Port YUY2ToYRow_AVX2 and UYVYToYRow_AVX2 to gcc/NaCL from Windows AVX code.

BUG=269
TESTED=ncval
R=brucedawson@google.com, harryjin@google.com

Review URL: https://webrtc-codereview.appspot.com/25039004

git-svn-id: http://libyuv.googlecode.com/svn/trunk@1151 16f28f9a-4ce2-e073-06de-1de4eb20be90

README.chromium

 Name: libyuv
 URL: http://code.google.com/p/libyuv/
-Version: 1148
+Version: 1149
 License: BSD
 License File: LICENSE
 

include/libyuv/version.h

@@ -11,6 +11,6 @@
 #ifndef INCLUDE_LIBYUV_VERSION_H_  // NOLINT
 #define INCLUDE_LIBYUV_VERSION_H_
 
-#define LIBYUV_VERSION 1148
+#define LIBYUV_VERSION 1149
 
 #endif  // INCLUDE_LIBYUV_VERSION_H_  NOLINT

source/row_any.cc

@@ -243,9 +243,13 @@ BAYERANY(ARGBToBayerGGRow_Any_NEON, ARGBToBayerGGRow_NEON, ARGBToBayerGGRow_C,
 #ifdef HAS_ARGBTOYROW_AVX2
 YANY(ARGBToYRow_Any_AVX2, ARGBToYRow_AVX2, ARGBToYRow_C, 4, 1, 31)
 YANY(ARGBToYJRow_Any_AVX2, ARGBToYJRow_AVX2, ARGBToYJRow_C, 4, 1, 31)
-YANY(YUY2ToYRow_Any_AVX2, YUY2ToYRow_AVX2, YUY2ToYRow_C, 2, 1, 31)
+#endif
+#ifdef HAS_UYVYTOYROW_AVX2
 YANY(UYVYToYRow_Any_AVX2, UYVYToYRow_AVX2, UYVYToYRow_C, 2, 1, 31)
 #endif
+#ifdef HAS_YUY2TOYROW_AVX2
+YANY(YUY2ToYRow_Any_AVX2, YUY2ToYRow_AVX2, YUY2ToYRow_C, 2, 1, 31)
+#endif
 #ifdef HAS_ARGBTOYROW_SSSE3
 YANY(ARGBToYRow_Any_SSSE3, ARGBToYRow_SSSE3, ARGBToYRow_C, 4, 1, 15)
 #endif

source/row_posix.cc

@@ -2853,6 +2853,64 @@ void UYVYToUV422Row_SSE2(const uint8* src_uyvy,
 }
 #endif  // HAS_YUY2TOYROW_SSE2
 
+#ifdef HAS_YUY2TOYROW_AVX2
+void YUY2ToYRow_AVX2(const uint8* src_yuy2, uint8* dst_y, int pix) {
+  asm volatile (
+    "vpcmpeqb  %%ymm5,%%ymm5,%%ymm5            \n"
+    "vpsrlw    $0x8,%%ymm5,%%ymm5              \n"
+    LABELALIGN
+  "1:                                          \n"
+    "vmovdqu   " MEMACCESS(0) ",%%ymm0         \n"
+    "vmovdqu   " MEMACCESS2(0x20,0) ",%%ymm1   \n"
+    "lea       " MEMLEA(0x40,0) ",%0           \n"
+    "vpand     %%ymm5,%%ymm0,%%ymm0            \n"
+    "vpand     %%ymm5,%%ymm1,%%ymm1            \n"
+    "vpackuswb %%ymm1,%%ymm0,%%ymm0            \n"
+    "vpermq    $0xd8,%%ymm0,%%ymm0             \n"
+    "sub       $0x20,%2                        \n"
+    "vmovdqu   %%ymm0," MEMACCESS(1) "         \n"
+    "lea      " MEMLEA(0x20,1) ",%1            \n"
+    "jg        1b                              \n"
+  : "+r"(src_yuy2),  // %0
+    "+r"(dst_y),     // %1
+    "+r"(pix)        // %2
+  :
+  : "memory", "cc"
+#if defined(__SSE2__)
+    , "xmm0", "xmm1", "xmm5"
+#endif
+  );
+}
+
+void UYVYToYRow_AVX2(const uint8* src_uyvy, uint8* dst_y, int pix) {
+  asm volatile (
+    LABELALIGN
+  "1:                                          \n"
+    "vmovdqu   " MEMACCESS(0) ",%%ymm0         \n"
+    "vmovdqu   " MEMACCESS2(0x20,0) ",%%ymm1   \n"
+    "lea       " MEMLEA(0x40,0) ",%0           \n"
+    "vpsrlw    $0x8,%%ymm0,%%ymm0              \n"
+    "vpsrlw    $0x8,%%ymm1,%%ymm1              \n"
+    "vpackuswb %%ymm1,%%ymm0,%%ymm0            \n"
+    "vpermq    $0xd8,%%ymm0,%%ymm0             \n"
+    "sub       $0x20,%2                     \n"
+    "vmovdqu   %%ymm0," MEMACCESS(1) "         \n"
+    "lea      " MEMLEA(0x20,1) ",%1            \n"
+    "jg        1b                              \n"
+    "vzeroupper                                \n"
+    "ret                                       \n"
+  : "+r"(src_uyvy),  // %0
+    "+r"(dst_y),     // %1
+    "+r"(pix)        // %2
+  :
+  : "memory", "cc"
+#if defined(__SSE2__)
+    , "xmm0", "xmm1", "xmm5"
+#endif
+  );
+}
+#endif  // HAS_YUY2TOYROW_AVX2
+
 #ifdef HAS_ARGBBLENDROW_SSE2
 // Blend 8 pixels at a time.
 void ARGBBlendRow_SSE2(const uint8* src_argb0, const uint8* src_argb1,