mirror of
https://github.com/NVIDIA/cuda-samples.git
synced 2026-10-23 13:18:16 +08:00
add and update samples for CUDA 11.5
This commit is contained in:
Rutwik Choughule
18
Samples/threadMigration/.vscode/c_cpp_properties.json
vendored
Normal file
18
Samples/threadMigration/.vscode/c_cpp_properties.json
vendored
Normal file
@@ -0,0 +1,18 @@
|
||||
{
|
||||
"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
|
||||
}
|
||||
7
Samples/threadMigration/.vscode/extensions.json
vendored
Normal file
7
Samples/threadMigration/.vscode/extensions.json
vendored
Normal file
@@ -0,0 +1,7 @@
|
||||
{
|
||||
"recommendations": [
|
||||
"nvidia.nsight-vscode-edition",
|
||||
"ms-vscode.cpptools",
|
||||
"ms-vscode.makefile-tools"
|
||||
]
|
||||
}
|
||||
10
Samples/threadMigration/.vscode/launch.json
vendored
Normal file
10
Samples/threadMigration/.vscode/launch.json
vendored
Normal file
@@ -0,0 +1,10 @@
|
||||
{
|
||||
"configurations": [
|
||||
{
|
||||
"name": "CUDA C++: Launch",
|
||||
"type": "cuda-gdb",
|
||||
"request": "launch",
|
||||
"program": "${workspaceFolder}/threadMigration"
|
||||
}
|
||||
]
|
||||
}
|
||||
15
Samples/threadMigration/.vscode/tasks.json
vendored
Normal file
15
Samples/threadMigration/.vscode/tasks.json
vendored
Normal file
@@ -0,0 +1,15 @@
|
||||
{
|
||||
"version": "2.0.0",
|
||||
"tasks": [
|
||||
{
|
||||
"label": "sample",
|
||||
"type": "shell",
|
||||
"command": "make dbg=1",
|
||||
"problemMatcher": ["$nvcc"],
|
||||
"group": {
|
||||
"kind": "build",
|
||||
"isDefault": true
|
||||
}
|
||||
}
|
||||
]
|
||||
}
|
||||
434
Samples/threadMigration/Makefile
Normal file
434
Samples/threadMigration/Makefile
Normal file
@@ -0,0 +1,434 @@
|
||||
################################################################################
|
||||
# Copyright (c) 2019, 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 $(TARGET_OS) for TARGET_OS!)
|
||||
endif
|
||||
|
||||
# host compiler
|
||||
ifeq ($(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
|
||||
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 += --qpp-config 5.4.0,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
|
||||
|
||||
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 :=
|
||||
|
||||
################################################################################
|
||||
|
||||
FATBIN_FILE := threadMigration_kernel${TARGET_SIZE}.fatbin
|
||||
|
||||
#Detect if installed version of GCC supports required C++11
|
||||
ifeq ($(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 $(GCCVERSION)` \>= 51000)
|
||||
|
||||
ifeq ($(IS_MIN_VERSION), 1)
|
||||
$(info >>> GCC Version is greater or equal to 5.1.0 <<<)
|
||||
else
|
||||
$(info >>> Waiving build. Minimum GCC version required is 5.1.0<<<)
|
||||
SAMPLE_ENABLED := 0
|
||||
endif
|
||||
endif
|
||||
|
||||
# Gencode arguments
|
||||
ifeq ($(TARGET_ARCH),$(filter $(TARGET_ARCH),armv7l aarch64))
|
||||
SMS ?= 35 37 50 52 60 61 70 72 75 80 86
|
||||
else
|
||||
SMS ?= 35 37 50 52 60 61 70 75 80 86
|
||||
endif
|
||||
|
||||
ifeq ($(GENCODE_FLAGS),)
|
||||
# Generate SASS code for each SM architecture listed in $(SMS)
|
||||
$(foreach sm,$(SMS),$(eval GENCODE_FLAGS += -gencode arch=compute_$(sm),code=sm_$(sm)))
|
||||
|
||||
ifeq ($(SMS),)
|
||||
# Generate PTX code from SM 35
|
||||
GENCODE_FLAGS += -gencode arch=compute_35,code=compute_35
|
||||
endif
|
||||
|
||||
# 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
|
||||
|
||||
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 += $(CUDA_PATH)/targets/x86_64-linux/lib/stubs
|
||||
endif
|
||||
|
||||
ifeq ($(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
|
||||
|
||||
LIBRARIES += -lcudart_static
|
||||
|
||||
ifeq ($(SAMPLE_ENABLED),0)
|
||||
EXEC ?= @echo "[@]"
|
||||
endif
|
||||
|
||||
################################################################################
|
||||
|
||||
# Target rules
|
||||
all: build
|
||||
|
||||
build: threadMigration $(FATBIN_FILE)
|
||||
|
||||
check.deps:
|
||||
ifeq ($(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
|
||||
|
||||
$(FATBIN_FILE): threadMigration_kernel.cu
|
||||
$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -fatbin $<
|
||||
$(EXEC) mkdir -p data
|
||||
$(EXEC) cp -f $@ ./data
|
||||
$(EXEC) mkdir -p ../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)
|
||||
$(EXEC) cp -f $@ ../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)
|
||||
|
||||
threadMigration.o:threadMigration.cpp
|
||||
$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $<
|
||||
|
||||
threadMigration: threadMigration.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
|
||||
$(EXEC) ./threadMigration
|
||||
|
||||
clean:
|
||||
rm -f threadMigration threadMigration.o data/$(FATBIN_FILE) $(FATBIN_FILE)
|
||||
rm -rf ../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)/threadMigration
|
||||
|
||||
rm -rf ../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)/$(FATBIN_FILE)
|
||||
|
||||
clobber: clean
|
||||
70
Samples/threadMigration/README.md
Normal file
70
Samples/threadMigration/README.md
Normal file
@@ -0,0 +1,70 @@
|
||||
# threadMigration - CUDA Context Thread Management
|
||||
|
||||
## Description
|
||||
|
||||
Simple program illustrating how to the CUDA Context Management API and uses the new CUDA 4.0 parameter passing and CUDA launch API. CUDA contexts can be created separately and attached independently to different threads.
|
||||
|
||||
## Key Concepts
|
||||
|
||||
CUDA Driver API
|
||||
|
||||
## Supported SM Architectures
|
||||
|
||||
[SM 3.5 ](https://developer.nvidia.com/cuda-gpus) [SM 3.7 ](https://developer.nvidia.com/cuda-gpus) [SM 5.0 ](https://developer.nvidia.com/cuda-gpus) [SM 5.2 ](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)
|
||||
|
||||
## Supported OSes
|
||||
|
||||
Linux, Windows
|
||||
|
||||
## Supported CPU Architecture
|
||||
|
||||
x86_64, ppc64le, armv7l
|
||||
|
||||
## CUDA APIs involved
|
||||
|
||||
### [CUDA Driver API](http://docs.nvidia.com/cuda/cuda-driver-api/index.html)
|
||||
cuCtxCreate, cuCtxDestroy, cuModuleLoad, cuModuleLoadDataEx, cuModuleGetFunction, cuLaunchKernel, cuMemcpyDtoH, cuCtxPushCurrent, cuCtxPopCurrent
|
||||
|
||||
## Prerequisites
|
||||
|
||||
Download and install the [CUDA Toolkit 11.5](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<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, 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.<br/>
|
||||
`$ make TARGET_ARCH=x86_64` <br/> `$ make TARGET_ARCH=ppc64le` <br/> `$ make TARGET_ARCH=armv7l` <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)
|
||||
|
||||
430
Samples/threadMigration/threadMigration.cpp
Normal file
430
Samples/threadMigration/threadMigration.cpp
Normal file
@@ -0,0 +1,430 @@
|
||||
/* Copyright (c) 2021, 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.
|
||||
*/
|
||||
|
||||
/******************************************************************************
|
||||
*
|
||||
* Module: threadMigration.cpp
|
||||
*
|
||||
* Description:
|
||||
* Simple sample demonstrating multi-GPU/multithread functionality using
|
||||
* the CUDA Context Management API. This API allows the a CUDA context to
|
||||
* be associated with a CPU process. A host thread may have only one device
|
||||
* context current at a time.
|
||||
*
|
||||
* Refer to the CUDA programming guide 4.5.3.3 on Context Management
|
||||
*
|
||||
******************************************************************************/
|
||||
|
||||
#define MAXTHREADS 256
|
||||
#define NUM_INTS 32
|
||||
|
||||
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
|
||||
// Windows threads use different data structures
|
||||
#include <windows.h>
|
||||
DWORD rgdwThreadIds[MAXTHREADS];
|
||||
HANDLE rghThreads[MAXTHREADS];
|
||||
CRITICAL_SECTION g_cs;
|
||||
|
||||
#define ENTERCRITICALSECTION EnterCriticalSection(&g_cs);
|
||||
#define LEAVECRITICALSECTION LeaveCriticalSection(&g_cs);
|
||||
#define STRICMP stricmp
|
||||
#else
|
||||
|
||||
// Includes POSIX thread headers for Linux thread support
|
||||
#include <pthread.h>
|
||||
#include <stdint.h>
|
||||
pthread_t rghThreads[MAXTHREADS];
|
||||
pthread_mutex_t g_mutex;
|
||||
|
||||
#define ENTERCRITICALSECTION pthread_mutex_lock(&g_mutex);
|
||||
#define LEAVECRITICALSECTION pthread_mutex_unlock(&g_mutex);
|
||||
#define STRICMP strcasecmp
|
||||
#endif
|
||||
|
||||
#include <stdlib.h>
|
||||
#include <stdio.h>
|
||||
#include <cuda.h>
|
||||
#include <cuda_runtime_api.h>
|
||||
#include <helper_cuda_drvapi.h>
|
||||
|
||||
#include <iostream>
|
||||
#include <cstring>
|
||||
|
||||
using namespace std;
|
||||
|
||||
int NumThreads;
|
||||
int ThreadLaunchCount;
|
||||
|
||||
typedef struct _CUDAContext_st {
|
||||
CUcontext hcuContext;
|
||||
CUmodule hcuModule;
|
||||
CUfunction hcuFunction;
|
||||
CUdeviceptr dptr;
|
||||
int deviceID;
|
||||
int threadNum;
|
||||
} CUDAContext;
|
||||
|
||||
CUDAContext g_ThreadParams[MAXTHREADS];
|
||||
|
||||
// define input fatbin file
|
||||
#ifndef FATBIN_FILE
|
||||
#define FATBIN_FILE "threadMigration_kernel64.fatbin"
|
||||
#endif
|
||||
|
||||
bool gbAutoQuit = false;
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// declaration, forward
|
||||
bool runTest(int argc, char **argv);
|
||||
|
||||
#define CLEANUP_ON_ERROR(dptr, hcuModule, hcuContext, status) \
|
||||
if (dptr) cuMemFree(dptr); \
|
||||
if (hcuModule) cuModuleUnload(hcuModule); \
|
||||
if (hcuContext) cuCtxDestroy(hcuContext); \
|
||||
return status;
|
||||
|
||||
#define THREAD_QUIT \
|
||||
printf("Error\n"); \
|
||||
return 0;
|
||||
|
||||
// This sample uses the Driver API interface. The CUDA context needs
|
||||
// to be setup and the CUDA module (CUBIN) is built by NVCC
|
||||
static CUresult InitCUDAContext(CUDAContext *pContext, CUdevice hcuDevice,
|
||||
int deviceID, char **argv) {
|
||||
CUcontext hcuContext = 0;
|
||||
CUmodule hcuModule = 0;
|
||||
CUfunction hcuFunction = 0;
|
||||
CUdeviceptr dptr = 0;
|
||||
|
||||
// cuCtxCreate: Function works on floating contexts and current context
|
||||
CUresult status = cuCtxCreate(&hcuContext, 0, hcuDevice);
|
||||
|
||||
if (CUDA_SUCCESS != status) {
|
||||
fprintf(stderr, "cuCtxCreate for <deviceID=%d> failed %d\n", deviceID,
|
||||
status);
|
||||
CLEANUP_ON_ERROR(dptr, hcuModule, hcuContext, status);
|
||||
}
|
||||
|
||||
status = CUDA_ERROR_INVALID_IMAGE;
|
||||
|
||||
string module_path, ptx_source;
|
||||
std::ostringstream fatbin;
|
||||
|
||||
if (!findFatbinPath(FATBIN_FILE, module_path, argv, fatbin)) {
|
||||
exit(EXIT_FAILURE);
|
||||
} else {
|
||||
printf("> initCUDA loading module: <%s>\n", module_path.c_str());
|
||||
}
|
||||
|
||||
if (!fatbin.str().size()) {
|
||||
printf("fatbin file empty. exiting..\n");
|
||||
exit(EXIT_FAILURE);
|
||||
}
|
||||
|
||||
// Create module from binary file (FATBIN)
|
||||
checkCudaErrors(cuModuleLoadData(&hcuModule, fatbin.str().c_str()));
|
||||
|
||||
status = cuModuleGetFunction(&hcuFunction, hcuModule, "kernelFunction");
|
||||
|
||||
if (CUDA_SUCCESS != status) {
|
||||
fprintf(stderr, "cuModuleGetFunction failed %d\n", status);
|
||||
CLEANUP_ON_ERROR(dptr, hcuModule, hcuContext, status);
|
||||
}
|
||||
|
||||
// Here we must release the CUDA context from the thread context
|
||||
status = cuCtxPopCurrent(NULL);
|
||||
|
||||
if (CUDA_SUCCESS != status) {
|
||||
fprintf(stderr, "cuCtxPopCurrent failed %d\n", status);
|
||||
CLEANUP_ON_ERROR(dptr, hcuModule, hcuContext, status);
|
||||
}
|
||||
|
||||
pContext->hcuContext = hcuContext;
|
||||
pContext->hcuModule = hcuModule;
|
||||
pContext->hcuFunction = hcuFunction;
|
||||
pContext->deviceID = deviceID;
|
||||
|
||||
return CUDA_SUCCESS;
|
||||
}
|
||||
|
||||
// ThreadProc launches the CUDA kernel on a CUDA context.
|
||||
// We have more than one thread that talks to a CUDA context
|
||||
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
|
||||
DWORD WINAPI ThreadProc(CUDAContext *pParams)
|
||||
#else
|
||||
void *ThreadProc(CUDAContext *pParams)
|
||||
#endif
|
||||
{
|
||||
int wrong = 0;
|
||||
int *pInt = 0;
|
||||
|
||||
printf("<CUDA Device=%d, Context=%p, Thread=%d> - ThreadProc() Launched...\n",
|
||||
pParams->deviceID, pParams->hcuContext, pParams->threadNum);
|
||||
|
||||
// cuCtxPushCurrent: Attach the caller CUDA context to the thread context.
|
||||
CUresult status = cuCtxPushCurrent(pParams->hcuContext);
|
||||
|
||||
if (CUDA_SUCCESS != status) {
|
||||
THREAD_QUIT;
|
||||
}
|
||||
checkCudaErrors(cuMemAlloc(&pParams->dptr, NUM_INTS * sizeof(int)));
|
||||
|
||||
// There are two ways to launch CUDA kernels via the Driver API.
|
||||
// In this CUDA Sample, we illustrate both ways to pass parameters
|
||||
// and specify parameters. By default we use the simpler method.
|
||||
|
||||
if (1) {
|
||||
// This is the new CUDA 4.0 API for Kernel Parameter passing and Kernel
|
||||
// Launching (simpler method)
|
||||
void *args[5] = {&pParams->dptr};
|
||||
|
||||
// new CUDA 4.0 Driver API Kernel launch call
|
||||
status = cuLaunchKernel(pParams->hcuFunction, 1, 1, 1, 32, 1, 1, 0, NULL,
|
||||
args, NULL);
|
||||
|
||||
if (CUDA_SUCCESS != status) {
|
||||
fprintf(stderr, "cuLaunch failed %d\n", status);
|
||||
THREAD_QUIT;
|
||||
}
|
||||
} else {
|
||||
// This is the new CUDA 4.0 API for Kernel Parameter passing and Kernel
|
||||
// Launching (advanced method)
|
||||
int offset = 0;
|
||||
char argBuffer[256];
|
||||
|
||||
// pass in launch parameters (not actually de-referencing CUdeviceptr).
|
||||
// CUdeviceptr is storing the value of the parameters
|
||||
*((CUdeviceptr *)&argBuffer[offset]) = pParams->dptr;
|
||||
offset += sizeof(CUdeviceptr);
|
||||
|
||||
void *kernel_launch_config[5] = {CU_LAUNCH_PARAM_BUFFER_POINTER, argBuffer,
|
||||
CU_LAUNCH_PARAM_BUFFER_SIZE, &offset,
|
||||
CU_LAUNCH_PARAM_END};
|
||||
|
||||
// new CUDA 4.0 Driver API Kernel launch call
|
||||
status = cuLaunchKernel(pParams->hcuFunction, 1, 1, 1, 32, 1, 1, 0, 0, NULL,
|
||||
(void **)&kernel_launch_config);
|
||||
|
||||
if (CUDA_SUCCESS != status) {
|
||||
fprintf(stderr, "cuLaunch failed %d\n", status);
|
||||
THREAD_QUIT;
|
||||
}
|
||||
}
|
||||
|
||||
pInt = (int *)malloc(NUM_INTS * sizeof(int));
|
||||
|
||||
if (!pInt) return 0;
|
||||
|
||||
if (CUDA_SUCCESS ==
|
||||
cuMemcpyDtoH(pInt, pParams->dptr, NUM_INTS * sizeof(int))) {
|
||||
for (int i = 0; i < NUM_INTS; i++) {
|
||||
if (pInt[i] != 32 - i) {
|
||||
printf("<CUDA Device=%d, Context=%p, Thread=%d> error [%d]=%d!\n",
|
||||
pParams->deviceID, pParams->hcuContext, pParams->threadNum, i,
|
||||
pInt[i]);
|
||||
wrong++;
|
||||
}
|
||||
}
|
||||
|
||||
ENTERCRITICALSECTION
|
||||
|
||||
if (!wrong) ThreadLaunchCount += 1;
|
||||
|
||||
LEAVECRITICALSECTION
|
||||
}
|
||||
|
||||
free(pInt);
|
||||
fflush(stdout);
|
||||
checkCudaErrors(cuMemFree(pParams->dptr));
|
||||
|
||||
// cuCtxPopCurrent: Detach the current CUDA context from the calling thread.
|
||||
checkCudaErrors(cuCtxPopCurrent(NULL));
|
||||
|
||||
printf("<CUDA Device=%d, Context=%p, Thread=%d> - ThreadProc() Finished!\n\n",
|
||||
pParams->deviceID, pParams->hcuContext, pParams->threadNum);
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
bool FinalErrorCheck(CUDAContext *pContext, int NumThreads, int deviceCount) {
|
||||
if (ThreadLaunchCount != NumThreads * deviceCount) {
|
||||
printf("<Expected=%d, Actual=%d> ThreadLaunchCounts(s)\n",
|
||||
NumThreads * deviceCount, ThreadLaunchCount);
|
||||
return false;
|
||||
} else {
|
||||
for (int iDevice = 0; iDevice < deviceCount; iDevice++) {
|
||||
// cuCtxDestroy called on current context or a floating context
|
||||
if (CUDA_SUCCESS != cuCtxDestroy(pContext[iDevice].hcuContext))
|
||||
return false;
|
||||
}
|
||||
|
||||
return true;
|
||||
}
|
||||
}
|
||||
|
||||
int main(int argc, char **argv) {
|
||||
printf("Starting threadMigration\n");
|
||||
|
||||
bool bTestResult = runTest(argc, argv);
|
||||
|
||||
exit(bTestResult ? EXIT_SUCCESS : EXIT_FAILURE);
|
||||
}
|
||||
|
||||
bool runTest(int argc, char **argv) {
|
||||
printf("[ threadMigration ] API test...\n");
|
||||
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
|
||||
InitializeCriticalSection(&g_cs);
|
||||
#else
|
||||
pthread_mutex_init(&g_mutex, NULL);
|
||||
#endif
|
||||
// By default, we will launch 2 CUDA threads for each device
|
||||
NumThreads = 2;
|
||||
|
||||
if (argc > 1) {
|
||||
// If we are doing the QAtest or automated testing, we quit without
|
||||
// prompting
|
||||
if (checkCmdLineFlag(argc, (const char **)argv, "qatest") ||
|
||||
checkCmdLineFlag(argc, (const char **)argv, "noprompt")) {
|
||||
gbAutoQuit = true;
|
||||
}
|
||||
|
||||
if (checkCmdLineFlag(argc, (const char **)argv, "numthreads")) {
|
||||
NumThreads =
|
||||
getCmdLineArgumentInt(argc, (const char **)argv, "numthreads");
|
||||
|
||||
if (NumThreads < 1 || NumThreads > 15) {
|
||||
printf(
|
||||
"Usage: \"threadMigration -n=<threads>\", <threads> ranges 1-15\n");
|
||||
return 1;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
int deviceCount;
|
||||
int hcuDevice = 0;
|
||||
CUresult status;
|
||||
status = cuInit(0);
|
||||
|
||||
if (CUDA_SUCCESS != status) return false;
|
||||
|
||||
status = cuDeviceGetCount(&deviceCount);
|
||||
|
||||
if (CUDA_SUCCESS != status) return false;
|
||||
|
||||
printf("> %d CUDA device(s), %d Thread(s)/device to launched\n\n",
|
||||
deviceCount, NumThreads);
|
||||
|
||||
if (deviceCount == 0) {
|
||||
return false;
|
||||
}
|
||||
|
||||
int ihThread = 0;
|
||||
int ThreadIndex = 0;
|
||||
|
||||
CUDAContext *pContext =
|
||||
(CUDAContext *)malloc(sizeof(CUDAContext) * deviceCount);
|
||||
|
||||
for (int iDevice = 0; iDevice < deviceCount; iDevice++) {
|
||||
char szName[256];
|
||||
status = cuDeviceGet(&hcuDevice, iDevice);
|
||||
|
||||
if (CUDA_SUCCESS != status) return false;
|
||||
|
||||
status = cuDeviceGetName(szName, 256, hcuDevice);
|
||||
|
||||
if (CUDA_SUCCESS != status) return false;
|
||||
|
||||
{
|
||||
int major = 0, minor = 0;
|
||||
checkCudaErrors(cuDeviceGetAttribute(
|
||||
&major, CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MAJOR, hcuDevice));
|
||||
checkCudaErrors(cuDeviceGetAttribute(
|
||||
&minor, CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MINOR, hcuDevice));
|
||||
int sharedMemPerBlock;
|
||||
checkCudaErrors(cuDeviceGetAttribute(
|
||||
&sharedMemPerBlock, CU_DEVICE_ATTRIBUTE_MAX_SHARED_MEMORY_PER_BLOCK,
|
||||
hcuDevice));
|
||||
int totalConstantMemory;
|
||||
checkCudaErrors(cuDeviceGetAttribute(
|
||||
&totalConstantMemory, CU_DEVICE_ATTRIBUTE_TOTAL_CONSTANT_MEMORY,
|
||||
hcuDevice));
|
||||
int regsPerBlock;
|
||||
checkCudaErrors(cuDeviceGetAttribute(
|
||||
®sPerBlock, CU_DEVICE_ATTRIBUTE_MAX_REGISTERS_PER_BLOCK,
|
||||
hcuDevice));
|
||||
int clockRate;
|
||||
checkCudaErrors(cuDeviceGetAttribute(
|
||||
&clockRate, CU_DEVICE_ATTRIBUTE_CLOCK_RATE, hcuDevice));
|
||||
printf("Device %d: \"%s\" (Compute %d.%d)\n", iDevice, szName, major,
|
||||
minor);
|
||||
printf("\tsharedMemPerBlock: %d\n", sharedMemPerBlock);
|
||||
printf("\tconstantMemory : %d\n", totalConstantMemory);
|
||||
printf("\tregsPerBlock : %d\n", regsPerBlock);
|
||||
printf("\tclockRate : %d\n", clockRate);
|
||||
printf("\n");
|
||||
}
|
||||
|
||||
if (CUDA_SUCCESS !=
|
||||
InitCUDAContext(&pContext[iDevice], hcuDevice, iDevice, argv)) {
|
||||
return FinalErrorCheck(pContext, NumThreads, deviceCount);
|
||||
} else {
|
||||
for (int iThread = 0; iThread < NumThreads; iThread++, ihThread++) {
|
||||
g_ThreadParams[ThreadIndex].hcuContext = pContext[iDevice].hcuContext;
|
||||
g_ThreadParams[ThreadIndex].hcuModule = pContext[iDevice].hcuModule;
|
||||
g_ThreadParams[ThreadIndex].hcuFunction = pContext[iDevice].hcuFunction;
|
||||
g_ThreadParams[ThreadIndex].deviceID = pContext[iDevice].deviceID;
|
||||
g_ThreadParams[ThreadIndex].threadNum = iThread;
|
||||
// Launch (NumThreads) for each CUDA context
|
||||
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
|
||||
rghThreads[ThreadIndex] = CreateThread(
|
||||
NULL, 0, (LPTHREAD_START_ROUTINE)ThreadProc,
|
||||
&g_ThreadParams[ThreadIndex], 0, &rgdwThreadIds[ThreadIndex]);
|
||||
#else // Assume we are running linux
|
||||
pthread_create(&rghThreads[ThreadIndex], NULL,
|
||||
(void *(*)(void *))ThreadProc,
|
||||
&g_ThreadParams[ThreadIndex]);
|
||||
#endif
|
||||
ThreadIndex += 1;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// Wait until all workers are done
|
||||
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
|
||||
WaitForMultipleObjects(ThreadIndex, rghThreads, TRUE, INFINITE);
|
||||
#else
|
||||
|
||||
for (int i = 0; i < ThreadIndex; i++) {
|
||||
pthread_join(rghThreads[i], NULL);
|
||||
}
|
||||
|
||||
#endif
|
||||
|
||||
bool ret_status = FinalErrorCheck(pContext, NumThreads, deviceCount);
|
||||
free(pContext);
|
||||
return ret_status;
|
||||
}
|
||||
30
Samples/threadMigration/threadMigration_kernel.cu
Normal file
30
Samples/threadMigration/threadMigration_kernel.cu
Normal file
@@ -0,0 +1,30 @@
|
||||
/* Copyright (c) 2021, 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.
|
||||
*/
|
||||
|
||||
extern "C" __global__ void kernelFunction(int *input) {
|
||||
input[threadIdx.x] = 32 - threadIdx.x;
|
||||
}
|
||||
20
Samples/threadMigration/threadMigration_vs2017.sln
Normal file
20
Samples/threadMigration/threadMigration_vs2017.sln
Normal file
@@ -0,0 +1,20 @@
|
||||
|
||||
Microsoft Visual Studio Solution File, Format Version 12.00
|
||||
# Visual Studio 2017
|
||||
Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "threadMigration", "threadMigration_vs2017.vcxproj", "{997E0757-EA74-4A4E-A0FC-47D8C8831A15}"
|
||||
EndProject
|
||||
Global
|
||||
GlobalSection(SolutionConfigurationPlatforms) = preSolution
|
||||
Debug|x64 = Debug|x64
|
||||
Release|x64 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(ProjectConfigurationPlatforms) = postSolution
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.ActiveCfg = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.Build.0 = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.ActiveCfg = Release|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.Build.0 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(SolutionProperties) = preSolution
|
||||
HideSolutionNode = FALSE
|
||||
EndGlobalSection
|
||||
EndGlobal
|
||||
117
Samples/threadMigration/threadMigration_vs2017.vcxproj
Normal file
117
Samples/threadMigration/threadMigration_vs2017.vcxproj
Normal file
@@ -0,0 +1,117 @@
|
||||
<?xml version="1.0" encoding="utf-8"?>
|
||||
<Project DefaultTargets="Build" ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
|
||||
<PropertyGroup>
|
||||
<CUDAPropsPath Condition="'$(CUDAPropsPath)'==''">$(VCTargetsPath)\BuildCustomizations</CUDAPropsPath>
|
||||
</PropertyGroup>
|
||||
<ItemGroup Label="ProjectConfigurations">
|
||||
<ProjectConfiguration Include="Debug|x64">
|
||||
<Configuration>Debug</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
<ProjectConfiguration Include="Release|x64">
|
||||
<Configuration>Release</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
</ItemGroup>
|
||||
<PropertyGroup Label="Globals">
|
||||
<ProjectGuid>{997E0757-EA74-4A4E-A0FC-47D8C8831A15}</ProjectGuid>
|
||||
<RootNamespace>threadMigration_vs2017</RootNamespace>
|
||||
<ProjectName>threadMigration</ProjectName>
|
||||
<CudaToolkitCustomDir />
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(WindowsTargetPlatformVersion)'==''">
|
||||
<LatestTargetPlatformVersion>$([Microsoft.Build.Utilities.ToolLocationHelper]::GetLatestSDKTargetPlatformVersion('Windows', '10.0'))</LatestTargetPlatformVersion>
|
||||
<WindowsTargetPlatformVersion Condition="'$(WindowsTargetPlatformVersion)' == ''">$(LatestTargetPlatformVersion)</WindowsTargetPlatformVersion>
|
||||
<TargetPlatformVersion>$(WindowsTargetPlatformVersion)</TargetPlatformVersion>
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
|
||||
<PropertyGroup>
|
||||
<ConfigurationType>Application</ConfigurationType>
|
||||
<CharacterSet>MultiByte</CharacterSet>
|
||||
<PlatformToolset>v141</PlatformToolset>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<UseDebugLibraries>true</UseDebugLibraries>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Release'">
|
||||
<WholeProgramOptimization>true</WholeProgramOptimization>
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" />
|
||||
<ImportGroup Label="ExtensionSettings">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.5.props" />
|
||||
</ImportGroup>
|
||||
<ImportGroup Label="PropertySheets">
|
||||
<Import Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" />
|
||||
</ImportGroup>
|
||||
<PropertyGroup Label="UserMacros" />
|
||||
<PropertyGroup>
|
||||
<IntDir>$(Platform)/$(Configuration)/</IntDir>
|
||||
<IncludePath>$(IncludePath)</IncludePath>
|
||||
<CodeAnalysisRuleSet>AllRules.ruleset</CodeAnalysisRuleSet>
|
||||
<CodeAnalysisRules />
|
||||
<CodeAnalysisRuleAssemblies />
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Platform)'=='x64'">
|
||||
<OutDir>../../bin/win64/$(Configuration)/</OutDir>
|
||||
</PropertyGroup>
|
||||
<ItemDefinitionGroup>
|
||||
<ClCompile>
|
||||
<WarningLevel>Level3</WarningLevel>
|
||||
<PreprocessorDefinitions>WIN32;_MBCS;%(PreprocessorDefinitions)</PreprocessorDefinitions>
|
||||
<AdditionalIncludeDirectories>./;$(CudaToolkitDir)/include;../../Common;$(CudaToolkitIncludeDir);</AdditionalIncludeDirectories>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<SubSystem>Console</SubSystem>
|
||||
<AdditionalDependencies>cuda.lib;cudart_static.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
|
||||
<AdditionalLibraryDirectories>$(CudaToolkitLibDir);</AdditionalLibraryDirectories>
|
||||
<OutputFile>$(OutDir)/threadMigration.exe</OutputFile>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<CodeGeneration>compute_35,compute_35;compute_35,sm_35;compute_37,sm_37;compute_50,sm_50;compute_52,sm_52;compute_60,sm_60;compute_61,sm_61;compute_70,sm_70;compute_75,sm_75;compute_80,sm_80;compute_86,sm_86;</CodeGeneration>
|
||||
<AdditionalOptions>-Xcompiler "/wd 4819" %(AdditionalOptions)</AdditionalOptions>
|
||||
<Include>./;../../Common</Include>
|
||||
<Defines>WIN32</Defines>
|
||||
<AdditionalOptions>--threads 0</AdditionalOptions>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<ClCompile>
|
||||
<Optimization>Disabled</Optimization>
|
||||
<RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>true</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>Default</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MTd</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Release'">
|
||||
<ClCompile>
|
||||
<Optimization>MaxSpeed</Optimization>
|
||||
<RuntimeLibrary>MultiThreaded</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>false</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>UseLinkTimeCodeGeneration</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MT</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemGroup>
|
||||
<ClCompile Include="threadMigration.cpp" />
|
||||
<CudaCompile Include="threadMigration_kernel.cu">
|
||||
<CompileOut Condition="'$(Platform)'=='x64'">data/%(Filename)64.fatbin</CompileOut>
|
||||
<NvccCompilation>fatbin</NvccCompilation>
|
||||
</CudaCompile>
|
||||
|
||||
</ItemGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
|
||||
<ImportGroup Label="ExtensionTargets">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.5.targets" />
|
||||
</ImportGroup>
|
||||
</Project>
|
||||
20
Samples/threadMigration/threadMigration_vs2019.sln
Normal file
20
Samples/threadMigration/threadMigration_vs2019.sln
Normal file
@@ -0,0 +1,20 @@
|
||||
|
||||
Microsoft Visual Studio Solution File, Format Version 12.00
|
||||
# Visual Studio 2019
|
||||
Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "threadMigration", "threadMigration_vs2019.vcxproj", "{997E0757-EA74-4A4E-A0FC-47D8C8831A15}"
|
||||
EndProject
|
||||
Global
|
||||
GlobalSection(SolutionConfigurationPlatforms) = preSolution
|
||||
Debug|x64 = Debug|x64
|
||||
Release|x64 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(ProjectConfigurationPlatforms) = postSolution
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.ActiveCfg = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.Build.0 = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.ActiveCfg = Release|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.Build.0 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(SolutionProperties) = preSolution
|
||||
HideSolutionNode = FALSE
|
||||
EndGlobalSection
|
||||
EndGlobal
|
||||
113
Samples/threadMigration/threadMigration_vs2019.vcxproj
Normal file
113
Samples/threadMigration/threadMigration_vs2019.vcxproj
Normal file
@@ -0,0 +1,113 @@
|
||||
<?xml version="1.0" encoding="utf-8"?>
|
||||
<Project DefaultTargets="Build" ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
|
||||
<PropertyGroup>
|
||||
<CUDAPropsPath Condition="'$(CUDAPropsPath)'==''">$(VCTargetsPath)\BuildCustomizations</CUDAPropsPath>
|
||||
</PropertyGroup>
|
||||
<ItemGroup Label="ProjectConfigurations">
|
||||
<ProjectConfiguration Include="Debug|x64">
|
||||
<Configuration>Debug</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
<ProjectConfiguration Include="Release|x64">
|
||||
<Configuration>Release</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
</ItemGroup>
|
||||
<PropertyGroup Label="Globals">
|
||||
<ProjectGuid>{997E0757-EA74-4A4E-A0FC-47D8C8831A15}</ProjectGuid>
|
||||
<RootNamespace>threadMigration_vs2019</RootNamespace>
|
||||
<ProjectName>threadMigration</ProjectName>
|
||||
<CudaToolkitCustomDir />
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
|
||||
<PropertyGroup>
|
||||
<ConfigurationType>Application</ConfigurationType>
|
||||
<CharacterSet>MultiByte</CharacterSet>
|
||||
<PlatformToolset>v142</PlatformToolset>
|
||||
<WindowsTargetPlatformVersion>10.0</WindowsTargetPlatformVersion>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<UseDebugLibraries>true</UseDebugLibraries>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Release'">
|
||||
<WholeProgramOptimization>true</WholeProgramOptimization>
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" />
|
||||
<ImportGroup Label="ExtensionSettings">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.5.props" />
|
||||
</ImportGroup>
|
||||
<ImportGroup Label="PropertySheets">
|
||||
<Import Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" />
|
||||
</ImportGroup>
|
||||
<PropertyGroup Label="UserMacros" />
|
||||
<PropertyGroup>
|
||||
<IntDir>$(Platform)/$(Configuration)/</IntDir>
|
||||
<IncludePath>$(IncludePath)</IncludePath>
|
||||
<CodeAnalysisRuleSet>AllRules.ruleset</CodeAnalysisRuleSet>
|
||||
<CodeAnalysisRules />
|
||||
<CodeAnalysisRuleAssemblies />
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Platform)'=='x64'">
|
||||
<OutDir>../../bin/win64/$(Configuration)/</OutDir>
|
||||
</PropertyGroup>
|
||||
<ItemDefinitionGroup>
|
||||
<ClCompile>
|
||||
<WarningLevel>Level3</WarningLevel>
|
||||
<PreprocessorDefinitions>WIN32;_MBCS;%(PreprocessorDefinitions)</PreprocessorDefinitions>
|
||||
<AdditionalIncludeDirectories>./;$(CudaToolkitDir)/include;../../Common;$(CudaToolkitIncludeDir);</AdditionalIncludeDirectories>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<SubSystem>Console</SubSystem>
|
||||
<AdditionalDependencies>cuda.lib;cudart_static.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
|
||||
<AdditionalLibraryDirectories>$(CudaToolkitLibDir);</AdditionalLibraryDirectories>
|
||||
<OutputFile>$(OutDir)/threadMigration.exe</OutputFile>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<CodeGeneration>compute_35,compute_35;compute_35,sm_35;compute_37,sm_37;compute_50,sm_50;compute_52,sm_52;compute_60,sm_60;compute_61,sm_61;compute_70,sm_70;compute_75,sm_75;compute_80,sm_80;compute_86,sm_86;</CodeGeneration>
|
||||
<AdditionalOptions>-Xcompiler "/wd 4819" %(AdditionalOptions)</AdditionalOptions>
|
||||
<Include>./;../../Common</Include>
|
||||
<Defines>WIN32</Defines>
|
||||
<AdditionalOptions>--threads 0</AdditionalOptions>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<ClCompile>
|
||||
<Optimization>Disabled</Optimization>
|
||||
<RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>true</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>Default</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MTd</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Release'">
|
||||
<ClCompile>
|
||||
<Optimization>MaxSpeed</Optimization>
|
||||
<RuntimeLibrary>MultiThreaded</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>false</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>UseLinkTimeCodeGeneration</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MT</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemGroup>
|
||||
<ClCompile Include="threadMigration.cpp" />
|
||||
<CudaCompile Include="threadMigration_kernel.cu">
|
||||
<CompileOut Condition="'$(Platform)'=='x64'">data/%(Filename)64.fatbin</CompileOut>
|
||||
<NvccCompilation>fatbin</NvccCompilation>
|
||||
</CudaCompile>
|
||||
|
||||
</ItemGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
|
||||
<ImportGroup Label="ExtensionTargets">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.5.targets" />
|
||||
</ImportGroup>
|
||||
</Project>
|
||||
Reference in New Issue
Block a user