libavformat/tee: Add fifo support for tee
[ffmpeg.git] / libswscale / input.c
index 9d03c68..1f4ea18 100644 (file)
@@ -18,7 +18,6 @@
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
-#include <assert.h>
 #include <math.h>
 #include <stdint.h>
 #include <stdio.h>
 #include "libavutil/intreadwrite.h"
 #include "libavutil/mathematics.h"
 #include "libavutil/pixdesc.h"
+#include "libavutil/avassert.h"
 #include "config.h"
 #include "rgb2rgb.h"
 #include "swscale.h"
 #include "swscale_internal.h"
 
-#define RGB2YUV_SHIFT 15
-#define BY  ((int)(0.114 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
-#define BV (-(int)(0.081 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
-#define BU  ((int)(0.500 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
-#define GY  ((int)(0.587 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
-#define GV (-(int)(0.419 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
-#define GU (-(int)(0.331 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
-#define RY  ((int)(0.299 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
-#define RV  ((int)(0.500 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
-#define RU (-(int)(0.169 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
-
 #define input_pixel(pos) (isBE(origin) ? AV_RB16(pos) : AV_RL16(pos))
 
-#define r ((origin == PIX_FMT_BGR48BE || origin == PIX_FMT_BGR48LE) ? b_r : r_b)
-#define b ((origin == PIX_FMT_BGR48BE || origin == PIX_FMT_BGR48LE) ? r_b : b_r)
+#define r ((origin == AV_PIX_FMT_BGR48BE || origin == AV_PIX_FMT_BGR48LE || origin == AV_PIX_FMT_BGRA64BE || origin == AV_PIX_FMT_BGRA64LE) ? b_r : r_b)
+#define b ((origin == AV_PIX_FMT_BGR48BE || origin == AV_PIX_FMT_BGR48LE || origin == AV_PIX_FMT_BGRA64BE || origin == AV_PIX_FMT_BGRA64LE) ? r_b : b_r)
 
 static av_always_inline void
 rgb64ToY_c_template(uint16_t *dst, const uint16_t *src, int width,
-                    enum PixelFormat origin)
+                    enum AVPixelFormat origin, int32_t *rgb2yuv)
 {
+    int32_t ry = rgb2yuv[RY_IDX], gy = rgb2yuv[GY_IDX], by = rgb2yuv[BY_IDX];
     int i;
     for (i = 0; i < width; i++) {
         unsigned int r_b = input_pixel(&src[i*4+0]);
         unsigned int   g = input_pixel(&src[i*4+1]);
         unsigned int b_r = input_pixel(&src[i*4+2]);
 
-        dst[i] = (RY*r + GY*g + BY*b + (0x2001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
+        dst[i] = (ry*r + gy*g + by*b + (0x2001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
     }
 }
 
 static av_always_inline void
 rgb64ToUV_c_template(uint16_t *dstU, uint16_t *dstV,
                     const uint16_t *src1, const uint16_t *src2,
-                    int width, enum PixelFormat origin)
+                    int width, enum AVPixelFormat origin, int32_t *rgb2yuv)
 {
     int i;
-    assert(src1==src2);
+    int32_t ru = rgb2yuv[RU_IDX], gu = rgb2yuv[GU_IDX], bu = rgb2yuv[BU_IDX];
+    int32_t rv = rgb2yuv[RV_IDX], gv = rgb2yuv[GV_IDX], bv = rgb2yuv[BV_IDX];
+    av_assert1(src1==src2);
     for (i = 0; i < width; i++) {
         int r_b = input_pixel(&src1[i*4+0]);
         int   g = input_pixel(&src1[i*4+1]);
         int b_r = input_pixel(&src1[i*4+2]);
 
-        dstU[i] = (RU*r + GU*g + BU*b + (0x10001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
-        dstV[i] = (RV*r + GV*g + BV*b + (0x10001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
+        dstU[i] = (ru*r + gu*g + bu*b + (0x10001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
+        dstV[i] = (rv*r + gv*g + bv*b + (0x10001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
     }
 }
 
 static av_always_inline void
 rgb64ToUV_half_c_template(uint16_t *dstU, uint16_t *dstV,
                           const uint16_t *src1, const uint16_t *src2,
-                          int width, enum PixelFormat origin)
+                          int width, enum AVPixelFormat origin, int32_t *rgb2yuv)
 {
     int i;
-    assert(src1==src2);
+    int32_t ru = rgb2yuv[RU_IDX], gu = rgb2yuv[GU_IDX], bu = rgb2yuv[BU_IDX];
+    int32_t rv = rgb2yuv[RV_IDX], gv = rgb2yuv[GV_IDX], bv = rgb2yuv[BV_IDX];
+    av_assert1(src1==src2);
     for (i = 0; i < width; i++) {
         int r_b = (input_pixel(&src1[8 * i + 0]) + input_pixel(&src1[8 * i + 4]) + 1) >> 1;
         int   g = (input_pixel(&src1[8 * i + 1]) + input_pixel(&src1[8 * i + 5]) + 1) >> 1;
         int b_r = (input_pixel(&src1[8 * i + 2]) + input_pixel(&src1[8 * i + 6]) + 1) >> 1;
 
-        dstU[i]= (RU*r + GU*g + BU*b + (0x10001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
-        dstV[i]= (RV*r + GV*g + BV*b + (0x10001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
+        dstU[i]= (ru*r + gu*g + bu*b + (0x10001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
+        dstV[i]= (rv*r + gv*g + bv*b + (0x10001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
     }
 }
 
 #define rgb64funcs(pattern, BE_LE, origin) \
 static void pattern ## 64 ## BE_LE ## ToY_c(uint8_t *_dst, const uint8_t *_src, const uint8_t *unused0, const uint8_t *unused1,\
-                                    int width, uint32_t *unused) \
+                                    int width, uint32_t *rgb2yuv) \
 { \
     const uint16_t *src = (const uint16_t *) _src; \
     uint16_t *dst = (uint16_t *) _dst; \
-    rgb64ToY_c_template(dst, src, width, origin); \
+    rgb64ToY_c_template(dst, src, width, origin, rgb2yuv); \
 } \
  \
 static void pattern ## 64 ## BE_LE ## ToUV_c(uint8_t *_dstU, uint8_t *_dstV, \
                                     const uint8_t *unused0, const uint8_t *_src1, const uint8_t *_src2, \
-                                    int width, uint32_t *unused) \
+                                    int width, uint32_t *rgb2yuv) \
 { \
     const uint16_t *src1 = (const uint16_t *) _src1, \
                    *src2 = (const uint16_t *) _src2; \
     uint16_t *dstU = (uint16_t *) _dstU, *dstV = (uint16_t *) _dstV; \
-    rgb64ToUV_c_template(dstU, dstV, src1, src2, width, origin); \
+    rgb64ToUV_c_template(dstU, dstV, src1, src2, width, origin, rgb2yuv); \
 } \
  \
 static void pattern ## 64 ## BE_LE ## ToUV_half_c(uint8_t *_dstU, uint8_t *_dstV, \
                                     const uint8_t *unused0, const uint8_t *_src1, const uint8_t *_src2, \
-                                    int width, uint32_t *unused) \
+                                    int width, uint32_t *rgb2yuv) \
 { \
     const uint16_t *src1 = (const uint16_t *) _src1, \
                    *src2 = (const uint16_t *) _src2; \
     uint16_t *dstU = (uint16_t *) _dstU, *dstV = (uint16_t *) _dstV; \
-    rgb64ToUV_half_c_template(dstU, dstV, src1, src2, width, origin); \
+    rgb64ToUV_half_c_template(dstU, dstV, src1, src2, width, origin, rgb2yuv); \
 }
 
-rgb64funcs(rgb, LE, PIX_FMT_RGBA64LE)
-rgb64funcs(rgb, BE, PIX_FMT_RGBA64BE)
+rgb64funcs(rgb, LE, AV_PIX_FMT_RGBA64LE)
+rgb64funcs(rgb, BE, AV_PIX_FMT_RGBA64BE)
+rgb64funcs(bgr, LE, AV_PIX_FMT_BGRA64LE)
+rgb64funcs(bgr, BE, AV_PIX_FMT_BGRA64BE)
 
 static av_always_inline void rgb48ToY_c_template(uint16_t *dst,
                                                  const uint16_t *src, int width,
-                                                 enum PixelFormat origin)
+                                                 enum AVPixelFormat origin,
+                                                 int32_t *rgb2yuv)
 {
+    int32_t ry = rgb2yuv[RY_IDX], gy = rgb2yuv[GY_IDX], by = rgb2yuv[BY_IDX];
     int i;
     for (i = 0; i < width; i++) {
         unsigned int r_b = input_pixel(&src[i * 3 + 0]);
         unsigned int g   = input_pixel(&src[i * 3 + 1]);
         unsigned int b_r = input_pixel(&src[i * 3 + 2]);
 
-        dst[i] = (RY * r + GY * g + BY * b + (0x2001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
+        dst[i] = (ry*r + gy*g + by*b + (0x2001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
     }
 }
 
@@ -150,17 +148,20 @@ static av_always_inline void rgb48ToUV_c_template(uint16_t *dstU,
                                                   const uint16_t *src1,
                                                   const uint16_t *src2,
                                                   int width,
-                                                  enum PixelFormat origin)
+                                                  enum AVPixelFormat origin,
+                                                  int32_t *rgb2yuv)
 {
     int i;
-    assert(src1 == src2);
+    int32_t ru = rgb2yuv[RU_IDX], gu = rgb2yuv[GU_IDX], bu = rgb2yuv[BU_IDX];
+    int32_t rv = rgb2yuv[RV_IDX], gv = rgb2yuv[GV_IDX], bv = rgb2yuv[BV_IDX];
+    av_assert1(src1 == src2);
     for (i = 0; i < width; i++) {
         int r_b = input_pixel(&src1[i * 3 + 0]);
         int g   = input_pixel(&src1[i * 3 + 1]);
         int b_r = input_pixel(&src1[i * 3 + 2]);
 
-        dstU[i] = (RU * r + GU * g + BU * b + (0x10001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
-        dstV[i] = (RV * r + GV * g + BV * b + (0x10001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
+        dstU[i] = (ru*r + gu*g + bu*b + (0x10001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
+        dstV[i] = (rv*r + gv*g + bv*b + (0x10001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
     }
 }
 
@@ -169,10 +170,13 @@ static av_always_inline void rgb48ToUV_half_c_template(uint16_t *dstU,
                                                        const uint16_t *src1,
                                                        const uint16_t *src2,
                                                        int width,
-                                                       enum PixelFormat origin)
+                                                       enum AVPixelFormat origin,
+                                                       int32_t *rgb2yuv)
 {
     int i;
-    assert(src1 == src2);
+    int32_t ru = rgb2yuv[RU_IDX], gu = rgb2yuv[GU_IDX], bu = rgb2yuv[BU_IDX];
+    int32_t rv = rgb2yuv[RV_IDX], gv = rgb2yuv[GV_IDX], bv = rgb2yuv[BV_IDX];
+    av_assert1(src1 == src2);
     for (i = 0; i < width; i++) {
         int r_b = (input_pixel(&src1[6 * i + 0]) +
                    input_pixel(&src1[6 * i + 3]) + 1) >> 1;
@@ -181,8 +185,8 @@ static av_always_inline void rgb48ToUV_half_c_template(uint16_t *dstU,
         int b_r = (input_pixel(&src1[6 * i + 2]) +
                    input_pixel(&src1[6 * i + 5]) + 1) >> 1;
 
-        dstU[i] = (RU * r + GU * g + BU * b + (0x10001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
-        dstV[i] = (RV * r + GV * g + BV * b + (0x10001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
+        dstU[i] = (ru*r + gu*g + bu*b + (0x10001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
+        dstV[i] = (rv*r + gv*g + bv*b + (0x10001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
     }
 }
 
@@ -195,11 +199,11 @@ static void pattern ## 48 ## BE_LE ## ToY_c(uint8_t *_dst,              \
                                             const uint8_t *_src,        \
                                             const uint8_t *unused0, const uint8_t *unused1,\
                                             int width,                  \
-                                            uint32_t *unused)           \
+                                            uint32_t *rgb2yuv)          \
 {                                                                       \
     const uint16_t *src = (const uint16_t *)_src;                       \
     uint16_t *dst       = (uint16_t *)_dst;                             \
-    rgb48ToY_c_template(dst, src, width, origin);                       \
+    rgb48ToY_c_template(dst, src, width, origin, rgb2yuv);              \
 }                                                                       \
                                                                         \
 static void pattern ## 48 ## BE_LE ## ToUV_c(uint8_t *_dstU,            \
@@ -208,13 +212,13 @@ static void pattern ## 48 ## BE_LE ## ToUV_c(uint8_t *_dstU,            \
                                              const uint8_t *_src1,      \
                                              const uint8_t *_src2,      \
                                              int width,                 \
-                                             uint32_t *unused)          \
+                                             uint32_t *rgb2yuv)         \
 {                                                                       \
     const uint16_t *src1 = (const uint16_t *)_src1,                     \
                    *src2 = (const uint16_t *)_src2;                     \
     uint16_t *dstU = (uint16_t *)_dstU,                                 \
              *dstV = (uint16_t *)_dstV;                                 \
-    rgb48ToUV_c_template(dstU, dstV, src1, src2, width, origin);        \
+    rgb48ToUV_c_template(dstU, dstV, src1, src2, width, origin, rgb2yuv);        \
 }                                                                       \
                                                                         \
 static void pattern ## 48 ## BE_LE ## ToUV_half_c(uint8_t *_dstU,       \
@@ -223,24 +227,24 @@ static void pattern ## 48 ## BE_LE ## ToUV_half_c(uint8_t *_dstU,       \
                                                   const uint8_t *_src1, \
                                                   const uint8_t *_src2, \
                                                   int width,            \
-                                                  uint32_t *unused)     \
+                                                  uint32_t *rgb2yuv)    \
 {                                                                       \
     const uint16_t *src1 = (const uint16_t *)_src1,                     \
                    *src2 = (const uint16_t *)_src2;                     \
     uint16_t *dstU = (uint16_t *)_dstU,                                 \
              *dstV = (uint16_t *)_dstV;                                 \
-    rgb48ToUV_half_c_template(dstU, dstV, src1, src2, width, origin);   \
+    rgb48ToUV_half_c_template(dstU, dstV, src1, src2, width, origin, rgb2yuv);   \
 }
 
-rgb48funcs(rgb, LE, PIX_FMT_RGB48LE)
-rgb48funcs(rgb, BE, PIX_FMT_RGB48BE)
-rgb48funcs(bgr, LE, PIX_FMT_BGR48LE)
-rgb48funcs(bgr, BE, PIX_FMT_BGR48BE)
+rgb48funcs(rgb, LE, AV_PIX_FMT_RGB48LE)
+rgb48funcs(rgb, BE, AV_PIX_FMT_RGB48BE)
+rgb48funcs(bgr, LE, AV_PIX_FMT_BGR48LE)
+rgb48funcs(bgr, BE, AV_PIX_FMT_BGR48BE)
 
-#define input_pixel(i) ((origin == PIX_FMT_RGBA ||                      \
-                         origin == PIX_FMT_BGRA ||                      \
-                         origin == PIX_FMT_ARGB ||                      \
-                         origin == PIX_FMT_ABGR)                        \
+#define input_pixel(i) ((origin == AV_PIX_FMT_RGBA ||                      \
+                         origin == AV_PIX_FMT_BGRA ||                      \
+                         origin == AV_PIX_FMT_ARGB ||                      \
+                         origin == AV_PIX_FMT_ABGR)                        \
                         ? AV_RN32A(&src[(i) * 4])                       \
                         : (isBE(origin) ? AV_RB16(&src[(i) * 2])        \
                                         : AV_RL16(&src[(i) * 2])))
@@ -248,14 +252,15 @@ rgb48funcs(bgr, BE, PIX_FMT_BGR48BE)
 static av_always_inline void rgb16_32ToY_c_template(int16_t *dst,
                                                     const uint8_t *src,
                                                     int width,
-                                                    enum PixelFormat origin,
+                                                    enum AVPixelFormat origin,
                                                     int shr, int shg,
                                                     int shb, int shp,
                                                     int maskr, int maskg,
                                                     int maskb, int rsh,
-                                                    int gsh, int bsh, int S)
+                                                    int gsh, int bsh, int S,
+                                                    int32_t *rgb2yuv)
 {
-    const int ry       = RY << rsh, gy = GY << gsh, by = BY << bsh;
+    const int ry       = rgb2yuv[RY_IDX]<<rsh, gy = rgb2yuv[GY_IDX]<<gsh, by = rgb2yuv[BY_IDX]<<bsh;
     const unsigned rnd = (32<<((S)-1)) + (1<<(S-7));
     int i;
 
@@ -273,15 +278,16 @@ static av_always_inline void rgb16_32ToUV_c_template(int16_t *dstU,
                                                      int16_t *dstV,
                                                      const uint8_t *src,
                                                      int width,
-                                                     enum PixelFormat origin,
+                                                     enum AVPixelFormat origin,
                                                      int shr, int shg,
                                                      int shb, int shp,
                                                      int maskr, int maskg,
                                                      int maskb, int rsh,
-                                                     int gsh, int bsh, int S)
+                                                     int gsh, int bsh, int S,
+                                                     int32_t *rgb2yuv)
 {
-    const int ru       = RU << rsh, gu = GU << gsh, bu = BU << bsh,
-              rv       = RV << rsh, gv = GV << gsh, bv = BV << bsh;
+    const int ru       = rgb2yuv[RU_IDX] << rsh, gu = rgb2yuv[GU_IDX] << gsh, bu = rgb2yuv[BU_IDX] << bsh,
+              rv       = rgb2yuv[RV_IDX] << rsh, gv = rgb2yuv[GV_IDX] << gsh, bv = rgb2yuv[BV_IDX] << bsh;
     const unsigned rnd = (256u<<((S)-1)) + (1<<(S-7));
     int i;
 
@@ -300,15 +306,16 @@ static av_always_inline void rgb16_32ToUV_half_c_template(int16_t *dstU,
                                                           int16_t *dstV,
                                                           const uint8_t *src,
                                                           int width,
-                                                          enum PixelFormat origin,
+                                                          enum AVPixelFormat origin,
                                                           int shr, int shg,
                                                           int shb, int shp,
                                                           int maskr, int maskg,
                                                           int maskb, int rsh,
-                                                          int gsh, int bsh, int S)
+                                                          int gsh, int bsh, int S,
+                                                          int32_t *rgb2yuv)
 {
-    const int ru       = RU << rsh, gu = GU << gsh, bu = BU << bsh,
-              rv       = RV << rsh, gv = GV << gsh, bv = BV << bsh,
+    const int ru       = rgb2yuv[RU_IDX] << rsh, gu = rgb2yuv[GU_IDX] << gsh, bu = rgb2yuv[BU_IDX] << bsh,
+              rv       = rgb2yuv[RV_IDX] << rsh, gv = rgb2yuv[GV_IDX] << gsh, bv = rgb2yuv[BV_IDX] << bsh,
               maskgx   = ~(maskr | maskb);
     const unsigned rnd = (256U<<(S)) + (1<<(S-6));
     int i;
@@ -317,15 +324,15 @@ static av_always_inline void rgb16_32ToUV_half_c_template(int16_t *dstU,
     maskb |= maskb << 1;
     maskg |= maskg << 1;
     for (i = 0; i < width; i++) {
-        int px0 = input_pixel(2 * i + 0) >> shp;
-        int px1 = input_pixel(2 * i + 1) >> shp;
+        unsigned px0 = input_pixel(2 * i + 0) >> shp;
+        unsigned px1 = input_pixel(2 * i + 1) >> shp;
         int b, r, g = (px0 & maskgx) + (px1 & maskgx);
         int rb = px0 + px1 - g;
 
         b = (rb & maskb) >> shb;
         if (shp ||
-            origin == PIX_FMT_BGR565LE || origin == PIX_FMT_BGR565BE ||
-            origin == PIX_FMT_RGB565LE || origin == PIX_FMT_RGB565BE) {
+            origin == AV_PIX_FMT_BGR565LE || origin == AV_PIX_FMT_BGR565BE ||
+            origin == AV_PIX_FMT_RGB565LE || origin == AV_PIX_FMT_RGB565BE) {
             g >>= shg;
         } else {
             g = (g & maskg) >> shg;
@@ -342,90 +349,110 @@ static av_always_inline void rgb16_32ToUV_half_c_template(int16_t *dstU,
 #define rgb16_32_wrapper(fmt, name, shr, shg, shb, shp, maskr,          \
                          maskg, maskb, rsh, gsh, bsh, S)                \
 static void name ## ToY_c(uint8_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2,            \
-                          int width, uint32_t *unused)                  \
+                          int width, uint32_t *tab)                     \
 {                                                                       \
     rgb16_32ToY_c_template((int16_t*)dst, src, width, fmt, shr, shg, shb, shp,    \
-                           maskr, maskg, maskb, rsh, gsh, bsh, S);      \
+                           maskr, maskg, maskb, rsh, gsh, bsh, S, tab); \
 }                                                                       \
                                                                         \
 static void name ## ToUV_c(uint8_t *dstU, uint8_t *dstV,                \
                            const uint8_t *unused0, const uint8_t *src, const uint8_t *dummy,    \
-                           int width, uint32_t *unused)                 \
+                           int width, uint32_t *tab)                    \
 {                                                                       \
     rgb16_32ToUV_c_template((int16_t*)dstU, (int16_t*)dstV, src, width, fmt,                \
                             shr, shg, shb, shp,                         \
-                            maskr, maskg, maskb, rsh, gsh, bsh, S);     \
+                            maskr, maskg, maskb, rsh, gsh, bsh, S, tab);\
 }                                                                       \
                                                                         \
 static void name ## ToUV_half_c(uint8_t *dstU, uint8_t *dstV,           \
                                 const uint8_t *unused0, const uint8_t *src,                     \
                                 const uint8_t *dummy,                   \
-                                int width, uint32_t *unused)            \
+                                int width, uint32_t *tab)               \
 {                                                                       \
     rgb16_32ToUV_half_c_template((int16_t*)dstU, (int16_t*)dstV, src, width, fmt,           \
                                  shr, shg, shb, shp,                    \
                                  maskr, maskg, maskb,                   \
-                                 rsh, gsh, bsh, S);                     \
+                                 rsh, gsh, bsh, S, tab);                \
 }
 
-rgb16_32_wrapper(PIX_FMT_BGR32,    bgr32,  16, 0,  0, 0, 0xFF0000, 0xFF00,   0x00FF,  8, 0,  8, RGB2YUV_SHIFT + 8)
-rgb16_32_wrapper(PIX_FMT_BGR32_1,  bgr321, 16, 0,  0, 8, 0xFF0000, 0xFF00,   0x00FF,  8, 0,  8, RGB2YUV_SHIFT + 8)
-rgb16_32_wrapper(PIX_FMT_RGB32,    rgb32,   0, 0, 16, 0,   0x00FF, 0xFF00, 0xFF0000,  8, 0,  8, RGB2YUV_SHIFT + 8)
-rgb16_32_wrapper(PIX_FMT_RGB32_1,  rgb321,  0, 0, 16, 8,   0x00FF, 0xFF00, 0xFF0000,  8, 0,  8, RGB2YUV_SHIFT + 8)
-rgb16_32_wrapper(PIX_FMT_BGR565LE, bgr16le, 0, 0,  0, 0,   0x001F, 0x07E0,   0xF800, 11, 5,  0, RGB2YUV_SHIFT + 8)
-rgb16_32_wrapper(PIX_FMT_BGR555LE, bgr15le, 0, 0,  0, 0,   0x001F, 0x03E0,   0x7C00, 10, 5,  0, RGB2YUV_SHIFT + 7)
-rgb16_32_wrapper(PIX_FMT_BGR444LE, bgr12le, 0, 0,  0, 0,   0x000F, 0x00F0,   0x0F00,  8, 4,  0, RGB2YUV_SHIFT + 4)
-rgb16_32_wrapper(PIX_FMT_RGB565LE, rgb16le, 0, 0,  0, 0,   0xF800, 0x07E0,   0x001F,  0, 5, 11, RGB2YUV_SHIFT + 8)
-rgb16_32_wrapper(PIX_FMT_RGB555LE, rgb15le, 0, 0,  0, 0,   0x7C00, 0x03E0,   0x001F,  0, 5, 10, RGB2YUV_SHIFT + 7)
-rgb16_32_wrapper(PIX_FMT_RGB444LE, rgb12le, 0, 0,  0, 0,   0x0F00, 0x00F0,   0x000F,  0, 4,  8, RGB2YUV_SHIFT + 4)
-rgb16_32_wrapper(PIX_FMT_BGR565BE, bgr16be, 0, 0,  0, 0,   0x001F, 0x07E0,   0xF800, 11, 5,  0, RGB2YUV_SHIFT + 8)
-rgb16_32_wrapper(PIX_FMT_BGR555BE, bgr15be, 0, 0,  0, 0,   0x001F, 0x03E0,   0x7C00, 10, 5,  0, RGB2YUV_SHIFT + 7)
-rgb16_32_wrapper(PIX_FMT_BGR444BE, bgr12be, 0, 0,  0, 0,   0x000F, 0x00F0,   0x0F00,  8, 4,  0, RGB2YUV_SHIFT + 4)
-rgb16_32_wrapper(PIX_FMT_RGB565BE, rgb16be, 0, 0,  0, 0,   0xF800, 0x07E0,   0x001F,  0, 5, 11, RGB2YUV_SHIFT + 8)
-rgb16_32_wrapper(PIX_FMT_RGB555BE, rgb15be, 0, 0,  0, 0,   0x7C00, 0x03E0,   0x001F,  0, 5, 10, RGB2YUV_SHIFT + 7)
-rgb16_32_wrapper(PIX_FMT_RGB444BE, rgb12be, 0, 0,  0, 0,   0x0F00, 0x00F0,   0x000F,  0, 4,  8, RGB2YUV_SHIFT + 4)
-
-static void gbr24pToUV_half_c(uint16_t *dstU, uint16_t *dstV,
+rgb16_32_wrapper(AV_PIX_FMT_BGR32,    bgr32,  16, 0,  0, 0, 0xFF0000, 0xFF00,   0x00FF,  8, 0,  8, RGB2YUV_SHIFT + 8)
+rgb16_32_wrapper(AV_PIX_FMT_BGR32_1,  bgr321, 16, 0,  0, 8, 0xFF0000, 0xFF00,   0x00FF,  8, 0,  8, RGB2YUV_SHIFT + 8)
+rgb16_32_wrapper(AV_PIX_FMT_RGB32,    rgb32,   0, 0, 16, 0,   0x00FF, 0xFF00, 0xFF0000,  8, 0,  8, RGB2YUV_SHIFT + 8)
+rgb16_32_wrapper(AV_PIX_FMT_RGB32_1,  rgb321,  0, 0, 16, 8,   0x00FF, 0xFF00, 0xFF0000,  8, 0,  8, RGB2YUV_SHIFT + 8)
+rgb16_32_wrapper(AV_PIX_FMT_BGR565LE, bgr16le, 0, 0,  0, 0,   0x001F, 0x07E0,   0xF800, 11, 5,  0, RGB2YUV_SHIFT + 8)
+rgb16_32_wrapper(AV_PIX_FMT_BGR555LE, bgr15le, 0, 0,  0, 0,   0x001F, 0x03E0,   0x7C00, 10, 5,  0, RGB2YUV_SHIFT + 7)
+rgb16_32_wrapper(AV_PIX_FMT_BGR444LE, bgr12le, 0, 0,  0, 0,   0x000F, 0x00F0,   0x0F00,  8, 4,  0, RGB2YUV_SHIFT + 4)
+rgb16_32_wrapper(AV_PIX_FMT_RGB565LE, rgb16le, 0, 0,  0, 0,   0xF800, 0x07E0,   0x001F,  0, 5, 11, RGB2YUV_SHIFT + 8)
+rgb16_32_wrapper(AV_PIX_FMT_RGB555LE, rgb15le, 0, 0,  0, 0,   0x7C00, 0x03E0,   0x001F,  0, 5, 10, RGB2YUV_SHIFT + 7)
+rgb16_32_wrapper(AV_PIX_FMT_RGB444LE, rgb12le, 0, 0,  0, 0,   0x0F00, 0x00F0,   0x000F,  0, 4,  8, RGB2YUV_SHIFT + 4)
+rgb16_32_wrapper(AV_PIX_FMT_BGR565BE, bgr16be, 0, 0,  0, 0,   0x001F, 0x07E0,   0xF800, 11, 5,  0, RGB2YUV_SHIFT + 8)
+rgb16_32_wrapper(AV_PIX_FMT_BGR555BE, bgr15be, 0, 0,  0, 0,   0x001F, 0x03E0,   0x7C00, 10, 5,  0, RGB2YUV_SHIFT + 7)
+rgb16_32_wrapper(AV_PIX_FMT_BGR444BE, bgr12be, 0, 0,  0, 0,   0x000F, 0x00F0,   0x0F00,  8, 4,  0, RGB2YUV_SHIFT + 4)
+rgb16_32_wrapper(AV_PIX_FMT_RGB565BE, rgb16be, 0, 0,  0, 0,   0xF800, 0x07E0,   0x001F,  0, 5, 11, RGB2YUV_SHIFT + 8)
+rgb16_32_wrapper(AV_PIX_FMT_RGB555BE, rgb15be, 0, 0,  0, 0,   0x7C00, 0x03E0,   0x001F,  0, 5, 10, RGB2YUV_SHIFT + 7)
+rgb16_32_wrapper(AV_PIX_FMT_RGB444BE, rgb12be, 0, 0,  0, 0,   0x0F00, 0x00F0,   0x000F,  0, 4,  8, RGB2YUV_SHIFT + 4)
+
+static void gbr24pToUV_half_c(uint8_t *_dstU, uint8_t *_dstV,
                          const uint8_t *gsrc, const uint8_t *bsrc, const uint8_t *rsrc,
-                         int width, enum PixelFormat origin)
+                         int width, uint32_t *rgb2yuv)
 {
+    uint16_t *dstU = (uint16_t *)_dstU;
+    uint16_t *dstV = (uint16_t *)_dstV;
+    int32_t ru = rgb2yuv[RU_IDX], gu = rgb2yuv[GU_IDX], bu = rgb2yuv[BU_IDX];
+    int32_t rv = rgb2yuv[RV_IDX], gv = rgb2yuv[GV_IDX], bv = rgb2yuv[BV_IDX];
+
     int i;
     for (i = 0; i < width; i++) {
         unsigned int g   = gsrc[2*i] + gsrc[2*i+1];
         unsigned int b   = bsrc[2*i] + bsrc[2*i+1];
         unsigned int r   = rsrc[2*i] + rsrc[2*i+1];
 
-        dstU[i] = (RU*r + GU*g + BU*b + (0x4001<<(RGB2YUV_SHIFT-6))) >> (RGB2YUV_SHIFT-6+1);
-        dstV[i] = (RV*r + GV*g + BV*b + (0x4001<<(RGB2YUV_SHIFT-6))) >> (RGB2YUV_SHIFT-6+1);
+        dstU[i] = (ru*r + gu*g + bu*b + (0x4001<<(RGB2YUV_SHIFT-6))) >> (RGB2YUV_SHIFT-6+1);
+        dstV[i] = (rv*r + gv*g + bv*b + (0x4001<<(RGB2YUV_SHIFT-6))) >> (RGB2YUV_SHIFT-6+1);
     }
 }
 
-static void rgba64ToA_c(int16_t *dst, const uint16_t *src, const uint8_t *unused1,
-                        const uint8_t *unused2, int width, uint32_t *unused)
+static void rgba64leToA_c(uint8_t *_dst, const uint8_t *_src, const uint8_t *unused1,
+                          const uint8_t *unused2, int width, uint32_t *unused)
 {
+    int16_t *dst = (int16_t *)_dst;
+    const uint16_t *src = (const uint16_t *)_src;
     int i;
     for (i = 0; i < width; i++)
-        dst[i] = src[4 * i + 3];
+        dst[i] = AV_RL16(src + 4 * i + 3);
 }
 
-static void abgrToA_c(int16_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width, uint32_t *unused)
+static void rgba64beToA_c(uint8_t *_dst, const uint8_t *_src, const uint8_t *unused1,
+                          const uint8_t *unused2, int width, uint32_t *unused)
 {
+    int16_t *dst = (int16_t *)_dst;
+    const uint16_t *src = (const uint16_t *)_src;
+    int i;
+    for (i = 0; i < width; i++)
+        dst[i] = AV_RB16(src + 4 * i + 3);
+}
+
+static void abgrToA_c(uint8_t *_dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width, uint32_t *unused)
+{
+    int16_t *dst = (int16_t *)_dst;
     int i;
     for (i=0; i<width; i++) {
         dst[i]= src[4*i]<<6;
     }
 }
 
-static void rgbaToA_c(int16_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width, uint32_t *unused)
+static void rgbaToA_c(uint8_t *_dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width, uint32_t *unused)
 {
+    int16_t *dst = (int16_t *)_dst;
     int i;
     for (i=0; i<width; i++) {
         dst[i]= src[4*i+3]<<6;
     }
 }
 
-static void palToA_c(int16_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width, uint32_t *pal)
+static void palToA_c(uint8_t *_dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width, uint32_t *pal)
 {
+    int16_t *dst = (int16_t *)_dst;
     int i;
     for (i=0; i<width; i++) {
         int d= src[i];
@@ -434,8 +461,9 @@ static void palToA_c(int16_t *dst, const uint8_t *src, const uint8_t *unused1, c
     }
 }
 
-static void palToY_c(int16_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, long width, uint32_t *pal)
+static void palToY_c(uint8_t *_dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width, uint32_t *pal)
 {
+    int16_t *dst = (int16_t *)_dst;
     int i;
     for (i = 0; i < width; i++) {
         int d = src[i];
@@ -444,12 +472,14 @@ static void palToY_c(int16_t *dst, const uint8_t *src, const uint8_t *unused1, c
     }
 }
 
-static void palToUV_c(uint16_t *dstU, int16_t *dstV,
+static void palToUV_c(uint8_t *_dstU, uint8_t *_dstV,
                            const uint8_t *unused0, const uint8_t *src1, const uint8_t *src2,
                       int width, uint32_t *pal)
 {
+    uint16_t *dstU = (uint16_t *)_dstU;
+    int16_t *dstV = (int16_t *)_dstV;
     int i;
-    assert(src1 == src2);
+    av_assert1(src1 == src2);
     for (i = 0; i < width; i++) {
         int p = pal[src1[i]];
 
@@ -458,8 +488,9 @@ static void palToUV_c(uint16_t *dstU, int16_t *dstV,
     }
 }
 
-static void monowhite2Y_c(int16_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2,  int width, uint32_t *unused)
+static void monowhite2Y_c(uint8_t *_dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2,  int width, uint32_t *unused)
 {
+    int16_t *dst = (int16_t *)_dst;
     int i, j;
     width = (width + 7) >> 3;
     for (i = 0; i < width; i++) {
@@ -474,8 +505,9 @@ static void monowhite2Y_c(int16_t *dst, const uint8_t *src, const uint8_t *unuse
     }
 }
 
-static void monoblack2Y_c(int16_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2,  int width, uint32_t *unused)
+static void monoblack2Y_c(uint8_t *_dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2,  int width, uint32_t *unused)
 {
+    int16_t *dst = (int16_t *)_dst;
     int i, j;
     width = (width + 7) >> 3;
     for (i = 0; i < width; i++) {
@@ -506,10 +538,21 @@ static void yuy2ToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *unused0, con
         dstU[i] = src1[4 * i + 1];
         dstV[i] = src1[4 * i + 3];
     }
-    assert(src1 == src2);
+    av_assert1(src1 == src2);
+}
+
+static void yvy2ToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *unused0, const uint8_t *src1,
+                       const uint8_t *src2, int width, uint32_t *unused)
+{
+    int i;
+    for (i = 0; i < width; i++) {
+        dstV[i] = src1[4 * i + 1];
+        dstU[i] = src1[4 * i + 3];
+    }
+    av_assert1(src1 == src2);
 }
 
-static void bswap16Y_c(uint8_t *_dst, const uint8_t *_src, const uint8_t *unused1, const uint8_t *unused2,  int width,
+static void bswap16Y_c(uint8_t *_dst, const uint8_t *_src, const uint8_t *unused1, const uint8_t *unused2, int width,
                        uint32_t *unused)
 {
     int i;
@@ -532,6 +575,65 @@ static void bswap16UV_c(uint8_t *_dstU, uint8_t *_dstV, const uint8_t *unused0,
     }
 }
 
+static void read_ya16le_gray_c(uint8_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width,
+                               uint32_t *unused)
+{
+    int i;
+    for (i = 0; i < width; i++)
+        AV_WN16(dst + i * 2, AV_RL16(src + i * 4));
+}
+
+static void read_ya16le_alpha_c(uint8_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width,
+                                uint32_t *unused)
+{
+    int i;
+    for (i = 0; i < width; i++)
+        AV_WN16(dst + i * 2, AV_RL16(src + i * 4 + 2));
+}
+
+static void read_ya16be_gray_c(uint8_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width,
+                               uint32_t *unused)
+{
+    int i;
+    for (i = 0; i < width; i++)
+        AV_WN16(dst + i * 2, AV_RB16(src + i * 4));
+}
+
+static void read_ya16be_alpha_c(uint8_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width,
+                                uint32_t *unused)
+{
+    int i;
+    for (i = 0; i < width; i++)
+        AV_WN16(dst + i * 2, AV_RB16(src + i * 4 + 2));
+}
+
+static void read_ayuv64le_Y_c(uint8_t *dst, const uint8_t *src, const uint8_t *unused0, const uint8_t *unused1, int width,
+                               uint32_t *unused2)
+{
+    int i;
+    for (i = 0; i < width; i++)
+        AV_WN16(dst + i * 2, AV_RL16(src + i * 8 + 2));
+}
+
+
+static void read_ayuv64le_UV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *unused0, const uint8_t *src,
+                               const uint8_t *unused1, int width, uint32_t *unused2)
+{
+    int i;
+    for (i = 0; i < width; i++) {
+        AV_WN16(dstU + i * 2, AV_RL16(src + i * 8 + 4));
+        AV_WN16(dstV + i * 2, AV_RL16(src + i * 8 + 6));
+    }
+}
+
+static void read_ayuv64le_A_c(uint8_t *dst, const uint8_t *src, const uint8_t *unused0, const uint8_t *unused1, int width,
+                                uint32_t *unused2)
+{
+    int i;
+    for (i = 0; i < width; i++)
+        AV_WN16(dst + i * 2, AV_RL16(src + i * 8));
+}
+
 /* This is almost identical to the previous, end exists only because
  * yuy2ToY/UV)(dst, src + 1, ...) would have 100% unaligned accesses. */
 static void uyvyToY_c(uint8_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2,  int width,
@@ -550,7 +652,7 @@ static void uyvyToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *unused0, con
         dstU[i] = src1[4 * i + 0];
         dstV[i] = src1[4 * i + 2];
     }
-    assert(src1 == src2);
+    av_assert1(src1 == src2);
 }
 
 static av_always_inline void nvXXtoUV_c(uint8_t *dst1, uint8_t *dst2,
@@ -577,511 +679,576 @@ static void nv21ToUV_c(uint8_t *dstU, uint8_t *dstV,
     nvXXtoUV_c(dstV, dstU, src1, width);
 }
 
-#define input_pixel(pos) (isBE(origin) ? AV_RB16(pos) : AV_RL16(pos))
+static void p010LEToY_c(uint8_t *dst, const uint8_t *src, const uint8_t *unused1,
+                        const uint8_t *unused2, int width, uint32_t *unused)
+{
+    int i;
+    for (i = 0; i < width; i++) {
+        AV_WN16(dst + i * 2, AV_RL16(src + i * 2) >> 6);
+    }
+}
 
-static void bgr24ToY_c(int16_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2,
+static void p010BEToY_c(uint8_t *dst, const uint8_t *src, const uint8_t *unused1,
+                        const uint8_t *unused2, int width, uint32_t *unused)
+{
+    int i;
+    for (i = 0; i < width; i++) {
+        AV_WN16(dst + i * 2, AV_RB16(src + i * 2) >> 6);
+    }
+}
+
+static void p010LEToUV_c(uint8_t *dstU, uint8_t *dstV,
+                       const uint8_t *unused0, const uint8_t *src1, const uint8_t *src2,
+                       int width, uint32_t *unused)
+{
+    int i;
+    for (i = 0; i < width; i++) {
+        AV_WN16(dstU + i * 2, AV_RL16(src1 + i * 4 + 0) >> 6);
+        AV_WN16(dstV + i * 2, AV_RL16(src1 + i * 4 + 2) >> 6);
+    }
+}
+
+static void p010BEToUV_c(uint8_t *dstU, uint8_t *dstV,
+                       const uint8_t *unused0, const uint8_t *src1, const uint8_t *src2,
                        int width, uint32_t *unused)
 {
     int i;
     for (i = 0; i < width; i++) {
+        AV_WN16(dstU + i * 2, AV_RB16(src1 + i * 4 + 0) >> 6);
+        AV_WN16(dstV + i * 2, AV_RB16(src1 + i * 4 + 2) >> 6);
+    }
+}
+
+#define input_pixel(pos) (isBE(origin) ? AV_RB16(pos) : AV_RL16(pos))
+
+static void bgr24ToY_c(uint8_t *_dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2,
+                       int width, uint32_t *rgb2yuv)
+{
+    int16_t *dst = (int16_t *)_dst;
+    int32_t ry = rgb2yuv[RY_IDX], gy = rgb2yuv[GY_IDX], by = rgb2yuv[BY_IDX];
+    int i;
+    for (i = 0; i < width; i++) {
         int b = src[i * 3 + 0];
         int g = src[i * 3 + 1];
         int r = src[i * 3 + 2];
 
-        dst[i] = ((RY*r + GY*g + BY*b + (32<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6));
+        dst[i] = ((ry*r + gy*g + by*b + (32<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6));
     }
 }
 
-static void bgr24ToUV_c(int16_t *dstU, int16_t *dstV, const uint8_t *unused0, const uint8_t *src1,
-                        const uint8_t *src2, int width, uint32_t *unused)
+static void bgr24ToUV_c(uint8_t *_dstU, uint8_t *_dstV, const uint8_t *unused0, const uint8_t *src1,
+                        const uint8_t *src2, int width, uint32_t *rgb2yuv)
 {
+    int16_t *dstU = (int16_t *)_dstU;
+    int16_t *dstV = (int16_t *)_dstV;
+    int32_t ru = rgb2yuv[RU_IDX], gu = rgb2yuv[GU_IDX], bu = rgb2yuv[BU_IDX];
+    int32_t rv = rgb2yuv[RV_IDX], gv = rgb2yuv[GV_IDX], bv = rgb2yuv[BV_IDX];
     int i;
     for (i = 0; i < width; i++) {
         int b = src1[3 * i + 0];
         int g = src1[3 * i + 1];
         int r = src1[3 * i + 2];
 
-        dstU[i] = (RU*r + GU*g + BU*b + (256<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6);
-        dstV[i] = (RV*r + GV*g + BV*b + (256<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6);
+        dstU[i] = (ru*r + gu*g + bu*b + (256<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6);
+        dstV[i] = (rv*r + gv*g + bv*b + (256<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6);
     }
-    assert(src1 == src2);
+    av_assert1(src1 == src2);
 }
 
-static void bgr24ToUV_half_c(int16_t *dstU, int16_t *dstV, const uint8_t *unused0, const uint8_t *src1,
-                             const uint8_t *src2, int width, uint32_t *unused)
+static void bgr24ToUV_half_c(uint8_t *_dstU, uint8_t *_dstV, const uint8_t *unused0, const uint8_t *src1,
+                             const uint8_t *src2, int width, uint32_t *rgb2yuv)
 {
+    int16_t *dstU = (int16_t *)_dstU;
+    int16_t *dstV = (int16_t *)_dstV;
     int i;
+    int32_t ru = rgb2yuv[RU_IDX], gu = rgb2yuv[GU_IDX], bu = rgb2yuv[BU_IDX];
+    int32_t rv = rgb2yuv[RV_IDX], gv = rgb2yuv[GV_IDX], bv = rgb2yuv[BV_IDX];
     for (i = 0; i < width; i++) {
         int b = src1[6 * i + 0] + src1[6 * i + 3];
         int g = src1[6 * i + 1] + src1[6 * i + 4];
         int r = src1[6 * i + 2] + src1[6 * i + 5];
 
-        dstU[i] = (RU*r + GU*g + BU*b + (256<<RGB2YUV_SHIFT) + (1<<(RGB2YUV_SHIFT-6)))>>(RGB2YUV_SHIFT-5);
-        dstV[i] = (RV*r + GV*g + BV*b + (256<<RGB2YUV_SHIFT) + (1<<(RGB2YUV_SHIFT-6)))>>(RGB2YUV_SHIFT-5);
+        dstU[i] = (ru*r + gu*g + bu*b + (256<<RGB2YUV_SHIFT) + (1<<(RGB2YUV_SHIFT-6)))>>(RGB2YUV_SHIFT-5);
+        dstV[i] = (rv*r + gv*g + bv*b + (256<<RGB2YUV_SHIFT) + (1<<(RGB2YUV_SHIFT-6)))>>(RGB2YUV_SHIFT-5);
     }
-    assert(src1 == src2);
+    av_assert1(src1 == src2);
 }
 
-static void rgb24ToY_c(int16_t *dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width,
-                       uint32_t *unused)
+static void rgb24ToY_c(uint8_t *_dst, const uint8_t *src, const uint8_t *unused1, const uint8_t *unused2, int width,
+                       uint32_t *rgb2yuv)
 {
+    int16_t *dst = (int16_t *)_dst;
+    int32_t ry = rgb2yuv[RY_IDX], gy = rgb2yuv[GY_IDX], by = rgb2yuv[BY_IDX];
     int i;
     for (i = 0; i < width; i++) {
         int r = src[i * 3 + 0];
         int g = src[i * 3 + 1];
         int b = src[i * 3 + 2];
 
-        dst[i] = ((RY*r + GY*g + BY*b + (32<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6));
+        dst[i] = ((ry*r + gy*g + by*b + (32<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6));
     }
 }
 
-static void rgb24ToUV_c(int16_t *dstU, int16_t *dstV, const uint8_t *unused0, const uint8_t *src1,
-                        const uint8_t *src2, int width, uint32_t *unused)
+static void rgb24ToUV_c(uint8_t *_dstU, uint8_t *_dstV, const uint8_t *unused0, const uint8_t *src1,
+                        const uint8_t *src2, int width, uint32_t *rgb2yuv)
 {
+    int16_t *dstU = (int16_t *)_dstU;
+    int16_t *dstV = (int16_t *)_dstV;
     int i;
-    assert(src1 == src2);
+    int32_t ru = rgb2yuv[RU_IDX], gu = rgb2yuv[GU_IDX], bu = rgb2yuv[BU_IDX];
+    int32_t rv = rgb2yuv[RV_IDX], gv = rgb2yuv[GV_IDX], bv = rgb2yuv[BV_IDX];
+    av_assert1(src1 == src2);
     for (i = 0; i < width; i++) {
         int r = src1[3 * i + 0];
         int g = src1[3 * i + 1];
         int b = src1[3 * i + 2];
 
-        dstU[i] = (RU*r + GU*g + BU*b + (256<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6);
-        dstV[i] = (RV*r + GV*g + BV*b + (256<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6);
+        dstU[i] = (ru*r + gu*g + bu*b + (256<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6);
+        dstV[i] = (rv*r + gv*g + bv*b + (256<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6);
     }
 }
 
-static void rgb24ToUV_half_c(int16_t *dstU, int16_t *dstV, const uint8_t *unused0, const uint8_t *src1,
-                             const uint8_t *src2, int width, uint32_t *unused)
+static void rgb24ToUV_half_c(uint8_t *_dstU, uint8_t *_dstV, const uint8_t *unused0, const uint8_t *src1,
+                             const uint8_t *src2, int width, uint32_t *rgb2yuv)
 {
+    int16_t *dstU = (int16_t *)_dstU;
+    int16_t *dstV = (int16_t *)_dstV;
     int i;
-    assert(src1 == src2);
+    int32_t ru = rgb2yuv[RU_IDX], gu = rgb2yuv[GU_IDX], bu = rgb2yuv[BU_IDX];
+    int32_t rv = rgb2yuv[RV_IDX], gv = rgb2yuv[GV_IDX], bv = rgb2yuv[BV_IDX];
+    av_assert1(src1 == src2);
     for (i = 0; i < width; i++) {
         int r = src1[6 * i + 0] + src1[6 * i + 3];
         int g = src1[6 * i + 1] + src1[6 * i + 4];
         int b = src1[6 * i + 2] + src1[6 * i + 5];
 
-        dstU[i] = (RU*r + GU*g + BU*b + (256<<RGB2YUV_SHIFT) + (1<<(RGB2YUV_SHIFT-6)))>>(RGB2YUV_SHIFT-5);
-        dstV[i] = (RV*r + GV*g + BV*b + (256<<RGB2YUV_SHIFT) + (1<<(RGB2YUV_SHIFT-6)))>>(RGB2YUV_SHIFT-5);
+        dstU[i] = (ru*r + gu*g + bu*b + (256<<RGB2YUV_SHIFT) + (1<<(RGB2YUV_SHIFT-6)))>>(RGB2YUV_SHIFT-5);
+        dstV[i] = (rv*r + gv*g + bv*b + (256<<RGB2YUV_SHIFT) + (1<<(RGB2YUV_SHIFT-6)))>>(RGB2YUV_SHIFT-5);
     }
 }
 
-static void planar_rgb_to_y(uint16_t *dst, const uint8_t *src[4], int width)
+static void planar_rgb_to_y(uint8_t *_dst, const uint8_t *src[4], int width, int32_t *rgb2yuv)
 {
+    uint16_t *dst = (uint16_t *)_dst;
+    int32_t ry = rgb2yuv[RY_IDX], gy = rgb2yuv[GY_IDX], by = rgb2yuv[BY_IDX];
     int i;
     for (i = 0; i < width; i++) {
         int g = src[0][i];
         int b = src[1][i];
         int r = src[2][i];
 
-        dst[i] = (RY*r + GY*g + BY*b + (0x801<<(RGB2YUV_SHIFT-7))) >> (RGB2YUV_SHIFT-6);
+        dst[i] = (ry*r + gy*g + by*b + (0x801<<(RGB2YUV_SHIFT-7))) >> (RGB2YUV_SHIFT-6);
     }
 }
 
-static void planar_rgb_to_uv(uint16_t *dstU, uint16_t *dstV, const uint8_t *src[4], int width)
+static void planar_rgb_to_a(uint8_t *_dst, const uint8_t *src[4], int width, int32_t *unused)
+{
+    uint16_t *dst = (uint16_t *)_dst;
+    int i;
+    for (i = 0; i < width; i++)
+        dst[i] = src[3][i] << 6;
+}
+
+static void planar_rgb_to_uv(uint8_t *_dstU, uint8_t *_dstV, const uint8_t *src[4], int width, int32_t *rgb2yuv)
 {
+    uint16_t *dstU = (uint16_t *)_dstU;
+    uint16_t *dstV = (uint16_t *)_dstV;
+    int32_t ru = rgb2yuv[RU_IDX], gu = rgb2yuv[GU_IDX], bu = rgb2yuv[BU_IDX];
+    int32_t rv = rgb2yuv[RV_IDX], gv = rgb2yuv[GV_IDX], bv = rgb2yuv[BV_IDX];
     int i;
     for (i = 0; i < width; i++) {
         int g = src[0][i];
         int b = src[1][i];
         int r = src[2][i];
 
-        dstU[i] = (RU*r + GU*g + BU*b + (0x4001<<(RGB2YUV_SHIFT-7))) >> (RGB2YUV_SHIFT-6);
-        dstV[i] = (RV*r + GV*g + BV*b + (0x4001<<(RGB2YUV_SHIFT-7))) >> (RGB2YUV_SHIFT-6);
+        dstU[i] = (ru*r + gu*g + bu*b + (0x4001<<(RGB2YUV_SHIFT-7))) >> (RGB2YUV_SHIFT-6);
+        dstV[i] = (rv*r + gv*g + bv*b + (0x4001<<(RGB2YUV_SHIFT-7))) >> (RGB2YUV_SHIFT-6);
     }
 }
 
 #define rdpx(src) \
     is_be ? AV_RB16(src) : AV_RL16(src)
 static av_always_inline void planar_rgb16_to_y(uint8_t *_dst, const uint8_t *_src[4],
-                                               int width, int bpc, int is_be)
+                                               int width, int bpc, int is_be, int32_t *rgb2yuv)
 {
     int i;
     const uint16_t **src = (const uint16_t **)_src;
     uint16_t *dst        = (uint16_t *)_dst;
+    int32_t ry = rgb2yuv[RY_IDX], gy = rgb2yuv[GY_IDX], by = rgb2yuv[BY_IDX];
+    int shift = bpc < 16 ? bpc : 14;
     for (i = 0; i < width; i++) {
         int g = rdpx(src[0] + i);
         int b = rdpx(src[1] + i);
         int r = rdpx(src[2] + i);
 
-        dst[i] = ((RY * r + GY * g + BY * b + (33 << (RGB2YUV_SHIFT + bpc - 9))) >> RGB2YUV_SHIFT);
+        dst[i] = ((ry*r + gy*g + by*b + (33 << (RGB2YUV_SHIFT + bpc - 9))) >> (RGB2YUV_SHIFT + shift - 14));
     }
 }
 
-static void planar_rgb9le_to_y(uint8_t *dst, const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_y(dst, src, w, 9, 0);
-}
-
-static void planar_rgb9be_to_y(uint8_t *dst, const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_y(dst, src, w, 9, 1);
-}
-
-static void planar_rgb10le_to_y(uint8_t *dst, const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_y(dst, src, w, 10, 0);
-}
-
-static void planar_rgb10be_to_y(uint8_t *dst, const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_y(dst, src, w, 10, 1);
-}
-
-static void planar_rgb12le_to_y(uint8_t *dst, const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_y(dst, src, w, 12, 0);
-}
-
-static void planar_rgb12be_to_y(uint8_t *dst, const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_y(dst, src, w, 12, 1);
-}
-
-static void planar_rgb14le_to_y(uint8_t *dst, const uint8_t *src[4], int w)
+static av_always_inline void planar_rgb16_to_a(uint8_t *_dst, const uint8_t *_src[4],
+                                               int width, int bpc, int is_be, int32_t *rgb2yuv)
 {
-    planar_rgb16_to_y(dst, src, w, 14, 0);
-}
-
-static void planar_rgb14be_to_y(uint8_t *dst, const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_y(dst, src, w, 14, 1);
-}
-
-static void planar_rgb16le_to_y(uint8_t *dst, const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_y(dst, src, w, 16, 0);
-}
+    int i;
+    const uint16_t **src = (const uint16_t **)_src;
+    uint16_t *dst        = (uint16_t *)_dst;
+    int shift = bpc < 16 ? bpc : 14;
 
-static void planar_rgb16be_to_y(uint8_t *dst, const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_y(dst, src, w, 16, 1);
+    for (i = 0; i < width; i++) {
+        dst[i] = rdpx(src[3] + i) << (14 - shift);
+    }
 }
 
 static av_always_inline void planar_rgb16_to_uv(uint8_t *_dstU, uint8_t *_dstV,
                                                 const uint8_t *_src[4], int width,
-                                                int bpc, int is_be)
+                                                int bpc, int is_be, int32_t *rgb2yuv)
 {
     int i;
     const uint16_t **src = (const uint16_t **)_src;
     uint16_t *dstU       = (uint16_t *)_dstU;
     uint16_t *dstV       = (uint16_t *)_dstV;
+    int32_t ru = rgb2yuv[RU_IDX], gu = rgb2yuv[GU_IDX], bu = rgb2yuv[BU_IDX];
+    int32_t rv = rgb2yuv[RV_IDX], gv = rgb2yuv[GV_IDX], bv = rgb2yuv[BV_IDX];
+    int shift = bpc < 16 ? bpc : 14;
     for (i = 0; i < width; i++) {
         int g = rdpx(src[0] + i);
         int b = rdpx(src[1] + i);
         int r = rdpx(src[2] + i);
 
-        dstU[i] = (RU * r + GU * g + BU * b + (257 << (RGB2YUV_SHIFT + bpc - 9))) >> RGB2YUV_SHIFT;
-        dstV[i] = (RV * r + GV * g + BV * b + (257 << (RGB2YUV_SHIFT + bpc - 9))) >> RGB2YUV_SHIFT;
+        dstU[i] = (ru*r + gu*g + bu*b + (257 << (RGB2YUV_SHIFT + bpc - 9))) >> (RGB2YUV_SHIFT + shift - 14);
+        dstV[i] = (rv*r + gv*g + bv*b + (257 << (RGB2YUV_SHIFT + bpc - 9))) >> (RGB2YUV_SHIFT + shift - 14);
     }
 }
 #undef rdpx
 
-static void planar_rgb9le_to_uv(uint8_t *dstU, uint8_t *dstV,
-                                const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_uv(dstU, dstV, src, w, 9, 0);
-}
-
-static void planar_rgb9be_to_uv(uint8_t *dstU, uint8_t *dstV,
-                                const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_uv(dstU, dstV, src, w, 9, 1);
-}
-
-static void planar_rgb10le_to_uv(uint8_t *dstU, uint8_t *dstV,
-                                 const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_uv(dstU, dstV, src, w, 10, 0);
-}
-
-static void planar_rgb10be_to_uv(uint8_t *dstU, uint8_t *dstV,
-                                 const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_uv(dstU, dstV, src, w, 10, 1);
-}
-
-static void planar_rgb12le_to_uv(uint8_t *dstU, uint8_t *dstV,
-                                 const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_uv(dstU, dstV, src, w, 12, 0);
-}
-
-static void planar_rgb12be_to_uv(uint8_t *dstU, uint8_t *dstV,
-                                 const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_uv(dstU, dstV, src, w, 12, 1);
-}
-
-static void planar_rgb14le_to_uv(uint8_t *dstU, uint8_t *dstV,
-                                 const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_uv(dstU, dstV, src, w, 14, 0);
-}
-
-static void planar_rgb14be_to_uv(uint8_t *dstU, uint8_t *dstV,
-                                 const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_uv(dstU, dstV, src, w, 14, 1);
-}
-
-static void planar_rgb16le_to_uv(uint8_t *dstU, uint8_t *dstV,
-                                 const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_uv(dstU, dstV, src, w, 16, 0);
-}
-
-static void planar_rgb16be_to_uv(uint8_t *dstU, uint8_t *dstV,
-                                 const uint8_t *src[4], int w)
-{
-    planar_rgb16_to_uv(dstU, dstV, src, w, 16, 1);
-}
+#define rgb9plus_planar_funcs_endian(nbits, endian_name, endian)                                    \
+static void planar_rgb##nbits##endian_name##_to_y(uint8_t *dst, const uint8_t *src[4],              \
+                                                  int w, int32_t *rgb2yuv)                          \
+{                                                                                                   \
+    planar_rgb16_to_y(dst, src, w, nbits, endian, rgb2yuv);                                         \
+}                                                                                                   \
+static void planar_rgb##nbits##endian_name##_to_a(uint8_t *dst, const uint8_t *src[4],              \
+                                                  int w, int32_t *rgb2yuv)                          \
+{                                                                                                   \
+    planar_rgb16_to_a(dst, src, w, nbits, endian, rgb2yuv);                                         \
+}                                                                                                   \
+static void planar_rgb##nbits##endian_name##_to_uv(uint8_t *dstU, uint8_t *dstV,                    \
+                                                   const uint8_t *src[4], int w, int32_t *rgb2yuv)  \
+{                                                                                                   \
+    planar_rgb16_to_uv(dstU, dstV, src, w, nbits, endian, rgb2yuv);                                 \
+}                                                                                                   \
+
+#define rgb9plus_planar_funcs(nbits)            \
+    rgb9plus_planar_funcs_endian(nbits, le, 0)  \
+    rgb9plus_planar_funcs_endian(nbits, be, 1)
+
+rgb9plus_planar_funcs(9)
+rgb9plus_planar_funcs(10)
+rgb9plus_planar_funcs(12)
+rgb9plus_planar_funcs(14)
+rgb9plus_planar_funcs(16)
 
 av_cold void ff_sws_init_input_funcs(SwsContext *c)
 {
-    enum PixelFormat srcFormat = c->srcFormat;
+    enum AVPixelFormat srcFormat = c->srcFormat;
 
     c->chrToYV12 = NULL;
     switch (srcFormat) {
-    case PIX_FMT_YUYV422:
+    case AV_PIX_FMT_YUYV422:
         c->chrToYV12 = yuy2ToUV_c;
         break;
-    case PIX_FMT_UYVY422:
+    case AV_PIX_FMT_YVYU422:
+        c->chrToYV12 = yvy2ToUV_c;
+        break;
+    case AV_PIX_FMT_UYVY422:
         c->chrToYV12 = uyvyToUV_c;
         break;
-    case PIX_FMT_NV12:
+    case AV_PIX_FMT_NV12:
         c->chrToYV12 = nv12ToUV_c;
         break;
-    case PIX_FMT_NV21:
+    case AV_PIX_FMT_NV21:
         c->chrToYV12 = nv21ToUV_c;
         break;
-    case PIX_FMT_RGB8:
-    case PIX_FMT_BGR8:
-    case PIX_FMT_PAL8:
-    case PIX_FMT_BGR4_BYTE:
-    case PIX_FMT_RGB4_BYTE:
+    case AV_PIX_FMT_RGB8:
+    case AV_PIX_FMT_BGR8:
+    case AV_PIX_FMT_PAL8:
+    case AV_PIX_FMT_BGR4_BYTE:
+    case AV_PIX_FMT_RGB4_BYTE:
         c->chrToYV12 = palToUV_c;
         break;
-    case PIX_FMT_GBRP9LE:
+    case AV_PIX_FMT_GBRP9LE:
         c->readChrPlanar = planar_rgb9le_to_uv;
         break;
-    case PIX_FMT_GBRP10LE:
+    case AV_PIX_FMT_GBRAP10LE:
+    case AV_PIX_FMT_GBRP10LE:
         c->readChrPlanar = planar_rgb10le_to_uv;
         break;
-    case PIX_FMT_GBRP12LE:
+    case AV_PIX_FMT_GBRAP12LE:
+    case AV_PIX_FMT_GBRP12LE:
         c->readChrPlanar = planar_rgb12le_to_uv;
         break;
-    case PIX_FMT_GBRP14LE:
+    case AV_PIX_FMT_GBRP14LE:
         c->readChrPlanar = planar_rgb14le_to_uv;
         break;
-    case PIX_FMT_GBRP16LE:
+    case AV_PIX_FMT_GBRAP16LE:
+    case AV_PIX_FMT_GBRP16LE:
         c->readChrPlanar = planar_rgb16le_to_uv;
         break;
-    case PIX_FMT_GBRP9BE:
+    case AV_PIX_FMT_GBRP9BE:
         c->readChrPlanar = planar_rgb9be_to_uv;
         break;
-    case PIX_FMT_GBRP10BE:
+    case AV_PIX_FMT_GBRAP10BE:
+    case AV_PIX_FMT_GBRP10BE:
         c->readChrPlanar = planar_rgb10be_to_uv;
         break;
-    case PIX_FMT_GBRP12BE:
+    case AV_PIX_FMT_GBRAP12BE:
+    case AV_PIX_FMT_GBRP12BE:
         c->readChrPlanar = planar_rgb12be_to_uv;
         break;
-    case PIX_FMT_GBRP14BE:
+    case AV_PIX_FMT_GBRP14BE:
         c->readChrPlanar = planar_rgb14be_to_uv;
         break;
-    case PIX_FMT_GBRP16BE:
+    case AV_PIX_FMT_GBRAP16BE:
+    case AV_PIX_FMT_GBRP16BE:
         c->readChrPlanar = planar_rgb16be_to_uv;
         break;
-    case PIX_FMT_GBRP:
+    case AV_PIX_FMT_GBRAP:
+    case AV_PIX_FMT_GBRP:
         c->readChrPlanar = planar_rgb_to_uv;
         break;
 #if HAVE_BIGENDIAN
-    case PIX_FMT_YUV444P9LE:
-    case PIX_FMT_YUV422P9LE:
-    case PIX_FMT_YUV420P9LE:
-    case PIX_FMT_YUV422P10LE:
-    case PIX_FMT_YUV444P10LE:
-    case PIX_FMT_YUV420P10LE:
-    case PIX_FMT_YUV422P12LE:
-    case PIX_FMT_YUV444P12LE:
-    case PIX_FMT_YUV420P12LE:
-    case PIX_FMT_YUV422P14LE:
-    case PIX_FMT_YUV444P14LE:
-    case PIX_FMT_YUV420P14LE:
-    case PIX_FMT_YUV420P16LE:
-    case PIX_FMT_YUV422P16LE:
-    case PIX_FMT_YUV444P16LE:
+    case AV_PIX_FMT_YUV444P9LE:
+    case AV_PIX_FMT_YUV422P9LE:
+    case AV_PIX_FMT_YUV420P9LE:
+    case AV_PIX_FMT_YUV422P10LE:
+    case AV_PIX_FMT_YUV440P10LE:
+    case AV_PIX_FMT_YUV444P10LE:
+    case AV_PIX_FMT_YUV420P10LE:
+    case AV_PIX_FMT_YUV422P12LE:
+    case AV_PIX_FMT_YUV440P12LE:
+    case AV_PIX_FMT_YUV444P12LE:
+    case AV_PIX_FMT_YUV420P12LE:
+    case AV_PIX_FMT_YUV422P14LE:
+    case AV_PIX_FMT_YUV444P14LE:
+    case AV_PIX_FMT_YUV420P14LE:
+    case AV_PIX_FMT_YUV420P16LE:
+    case AV_PIX_FMT_YUV422P16LE:
+    case AV_PIX_FMT_YUV444P16LE:
+
+    case AV_PIX_FMT_YUVA444P9LE:
+    case AV_PIX_FMT_YUVA422P9LE:
+    case AV_PIX_FMT_YUVA420P9LE:
+    case AV_PIX_FMT_YUVA444P10LE:
+    case AV_PIX_FMT_YUVA422P10LE:
+    case AV_PIX_FMT_YUVA420P10LE:
+    case AV_PIX_FMT_YUVA420P16LE:
+    case AV_PIX_FMT_YUVA422P16LE:
+    case AV_PIX_FMT_YUVA444P16LE:
         c->chrToYV12 = bswap16UV_c;
         break;
 #else
-    case PIX_FMT_YUV444P9BE:
-    case PIX_FMT_YUV422P9BE:
-    case PIX_FMT_YUV420P9BE:
-    case PIX_FMT_YUV444P10BE:
-    case PIX_FMT_YUV422P10BE:
-    case PIX_FMT_YUV420P10BE:
-    case PIX_FMT_YUV444P12BE:
-    case PIX_FMT_YUV422P12BE:
-    case PIX_FMT_YUV420P12BE:
-    case PIX_FMT_YUV444P14BE:
-    case PIX_FMT_YUV422P14BE:
-    case PIX_FMT_YUV420P14BE:
-    case PIX_FMT_YUV420P16BE:
-    case PIX_FMT_YUV422P16BE:
-    case PIX_FMT_YUV444P16BE:
+    case AV_PIX_FMT_YUV444P9BE:
+    case AV_PIX_FMT_YUV422P9BE:
+    case AV_PIX_FMT_YUV420P9BE:
+    case AV_PIX_FMT_YUV440P10BE:
+    case AV_PIX_FMT_YUV444P10BE:
+    case AV_PIX_FMT_YUV422P10BE:
+    case AV_PIX_FMT_YUV420P10BE:
+    case AV_PIX_FMT_YUV440P12BE:
+    case AV_PIX_FMT_YUV444P12BE:
+    case AV_PIX_FMT_YUV422P12BE:
+    case AV_PIX_FMT_YUV420P12BE:
+    case AV_PIX_FMT_YUV444P14BE:
+    case AV_PIX_FMT_YUV422P14BE:
+    case AV_PIX_FMT_YUV420P14BE:
+    case AV_PIX_FMT_YUV420P16BE:
+    case AV_PIX_FMT_YUV422P16BE:
+    case AV_PIX_FMT_YUV444P16BE:
+
+    case AV_PIX_FMT_YUVA444P9BE:
+    case AV_PIX_FMT_YUVA422P9BE:
+    case AV_PIX_FMT_YUVA420P9BE:
+    case AV_PIX_FMT_YUVA444P10BE:
+    case AV_PIX_FMT_YUVA422P10BE:
+    case AV_PIX_FMT_YUVA420P10BE:
+    case AV_PIX_FMT_YUVA420P16BE:
+    case AV_PIX_FMT_YUVA422P16BE:
+    case AV_PIX_FMT_YUVA444P16BE:
         c->chrToYV12 = bswap16UV_c;
         break;
 #endif
+    case AV_PIX_FMT_AYUV64LE:
+        c->chrToYV12 = read_ayuv64le_UV_c;
+        break;
+    case AV_PIX_FMT_P010LE:
+        c->chrToYV12 = p010LEToUV_c;
+        break;
+    case AV_PIX_FMT_P010BE:
+        c->chrToYV12 = p010BEToUV_c;
+        break;
     }
     if (c->chrSrcHSubSample) {
         switch (srcFormat) {
-        case PIX_FMT_RGBA64BE:
+        case AV_PIX_FMT_RGBA64BE:
             c->chrToYV12 = rgb64BEToUV_half_c;
             break;
-        case PIX_FMT_RGBA64LE:
+        case AV_PIX_FMT_RGBA64LE:
             c->chrToYV12 = rgb64LEToUV_half_c;
             break;
-        case PIX_FMT_RGB48BE:
+        case AV_PIX_FMT_BGRA64BE:
+            c->chrToYV12 = bgr64BEToUV_half_c;
+            break;
+        case AV_PIX_FMT_BGRA64LE:
+            c->chrToYV12 = bgr64LEToUV_half_c;
+            break;
+        case AV_PIX_FMT_RGB48BE:
             c->chrToYV12 = rgb48BEToUV_half_c;
             break;
-        case PIX_FMT_RGB48LE:
+        case AV_PIX_FMT_RGB48LE:
             c->chrToYV12 = rgb48LEToUV_half_c;
             break;
-        case PIX_FMT_BGR48BE:
+        case AV_PIX_FMT_BGR48BE:
             c->chrToYV12 = bgr48BEToUV_half_c;
             break;
-        case PIX_FMT_BGR48LE:
+        case AV_PIX_FMT_BGR48LE:
             c->chrToYV12 = bgr48LEToUV_half_c;
             break;
-        case PIX_FMT_RGB32:
+        case AV_PIX_FMT_RGB32:
             c->chrToYV12 = bgr32ToUV_half_c;
             break;
-        case PIX_FMT_RGB32_1:
+        case AV_PIX_FMT_RGB32_1:
             c->chrToYV12 = bgr321ToUV_half_c;
             break;
-        case PIX_FMT_BGR24:
+        case AV_PIX_FMT_BGR24:
             c->chrToYV12 = bgr24ToUV_half_c;
             break;
-        case PIX_FMT_BGR565LE:
+        case AV_PIX_FMT_BGR565LE:
             c->chrToYV12 = bgr16leToUV_half_c;
             break;
-        case PIX_FMT_BGR565BE:
+        case AV_PIX_FMT_BGR565BE:
             c->chrToYV12 = bgr16beToUV_half_c;
             break;
-        case PIX_FMT_BGR555LE:
+        case AV_PIX_FMT_BGR555LE:
             c->chrToYV12 = bgr15leToUV_half_c;
             break;
-        case PIX_FMT_BGR555BE:
+        case AV_PIX_FMT_BGR555BE:
             c->chrToYV12 = bgr15beToUV_half_c;
             break;
-        case PIX_FMT_GBR24P  :
+        case AV_PIX_FMT_GBRAP:
+        case AV_PIX_FMT_GBRP:
             c->chrToYV12 = gbr24pToUV_half_c;
             break;
-        case PIX_FMT_BGR444LE:
+        case AV_PIX_FMT_BGR444LE:
             c->chrToYV12 = bgr12leToUV_half_c;
             break;
-        case PIX_FMT_BGR444BE:
+        case AV_PIX_FMT_BGR444BE:
             c->chrToYV12 = bgr12beToUV_half_c;
             break;
-        case PIX_FMT_BGR32:
+        case AV_PIX_FMT_BGR32:
             c->chrToYV12 = rgb32ToUV_half_c;
             break;
-        case PIX_FMT_BGR32_1:
+        case AV_PIX_FMT_BGR32_1:
             c->chrToYV12 = rgb321ToUV_half_c;
             break;
-        case PIX_FMT_RGB24:
+        case AV_PIX_FMT_RGB24:
             c->chrToYV12 = rgb24ToUV_half_c;
             break;
-        case PIX_FMT_RGB565LE:
+        case AV_PIX_FMT_RGB565LE:
             c->chrToYV12 = rgb16leToUV_half_c;
             break;
-        case PIX_FMT_RGB565BE:
+        case AV_PIX_FMT_RGB565BE:
             c->chrToYV12 = rgb16beToUV_half_c;
             break;
-        case PIX_FMT_RGB555LE:
+        case AV_PIX_FMT_RGB555LE:
             c->chrToYV12 = rgb15leToUV_half_c;
             break;
-        case PIX_FMT_RGB555BE:
+        case AV_PIX_FMT_RGB555BE:
             c->chrToYV12 = rgb15beToUV_half_c;
             break;
-        case PIX_FMT_RGB444LE:
+        case AV_PIX_FMT_RGB444LE:
             c->chrToYV12 = rgb12leToUV_half_c;
             break;
-        case PIX_FMT_RGB444BE:
+        case AV_PIX_FMT_RGB444BE:
             c->chrToYV12 = rgb12beToUV_half_c;
             break;
         }
     } else {
         switch (srcFormat) {
-        case PIX_FMT_RGBA64BE:
+        case AV_PIX_FMT_RGBA64BE:
             c->chrToYV12 = rgb64BEToUV_c;
             break;
-        case PIX_FMT_RGBA64LE:
+        case AV_PIX_FMT_RGBA64LE:
             c->chrToYV12 = rgb64LEToUV_c;
             break;
-        case PIX_FMT_RGB48BE:
+        case AV_PIX_FMT_BGRA64BE:
+            c->chrToYV12 = bgr64BEToUV_c;
+            break;
+        case AV_PIX_FMT_BGRA64LE:
+            c->chrToYV12 = bgr64LEToUV_c;
+            break;
+        case AV_PIX_FMT_RGB48BE:
             c->chrToYV12 = rgb48BEToUV_c;
             break;
-        case PIX_FMT_RGB48LE:
+        case AV_PIX_FMT_RGB48LE:
             c->chrToYV12 = rgb48LEToUV_c;
             break;
-        case PIX_FMT_BGR48BE:
+        case AV_PIX_FMT_BGR48BE:
             c->chrToYV12 = bgr48BEToUV_c;
             break;
-        case PIX_FMT_BGR48LE:
+        case AV_PIX_FMT_BGR48LE:
             c->chrToYV12 = bgr48LEToUV_c;
             break;
-        case PIX_FMT_RGB32:
+        case AV_PIX_FMT_RGB32:
             c->chrToYV12 = bgr32ToUV_c;
             break;
-        case PIX_FMT_RGB32_1:
+        case AV_PIX_FMT_RGB32_1:
             c->chrToYV12 = bgr321ToUV_c;
             break;
-        case PIX_FMT_BGR24:
+        case AV_PIX_FMT_BGR24:
             c->chrToYV12 = bgr24ToUV_c;
             break;
-        case PIX_FMT_BGR565LE:
+        case AV_PIX_FMT_BGR565LE:
             c->chrToYV12 = bgr16leToUV_c;
             break;
-        case PIX_FMT_BGR565BE:
+        case AV_PIX_FMT_BGR565BE:
             c->chrToYV12 = bgr16beToUV_c;
             break;
-        case PIX_FMT_BGR555LE:
+        case AV_PIX_FMT_BGR555LE:
             c->chrToYV12 = bgr15leToUV_c;
             break;
-        case PIX_FMT_BGR555BE:
+        case AV_PIX_FMT_BGR555BE:
             c->chrToYV12 = bgr15beToUV_c;
             break;
-        case PIX_FMT_BGR444LE:
+        case AV_PIX_FMT_BGR444LE:
             c->chrToYV12 = bgr12leToUV_c;
             break;
-        case PIX_FMT_BGR444BE:
+        case AV_PIX_FMT_BGR444BE:
             c->chrToYV12 = bgr12beToUV_c;
             break;
-        case PIX_FMT_BGR32:
+        case AV_PIX_FMT_BGR32:
             c->chrToYV12 = rgb32ToUV_c;
             break;
-        case PIX_FMT_BGR32_1:
+        case AV_PIX_FMT_BGR32_1:
             c->chrToYV12 = rgb321ToUV_c;
             break;
-        case PIX_FMT_RGB24:
+        case AV_PIX_FMT_RGB24:
             c->chrToYV12 = rgb24ToUV_c;
             break;
-        case PIX_FMT_RGB565LE:
+        case AV_PIX_FMT_RGB565LE:
             c->chrToYV12 = rgb16leToUV_c;
             break;
-        case PIX_FMT_RGB565BE:
+        case AV_PIX_FMT_RGB565BE:
             c->chrToYV12 = rgb16beToUV_c;
             break;
-        case PIX_FMT_RGB555LE:
+        case AV_PIX_FMT_RGB555LE:
             c->chrToYV12 = rgb15leToUV_c;
             break;
-        case PIX_FMT_RGB555BE:
+        case AV_PIX_FMT_RGB555BE:
             c->chrToYV12 = rgb15beToUV_c;
             break;
-        case PIX_FMT_RGB444LE:
+        case AV_PIX_FMT_RGB444LE:
             c->chrToYV12 = rgb12leToUV_c;
             break;
-        case PIX_FMT_RGB444BE:
+        case AV_PIX_FMT_RGB444BE:
             c->chrToYV12 = rgb12beToUV_c;
             break;
         }
@@ -1090,187 +1257,272 @@ av_cold void ff_sws_init_input_funcs(SwsContext *c)
     c->lumToYV12 = NULL;
     c->alpToYV12 = NULL;
     switch (srcFormat) {
-    case PIX_FMT_GBRP9LE:
+    case AV_PIX_FMT_GBRP9LE:
         c->readLumPlanar = planar_rgb9le_to_y;
         break;
-    case PIX_FMT_GBRP10LE:
+    case AV_PIX_FMT_GBRAP10LE:
+        c->readAlpPlanar = planar_rgb10le_to_a;
+    case AV_PIX_FMT_GBRP10LE:
         c->readLumPlanar = planar_rgb10le_to_y;
         break;
-    case PIX_FMT_GBRP12LE:
+    case AV_PIX_FMT_GBRAP12LE:
+        c->readAlpPlanar = planar_rgb12le_to_a;
+    case AV_PIX_FMT_GBRP12LE:
         c->readLumPlanar = planar_rgb12le_to_y;
         break;
-    case PIX_FMT_GBRP14LE:
+    case AV_PIX_FMT_GBRP14LE:
         c->readLumPlanar = planar_rgb14le_to_y;
         break;
-    case PIX_FMT_GBRP16LE:
+    case AV_PIX_FMT_GBRAP16LE:
+        c->readAlpPlanar = planar_rgb16le_to_a;
+    case AV_PIX_FMT_GBRP16LE:
         c->readLumPlanar = planar_rgb16le_to_y;
         break;
-    case PIX_FMT_GBRP9BE:
+    case AV_PIX_FMT_GBRP9BE:
         c->readLumPlanar = planar_rgb9be_to_y;
         break;
-    case PIX_FMT_GBRP10BE:
+    case AV_PIX_FMT_GBRAP10BE:
+        c->readAlpPlanar = planar_rgb10be_to_a;
+    case AV_PIX_FMT_GBRP10BE:
         c->readLumPlanar = planar_rgb10be_to_y;
         break;
-    case PIX_FMT_GBRP12BE:
+    case AV_PIX_FMT_GBRAP12BE:
+        c->readAlpPlanar = planar_rgb12be_to_a;
+    case AV_PIX_FMT_GBRP12BE:
         c->readLumPlanar = planar_rgb12be_to_y;
         break;
-    case PIX_FMT_GBRP14BE:
+    case AV_PIX_FMT_GBRP14BE:
         c->readLumPlanar = planar_rgb14be_to_y;
         break;
-    case PIX_FMT_GBRP16BE:
+    case AV_PIX_FMT_GBRAP16BE:
+        c->readAlpPlanar = planar_rgb16be_to_a;
+    case AV_PIX_FMT_GBRP16BE:
         c->readLumPlanar = planar_rgb16be_to_y;
         break;
-    case PIX_FMT_GBRP:
+    case AV_PIX_FMT_GBRAP:
+        c->readAlpPlanar = planar_rgb_to_a;
+    case AV_PIX_FMT_GBRP:
         c->readLumPlanar = planar_rgb_to_y;
         break;
 #if HAVE_BIGENDIAN
-    case PIX_FMT_YUV444P9LE:
-    case PIX_FMT_YUV422P9LE:
-    case PIX_FMT_YUV420P9LE:
-    case PIX_FMT_YUV444P10LE:
-    case PIX_FMT_YUV422P10LE:
-    case PIX_FMT_YUV420P10LE:
-    case PIX_FMT_YUV444P12LE:
-    case PIX_FMT_YUV422P12LE:
-    case PIX_FMT_YUV420P12LE:
-    case PIX_FMT_YUV444P14LE:
-    case PIX_FMT_YUV422P14LE:
-    case PIX_FMT_YUV420P14LE:
-    case PIX_FMT_YUV420P16LE:
-    case PIX_FMT_YUV422P16LE:
-    case PIX_FMT_YUV444P16LE:
-    case PIX_FMT_GRAY16LE:
+    case AV_PIX_FMT_YUV444P9LE:
+    case AV_PIX_FMT_YUV422P9LE:
+    case AV_PIX_FMT_YUV420P9LE:
+    case AV_PIX_FMT_YUV444P10LE:
+    case AV_PIX_FMT_YUV440P10LE:
+    case AV_PIX_FMT_YUV422P10LE:
+    case AV_PIX_FMT_YUV420P10LE:
+    case AV_PIX_FMT_YUV444P12LE:
+    case AV_PIX_FMT_YUV440P12LE:
+    case AV_PIX_FMT_YUV422P12LE:
+    case AV_PIX_FMT_YUV420P12LE:
+    case AV_PIX_FMT_YUV444P14LE:
+    case AV_PIX_FMT_YUV422P14LE:
+    case AV_PIX_FMT_YUV420P14LE:
+    case AV_PIX_FMT_YUV420P16LE:
+    case AV_PIX_FMT_YUV422P16LE:
+    case AV_PIX_FMT_YUV444P16LE:
+
+    case AV_PIX_FMT_GRAY10LE:
+    case AV_PIX_FMT_GRAY12LE:
+    case AV_PIX_FMT_GRAY16LE:
         c->lumToYV12 = bswap16Y_c;
         break;
+    case AV_PIX_FMT_YUVA444P9LE:
+    case AV_PIX_FMT_YUVA422P9LE:
+    case AV_PIX_FMT_YUVA420P9LE:
+    case AV_PIX_FMT_YUVA444P10LE:
+    case AV_PIX_FMT_YUVA422P10LE:
+    case AV_PIX_FMT_YUVA420P10LE:
+    case AV_PIX_FMT_YUVA420P16LE:
+    case AV_PIX_FMT_YUVA422P16LE:
+    case AV_PIX_FMT_YUVA444P16LE:
+        c->lumToYV12 = bswap16Y_c;
+        c->alpToYV12 = bswap16Y_c;
+        break;
 #else
-    case PIX_FMT_YUV444P9BE:
-    case PIX_FMT_YUV422P9BE:
-    case PIX_FMT_YUV420P9BE:
-    case PIX_FMT_YUV444P10BE:
-    case PIX_FMT_YUV422P10BE:
-    case PIX_FMT_YUV420P10BE:
-    case PIX_FMT_YUV444P12BE:
-    case PIX_FMT_YUV422P12BE:
-    case PIX_FMT_YUV420P12BE:
-    case PIX_FMT_YUV444P14BE:
-    case PIX_FMT_YUV422P14BE:
-    case PIX_FMT_YUV420P14BE:
-    case PIX_FMT_YUV420P16BE:
-    case PIX_FMT_YUV422P16BE:
-    case PIX_FMT_YUV444P16BE:
-    case PIX_FMT_GRAY16BE:
+    case AV_PIX_FMT_YUV444P9BE:
+    case AV_PIX_FMT_YUV422P9BE:
+    case AV_PIX_FMT_YUV420P9BE:
+    case AV_PIX_FMT_YUV444P10BE:
+    case AV_PIX_FMT_YUV440P10BE:
+    case AV_PIX_FMT_YUV422P10BE:
+    case AV_PIX_FMT_YUV420P10BE:
+    case AV_PIX_FMT_YUV444P12BE:
+    case AV_PIX_FMT_YUV440P12BE:
+    case AV_PIX_FMT_YUV422P12BE:
+    case AV_PIX_FMT_YUV420P12BE:
+    case AV_PIX_FMT_YUV444P14BE:
+    case AV_PIX_FMT_YUV422P14BE:
+    case AV_PIX_FMT_YUV420P14BE:
+    case AV_PIX_FMT_YUV420P16BE:
+    case AV_PIX_FMT_YUV422P16BE:
+    case AV_PIX_FMT_YUV444P16BE:
+
+    case AV_PIX_FMT_GRAY10BE:
+    case AV_PIX_FMT_GRAY12BE:
+    case AV_PIX_FMT_GRAY16BE:
         c->lumToYV12 = bswap16Y_c;
         break;
+    case AV_PIX_FMT_YUVA444P9BE:
+    case AV_PIX_FMT_YUVA422P9BE:
+    case AV_PIX_FMT_YUVA420P9BE:
+    case AV_PIX_FMT_YUVA444P10BE:
+    case AV_PIX_FMT_YUVA422P10BE:
+    case AV_PIX_FMT_YUVA420P10BE:
+    case AV_PIX_FMT_YUVA420P16BE:
+    case AV_PIX_FMT_YUVA422P16BE:
+    case AV_PIX_FMT_YUVA444P16BE:
+        c->lumToYV12 = bswap16Y_c;
+        c->alpToYV12 = bswap16Y_c;
+        break;
 #endif
-    case PIX_FMT_YUYV422:
-    case PIX_FMT_Y400A:
+    case AV_PIX_FMT_YA16LE:
+        c->lumToYV12 = read_ya16le_gray_c;
+        break;
+    case AV_PIX_FMT_YA16BE:
+        c->lumToYV12 = read_ya16be_gray_c;
+        break;
+    case AV_PIX_FMT_AYUV64LE:
+        c->lumToYV12 = read_ayuv64le_Y_c;
+        break;
+    case AV_PIX_FMT_YUYV422:
+    case AV_PIX_FMT_YVYU422:
+    case AV_PIX_FMT_YA8:
         c->lumToYV12 = yuy2ToY_c;
         break;
-    case PIX_FMT_UYVY422:
+    case AV_PIX_FMT_UYVY422:
         c->lumToYV12 = uyvyToY_c;
         break;
-    case PIX_FMT_BGR24:
+    case AV_PIX_FMT_BGR24:
         c->lumToYV12 = bgr24ToY_c;
         break;
-    case PIX_FMT_BGR565LE:
+    case AV_PIX_FMT_BGR565LE:
         c->lumToYV12 = bgr16leToY_c;
         break;
-    case PIX_FMT_BGR565BE:
+    case AV_PIX_FMT_BGR565BE:
         c->lumToYV12 = bgr16beToY_c;
         break;
-    case PIX_FMT_BGR555LE:
+    case AV_PIX_FMT_BGR555LE:
         c->lumToYV12 = bgr15leToY_c;
         break;
-    case PIX_FMT_BGR555BE:
+    case AV_PIX_FMT_BGR555BE:
         c->lumToYV12 = bgr15beToY_c;
         break;
-    case PIX_FMT_BGR444LE:
+    case AV_PIX_FMT_BGR444LE:
         c->lumToYV12 = bgr12leToY_c;
         break;
-    case PIX_FMT_BGR444BE:
+    case AV_PIX_FMT_BGR444BE:
         c->lumToYV12 = bgr12beToY_c;
         break;
-    case PIX_FMT_RGB24:
+    case AV_PIX_FMT_RGB24:
         c->lumToYV12 = rgb24ToY_c;
         break;
-    case PIX_FMT_RGB565LE:
+    case AV_PIX_FMT_RGB565LE:
         c->lumToYV12 = rgb16leToY_c;
         break;
-    case PIX_FMT_RGB565BE:
+    case AV_PIX_FMT_RGB565BE:
         c->lumToYV12 = rgb16beToY_c;
         break;
-    case PIX_FMT_RGB555LE:
+    case AV_PIX_FMT_RGB555LE:
         c->lumToYV12 = rgb15leToY_c;
         break;
-    case PIX_FMT_RGB555BE:
+    case AV_PIX_FMT_RGB555BE:
         c->lumToYV12 = rgb15beToY_c;
         break;
-    case PIX_FMT_RGB444LE:
+    case AV_PIX_FMT_RGB444LE:
         c->lumToYV12 = rgb12leToY_c;
         break;
-    case PIX_FMT_RGB444BE:
+    case AV_PIX_FMT_RGB444BE:
         c->lumToYV12 = rgb12beToY_c;
         break;
-    case PIX_FMT_RGB8:
-    case PIX_FMT_BGR8:
-    case PIX_FMT_PAL8:
-    case PIX_FMT_BGR4_BYTE:
-    case PIX_FMT_RGB4_BYTE:
+    case AV_PIX_FMT_RGB8:
+    case AV_PIX_FMT_BGR8:
+    case AV_PIX_FMT_PAL8:
+    case AV_PIX_FMT_BGR4_BYTE:
+    case AV_PIX_FMT_RGB4_BYTE:
         c->lumToYV12 = palToY_c;
         break;
-    case PIX_FMT_MONOBLACK:
+    case AV_PIX_FMT_MONOBLACK:
         c->lumToYV12 = monoblack2Y_c;
         break;
-    case PIX_FMT_MONOWHITE:
+    case AV_PIX_FMT_MONOWHITE:
         c->lumToYV12 = monowhite2Y_c;
         break;
-    case PIX_FMT_RGB32:
+    case AV_PIX_FMT_RGB32:
         c->lumToYV12 = bgr32ToY_c;
         break;
-    case PIX_FMT_RGB32_1:
+    case AV_PIX_FMT_RGB32_1:
         c->lumToYV12 = bgr321ToY_c;
         break;
-    case PIX_FMT_BGR32:
+    case AV_PIX_FMT_BGR32:
         c->lumToYV12 = rgb32ToY_c;
         break;
-    case PIX_FMT_BGR32_1:
+    case AV_PIX_FMT_BGR32_1:
         c->lumToYV12 = rgb321ToY_c;
         break;
-    case PIX_FMT_RGB48BE:
+    case AV_PIX_FMT_RGB48BE:
         c->lumToYV12 = rgb48BEToY_c;
         break;
-    case PIX_FMT_RGB48LE:
+    case AV_PIX_FMT_RGB48LE:
         c->lumToYV12 = rgb48LEToY_c;
         break;
-    case PIX_FMT_BGR48BE:
+    case AV_PIX_FMT_BGR48BE:
         c->lumToYV12 = bgr48BEToY_c;
         break;
-    case PIX_FMT_BGR48LE:
+    case AV_PIX_FMT_BGR48LE:
         c->lumToYV12 = bgr48LEToY_c;
         break;
-    case PIX_FMT_RGBA64BE:
+    case AV_PIX_FMT_RGBA64BE:
         c->lumToYV12 = rgb64BEToY_c;
         break;
-    case PIX_FMT_RGBA64LE:
+    case AV_PIX_FMT_RGBA64LE:
         c->lumToYV12 = rgb64LEToY_c;
         break;
+    case AV_PIX_FMT_BGRA64BE:
+        c->lumToYV12 = bgr64BEToY_c;
+        break;
+    case AV_PIX_FMT_BGRA64LE:
+        c->lumToYV12 = bgr64LEToY_c;
+        break;
+    case AV_PIX_FMT_P010LE:
+        c->lumToYV12 = p010LEToY_c;
+        break;
+    case AV_PIX_FMT_P010BE:
+        c->lumToYV12 = p010BEToY_c;
+        break;
     }
-    if (c->alpPixBuf) {
+    if (c->needAlpha) {
+        if (is16BPS(srcFormat) || isNBPS(srcFormat)) {
+            if (HAVE_BIGENDIAN == !isBE(srcFormat) && !c->readAlpPlanar)
+                c->alpToYV12 = bswap16Y_c;
+        }
         switch (srcFormat) {
-        case PIX_FMT_RGBA64LE:
-        case PIX_FMT_RGBA64BE:  c->alpToYV12 = rgba64ToA_c; break;
-        case PIX_FMT_BGRA:
-        case PIX_FMT_RGBA:
+        case AV_PIX_FMT_BGRA64LE:
+        case AV_PIX_FMT_RGBA64LE:  c->alpToYV12 = rgba64leToA_c; break;
+        case AV_PIX_FMT_BGRA64BE:
+        case AV_PIX_FMT_RGBA64BE:  c->alpToYV12 = rgba64beToA_c; break;
+        case AV_PIX_FMT_BGRA:
+        case AV_PIX_FMT_RGBA:
             c->alpToYV12 = rgbaToA_c;
             break;
-        case PIX_FMT_ABGR:
-        case PIX_FMT_ARGB:
+        case AV_PIX_FMT_ABGR:
+        case AV_PIX_FMT_ARGB:
             c->alpToYV12 = abgrToA_c;
             break;
-        case PIX_FMT_Y400A:
+        case AV_PIX_FMT_YA8:
             c->alpToYV12 = uyvyToY_c;
             break;
-        case PIX_FMT_PAL8 :
+        case AV_PIX_FMT_YA16LE:
+            c->alpToYV12 = read_ya16le_alpha_c;
+            break;
+        case AV_PIX_FMT_YA16BE:
+            c->alpToYV12 = read_ya16be_alpha_c;
+            break;
+        case AV_PIX_FMT_AYUV64LE:
+            c->alpToYV12 = read_ayuv64le_A_c;
+            break;
+        case AV_PIX_FMT_PAL8 :
             c->alpToYV12 = palToA_c;
             break;
         }