mirror of
https://github.com/NVIDIA/cuda-samples.git
synced 2025-11-04 15:47:50 +08:00
Remove simpleTemplates_nvrtc
This commit is contained in:
Rob Armstrong
parent
769a225af3
commit
a461e61485
@ -9,7 +9,7 @@
|
||||
* `cppIntegration` demonstrating calling between .cu and .cpp files (reason: obsolete)
|
||||
* `cppOverload` demonstrating C++ function overloading (reason: obsolete)
|
||||
* `simpleSeparateCompilation` demonstrating NVCC compilation to a static library (reason: trivial)
|
||||
*
|
||||
* `simpleTemplates_nvrtc` demonstrating NVRTC usage for `simpleTemplates` sample (reason: redundant)
|
||||
|
||||
### CUDA 12.5
|
||||
|
||||
|
||||
@ -32,8 +32,7 @@ add_subdirectory(simplePitchLinearTexture)
|
||||
add_subdirectory(simplePrintf)
|
||||
add_subdirectory(simpleStreams)
|
||||
add_subdirectory(simpleSurfaceWrite)
|
||||
#add_subdirectory(simpleTemplates)
|
||||
#add_subdirectory(simpleTemplates_nvrtc)
|
||||
add_subdirectory(simpleTemplates)
|
||||
#add_subdirectory(simpleTexture)
|
||||
#add_subdirectory(simpleTexture3D)
|
||||
#add_subdirectory(simpleTextureDrv)
|
||||
|
||||
@ -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
|
||||
}
|
||||
@ -1,7 +0,0 @@
|
||||
{
|
||||
"recommendations": [
|
||||
"nvidia.nsight-vscode-edition",
|
||||
"ms-vscode.cpptools",
|
||||
"ms-vscode.makefile-tools"
|
||||
]
|
||||
}
|
||||
@ -1,10 +0,0 @@
|
||||
{
|
||||
"configurations": [
|
||||
{
|
||||
"name": "CUDA C++: Launch",
|
||||
"type": "cuda-gdb",
|
||||
"request": "launch",
|
||||
"program": "${workspaceFolder}/simpleTemplates_nvrtc"
|
||||
}
|
||||
]
|
||||
}
|
||||
@ -1,15 +0,0 @@
|
||||
{
|
||||
"version": "2.0.0",
|
||||
"tasks": [
|
||||
{
|
||||
"label": "sample",
|
||||
"type": "shell",
|
||||
"command": "make dbg=1",
|
||||
"problemMatcher": ["$nvcc"],
|
||||
"group": {
|
||||
"kind": "build",
|
||||
"isDefault": true
|
||||
}
|
||||
}
|
||||
]
|
||||
}
|
||||
@ -1,409 +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))
|
||||
|
||||
UBUNTU = $(shell lsb_release -i -s 2>/dev/null | grep -i ubuntu)
|
||||
|
||||
SAMPLE_ENABLED := 1
|
||||
|
||||
# This sample is not supported on ARMv7
|
||||
ifeq ($(TARGET_ARCH),armv7l)
|
||||
$(info >>> WARNING - simpleTemplates_nvrtc is not supported on ARMv7 - 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 :=
|
||||
|
||||
################################################################################
|
||||
|
||||
# libNVRTC specific libraries
|
||||
ifeq ($(TARGET_OS),darwin)
|
||||
LDFLAGS += -L$(CUDA_PATH)/lib -F/Library/Frameworks -framework CUDA
|
||||
endif
|
||||
|
||||
ifeq ($(TARGET_OS),darwin)
|
||||
ALL_LDFLAGS += -Xcompiler -F/Library/Frameworks -Xlinker -framework -Xlinker CUDA
|
||||
else
|
||||
ifeq ($(TARGET_ARCH),x86_64)
|
||||
CUDA_SEARCH_PATH ?= $(CUDA_PATH)/lib64/stubs
|
||||
CUDA_SEARCH_PATH += $(CUDA_PATH)/lib/stubs
|
||||
CUDA_SEARCH_PATH += $(CUDA_PATH)/targets/x86_64-linux/lib/stubs
|
||||
endif
|
||||
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-linux)
|
||||
CUDA_SEARCH_PATH ?= $(CUDA_PATH)/targets/armv7-linux-gnueabihf/lib/stubs
|
||||
endif
|
||||
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-linux)
|
||||
CUDA_SEARCH_PATH ?= $(CUDA_PATH)/targets/aarch64-linux/lib/stubs
|
||||
endif
|
||||
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),sbsa-linux)
|
||||
CUDA_SEARCH_PATH ?= $(CUDA_PATH)/targets/sbsa-linux/lib/stubs
|
||||
endif
|
||||
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-android)
|
||||
CUDA_SEARCH_PATH ?= $(CUDA_PATH)/targets/armv7-linux-androideabi/lib/stubs
|
||||
endif
|
||||
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-android)
|
||||
CUDA_SEARCH_PATH ?= $(CUDA_PATH)/targets/aarch64-linux-androideabi/lib/stubs
|
||||
endif
|
||||
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-qnx)
|
||||
CUDA_SEARCH_PATH ?= $(CUDA_PATH)/targets/ARMv7-linux-QNX/lib/stubs
|
||||
endif
|
||||
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-qnx)
|
||||
CUDA_SEARCH_PATH ?= $(CUDA_PATH)/targets/aarch64-qnx/lib/stubs
|
||||
ifdef TARGET_OVERRIDE
|
||||
CUDA_SEARCH_PATH := $(CUDA_PATH)/targets/$(TARGET_OVERRIDE)/lib/stubs
|
||||
endif
|
||||
endif
|
||||
|
||||
ifeq ($(TARGET_ARCH),ppc64le)
|
||||
CUDA_SEARCH_PATH ?= $(CUDA_PATH)/targets/ppc64le-linux/lib/stubs
|
||||
endif
|
||||
|
||||
ifeq ($(HOST_ARCH),ppc64le)
|
||||
CUDA_SEARCH_PATH += $(CUDA_PATH)/lib64/stubs
|
||||
endif
|
||||
|
||||
CUDALIB ?= $(shell find -L $(CUDA_SEARCH_PATH) -maxdepth 1 -name libcuda.so 2> /dev/null)
|
||||
ifeq ("$(CUDALIB)","")
|
||||
$(info >>> WARNING - libcuda.so not found, CUDA Driver is not installed. Please re-install the driver. <<<)
|
||||
SAMPLE_ENABLED := 0
|
||||
else
|
||||
CUDALIB := $(shell echo $(CUDALIB) | sed "s/ .*//" | sed "s/\/libcuda.so//" )
|
||||
LIBRARIES += -L$(CUDALIB) -lcuda
|
||||
endif
|
||||
endif
|
||||
|
||||
ALL_CCFLAGS += --threads 0 --std=c++11
|
||||
|
||||
INCLUDES += -I$(CUDA_PATH)/include
|
||||
|
||||
LIBRARIES += -lnvrtc
|
||||
|
||||
ifeq ($(SAMPLE_ENABLED),0)
|
||||
EXEC ?= @echo "[@]"
|
||||
endif
|
||||
|
||||
################################################################################
|
||||
|
||||
# Target rules
|
||||
all: build
|
||||
|
||||
build: simpleTemplates_nvrtc
|
||||
|
||||
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
|
||||
|
||||
simpleTemplates.o:simpleTemplates.cpp
|
||||
$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $<
|
||||
|
||||
simpleTemplates_nvrtc: simpleTemplates.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
|
||||
$(EXEC) ./simpleTemplates_nvrtc
|
||||
|
||||
testrun: build
|
||||
|
||||
clean:
|
||||
rm -f simpleTemplates_nvrtc simpleTemplates.o
|
||||
rm -rf ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)/simpleTemplates_nvrtc
|
||||
|
||||
clobber: clean
|
||||
@ -1,74 +0,0 @@
|
||||
# simpleTemplates_nvrtc - Simple Templates with libNVRTC
|
||||
|
||||
## Description
|
||||
|
||||
This sample is a templatized version of the template project. It also shows how to correctly templatize dynamically allocated shared memory arrays.
|
||||
|
||||
## Key Concepts
|
||||
|
||||
C++ Templates, Runtime Compilation
|
||||
|
||||
## 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, QNX
|
||||
|
||||
## Supported CPU Architecture
|
||||
|
||||
x86_64, ppc64le, aarch64
|
||||
|
||||
## CUDA APIs involved
|
||||
|
||||
### [CUDA Driver API](http://docs.nvidia.com/cuda/cuda-driver-api/index.html)
|
||||
cuMemcpyDtoH, cuLaunchKernel, cuMemcpyHtoD, cuCtxSynchronize, cuMemAlloc, cuMemFree, cuModuleGetFunction
|
||||
|
||||
## Dependencies needed to build/run
|
||||
[NVRTC](../../../README.md#nvrtc)
|
||||
|
||||
## Prerequisites
|
||||
|
||||
Download and install the [CUDA Toolkit 12.5](https://developer.nvidia.com/cuda-downloads) for your corresponding platform.
|
||||
Make sure the dependencies mentioned in [Dependencies]() section above are installed.
|
||||
|
||||
## 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<version>.sln - for Visual Studio <version>
|
||||
```
|
||||
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 <sample_dir>
|
||||
$ make
|
||||
```
|
||||
The samples makefiles can take advantage of certain options:
|
||||
* **TARGET_ARCH=<arch>** - cross-compile targeting a specific architecture. Allowed architectures are x86_64, ppc64le, aarch64.
|
||||
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.<br/>
|
||||
`$ make TARGET_ARCH=x86_64` <br/> `$ make TARGET_ARCH=ppc64le` <br/> `$ make TARGET_ARCH=aarch64` <br/>
|
||||
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=<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)
|
||||
|
||||
@ -1,177 +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.
|
||||
*/
|
||||
|
||||
#ifndef _SHAREDMEM_H_
|
||||
#define _SHAREDMEM_H_
|
||||
|
||||
//****************************************************************************
|
||||
// Because dynamically sized shared memory arrays are declared "extern",
|
||||
// we can't templatize them directly. To get around this, we declare a
|
||||
// simple wrapper struct that will declare the extern array with a different
|
||||
// name depending on the type. This avoids compiler errors about duplicate
|
||||
// definitions.
|
||||
//
|
||||
|
||||
// To use dynamically allocated shared memory in a templatized __global__ or
|
||||
// __device__ function, just replace code like this:
|
||||
//
|
||||
|
||||
//
|
||||
// template<class T>
|
||||
// __global__ void
|
||||
// foo( T* g_idata, T* g_odata)
|
||||
// {
|
||||
// // Shared mem size is determined by the host app at run time
|
||||
// extern __shared__ T sdata[];
|
||||
// ...
|
||||
// doStuff(sdata);
|
||||
// ...
|
||||
// }
|
||||
|
||||
//
|
||||
// With this
|
||||
// template<class T>
|
||||
// __global__ void
|
||||
// foo( T* g_idata, T* g_odata)
|
||||
// {
|
||||
// // Shared mem size is determined by the host app at run time
|
||||
// SharedMemory<T> smem;
|
||||
// T* sdata = smem.getPointer();
|
||||
// ...
|
||||
// doStuff(sdata);
|
||||
// ...
|
||||
// }
|
||||
//****************************************************************************
|
||||
|
||||
// This is the un-specialized struct. Note that we prevent instantiation of
|
||||
// this
|
||||
// struct by putting an undefined symbol in the function body so it won't
|
||||
// compile.
|
||||
|
||||
template <typename T>
|
||||
struct SharedMemory {
|
||||
// Ensure that we won't compile any un-specialized types
|
||||
__device__ T *getPointer() {
|
||||
extern __device__ void error(void);
|
||||
error();
|
||||
return NULL;
|
||||
}
|
||||
};
|
||||
|
||||
// Following are the specializations for the following types.
|
||||
// int, uint, char, uchar, short, ushort, long, ulong, bool, float, and double
|
||||
// One could also specialize it for user-defined types.
|
||||
|
||||
template <>
|
||||
struct SharedMemory<int> {
|
||||
__device__ int *getPointer() {
|
||||
extern __shared__ int s_int[];
|
||||
return s_int;
|
||||
}
|
||||
};
|
||||
|
||||
template <>
|
||||
struct SharedMemory<unsigned int> {
|
||||
__device__ unsigned int *getPointer() {
|
||||
extern __shared__ unsigned int s_uint[];
|
||||
return s_uint;
|
||||
}
|
||||
};
|
||||
|
||||
template <>
|
||||
struct SharedMemory<char> {
|
||||
__device__ char *getPointer() {
|
||||
extern __shared__ char s_char[];
|
||||
return s_char;
|
||||
}
|
||||
};
|
||||
|
||||
template <>
|
||||
struct SharedMemory<unsigned char> {
|
||||
__device__ unsigned char *getPointer() {
|
||||
extern __shared__ unsigned char s_uchar[];
|
||||
return s_uchar;
|
||||
}
|
||||
};
|
||||
|
||||
template <>
|
||||
struct SharedMemory<short> {
|
||||
__device__ short *getPointer() {
|
||||
extern __shared__ short s_short[];
|
||||
return s_short;
|
||||
}
|
||||
};
|
||||
|
||||
template <>
|
||||
struct SharedMemory<unsigned short> {
|
||||
__device__ unsigned short *getPointer() {
|
||||
extern __shared__ unsigned short s_ushort[];
|
||||
return s_ushort;
|
||||
}
|
||||
};
|
||||
|
||||
template <>
|
||||
struct SharedMemory<long> {
|
||||
__device__ long *getPointer() {
|
||||
extern __shared__ long s_long[];
|
||||
return s_long;
|
||||
}
|
||||
};
|
||||
|
||||
template <>
|
||||
struct SharedMemory<unsigned long> {
|
||||
__device__ unsigned long *getPointer() {
|
||||
extern __shared__ unsigned long s_ulong[];
|
||||
return s_ulong;
|
||||
}
|
||||
};
|
||||
|
||||
template <>
|
||||
struct SharedMemory<bool> {
|
||||
__device__ bool *getPointer() {
|
||||
extern __shared__ bool s_bool[];
|
||||
return s_bool;
|
||||
}
|
||||
};
|
||||
|
||||
template <>
|
||||
struct SharedMemory<float> {
|
||||
__device__ float *getPointer() {
|
||||
extern __shared__ float s_float[];
|
||||
return s_float;
|
||||
}
|
||||
};
|
||||
|
||||
template <>
|
||||
struct SharedMemory<double> {
|
||||
__device__ double *getPointer() {
|
||||
extern __shared__ double s_double[];
|
||||
return s_double;
|
||||
}
|
||||
};
|
||||
|
||||
#endif //_SHAREDMEM_H_
|
||||
@ -1,272 +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.
|
||||
*/
|
||||
|
||||
/* This sample is a templatized version of the template project.
|
||||
* It also shows how to correctly templatize dynamically allocated shared
|
||||
* memory arrays.
|
||||
* Host code.
|
||||
*/
|
||||
|
||||
// System includes
|
||||
#include <stdio.h>
|
||||
#include <assert.h>
|
||||
#include <string.h>
|
||||
#include <math.h>
|
||||
|
||||
// CUDA runtime
|
||||
#include <cuda_runtime.h>
|
||||
|
||||
// helper functions and utilities to work with CUDA
|
||||
#include <helper_functions.h>
|
||||
#include <nvrtc_helper.h>
|
||||
|
||||
#ifndef MAX
|
||||
#define MAX(a, b) (a > b ? a : b)
|
||||
#endif
|
||||
|
||||
int g_TotalFailures = 0;
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// declaration, forward
|
||||
|
||||
template <class T>
|
||||
void runTest(int argc, char **argv, int len);
|
||||
|
||||
template <class T>
|
||||
void computeGold(T *reference, T *idata, const unsigned int len) {
|
||||
const T T_len = static_cast<T>(len);
|
||||
|
||||
for (unsigned int i = 0; i < len; ++i) {
|
||||
reference[i] = idata[i] * T_len;
|
||||
}
|
||||
}
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Program main
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
int main(int argc, char **argv) {
|
||||
printf("> runTest<float,32>\n");
|
||||
|
||||
runTest<float>(argc, argv, 32);
|
||||
|
||||
printf("> runTest<int,64>\n");
|
||||
|
||||
runTest<int>(argc, argv, 64);
|
||||
|
||||
printf("\n[simpleTemplates_nvrtc] -> Test Results: %d Failures\n",
|
||||
g_TotalFailures);
|
||||
|
||||
exit(g_TotalFailures == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
|
||||
}
|
||||
|
||||
// To completely templatize runTest (below) with cutil, we need to use
|
||||
// template specialization to wrap up CUTIL's array comparison and file writing
|
||||
// functions for different types.
|
||||
|
||||
// Here's the generic wrapper for cutCompare*
|
||||
template <class T>
|
||||
class ArrayComparator {
|
||||
public:
|
||||
bool compare(const T *reference, T *data, unsigned int len) {
|
||||
fprintf(stderr,
|
||||
"Error: no comparison function implemented for this type\n");
|
||||
return false;
|
||||
}
|
||||
};
|
||||
|
||||
// Here's the specialization for ints:
|
||||
template <>
|
||||
class ArrayComparator<int> {
|
||||
public:
|
||||
bool compare(const int *reference, int *data, unsigned int len) {
|
||||
return compareData(reference, data, len, 0.15f, 0.0f);
|
||||
}
|
||||
};
|
||||
|
||||
// Here's the specialization for floats:
|
||||
template <>
|
||||
class ArrayComparator<float> {
|
||||
public:
|
||||
bool compare(const float *reference, float *data, unsigned int len) {
|
||||
return compareData(reference, data, len, 0.15f, 0.15f);
|
||||
}
|
||||
};
|
||||
|
||||
// Here's the generic wrapper for cutWriteFile*
|
||||
template <class T>
|
||||
class ArrayFileWriter {
|
||||
public:
|
||||
bool write(const char *filename, T *data, unsigned int len, float epsilon) {
|
||||
fprintf(stderr,
|
||||
"Error: no file write function implemented for this type\n");
|
||||
return false;
|
||||
}
|
||||
};
|
||||
|
||||
// Here's the specialization for ints:
|
||||
template <>
|
||||
class ArrayFileWriter<int> {
|
||||
public:
|
||||
bool write(const char *filename, int *data, unsigned int len, float epsilon) {
|
||||
return sdkWriteFile(filename, data, len, epsilon, false);
|
||||
}
|
||||
};
|
||||
|
||||
// Here's the specialization for floats:
|
||||
template <>
|
||||
class ArrayFileWriter<float> {
|
||||
public:
|
||||
bool write(const char *filename, float *data, unsigned int len,
|
||||
float epsilon) {
|
||||
return sdkWriteFile(filename, data, len, epsilon, false);
|
||||
}
|
||||
};
|
||||
|
||||
template <typename T>
|
||||
CUfunction getKernel(CUmodule in);
|
||||
|
||||
template <>
|
||||
CUfunction getKernel<int>(CUmodule in) {
|
||||
CUfunction kernel_addr;
|
||||
checkCudaErrors(cuModuleGetFunction(&kernel_addr, in, "testInt"));
|
||||
|
||||
return kernel_addr;
|
||||
}
|
||||
|
||||
template <>
|
||||
CUfunction getKernel<float>(CUmodule in) {
|
||||
CUfunction kernel_addr;
|
||||
checkCudaErrors(cuModuleGetFunction(&kernel_addr, in, "testFloat"));
|
||||
|
||||
return kernel_addr;
|
||||
}
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
//! Run a simple test for CUDA
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
static bool moduleLoaded = false;
|
||||
CUmodule module;
|
||||
char *cubin, *kernel_file;
|
||||
size_t cubinSize;
|
||||
|
||||
template <class T>
|
||||
void runTest(int argc, char **argv, int len) {
|
||||
if (!moduleLoaded) {
|
||||
kernel_file = sdkFindFilePath("simpleTemplates_kernel.cu", argv[0]);
|
||||
compileFileToCUBIN(kernel_file, argc, argv, &cubin, &cubinSize, 0);
|
||||
module = loadCUBIN(cubin, argc, argv);
|
||||
moduleLoaded = true;
|
||||
}
|
||||
|
||||
// create and start timer
|
||||
StopWatchInterface *timer = NULL;
|
||||
sdkCreateTimer(&timer);
|
||||
|
||||
// start the timer
|
||||
sdkStartTimer(&timer);
|
||||
|
||||
unsigned int num_threads = len;
|
||||
unsigned int mem_size = sizeof(float) * num_threads;
|
||||
|
||||
// allocate host memory
|
||||
T *h_idata = (T *)malloc(mem_size);
|
||||
|
||||
// initialize the memory
|
||||
for (unsigned int i = 0; i < num_threads; ++i) {
|
||||
h_idata[i] = (T)i;
|
||||
}
|
||||
|
||||
// allocate device memory
|
||||
CUdeviceptr d_idata;
|
||||
checkCudaErrors(cuMemAlloc(&d_idata, mem_size));
|
||||
|
||||
// copy host memory to device
|
||||
checkCudaErrors(cuMemcpyHtoD(d_idata, h_idata, mem_size));
|
||||
|
||||
// allocate device memory for result
|
||||
CUdeviceptr d_odata;
|
||||
checkCudaErrors(cuMemAlloc(&d_odata, mem_size));
|
||||
|
||||
// setup execution parameters
|
||||
dim3 grid(1, 1, 1);
|
||||
dim3 threads(num_threads, 1, 1);
|
||||
|
||||
// execute the kernel
|
||||
CUfunction kernel_addr = getKernel<T>(module);
|
||||
|
||||
void *arr[] = {(void *)&d_idata, (void *)&d_odata};
|
||||
checkCudaErrors(
|
||||
cuLaunchKernel(kernel_addr, grid.x, grid.y, grid.z, /* grid dim */
|
||||
threads.x, threads.y, threads.z, /* block dim */
|
||||
mem_size, 0, /* shared mem, stream */
|
||||
&arr[0], /* arguments */
|
||||
0));
|
||||
|
||||
// check if kernel execution generated and error
|
||||
checkCudaErrors(cuCtxSynchronize());
|
||||
|
||||
// allocate mem for the result on host side
|
||||
T *h_odata = (T *)malloc(mem_size);
|
||||
|
||||
// copy result from device to host
|
||||
checkCudaErrors(cuMemcpyDtoH(h_odata, d_odata, sizeof(T) * num_threads));
|
||||
|
||||
sdkStopTimer(&timer);
|
||||
printf("Processing time: %f (ms)\n", sdkGetTimerValue(&timer));
|
||||
sdkDeleteTimer(&timer);
|
||||
|
||||
// compute reference solution
|
||||
T *reference = (T *)malloc(mem_size);
|
||||
|
||||
computeGold<T>(reference, h_idata, num_threads);
|
||||
|
||||
ArrayComparator<T> comparator;
|
||||
ArrayFileWriter<T> writer;
|
||||
|
||||
// check result
|
||||
if (checkCmdLineFlag(argc, (const char **)argv, "regression")) {
|
||||
// write file for regression test
|
||||
writer.write("./data/regression.dat", h_odata, num_threads, 0.0f);
|
||||
} else {
|
||||
// custom output handling when no regression test running
|
||||
// in this case check if the result is equivalent to the expected solution
|
||||
bool res = comparator.compare(reference, h_odata, num_threads);
|
||||
|
||||
printf("Compare %s\n\n", (1 == res) ? "OK" : "MISMATCH");
|
||||
|
||||
g_TotalFailures += (1 != res);
|
||||
}
|
||||
|
||||
// cleanup memory
|
||||
free(h_idata);
|
||||
free(h_odata);
|
||||
free(reference);
|
||||
checkCudaErrors(cuMemFree(d_idata));
|
||||
checkCudaErrors(cuMemFree(d_odata));
|
||||
}
|
||||
@ -1,70 +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.
|
||||
*/
|
||||
|
||||
// includes, kernels
|
||||
#include "sharedmem.cuh"
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
//! Simple test kernel for device functionality
|
||||
//! @param g_idata input data in global memory
|
||||
//! @param g_odata output data in global memory
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
template <class T>
|
||||
__device__ void testKernel(T *g_idata, T *g_odata) {
|
||||
// Shared mem size is determined by the host app at run time
|
||||
SharedMemory<T> smem;
|
||||
|
||||
T *sdata = smem.getPointer();
|
||||
|
||||
// access thread id
|
||||
const unsigned int tid = threadIdx.x;
|
||||
|
||||
// access number of threads in this block
|
||||
const unsigned int num_threads = blockDim.x;
|
||||
|
||||
// read in input data from global memory
|
||||
sdata[tid] = g_idata[tid];
|
||||
|
||||
__syncthreads();
|
||||
|
||||
// perform some computations
|
||||
sdata[tid] = (T)num_threads * sdata[tid];
|
||||
|
||||
__syncthreads();
|
||||
|
||||
// write data to global memory
|
||||
g_odata[tid] = sdata[tid];
|
||||
}
|
||||
|
||||
extern "C" __global__ void testFloat(float *p1, float *p2) {
|
||||
testKernel<float>(p1, p2);
|
||||
}
|
||||
|
||||
extern "C" __global__ void testInt(int *p1, int *p2) {
|
||||
testKernel<int>(p1, p2);
|
||||
}
|
||||
Loading…
x
Reference in New Issue
Block a user