diff options
| author | Manojkumar Bhosale <manojkumar.bhosale@imgtec.com> | 2017-02-01 10:31:28 +0530 |
|---|---|---|
| committer | Manojkumar Bhosale <manojkumar.bhosale@imgtec.com> | 2017-02-01 10:31:28 +0530 |
| commit | 54ce8f23d6d82d671ec05229812842548faf14ef (patch) | |
| tree | 8f4f7680b8aad4fa443b38ee59ab560c767f2f8f /source/convert_from_argb.cc | |
| parent | 03510421970e0029bf744feac5c128c2ee22c90c (diff) | |
| download | libyuv-54ce8f23d6d82d671ec05229812842548faf14ef.tar.gz | |
Add MSA optimized ARGB/ABGR/BGRA/RGBA To Y/UV row functions
R=fbarchard@google.com
BUG=libyuv:634
Performance Gain (vs C auto-vectorized)
ARGBToYJRow_MSA - ~3.2x
ARGBToYJRow_Any_MSA - ~2.7x
BGRAToYRow_MSA - ~3.2x
BGRAToYRow_Any_MSA - ~2.7x
ABGRToYRow_MSA - ~3.2x
ABGRToYRow_Any_MSA - ~2.6x
RGBAToYRow_MSA - ~3.1x
RGBAToYRow_Any_MSA - ~2.7x
ARGBToUVJRow_MSA - ~5.5x
ARGBToUVJRow_Any_MSA - ~4.5x
BGRAToUVRow_MSA - ~2.1x
BGRAToUVRow_Any_MSA - ~2.0x
ABGRToUVRow_MSA - ~2.1x
ABGRToUVRow_Any_MSA - ~1.9x
RGBAToUVRow_MSA - ~2.2x
RGBAToUVRow_Any_MSA - ~1.9x
Performance Gain (vs C non-vectorized)
ARGBToYJRow_MSA - ~10.9x
ARGBToYJRow_Any_MSA - ~9.2x
BGRAToYRow_MSA - ~10.9x
BGRAToYRow_Any_MSA - ~9.3x
ABGRToYRow_MSA - ~11.0x
ABGRToYRow_Any_MSA - ~9.3x
RGBAToYRow_MSA - ~10.9x
RGBAToYRow_Any_MSA - ~9.1x
ARGBToUVJRow_MSA - ~12.4x
ARGBToUVJRow_Any_MSA - ~10.5x
BGRAToUVRow_MSA - ~4.7x
BGRAToUVRow_Any_MSA - ~4.4x
ABGRToUVRow_MSA - ~4.7x
ABGRToUVRow_Any_MSA - ~4.5x
RGBAToUVRow_MSA - ~4.8x
RGBAToUVRow_Any_MSA - ~4.4x
Review-Url: https://codereview.chromium.org/2641153003 .
Diffstat (limited to 'source/convert_from_argb.cc')
| -rw-r--r-- | source/convert_from_argb.cc | 40 |
1 files changed, 40 insertions, 0 deletions
diff --git a/source/convert_from_argb.cc b/source/convert_from_argb.cc index e45bb230..23ae66b2 100644 --- a/source/convert_from_argb.cc +++ b/source/convert_from_argb.cc @@ -1352,6 +1352,22 @@ int ARGBToJ420(const uint8* src_argb, } } #endif +#if defined(HAS_ARGBTOYJROW_MSA) + if (TestCpuFlag(kCpuHasMSA)) { + ARGBToYJRow = ARGBToYJRow_Any_MSA; + if (IS_ALIGNED(width, 16)) { + ARGBToYJRow = ARGBToYJRow_MSA; + } + } +#endif +#if defined(HAS_ARGBTOUVJROW_MSA) + if (TestCpuFlag(kCpuHasMSA)) { + ARGBToUVJRow = ARGBToUVJRow_Any_MSA; + if (IS_ALIGNED(width, 32)) { + ARGBToUVJRow = ARGBToUVJRow_MSA; + } + } +#endif for (y = 0; y < height - 1; y += 2) { ARGBToUVJRow(src_argb, src_stride_argb, dst_u, dst_v, width); @@ -1436,6 +1452,22 @@ int ARGBToJ422(const uint8* src_argb, } } #endif +#if defined(HAS_ARGBTOYJROW_MSA) + if (TestCpuFlag(kCpuHasMSA)) { + ARGBToYJRow = ARGBToYJRow_Any_MSA; + if (IS_ALIGNED(width, 16)) { + ARGBToYJRow = ARGBToYJRow_MSA; + } + } +#endif +#if defined(HAS_ARGBTOUVJROW_MSA) + if (TestCpuFlag(kCpuHasMSA)) { + ARGBToUVJRow = ARGBToUVJRow_Any_MSA; + if (IS_ALIGNED(width, 32)) { + ARGBToUVJRow = ARGBToUVJRow_MSA; + } + } +#endif for (y = 0; y < height; ++y) { ARGBToUVJRow(src_argb, 0, dst_u, dst_v, width); @@ -1497,6 +1529,14 @@ int ARGBToJ400(const uint8* src_argb, } } #endif +#if defined(HAS_ARGBTOYJROW_MSA) + if (TestCpuFlag(kCpuHasMSA)) { + ARGBToYJRow = ARGBToYJRow_Any_MSA; + if (IS_ALIGNED(width, 16)) { + ARGBToYJRow = ARGBToYJRow_MSA; + } + } +#endif for (y = 0; y < height; ++y) { ARGBToYJRow(src_argb, dst_yj, width); |