I422ToRGB24, I422ToRAW, I422ToRGB24MatrixFilter conversion functions added.

- YUV to RGB use linear for first and last row.
- add assert(yuvconstants)
- rename pointers to match row functions.
- use macros that match row functions.
- use 12 bit upsampler for conversions of 10 and 12 bits

Cortex A53 AArch32
I420ToRGB24_Opt (3627 ms)
I422ToRGB24_Opt (4099 ms)
I444ToRGB24_Opt (4186 ms)
I420ToRGB24Filter_Opt (5451 ms)
I422ToRGB24Filter_Opt (5430 ms)

AVX2
Was I420ToRGB24Filter_Opt (583 ms)
Now I420ToRGB24Filter_Opt (560 ms)

Neon Cortex A7
Was I420ToRGB24Filter_Opt (5447 ms)
Now I420ToRGB24Filter_Opt (5439 ms)

Bug: libyuv:938


Change-Id: I1731f2dd591073ae11a756f06574103ba0f803c7
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3906082
Reviewed-by: Justin Green <greenjustin@google.com>
Commit-Queue: Frank Barchard <fbarchard@chromium.org>

1 files changed, 43 insertions, 0 deletions

diff --git a/source/row_common.cc b/source/row_common.cc
index 4e1141f7..2531c85b 100644
--- a/source/row_common.cc
+++ b/source/row_common.cc
@@ -1863,6 +1863,23 @@ void I444ToARGBRow_C(const uint8_t* src_y,
   }
 }
 
+void I444ToRGB24Row_C(const uint8_t* src_y,
+                      const uint8_t* src_u,
+                      const uint8_t* src_v,
+                      uint8_t* rgb_buf,
+                      const struct YuvConstants* yuvconstants,
+                      int width) {
+  int x;
+  for (x = 0; x < width; ++x) {
+    YuvPixel(src_y[0], src_u[0], src_v[0], rgb_buf + 0, rgb_buf + 1,
+             rgb_buf + 2, yuvconstants);
+    src_y += 1;
+    src_u += 1;
+    src_v += 1;
+    rgb_buf += 3;  // Advance 1 pixel.
+  }
+}
+
 // Also used for 420
 void I422ToARGBRow_C(const uint8_t* src_y,
                      const uint8_t* src_u,
@@ -4061,6 +4078,32 @@ void I422ToRGB24Row_AVX2(const uint8_t* src_y,
 }
 #endif
 
+#if defined(HAS_I444TORGB24ROW_AVX2)
+void I444ToRGB24Row_AVX2(const uint8_t* src_y,
+                         const uint8_t* src_u,
+                         const uint8_t* src_v,
+                         uint8_t* dst_rgb24,
+                         const struct YuvConstants* yuvconstants,
+                         int width) {
+  // Row buffer for intermediate ARGB pixels.
+  SIMD_ALIGNED(uint8_t row[MAXTWIDTH * 4]);
+  while (width > 0) {
+    int twidth = width > MAXTWIDTH ? MAXTWIDTH : width;
+    I444ToARGBRow_AVX2(src_y, src_u, src_v, row, yuvconstants, twidth);
+#if defined(HAS_ARGBTORGB24ROW_AVX2)
+    ARGBToRGB24Row_AVX2(row, dst_rgb24, twidth);
+#else
+    ARGBToRGB24Row_SSSE3(row, dst_rgb24, twidth);
+#endif
+    src_y += twidth;
+    src_u += twidth;
+    src_v += twidth;
+    dst_rgb24 += twidth * 3;
+    width -= twidth;
+  }
+}
+#endif
+
 #if defined(HAS_NV12TORGB565ROW_AVX2)
 void NV12ToRGB565Row_AVX2(const uint8_t* src_y,
                           const uint8_t* src_uv,