Add MSA optimized ARGB Multiply/Add/Subtract row functions

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 372 matching lines @@
 #define HAS_I422TOARGBROW_MSA
 #define HAS_I422TORGBAROW_MSA
 #define HAS_I422ALPHATOARGBROW_MSA
 #define HAS_I422TORGB24ROW_MSA
 #define HAS_ARGBTORGB24ROW_MSA
 #define HAS_ARGBTORAWROW_MSA
 #define HAS_ARGBTORGB565ROW_MSA
 #define HAS_ARGBTOARGB1555ROW_MSA
 #define HAS_ARGBTOARGB4444ROW_MSA
 #define HAS_ARGBTOUV444ROW_MSA
+#define HAS_ARGBMULTIPLYROW_MSA
+#define HAS_ARGBADDROW_MSA
+#define HAS_ARGBSUBTRACTROW_MSA
 #endif
 
 #if defined(_MSC_VER) && !defined(__CLR_VER) && !defined(__clang__)
 #if defined(VISUALC_HAS_AVX2)
 #define SIMD_ALIGNED(var) __declspec(align(32)) var
 #else
 #define SIMD_ALIGNED(var) __declspec(align(16)) var
 #endif
 typedef __declspec(align(16)) int16 vec16[8];
 typedef __declspec(align(16)) int32 vec32[4];
@@ skipping 1399 matching lines @@
                               uint8* dst_argb,
                               int width);
 void ARGBMultiplyRow_NEON(const uint8* src_argb,
                           const uint8* src_argb1,
                           uint8* dst_argb,
                           int width);
 void ARGBMultiplyRow_Any_NEON(const uint8* src_argb,
                               const uint8* src_argb1,
                               uint8* dst_argb,
                               int width);
+void ARGBMultiplyRow_MSA(const uint8* src_argb,
+                         const uint8* src_argb1,
+                         uint8* dst_argb,
+                         int width);
+void ARGBMultiplyRow_Any_MSA(const uint8* src_argb,
+                             const uint8* src_argb1,
+                             uint8* dst_argb,
+                             int width);
 
 // ARGB add images.
 void ARGBAddRow_C(const uint8* src_argb,
                   const uint8* src_argb1,
                   uint8* dst_argb,
                   int width);
 void ARGBAddRow_SSE2(const uint8* src_argb,
                      const uint8* src_argb1,
                      uint8* dst_argb,
                      int width);
@@ skipping 10 matching lines @@
                          uint8* dst_argb,
                          int width);
 void ARGBAddRow_NEON(const uint8* src_argb,
                      const uint8* src_argb1,
                      uint8* dst_argb,
                      int width);
 void ARGBAddRow_Any_NEON(const uint8* src_argb,
                          const uint8* src_argb1,
                          uint8* dst_argb,
                          int width);
+void ARGBAddRow_MSA(const uint8* src_argb,
+                    const uint8* src_argb1,
+                    uint8* dst_argb,
+                    int width);
+void ARGBAddRow_Any_MSA(const uint8* src_argb,
+                        const uint8* src_argb1,
+                        uint8* dst_argb,
+                        int width);
 
 // ARGB subtract images. Same API as Blend, but these require
 // pointer and width alignment for SSE2.
 void ARGBSubtractRow_C(const uint8* src_argb,
                        const uint8* src_argb1,
                        uint8* dst_argb,
                        int width);
 void ARGBSubtractRow_SSE2(const uint8* src_argb,
                           const uint8* src_argb1,
                           uint8* dst_argb,
@@ skipping 11 matching lines @@
                               uint8* dst_argb,
                               int width);
 void ARGBSubtractRow_NEON(const uint8* src_argb,
                           const uint8* src_argb1,
                           uint8* dst_argb,
                           int width);
 void ARGBSubtractRow_Any_NEON(const uint8* src_argb,
                               const uint8* src_argb1,
                               uint8* dst_argb,
                               int width);
+void ARGBSubtractRow_MSA(const uint8* src_argb,
+                         const uint8* src_argb1,
+                         uint8* dst_argb,
+                         int width);
+void ARGBSubtractRow_Any_MSA(const uint8* src_argb,
+                             const uint8* src_argb1,
+                             uint8* dst_argb,
+                             int width);
 
 void ARGBToRGB24Row_Any_SSSE3(const uint8* src_argb, uint8* dst_rgb, int width);
 void ARGBToRAWRow_Any_SSSE3(const uint8* src_argb, uint8* dst_rgb, int width);
 void ARGBToRGB565Row_Any_SSE2(const uint8* src_argb, uint8* dst_rgb, int width);
 void ARGBToARGB1555Row_Any_SSE2(const uint8* src_argb,
                                 uint8* dst_rgb,
                                 int width);
 void ARGBToARGB4444Row_Any_SSE2(const uint8* src_argb,
                                 uint8* dst_rgb,
                                 int width);
@@ skipping 749 matching lines @@
                                  int width,
                                  const uint8* luma,
                                  uint32 lumacoeff);
 
 #ifdef __cplusplus
 }  // extern "C"
 }  // namespace libyuv
 #endif
 
 #endif  // INCLUDE_LIBYUV_ROW_H_