diff --git a/README.chromium b/README.chromium
index f8834f228..8ca902f4b 100644
--- a/README.chromium
+++ b/README.chromium
@@ -1,6 +1,6 @@
 Name: libyuv
 URL: http://code.google.com/p/libyuv/
-Version: 1198
+Version: 1200
 License: BSD
 License File: LICENSE
 
diff --git a/include/libyuv/row.h b/include/libyuv/row.h
index 02958a7b6..7dbb5e569 100644
--- a/include/libyuv/row.h
+++ b/include/libyuv/row.h
@@ -189,7 +189,7 @@ extern "C" {
 #define HAS_ARGBMIRRORROW_AVX2
 #define HAS_ARGBPOLYNOMIALROW_AVX2
 #define HAS_ARGBSHUFFLEROW_AVX2
-//#define HAS_ARGBTOUVROW_AVX2
+#define HAS_ARGBTOUVROW_AVX2
 #define HAS_ARGBTOYJROW_AVX2
 #define HAS_ARGBTOYROW_AVX2
 #define HAS_COPYROW_AVX
diff --git a/include/libyuv/version.h b/include/libyuv/version.h
index c975cc9d8..783e628f0 100644
--- a/include/libyuv/version.h
+++ b/include/libyuv/version.h
@@ -11,6 +11,6 @@
 #ifndef INCLUDE_LIBYUV_VERSION_H_  // NOLINT
 #define INCLUDE_LIBYUV_VERSION_H_
 
-#define LIBYUV_VERSION 1198
+#define LIBYUV_VERSION 1200
 
 #endif  // INCLUDE_LIBYUV_VERSION_H_  NOLINT
diff --git a/source/row_posix.cc b/source/row_posix.cc
index 5a5d14835..78997edc2 100644
--- a/source/row_posix.cc
+++ b/source/row_posix.cc
@@ -1006,9 +1006,9 @@ void ARGBToUVRow_AVX2(const uint8* src_argb0, int src_stride_argb,
     "+r"(dst_v),           // %2
     "+rm"(width)           // %3
   : "r"((intptr_t)(src_stride_argb)), // %4
-    "m"(kARGBToV),   // %5
-    "m"(kARGBToU),   // %6
-    "m"(kAddUV128),  // %7
+    "m"(kAddUV128),  // %5
+    "m"(kARGBToV),   // %6
+    "m"(kARGBToU),   // %7
     "m"(kShufARGBToUV_AVX)  // %8
   : "memory", "cc"
 #if defined(__native_client__) && defined(__x86_64__)
diff --git a/source/row_win.cc b/source/row_win.cc
index fae111cbf..6b6776659 100644
--- a/source/row_win.cc
+++ b/source/row_win.cc
@@ -847,8 +847,8 @@ void BGRAToYRow_SSSE3(const uint8* src_argb, uint8* dst_y, int pix) {
     mov        eax, [esp + 4]   /* src_argb */
     mov        edx, [esp + 8]   /* dst_y */
     mov        ecx, [esp + 12]  /* pix */
-    movdqa     xmm5, kAddY16
     movdqa     xmm4, kBGRAToY
+    movdqa     xmm5, kAddY16
 
  convertloop:
     movdqu     xmm0, [eax]
@@ -880,8 +880,8 @@ void ABGRToYRow_SSSE3(const uint8* src_argb, uint8* dst_y, int pix) {
     mov        eax, [esp + 4]   /* src_argb */
     mov        edx, [esp + 8]   /* dst_y */
     mov        ecx, [esp + 12]  /* pix */
-    movdqa     xmm5, kAddY16
     movdqa     xmm4, kABGRToY
+    movdqa     xmm5, kAddY16
 
  convertloop:
     movdqu     xmm0, [eax]
@@ -913,8 +913,8 @@ void RGBAToYRow_SSSE3(const uint8* src_argb, uint8* dst_y, int pix) {
     mov        eax, [esp + 4]   /* src_argb */
     mov        edx, [esp + 8]   /* dst_y */
     mov        ecx, [esp + 12]  /* pix */
-    movdqa     xmm5, kAddY16
     movdqa     xmm4, kRGBAToY
+    movdqa     xmm5, kAddY16
 
  convertloop:
     movdqu     xmm0, [eax]
@@ -951,9 +951,9 @@ void ARGBToUVRow_SSSE3(const uint8* src_argb0, int src_stride_argb,
     mov        edx, [esp + 8 + 12]  // dst_u
     mov        edi, [esp + 8 + 16]  // dst_v
     mov        ecx, [esp + 8 + 20]  // pix
-    movdqa     xmm7, kARGBToU
-    movdqa     xmm6, kARGBToV
     movdqa     xmm5, kAddUV128
+    movdqa     xmm6, kARGBToV
+    movdqa     xmm7, kARGBToU
     sub        edi, edx             // stride from u to v
 
  convertloop:
@@ -1021,9 +1021,9 @@ void ARGBToUVJRow_SSSE3(const uint8* src_argb0, int src_stride_argb,
     mov        edx, [esp + 8 + 12]  // dst_u
     mov        edi, [esp + 8 + 16]  // dst_v
     mov        ecx, [esp + 8 + 20]  // pix
-    movdqa     xmm7, kARGBToUJ
-    movdqa     xmm6, kARGBToVJ
     movdqa     xmm5, kAddUVJ128
+    movdqa     xmm6, kARGBToVJ
+    movdqa     xmm7, kARGBToUJ
     sub        edi, edx             // stride from u to v
 
  convertloop:
@@ -1156,9 +1156,9 @@ void ARGBToUV444Row_SSSE3(const uint8* src_argb0,
     mov        edx, [esp + 4 + 8]   // dst_u
     mov        edi, [esp + 4 + 12]  // dst_v
     mov        ecx, [esp + 4 + 16]  // pix
-    movdqa     xmm7, kARGBToU
-    movdqa     xmm6, kARGBToV
     movdqa     xmm5, kAddUV128
+    movdqa     xmm6, kARGBToV
+    movdqa     xmm7, kARGBToU
     sub        edi, edx             // stride from u to v
 
  convertloop:
@@ -1213,9 +1213,9 @@ void ARGBToUV422Row_SSSE3(const uint8* src_argb0,
     mov        edx, [esp + 4 + 8]   // dst_u
     mov        edi, [esp + 4 + 12]  // dst_v
     mov        ecx, [esp + 4 + 16]  // pix
-    movdqa     xmm7, kARGBToU
-    movdqa     xmm6, kARGBToV
     movdqa     xmm5, kAddUV128
+    movdqa     xmm6, kARGBToV
+    movdqa     xmm7, kARGBToU
     sub        edi, edx             // stride from u to v
 
  convertloop:
@@ -1273,9 +1273,9 @@ void BGRAToUVRow_SSSE3(const uint8* src_argb0, int src_stride_argb,
     mov        edx, [esp + 8 + 12]  // dst_u
     mov        edi, [esp + 8 + 16]  // dst_v
     mov        ecx, [esp + 8 + 20]  // pix
-    movdqa     xmm7, kBGRAToU
-    movdqa     xmm6, kBGRAToV
     movdqa     xmm5, kAddUV128
+    movdqa     xmm6, kBGRAToV
+    movdqa     xmm7, kBGRAToU
     sub        edi, edx             // stride from u to v
 
  convertloop:
@@ -1343,9 +1343,9 @@ void ABGRToUVRow_SSSE3(const uint8* src_argb0, int src_stride_argb,
     mov        edx, [esp + 8 + 12]  // dst_u
     mov        edi, [esp + 8 + 16]  // dst_v
     mov        ecx, [esp + 8 + 20]  // pix
-    movdqa     xmm7, kABGRToU
-    movdqa     xmm6, kABGRToV
     movdqa     xmm5, kAddUV128
+    movdqa     xmm6, kABGRToV
+    movdqa     xmm7, kABGRToU
     sub        edi, edx             // stride from u to v
 
  convertloop:
@@ -1413,9 +1413,9 @@ void RGBAToUVRow_SSSE3(const uint8* src_argb0, int src_stride_argb,
     mov        edx, [esp + 8 + 12]  // dst_u
     mov        edi, [esp + 8 + 16]  // dst_v
     mov        ecx, [esp + 8 + 20]  // pix
-    movdqa     xmm7, kRGBAToU
-    movdqa     xmm6, kRGBAToV
     movdqa     xmm5, kAddUV128
+    movdqa     xmm6, kRGBAToV
+    movdqa     xmm7, kRGBAToU
     sub        edi, edx             // stride from u to v
 
  convertloop: