neon yuv matrix function

source/row_neon.cc

 /*
  *  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 132 matching lines @@
 #define UB -128 /* -min(128, round(2.018 * 64)) */
 #define UG 25 /* -round(-0.391 * 64) */
 #define VG 52 /* -round(-0.813 * 64) */
 #define VR -102 /* -round(1.596 * 64) */
 
 // Bias values to subtract 16 from Y and 128 from U and V.
 #define BB (UB * 128            - YGB)
 #define BG (UG * 128 + VG * 128 - YGB)
 #define BR            (VR * 128 - YGB)
 
+YuvConstantsNEON SIMD_ALIGNED(kYuvConstantsNEON) = {
+  { 128, 128, 128, 128, 102, 102, 102, 102, 0, 0, 0, 0, 0, 0, 0, 0 },
+  { 25, 25, 25, 25, 52, 52, 52, 52, 0, 0, 0, 0, 0, 0, 0, 0 },
+  { BB, BG, BR, 0, 0, 0, 0, 0 },
+  { 0x0101 * YG, 0, 0, 0 }
+};
+
 static uvec8 kUVToRB  = { 128, 128, 128, 128, 102, 102, 102, 102,
                           0, 0, 0, 0, 0, 0, 0, 0 };
 static uvec8 kUVToG = { 25, 25, 25, 25, 52, 52, 52, 52,
                         0, 0, 0, 0, 0, 0, 0, 0 };
 static vec16 kUVBiasBGR = { BB, BG, BR, 0, 0, 0, 0, 0 };
 static vec32 kYToRgb = { 0x0101 * YG, 0, 0, 0 };
 
 #undef YG
 #undef YGB
 #undef UB
@@ skipping 26 matching lines @@
       "+r"(width)      // %4
     : [kUVToRB]"r"(&kUVToRB),   // %5
       [kUVToG]"r"(&kUVToG),     // %6
       [kUVBiasBGR]"r"(&kUVBiasBGR),
       [kYToRgb]"r"(&kYToRgb)
     : "cc", "memory", "q0", "q1", "q2", "q3", "q4",
       "q8", "q9", "q10", "q11", "q12", "q13", "q14", "q15"
   );
 }
 
-void I422ToARGBRow_NEON(const uint8* src_y,
-                        const uint8* src_u,
-                        const uint8* src_v,
-                        uint8* dst_argb,
-                        int width) {
+void I422ToARGBMatrixRow_NEON(const uint8* src_y,
+                              const uint8* src_u,
+                              const uint8* src_v,
+                              uint8* dst_argb,
+                              struct YuvConstantsNEON* YuvConstants,
+                              int width) {
   asm volatile (
     YUV422TORGB_SETUP_REG
   "1:                                          \n"
     READYUV422
     YUV422TORGB
     "subs       %4, %4, #8                     \n"
     "vmov.u8    d23, #255                      \n"
     MEMACCESS(3)
     "vst4.8     {d20, d21, d22, d23}, [%3]!    \n"
     "bgt        1b                             \n"
     : "+r"(src_y),     // %0
       "+r"(src_u),     // %1
       "+r"(src_v),     // %2
       "+r"(dst_argb),  // %3
       "+r"(width)      // %4
-    : [kUVToRB]"r"(&kUVToRB),   // %5
-      [kUVToG]"r"(&kUVToG),     // %6
-      [kUVBiasBGR]"r"(&kUVBiasBGR),
-      [kYToRgb]"r"(&kYToRgb)
+    : [kUVToRB]"r"(&YuvConstants->kUVToRB),   // %5
+      [kUVToG]"r"(&YuvConstants->kUVToG),     // %6
+      [kUVBiasBGR]"r"(&YuvConstants->kUVBiasBGR),
+      [kYToRgb]"r"(&YuvConstants->kYToRgb)
     : "cc", "memory", "q0", "q1", "q2", "q3", "q4",
       "q8", "q9", "q10", "q11", "q12", "q13", "q14", "q15"
   );
 }
 
 void I411ToARGBRow_NEON(const uint8* src_y,
                         const uint8* src_u,
                         const uint8* src_v,
                         uint8* dst_argb,
                         int width) {
@@ skipping 2762 matching lines @@
     "r"(6)             // %5
   : "cc", "memory", "q0", "q1"  // Clobber List
   );
 }
 #endif  // defined(__ARM_NEON__) && !defined(__aarch64__)
 
 #ifdef __cplusplus
 }  // extern "C"
 }  // namespace libyuv
 #endif