#include #include #include #include #include #include #include namespace mask { template struct type { genType Value; genType Return; }; inline int mask_zero(int Bits) { return ~((~0) << Bits); } inline int mask_mix(int Bits) { return Bits >= sizeof(int) * 8 ? 0xffffffff : (static_cast(1) << Bits) - static_cast(1); } inline int mask_half(int Bits) { // We do the shift in two steps because 1 << 32 on an int is undefined. int const Half = Bits >> 1; int const Fill = ~0; int const ShiftHaft = (Fill << Half); int const Rest = Bits - Half; int const Reversed = ShiftHaft << Rest; return ~Reversed; } inline int mask_loop(int Bits) { int Mask = 0; for(int Bit = 0; Bit < Bits; ++Bit) Mask |= (static_cast(1) << Bit); return Mask; } int perf() { int const Count = 100000000; std::clock_t Timestamp1 = std::clock(); { std::vector Mask; Mask.resize(Count); for(int i = 0; i < Count; ++i) Mask[i] = mask_mix(i % 32); } std::clock_t Timestamp2 = std::clock(); { std::vector Mask; Mask.resize(Count); for(int i = 0; i < Count; ++i) Mask[i] = mask_loop(i % 32); } std::clock_t Timestamp3 = std::clock(); { std::vector Mask; Mask.resize(Count); for(int i = 0; i < Count; ++i) Mask[i] = glm::mask(i % 32); } std::clock_t Timestamp4 = std::clock(); { std::vector Mask; Mask.resize(Count); for(int i = 0; i < Count; ++i) Mask[i] = mask_zero(i % 32); } std::clock_t Timestamp5 = std::clock(); { std::vector Mask; Mask.resize(Count); for(int i = 0; i < Count; ++i) Mask[i] = mask_half(i % 32); } std::clock_t Timestamp6 = std::clock(); std::clock_t TimeMix = Timestamp2 - Timestamp1; std::clock_t TimeLoop = Timestamp3 - Timestamp2; std::clock_t TimeDefault = Timestamp4 - Timestamp3; std::clock_t TimeZero = Timestamp5 - Timestamp4; std::clock_t TimeHalf = Timestamp6 - Timestamp5; printf("mask[mix]: %d\n", static_cast(TimeMix)); printf("mask[loop]: %d\n", static_cast(TimeLoop)); printf("mask[default]: %d\n", static_cast(TimeDefault)); printf("mask[zero]: %d\n", static_cast(TimeZero)); printf("mask[half]: %d\n", static_cast(TimeHalf)); return TimeDefault < TimeLoop ? 0 : 1; } int test_uint() { type const Data[] = { { 0, 0x00000000}, { 1, 0x00000001}, { 2, 0x00000003}, { 3, 0x00000007}, {31, 0x7fffffff}, {32, 0xffffffff} }; int Error(0); /* mask_zero is sadly not a correct code for(std::size_t i = 0; i < sizeof(Data) / sizeof(type); ++i) { int Result = mask_zero(Data[i].Value); Error += Data[i].Return == Result ? 0 : 1; } */ for(std::size_t i = 0; i < sizeof(Data) / sizeof(type); ++i) { int Result = mask_mix(Data[i].Value); Error += Data[i].Return == Result ? 0 : 1; } for(std::size_t i = 0; i < sizeof(Data) / sizeof(type); ++i) { int Result = mask_half(Data[i].Value); Error += Data[i].Return == Result ? 0 : 1; } for(std::size_t i = 0; i < sizeof(Data) / sizeof(type); ++i) { int Result = mask_loop(Data[i].Value); Error += Data[i].Return == Result ? 0 : 1; } for(std::size_t i = 0; i < sizeof(Data) / sizeof(type); ++i) { int Result = glm::mask(Data[i].Value); Error += Data[i].Return == Result ? 0 : 1; } return Error; } int test_uvec4() { type const Data[] = { {glm::ivec4( 0), glm::ivec4(0x00000000)}, {glm::ivec4( 1), glm::ivec4(0x00000001)}, {glm::ivec4( 2), glm::ivec4(0x00000003)}, {glm::ivec4( 3), glm::ivec4(0x00000007)}, {glm::ivec4(31), glm::ivec4(0x7fffffff)}, {glm::ivec4(32), glm::ivec4(0xffffffff)} }; int Error(0); for(std::size_t i = 0, n = sizeof(Data) / sizeof(type); i < n; ++i) { glm::ivec4 Result = glm::mask(Data[i].Value); Error += glm::all(glm::equal(Data[i].Return, Result)) ? 0 : 1; } return Error; } int test() { int Error(0); Error += test_uint(); Error += test_uvec4(); return Error; } }//namespace mask namespace bitfieldInterleave3 { template inline RET refBitfieldInterleave(PARAM x, PARAM y, PARAM z) { RET Result = 0; for(RET i = 0; i < sizeof(PARAM) * 8; ++i) { Result |= ((RET(x) & (RET(1U) << i)) << ((i << 1) + 0)); Result |= ((RET(y) & (RET(1U) << i)) << ((i << 1) + 1)); Result |= ((RET(z) & (RET(1U) << i)) << ((i << 1) + 2)); } return Result; } int test() { int Error(0); glm::uint16 x_max = 1 << 11; glm::uint16 y_max = 1 << 11; glm::uint16 z_max = 1 << 11; for(glm::uint16 z = 0; z < z_max; z += 27) for(glm::uint16 y = 0; y < y_max; y += 27) for(glm::uint16 x = 0; x < x_max; x += 27) { glm::uint64 ResultA = refBitfieldInterleave(x, y, z); glm::uint64 ResultB = glm::bitfieldInterleave(x, y, z); Error += ResultA == ResultB ? 0 : 1; } return Error; } } namespace bitfieldInterleave4 { template inline RET loopBitfieldInterleave(PARAM x, PARAM y, PARAM z, PARAM w) { RET const v[4] = {x, y, z, w}; RET Result = 0; for(RET i = 0; i < sizeof(PARAM) * 8; i++) { Result |= ((((v[0] >> i) & 1U)) << ((i << 2) + 0)); Result |= ((((v[1] >> i) & 1U)) << ((i << 2) + 1)); Result |= ((((v[2] >> i) & 1U)) << ((i << 2) + 2)); Result |= ((((v[3] >> i) & 1U)) << ((i << 2) + 3)); } return Result; } int test() { int Error(0); glm::uint16 x_max = 1 << 11; glm::uint16 y_max = 1 << 11; glm::uint16 z_max = 1 << 11; glm::uint16 w_max = 1 << 11; for(glm::uint16 w = 0; w < w_max; w += 27) for(glm::uint16 z = 0; z < z_max; z += 27) for(glm::uint16 y = 0; y < y_max; y += 27) for(glm::uint16 x = 0; x < x_max; x += 27) { glm::uint64 ResultA = loopBitfieldInterleave(x, y, z, w); glm::uint64 ResultB = glm::bitfieldInterleave(x, y, z, w); Error += ResultA == ResultB ? 0 : 1; } return Error; } } namespace bitfieldInterleave { inline glm::uint64 fastBitfieldInterleave(glm::uint32 x, glm::uint32 y) { glm::uint64 REG1; glm::uint64 REG2; REG1 = x; REG1 = ((REG1 << 16) | REG1) & glm::uint64(0x0000FFFF0000FFFF); REG1 = ((REG1 << 8) | REG1) & glm::uint64(0x00FF00FF00FF00FF); REG1 = ((REG1 << 4) | REG1) & glm::uint64(0x0F0F0F0F0F0F0F0F); REG1 = ((REG1 << 2) | REG1) & glm::uint64(0x3333333333333333); REG1 = ((REG1 << 1) | REG1) & glm::uint64(0x5555555555555555); REG2 = y; REG2 = ((REG2 << 16) | REG2) & glm::uint64(0x0000FFFF0000FFFF); REG2 = ((REG2 << 8) | REG2) & glm::uint64(0x00FF00FF00FF00FF); REG2 = ((REG2 << 4) | REG2) & glm::uint64(0x0F0F0F0F0F0F0F0F); REG2 = ((REG2 << 2) | REG2) & glm::uint64(0x3333333333333333); REG2 = ((REG2 << 1) | REG2) & glm::uint64(0x5555555555555555); return REG1 | (REG2 << 1); } inline glm::uint64 interleaveBitfieldInterleave(glm::uint32 x, glm::uint32 y) { glm::uint64 REG1; glm::uint64 REG2; REG1 = x; REG2 = y; REG1 = ((REG1 << 16) | REG1) & glm::uint64(0x0000FFFF0000FFFF); REG2 = ((REG2 << 16) | REG2) & glm::uint64(0x0000FFFF0000FFFF); REG1 = ((REG1 << 8) | REG1) & glm::uint64(0x00FF00FF00FF00FF); REG2 = ((REG2 << 8) | REG2) & glm::uint64(0x00FF00FF00FF00FF); REG1 = ((REG1 << 4) | REG1) & glm::uint64(0x0F0F0F0F0F0F0F0F); REG2 = ((REG2 << 4) | REG2) & glm::uint64(0x0F0F0F0F0F0F0F0F); REG1 = ((REG1 << 2) | REG1) & glm::uint64(0x3333333333333333); REG2 = ((REG2 << 2) | REG2) & glm::uint64(0x3333333333333333); REG1 = ((REG1 << 1) | REG1) & glm::uint64(0x5555555555555555); REG2 = ((REG2 << 1) | REG2) & glm::uint64(0x5555555555555555); return REG1 | (REG2 << 1); } /* inline glm::uint64 loopBitfieldInterleave(glm::uint32 x, glm::uint32 y) { static glm::uint64 const Mask[5] = { 0x5555555555555555, 0x3333333333333333, 0x0F0F0F0F0F0F0F0F, 0x00FF00FF00FF00FF, 0x0000FFFF0000FFFF }; glm::uint64 REG1 = x; glm::uint64 REG2 = y; for(int i = 4; i >= 0; --i) { REG1 = ((REG1 << (1 << i)) | REG1) & Mask[i]; REG2 = ((REG2 << (1 << i)) | REG2) & Mask[i]; } return REG1 | (REG2 << 1); } */ #if(GLM_ARCH != GLM_ARCH_PURE) inline glm::uint64 sseBitfieldInterleave(glm::uint32 x, glm::uint32 y) { GLM_ALIGN(16) glm::uint32 const Array[4] = {x, 0, y, 0}; __m128i const Mask4 = _mm_set1_epi32(0x0000FFFF); __m128i const Mask3 = _mm_set1_epi32(0x00FF00FF); __m128i const Mask2 = _mm_set1_epi32(0x0F0F0F0F); __m128i const Mask1 = _mm_set1_epi32(0x33333333); __m128i const Mask0 = _mm_set1_epi32(0x55555555); __m128i Reg1; __m128i Reg2; // REG1 = x; // REG2 = y; Reg1 = _mm_load_si128((__m128i*)Array); //REG1 = ((REG1 << 16) | REG1) & glm::uint64(0x0000FFFF0000FFFF); //REG2 = ((REG2 << 16) | REG2) & glm::uint64(0x0000FFFF0000FFFF); Reg2 = _mm_slli_si128(Reg1, 2); Reg1 = _mm_or_si128(Reg2, Reg1); Reg1 = _mm_and_si128(Reg1, Mask4); //REG1 = ((REG1 << 8) | REG1) & glm::uint64(0x00FF00FF00FF00FF); //REG2 = ((REG2 << 8) | REG2) & glm::uint64(0x00FF00FF00FF00FF); Reg2 = _mm_slli_si128(Reg1, 1); Reg1 = _mm_or_si128(Reg2, Reg1); Reg1 = _mm_and_si128(Reg1, Mask3); //REG1 = ((REG1 << 4) | REG1) & glm::uint64(0x0F0F0F0F0F0F0F0F); //REG2 = ((REG2 << 4) | REG2) & glm::uint64(0x0F0F0F0F0F0F0F0F); Reg2 = _mm_slli_epi32(Reg1, 4); Reg1 = _mm_or_si128(Reg2, Reg1); Reg1 = _mm_and_si128(Reg1, Mask2); //REG1 = ((REG1 << 2) | REG1) & glm::uint64(0x3333333333333333); //REG2 = ((REG2 << 2) | REG2) & glm::uint64(0x3333333333333333); Reg2 = _mm_slli_epi32(Reg1, 2); Reg1 = _mm_or_si128(Reg2, Reg1); Reg1 = _mm_and_si128(Reg1, Mask1); //REG1 = ((REG1 << 1) | REG1) & glm::uint64(0x5555555555555555); //REG2 = ((REG2 << 1) | REG2) & glm::uint64(0x5555555555555555); Reg2 = _mm_slli_epi32(Reg1, 1); Reg1 = _mm_or_si128(Reg2, Reg1); Reg1 = _mm_and_si128(Reg1, Mask0); //return REG1 | (REG2 << 1); Reg2 = _mm_slli_epi32(Reg1, 1); Reg2 = _mm_srli_si128(Reg2, 8); Reg1 = _mm_or_si128(Reg1, Reg2); GLM_ALIGN(16) glm::uint64 Result[2]; _mm_store_si128((__m128i*)Result, Reg1); return Result[0]; } inline glm::uint64 sseUnalignedBitfieldInterleave(glm::uint32 x, glm::uint32 y) { glm::uint32 const Array[4] = {x, 0, y, 0}; __m128i const Mask4 = _mm_set1_epi32(0x0000FFFF); __m128i const Mask3 = _mm_set1_epi32(0x00FF00FF); __m128i const Mask2 = _mm_set1_epi32(0x0F0F0F0F); __m128i const Mask1 = _mm_set1_epi32(0x33333333); __m128i const Mask0 = _mm_set1_epi32(0x55555555); __m128i Reg1; __m128i Reg2; // REG1 = x; // REG2 = y; Reg1 = _mm_loadu_si128((__m128i*)Array); //REG1 = ((REG1 << 16) | REG1) & glm::uint64(0x0000FFFF0000FFFF); //REG2 = ((REG2 << 16) | REG2) & glm::uint64(0x0000FFFF0000FFFF); Reg2 = _mm_slli_si128(Reg1, 2); Reg1 = _mm_or_si128(Reg2, Reg1); Reg1 = _mm_and_si128(Reg1, Mask4); //REG1 = ((REG1 << 8) | REG1) & glm::uint64(0x00FF00FF00FF00FF); //REG2 = ((REG2 << 8) | REG2) & glm::uint64(0x00FF00FF00FF00FF); Reg2 = _mm_slli_si128(Reg1, 1); Reg1 = _mm_or_si128(Reg2, Reg1); Reg1 = _mm_and_si128(Reg1, Mask3); //REG1 = ((REG1 << 4) | REG1) & glm::uint64(0x0F0F0F0F0F0F0F0F); //REG2 = ((REG2 << 4) | REG2) & glm::uint64(0x0F0F0F0F0F0F0F0F); Reg2 = _mm_slli_epi32(Reg1, 4); Reg1 = _mm_or_si128(Reg2, Reg1); Reg1 = _mm_and_si128(Reg1, Mask2); //REG1 = ((REG1 << 2) | REG1) & glm::uint64(0x3333333333333333); //REG2 = ((REG2 << 2) | REG2) & glm::uint64(0x3333333333333333); Reg2 = _mm_slli_epi32(Reg1, 2); Reg1 = _mm_or_si128(Reg2, Reg1); Reg1 = _mm_and_si128(Reg1, Mask1); //REG1 = ((REG1 << 1) | REG1) & glm::uint64(0x5555555555555555); //REG2 = ((REG2 << 1) | REG2) & glm::uint64(0x5555555555555555); Reg2 = _mm_slli_epi32(Reg1, 1); Reg1 = _mm_or_si128(Reg2, Reg1); Reg1 = _mm_and_si128(Reg1, Mask0); //return REG1 | (REG2 << 1); Reg2 = _mm_slli_epi32(Reg1, 1); Reg2 = _mm_srli_si128(Reg2, 8); Reg1 = _mm_or_si128(Reg1, Reg2); glm::uint64 Result[2]; _mm_storeu_si128((__m128i*)Result, Reg1); return Result[0]; } #endif//(GLM_ARCH != GLM_ARCH_PURE) int test() { { for(glm::uint32 y = 0; y < (1 << 10); ++y) for(glm::uint32 x = 0; x < (1 << 10); ++x) { glm::uint64 A = glm::bitfieldInterleave(x, y); glm::uint64 B = fastBitfieldInterleave(x, y); //glm::uint64 C = loopBitfieldInterleave(x, y); glm::uint64 D = interleaveBitfieldInterleave(x, y); assert(A == B); //assert(A == C); assert(A == D); # if GLM_ARCH & GLM_ARCH_SSE2_BIT glm::uint64 E = sseBitfieldInterleave(x, y); glm::uint64 F = sseUnalignedBitfieldInterleave(x, y); assert(A == E); assert(A == F); __m128i G = glm_i128_interleave(_mm_set_epi32(0, y, 0, x)); glm::uint64 Result[2]; _mm_storeu_si128((__m128i*)Result, G); assert(A == Result[0]); # endif//GLM_ARCH & GLM_ARCH_SSE2_BIT } } { for(glm::uint8 y = 0; y < 127; ++y) for(glm::uint8 x = 0; x < 127; ++x) { glm::uint64 A(glm::bitfieldInterleave(glm::uint8(x), glm::uint8(y))); glm::uint64 B(glm::bitfieldInterleave(glm::uint16(x), glm::uint16(y))); glm::uint64 C(glm::bitfieldInterleave(glm::uint32(x), glm::uint32(y))); glm::int64 D(glm::bitfieldInterleave(glm::int8(x), glm::int8(y))); glm::int64 E(glm::bitfieldInterleave(glm::int16(x), glm::int16(y))); glm::int64 F(glm::bitfieldInterleave(glm::int32(x), glm::int32(y))); assert(D == E); assert(D == F); } } return 0; } int perf() { glm::uint32 x_max = 1 << 11; glm::uint32 y_max = 1 << 10; // ALU std::vector Data(x_max * y_max); std::vector Param(x_max * y_max); for(glm::uint32 i = 0; i < Param.size(); ++i) Param[i] = glm::u32vec2(i % x_max, i / y_max); { std::clock_t LastTime = std::clock(); for(std::size_t i = 0; i < Data.size(); ++i) Data[i] = glm::bitfieldInterleave(Param[i].x, Param[i].y); std::clock_t Time = std::clock() - LastTime; std::printf("glm::bitfieldInterleave Time %d clocks\n", static_cast(Time)); } { std::clock_t LastTime = std::clock(); for(std::size_t i = 0; i < Data.size(); ++i) Data[i] = fastBitfieldInterleave(Param[i].x, Param[i].y); std::clock_t Time = std::clock() - LastTime; std::printf("fastBitfieldInterleave Time %d clocks\n", static_cast(Time)); } /* { std::clock_t LastTime = std::clock(); for(std::size_t i = 0; i < Data.size(); ++i) Data[i] = loopBitfieldInterleave(Param[i].x, Param[i].y); std::clock_t Time = std::clock() - LastTime; std::printf("loopBitfieldInterleave Time %d clocks\n", static_cast(Time)); } */ { std::clock_t LastTime = std::clock(); for(std::size_t i = 0; i < Data.size(); ++i) Data[i] = interleaveBitfieldInterleave(Param[i].x, Param[i].y); std::clock_t Time = std::clock() - LastTime; std::printf("interleaveBitfieldInterleave Time %d clocks\n", static_cast(Time)); } # if(GLM_ARCH != GLM_ARCH_PURE) { std::clock_t LastTime = std::clock(); for(std::size_t i = 0; i < Data.size(); ++i) Data[i] = sseBitfieldInterleave(Param[i].x, Param[i].y); std::clock_t Time = std::clock() - LastTime; std::printf("sseBitfieldInterleave Time %d clocks\n", static_cast(Time)); } { std::clock_t LastTime = std::clock(); for(std::size_t i = 0; i < Data.size(); ++i) Data[i] = sseUnalignedBitfieldInterleave(Param[i].x, Param[i].y); std::clock_t Time = std::clock() - LastTime; std::printf("sseUnalignedBitfieldInterleave Time %d clocks\n", static_cast(Time)); } # endif//(GLM_ARCH != GLM_ARCH_PURE) { std::clock_t LastTime = std::clock(); for(std::size_t i = 0; i < Data.size(); ++i) Data[i] = glm::bitfieldInterleave(Param[i].x, Param[i].y, Param[i].x); std::clock_t Time = std::clock() - LastTime; std::printf("glm::detail::bitfieldInterleave Time %d clocks\n", static_cast(Time)); } # if(GLM_ARCH & GLM_ARCH_SSE2_BIT && !(GLM_COMPILER & GLM_COMPILER_GCC)) { // SIMD std::vector<__m128i> SimdData; SimdData.resize(x_max * y_max); std::vector<__m128i> SimdParam; SimdParam.resize(x_max * y_max); for(int i = 0; i < SimdParam.size(); ++i) SimdParam[i] = _mm_set_epi32(i % x_max, 0, i / y_max, 0); std::clock_t LastTime = std::clock(); for(std::size_t i = 0; i < SimdData.size(); ++i) SimdData[i] = glm_i128_interleave(SimdParam[i]); std::clock_t Time = std::clock() - LastTime; std::printf("_mm_bit_interleave_si128 Time %d clocks\n", static_cast(Time)); } # endif//GLM_ARCH & GLM_ARCH_SSE2_BIT return 0; } }//namespace bitfieldInterleave int main() { int Error(0); Error += ::mask::test(); Error += ::bitfieldInterleave3::test(); Error += ::bitfieldInterleave4::test(); Error += ::bitfieldInterleave::test(); //Error += ::bitRevert::test(); # ifdef NDEBUG Error += ::mask::perf(); Error += ::bitfieldInterleave::perf(); # endif//NDEBUG return Error; }