v1.16 release: OpenCL support, removing global selector palette code …

…(it was unused/dead code), ETC1S encoding optimizations, -parallel command line option

.gitignore

@@ -1,4 +1,3 @@
-bin/
 build/
 
 # VS project and working files

.reuse/dep5

@@ -7,22 +7,11 @@ Files: *
 Copyright: 2019-2021 Binomial LLC
 License: Apache-2.0
 
-Files: encoder/apg_bmp.*
-Copyright: 2019 Anton Gerdelan
-License: Apache-2.0
-
-Files: encoder/basisu_astc_decomp.*
-Copyright: 2016 The Android Open Source Project
-License: Apache-2.0
-
-Files: encoder/cppspmd_sse.h
-Copyright: Copyright 2016 Nicolas Guillemot
-License: Apache-2.0
-
-Files: encoder/lodepng.*
-Copyright: 2005-2019 Lode Vandevenne
-License: Zlib
-
 Files: zstd/*
 Copyright: 2016-present, Facebook, Inc.
 License: BSD
+
+Files: OpenCL/*
+Copyright: Copyright (c) 2008-2020 The Khronos Group Inc.
+License: Apache-2.0
+

CMakeLists.txt

@@ -1,15 +1,38 @@
 project(basisu)
 
 cmake_minimum_required(VERSION 3.0)
-option(BUILD_X64 "build 64-bit" TRUE)
 option(STATIC "static linking" FALSE)
-option(SSE "SSE 4.1 support" FALSE)
+
+# For MSVC builds default to SSE enabled, and determine if it's a 64-bit (-A x64) vs. 32-bit (-A Win32) build.
+if (MSVC)
+	option(SSE "SSE 4.1 support" TRUE)
+	if ( CMAKE_GENERATOR_PLATFORM STREQUAL Win32 )	
+		set(BUILD_X64 0)
+	else()
+		set(BUILD_X64 1)
+	endif()
+else()
+	option(SSE "SSE 4.1 support" FALSE)
+	option(BUILD_X64 "build 64-bit" TRUE)
+endif()
+
 option(ZSTD "ZSTD support for KTX2 transcoding/encoding" TRUE)
+option(OPENCL "OpenCL support in encoder" FALSE)
 
 message("Initial BUILD_X64=${BUILD_X64}")
 message("Initial CMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE}")
 message("Initial SSE=${SSE}")
 message("Initial ZSTD=${ZSTD}")
+message("Initial OPENCL=${OPENCL}")
+
+if (NOT MSVC)
+	# With MSVC builds we use the Khronos lib/include files in the project's "OpenCL" directory, to completely avoid requiring fiddly to install vendor SDK's.
+	# Otherwise we use the system's (if any).
+	find_package( OpenCL )
+	message(STATUS "OpenCL found: ${OPENCL_FOUND}")
+	message(STATUS "OpenCL includes: ${OpenCL_INCLUDE_DIRS}")
+	message(STATUS "OpenCL libraries: ${OpenCL_LIBRARIES}")
+endif()
 
 if( NOT CMAKE_BUILD_TYPE )
   set( CMAKE_BUILD_TYPE Release )
@@ -101,19 +124,17 @@ set(BASISU_SRC_LIST ${COMMON_SRC_LIST}
 	encoder/basisu_enc.cpp
 	encoder/basisu_etc.cpp
 	encoder/basisu_frontend.cpp
-	encoder/basisu_global_selector_palette_helpers.cpp
 	encoder/basisu_gpu_texture.cpp
 	encoder/basisu_pvrtc1_4.cpp
 	encoder/basisu_resampler.cpp
 	encoder/basisu_resample_filters.cpp
 	encoder/basisu_ssim.cpp
-	encoder/basisu_astc_decomp.cpp
 	encoder/basisu_uastc_enc.cpp
 	encoder/basisu_bc7enc.cpp
-	encoder/lodepng.cpp
-	encoder/apg_bmp.c
 	encoder/jpgd.cpp
 	encoder/basisu_kernels_sse.cpp
+	encoder/basisu_opencl.cpp
+	encoder/pvpngreader.cpp
 	transcoder/basisu_transcoder.cpp
 	)
 
@@ -138,7 +159,34 @@ else()
 endif()
 
 if (NOT MSVC)
-   target_link_libraries(basisu m pthread)
+	# For Non-Windows builds, let cmake try and find the system OpenCL headers/libs for us.
+	if (OPENCL_FOUND)
+		set(CMAKE_C_FLAGS  "${CMAKE_C_FLAGS} -DBASISU_SUPPORT_OPENCL=1")
+		set(CMAKE_CXX_FLAGS  "${CMAKE_CXX_FLAGS} -DBASISU_SUPPORT_OPENCL=1")	
+
+		target_include_directories( basisu PRIVATE ${OpenCL_INCLUDE_DIRS} )
+		set(BASISU_EXTRA_LIBS ${OpenCL_LIBRARIES})
+	endif()
+
+else()
+	# For Windows builds, we use our local copies of the OpenCL import lib and Khronos headers.
+	if (OPENCL)
+		set(CMAKE_C_FLAGS  "${CMAKE_C_FLAGS} -DBASISU_SUPPORT_OPENCL=1")
+		set(CMAKE_CXX_FLAGS  "${CMAKE_CXX_FLAGS} -DBASISU_SUPPORT_OPENCL=1")	
+
+		target_include_directories( basisu PRIVATE "OpenCL" )
+
+		if ( BUILD_X64 )
+			target_link_libraries( basisu PRIVATE "OpenCL/lib/OpenCL64" )
+		else()
+			target_link_libraries( basisu PRIVATE "OpenCL/lib/OpenCL" )
+		endif()
+
+	endif()
+endif()	
+
+if (NOT MSVC)
+   target_link_libraries(basisu m pthread ${BASISU_EXTRA_LIBS})
 endif()
 
 if (NOT EMSCRIPTEN)

basisu.vcxproj

@@ -27,26 +27,26 @@
   <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'" Label="Configuration">
     <ConfigurationType>Application</ConfigurationType>
     <UseDebugLibraries>true</UseDebugLibraries>
-    <PlatformToolset>v142</PlatformToolset>
+    <PlatformToolset>v143</PlatformToolset>
     <CharacterSet>MultiByte</CharacterSet>
   </PropertyGroup>
   <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'" Label="Configuration">
     <ConfigurationType>Application</ConfigurationType>
     <UseDebugLibraries>false</UseDebugLibraries>
-    <PlatformToolset>v142</PlatformToolset>
+    <PlatformToolset>v143</PlatformToolset>
     <WholeProgramOptimization>true</WholeProgramOptimization>
     <CharacterSet>MultiByte</CharacterSet>
   </PropertyGroup>
   <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'" Label="Configuration">
     <ConfigurationType>Application</ConfigurationType>
     <UseDebugLibraries>true</UseDebugLibraries>
-    <PlatformToolset>v142</PlatformToolset>
+    <PlatformToolset>v143</PlatformToolset>
     <CharacterSet>MultiByte</CharacterSet>
   </PropertyGroup>
   <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'" Label="Configuration">
     <ConfigurationType>Application</ConfigurationType>
     <UseDebugLibraries>false</UseDebugLibraries>
-    <PlatformToolset>v142</PlatformToolset>
+    <PlatformToolset>v143</PlatformToolset>
     <WholeProgramOptimization>true</WholeProgramOptimization>
     <CharacterSet>MultiByte</CharacterSet>
   </PropertyGroup>
@@ -87,27 +87,30 @@
       <Optimization>Disabled</Optimization>
       <SDLCheck>true</SDLCheck>
       <OpenMPSupport>true</OpenMPSupport>
-      <AdditionalIncludeDirectories>
-      </AdditionalIncludeDirectories>
-      <PreprocessorDefinitions>_MBCS;%(PreprocessorDefinitions);BASISU_SUPPORT_SSE=1</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>OpenCL</AdditionalIncludeDirectories>
+      <PreprocessorDefinitions>_MBCS;%(PreprocessorDefinitions);BASISU_SUPPORT_SSE=1;BASISU_SUPPORT_OPENCL=1</PreprocessorDefinitions>
       <EnableEnhancedInstructionSet>StreamingSIMDExtensions</EnableEnhancedInstructionSet>
     </ClCompile>
     <Link>
       <SubSystem>Console</SubSystem>
+      <AdditionalLibraryDirectories>OpenCL\lib</AdditionalLibraryDirectories>
+      <AdditionalDependencies>opencl.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
     </Link>
   </ItemDefinitionGroup>
   <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
     <ClCompile>
       <WarningLevel>Level4</WarningLevel>
       <Optimization>Disabled</Optimization>
       <SDLCheck>true</SDLCheck>
-      <AdditionalIncludeDirectories>ispc_texcomp</AdditionalIncludeDirectories>
+      <AdditionalIncludeDirectories>OpenCL</AdditionalIncludeDirectories>
       <MultiProcessorCompilation>true</MultiProcessorCompilation>
       <OpenMPSupport>true</OpenMPSupport>
-      <PreprocessorDefinitions>_MBCS;%(PreprocessorDefinitions);BASISU_SUPPORT_SSE=1;</PreprocessorDefinitions>
+      <PreprocessorDefinitions>_MBCS;%(PreprocessorDefinitions);BASISU_SUPPORT_SSE=1;BASISU_SUPPORT_OPENCL=1</PreprocessorDefinitions>
     </ClCompile>
     <Link>
       <SubSystem>Console</SubSystem>
+      <AdditionalLibraryDirectories>OpenCL\lib</AdditionalLibraryDirectories>
+      <AdditionalDependencies>opencl64.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
     </Link>
   </ItemDefinitionGroup>
   <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
@@ -117,9 +120,8 @@
       <FunctionLevelLinking>true</FunctionLevelLinking>
       <IntrinsicFunctions>true</IntrinsicFunctions>
       <OpenMPSupport>true</OpenMPSupport>
-      <AdditionalIncludeDirectories>
-      </AdditionalIncludeDirectories>
-      <PreprocessorDefinitions>NDEBUG;_HAS_EXCEPTIONS=0;%(PreprocessorDefinitions);BASISU_SUPPORT_SSE=1</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>OpenCL</AdditionalIncludeDirectories>
+      <PreprocessorDefinitions>NDEBUG;_HAS_EXCEPTIONS=0;%(PreprocessorDefinitions);BASISU_SUPPORT_SSE=1;BASISU_SUPPORT_OPENCL=1</PreprocessorDefinitions>
       <BufferSecurityCheck>false</BufferSecurityCheck>
       <InlineFunctionExpansion>AnySuitable</InlineFunctionExpansion>
       <EnableEnhancedInstructionSet>StreamingSIMDExtensions</EnableEnhancedInstructionSet>
@@ -131,6 +133,8 @@
       <EnableCOMDATFolding>true</EnableCOMDATFolding>
       <OptimizeReferences>true</OptimizeReferences>
       <SubSystem>Console</SubSystem>
+      <AdditionalLibraryDirectories>OpenCL\lib</AdditionalLibraryDirectories>
+      <AdditionalDependencies>opencl.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
     </Link>
   </ItemDefinitionGroup>
   <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
@@ -139,8 +143,8 @@
       <Optimization>Full</Optimization>
       <FunctionLevelLinking>true</FunctionLevelLinking>
       <IntrinsicFunctions>true</IntrinsicFunctions>
-      <AdditionalIncludeDirectories>ispc_texcomp</AdditionalIncludeDirectories>
-      <PreprocessorDefinitions>NDEBUG;_HAS_EXCEPTIONS=0;%(PreprocessorDefinitions);BASISU_SUPPORT_SSE=1;</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>OpenCL</AdditionalIncludeDirectories>
+      <PreprocessorDefinitions>NDEBUG;_HAS_EXCEPTIONS=0;%(PreprocessorDefinitions);BASISU_SUPPORT_SSE=1;BASISU_SUPPORT_OPENCL=1</PreprocessorDefinitions>
       <BufferSecurityCheck>false</BufferSecurityCheck>
       <MultiProcessorCompilation>true</MultiProcessorCompilation>
       <InlineFunctionExpansion>AnySuitable</InlineFunctionExpansion>
@@ -153,41 +157,41 @@
       <EnableCOMDATFolding>true</EnableCOMDATFolding>
       <OptimizeReferences>true</OptimizeReferences>
       <SubSystem>Console</SubSystem>
+      <AdditionalLibraryDirectories>OpenCL\lib</AdditionalLibraryDirectories>
+      <AdditionalDependencies>opencl64.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
     </Link>
   </ItemDefinitionGroup>
   <ItemGroup>
-    <ClCompile Include="encoder\apg_bmp.c" />
-    <ClCompile Include="encoder\basisu_astc_decomp.cpp" />
     <ClCompile Include="encoder\basisu_backend.cpp" />
     <ClCompile Include="encoder\basisu_basis_file.cpp" />
     <ClCompile Include="encoder\basisu_bc7enc.cpp" />
     <ClCompile Include="encoder\basisu_comp.cpp" />
     <ClCompile Include="encoder\basisu_enc.cpp" />
     <ClCompile Include="encoder\basisu_etc.cpp" />
     <ClCompile Include="encoder\basisu_frontend.cpp" />
-    <ClCompile Include="encoder\basisu_global_selector_palette_helpers.cpp" />
     <ClCompile Include="encoder\basisu_gpu_texture.cpp" />
     <ClCompile Include="encoder\basisu_kernels_sse.cpp" />
+    <ClCompile Include="encoder\basisu_opencl.cpp" />
     <ClCompile Include="encoder\basisu_pvrtc1_4.cpp" />
     <ClCompile Include="encoder\basisu_resampler.cpp" />
     <ClCompile Include="encoder\basisu_resample_filters.cpp" />
     <ClCompile Include="encoder\basisu_ssim.cpp" />
     <ClCompile Include="encoder\basisu_uastc_enc.cpp" />
     <ClCompile Include="encoder\jpgd.cpp" />
-    <ClCompile Include="encoder\lodepng.cpp" />
+    <ClCompile Include="encoder\pvpngreader.cpp" />
     <ClCompile Include="zstd\zstd.c" />
+    <ClInclude Include="encoder\basisu_ocl_kernels.h" />
+    <ClInclude Include="encoder\basisu_opencl.h" />
+    <ClInclude Include="encoder\pvpngreader.h" />
     <ClInclude Include="transcoder\basisu_containers_impl.h" />
     <ClCompile Include="transcoder\basisu_transcoder.cpp" />
-    <ClInclude Include="encoder\apg_bmp.h" />
-    <ClInclude Include="encoder\basisu_astc_decomp.h" />
     <ClInclude Include="encoder\basisu_backend.h" />
     <ClInclude Include="encoder\basisu_basis_file.h" />
     <ClInclude Include="encoder\basisu_bc7enc.h" />
     <ClInclude Include="encoder\basisu_comp.h" />
     <ClInclude Include="encoder\basisu_enc.h" />
     <ClInclude Include="encoder\basisu_etc.h" />
     <ClInclude Include="encoder\basisu_frontend.h" />
-    <ClInclude Include="encoder\basisu_global_selector_palette_helpers.h" />
     <ClInclude Include="encoder\basisu_gpu_texture.h" />
     <ClInclude Include="encoder\basisu_kernels_declares.h" />
     <ClInclude Include="encoder\basisu_kernels_imp.h" />
@@ -203,19 +207,18 @@
     <ClInclude Include="encoder\cppspmd_sse.h" />
     <ClInclude Include="encoder\cppspmd_type_aliases.h" />
     <ClInclude Include="encoder\jpgd.h" />
-    <ClInclude Include="encoder\lodepng.h" />
     <ClInclude Include="transcoder\basisu.h" />
     <ClInclude Include="transcoder\basisu_containers.h" />
     <ClInclude Include="transcoder\basisu_file_headers.h" />
     <ClInclude Include="transcoder\basisu_transcoder.h" />
     <ClInclude Include="transcoder\basisu_transcoder_internal.h" />
-    <ClInclude Include="transcoder\basisu_global_selector_palette.h" />
     <ClInclude Include="transcoder\basisu_transcoder_uastc.h" />
   </ItemGroup>
   <ItemGroup>
     <ClCompile Include="basisu_tool.cpp" />
   </ItemGroup>
   <ItemGroup>
+    <None Include="bin\ocl_kernels.cl" />
     <None Include="transcoder\basisu_transcoder_tables_astc.inc" />
     <None Include="transcoder\basisu_transcoder_tables_astc_0_255.inc" />
     <None Include="transcoder\basisu_transcoder_tables_atc_55.inc" />