Merge 228c97725a into e9a8630a3f

CMakeLists.txt

@@ -1,15 +1,10 @@
-CMAKE_MINIMUM_REQUIRED(VERSION 2.8.0)
+CMAKE_MINIMUM_REQUIRED(VERSION 2.8.12)
 PROJECT(NV)
 ENABLE_TESTING()
 
 SET(NV_CMAKE_DIR "${NV_SOURCE_DIR}/cmake")
 SET(CMAKE_MODULE_PATH ${CMAKE_MODULE_PATH} "${NV_CMAKE_DIR}")
 
-# GCC check (needs -std:c++11 flag)
-if(CMAKE_COMPILER_IS_GNUCC)
-	ADD_DEFINITIONS("-std=c++11")
-ENDIF(CMAKE_COMPILER_IS_GNUCC)
-
 IF(WIN32)
 	# gnuwin32 paths:
 	SET(GNUWIN32_PATH "${NV_SOURCE_DIR}/extern/gnuwin32")
@@ -29,6 +24,10 @@ MESSAGE(STATUS "Setting optimal options")
 MESSAGE(STATUS "  Processor: ${NV_SYSTEM_PROCESSOR}")
 MESSAGE(STATUS "  Compiler Flags: ${CMAKE_CXX_FLAGS}")
 
+IF(CMAKE_BUILD_TYPE MATCHES "debug")
+	SET(CMAKE_DEBUG_POSTFIX "_d" CACHE STRING "Postfix for debug build libraries.")
+ENDIF()
+
 IF(NVTT_SHARED)
 	SET(NVCORE_SHARED TRUE)
 	SET(NVMATH_SHARED TRUE)

cmake/run_nvcc.cmake

@@ -88,6 +88,7 @@ endif()
 # and other return variables are present after executing the process.
 macro(cuda_execute_process status command)
   set(_command ${command})
+  cmake_policy(SET CMP0054 OLD)
   if(NOT _command STREQUAL "COMMAND")
     message(FATAL_ERROR "Malformed call to cuda_execute_process.  Missing COMMAND as second argument. (command = ${command})")
   endif()

configure

@@ -61,15 +61,15 @@ echo ""
 
 cat > Makefile << EOF
 all:
-	@make --no-print-directory -C build-$build/
+	@+make --no-print-directory -C build-$build/
 install:
-	@make install --no-print-directory -C build-$build/
+	@+make install --no-print-directory -C build-$build/
 package:
-	@make package --no-print-directory -C build-$build/
+	@+make package --no-print-directory -C build-$build/
 test:
-	@make test --no-print-directory -C build-$build/
+	@+make test --no-print-directory -C build-$build/
 clean:
-	@make clean --no-print-directory -C build-$build/
+	@+make clean --no-print-directory -C build-$build/
 distclean:
 	@rm -Rf build-$build/
 EOF

project/vc12/nvcore/nvcore.vcxproj

@@ -157,6 +157,7 @@
     <ClInclude Include="..\..\..\src\nvcore\Debug.h" />
     <ClInclude Include="..\..\..\src\nvcore\DefsVcWin32.h" />
     <ClInclude Include="..\..\..\src\nvcore\FileSystem.h" />
+    <ClInclude Include="..\..\..\src\nvcore\Library.h" />
     <ClInclude Include="..\..\..\src\nvcore\Memory.h" />
     <ClInclude Include="..\..\..\src\nvcore\nvcore.h" />
     <ClInclude Include="..\..\..\src\nvcore\Ptr.h" />
@@ -168,6 +169,7 @@
   <ItemGroup>
     <ClCompile Include="..\..\..\src\nvcore\Debug.cpp" />
     <ClCompile Include="..\..\..\src\nvcore\FileSystem.cpp" />
+    <ClCompile Include="..\..\..\src\nvcore\Library.cpp" />
     <ClCompile Include="..\..\..\src\nvcore\Memory.cpp" />
     <ClCompile Include="..\..\..\src\nvcore\StrLib.cpp" />
     <ClCompile Include="..\..\..\src\nvcore\TextWriter.cpp" />

project/vc12/nvdecompress/nvdecompress.vcxproj

@@ -229,14 +229,6 @@
       <Project>{50c465fe-b308-42bc-894d-89484482af06}</Project>
       <ReferenceOutputAssembly>false</ReferenceOutputAssembly>
     </ProjectReference>
-    <ProjectReference Include="..\nvtt\nvtt.vcxproj">
-      <Project>{1aeb7681-57d8-48ee-813d-5c41cc38b647}</Project>
-      <Private>false</Private>
-      <ReferenceOutputAssembly>false</ReferenceOutputAssembly>
-      <CopyLocalSatelliteAssemblies>false</CopyLocalSatelliteAssemblies>
-      <LinkLibraryDependencies>true</LinkLibraryDependencies>
-      <UseLibraryDependencyInputs>false</UseLibraryDependencyInputs>
-    </ProjectReference>
   </ItemGroup>
   <Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
   <ImportGroup Label="ExtensionTargets">

project/vc12/nvimgdiff/nvimgdiff.vcxproj

@@ -231,14 +231,6 @@
       <Project>{50c465fe-b308-42bc-894d-89484482af06}</Project>
       <ReferenceOutputAssembly>false</ReferenceOutputAssembly>
     </ProjectReference>
-    <ProjectReference Include="..\nvtt\nvtt.vcxproj">
-      <Project>{1aeb7681-57d8-48ee-813d-5c41cc38b647}</Project>
-      <Private>false</Private>
-      <ReferenceOutputAssembly>false</ReferenceOutputAssembly>
-      <CopyLocalSatelliteAssemblies>false</CopyLocalSatelliteAssemblies>
-      <LinkLibraryDependencies>true</LinkLibraryDependencies>
-      <UseLibraryDependencyInputs>false</UseLibraryDependencyInputs>
-    </ProjectReference>
   </ItemGroup>
   <Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
   <ImportGroup Label="ExtensionTargets">

project/vc12/nvtt.sln

@@ -378,4 +378,7 @@ Global
 	GlobalSection(SolutionProperties) = preSolution
 		HideSolutionNode = FALSE
 	EndGlobalSection
+	GlobalSection(Performance) = preSolution
+		HasPerformanceSessions = true
+	EndGlobalSection
 EndGlobal

project/vc9/nvcore/nvcore.vcproj

@@ -290,10 +290,6 @@
 			RelativePath="..\..\..\src\nvcore\Array.h"
 			>
 		</File>
-		<File
-			RelativePath="..\..\..\src\nvcore\Array.inl"
-			>
-		</File>
 		<File
 			RelativePath="..\..\..\src\nvcore\Debug.cpp"
 			>
@@ -315,11 +311,11 @@
 			>
 		</File>
 		<File
-			RelativePath="..\..\..\src\nvcore\ForEach.h"
+			RelativePath="..\..\..\src\nvcore\Library.cpp"
 			>
 		</File>
 		<File
-			RelativePath="..\..\..\src\nvcore\Hash.h"
+			RelativePath="..\..\..\src\nvcore\Library.h"
 			>
 		</File>
 		<File

src/CMakeLists.txt

@@ -3,6 +3,10 @@ INCLUDE_DIRECTORIES(${CMAKE_CURRENT_SOURCE_DIR} ${CMAKE_CURRENT_BINARY_DIR})
 INCLUDE_DIRECTORIES(${NV_SOURCE_DIR}/extern/poshlib)
 INCLUDE_DIRECTORIES(${NV_SOURCE_DIR}/extern/stb)
 
+if(CMAKE_COMPILER_IS_GNUCC)
+       ADD_COMPILE_OPTIONS(-std=c++11)
+ENDIF(CMAKE_COMPILER_IS_GNUCC)
+
 SUBDIRS(nvcore)
 SUBDIRS(nvmath)
 SUBDIRS(nvimage)
@@ -67,7 +71,7 @@ ELSE(CUDA_FOUND)
 ENDIF(CUDA_FOUND)
 
 # Maya
-#INCLUDE(${NV_CMAKE_DIR}/FindMaya.cmake)
+INCLUDE(${NV_CMAKE_DIR}/FindMaya.cmake)
 #IF(MAYA_FOUND)
 #	SET(HAVE_MAYA ${MAYA_FOUND} CACHE BOOL "Set to TRUE if Maya is found, FALSE otherwise")
 #	MESSAGE(STATUS "Looking for Maya - found")
@@ -76,7 +80,7 @@ ENDIF(CUDA_FOUND)
 #ENDIF(MAYA_FOUND)
 
 # FreeImage
-#INCLUDE(${NV_CMAKE_DIR}/FindFreeImage.cmake)
+INCLUDE(${NV_CMAKE_DIR}/FindFreeImage.cmake)
 IF(FREEIMAGE_FOUND)
 	SET(HAVE_FREEIMAGE ${FREEIMAGE_FOUND} CACHE BOOL "Set to TRUE if FreeImage is found, FALSE otherwise")
 	MESSAGE(STATUS "Looking for FreeImage - found")
@@ -85,7 +89,7 @@ ELSE(FREEIMAGE_FOUND)
 ENDIF(FREEIMAGE_FOUND)
 
 # JPEG
-#INCLUDE(FindJPEG)
+INCLUDE(FindJPEG)
 IF(JPEG_FOUND)
 	SET(HAVE_JPEG ${JPEG_FOUND} CACHE BOOL "Set to TRUE if JPEG is found, FALSE otherwise")
 	MESSAGE(STATUS "Looking for JPEG - found")
@@ -94,7 +98,7 @@ ELSE(JPEG_FOUND)
 ENDIF(JPEG_FOUND)
 
 # PNG
-#INCLUDE(FindPNG)
+INCLUDE(FindPNG)
 IF(PNG_FOUND)
 	SET(HAVE_PNG ${PNG_FOUND} CACHE BOOL "Set to TRUE if PNG is found, FALSE otherwise")
 	MESSAGE(STATUS "Looking for PNG - found")
@@ -113,7 +117,7 @@ ELSE(TIFF_FOUND)
 ENDIF(TIFF_FOUND)
 
 # OpenEXR
-#INCLUDE(${NV_CMAKE_DIR}/FindOpenEXR.cmake)
+INCLUDE(${NV_CMAKE_DIR}/FindOpenEXR.cmake)
 IF(OPENEXR_FOUND)
 	SET(HAVE_OPENEXR ${OPENEXR_FOUND} CACHE BOOL "Set to TRUE if OpenEXR is found, FALSE otherwise")
 	MESSAGE(STATUS "Looking for OpenEXR - found")
@@ -132,9 +136,6 @@ ELSE(OPENMP_FOUND)
 	MESSAGE(STATUS "Looking for OpenMP - not found")
 ENDIF(OPENMP_FOUND)
 	
-# Qt
-FIND_PACKAGE(Qt4)
-
 # Threads
 FIND_PACKAGE(Threads REQUIRED)
 MESSAGE(STATUS "Use thread library: ${CMAKE_THREAD_LIBS_INIT}")

src/nvcore/CMakeLists.txt

@@ -10,6 +10,7 @@ SET(CORE_SRCS
     DefsVcWin32.h
     FileSystem.h FileSystem.cpp
     ForEach.h
+    Library.h Library.cpp
     Memory.h Memory.cpp
     Ptr.h
     RefCounted.h

src/nvcore/Debug.cpp

@@ -586,7 +586,7 @@ namespace
         writeStackTrace(trace, size, 1, lines);
 
         for (uint i = 0; i < lines.count(); i++) {
-            nvDebug(lines[i]);
+            nvDebug("%s", lines[i]);
             delete lines[i];
         }
 
@@ -992,7 +992,7 @@ void debug::dumpInfo()
         writeStackTrace(trace, size, 1, lines);
 
         for (uint i = 0; i < lines.count(); i++) {
-            nvDebug(lines[i]);
+            nvDebug("%s", lines[i]);
             delete lines[i];
         }
     }

src/nvcore/Library.cpp

@@ -0,0 +1,45 @@
+
+#include "Library.h"
+#include "Debug.h"
+
+#if NV_OS_WIN32
+#define WIN32_LEAN_AND_MEAN
+#define VC_EXTRALEAN
+#include <windows.h>
+#elif NV_OS_XBOX
+#include <Xtl.h>
+#else
+#include <dlfcn.h>
+#endif
+
+
+
+void * nvLoadLibrary(const char * name)
+{
+#if NV_OS_WIN32
+	return (void *)LoadLibraryExA( name, NULL, 0 );
+#elif NV_OS_XBOX
+    return (void *)LoadLibraryA( name );
+#else
+	return dlopen(name, RTLD_LAZY);
+#endif
+}
+
+void nvUnloadLibrary(void * handle)
+{
+	nvDebugCheck(handle != NULL);
+#if NV_OS_WIN32 || NV_OS_XBOX
+	FreeLibrary((HMODULE)handle);
+#else
+	dlclose(handle);
+#endif
+}
+
+void * nvBindSymbol(void * handle, const char * symbol)
+{
+#if NV_OS_WIN32 || NV_OS_XBOX
+	return (void *)GetProcAddress((HMODULE)handle, symbol);
+#else
+	return (void *)dlsym(handle, symbol);
+#endif	
+}

src/nvcore/Library.h

@@ -0,0 +1,51 @@
+// This code is in the public domain -- castano@gmail.com
+
+#pragma once
+#ifndef NV_CORE_LIBRARY_H
+#define NV_CORE_LIBRARY_H
+
+#include "nvcore.h"
+
+#if NV_OS_WIN32
+#define LIBRARY_NAME(name)	#name ".dll"
+#elif NV_OS_DARWIN
+#define NV_LIBRARY_NAME(name)	"lib" #name ".dylib"
+#else
+#define NV_LIBRARY_NAME(name)	"lib" #name ".so"
+#endif
+
+NVCORE_API void * nvLoadLibrary(const char * name);
+NVCORE_API void nvUnloadLibrary(void * lib);
+NVCORE_API void * nvBindSymbol(void * lib, const char * symbol);
+
+class NVCORE_CLASS Library
+{
+public:
+	Library(const char * name)
+	{
+		handle = nvLoadLibrary(name);
+	}
+	~Library()
+	{
+		if (isValid())
+		{
+			nvUnloadLibrary(handle);
+		}
+	}
+	
+	bool isValid() const
+	{
+		return handle != NULL;
+	}
+	
+	void * bindSymbol(const char * symbol)
+	{
+		return nvBindSymbol(handle, symbol);
+	}
+	
+private:
+	void * handle;
+};
+
+
+#endif // NV_CORE_LIBRARY_H

src/nvimage/CMakeLists.txt

@@ -53,7 +53,7 @@ ELSE(NVIMAGE_SHARED)
     ADD_LIBRARY(nvimage ${IMAGE_SRCS})
 ENDIF(NVIMAGE_SHARED)
 
-TARGET_LINK_LIBRARIES(nvimage ${LIBS} nvcore nvmath posh bc6h bc7)
+TARGET_LINK_LIBRARIES(nvimage ${LIBS} nvcore posh bc6h bc7 nvmath)
 
 INSTALL(TARGETS nvimage
     RUNTIME DESTINATION bin

src/nvthread/ThreadPool.cpp

@@ -120,7 +120,10 @@ ThreadPool::ThreadPool(uint workerCount/*=processorCount()*/, bool useThreadAffi
         StringBuilder name;
         name.format("worker %d", i);
         workers[i].setName(name.release());     // @Leak
-        workers[i].start(workerFunc, (void *)i);
+        // This code was just wrong.  Is this what was intended?
+        uint* p = new uint; // @Leak
+        *p = i;
+        workers[i].start(workerFunc, (void *)p);
     }
 
     allIdle = true;

src/nvtt/CMakeLists.txt

@@ -47,7 +47,7 @@ ELSE(NVTT_SHARED)
     ADD_LIBRARY(nvtt ${NVTT_SRCS})
 ENDIF(NVTT_SHARED)
 
-TARGET_LINK_LIBRARIES(nvtt ${LIBS} nvcore nvmath nvimage nvthread squish bc6h bc7)
+TARGET_LINK_LIBRARIES(nvtt ${LIBS} nvcore nvimage nvthread squish bc6h bc7 nvmath)
 
 INSTALL(TARGETS nvtt 
     RUNTIME DESTINATION bin

src/nvtt/cuda/CompressKernel.cu

@@ -1464,13 +1464,13 @@ __global__ void compressDXT1(uint firstBlock, uint blockWidth, const uint * perm
     const int minIdx = findMinError(errors);
 
     __shared__ uint s_bestEndPoints;
-    __shared__ uint s_bestPermutation;
+    //__shared__ uint s_bestPermutation;
 
     // Only write the result of the winner thread.
     if (threadIdx.x == minIdx)
     {
         s_bestEndPoints = (bestEnd << 16) | bestStart;
-        s_bestPermutation = (bestStart != bestEnd) ? bestPermutation : 0;
+        //s_bestPermutation = (bestStart != bestEnd) ? bestPermutation : 0;
     }
 
     __syncthreads();

src/nvtt/cuda/CudaUtils.cpp

@@ -23,6 +23,7 @@
 // OTHER DEALINGS IN THE SOFTWARE.
 
 #include "nvcore/Debug.h"
+#include "nvcore/Library.h"
 #include "CudaUtils.h"
 
 #if defined HAVE_CUDA

src/nvtt/squish/fastclusterfit.cpp

@@ -143,11 +143,11 @@ void FastClusterFit::Compress3( void* block )
 	int i = 0;
 
 	// check all possible clusters for this total order
-	for( int c0 = 0; c0 <= 16; c0++)
+	for( int c0 = 0; c0 < 16; c0++)
 	{	
 		x1 = zero;
 		
-		for( int c1 = 0; c1 <= 16-c0; c1++)
+		for( int c1 = 0; c1 < 16-c0; c1++)
 		{
 			Vec4 const constants = Vec4((const float *)&s_threeElement[i]);
 			Vec4 const alpha2_sum = constants.SplatX();
@@ -256,15 +256,15 @@ void FastClusterFit::Compress4( void* block )
 	int i = 0;
 
 	// check all possible clusters for this total order
-	for( int c0 = 0; c0 <= 16; c0++)
+	for( int c0 = 0; c0 < 16; c0++)
 	{	
 		Vec4 x1 = zero;
 		
-		for( int c1 = 0; c1 <= 16-c0; c1++)
+		for( int c1 = 0; c1 < 16-c0; c1++)
 		{	
 			Vec4 x2 = zero;
 			
-			for( int c2 = 0; c2 <= 16-c0-c1; c2++)
+			for( int c2 = 0; c2 < 16-c0-c1; c2++)
 			{
 				Vec4 const constants = Vec4((const float *)&s_fourElement[i]);
 				Vec4 const alpha2_sum = constants.SplatX();

src/nvtt/tools/CMakeLists.txt

@@ -22,7 +22,7 @@ SET(TOOLS nvcompress nvdecompress nvddsinfo nvassemble nvzoom)
 IF(GLEW_FOUND AND GLUT_FOUND AND OPENGL_FOUND)
     INCLUDE_DIRECTORIES(${GLEW_INCLUDE_PATH} ${GLUT_INCLUDE_DIR} ${OPENGL_INCLUDE_DIR})
     ADD_EXECUTABLE(nvddsview ddsview.cpp cmdline.h)
-    TARGET_LINK_LIBRARIES(nvddsview nvcore nvmath nvimage ${GLEW_LIBRARY} ${GLUT_LIBRARY} ${OPENGL_LIBRARY})
+    TARGET_LINK_LIBRARIES(nvddsview nvcore nvimage ${GLEW_LIBRARY} ${GLUT_LIBRARY} ${OPENGL_LIBRARY})
     SET(TOOLS ${TOOLS} nvddsview)
 ENDIF(GLEW_FOUND AND GLUT_FOUND AND OPENGL_FOUND)
 
@@ -47,7 +47,8 @@ IF(GCONFTOOL2)
  	INSTALL(CODE "EXECUTE_PROCESS(COMMAND ${GCONFTOOL2} --makefile-install-rule ${CMAKE_CURRENT_BINARY_DIR}/nvtt-thumbnailer.schema)")
 ENDIF(GCONFTOOL2)
 
-# UI tools
+# UI tools (not functional)
+#FIND_PACKAGE(Qt4)
 IF(QT4_FOUND)
 	SET(QT_USE_QTOPENGL TRUE)
 	INCLUDE_DIRECTORIES(${QT_INCLUDE_DIR} ${CMAKE_CURRENT_BINARY_DIR})
@@ -64,10 +65,10 @@ IF(QT4_FOUND)
 
         QT4_WRAP_UI(UICS compressdialog.ui)
         QT4_WRAP_CPP(MOCS compressdialog.h)
-	#QT4_ADD_RESOURCES(RCCS ui/configdialog.rc)
+	QT4_ADD_RESOURCES(RCCS ui/configdialog.rc)
 
-	#ADD_EXECUTABLE(nvtt-diag MACOSX_BUNDLE ${SRCS} ${UICS} ${MOCS})
+	ADD_EXECUTABLE(nvtt-diag MACOSX_BUNDLE ${SRCS} ${UICS} ${MOCS})
-	#TARGET_LINK_LIBRARIES(nvtt-diag ${LIBS})
+	TARGET_LINK_LIBRARIES(nvtt-diag ${LIBS})
 
 ENDIF(QT4_FOUND)