AVX2 YUV alpha blender and improved unittests

include/libyuv/row.h

 /*
  *  Copyright 2011 The LibYuv Project Authors. All rights reserved.
  *
  *  Use of this source code is governed by a BSD-style license
  *  that can be found in the LICENSE file in the root of the source
  *  tree. An additional intellectual property rights grant can be found
  *  in the file PATENTS. All contributing project authors may
  *  be found in the AUTHORS file in the root of the source tree.
  */
 
@@ skipping 215 matching lines @@
 #define HAS_YUY2TOUV422ROW_AVX2
 #define HAS_YUY2TOUVROW_AVX2
 #define HAS_YUY2TOYROW_AVX2
 
 // Effects:
 #define HAS_ARGBADDROW_AVX2
 #define HAS_ARGBATTENUATEROW_AVX2
 #define HAS_ARGBMULTIPLYROW_AVX2
 #define HAS_ARGBSUBTRACTROW_AVX2
 #define HAS_ARGBUNATTENUATEROW_AVX2
+#define HAS_BLENDPLANEROW_AVX2
 #endif
 
 // The following are available for AVX2 Visual C and clangcl 32 bit:
 // TODO(fbarchard): Port to gcc.
 #if !defined(LIBYUV_DISABLE_X86) && defined(_M_IX86) && \
     (defined(VISUALC_HAS_AVX2) || defined(CLANG_HAS_AVX2))
 #define HAS_ARGB1555TOARGBROW_AVX2
 #define HAS_ARGB4444TOARGBROW_AVX2
 #define HAS_ARGBTOARGB1555ROW_AVX2
 #define HAS_ARGBTOARGB4444ROW_AVX2
 #define HAS_ARGBTORGB565ROW_AVX2
 #define HAS_I411TOARGBROW_AVX2
 #define HAS_I422TOARGB1555ROW_AVX2
 #define HAS_I422TOARGB4444ROW_AVX2
 #define HAS_I422TORGB565ROW_AVX2
 #define HAS_J400TOARGBROW_AVX2
 #define HAS_NV12TORGB565ROW_AVX2
 #define HAS_RGB565TOARGBROW_AVX2
 #endif
 
-// The following are available for 32 bit Visual C and clangcl 32 bit:
-// TODO(fbarchard): Port to gcc.
-#if !defined(LIBYUV_DISABLE_X86) && defined(_M_IX86)
-#define HAS_BLENDPLANEROW_SSSE3
-#endif
-
 // The following are also available on x64 Visual C.
 #if !defined(LIBYUV_DISABLE_X86) && defined (_M_X64) && \
     (!defined(__clang__) || defined(__SSSE3__))
 #define HAS_I422ALPHATOARGBROW_SSSE3
 #define HAS_I422TOARGBROW_SSSE3
 #endif
 
 // The following are available on Neon platforms:
 #if !defined(LIBYUV_DISABLE_NEON) && \
     (defined(__aarch64__) || defined(__ARM_NEON__) || defined(LIBYUV_NEON))
@@ skipping 1185 matching lines @@
 void ARGBBlendRow_SSSE3(const uint8* src_argb, const uint8* src_argb1,
                         uint8* dst_argb, int width);
 void ARGBBlendRow_NEON(const uint8* src_argb, const uint8* src_argb1,
                        uint8* dst_argb, int width);
 void ARGBBlendRow_C(const uint8* src_argb, const uint8* src_argb1,
                     uint8* dst_argb, int width);
 
 // Unattenuated planar alpha blend.
 void BlendPlaneRow_SSSE3(const uint8* src0, const uint8* src1,
                          const uint8* alpha, uint8* dst, int width);
+void BlendPlaneRow_Any_SSSE3(const uint8* src0, const uint8* src1,
+                             const uint8* alpha, uint8* dst, int width);
+void BlendPlaneRow_AVX2(const uint8* src0, const uint8* src1,
+                        const uint8* alpha, uint8* dst, int width);
+void BlendPlaneRow_Any_AVX2(const uint8* src0, const uint8* src1,
+                            const uint8* alpha, uint8* dst, int width);
 void BlendPlaneRow_C(const uint8* src0, const uint8* src1,
                      const uint8* alpha, uint8* dst, int width);
 
 // ARGB multiply images. Same API as Blend, but these require
 // pointer and width alignment for SSE2.
 void ARGBMultiplyRow_C(const uint8* src_argb, const uint8* src_argb1,
                        uint8* dst_argb, int width);
 void ARGBMultiplyRow_SSE2(const uint8* src_argb, const uint8* src_argb1,
                           uint8* dst_argb, int width);
 void ARGBMultiplyRow_Any_SSE2(const uint8* src_argb, const uint8* src_argb1,
@@ skipping 453 matching lines @@
 void ARGBLumaColorTableRow_SSSE3(const uint8* src_argb, uint8* dst_argb,
                                  int width,
                                  const uint8* luma, uint32 lumacoeff);
 
 #ifdef __cplusplus
 }  // extern "C"
 }  // namespace libyuv
 #endif
 
 #endif  // INCLUDE_LIBYUV_ROW_H_  NOLINT