Keep track of relative pixel offsets and utilize pshufb to efficiently extract relevant pixels for horizontal scaling ratios <= 8. Because pshufb does not cross 128-bit lanes, the overhead of address calculations and loads is relatively greater as compared with an SSSE3 implementation. Fall back to a generic approach for ratios > 8. The implementation assumes that data beyond the end of each line, before the next line begins, can be dirtied; which AFAICT is safe with the current usage of these routines. Speedup is ~10.42x/~5.23x (32-bit/64-bit) for horizontal ratios <= 2, ~9.49x/~4.64x for ratios within (2, 4], ~6.43x/~3.18x for ratios within (4, 8], and ~5.42x/~2.50x for ratios > 8 when not memory-bound on Haswell as compared with the current SSE2 implementation.
Showing the changes to codec/processing/src/downsample (3 of 5 files). Show all
Unified · Split
| @@ -109,6 +109,9 @@ void CDownsampling::InitDownsampleFuncs (SDownsampleFuncs& sDownsampleFunc, int · 108 unmodified lines | |||
| 109 | sDownsampleFunc.pfQuarterDownsampler = DyadicBilinearQuarterDownsampler_sse4; | 109 | sDownsampleFunc.pfQuarterDownsampler = DyadicBilinearQuarterDownsampler_sse4; |
| 110 | sDownsampleFunc.pfGeneralRatioChroma = GeneralBilinearAccurateDownsamplerWrap_sse41; | 110 | sDownsampleFunc.pfGeneralRatioChroma = GeneralBilinearAccurateDownsamplerWrap_sse41; |
| 111 | } | 111 | } |
| 112 | if (iCpuFlag & WELS_CPU_AVX2) { | ||
| 113 | sDownsampleFunc.pfGeneralRatioLuma = GeneralBilinearFastDownsamplerWrap_avx2; | ||
| 114 | } | ||
| 112 | #endif//X86_ASM | 115 | #endif//X86_ASM |
| 113 | 116 | ||
| 114 | #if defined(HAVE_NEON) | 117 | #if defined(HAVE_NEON) |
| @@ -103,6 +103,7 @@ GeneralDownsampleFunc GeneralBilinearFastDownsamplerWrap_sse2; · 102 unmodified lines | |||
| 103 | GeneralDownsampleFunc GeneralBilinearAccurateDownsamplerWrap_sse2; | 103 | GeneralDownsampleFunc GeneralBilinearAccurateDownsamplerWrap_sse2; |
| 104 | GeneralDownsampleFunc GeneralBilinearFastDownsamplerWrap_ssse3; | 104 | GeneralDownsampleFunc GeneralBilinearFastDownsamplerWrap_ssse3; |
| 105 | GeneralDownsampleFunc GeneralBilinearAccurateDownsamplerWrap_sse41; | 105 | GeneralDownsampleFunc GeneralBilinearAccurateDownsamplerWrap_sse41; |
| 106 | GeneralDownsampleFunc GeneralBilinearFastDownsamplerWrap_avx2; | ||
| 106 | 107 | ||
| 107 | SpecificDownsampleFunc DyadicBilinearOneThirdDownsampler_ssse3; | 108 | SpecificDownsampleFunc DyadicBilinearOneThirdDownsampler_ssse3; |
| 108 | SpecificDownsampleFunc DyadicBilinearOneThirdDownsampler_sse4; | 109 | SpecificDownsampleFunc DyadicBilinearOneThirdDownsampler_sse4; |
| @@ -122,6 +123,9 @@ void GeneralBilinearFastDownsampler_ssse3 (uint8_t* pDst, int32_t iDstStride, in · 13 unmodified lines | |||
| 122 | void GeneralBilinearAccurateDownsampler_sse41 (uint8_t* pDst, int32_t iDstStride, int32_t iDstWidth, | 123 | void GeneralBilinearAccurateDownsampler_sse41 (uint8_t* pDst, int32_t iDstStride, int32_t iDstWidth, |
| 123 | int32_t iDstHeight, uint8_t* pSrc, int32_t iSrcStride, uint32_t uiScaleX, | 124 | int32_t iDstHeight, uint8_t* pSrc, int32_t iSrcStride, uint32_t uiScaleX, |
| 124 | uint32_t uiScaleY); | 125 | uint32_t uiScaleY); |
| 126 | void GeneralBilinearFastDownsampler_avx2 (uint8_t* pDst, int32_t iDstStride, int32_t iDstWidth, | ||
| 127 | int32_t iDstHeight, uint8_t* pSrc, int32_t iSrcStride, uint32_t uiScaleX, | ||
| 128 | uint32_t uiScaleY); | ||
| 125 | 129 | ||
| 126 | WELSVP_EXTERN_C_END | 130 | WELSVP_EXTERN_C_END |
| 127 | #endif | 131 | #endif |
| @@ -284,6 +284,7 @@ DEFINE_GENERAL_BILINEAR_FAST_DOWNSAMPLER_WRAP (sse2) · 283 unmodified lines | |||
| 284 | DEFINE_GENERAL_BILINEAR_ACCURATE_DOWNSAMPLER_WRAP (sse2) | 284 | DEFINE_GENERAL_BILINEAR_ACCURATE_DOWNSAMPLER_WRAP (sse2) |
| 285 | DEFINE_GENERAL_BILINEAR_FAST_DOWNSAMPLER_WRAP (ssse3) | 285 | DEFINE_GENERAL_BILINEAR_FAST_DOWNSAMPLER_WRAP (ssse3) |
| 286 | DEFINE_GENERAL_BILINEAR_ACCURATE_DOWNSAMPLER_WRAP (sse41) | 286 | DEFINE_GENERAL_BILINEAR_ACCURATE_DOWNSAMPLER_WRAP (sse41) |
| 287 | DEFINE_GENERAL_BILINEAR_FAST_DOWNSAMPLER_WRAP (avx2) | ||
| 287 | #endif //X86_ASM | 288 | #endif //X86_ASM |
| 288 | 289 | ||
| 289 | #ifdef HAVE_NEON | 290 | #ifdef HAVE_NEON |