Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(18)

Issues Search
    My Issues | Starred     Open | Closed | All

Unified Diff: source/row_win.cc

Issue 1355393002: yuy2 to rgb gcc versions (Closed) Base URL: https://chromium.googlesource.com/libyuv/libyuv@master
Patch Set: Created 5 years, 3 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View side-by-side diff with in-line comments
Download patch
« no previous file with comments | « source/row_gcc.cc ('k') | no next file » | no next file with comments »
Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
Index: source/row_win.cc
diff --git a/source/row_win.cc b/source/row_win.cc
index f09d2a75be433942f5d816f572837cab344eaa75..3bc9f9bf9061f4d7dafb42b1d684b54f9c8592f1 100644
--- a/source/row_win.cc
+++ b/source/row_win.cc
@@ -34,7 +34,10 @@ extern "C" {
xmm1 = _mm_cvtsi32_si128(*(uint32*)(u_buf + offset)); \
xmm0 = _mm_unpacklo_epi8(xmm0, xmm1); \
xmm0 = _mm_unpacklo_epi16(xmm0, xmm0); \
- u_buf += 4;
+ u_buf += 4; \
+ xmm4 = _mm_loadl_epi64((__m128i*)y_buf); \
+ xmm4 = _mm_unpacklo_epi8(xmm4, xmm4); \
+ y_buf += 8; \
// Convert 8 pixels: 8 UV and 8 Y.
#define YUVTORGB(YuvConstants) \
@@ -46,13 +49,10 @@ extern "C" {
xmm0 = _mm_sub_epi16(*(__m128i*)YuvConstants->kUVBiasB, xmm0); \
xmm1 = _mm_sub_epi16(*(__m128i*)YuvConstants->kUVBiasG, xmm1); \
xmm2 = _mm_sub_epi16(*(__m128i*)YuvConstants->kUVBiasR, xmm2); \
- xmm3 = _mm_loadl_epi64((__m128i*)y_buf); \
- y_buf += 8; \
- xmm3 = _mm_unpacklo_epi8(xmm3, xmm3); \
- xmm3 = _mm_mulhi_epu16(xmm3, *(__m128i*)YuvConstants->kYToRgb); \
- xmm0 = _mm_adds_epi16(xmm0, xmm3); \
- xmm1 = _mm_adds_epi16(xmm1, xmm3); \
- xmm2 = _mm_adds_epi16(xmm2, xmm3); \
+ xmm4 = _mm_mulhi_epu16(xmm4, *(__m128i*)YuvConstants->kYToRgb); \
+ xmm0 = _mm_adds_epi16(xmm0, xmm4); \
+ xmm1 = _mm_adds_epi16(xmm1, xmm4); \
+ xmm2 = _mm_adds_epi16(xmm2, xmm4); \
xmm0 = _mm_srai_epi16(xmm0, 6); \
xmm1 = _mm_srai_epi16(xmm1, 6); \
xmm2 = _mm_srai_epi16(xmm2, 6); \
@@ -90,12 +90,12 @@ void I422ToARGBRow_SSSE3(const uint8* y_buf,
uint8* dst_argb,
struct YuvConstants* yuvconstants,
int width) {
- __m128i xmm0, xmm1, xmm2, xmm3;
+ __m128i xmm0, xmm1, xmm2, xmm4;
const __m128i xmm5 = _mm_set1_epi8(-1);
const ptrdiff_t offset = (uint8*)v_buf - (uint8*)u_buf;
while (width > 0) {
READYUV422
- YUVTORGB(YuvConstants)
+ YUVTORGB(yuvconstants)
STOREARGB
width -= 8;
}
@@ -109,12 +109,12 @@ void I422ToABGRRow_SSSE3(const uint8* y_buf,
uint8* dst_argb,
struct YuvConstants* yuvconstants,
int width) {
- __m128i xmm0, xmm1, xmm2, xmm3;
+ __m128i xmm0, xmm1, xmm2, xmm4;
const __m128i xmm5 = _mm_set1_epi8(-1);
const ptrdiff_t offset = (uint8*)v_buf - (uint8*)u_buf;
while (width > 0) {
READYUV422
- YUVTORGB(YuvConstants)
+ YUVTORGB(yuvconstants)
STOREABGR
width -= 8;
}
@@ -1852,6 +1852,10 @@ void RGBAToUVRow_SSSE3(const uint8* src_argb0, int src_stride_argb,
__asm vpermq ymm0, ymm0, 0xd8 \
__asm vpermq ymm1, ymm1, 0xd8 \
__asm vpunpcklbw ymm0, ymm0, ymm1 /* UV */ \
+ __asm vmovdqu xmm4, [eax] /* Y */ \
+ __asm vpermq ymm4, ymm4, 0xd8 \
+ __asm vpunpcklbw ymm4, ymm4, ymm4 \
+ __asm lea eax, [eax + 16] \
}
// Read 8 UV from 422, upsample to 16 UV.
@@ -1862,6 +1866,10 @@ void RGBAToUVRow_SSSE3(const uint8* src_argb0, int src_stride_argb,
__asm vpunpcklbw ymm0, ymm0, ymm1 /* UV */ \
__asm vpermq ymm0, ymm0, 0xd8 \
__asm vpunpcklwd ymm0, ymm0, ymm0 /* UVUV (upsample) */ \
+ __asm vmovdqu xmm4, [eax] /* Y */ \
+ __asm vpermq ymm4, ymm4, 0xd8 \
+ __asm vpunpcklbw ymm4, ymm4, ymm4 \
+ __asm lea eax, [eax + 16] \
}
// Read 4 UV from 411, upsample to 16 UV.
@@ -1873,6 +1881,10 @@ void RGBAToUVRow_SSSE3(const uint8* src_argb0, int src_stride_argb,
__asm vpunpcklwd ymm0, ymm0, ymm0 /* UVUV (upsample) */ \
__asm vpermq ymm0, ymm0, 0xd8 \
__asm vpunpckldq ymm0, ymm0, ymm0 /* UVUVUVUV (upsample) */ \
+ __asm vmovdqu xmm4, [eax] /* Y */ \
+ __asm vpermq ymm4, ymm4, 0xd8 \
+ __asm vpunpcklbw ymm4, ymm4, ymm4 \
+ __asm lea eax, [eax + 16] \
}
// Read 8 UV from NV12, upsample to 16 UV.
@@ -1881,6 +1893,10 @@ void RGBAToUVRow_SSSE3(const uint8* src_argb0, int src_stride_argb,
__asm lea esi, [esi + 16] \
__asm vpermq ymm0, ymm0, 0xd8 \
__asm vpunpcklwd ymm0, ymm0, ymm0 /* UVUV (upsample) */ \
+ __asm vmovdqu xmm4, [eax] /* Y */ \
+ __asm vpermq ymm4, ymm4, 0xd8 \
+ __asm vpunpcklbw ymm4, ymm4, ymm4 \
+ __asm lea eax, [eax + 16] \
}
// Convert 16 pixels: 16 UV and 16 Y.
@@ -1895,14 +1911,10 @@ void RGBAToUVRow_SSSE3(const uint8* src_argb0, int src_stride_argb,
__asm vmovdqu ymm3, ymmword ptr [YuvConstants + KUVBIASB] \
__asm vpsubw ymm0, ymm3, ymm0 \
/* Step 2: Find Y contribution to 16 R,G,B values */ \
- __asm vmovdqu xmm3, [eax] \
- __asm lea eax, [eax + 16] \
- __asm vpermq ymm3, ymm3, 0xd8 \
- __asm vpunpcklbw ymm3, ymm3, ymm3 \
- __asm vpmulhuw ymm3, ymm3, ymmword ptr [YuvConstants + KYTORGB] \
- __asm vpaddsw ymm0, ymm0, ymm3 /* B += Y */ \
- __asm vpaddsw ymm1, ymm1, ymm3 /* G += Y */ \
- __asm vpaddsw ymm2, ymm2, ymm3 /* R += Y */ \
+ __asm vpmulhuw ymm4, ymm4, ymmword ptr [YuvConstants + KYTORGB] \
+ __asm vpaddsw ymm0, ymm0, ymm4 /* B += Y */ \
+ __asm vpaddsw ymm1, ymm1, ymm4 /* G += Y */ \
+ __asm vpaddsw ymm2, ymm2, ymm4 /* R += Y */ \
__asm vpsraw ymm0, ymm0, 6 \
__asm vpsraw ymm1, ymm1, 6 \
__asm vpsraw ymm2, ymm2, 6 \
@@ -1981,7 +1993,7 @@ void I422ToARGBRow_AVX2(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // argb
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
vpcmpeqb ymm5, ymm5, ymm5 // generate 0xffffffffffffffff for alpha
@@ -2021,7 +2033,7 @@ void I444ToARGBRow_AVX2(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // argb
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
vpcmpeqb ymm5, ymm5, ymm5 // generate 0xffffffffffffffff for alpha
@@ -2060,7 +2072,7 @@ void I444ToABGRRow_AVX2(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // abgr
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
vpcmpeqb ymm5, ymm5, ymm5 // generate 0xffffffffffffffff for alpha
@@ -2099,7 +2111,7 @@ void I411ToARGBRow_AVX2(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // abgr
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
vpcmpeqb ymm5, ymm5, ymm5 // generate 0xffffffffffffffff for alpha
@@ -2136,7 +2148,7 @@ void NV12ToARGBRow_AVX2(const uint8* y_buf,
mov eax, [esp + 8 + 4] // Y
mov esi, [esp + 8 + 8] // UV
mov edx, [esp + 8 + 12] // argb
- mov ebp, [esp + 8 + 16] // YuvConstants
+ mov ebp, [esp + 8 + 16] // yuvconstants
mov ecx, [esp + 8 + 20] // width
vpcmpeqb ymm5, ymm5, ymm5 // generate 0xffffffffffffffff for alpha
@@ -2175,7 +2187,7 @@ void I422ToBGRARow_AVX2(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // abgr
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
vpcmpeqb ymm5, ymm5, ymm5 // generate 0xffffffffffffffff for alpha
@@ -2215,7 +2227,7 @@ void I422ToRGBARow_AVX2(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // abgr
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
vpcmpeqb ymm5, ymm5, ymm5 // generate 0xffffffffffffffff for alpha
@@ -2255,7 +2267,7 @@ void I422ToABGRRow_AVX2(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // argb
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
vpcmpeqb ymm5, ymm5, ymm5 // generate 0xffffffffffffffff for alpha
@@ -2286,6 +2298,9 @@ void I422ToABGRRow_AVX2(const uint8* y_buf,
__asm movq xmm1, qword ptr [esi + edi] /* V */ \
__asm lea esi, [esi + 8] \
__asm punpcklbw xmm0, xmm1 /* UV */ \
+ __asm movq xmm4, qword ptr [eax] \
+ __asm punpcklbw xmm4, xmm4 \
+ __asm lea eax, [eax + 8] \
}
// Read 4 UV from 422, upsample to 8 UV.
@@ -2295,6 +2310,9 @@ void I422ToABGRRow_AVX2(const uint8* y_buf,
__asm lea esi, [esi + 4] \
__asm punpcklbw xmm0, xmm1 /* UV */ \
__asm punpcklwd xmm0, xmm0 /* UVUV (upsample) */ \
+ __asm movq xmm4, qword ptr [eax] \
+ __asm punpcklbw xmm4, xmm4 \
+ __asm lea eax, [eax + 8] \
}
// Read 2 UV from 411, upsample to 8 UV.
@@ -2305,6 +2323,9 @@ void I422ToABGRRow_AVX2(const uint8* y_buf,
__asm punpcklbw xmm0, xmm1 /* UV */ \
__asm punpcklwd xmm0, xmm0 /* UVUV (upsample) */ \
__asm punpckldq xmm0, xmm0 /* UVUVUVUV (upsample) */ \
+ __asm movq xmm4, qword ptr [eax] \
+ __asm punpcklbw xmm4, xmm4 \
+ __asm lea eax, [eax + 8] \
}
// Read 4 UV from NV12, upsample to 8 UV.
@@ -2312,6 +2333,47 @@ void I422ToABGRRow_AVX2(const uint8* y_buf,
__asm movq xmm0, qword ptr [esi] /* UV */ \
__asm lea esi, [esi + 8] \
__asm punpcklwd xmm0, xmm0 /* UVUV (upsample) */ \
+ __asm movq xmm4, qword ptr [eax] \
+ __asm punpcklbw xmm4, xmm4 \
+ __asm lea eax, [eax + 8] \
+ }
+
+// YUY2 shuf 8 Y to 16 Y.
+static const vec8 kShuffleYUY2Y = {
+ 0, 0, 2, 2, 4, 4, 6, 6, 8, 8, 10, 10, 12, 12, 14, 14
+};
+
+// YUY2 shuf 4 UV to 8 UV.
+static const vec8 kShuffleYUY2UV = {
+ 1, 3, 1, 3, 5, 7, 5, 7, 9, 11, 9, 11, 13, 15, 13, 15
+};
+
+// Read 4 YUY2 with 8 Y and update 4 UV to 8 UV.
+#define READYUY2 __asm { \
+ __asm movdqu xmm4, [eax] /* YUY2 */ \
+ __asm pshufb xmm4, xmmword ptr kShuffleYUY2Y \
+ __asm movdqu xmm0, [eax] /* UV */ \
+ __asm pshufb xmm0, xmmword ptr kShuffleYUY2UV \
+ __asm lea eax, [eax + 16] \
+ }
+
+// UYVY shuf 8 Y to 16 Y.
+static const vec8 kShuffleUYVYY = {
+ 1, 1, 3, 3, 5, 5, 7, 7, 9, 9, 11, 11, 13, 13, 15, 15
+};
+
+// UYVY shuf 4 UV to 8 UV.
+static const vec8 kShuffleUYVYUV = {
+ 0, 2, 0, 2, 4, 6, 4, 6, 8, 10, 8, 10, 12, 14, 12, 14
+};
+
+// Read 4 UYVY with 8 Y and update 4 UV to 8 UV.
+#define READUYVY __asm { \
+ __asm movdqu xmm4, [eax] /* UYVY */ \
+ __asm pshufb xmm4, xmmword ptr kShuffleUYVYY \
+ __asm movdqu xmm0, [eax] /* UV */ \
+ __asm pshufb xmm0, xmmword ptr kShuffleUYVYUV \
+ __asm lea eax, [eax + 16] \
}
// Convert 8 pixels: 8 UV and 8 Y.
@@ -2328,13 +2390,10 @@ void I422ToABGRRow_AVX2(const uint8* y_buf,
__asm movdqa xmm2, xmmword ptr [YuvConstants + KUVBIASR] \
__asm pmaddubsw xmm3, xmmword ptr [YuvConstants + KUVTOR] \
__asm psubw xmm2, xmm3 \
- __asm movq xmm3, qword ptr [eax] \
- __asm lea eax, [eax + 8] \
- __asm punpcklbw xmm3, xmm3 \
- __asm pmulhuw xmm3, xmmword ptr [YuvConstants + KYTORGB] \
- __asm paddsw xmm0, xmm3 /* B += Y */ \
- __asm paddsw xmm1, xmm3 /* G += Y */ \
- __asm paddsw xmm2, xmm3 /* R += Y */ \
+ __asm pmulhuw xmm4, xmmword ptr [YuvConstants + KYTORGB] \
+ __asm paddsw xmm0, xmm4 /* B += Y */ \
+ __asm paddsw xmm1, xmm4 /* G += Y */ \
+ __asm paddsw xmm2, xmm4 /* R += Y */ \
__asm psraw xmm0, 6 \
__asm psraw xmm1, 6 \
__asm psraw xmm2, 6 \
@@ -2480,7 +2539,7 @@ void I444ToARGBRow_SSSE3(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // argb
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
pcmpeqb xmm5, xmm5 // generate 0xffffffff for alpha
@@ -2517,7 +2576,7 @@ void I444ToABGRRow_SSSE3(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // abgr
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
pcmpeqb xmm5, xmm5 // generate 0xffffffff for alpha
@@ -2554,7 +2613,7 @@ void I422ToRGB24Row_SSSE3(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // argb
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
movdqa xmm5, xmmword ptr kShuffleMaskARGBToRGB24_0
@@ -2592,7 +2651,7 @@ void I422ToRAWRow_SSSE3(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // argb
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
movdqa xmm5, xmmword ptr kShuffleMaskARGBToRAW_0
@@ -2630,7 +2689,7 @@ void I422ToRGB565Row_SSSE3(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // argb
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
pcmpeqb xmm5, xmm5 // generate mask 0x0000001f
@@ -2673,7 +2732,7 @@ void I422ToARGBRow_SSSE3(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // argb
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
pcmpeqb xmm5, xmm5 // generate 0xffffffff for alpha
@@ -2711,7 +2770,7 @@ void I411ToARGBRow_SSSE3(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // abgr
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
pcmpeqb xmm5, xmm5 // generate 0xffffffff for alpha
@@ -2745,7 +2804,7 @@ void NV12ToARGBRow_SSSE3(const uint8* y_buf,
mov eax, [esp + 8 + 4] // Y
mov esi, [esp + 8 + 8] // UV
mov edx, [esp + 8 + 12] // argb
- mov ebp, [esp + 8 + 16] // YuvConstants
+ mov ebp, [esp + 8 + 16] // yuvconstants
mov ecx, [esp + 8 + 20] // width
pcmpeqb xmm5, xmm5 // generate 0xffffffff for alpha
@@ -2763,6 +2822,62 @@ void NV12ToARGBRow_SSSE3(const uint8* y_buf,
}
}
+// 8 pixels.
+// 4 YUY2 values with 8 Y and 4 UV producing 8 ARGB (32 bytes).
+__declspec(naked)
+void YUY2ToARGBRow_SSSE3(const uint8* src_yuy2,
+ uint8* dst_argb,
+ struct YuvConstants* yuvconstants,
+ int width) {
+ __asm {
+ push ebp
+ mov eax, [esp + 4 + 4] // yuy2
+ mov edx, [esp + 4 + 8] // argb
+ mov ebp, [esp + 4 + 12] // yuvconstants
+ mov ecx, [esp + 4 + 16] // width
+ pcmpeqb xmm5, xmm5 // generate 0xffffffff for alpha
+
+ convertloop:
+ READYUY2
+ YUVTORGB(ebp)
+ STOREARGB
+
+ sub ecx, 8
+ jg convertloop
+
+ pop ebp
+ ret
+ }
+}
+
+// 8 pixels.
+// 4 UYVY values with 8 Y and 4 UV producing 8 ARGB (32 bytes).
+__declspec(naked)
+void UYVYToARGBRow_SSSE3(const uint8* src_uyvy,
+ uint8* dst_argb,
+ struct YuvConstants* yuvconstants,
+ int width) {
+ __asm {
+ push ebp
+ mov eax, [esp + 4 + 4] // uyvy
+ mov edx, [esp + 4 + 8] // argb
+ mov ebp, [esp + 4 + 12] // yuvconstants
+ mov ecx, [esp + 4 + 16] // width
+ pcmpeqb xmm5, xmm5 // generate 0xffffffff for alpha
+
+ convertloop:
+ READUYVY
+ YUVTORGB(ebp)
+ STOREARGB
+
+ sub ecx, 8
+ jg convertloop
+
+ pop ebp
+ ret
+ }
+}
+
__declspec(naked)
void I422ToBGRARow_SSSE3(const uint8* y_buf,
const uint8* u_buf,
@@ -2778,7 +2893,7 @@ void I422ToBGRARow_SSSE3(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // argb
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
@@ -2812,7 +2927,7 @@ void I422ToABGRRow_SSSE3(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // argb
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
pcmpeqb xmm5, xmm5 // generate 0xffffffff for alpha
@@ -2847,7 +2962,7 @@ void I422ToRGBARow_SSSE3(const uint8* y_buf,
mov esi, [esp + 12 + 8] // U
mov edi, [esp + 12 + 12] // V
mov edx, [esp + 12 + 16] // argb
- mov ebp, [esp + 12 + 20] // YuvConstants
+ mov ebp, [esp + 12 + 20] // yuvconstants
mov ecx, [esp + 12 + 24] // width
sub edi, esi
@@ -3512,8 +3627,7 @@ void ARGBSetRow_X86(uint8* dst_argb, uint32 v32, int count) {
#ifdef HAS_YUY2TOYROW_AVX2
__declspec(naked)
-void YUY2ToYRow_AVX2(const uint8* src_yuy2,
- uint8* dst_y, int pix) {
+void YUY2ToYRow_AVX2(const uint8* src_yuy2, uint8* dst_y, int pix) {
__asm {
mov eax, [esp + 4] // src_yuy2
mov edx, [esp + 8] // dst_y
« no previous file with comments | « source/row_gcc.cc ('k') | no next file » | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698