diff --git a/Samples/2_Concepts_and_Techniques/CMakeLists.txt b/Samples/2_Concepts_and_Techniques/CMakeLists.txt index 931ae223..6ac2fc1c 100644 --- a/Samples/2_Concepts_and_Techniques/CMakeLists.txt +++ b/Samples/2_Concepts_and_Techniques/CMakeLists.txt @@ -16,8 +16,8 @@ add_subdirectory(histogram) add_subdirectory(imageDenoising) add_subdirectory(inlinePTX) add_subdirectory(inlinePTX_nvrtc) -#add_subdirectory(interval) -#add_subdirectory(particles) +add_subdirectory(interval) +add_subdirectory(particles) #add_subdirectory(radixSortThrust) #add_subdirectory(reduction) #add_subdirectory(reductionMultiBlockCG) diff --git a/Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/CMakeLists.txt b/Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/CMakeLists.txt new file mode 100644 index 00000000..1fffbaa7 --- /dev/null +++ b/Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/CMakeLists.txt @@ -0,0 +1,11 @@ +# Include directories and libraries +include_directories(../../../Common) + +# Source file +set(SRC_FILES + MC_SingleAsianOptionP.cu +) + +# Add target for MC_SingleAsianOptionP +add_executable(MC_SingleAsianOptionP ${SRC_FILES}) +set_target_properties(MC_SingleAsianOptionP PROPERTIES CUDA_SEPARABLE_COMPILATION ON) diff --git a/Samples/2_Concepts_and_Techniques/interval/CMakeLists.txt b/Samples/2_Concepts_and_Techniques/interval/CMakeLists.txt new file mode 100644 index 00000000..f825af84 --- /dev/null +++ b/Samples/2_Concepts_and_Techniques/interval/CMakeLists.txt @@ -0,0 +1,15 @@ +# Include directories and libraries +include_directories(../../../Common) + +# Source file +set(SRC_FILES + interval.cu +) + +# Add target for interval +add_executable(interval ${SRC_FILES}) +set_target_properties(interval PROPERTIES CUDA_SEPARABLE_COMPILATION ON) + +target_include_directories(interval PUBLIC + ${CMAKE_CURRENT_SOURCE_DIR} +) diff --git a/Samples/2_Concepts_and_Techniques/interval/Makefile b/Samples/2_Concepts_and_Techniques/interval/Makefile deleted file mode 100644 index b2e314c7..00000000 --- a/Samples/2_Concepts_and_Techniques/interval/Makefile +++ /dev/null @@ -1,359 +0,0 @@ -################################################################################ -# Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. -# -# Redistribution and use in source and binary forms, with or without -# modification, are permitted provided that the following conditions -# are met: -# * Redistributions of source code must retain the above copyright -# notice, this list of conditions and the following disclaimer. -# * Redistributions in binary form must reproduce the above copyright -# notice, this list of conditions and the following disclaimer in the -# documentation and/or other materials provided with the distribution. -# * Neither the name of NVIDIA CORPORATION nor the names of its -# contributors may be used to endorse or promote products derived -# from this software without specific prior written permission. -# -# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY -# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE -# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR -# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR -# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, -# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, -# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR -# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY -# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT -# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE -# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. -# -################################################################################ -# -# Makefile project only supported on Mac OS X and Linux Platforms) -# -################################################################################ - -# Location of the CUDA Toolkit -CUDA_PATH ?= /usr/local/cuda - -############################## -# start deprecated interface # -############################## -ifeq ($(x86_64),1) - $(info WARNING - x86_64 variable has been deprecated) - $(info WARNING - please use TARGET_ARCH=x86_64 instead) - TARGET_ARCH ?= x86_64 -endif -ifeq ($(ARMv7),1) - $(info WARNING - ARMv7 variable has been deprecated) - $(info WARNING - please use TARGET_ARCH=armv7l instead) - TARGET_ARCH ?= armv7l -endif -ifeq ($(aarch64),1) - $(info WARNING - aarch64 variable has been deprecated) - $(info WARNING - please use TARGET_ARCH=aarch64 instead) - TARGET_ARCH ?= aarch64 -endif -ifeq ($(ppc64le),1) - $(info WARNING - ppc64le variable has been deprecated) - $(info WARNING - please use TARGET_ARCH=ppc64le instead) - TARGET_ARCH ?= ppc64le -endif -ifneq ($(GCC),) - $(info WARNING - GCC variable has been deprecated) - $(info WARNING - please use HOST_COMPILER=$(GCC) instead) - HOST_COMPILER ?= $(GCC) -endif -ifneq ($(abi),) - $(error ERROR - abi variable has been removed) -endif -############################ -# end deprecated interface # -############################ - -# architecture -HOST_ARCH := $(shell uname -m) -TARGET_ARCH ?= $(HOST_ARCH) -ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 sbsa ppc64le armv7l)) - ifneq ($(TARGET_ARCH),$(HOST_ARCH)) - ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 sbsa ppc64le)) - TARGET_SIZE := 64 - else ifneq (,$(filter $(TARGET_ARCH),armv7l)) - TARGET_SIZE := 32 - endif - else - TARGET_SIZE := $(shell getconf LONG_BIT) - endif -else - $(error ERROR - unsupported value $(TARGET_ARCH) for TARGET_ARCH!) -endif - -# sbsa and aarch64 systems look similar. Need to differentiate them at host level for now. -ifeq ($(HOST_ARCH),aarch64) - ifeq ($(CUDA_PATH)/targets/sbsa-linux,$(shell ls -1d $(CUDA_PATH)/targets/sbsa-linux 2>/dev/null)) - HOST_ARCH := sbsa - TARGET_ARCH := sbsa - endif -endif - -ifneq ($(TARGET_ARCH),$(HOST_ARCH)) - ifeq (,$(filter $(HOST_ARCH)-$(TARGET_ARCH),aarch64-armv7l x86_64-armv7l x86_64-aarch64 x86_64-sbsa x86_64-ppc64le)) - $(error ERROR - cross compiling from $(HOST_ARCH) to $(TARGET_ARCH) is not supported!) - endif -endif - -# When on native aarch64 system with userspace of 32-bit, change TARGET_ARCH to armv7l -ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_SIZE),aarch64-aarch64-32) - TARGET_ARCH = armv7l -endif - -# operating system -HOST_OS := $(shell uname -s 2>/dev/null | tr "[:upper:]" "[:lower:]") -TARGET_OS ?= $(HOST_OS) -ifeq (,$(filter $(TARGET_OS),linux darwin qnx android)) - $(error ERROR - unsupported value $(TARGET_OS) for TARGET_OS!) -endif - -# host compiler -ifdef HOST_COMPILER - CUSTOM_HOST_COMPILER = 1 -endif - -ifeq ($(TARGET_OS),darwin) - ifeq ($(shell expr `xcodebuild -version | grep -i xcode | awk '{print $$2}' | cut -d'.' -f1` \>= 5),1) - HOST_COMPILER ?= clang++ - endif -else ifneq ($(TARGET_ARCH),$(HOST_ARCH)) - ifeq ($(HOST_ARCH)-$(TARGET_ARCH),x86_64-armv7l) - ifeq ($(TARGET_OS),linux) - HOST_COMPILER ?= arm-linux-gnueabihf-g++ - else ifeq ($(TARGET_OS),qnx) - ifeq ($(QNX_HOST),) - $(error ERROR - QNX_HOST must be passed to the QNX host toolchain) - endif - ifeq ($(QNX_TARGET),) - $(error ERROR - QNX_TARGET must be passed to the QNX target toolchain) - endif - export QNX_HOST - export QNX_TARGET - HOST_COMPILER ?= $(QNX_HOST)/usr/bin/arm-unknown-nto-qnx6.6.0eabi-g++ - else ifeq ($(TARGET_OS),android) - HOST_COMPILER ?= arm-linux-androideabi-g++ - endif - else ifeq ($(TARGET_ARCH),aarch64) - ifeq ($(TARGET_OS), linux) - HOST_COMPILER ?= aarch64-linux-gnu-g++ - else ifeq ($(TARGET_OS),qnx) - ifeq ($(QNX_HOST),) - $(error ERROR - QNX_HOST must be passed to the QNX host toolchain) - endif - ifeq ($(QNX_TARGET),) - $(error ERROR - QNX_TARGET must be passed to the QNX target toolchain) - endif - export QNX_HOST - export QNX_TARGET - HOST_COMPILER ?= $(QNX_HOST)/usr/bin/q++ - else ifeq ($(TARGET_OS), android) - HOST_COMPILER ?= aarch64-linux-android-clang++ - endif - else ifeq ($(TARGET_ARCH),sbsa) - HOST_COMPILER ?= aarch64-linux-gnu-g++ - else ifeq ($(TARGET_ARCH),ppc64le) - HOST_COMPILER ?= powerpc64le-linux-gnu-g++ - endif -endif -HOST_COMPILER ?= g++ -NVCC := $(CUDA_PATH)/bin/nvcc -ccbin $(HOST_COMPILER) - -# internal flags -NVCCFLAGS := -m${TARGET_SIZE} -CCFLAGS := -LDFLAGS := - -# build flags - -# Link flag for customized HOST_COMPILER with gcc realpath -GCC_PATH := $(shell which gcc) -ifeq ($(CUSTOM_HOST_COMPILER),1) - ifneq ($(filter /%,$(HOST_COMPILER)),) - ifneq ($(findstring gcc,$(HOST_COMPILER)),) - ifneq ($(GCC_PATH),$(HOST_COMPILER)) - LDFLAGS += -lstdc++ - endif - endif - endif -endif - -ifeq ($(TARGET_OS),darwin) - LDFLAGS += -rpath $(CUDA_PATH)/lib - CCFLAGS += -arch $(HOST_ARCH) -else ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_OS),x86_64-armv7l-linux) - LDFLAGS += --dynamic-linker=/lib/ld-linux-armhf.so.3 - CCFLAGS += -mfloat-abi=hard -else ifeq ($(TARGET_OS),android) - LDFLAGS += -pie - CCFLAGS += -fpie -fpic -fexceptions -endif - -ifneq ($(TARGET_ARCH),$(HOST_ARCH)) - ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-linux) - ifneq ($(TARGET_FS),) - GCCVERSIONLTEQ46 := $(shell expr `$(HOST_COMPILER) -dumpversion` \<= 4.6) - ifeq ($(GCCVERSIONLTEQ46),1) - CCFLAGS += --sysroot=$(TARGET_FS) - endif - LDFLAGS += --sysroot=$(TARGET_FS) - LDFLAGS += -rpath-link=$(TARGET_FS)/lib - LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib - LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib/arm-linux-gnueabihf - endif - endif - ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-linux) - ifneq ($(TARGET_FS),) - GCCVERSIONLTEQ46 := $(shell expr `$(HOST_COMPILER) -dumpversion` \<= 4.6) - ifeq ($(GCCVERSIONLTEQ46),1) - CCFLAGS += --sysroot=$(TARGET_FS) - endif - LDFLAGS += --sysroot=$(TARGET_FS) - LDFLAGS += -rpath-link=$(TARGET_FS)/lib -L$(TARGET_FS)/lib - LDFLAGS += -rpath-link=$(TARGET_FS)/lib/aarch64-linux-gnu -L$(TARGET_FS)/lib/aarch64-linux-gnu - LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib -L$(TARGET_FS)/usr/lib - LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib/aarch64-linux-gnu -L$(TARGET_FS)/usr/lib/aarch64-linux-gnu - LDFLAGS += --unresolved-symbols=ignore-in-shared-libs - CCFLAGS += -isystem=$(TARGET_FS)/usr/include -I$(TARGET_FS)/usr/include -I$(TARGET_FS)/usr/include/libdrm - CCFLAGS += -isystem=$(TARGET_FS)/usr/include/aarch64-linux-gnu -I$(TARGET_FS)/usr/include/aarch64-linux-gnu - endif - endif - ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-qnx) - NVCCFLAGS += -D_QNX_SOURCE - NVCCFLAGS += --qpp-config 8.3.0,gcc_ntoaarch64le - CCFLAGS += -DWIN_INTERFACE_CUSTOM -I/usr/include/aarch64-qnx-gnu - LDFLAGS += -lsocket - LDFLAGS += -L/usr/lib/aarch64-qnx-gnu - CCFLAGS += "-Wl\,-rpath-link\,/usr/lib/aarch64-qnx-gnu" - ifdef TARGET_OVERRIDE - LDFLAGS += -lslog2 - endif - - ifneq ($(TARGET_FS),) - LDFLAGS += -L$(TARGET_FS)/usr/lib - CCFLAGS += "-Wl\,-rpath-link\,$(TARGET_FS)/usr/lib" - LDFLAGS += -L$(TARGET_FS)/usr/libnvidia - CCFLAGS += "-Wl\,-rpath-link\,$(TARGET_FS)/usr/libnvidia" - CCFLAGS += -I$(TARGET_FS)/../include - endif - endif -endif - -ifdef TARGET_OVERRIDE # cuda toolkit targets override - NVCCFLAGS += -target-dir $(TARGET_OVERRIDE) -endif - -# Install directory of different arch -CUDA_INSTALL_TARGET_DIR := -ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-linux) - CUDA_INSTALL_TARGET_DIR = targets/armv7-linux-gnueabihf/ -else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-linux) - CUDA_INSTALL_TARGET_DIR = targets/aarch64-linux/ -else ifeq ($(TARGET_ARCH)-$(TARGET_OS),sbsa-linux) - CUDA_INSTALL_TARGET_DIR = targets/sbsa-linux/ -else ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-android) - CUDA_INSTALL_TARGET_DIR = targets/armv7-linux-androideabi/ -else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-android) - CUDA_INSTALL_TARGET_DIR = targets/aarch64-linux-androideabi/ -else ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-qnx) - CUDA_INSTALL_TARGET_DIR = targets/ARMv7-linux-QNX/ -else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-qnx) - CUDA_INSTALL_TARGET_DIR = targets/aarch64-qnx/ -else ifeq ($(TARGET_ARCH),ppc64le) - CUDA_INSTALL_TARGET_DIR = targets/ppc64le-linux/ -endif - -# Debug build flags -ifeq ($(dbg),1) - NVCCFLAGS += -g -G - BUILD_TYPE := debug -else - BUILD_TYPE := release -endif - -ALL_CCFLAGS := -ALL_CCFLAGS += $(NVCCFLAGS) -ALL_CCFLAGS += $(EXTRA_NVCCFLAGS) -ALL_CCFLAGS += $(addprefix -Xcompiler ,$(CCFLAGS)) -ALL_CCFLAGS += $(addprefix -Xcompiler ,$(EXTRA_CCFLAGS)) - -SAMPLE_ENABLED := 1 - -ALL_LDFLAGS := -ALL_LDFLAGS += $(ALL_CCFLAGS) -ALL_LDFLAGS += $(addprefix -Xlinker ,$(LDFLAGS)) -ALL_LDFLAGS += $(addprefix -Xlinker ,$(EXTRA_LDFLAGS)) - -# Common includes and paths for CUDA -INCLUDES := -I../../../Common -LIBRARIES := - -################################################################################ - -# Gencode arguments -ifeq ($(TARGET_ARCH),$(filter $(TARGET_ARCH),armv7l aarch64 sbsa)) -SMS ?= 53 61 70 72 75 80 86 87 90 -else -SMS ?= 50 52 60 61 70 75 80 86 89 90 -endif - -ifeq ($(SMS),) -$(info >>> WARNING - no SM architectures have been specified - waiving sample <<<) -SAMPLE_ENABLED := 0 -endif - -ifeq ($(GENCODE_FLAGS),) -# Generate SASS code for each SM architecture listed in $(SMS) -$(foreach sm,$(SMS),$(eval GENCODE_FLAGS += -gencode arch=compute_$(sm),code=sm_$(sm))) - -# Generate PTX code from the highest SM architecture in $(SMS) to guarantee forward-compatibility -HIGHEST_SM := $(lastword $(sort $(SMS))) -ifneq ($(HIGHEST_SM),) -GENCODE_FLAGS += -gencode arch=compute_$(HIGHEST_SM),code=compute_$(HIGHEST_SM) -endif -endif - -ALL_CCFLAGS += --threads 0 --std=c++11 - -INCLUDES += -I. - -ifeq ($(SAMPLE_ENABLED),0) -EXEC ?= @echo "[@]" -endif - -################################################################################ - -# Target rules -all: build - -build: interval - -check.deps: -ifeq ($(SAMPLE_ENABLED),0) - @echo "Sample will be waived due to the above missing dependencies" -else - @echo "Sample is ready - all dependencies have been met" -endif - -interval.o:interval.cu - $(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $< - -interval: interval.o - $(EXEC) $(NVCC) $(ALL_LDFLAGS) $(GENCODE_FLAGS) -o $@ $+ $(LIBRARIES) - $(EXEC) mkdir -p ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE) - $(EXEC) cp $@ ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE) - -run: build - $(EXEC) ./interval - -testrun: build - -clean: - rm -f interval interval.o - rm -rf ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)/interval - -clobber: clean diff --git a/Samples/2_Concepts_and_Techniques/interval/NsightEclipse.xml b/Samples/2_Concepts_and_Techniques/interval/NsightEclipse.xml deleted file mode 100644 index 1b42c009..00000000 --- a/Samples/2_Concepts_and_Techniques/interval/NsightEclipse.xml +++ /dev/null @@ -1,86 +0,0 @@ - - - - interval - - cudaMemcpy - cudaFuncSetCacheConfig - cudaMalloc - cudaFree - cudaGetLastError - cudaSetDevice - cudaDeviceSynchronize - cudaEventRecord - cudaDeviceSetLimit - cudaEventDestroy - cudaEventElapsedTime - cudaGetDeviceProperties - cudaEventCreate - - - whole - - . - ./ - ../ - ../../../Common - - - Recursion - Templates - - - newdelete arithmetic - rounding modes - - - - - - true - interval.cu - - 1:CUDA Advanced Topics - - sm50 - sm52 - sm53 - sm60 - sm61 - sm70 - sm72 - sm75 - sm80 - sm86 - sm87 - sm89 - sm90 - - - x86_64 - linux - - - windows7 - - - x86_64 - macosx - - - arm - - - sbsa - - - ppc64le - linux - - - - all - - Interval Computing - exe - diff --git a/Samples/2_Concepts_and_Techniques/particles/CMakeLists.txt b/Samples/2_Concepts_and_Techniques/particles/CMakeLists.txt new file mode 100644 index 00000000..6f6be38c --- /dev/null +++ b/Samples/2_Concepts_and_Techniques/particles/CMakeLists.txt @@ -0,0 +1,45 @@ +# Include directories and libraries +include_directories(../../../Common) + +find_package(OpenGL) +find_package(GLUT) + +# Source file +set(SRC_FILES + particleSystem.cpp + particleSystem_cuda.cu + particles.cpp + render_particles.cpp + shaders.cpp +) + +if(${OpenGL_FOUND}) + if (${GLUT_FOUND}) + # Add target for particles + add_executable(particles ${SRC_FILES}) + set_target_properties(particles PROPERTIES CUDA_SEPARABLE_COMPILATION ON) + + target_include_directories(particles PUBLIC + ${OPENGL_INCLUDE_DIR} + ${CUDAToolkit_INCLUDE_DIRS} + ${GLUT_INCLUDE_DIRS} + ) + + target_link_libraries(particles + ${OPENGL_LIBRARIES} + ${GLUT_LIBRARIES} + ) + + # Copy clock_kernel.cu to the output directory + add_custom_command(TARGET particles POST_BUILD + COMMAND ${CMAKE_COMMAND} -E copy_directory_if_different + ${CMAKE_CURRENT_SOURCE_DIR}/data + ${CMAKE_CURRENT_BINARY_DIR}/data + ) + + else() + message(STATUS "GLUT not found - will not build sample 'particles'") + endif() +else() + message(STATUS "OpenGL not found - will not build sample 'particles'") +endif() diff --git a/Samples/2_Concepts_and_Techniques/particles/Makefile b/Samples/2_Concepts_and_Techniques/particles/Makefile deleted file mode 100644 index f45b0588..00000000 --- a/Samples/2_Concepts_and_Techniques/particles/Makefile +++ /dev/null @@ -1,418 +0,0 @@ -################################################################################ -# Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. -# -# Redistribution and use in source and binary forms, with or without -# modification, are permitted provided that the following conditions -# are met: -# * Redistributions of source code must retain the above copyright -# notice, this list of conditions and the following disclaimer. -# * Redistributions in binary form must reproduce the above copyright -# notice, this list of conditions and the following disclaimer in the -# documentation and/or other materials provided with the distribution. -# * Neither the name of NVIDIA CORPORATION nor the names of its -# contributors may be used to endorse or promote products derived -# from this software without specific prior written permission. -# -# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY -# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE -# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR -# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR -# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, -# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, -# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR -# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY -# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT -# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE -# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. -# -################################################################################ -# -# Makefile project only supported on Mac OS X and Linux Platforms) -# -################################################################################ - -# Location of the CUDA Toolkit -CUDA_PATH ?= /usr/local/cuda - -############################## -# start deprecated interface # -############################## -ifeq ($(x86_64),1) - $(info WARNING - x86_64 variable has been deprecated) - $(info WARNING - please use TARGET_ARCH=x86_64 instead) - TARGET_ARCH ?= x86_64 -endif -ifeq ($(ARMv7),1) - $(info WARNING - ARMv7 variable has been deprecated) - $(info WARNING - please use TARGET_ARCH=armv7l instead) - TARGET_ARCH ?= armv7l -endif -ifeq ($(aarch64),1) - $(info WARNING - aarch64 variable has been deprecated) - $(info WARNING - please use TARGET_ARCH=aarch64 instead) - TARGET_ARCH ?= aarch64 -endif -ifeq ($(ppc64le),1) - $(info WARNING - ppc64le variable has been deprecated) - $(info WARNING - please use TARGET_ARCH=ppc64le instead) - TARGET_ARCH ?= ppc64le -endif -ifneq ($(GCC),) - $(info WARNING - GCC variable has been deprecated) - $(info WARNING - please use HOST_COMPILER=$(GCC) instead) - HOST_COMPILER ?= $(GCC) -endif -ifneq ($(abi),) - $(error ERROR - abi variable has been removed) -endif -############################ -# end deprecated interface # -############################ - -# architecture -HOST_ARCH := $(shell uname -m) -TARGET_ARCH ?= $(HOST_ARCH) -ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 sbsa ppc64le armv7l)) - ifneq ($(TARGET_ARCH),$(HOST_ARCH)) - ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 sbsa ppc64le)) - TARGET_SIZE := 64 - else ifneq (,$(filter $(TARGET_ARCH),armv7l)) - TARGET_SIZE := 32 - endif - else - TARGET_SIZE := $(shell getconf LONG_BIT) - endif -else - $(error ERROR - unsupported value $(TARGET_ARCH) for TARGET_ARCH!) -endif - -# sbsa and aarch64 systems look similar. Need to differentiate them at host level for now. -ifeq ($(HOST_ARCH),aarch64) - ifeq ($(CUDA_PATH)/targets/sbsa-linux,$(shell ls -1d $(CUDA_PATH)/targets/sbsa-linux 2>/dev/null)) - HOST_ARCH := sbsa - TARGET_ARCH := sbsa - endif -endif - -ifneq ($(TARGET_ARCH),$(HOST_ARCH)) - ifeq (,$(filter $(HOST_ARCH)-$(TARGET_ARCH),aarch64-armv7l x86_64-armv7l x86_64-aarch64 x86_64-sbsa x86_64-ppc64le)) - $(error ERROR - cross compiling from $(HOST_ARCH) to $(TARGET_ARCH) is not supported!) - endif -endif - -# When on native aarch64 system with userspace of 32-bit, change TARGET_ARCH to armv7l -ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_SIZE),aarch64-aarch64-32) - TARGET_ARCH = armv7l -endif - -# operating system -HOST_OS := $(shell uname -s 2>/dev/null | tr "[:upper:]" "[:lower:]") -TARGET_OS ?= $(HOST_OS) -ifeq (,$(filter $(TARGET_OS),linux darwin qnx android)) - $(error ERROR - unsupported value $(TARGET_OS) for TARGET_OS!) -endif - -# host compiler -ifdef HOST_COMPILER - CUSTOM_HOST_COMPILER = 1 -endif - -ifeq ($(TARGET_OS),darwin) - ifeq ($(shell expr `xcodebuild -version | grep -i xcode | awk '{print $$2}' | cut -d'.' -f1` \>= 5),1) - HOST_COMPILER ?= clang++ - endif -else ifneq ($(TARGET_ARCH),$(HOST_ARCH)) - ifeq ($(HOST_ARCH)-$(TARGET_ARCH),x86_64-armv7l) - ifeq ($(TARGET_OS),linux) - HOST_COMPILER ?= arm-linux-gnueabihf-g++ - else ifeq ($(TARGET_OS),qnx) - ifeq ($(QNX_HOST),) - $(error ERROR - QNX_HOST must be passed to the QNX host toolchain) - endif - ifeq ($(QNX_TARGET),) - $(error ERROR - QNX_TARGET must be passed to the QNX target toolchain) - endif - export QNX_HOST - export QNX_TARGET - HOST_COMPILER ?= $(QNX_HOST)/usr/bin/arm-unknown-nto-qnx6.6.0eabi-g++ - else ifeq ($(TARGET_OS),android) - HOST_COMPILER ?= arm-linux-androideabi-g++ - endif - else ifeq ($(TARGET_ARCH),aarch64) - ifeq ($(TARGET_OS), linux) - HOST_COMPILER ?= aarch64-linux-gnu-g++ - else ifeq ($(TARGET_OS),qnx) - ifeq ($(QNX_HOST),) - $(error ERROR - QNX_HOST must be passed to the QNX host toolchain) - endif - ifeq ($(QNX_TARGET),) - $(error ERROR - QNX_TARGET must be passed to the QNX target toolchain) - endif - export QNX_HOST - export QNX_TARGET - HOST_COMPILER ?= $(QNX_HOST)/usr/bin/q++ - else ifeq ($(TARGET_OS), android) - HOST_COMPILER ?= aarch64-linux-android-clang++ - endif - else ifeq ($(TARGET_ARCH),sbsa) - HOST_COMPILER ?= aarch64-linux-gnu-g++ - else ifeq ($(TARGET_ARCH),ppc64le) - HOST_COMPILER ?= powerpc64le-linux-gnu-g++ - endif -endif -HOST_COMPILER ?= g++ -NVCC := $(CUDA_PATH)/bin/nvcc -ccbin $(HOST_COMPILER) - -# internal flags -NVCCFLAGS := -m${TARGET_SIZE} -CCFLAGS := -LDFLAGS := - -# build flags - -# Link flag for customized HOST_COMPILER with gcc realpath -GCC_PATH := $(shell which gcc) -ifeq ($(CUSTOM_HOST_COMPILER),1) - ifneq ($(filter /%,$(HOST_COMPILER)),) - ifneq ($(findstring gcc,$(HOST_COMPILER)),) - ifneq ($(GCC_PATH),$(HOST_COMPILER)) - LDFLAGS += -lstdc++ - endif - endif - endif -endif - -ifeq ($(TARGET_OS),darwin) - LDFLAGS += -rpath $(CUDA_PATH)/lib - CCFLAGS += -arch $(HOST_ARCH) -else ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_OS),x86_64-armv7l-linux) - LDFLAGS += --dynamic-linker=/lib/ld-linux-armhf.so.3 - CCFLAGS += -mfloat-abi=hard -else ifeq ($(TARGET_OS),android) - LDFLAGS += -pie - CCFLAGS += -fpie -fpic -fexceptions -endif - -ifneq ($(TARGET_ARCH),$(HOST_ARCH)) - ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-linux) - ifneq ($(TARGET_FS),) - GCCVERSIONLTEQ46 := $(shell expr `$(HOST_COMPILER) -dumpversion` \<= 4.6) - ifeq ($(GCCVERSIONLTEQ46),1) - CCFLAGS += --sysroot=$(TARGET_FS) - endif - LDFLAGS += --sysroot=$(TARGET_FS) - LDFLAGS += -rpath-link=$(TARGET_FS)/lib - LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib - LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib/arm-linux-gnueabihf - endif - endif - ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-linux) - ifneq ($(TARGET_FS),) - GCCVERSIONLTEQ46 := $(shell expr `$(HOST_COMPILER) -dumpversion` \<= 4.6) - ifeq ($(GCCVERSIONLTEQ46),1) - CCFLAGS += --sysroot=$(TARGET_FS) - endif - LDFLAGS += --sysroot=$(TARGET_FS) - LDFLAGS += -rpath-link=$(TARGET_FS)/lib -L$(TARGET_FS)/lib - LDFLAGS += -rpath-link=$(TARGET_FS)/lib/aarch64-linux-gnu -L$(TARGET_FS)/lib/aarch64-linux-gnu - LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib -L$(TARGET_FS)/usr/lib - LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib/aarch64-linux-gnu -L$(TARGET_FS)/usr/lib/aarch64-linux-gnu - LDFLAGS += --unresolved-symbols=ignore-in-shared-libs - CCFLAGS += -isystem=$(TARGET_FS)/usr/include -I$(TARGET_FS)/usr/include -I$(TARGET_FS)/usr/include/libdrm - CCFLAGS += -isystem=$(TARGET_FS)/usr/include/aarch64-linux-gnu -I$(TARGET_FS)/usr/include/aarch64-linux-gnu - endif - endif - ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-qnx) - NVCCFLAGS += -D_QNX_SOURCE - NVCCFLAGS += --qpp-config 8.3.0,gcc_ntoaarch64le - CCFLAGS += -DWIN_INTERFACE_CUSTOM -I/usr/include/aarch64-qnx-gnu - LDFLAGS += -lsocket - LDFLAGS += -L/usr/lib/aarch64-qnx-gnu - CCFLAGS += "-Wl\,-rpath-link\,/usr/lib/aarch64-qnx-gnu" - ifdef TARGET_OVERRIDE - LDFLAGS += -lslog2 - endif - - ifneq ($(TARGET_FS),) - LDFLAGS += -L$(TARGET_FS)/usr/lib - CCFLAGS += "-Wl\,-rpath-link\,$(TARGET_FS)/usr/lib" - LDFLAGS += -L$(TARGET_FS)/usr/libnvidia - CCFLAGS += "-Wl\,-rpath-link\,$(TARGET_FS)/usr/libnvidia" - CCFLAGS += -I$(TARGET_FS)/../include - endif - endif -endif - -ifdef TARGET_OVERRIDE # cuda toolkit targets override - NVCCFLAGS += -target-dir $(TARGET_OVERRIDE) -endif - -# Install directory of different arch -CUDA_INSTALL_TARGET_DIR := -ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-linux) - CUDA_INSTALL_TARGET_DIR = targets/armv7-linux-gnueabihf/ -else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-linux) - CUDA_INSTALL_TARGET_DIR = targets/aarch64-linux/ -else ifeq ($(TARGET_ARCH)-$(TARGET_OS),sbsa-linux) - CUDA_INSTALL_TARGET_DIR = targets/sbsa-linux/ -else ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-android) - CUDA_INSTALL_TARGET_DIR = targets/armv7-linux-androideabi/ -else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-android) - CUDA_INSTALL_TARGET_DIR = targets/aarch64-linux-androideabi/ -else ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-qnx) - CUDA_INSTALL_TARGET_DIR = targets/ARMv7-linux-QNX/ -else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-qnx) - CUDA_INSTALL_TARGET_DIR = targets/aarch64-qnx/ -else ifeq ($(TARGET_ARCH),ppc64le) - CUDA_INSTALL_TARGET_DIR = targets/ppc64le-linux/ -endif - -# Debug build flags -ifeq ($(dbg),1) - NVCCFLAGS += -g -G - BUILD_TYPE := debug -else - BUILD_TYPE := release -endif - -ALL_CCFLAGS := -ALL_CCFLAGS += $(NVCCFLAGS) -ALL_CCFLAGS += $(EXTRA_NVCCFLAGS) -ALL_CCFLAGS += $(addprefix -Xcompiler ,$(CCFLAGS)) -ALL_CCFLAGS += $(addprefix -Xcompiler ,$(EXTRA_CCFLAGS)) - -SAMPLE_ENABLED := 1 - -# This sample is not supported on QNX -ifeq ($(TARGET_OS),qnx) - $(info >>> WARNING - particles is not supported on QNX - waiving sample <<<) - SAMPLE_ENABLED := 0 -endif - -ALL_LDFLAGS := -ALL_LDFLAGS += $(ALL_CCFLAGS) -ALL_LDFLAGS += $(addprefix -Xlinker ,$(LDFLAGS)) -ALL_LDFLAGS += $(addprefix -Xlinker ,$(EXTRA_LDFLAGS)) - -# Common includes and paths for CUDA -INCLUDES := -I../../../Common -LIBRARIES := - -################################################################################ - -# Makefile include to help find GL Libraries -include ./findgllib.mk - -# OpenGL specific libraries -ifeq ($(TARGET_OS),darwin) - # Mac OSX specific libraries and paths to include - LIBRARIES += -L/System/Library/Frameworks/OpenGL.framework/Libraries - LIBRARIES += -lGL -lGLU - ALL_LDFLAGS += -Xlinker -framework -Xlinker GLUT -else - LIBRARIES += $(GLLINK) - LIBRARIES += -lGL -lGLU -lglut -endif - -#Detect if installed version of GCC supports required C++14 -ifeq ($(TARGET_OS),linux) - empty := - space := $(empty) $(empty) - GCCVERSIONSTRING := $(shell expr `$(HOST_COMPILER) -dumpversion`) -#Create version number without "." - GCCVERSION := $(shell expr `echo $(GCCVERSIONSTRING)` | cut -f1 -d.) - GCCVERSION += $(shell expr `echo $(GCCVERSIONSTRING)` | cut -f2 -d.) - GCCVERSION += $(shell expr `echo $(GCCVERSIONSTRING)` | cut -f3 -d.) -# Make sure the version number has at least 3 decimals - GCCVERSION += 00 -# Remove spaces from the version number - GCCVERSION := $(subst $(space),$(empty),$(GCCVERSION)) -#$(warning $(GCCVERSION)) - - IS_MIN_VERSION := $(shell expr `echo $(GCCVERSION)` \>= 50000) - ifneq ($(CUSTOM_HOST_COMPILER), 1) - ifeq ($(IS_MIN_VERSION), 1) - $(info >>> GCC Version is greater or equal to 5.0.0 <<<) - else - $(info >>> Waiving build. Minimum GCC version required is 5.0.0<<<) - SAMPLE_ENABLED := 0 - endif - else - $(warning >>> Custom HOST_COMPILER set; skipping GCC version check. This may lead to unintended behavior. Please note the minimum equivalent GCC version is 5.0.0 <<<) - endif -endif - -# Gencode arguments -ifeq ($(TARGET_ARCH),$(filter $(TARGET_ARCH),armv7l aarch64 sbsa)) -SMS ?= 53 61 70 72 75 80 86 87 90 -else -SMS ?= 50 52 60 61 70 75 80 86 89 90 -endif - -ifeq ($(SMS),) -$(info >>> WARNING - no SM architectures have been specified - waiving sample <<<) -SAMPLE_ENABLED := 0 -endif - -ifeq ($(GENCODE_FLAGS),) -# Generate SASS code for each SM architecture listed in $(SMS) -$(foreach sm,$(SMS),$(eval GENCODE_FLAGS += -gencode arch=compute_$(sm),code=sm_$(sm))) - -# Generate PTX code from the highest SM architecture in $(SMS) to guarantee forward-compatibility -HIGHEST_SM := $(lastword $(sort $(SMS))) -ifneq ($(HIGHEST_SM),) -GENCODE_FLAGS += -gencode arch=compute_$(HIGHEST_SM),code=compute_$(HIGHEST_SM) -endif -endif - -ALL_CCFLAGS += --std=c++14 --threads 0 - -ifeq ($(SAMPLE_ENABLED),0) -EXEC ?= @echo "[@]" -endif - -################################################################################ - -# Target rules -all: build - -build: particles - -check.deps: -ifeq ($(SAMPLE_ENABLED),0) - @echo "Sample will be waived due to the above missing dependencies" -else - @echo "Sample is ready - all dependencies have been met" -endif - -particleSystem.o:particleSystem.cpp - $(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $< - -particleSystem_cuda.o:particleSystem_cuda.cu - $(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $< - -particles.o:particles.cpp - $(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $< - -render_particles.o:render_particles.cpp - $(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $< - -shaders.o:shaders.cpp - $(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $< - -particles: particleSystem.o particleSystem_cuda.o particles.o render_particles.o shaders.o - $(EXEC) $(NVCC) $(ALL_LDFLAGS) $(GENCODE_FLAGS) -o $@ $+ $(LIBRARIES) - $(EXEC) mkdir -p ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE) - $(EXEC) cp $@ ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE) - -run: build - $(EXEC) ./particles - -testrun: build - $(EXEC) ./particles -file=ref_particles.bin - -clean: - rm -f particles particleSystem.o particleSystem_cuda.o particles.o render_particles.o shaders.o - rm -rf ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)/particles - -clobber: clean diff --git a/Samples/2_Concepts_and_Techniques/particles/NsightEclipse.xml b/Samples/2_Concepts_and_Techniques/particles/NsightEclipse.xml deleted file mode 100644 index 07c2d597..00000000 --- a/Samples/2_Concepts_and_Techniques/particles/NsightEclipse.xml +++ /dev/null @@ -1,116 +0,0 @@ - - - - particles - - --std=c++14 - - - cudaGraphicsUnmapResources - cudaMemcpy - cudaFree - cudaGraphicsResourceGetMappedPointer - cudaGraphicsMapResources - cudaDeviceSynchronize - cudaMemset - cudaMemcpyToSymbol - cudaGraphicsGLRegisterBuffer - cudaGraphicsUnregisterResource - cudaMalloc - cudaInit - cudaGLInit - - " to the command line will allow users to set # of particles for simulation. This example implements a uniform grid data structure using either atomic operations or a fast radix sort from the Thrust library]]> - whole - - data\ref_particles.ppm - data\ref_particles.bin - - - ./ - ../ - ../../../Common - - - Graphics Interop - Data Parallel Algorithms - Physically-Based Simulation - Performance Strategies - - - CUDA - GPGPU - OpenGL - openGL - CPP14 - - - GLU - GL - GLUT - GLEW - glut - X11 - - - ../../../common/lib/linux/x86_64 - ../../../common/lib/linux/armv7l - ../../../common/lib/darwin - - true - particles.cpp - - -file=ref_particles.bin - - - X11 - GL - - - 2:Graphics Interop - 3:Physically-Based Simulation - - sm50 - sm52 - sm53 - sm60 - sm61 - sm70 - sm72 - sm75 - sm80 - sm86 - sm87 - sm89 - sm90 - - ..\..\..\Common\param.h - ..\..\..\Common\paramgl.h - ..\..\..\Common\rendercheck_gl.h - - - - x86_64 - linux - - - windows7 - - - x86_64 - macosx - - - arm - - - sbsa - - - - all - - Particles - exe - doc\particles.pdf - diff --git a/Samples/2_Concepts_and_Techniques/particles/findgllib.mk b/Samples/2_Concepts_and_Techniques/particles/findgllib.mk deleted file mode 100644 index 998fcf0f..00000000 --- a/Samples/2_Concepts_and_Techniques/particles/findgllib.mk +++ /dev/null @@ -1,136 +0,0 @@ -################################################################################ -# Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. -# -# Redistribution and use in source and binary forms, with or without -# modification, are permitted provided that the following conditions -# are met: -# * Redistributions of source code must retain the above copyright -# notice, this list of conditions and the following disclaimer. -# * Redistributions in binary form must reproduce the above copyright -# notice, this list of conditions and the following disclaimer in the -# documentation and/or other materials provided with the distribution. -# * Neither the name of NVIDIA CORPORATION nor the names of its -# contributors may be used to endorse or promote products derived -# from this software without specific prior written permission. -# -# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY -# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE -# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR -# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR -# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, -# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, -# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR -# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY -# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT -# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE -# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. -# -################################################################################ -# -# findgllib.mk is used to find the necessary GL Libraries for specific distributions -# this is supported on Mac OSX and Linux Platforms -# -################################################################################ - -# Determine OS platform and unix distribution -ifeq ("$(TARGET_OS)","linux") - # first search lsb_release - DISTRO := $(shell lsb_release -i -s 2>/dev/null | tr "[:upper:]" "[:lower:]") - ifeq ("$(DISTRO)","") - # second search and parse /etc/issue - DISTRO := $(shell awk '{print $$1}' /etc/issue | tr -d "[:space:]" | sed -e "/^$$/d" | tr "[:upper:]" "[:lower:]") - # ensure data from /etc/issue is valid - ifeq (,$(filter $(DISTRO),ubuntu fedora red rhel centos suse)) - DISTRO := - endif - ifeq ("$(DISTRO)","") - # third, we can search in /etc/os-release or /etc/{distro}-release - DISTRO := $(shell awk '/ID/' /etc/*-release | sed 's/ID=//' | grep -v "VERSION" | grep -v "ID" | grep -v "DISTRIB") - endif - endif -endif - -ifeq ("$(TARGET_OS)","linux") - # $(info) >> findgllib.mk -> LINUX path <<<) - # Each set of Linux Distros have different paths for where to find their OpenGL libraries reside - UBUNTU = $(shell echo $(DISTRO) | grep -i ubuntu >/dev/null 2>&1; echo $$?) - FEDORA = $(shell echo $(DISTRO) | grep -i fedora >/dev/null 2>&1; echo $$?) - RHEL = $(shell echo $(DISTRO) | grep -i 'red\|rhel' >/dev/null 2>&1; echo $$?) - CENTOS = $(shell echo $(DISTRO) | grep -i centos >/dev/null 2>&1; echo $$?) - SUSE = $(shell echo $(DISTRO) | grep -i 'suse\|sles' >/dev/null 2>&1; echo $$?) - KYLIN = $(shell echo $(DISTRO) | grep -i kylin >/dev/null 2>&1; echo $$?) - ifeq ("$(UBUNTU)","0") - ifeq ($(HOST_ARCH)-$(TARGET_ARCH),x86_64-armv7l) - GLPATH := /usr/arm-linux-gnueabihf/lib - GLLINK := -L/usr/arm-linux-gnueabihf/lib - ifneq ($(TARGET_FS),) - GLPATH += $(TARGET_FS)/usr/lib/arm-linux-gnueabihf - GLLINK += -L$(TARGET_FS)/usr/lib/arm-linux-gnueabihf - endif - else ifeq ($(HOST_ARCH)-$(TARGET_ARCH),x86_64-aarch64) - GLPATH := /usr/aarch64-linux-gnu/lib - GLLINK := -L/usr/aarch64-linux-gnu/lib - ifneq ($(TARGET_FS),) - GLPATH += $(TARGET_FS)/usr/lib - GLPATH += $(TARGET_FS)/usr/lib/aarch64-linux-gnu - GLLINK += -L$(TARGET_FS)/usr/lib/aarch64-linux-gnu - endif - else ifeq ($(HOST_ARCH)-$(TARGET_ARCH),x86_64-ppc64le) - GLPATH := /usr/powerpc64le-linux-gnu/lib - GLLINK := -L/usr/powerpc64le-linux-gnu/lib - else - UBUNTU_PKG_NAME = $(shell which dpkg >/dev/null 2>&1 && dpkg -l 'nvidia-*' | grep '^ii' | awk '{print $$2}' | head -1) - ifneq ("$(UBUNTU_PKG_NAME)","") - GLPATH ?= /usr/lib/$(UBUNTU_PKG_NAME) - GLLINK ?= -L/usr/lib/$(UBUNTU_PKG_NAME) - endif - - DFLT_PATH ?= /usr/lib - endif - endif - - ifeq ("$(SUSE)","0") - GLPATH ?= /usr/X11R6/lib64 - GLLINK ?= -L/usr/X11R6/lib64 - DFLT_PATH ?= /usr/lib64 - else - GLPATH ?= /usr/lib64/nvidia - GLLINK ?= -L/usr/lib64/nvidia - DFLT_PATH ?= /usr/lib64 - endif - - # find libGL, libGLU - GLLIB := $(shell find -L $(GLPATH) $(DFLT_PATH) -name libGL.so -print 2>/dev/null) - GLULIB := $(shell find -L $(GLPATH) $(DFLT_PATH) -name libGLU.so -print 2>/dev/null) - - ifeq ("$(GLLIB)","") - $(info >>> WARNING - libGL.so not found, refer to CUDA Getting Started Guide for how to find and install them. <<<) - SAMPLE_ENABLED := 0 - endif - ifeq ("$(GLULIB)","") - $(info >>> WARNING - libGLU.so not found, refer to CUDA Getting Started Guide for how to find and install them. <<<) - SAMPLE_ENABLED := 0 - endif - - HEADER_SEARCH_PATH ?= $(TARGET_FS)/usr/include - ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_OS),x86_64-armv7l-linux) - HEADER_SEARCH_PATH += /usr/arm-linux-gnueabihf/include - else ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_OS),x86_64-aarch64-linux) - HEADER_SEARCH_PATH += /usr/aarch64-linux-gnu/include - endif - - GLHEADER := $(shell find -L $(HEADER_SEARCH_PATH) -name gl.h -print 2>/dev/null) - GLUHEADER := $(shell find -L $(HEADER_SEARCH_PATH) -name glu.h -print 2>/dev/null) - - ifeq ("$(GLHEADER)","") - $(info >>> WARNING - gl.h not found, refer to CUDA Getting Started Guide for how to find and install them. <<<) - SAMPLE_ENABLED := 0 - endif - ifeq ("$(GLUHEADER)","") - $(info >>> WARNING - glu.h not found, refer to CUDA Getting Started Guide for how to find and install them. <<<) - SAMPLE_ENABLED := 0 - endif -else - # This would be the Mac OS X path if we had to do anything special -endif - diff --git a/Samples/2_Concepts_and_Techniques/particles/particleSystem_cuda.cu b/Samples/2_Concepts_and_Techniques/particles/particleSystem_cuda.cu index 8d043abc..f1c97348 100644 --- a/Samples/2_Concepts_and_Techniques/particles/particleSystem_cuda.cu +++ b/Samples/2_Concepts_and_Techniques/particles/particleSystem_cuda.cu @@ -52,153 +52,172 @@ #include "particles_kernel_impl.cuh" -extern "C" { +extern "C" +{ -void cudaInit(int argc, char **argv) { - int devID; + void cudaInit(int argc, char **argv) + { + int devID; - // use command-line specified CUDA device, otherwise use device with highest - // Gflops/s - devID = findCudaDevice(argc, (const char **)argv); + // use command-line specified CUDA device, otherwise use device with highest + // Gflops/s + devID = findCudaDevice(argc, (const char **)argv); - if (devID < 0) { - printf("No CUDA Capable devices found, exiting...\n"); - exit(EXIT_SUCCESS); - } -} - -void allocateArray(void **devPtr, size_t size) { - checkCudaErrors(cudaMalloc(devPtr, size)); -} - -void freeArray(void *devPtr) { checkCudaErrors(cudaFree(devPtr)); } - -void threadSync() { checkCudaErrors(cudaDeviceSynchronize()); } - -void copyArrayToDevice(void *device, const void *host, int offset, int size) { - checkCudaErrors( - cudaMemcpy((char *)device + offset, host, size, cudaMemcpyHostToDevice)); -} - -void registerGLBufferObject(uint vbo, - struct cudaGraphicsResource **cuda_vbo_resource) { - checkCudaErrors(cudaGraphicsGLRegisterBuffer(cuda_vbo_resource, vbo, - cudaGraphicsMapFlagsNone)); -} - -void unregisterGLBufferObject(struct cudaGraphicsResource *cuda_vbo_resource) { - checkCudaErrors(cudaGraphicsUnregisterResource(cuda_vbo_resource)); -} - -void *mapGLBufferObject(struct cudaGraphicsResource **cuda_vbo_resource) { - void *ptr; - checkCudaErrors(cudaGraphicsMapResources(1, cuda_vbo_resource, 0)); - size_t num_bytes; - checkCudaErrors(cudaGraphicsResourceGetMappedPointer( - (void **)&ptr, &num_bytes, *cuda_vbo_resource)); - return ptr; -} - -void unmapGLBufferObject(struct cudaGraphicsResource *cuda_vbo_resource) { - checkCudaErrors(cudaGraphicsUnmapResources(1, &cuda_vbo_resource, 0)); -} - -void copyArrayFromDevice(void *host, const void *device, - struct cudaGraphicsResource **cuda_vbo_resource, - int size) { - if (cuda_vbo_resource) { - device = mapGLBufferObject(cuda_vbo_resource); + if (devID < 0) + { + printf("No CUDA Capable devices found, exiting...\n"); + exit(EXIT_SUCCESS); + } } - checkCudaErrors(cudaMemcpy(host, device, size, cudaMemcpyDeviceToHost)); - - if (cuda_vbo_resource) { - unmapGLBufferObject(*cuda_vbo_resource); + void allocateArray(void **devPtr, size_t size) + { + checkCudaErrors(cudaMalloc(devPtr, size)); } -} -void setParameters(SimParams *hostParams) { - // copy parameters to constant memory - checkCudaErrors(cudaMemcpyToSymbol(params, hostParams, sizeof(SimParams))); -} + void freeArray(void *devPtr) { checkCudaErrors(cudaFree(devPtr)); } -// Round a / b to nearest higher integer value -uint iDivUp(uint a, uint b) { return (a % b != 0) ? (a / b + 1) : (a / b); } + void threadSync() { checkCudaErrors(cudaDeviceSynchronize()); } -// compute grid and thread block size for a given number of elements -void computeGridSize(uint n, uint blockSize, uint &numBlocks, - uint &numThreads) { - numThreads = min(blockSize, n); - numBlocks = iDivUp(n, numThreads); -} + void copyArrayToDevice(void *device, const void *host, int offset, int size) + { + checkCudaErrors( + cudaMemcpy((char *)device + offset, host, size, cudaMemcpyHostToDevice)); + } -void integrateSystem(float *pos, float *vel, float deltaTime, - uint numParticles) { - thrust::device_ptr d_pos4((float4 *)pos); - thrust::device_ptr d_vel4((float4 *)vel); + void registerGLBufferObject(uint vbo, + struct cudaGraphicsResource **cuda_vbo_resource) + { + checkCudaErrors(cudaGraphicsGLRegisterBuffer(cuda_vbo_resource, vbo, + cudaGraphicsMapFlagsNone)); + } - thrust::for_each( - thrust::make_zip_iterator(thrust::make_tuple(d_pos4, d_vel4)), - thrust::make_zip_iterator( - thrust::make_tuple(d_pos4 + numParticles, d_vel4 + numParticles)), - integrate_functor(deltaTime)); -} + void unregisterGLBufferObject(struct cudaGraphicsResource *cuda_vbo_resource) + { + checkCudaErrors(cudaGraphicsUnregisterResource(cuda_vbo_resource)); + } -void calcHash(uint *gridParticleHash, uint *gridParticleIndex, float *pos, - int numParticles) { - uint numThreads, numBlocks; - computeGridSize(numParticles, 256, numBlocks, numThreads); + void *mapGLBufferObject(struct cudaGraphicsResource **cuda_vbo_resource) + { + void *ptr; + checkCudaErrors(cudaGraphicsMapResources(1, cuda_vbo_resource, 0)); + size_t num_bytes; + checkCudaErrors(cudaGraphicsResourceGetMappedPointer( + (void **)&ptr, &num_bytes, *cuda_vbo_resource)); + return ptr; + } - // execute the kernel - calcHashD<<>>(gridParticleHash, gridParticleIndex, - (float4 *)pos, numParticles); + void unmapGLBufferObject(struct cudaGraphicsResource *cuda_vbo_resource) + { + checkCudaErrors(cudaGraphicsUnmapResources(1, &cuda_vbo_resource, 0)); + } - // check if kernel invocation generated an error - getLastCudaError("Kernel execution failed"); -} + void copyArrayFromDevice(void *host, const void *device, + struct cudaGraphicsResource **cuda_vbo_resource, + int size) + { + if (cuda_vbo_resource) + { + device = mapGLBufferObject(cuda_vbo_resource); + } -void reorderDataAndFindCellStart(uint *cellStart, uint *cellEnd, - float *sortedPos, float *sortedVel, - uint *gridParticleHash, - uint *gridParticleIndex, float *oldPos, - float *oldVel, uint numParticles, - uint numCells) { - uint numThreads, numBlocks; - computeGridSize(numParticles, 256, numBlocks, numThreads); + checkCudaErrors(cudaMemcpy(host, device, size, cudaMemcpyDeviceToHost)); - // set all cells to empty - checkCudaErrors(cudaMemset(cellStart, 0xffffffff, numCells * sizeof(uint))); + if (cuda_vbo_resource) + { + unmapGLBufferObject(*cuda_vbo_resource); + } + } - uint smemSize = sizeof(uint) * (numThreads + 1); - reorderDataAndFindCellStartD<<>>( - cellStart, cellEnd, (float4 *)sortedPos, (float4 *)sortedVel, - gridParticleHash, gridParticleIndex, (float4 *)oldPos, (float4 *)oldVel, - numParticles); - getLastCudaError("Kernel execution failed: reorderDataAndFindCellStartD"); -} + void setParameters(SimParams *hostParams) + { + // copy parameters to constant memory + checkCudaErrors(cudaMemcpyToSymbol(cudaParams, hostParams, sizeof(SimParams))); + } -void collide(float *newVel, float *sortedPos, float *sortedVel, - uint *gridParticleIndex, uint *cellStart, uint *cellEnd, - uint numParticles, uint numCells) { - // thread per particle - uint numThreads, numBlocks; - computeGridSize(numParticles, 64, numBlocks, numThreads); + // Round a / b to nearest higher integer value + uint iDivUp(uint a, uint b) { return (a % b != 0) ? (a / b + 1) : (a / b); } - // execute the kernel - collideD<<>>((float4 *)newVel, (float4 *)sortedPos, - (float4 *)sortedVel, gridParticleIndex, - cellStart, cellEnd, numParticles); + // compute grid and thread block size for a given number of elements + void computeGridSize(uint n, uint blockSize, uint &numBlocks, + uint &numThreads) + { + numThreads = min(blockSize, n); + numBlocks = iDivUp(n, numThreads); + } - // check if kernel invocation generated an error - getLastCudaError("Kernel execution failed"); -} + void integrateSystem(float *pos, float *vel, float deltaTime, + uint numParticles) + { + thrust::device_ptr d_pos4((float4 *)pos); + thrust::device_ptr d_vel4((float4 *)vel); -void sortParticles(uint *dGridParticleHash, uint *dGridParticleIndex, - uint numParticles) { - thrust::sort_by_key( - thrust::device_ptr(dGridParticleHash), - thrust::device_ptr(dGridParticleHash + numParticles), - thrust::device_ptr(dGridParticleIndex)); -} + thrust::for_each( + thrust::make_zip_iterator(thrust::make_tuple(d_pos4, d_vel4)), + thrust::make_zip_iterator( + thrust::make_tuple(d_pos4 + numParticles, d_vel4 + numParticles)), + integrate_functor(deltaTime)); + } -} // extern "C" + void calcHash(uint *gridParticleHash, uint *gridParticleIndex, float *pos, + int numParticles) + { + uint numThreads, numBlocks; + computeGridSize(numParticles, 256, numBlocks, numThreads); + + // execute the kernel + calcHashD<<>>(gridParticleHash, gridParticleIndex, + (float4 *)pos, numParticles); + + // check if kernel invocation generated an error + getLastCudaError("Kernel execution failed"); + } + + void reorderDataAndFindCellStart(uint *cellStart, uint *cellEnd, + float *sortedPos, float *sortedVel, + uint *gridParticleHash, + uint *gridParticleIndex, float *oldPos, + float *oldVel, uint numParticles, + uint numCells) + { + uint numThreads, numBlocks; + computeGridSize(numParticles, 256, numBlocks, numThreads); + + // set all cells to empty + checkCudaErrors(cudaMemset(cellStart, 0xffffffff, numCells * sizeof(uint))); + + uint smemSize = sizeof(uint) * (numThreads + 1); + reorderDataAndFindCellStartD<<>>( + cellStart, cellEnd, (float4 *)sortedPos, (float4 *)sortedVel, + gridParticleHash, gridParticleIndex, (float4 *)oldPos, (float4 *)oldVel, + numParticles); + getLastCudaError("Kernel execution failed: reorderDataAndFindCellStartD"); + } + + void collide(float *newVel, float *sortedPos, float *sortedVel, + uint *gridParticleIndex, uint *cellStart, uint *cellEnd, + uint numParticles, uint numCells) + { + // thread per particle + uint numThreads, numBlocks; + computeGridSize(numParticles, 64, numBlocks, numThreads); + + // execute the kernel + collideD<<>>((float4 *)newVel, (float4 *)sortedPos, + (float4 *)sortedVel, gridParticleIndex, + cellStart, cellEnd, numParticles); + + // check if kernel invocation generated an error + getLastCudaError("Kernel execution failed"); + } + + void sortParticles(uint *dGridParticleHash, uint *dGridParticleIndex, + uint numParticles) + { + thrust::sort_by_key( + thrust::device_ptr(dGridParticleHash), + thrust::device_ptr(dGridParticleHash + numParticles), + thrust::device_ptr(dGridParticleIndex)); + } + +} // extern "C" diff --git a/Samples/2_Concepts_and_Techniques/particles/particles_kernel_impl.cuh b/Samples/2_Concepts_and_Techniques/particles/particles_kernel_impl.cuh index ba316264..5047239c 100644 --- a/Samples/2_Concepts_and_Techniques/particles/particles_kernel_impl.cuh +++ b/Samples/2_Concepts_and_Techniques/particles/particles_kernel_impl.cuh @@ -42,23 +42,25 @@ namespace cg = cooperative_groups; #include "particles_kernel.cuh" // simulation parameters in constant memory -__constant__ SimParams params; +__constant__ SimParams cudaParams; -struct integrate_functor { +struct integrate_functor +{ float deltaTime; __host__ __device__ integrate_functor(float delta_time) : deltaTime(delta_time) {} template - __device__ void operator()(Tuple t) { + __device__ void operator()(Tuple t) + { volatile float4 posData = thrust::get<0>(t); volatile float4 velData = thrust::get<1>(t); float3 pos = make_float3(posData.x, posData.y, posData.z); float3 vel = make_float3(velData.x, velData.y, velData.z); - vel += params.gravity * deltaTime; - vel *= params.globalDamping; + vel += cudaParams.gravity * deltaTime; + vel *= cudaParams.globalDamping; // new position = old position + velocity * deltaTime pos += vel * deltaTime; @@ -66,36 +68,42 @@ struct integrate_functor { // set this to zero to disable collisions with cube sides #if 1 - if (pos.x > 1.0f - params.particleRadius) { - pos.x = 1.0f - params.particleRadius; - vel.x *= params.boundaryDamping; + if (pos.x > 1.0f - cudaParams.particleRadius) + { + pos.x = 1.0f - cudaParams.particleRadius; + vel.x *= cudaParams.boundaryDamping; } - if (pos.x < -1.0f + params.particleRadius) { - pos.x = -1.0f + params.particleRadius; - vel.x *= params.boundaryDamping; + if (pos.x < -1.0f + cudaParams.particleRadius) + { + pos.x = -1.0f + cudaParams.particleRadius; + vel.x *= cudaParams.boundaryDamping; } - if (pos.y > 1.0f - params.particleRadius) { - pos.y = 1.0f - params.particleRadius; - vel.y *= params.boundaryDamping; + if (pos.y > 1.0f - cudaParams.particleRadius) + { + pos.y = 1.0f - cudaParams.particleRadius; + vel.y *= cudaParams.boundaryDamping; } - if (pos.z > 1.0f - params.particleRadius) { - pos.z = 1.0f - params.particleRadius; - vel.z *= params.boundaryDamping; + if (pos.z > 1.0f - cudaParams.particleRadius) + { + pos.z = 1.0f - cudaParams.particleRadius; + vel.z *= cudaParams.boundaryDamping; } - if (pos.z < -1.0f + params.particleRadius) { - pos.z = -1.0f + params.particleRadius; - vel.z *= params.boundaryDamping; + if (pos.z < -1.0f + cudaParams.particleRadius) + { + pos.z = -1.0f + cudaParams.particleRadius; + vel.z *= cudaParams.boundaryDamping; } #endif - if (pos.y < -1.0f + params.particleRadius) { - pos.y = -1.0f + params.particleRadius; - vel.y *= params.boundaryDamping; + if (pos.y < -1.0f + cudaParams.particleRadius) + { + pos.y = -1.0f + cudaParams.particleRadius; + vel.y *= cudaParams.boundaryDamping; } // store new position and velocity @@ -105,32 +113,36 @@ struct integrate_functor { }; // calculate position in uniform grid -__device__ int3 calcGridPos(float3 p) { +__device__ int3 calcGridPos(float3 p) +{ int3 gridPos; - gridPos.x = floorf((p.x - params.worldOrigin.x) / params.cellSize.x); - gridPos.y = floorf((p.y - params.worldOrigin.y) / params.cellSize.y); - gridPos.z = floorf((p.z - params.worldOrigin.z) / params.cellSize.z); + gridPos.x = floorf((p.x - cudaParams.worldOrigin.x) / cudaParams.cellSize.x); + gridPos.y = floorf((p.y - cudaParams.worldOrigin.y) / cudaParams.cellSize.y); + gridPos.z = floorf((p.z - cudaParams.worldOrigin.z) / cudaParams.cellSize.z); return gridPos; } // calculate address in grid from position (clamping to edges) -__device__ uint calcGridHash(int3 gridPos) { +__device__ uint calcGridHash(int3 gridPos) +{ gridPos.x = gridPos.x & - (params.gridSize.x - 1); // wrap grid, assumes size is power of 2 - gridPos.y = gridPos.y & (params.gridSize.y - 1); - gridPos.z = gridPos.z & (params.gridSize.z - 1); - return __umul24(__umul24(gridPos.z, params.gridSize.y), params.gridSize.x) + - __umul24(gridPos.y, params.gridSize.x) + gridPos.x; + (cudaParams.gridSize.x - 1); // wrap grid, assumes size is power of 2 + gridPos.y = gridPos.y & (cudaParams.gridSize.y - 1); + gridPos.z = gridPos.z & (cudaParams.gridSize.z - 1); + return __umul24(__umul24(gridPos.z, cudaParams.gridSize.y), cudaParams.gridSize.x) + + __umul24(gridPos.y, cudaParams.gridSize.x) + gridPos.x; } // calculate grid hash value for each particle -__global__ void calcHashD(uint *gridParticleHash, // output - uint *gridParticleIndex, // output - float4 *pos, // input: positions - uint numParticles) { +__global__ void calcHashD(uint *gridParticleHash, // output + uint *gridParticleIndex, // output + float4 *pos, // input: positions + uint numParticles) +{ uint index = __umul24(blockIdx.x, blockDim.x) + threadIdx.x; - if (index >= numParticles) return; + if (index >= numParticles) + return; volatile float4 p = pos[index]; @@ -146,24 +158,26 @@ __global__ void calcHashD(uint *gridParticleHash, // output // rearrange particle data into sorted order, and find the start of each cell // in the sorted hash array __global__ void reorderDataAndFindCellStartD( - uint *cellStart, // output: cell start index - uint *cellEnd, // output: cell end index - float4 *sortedPos, // output: sorted positions - float4 *sortedVel, // output: sorted velocities - uint *gridParticleHash, // input: sorted grid hashes - uint *gridParticleIndex, // input: sorted particle indices - float4 *oldPos, // input: sorted position array - float4 *oldVel, // input: sorted velocity array - uint numParticles) { + uint *cellStart, // output: cell start index + uint *cellEnd, // output: cell end index + float4 *sortedPos, // output: sorted positions + float4 *sortedVel, // output: sorted velocities + uint *gridParticleHash, // input: sorted grid hashes + uint *gridParticleIndex, // input: sorted particle indices + float4 *oldPos, // input: sorted position array + float4 *oldVel, // input: sorted velocity array + uint numParticles) +{ // Handle to thread block group cg::thread_block cta = cg::this_thread_block(); - extern __shared__ uint sharedHash[]; // blockSize + 1 elements + extern __shared__ uint sharedHash[]; // blockSize + 1 elements uint index = __umul24(blockIdx.x, blockDim.x) + threadIdx.x; uint hash; // handle case when no. of particles not multiple of block size - if (index < numParticles) { + if (index < numParticles) + { hash = gridParticleHash[index]; // Load hash data into shared memory so that we can look @@ -171,7 +185,8 @@ __global__ void reorderDataAndFindCellStartD( // two hash values per thread sharedHash[threadIdx.x + 1] = hash; - if (index > 0 && threadIdx.x == 0) { + if (index > 0 && threadIdx.x == 0) + { // first thread in block must load neighbor particle hash sharedHash[0] = gridParticleHash[index - 1]; } @@ -179,20 +194,24 @@ __global__ void reorderDataAndFindCellStartD( cg::sync(cta); - if (index < numParticles) { + if (index < numParticles) + { // If this particle has a different cell index to the previous // particle then it must be the first particle in the cell, // so store the index of this particle in the cell. // As it isn't the first particle, it must also be the cell end of // the previous particle's cell - if (index == 0 || hash != sharedHash[threadIdx.x]) { + if (index == 0 || hash != sharedHash[threadIdx.x]) + { cellStart[hash] = index; - if (index > 0) cellEnd[sharedHash[threadIdx.x]] = index; + if (index > 0) + cellEnd[sharedHash[threadIdx.x]] = index; } - if (index == numParticles - 1) { + if (index == numParticles - 1) + { cellEnd[hash] = index + 1; } @@ -209,7 +228,8 @@ __global__ void reorderDataAndFindCellStartD( // collide two spheres using DEM method __device__ float3 collideSpheres(float3 posA, float3 posB, float3 velA, float3 velB, float radiusA, float radiusB, - float attraction) { + float attraction) +{ // calculate relative position float3 relPos = posB - posA; @@ -218,7 +238,8 @@ __device__ float3 collideSpheres(float3 posA, float3 posB, float3 velA, float3 force = make_float3(0.0f); - if (dist < collideDist) { + if (dist < collideDist) + { float3 norm = relPos / dist; // relative velocity @@ -228,11 +249,11 @@ __device__ float3 collideSpheres(float3 posA, float3 posB, float3 velA, float3 tanVel = relVel - (dot(relVel, norm) * norm); // spring force - force = -params.spring * (collideDist - dist) * norm; + force = -cudaParams.spring * (collideDist - dist) * norm; // dashpot (damping) force - force += params.damping * relVel; + force += cudaParams.damping * relVel; // tangential shear force - force += params.shear * tanVel; + force += cudaParams.shear * tanVel; // attraction force += attraction * relPos; } @@ -243,7 +264,8 @@ __device__ float3 collideSpheres(float3 posA, float3 posB, float3 velA, // collide a particle against all other particles in a given cell __device__ float3 collideCell(int3 gridPos, uint index, float3 pos, float3 vel, float4 *oldPos, float4 *oldVel, uint *cellStart, - uint *cellEnd) { + uint *cellEnd) +{ uint gridHash = calcGridHash(gridPos); // get start of bucket for this cell @@ -251,20 +273,21 @@ __device__ float3 collideCell(int3 gridPos, uint index, float3 pos, float3 vel, float3 force = make_float3(0.0f); - if (startIndex != 0xffffffff) // cell is not empty + if (startIndex != 0xffffffff) // cell is not empty { // iterate over particles in this cell uint endIndex = cellEnd[gridHash]; - for (uint j = startIndex; j < endIndex; j++) { - if (j != index) // check not colliding with self + for (uint j = startIndex; j < endIndex; j++) + { + if (j != index) // check not colliding with self { float3 pos2 = make_float3(oldPos[j]); float3 vel2 = make_float3(oldVel[j]); // collide two spheres - force += collideSpheres(pos, pos2, vel, vel2, params.particleRadius, - params.particleRadius, params.attraction); + force += collideSpheres(pos, pos2, vel, vel2, cudaParams.particleRadius, + cudaParams.particleRadius, cudaParams.attraction); } } } @@ -273,14 +296,16 @@ __device__ float3 collideCell(int3 gridPos, uint index, float3 pos, float3 vel, } __global__ void collideD( - float4 *newVel, // output: new velocity - float4 *oldPos, // input: sorted positions - float4 *oldVel, // input: sorted velocities - uint *gridParticleIndex, // input: sorted particle indices - uint *cellStart, uint *cellEnd, uint numParticles) { + float4 *newVel, // output: new velocity + float4 *oldPos, // input: sorted positions + float4 *oldVel, // input: sorted velocities + uint *gridParticleIndex, // input: sorted particle indices + uint *cellStart, uint *cellEnd, uint numParticles) +{ uint index = __mul24(blockIdx.x, blockDim.x) + threadIdx.x; - if (index >= numParticles) return; + if (index >= numParticles) + return; // read particle data from sorted arrays float3 pos = make_float3(oldPos[index]); @@ -292,9 +317,12 @@ __global__ void collideD( // examine neighbouring cells float3 force = make_float3(0.0f); - for (int z = -1; z <= 1; z++) { - for (int y = -1; y <= 1; y++) { - for (int x = -1; x <= 1; x++) { + for (int z = -1; z <= 1; z++) + { + for (int y = -1; y <= 1; y++) + { + for (int x = -1; x <= 1; x++) + { int3 neighbourPos = gridPos + make_int3(x, y, z); force += collideCell(neighbourPos, index, pos, vel, oldPos, oldVel, cellStart, cellEnd); @@ -303,9 +331,9 @@ __global__ void collideD( } // collide with cursor sphere - force += collideSpheres(pos, params.colliderPos, vel, - make_float3(0.0f, 0.0f, 0.0f), params.particleRadius, - params.colliderRadius, 0.0f); + force += collideSpheres(pos, cudaParams.colliderPos, vel, + make_float3(0.0f, 0.0f, 0.0f), cudaParams.particleRadius, + cudaParams.colliderRadius, 0.0f); // write new velocity back to original unsorted location uint originalIndex = gridParticleIndex[index];