diff --git a/CHANGELOG.md b/CHANGELOG.md index c3a12dba..1b3dc904 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -6,6 +6,7 @@ * `c++11_cuda` demonstrating CUDA and C++ 11 interoperability (reason: obsolete) * `concurrentKernels` demonstrating the ability to run multiple kernels simultaneously (reason: obsolete) * `cppIntegration` demonstrating calling between .cu and .cpp files (reason: obsolete) + * `cppOverload` demonstrating C++ function overloading (reason: obsolete) ### CUDA 12.5 diff --git a/Samples/0_Introduction/cppOverload/.vscode/c_cpp_properties.json b/Samples/0_Introduction/cppOverload/.vscode/c_cpp_properties.json deleted file mode 100644 index f0066b0f..00000000 --- a/Samples/0_Introduction/cppOverload/.vscode/c_cpp_properties.json +++ /dev/null @@ -1,18 +0,0 @@ -{ - "configurations": [ - { - "name": "Linux", - "includePath": [ - "${workspaceFolder}/**", - "${workspaceFolder}/../../../Common" - ], - "defines": [], - "compilerPath": "/usr/local/cuda/bin/nvcc", - "cStandard": "gnu17", - "cppStandard": "gnu++14", - "intelliSenseMode": "linux-gcc-x64", - "configurationProvider": "ms-vscode.makefile-tools" - } - ], - "version": 4 -} diff --git a/Samples/0_Introduction/cppOverload/.vscode/extensions.json b/Samples/0_Introduction/cppOverload/.vscode/extensions.json deleted file mode 100644 index c7eb54dc..00000000 --- a/Samples/0_Introduction/cppOverload/.vscode/extensions.json +++ /dev/null @@ -1,7 +0,0 @@ -{ - "recommendations": [ - "nvidia.nsight-vscode-edition", - "ms-vscode.cpptools", - "ms-vscode.makefile-tools" - ] -} diff --git a/Samples/0_Introduction/cppOverload/.vscode/launch.json b/Samples/0_Introduction/cppOverload/.vscode/launch.json deleted file mode 100644 index 91413785..00000000 --- a/Samples/0_Introduction/cppOverload/.vscode/launch.json +++ /dev/null @@ -1,10 +0,0 @@ -{ - "configurations": [ - { - "name": "CUDA C++: Launch", - "type": "cuda-gdb", - "request": "launch", - "program": "${workspaceFolder}/cppOverload" - } - ] -} diff --git a/Samples/0_Introduction/cppOverload/.vscode/tasks.json b/Samples/0_Introduction/cppOverload/.vscode/tasks.json deleted file mode 100644 index 4509aeb1..00000000 --- a/Samples/0_Introduction/cppOverload/.vscode/tasks.json +++ /dev/null @@ -1,15 +0,0 @@ -{ - "version": "2.0.0", - "tasks": [ - { - "label": "sample", - "type": "shell", - "command": "make dbg=1", - "problemMatcher": ["$nvcc"], - "group": { - "kind": "build", - "isDefault": true - } - } - ] -} diff --git a/Samples/0_Introduction/cppOverload/Makefile b/Samples/0_Introduction/cppOverload/Makefile deleted file mode 100644 index c6856a44..00000000 --- a/Samples/0_Introduction/cppOverload/Makefile +++ /dev/null @@ -1,357 +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 - -ifeq ($(SAMPLE_ENABLED),0) -EXEC ?= @echo "[@]" -endif - -################################################################################ - -# Target rules -all: build - -build: cppOverload - -check.deps: -ifeq ($(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 - -cppOverload.o:cppOverload.cu - $(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $< - -cppOverload: cppOverload.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 - $(EXEC) ./cppOverload - -testrun: build - -clean: - rm -f cppOverload cppOverload.o - rm -rf ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)/cppOverload - -clobber: clean diff --git a/Samples/0_Introduction/cppOverload/NsightEclipse.xml b/Samples/0_Introduction/cppOverload/NsightEclipse.xml deleted file mode 100644 index a4cf0c85..00000000 --- a/Samples/0_Introduction/cppOverload/NsightEclipse.xml +++ /dev/null @@ -1,83 +0,0 @@ - - - - cppOverload - - cudaMemcpy - cudaFuncSetCacheConfig - cudaFree - cudaMallocHost - cudaSetDevice - cudaGetDeviceProperties - cudaDeviceSynchronize - cudaFreeHost - cudaMalloc - cudaFuncGetAttributes - cudaGetDeviceCount - - - whole - - ./ - ../ - ../../../Common - - - C++ Function Overloading - CUDA Streams and Events - - - GPGPU - - - - - - true - cppOverload_kernel.cu - - 1:CUDA Basic Topics - 1:Performance Strategies - - 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 - - cppOverload - exe - diff --git a/Samples/0_Introduction/cppOverload/README.md b/Samples/0_Introduction/cppOverload/README.md deleted file mode 100644 index dafa4af6..00000000 --- a/Samples/0_Introduction/cppOverload/README.md +++ /dev/null @@ -1,70 +0,0 @@ -# cppOverload - cppOverload - -## Description - -This sample demonstrates how to use C++ function overloading on the GPU. - -## Key Concepts - -C++ Function Overloading, CUDA Streams and Events - -## Supported SM Architectures - -[SM 5.0 ](https://developer.nvidia.com/cuda-gpus) [SM 5.2 ](https://developer.nvidia.com/cuda-gpus) [SM 5.3 ](https://developer.nvidia.com/cuda-gpus) [SM 6.0 ](https://developer.nvidia.com/cuda-gpus) [SM 6.1 ](https://developer.nvidia.com/cuda-gpus) [SM 7.0 ](https://developer.nvidia.com/cuda-gpus) [SM 7.2 ](https://developer.nvidia.com/cuda-gpus) [SM 7.5 ](https://developer.nvidia.com/cuda-gpus) [SM 8.0 ](https://developer.nvidia.com/cuda-gpus) [SM 8.6 ](https://developer.nvidia.com/cuda-gpus) [SM 8.7 ](https://developer.nvidia.com/cuda-gpus) [SM 8.9 ](https://developer.nvidia.com/cuda-gpus) [SM 9.0 ](https://developer.nvidia.com/cuda-gpus) - -## Supported OSes - -Linux, Windows - -## Supported CPU Architecture - -x86_64, ppc64le, armv7l - -## CUDA APIs involved - -### [CUDA Runtime API](http://docs.nvidia.com/cuda/cuda-runtime-api/index.html) -cudaMemcpy, cudaFuncSetCacheConfig, cudaFree, cudaMallocHost, cudaSetDevice, cudaGetDeviceProperties, cudaDeviceSynchronize, cudaFreeHost, cudaMalloc, cudaFuncGetAttributes, cudaGetDeviceCount - -## Prerequisites - -Download and install the [CUDA Toolkit 12.5](https://developer.nvidia.com/cuda-downloads) for your corresponding platform. - -## Build and Run - -### Windows -The Windows samples are built using the Visual Studio IDE. Solution files (.sln) are provided for each supported version of Visual Studio, using the format: -``` -*_vs.sln - for Visual Studio -``` -Each individual sample has its own set of solution files in its directory: - -To build/examine all the samples at once, the complete solution files should be used. To build/examine a single sample, the individual sample solution files should be used. -> **Note:** Some samples require that the Microsoft DirectX SDK (June 2010 or newer) be installed and that the VC++ directory paths are properly set up (**Tools > Options...**). Check DirectX Dependencies section for details." - -### Linux -The Linux samples are built using makefiles. To use the makefiles, change the current directory to the sample directory you wish to build, and run make: -``` -$ cd -$ make -``` -The samples makefiles can take advantage of certain options: -* **TARGET_ARCH=** - cross-compile targeting a specific architecture. Allowed architectures are x86_64, ppc64le, armv7l. - By default, TARGET_ARCH is set to HOST_ARCH. On a x86_64 machine, not setting TARGET_ARCH is the equivalent of setting TARGET_ARCH=x86_64.
-`$ make TARGET_ARCH=x86_64`
`$ make TARGET_ARCH=ppc64le`
`$ make TARGET_ARCH=armv7l`
- See [here](http://docs.nvidia.com/cuda/cuda-samples/index.html#cross-samples) for more details. -* **dbg=1** - build with debug symbols - ``` - $ make dbg=1 - ``` -* **SMS="A B ..."** - override the SM architectures for which the sample will be built, where `"A B ..."` is a space-delimited list of SM architectures. For example, to generate SASS for SM 50 and SM 60, use `SMS="50 60"`. - ``` - $ make SMS="50 60" - ``` - -* **HOST_COMPILER=** - override the default g++ host compiler. See the [Linux Installation Guide](http://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html#system-requirements) for a list of supported host compilers. -``` - $ make HOST_COMPILER=g++ -``` - -## References (for more details) - diff --git a/Samples/0_Introduction/cppOverload/cppOverload.cu b/Samples/0_Introduction/cppOverload/cppOverload.cu deleted file mode 100644 index 9d191ae9..00000000 --- a/Samples/0_Introduction/cppOverload/cppOverload.cu +++ /dev/null @@ -1,188 +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. - */ - -#define THREAD_N 256 -#define N 1024 -#define DIV_UP(a, b) (((a) + (b) - 1) / (b)) - -// Includes, system -#include -#include -#include -#include -#include "cppOverload_kernel.cuh" - -const char *sampleName = "C++ Function Overloading"; - -#define OUTPUT_ATTR(attr) \ - printf("Shared Size: %d\n", (int)attr.sharedSizeBytes); \ - printf("Constant Size: %d\n", (int)attr.constSizeBytes); \ - printf("Local Size: %d\n", (int)attr.localSizeBytes); \ - printf("Max Threads Per Block: %d\n", attr.maxThreadsPerBlock); \ - printf("Number of Registers: %d\n", attr.numRegs); \ - printf("PTX Version: %d\n", attr.ptxVersion); \ - printf("Binary Version: %d\n", attr.binaryVersion); - -bool check_func1(int *hInput, int *hOutput, int a) { - for (int i = 0; i < N; ++i) { - int cpuRes = hInput[i] * a + i; - - if (hOutput[i] != cpuRes) { - return false; - } - } - - return true; -} - -bool check_func2(int2 *hInput, int *hOutput, int a) { - for (int i = 0; i < N; i++) { - int cpuRes = (hInput[i].x + hInput[i].y) * a + i; - - if (hOutput[i] != cpuRes) { - return false; - } - } - - return true; -} - -bool check_func3(int *hInput1, int *hInput2, int *hOutput, int a) { - for (int i = 0; i < N; i++) { - if (hOutput[i] != (hInput1[i] + hInput2[i]) * a + i) { - return false; - } - } - - return true; -} - -int main(int argc, const char *argv[]) { - int *hInput = NULL; - int *hOutput = NULL; - int *dInput = NULL; - int *dOutput = NULL; - - printf("%s starting...\n", sampleName); - - int deviceCount; - checkCudaErrors(cudaGetDeviceCount(&deviceCount)); - printf("Device Count: %d\n", deviceCount); - - int deviceID = findCudaDevice(argc, argv); - cudaDeviceProp prop; - checkCudaErrors(cudaGetDeviceProperties(&prop, deviceID)); - if (prop.major < 2) { - printf( - "ERROR: cppOverload requires GPU devices with compute SM 2.0 or " - "higher.\n"); - printf("Current GPU device has compute SM%d.%d, Exiting...", prop.major, - prop.minor); - exit(EXIT_WAIVED); - } - - checkCudaErrors(cudaSetDevice(deviceID)); - - // Allocate device memory - checkCudaErrors(cudaMalloc(&dInput, sizeof(int) * N * 2)); - checkCudaErrors(cudaMalloc(&dOutput, sizeof(int) * N)); - - // Allocate host memory - checkCudaErrors(cudaMallocHost(&hInput, sizeof(int) * N * 2)); - checkCudaErrors(cudaMallocHost(&hOutput, sizeof(int) * N)); - - for (int i = 0; i < N * 2; i++) { - hInput[i] = i; - } - - // Copy data from host to device - checkCudaErrors( - cudaMemcpy(dInput, hInput, sizeof(int) * N * 2, cudaMemcpyHostToDevice)); - - // Test C++ overloading - bool testResult = true; - bool funcResult = true; - int a = 1; - - void (*func1)(const int *, int *, int); - void (*func2)(const int2 *, int *, int); - void (*func3)(const int *, const int *, int *, int); - struct cudaFuncAttributes attr; - - // overload function 1 - func1 = simple_kernel; - memset(&attr, 0, sizeof(attr)); - checkCudaErrors(cudaFuncSetCacheConfig(*func1, cudaFuncCachePreferShared)); - checkCudaErrors(cudaFuncGetAttributes(&attr, *func1)); - OUTPUT_ATTR(attr); - (*func1)<<>>(dInput, dOutput, a); - checkCudaErrors( - cudaMemcpy(hOutput, dOutput, sizeof(int) * N, cudaMemcpyDeviceToHost)); - funcResult = check_func1(hInput, hOutput, a); - printf("simple_kernel(const int *pIn, int *pOut, int a) %s\n\n", - funcResult ? "PASSED" : "FAILED"); - testResult &= funcResult; - - // overload function 2 - func2 = simple_kernel; - memset(&attr, 0, sizeof(attr)); - checkCudaErrors(cudaFuncSetCacheConfig(*func2, cudaFuncCachePreferShared)); - checkCudaErrors(cudaFuncGetAttributes(&attr, *func2)); - OUTPUT_ATTR(attr); - (*func2)<<>>((int2 *)dInput, dOutput, a); - checkCudaErrors( - cudaMemcpy(hOutput, dOutput, sizeof(int) * N, cudaMemcpyDeviceToHost)); - funcResult = check_func2(reinterpret_cast(hInput), hOutput, a); - printf("simple_kernel(const int2 *pIn, int *pOut, int a) %s\n\n", - funcResult ? "PASSED" : "FAILED"); - testResult &= funcResult; - - // overload function 3 - func3 = simple_kernel; - memset(&attr, 0, sizeof(attr)); - checkCudaErrors(cudaFuncSetCacheConfig(*func3, cudaFuncCachePreferShared)); - checkCudaErrors(cudaFuncGetAttributes(&attr, *func3)); - OUTPUT_ATTR(attr); - (*func3)<<>>(dInput, dInput + N, dOutput, a); - checkCudaErrors( - cudaMemcpy(hOutput, dOutput, sizeof(int) * N, cudaMemcpyDeviceToHost)); - funcResult = check_func3(&hInput[0], &hInput[N], hOutput, a); - printf( - "simple_kernel(const int *pIn1, const int *pIn2, int *pOut, int a) " - "%s\n\n", - funcResult ? "PASSED" : "FAILED"); - testResult &= funcResult; - - checkCudaErrors(cudaFree(dInput)); - checkCudaErrors(cudaFree(dOutput)); - checkCudaErrors(cudaFreeHost(hOutput)); - checkCudaErrors(cudaFreeHost(hInput)); - - checkCudaErrors(cudaDeviceSynchronize()); - - exit(testResult ? EXIT_SUCCESS : EXIT_FAILURE); -} diff --git a/Samples/0_Introduction/cppOverload/cppOverload_kernel.cuh b/Samples/0_Introduction/cppOverload/cppOverload_kernel.cuh deleted file mode 100644 index 23545905..00000000 --- a/Samples/0_Introduction/cppOverload/cppOverload_kernel.cuh +++ /dev/null @@ -1,61 +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. - */ - -__global__ void simple_kernel(const int *pIn, int *pOut, int a) { - __shared__ int sData[THREAD_N]; - int tid = threadIdx.x + blockDim.x * blockIdx.x; - - sData[threadIdx.x] = pIn[tid]; - __syncthreads(); - - pOut[tid] = sData[threadIdx.x] * a + tid; - ; -} - -__global__ void simple_kernel(const int2 *pIn, int *pOut, int a) { - __shared__ int2 sData[THREAD_N]; - int tid = threadIdx.x + blockDim.x * blockIdx.x; - - sData[threadIdx.x] = pIn[tid]; - __syncthreads(); - - pOut[tid] = (sData[threadIdx.x].x + sData[threadIdx.x].y) * a + tid; - ; -} - -__global__ void simple_kernel(const int *pIn1, const int *pIn2, int *pOut, - int a) { - __shared__ int sData1[THREAD_N]; - __shared__ int sData2[THREAD_N]; - int tid = threadIdx.x + blockDim.x * blockIdx.x; - - sData1[threadIdx.x] = pIn1[tid]; - sData2[threadIdx.x] = pIn2[tid]; - __syncthreads(); - - pOut[tid] = (sData1[threadIdx.x] + sData2[threadIdx.x]) * a + tid; -}