From e510a287deb25f2542a68bf21382f2557740d70c Mon Sep 17 00:00:00 2001 From: jsmall-nvidia Date: Thu, 6 May 2021 12:45:00 -0400 Subject: Half texture support (#1836) * #include an absolute path didn't work - because paths were taken to always be relative. * Split out StringEscapeUtil. * Added StringEscapeUtil. * Fix typo in unix quoting type. * Small comment improvements. * Try to fix linux linking issue. * Fix typo. * Attempt to fix linux link issue. * Update VS proj even though nothing really changed. * Fix another typo issue. * Fix for windows issue. Fixed bug. * Make separate Utils for escaping. * Fix typo. * Split out into StringEscapeHandler. * Windows shell does handle removing quotes (so remove code to remove them). * Handle unescaping if not initiating using the shell. * Slight improvement around shell like decoding. * Simplify command extraction. * Add shared-library category type. * Fix bug in command extraction. * Typo in transcendental category. * Enable unit-test on in smoke test category. * Make parsing failing output as a failing test. * Fixes for transcendental tests. Disable tests that do not work. * Changed category parsing. * Removed the TestResult parameter from _gatherTestsForFile. Made testsList only output. * Remove testing if all tests were disabled. * Make args of CommandLine always unescaped. * Add category. * Don't need escaping on unix/linux. * Remove some no longer used functions. * Add requireSMVersion to CUDAExtensionTracker. * half-calc.slang now works for CUDA. * bit-cast-16-bit works on CUDA. * WIP handling of CUDA vector types. * Half swizzle CUDA. * Half vector test. * Fix swizzle half bug. * Fix compilation issue with narrowing to Index. * Add unary ops. * Add some vector scalar maths ops. * Add half vector conversions for CUDA. * Fix erroneous comment. * Support for half comparisons. * First pass test for half compare. * Fix bug in CUDA specialized emit control. Updated tests to have pre and post inc/dec. * Removed unneeded parts of the cuda prelude. * Half structured buffer works on CUDA. * Added name lookup for Gfx::Format * Support half texture type in test system. * Test for half reading on CUDA. * Add half formats to Vk and D3D utils. * Fix getAt for CUDA - where there might not be a .x member in a vector. --- tools/gfx/cpu/render-cpu.cpp | 73 ++++++++++++++++++++++++++------------------ 1 file changed, 44 insertions(+), 29 deletions(-) (limited to 'tools/gfx/cpu/render-cpu.cpp') diff --git a/tools/gfx/cpu/render-cpu.cpp b/tools/gfx/cpu/render-cpu.cpp index 2d7d858c4..0bdc06ad6 100644 --- a/tools/gfx/cpu/render-cpu.cpp +++ b/tools/gfx/cpu/render-cpu.cpp @@ -89,6 +89,18 @@ void _unpackFloatTexel(void const* texelData, void* outData, size_t outSize) memcpy(outData, temp, outSize); } +template +void _unpackFloat16Texel(void const* texelData, void* outData, size_t outSize) +{ + auto input = (int16_t const*)texelData; + + float temp[4] = { 0.0f, 0.0f, 0.0f, 1.0f }; + for (int i = 0; i < N; ++i) + temp[i] = HalfToFloat(input[i]); + + memcpy(outData, temp, outSize); +} + static inline float _unpackUnorm8Value(uint8_t value) { return value / 255.0f; @@ -143,42 +155,45 @@ void _unpackUInt32Texel(void const* texelData, void* outData, size_t outSize) memcpy(outData, temp, outSize); } -#define TEXTURE_FORMAT_INFO(FORMAT) static const CPUTextureFormatInfo kCPUTextureFormatInfo_##FORMAT +struct CPUFormatInfoMap +{ + CPUFormatInfoMap() + { + memset(m_infos, 0, sizeof(m_infos)); -TEXTURE_FORMAT_INFO(RGBA_Float32) = { &_unpackFloatTexel<4> }; -TEXTURE_FORMAT_INFO(RGB_Float32) = { &_unpackFloatTexel<3> }; -TEXTURE_FORMAT_INFO(RG_Float32) = { &_unpackFloatTexel<2> }; -TEXTURE_FORMAT_INFO(R_Float32) = { &_unpackFloatTexel<1> }; -TEXTURE_FORMAT_INFO(RGBA_Unorm_UInt8) = { &_unpackUnorm8Texel<4> }; -TEXTURE_FORMAT_INFO(BGRA_Unorm_UInt8) = { &_unpackUnormBGRA8Texel }; -TEXTURE_FORMAT_INFO(R_UInt16) = { &_unpackUInt16Texel<1> }; -TEXTURE_FORMAT_INFO(R_UInt32) = { &_unpackUInt32Texel<1> }; -TEXTURE_FORMAT_INFO(D_Float32) = { &_unpackFloatTexel<1> }; + set(Format::RGBA_Float32, &_unpackFloatTexel<4>); + set(Format::RGB_Float32, &_unpackFloatTexel<3>); -#undef TEXTURE_FORMAT_INFO + set(Format::RG_Float32, &_unpackFloatTexel<2>); + set(Format::R_Float32, &_unpackFloatTexel<1>); -static CPUTextureFormatInfo const* _getFormatInfo(Format format) -{ - switch(format) + set(Format::RGBA_Float16, &_unpackFloat16Texel<4>); + set(Format::RG_Float16, &_unpackFloat16Texel<2>); + set(Format::R_Float16, &_unpackFloat16Texel<1>); + + set(Format::RGBA_Unorm_UInt8, &_unpackUnorm8Texel<4>); + set(Format::BGRA_Unorm_UInt8, &_unpackUnormBGRA8Texel); + set(Format::R_UInt16, &_unpackUInt16Texel<1>); + set(Format::R_UInt32, &_unpackUInt32Texel<1>); + set(Format::D_Float32, &_unpackFloatTexel<1>); + } + + void set(Format format, CPUTextureUnpackFunc func) { - case Format::D_Unorm24_S8: - default: - return nullptr; + auto& info = m_infos[Index(format)]; + info.unpackFunc = func; + } + SLANG_FORCE_INLINE const CPUTextureFormatInfo& get(Format format) const { return m_infos[Index(format)]; } + CPUTextureFormatInfo m_infos[Index(Format::CountOf)]; +}; -#define CASE(FORMAT) case Format::FORMAT: return &kCPUTextureFormatInfo_##FORMAT; - CASE(RGBA_Float32) - CASE(RGB_Float32) - CASE(RG_Float32) - CASE(R_Float32) - CASE(RGBA_Unorm_UInt8) - CASE(BGRA_Unorm_UInt8) - CASE(R_UInt16) - CASE(R_UInt32) - CASE(D_Float32) +static const CPUFormatInfoMap g_formatInfoMap; -#undef CASE - } +static CPUTextureFormatInfo const* _getFormatInfo(Format format) +{ + const CPUTextureFormatInfo& info = g_formatInfoMap.get(format); + return info.unpackFunc ? &info : nullptr; } class CPUTextureResource : public TextureResource -- cgit v1.2.3