrcY);
+        }
+
         //Do horizontal scaling
         while(lastInLumBuf < lastLumSrcY) {
-            uint8_t *src1= src[0]+(lastInLumBuf + 1 - srcSliceY)*srcStride[0];
-            uint8_t *src2= src[3]+(lastInLumBuf + 1 - srcSliceY)*srcStride[3];
+            const uint8_t *src1= src[0]+(lastInLumBuf + 1 - srcSliceY)*srcStride[0];
+            const uint8_t *src2= src[3]+(lastInLumBuf + 1 - srcSliceY)*srcStride[3];
             lumBufIndex++;
-            DEBUG_BUFFERS("\t\tlumBufIndex %d: lastInLumBuf: %d\n",
-                               lumBufIndex,    lastInLumBuf);
             assert(lumBufIndex < 2*vLumBufSize);
             assert(lastInLumBuf + 1 - srcSliceY < srcSliceH);
             assert(lastInLumBuf + 1 - srcSliceY >= 0);
             RENAME(hyscale)(c, lumPixBuf[ lumBufIndex ], dstW, src1, srcW, lumXInc,
-                            flags, hLumFilter, hLumFilterPos, hLumFilterSize,
-                            c->srcFormat, formatConvBuffer,
+                            hLumFilter, hLumFilterPos, hLumFilterSize,
+                            formatConvBuffer,
                             pal, 0);
             if (CONFIG_SWSCALE_ALPHA && alpPixBuf)
                 RENAME(hyscale)(c, alpPixBuf[ lumBufIndex ], dstW, src2, srcW, lumXInc,
-                                flags, hLumFilter, hLumFilterPos, hLumFilterSize,
-                                c->srcFormat, formatConvBuffer,
+                                hLumFilter, hLumFilterPos, hLumFilterSize,
+                                formatConvBuffer,
                                 pal, 1);
             lastInLumBuf++;
+            DEBUG_BUFFERS("\t\tlumBufIndex %d: lastInLumBuf: %d\n",
+                               lumBufIndex,    lastInLumBuf);
         }
         while(lastInChrBuf < lastChrSrcY) {
-            uint8_t *src1= src[1]+(lastInChrBuf + 1 - chrSrcSliceY)*srcStride[1];
-            uint8_t *src2= src[2]+(lastInChrBuf + 1 - chrSrcSliceY)*srcStride[2];
+            const uint8_t *src1= src[1]+(lastInChrBuf + 1 - chrSrcSliceY)*srcStride[1];
+            const uint8_t *src2= src[2]+(lastInChrBuf + 1 - chrSrcSliceY)*srcStride[2];
             chrBufIndex++;
-            DEBUG_BUFFERS("\t\tchrBufIndex %d: lastInChrBuf: %d\n",
-                               chrBufIndex,    lastInChrBuf);
             assert(chrBufIndex < 2*vChrBufSize);
             assert(lastInChrBuf + 1 - chrSrcSliceY < (chrSrcSliceH));
             assert(lastInChrBuf + 1 - chrSrcSliceY >= 0);
             //FIXME replace parameters through context struct (some at least)
 
-            if (!(isGray(srcFormat) || isGray(dstFormat)))
+            if (c->needs_hcscale)
                 RENAME(hcscale)(c, chrPixBuf[ chrBufIndex ], chrDstW, src1, src2, chrSrcW, chrXInc,
-                                flags, hChrFilter, hChrFilterPos, hChrFilterSize,
-                                c->srcFormat, formatConvBuffer,
+                                hChrFilter, hChrFilterPos, hChrFilterSize,
+                                formatConvBuffer,
                                 pal);
             lastInChrBuf++;
+            DEBUG_BUFFERS("\t\tchrBufIndex %d: lastInChrBuf: %d\n",
+                               chrBufIndex,    lastInChrBuf);
         }
         //wrap buf index around to stay inside the ring buffer
         if (lumBufIndex >= vLumBufSize) lumBufIndex-= vLumBufSize;
@@ -2748,21 +2765,21 @@ static int RENAME(swScale)(SwsContext *c, uint8_t* src[], int srcStride[], int s
             if (flags & SWS_ACCURATE_RND) {
                 int s= APCK_SIZE / 8;
                 for (i=0; i<vLumFilterSize; i+=2) {
-                    *(void**)&lumMmxFilter[s*i              ]= lumSrcPtr[i  ];
-                    *(void**)&lumMmxFilter[s*i+APCK_PTR2/4  ]= lumSrcPtr[i+(vLumFilterSize>1)];
+                    *(const void**)&lumMmxFilter[s*i              ]= lumSrcPtr[i  ];
+                    *(const void**)&lumMmxFilter[s*i+APCK_PTR2/4  ]= lumSrcPtr[i+(vLumFilterSize>1)];
                               lumMmxFilter[s*i+APCK_COEF/4  ]=
                               lumMmxFilter[s*i+APCK_COEF/4+1]= vLumFilter[dstY*vLumFilterSize + i    ]
                         + (vLumFilterSize>1 ? vLumFilter[dstY*vLumFilterSize + i + 1]<<16 : 0);
                     if (CONFIG_SWSCALE_ALPHA && alpPixBuf) {
-                        *(void**)&alpMmxFilter[s*i              ]= alpSrcPtr[i  ];
-                        *(void**)&alpMmxFilter[s*i+APCK_PTR2/4  ]= alpSrcPtr[i+(vLumFilterSize>1)];
+                        *(const void**)&alpMmxFilter[s*i              ]= alpSrcPtr[i  ];
+                        *(const void**)&alpMmxFilter[s*i+APCK_PTR2/4  ]= alpSrcPtr[i+(vLumFilterSize>1)];
                                   alpMmxFilter[s*i+APCK_COEF/4  ]=
                                   alpMmxFilter[s*i+APCK_COEF/4+1]= lumMmxFilter[s*i+APCK_COEF/4  ];
                     }
                 }
                 for (i=0; i<vChrFilterSize; i+=2) {
-                    *(void**)&chrMmxFilter[s*i              ]= chrSrcPtr[i  ];
-                    *(void**)&chrMmxFilter[s*i+APCK_PTR2/4  ]= chrSrcPtr[i+(vChrFilterSize>1)];
+                    *(const void**)&chrMmxFilter[s*i              ]= chrSrcPtr[i  ];
+                    *(const void**)&chrMmxFilter[s*i+APCK_PTR2/4  ]= chrSrcPtr[i+(vChrFilterSize>1)];
                               chrMmxFilter[s*i+APCK_COEF/4  ]=
                               chrMmxFilter[s*i+APCK_COEF/4+1]= vChrFilter[chrDstY*vChrFilterSize + i    ]
                         + (vChrFilterSize>1 ? vChrFilter[chrDstY*vChrFilterSize + i + 1]<<16 : 0);
@@ -2807,9 +2824,9 @@ static int RENAME(swScale)(SwsContext *c, uint8_t* src[], int srcStride[], int s
                                   alpSrcPtr, (uint16_t *) dest, (uint16_t *) uDest, (uint16_t *) vDest, (uint16_t *) aDest, dstW, chrDstW,
                                   dstFormat);
                 } else if (vLumFilterSize == 1 && vChrFilterSize == 1) { // unscaled YV12
-                    int16_t *lumBuf = lumSrcPtr[0];
-                    int16_t *chrBuf= chrSrcPtr[0];
-                    int16_t *alpBuf= (CONFIG_SWSCALE_ALPHA && alpPixBuf) ? alpSrcPtr[0] : NULL;
+                    const int16_t *lumBuf = lumSrcPtr[0];
+                    const int16_t *chrBuf= chrSrcPtr[0];
+                    const int16_t *alpBuf= (CONFIG_SWSCALE_ALPHA && alpPixBuf) ? alpSrcPtr[0] : NULL;
                     c->yuv2yuv1(c, lumBuf, chrBuf, alpBuf, dest, uDest, vDest, aDest, dstW, chrDstW);
                 } else { //General YV12
                     c->yuv2yuvX(c,
@@ -2939,115 +2956,148 @@ static void RENAME(sws_init_swScale)(SwsContext *c)
 
     c->hScale       = RENAME(hScale      );
 
-    c->hyscale_fast = RENAME(hyscale_fast);
-    c->hcscale_fast = RENAME(hcscale_fast);
+#if COMPILE_TEMPLATE_MMX
+    // Use the new MMX scaler if the MMX2 one can't be used (it is faster than the x86 ASM one).
+    if (c->flags & SWS_FAST_BILINEAR && c->canMMX2BeUsed)
+#else
+    if (c->flags & SWS_FAST_BILINEAR)
+#endif
+    {
+        c->hyscale_fast = RENAME(hyscale_fast);
+        c->hcscale_fast = RENAME(hcscale_fast);
+    }
 
-    c->hcscale_internal = NULL;
+    c->chrToYV12 = NULL;
     switch(srcFormat) {
-        case PIX_FMT_YUYV422  : c->hcscale_internal = RENAME(yuy2ToUV); break;
-        case PIX_FMT_UYVY422  : c->hcscale_internal = RENAME(uyvyToUV); break;
-        case PIX_FMT_NV12     : c->hcscale_internal = RENAME(nv12ToUV); break;
-        case PIX_FMT_NV21     : c->hcscale_internal = RENAME(nv21ToUV); break;
+        case PIX_FMT_YUYV422  : c->chrToYV12 = RENAME(yuy2ToUV); break;
+        case PIX_FMT_UYVY422  : c->chrToYV12 = RENAME(uyvyToUV); break;
+        case PIX_FMT_NV12     : c->chrToYV12 = RENAME(nv12ToUV); break;
+        case PIX_FMT_NV21     : c->chrToYV12 = RENAME(nv21ToUV); break;
         case PIX_FMT_RGB8     :
         case PIX_FMT_BGR8     :
         case PIX_FMT_PAL8     :
         case PIX_FMT_BGR4_BYTE:
-        case PIX_FMT_RGB4_BYTE: c->hcscale_internal = palToUV; break;
+        case PIX_FMT_RGB4_BYTE: c->chrToYV12 = palToUV; break;
+        case PIX_FMT_YUV420P9 : c->chrToYV12 = (void*)RENAME(yuv9ToUV ); break;
+        case PIX_FMT_YUV420P10: c->chrToYV12 = (void*)RENAME(yuv10ToUV); break;
         case PIX_FMT_YUV420P16BE:
         case PIX_FMT_YUV422P16BE:
-        case PIX_FMT_YUV444P16BE: c->hcscale_internal = RENAME(BEToUV); break;
+        case PIX_FMT_YUV444P16BE: c->chrToYV12 = RENAME(BEToUV); break;
         case PIX_FMT_YUV420P16LE:
         case PIX_FMT_YUV422P16LE:
-        case PIX_FMT_YUV444P16LE: c->hcscale_internal = RENAME(LEToUV); break;
+        case PIX_FMT_YUV444P16LE: c->chrToYV12 = RENAME(LEToUV); break;
     }
     if (c->chrSrcHSubSample) {
         switch(srcFormat) {
         case PIX_FMT_RGB48BE:
-        case PIX_FMT_RGB48LE: c->hcscale_internal = rgb48ToUV_half; break;
-        case PIX_FMT_RGB32  :
-        case PIX_FMT_RGB32_1: c->hcscale_internal = bgr32ToUV_half; break;
-        case PIX_FMT_BGR24  : c->hcscale_internal = RENAME(bgr24ToUV_half); break;
-        case PIX_FMT_BGR565 : c->hcscale_internal = bgr16ToUV_half; break;
-        case PIX_FMT_BGR555 : c->hcscale_internal = bgr15ToUV_half; break;
-        case PIX_FMT_BGR32  :
-        case PIX_FMT_BGR32_1: c->hcscale_internal = rgb32ToUV_half; break;
-        case PIX_FMT_RGB24  : c->hcscale_internal = RENAME(rgb24ToUV_half); break;
-        case PIX_FMT_RGB565 : c->hcscale_internal = rgb16ToUV_half; break;
-        case PIX_FMT_RGB555 : c->hcscale_internal = rgb15ToUV_half; break;
+        case PIX_FMT_RGB48LE: c->chrToYV12 = rgb48ToUV_half; break;
+        case PIX_FMT_BGR48BE:
+        case PIX_FMT_BGR48LE: c->chrToYV12 = bgr48ToUV_half; break;
+        case PIX_FMT_RGB32  : c->chrToYV12 = bgr32ToUV_half;  break;
+        case PIX_FMT_RGB32_1: c->chrToYV12 = bgr321ToUV_half; break;
+        case PIX_FMT_BGR24  : c->chrToYV12 = RENAME(bgr24ToUV_half); break;
+        case PIX_FMT_BGR565 : c->chrToYV12 = bgr16ToUV_half; break;
+        case PIX_FMT_BGR555 : c->chrToYV12 = bgr15ToUV_half; break;
+        case PIX_FMT_BGR32  : c->chrToYV12 = rgb32ToUV_half;  break;
+        case PIX_FMT_BGR32_1: c->chrToYV12 = rgb321ToUV_half; break;
+        case PIX_FMT_RGB24  : c->chrToYV12 = RENAME(rgb24ToUV_half); break;
+        case PIX_FMT_RGB565 : c->chrToYV12 = rgb16ToUV_half; break;
+        case PIX_FMT_RGB555 : c->chrToYV12 = rgb15ToUV_half; break;
         }
     } else {
         switch(srcFormat) {
         case PIX_FMT_RGB48BE:
-        case PIX_FMT_RGB48LE: c->hcscale_internal = rgb48ToUV; break;
-        case PIX_FMT_RGB32  :
-        case PIX_FMT_RGB32_1: c->hcscale_internal = bgr32ToUV; break;
-        case PIX_FMT_BGR24  : c->hcscale_internal = RENAME(bgr24ToUV); break;
-        case PIX_FMT_BGR565 : c->hcscale_internal = bgr16ToUV; break;
-        case PIX_FMT_BGR555 : c->hcscale_internal = bgr15ToUV; break;
-        case PIX_FMT_BGR32  :
-        case PIX_FMT_BGR32_1: c->hcscale_internal = rgb32ToUV; break;
-        case PIX_FMT_RGB24  : c->hcscale_internal = RENAME(rgb24ToUV); break;
-        case PIX_FMT_RGB565 : c->hcscale_internal = rgb16ToUV; break;
-        case PIX_FMT_RGB555 : c->hcscale_internal = rgb15ToUV; break;
+        case PIX_FMT_RGB48LE: c->chrToYV12 = rgb48ToUV; break;
+        case PIX_FMT_BGR48BE:
+        case PIX_FMT_BGR48LE: c->chrToYV12 = bgr48ToUV; break;
+        case PIX_FMT_RGB32  : c->chrToYV12 = bgr32ToUV;  break;
+        case PIX_FMT_RGB32_1: c->chrToYV12 = bgr321ToUV; break;
+        case PIX_FMT_BGR24  : c->chrToYV12 = RENAME(bgr24ToUV); break;
+        case PIX_FMT_BGR565 : c->chrToYV12 = bgr16ToUV; break;
+        case PIX_FMT_BGR555 : c->chrToYV12 = bgr15ToUV; break;
+        case PIX_FMT_BGR32  : c->chrToYV12 = rgb32ToUV;  break;
+        case PIX_FMT_BGR32_1: c->chrToYV12 = rgb321ToUV; break;
+        case PIX_FMT_RGB24  : c->chrToYV12 = RENAME(rgb24ToUV); break;
+        case PIX_FMT_RGB565 : c->chrToYV12 = rgb16ToUV; break;
+        case PIX_FMT_RGB555 : c->chrToYV12 = rgb15ToUV; break;
         }
     }
 
-    c->hyscale_internal = NULL;
-    c->hascale_internal = NULL;
+    c->lumToYV12 = NULL;
+    c->alpToYV12 = NULL;
     switch (srcFormat) {
+    case PIX_FMT_YUV420P9 : c->lumToYV12 = (void*)RENAME(yuv9ToY ); break;
+    case PIX_FMT_YUV420P10: c->lumToYV12 = (void*)RENAME(yuv10ToY); break;
     case PIX_FMT_YUYV422  :
     case PIX_FMT_YUV420P16BE:
     case PIX_FMT_YUV422P16BE:
     case PIX_FMT_YUV444P16BE:
-    case PIX_FMT_GRAY16BE : c->hyscale_internal = RENAME(yuy2ToY); break;
+    case PIX_FMT_GRAY8A   :
+    case PIX_FMT_GRAY16BE : c->lumToYV12 = RENAME(yuy2ToY); break;
     case PIX_FMT_UYVY422  :
     case PIX_FMT_YUV420P16LE:
     case PIX_FMT_YUV422P16LE:
     case PIX_FMT_YUV444P16LE:
-    case PIX_FMT_GRAY16LE : c->hyscale_internal = RENAME(uyvyToY); break;
-    case PIX_FMT_BGR24    : c->hyscale_internal = RENAME(bgr24ToY); break;
-    case PIX_FMT_BGR565   : c->hyscale_internal = bgr16ToY; break;
-    case PIX_FMT_BGR555   : c->hyscale_internal = bgr15ToY; break;
-    case PIX_FMT_RGB24    : c->hyscale_internal = RENAME(rgb24ToY); break;
-    case PIX_FMT_RGB565   : c->hyscale_internal = rgb16ToY; break;
-    case PIX_FMT_RGB555   : c->hyscale_internal = rgb15ToY; break;
+    case PIX_FMT_GRAY16LE : c->lumToYV12 = RENAME(uyvyToY); break;
+    case PIX_FMT_BGR24    : c->lumToYV12 = RENAME(bgr24ToY); break;
+    case PIX_FMT_BGR565   : c->lumToYV12 = bgr16ToY; break;
+    case PIX_FMT_BGR555   : c->lumToYV12 = bgr15ToY; break;
+    case PIX_FMT_RGB24    : c->lumToYV12 = RENAME(rgb24ToY); break;
+    case PIX_FMT_RGB565   : c->lumToYV12 = rgb16ToY; break;
+    case PIX_FMT_RGB555   : c->lumToYV12 = rgb15ToY; break;
     case PIX_FMT_RGB8     :
     case PIX_FMT_BGR8     :
     case PIX_FMT_PAL8     :
     case PIX_FMT_BGR4_BYTE:
-    case PIX_FMT_RGB4_BYTE: c->hyscale_internal = palToY; break;
-    case PIX_FMT_MONOBLACK: c->hyscale_internal = monoblack2Y; break;
-    case PIX_FMT_MONOWHITE: c->hyscale_internal = monowhite2Y; break;
-    case PIX_FMT_RGB32  :
-    case PIX_FMT_RGB32_1: c->hyscale_internal = bgr32ToY; break;
-    case PIX_FMT_BGR32  :
-    case PIX_FMT_BGR32_1: c->hyscale_internal = rgb32ToY; break;
+    case PIX_FMT_RGB4_BYTE: c->lumToYV12 = palToY; break;
+    case PIX_FMT_MONOBLACK: c->lumToYV12 = monoblack2Y; break;
+    case PIX_FMT_MONOWHITE: c->lumToYV12 = monowhite2Y; break;
+    case PIX_FMT_RGB32  : c->lumToYV12 = bgr32ToY;  break;
+    case PIX_FMT_RGB32_1: c->lumToYV12 = bgr321ToY; break;
+    case PIX_FMT_BGR32  : c->lumToYV12 = rgb32ToY;  break;
+    case PIX_FMT_BGR32_1: c->lumToYV12 = rgb321ToY; break;
     case PIX_FMT_RGB48BE:
-    case PIX_FMT_RGB48LE: c->hyscale_internal = rgb48ToY; break;
+    case PIX_FMT_RGB48LE: c->lumToYV12 = rgb48ToY; break;
+    case PIX_FMT_BGR48BE:
+    case PIX_FMT_BGR48LE: c->lumToYV12 = bgr48ToY; break;
     }
     if (c->alpPixBuf) {
         switch (srcFormat) {
         case PIX_FMT_RGB32  :
         case PIX_FMT_RGB32_1:
         case PIX_FMT_BGR32  :
-        case PIX_FMT_BGR32_1: c->hascale_internal = abgrToA; break;
+        case PIX_FMT_BGR32_1: c->alpToYV12 = abgrToA; break;
+        case PIX_FMT_GRAY8A : c->alpToYV12 = RENAME(yuy2ToY); break;
         }
     }
 
     switch (srcFormat) {
+    case PIX_FMT_GRAY8A :
+        c->alpSrcOffset = 1;
+        break;
     case PIX_FMT_RGB32  :
     case PIX_FMT_BGR32  :
         c->alpSrcOffset = 3;
         break;
-    case PIX_FMT_RGB32_1:
-    case PIX_FMT_BGR32_1:
-        c->lumSrcOffset = ALT32_CORR;
-        c->chrSrcOffset = ALT32_CORR;
-        break;
     case PIX_FMT_RGB48LE:
+    case PIX_FMT_BGR48LE:
         c->lumSrcOffset = 1;
         c->chrSrcOffset = 1;
         c->alpSrcOffset = 1;
         break;
     }
+
+    if (c->srcRange != c->dstRange && !isAnyRGB(c->dstFormat)) {
+        if (c->srcRange) {
+            c->lumConvertRange = RENAME(lumRangeFromJpeg);
+            c->chrConvertRange = RENAME(chrRangeFromJpeg);
+        } else {
+            c->lumConvertRange = RENAME(lumRangeToJpeg);
+            c->chrConvertRange = RENAME(chrRangeToJpeg);
+        }
+    }
+
+    if (!(isGray(srcFormat) || isGray(c->dstFormat) ||
+          srcFormat == PIX_FMT_MONOBLACK || srcFormat == PIX_FMT_MONOWHITE))
+        c->needs_hcscale = 1;
 }