mirror of
https://github.com/NVIDIA/cuda-samples.git
synced 2025-04-23 00:01:23 +01:00
Update template, simpleVoteIntrinsics - remove simpleVoteIntrinsics_nvrtc
This commit is contained in:
Rob Armstrong
parent
56ad17f97c
commit
a14d734664
@ -10,6 +10,7 @@
|
|||||||
* `cppOverload` demonstrating C++ function overloading (reason: obsolete)
|
* `cppOverload` demonstrating C++ function overloading (reason: obsolete)
|
||||||
* `simpleSeparateCompilation` demonstrating NVCC compilation to a static library (reason: trivial)
|
* `simpleSeparateCompilation` demonstrating NVCC compilation to a static library (reason: trivial)
|
||||||
* `simpleTemplates_nvrtc` demonstrating NVRTC usage for `simpleTemplates` sample (reason: redundant)
|
* `simpleTemplates_nvrtc` demonstrating NVRTC usage for `simpleTemplates` sample (reason: redundant)
|
||||||
|
* `simpleVoteIntrinsics_nvrtc` demonstrating NVRTC usage for `simpleVoteIntrinsics` sample (reason: redundant)
|
||||||
|
|
||||||
### CUDA 12.5
|
### CUDA 12.5
|
||||||
|
|
||||||
|
|||||||
@ -36,8 +36,7 @@ add_subdirectory(simpleTemplates)
|
|||||||
add_subdirectory(simpleTexture)
|
add_subdirectory(simpleTexture)
|
||||||
add_subdirectory(simpleTexture3D)
|
add_subdirectory(simpleTexture3D)
|
||||||
add_subdirectory(simpleTextureDrv)
|
add_subdirectory(simpleTextureDrv)
|
||||||
#add_subdirectory(simpleVoteIntrinsics)
|
add_subdirectory(simpleVoteIntrinsics)
|
||||||
#add_subdirectory(simpleVoteIntrinsics_nvrtc)
|
|
||||||
add_subdirectory(simpleZeroCopy)
|
add_subdirectory(simpleZeroCopy)
|
||||||
add_subdirectory(systemWideAtomics)
|
add_subdirectory(systemWideAtomics)
|
||||||
#add_subdirectory(template)
|
#add_subdirectory(template)
|
||||||
|
|||||||
11
Samples/0_Introduction/simpleVoteIntrinsics/CMakeLists.txt
Normal file
11
Samples/0_Introduction/simpleVoteIntrinsics/CMakeLists.txt
Normal file
@ -0,0 +1,11 @@
|
|||||||
|
# Include directories and libraries
|
||||||
|
include_directories(../../../Common)
|
||||||
|
|
||||||
|
# Source file
|
||||||
|
set(SRC_FILES
|
||||||
|
simpleVoteIntrinsics.cu
|
||||||
|
)
|
||||||
|
|
||||||
|
# Add target for simpleVoteIntrinsics
|
||||||
|
add_executable(simpleVoteIntrinsics ${SRC_FILES})
|
||||||
|
set_target_properties(simpleVoteIntrinsics PROPERTIES CUDA_SEPARABLE_COMPILATION ON)
|
||||||
@ -1,357 +0,0 @@
|
|||||||
################################################################################
|
|
||||||
# Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
|
|
||||||
#
|
|
||||||
# Redistribution and use in source and binary forms, with or without
|
|
||||||
# modification, are permitted provided that the following conditions
|
|
||||||
# are met:
|
|
||||||
# * Redistributions of source code must retain the above copyright
|
|
||||||
# notice, this list of conditions and the following disclaimer.
|
|
||||||
# * Redistributions in binary form must reproduce the above copyright
|
|
||||||
# notice, this list of conditions and the following disclaimer in the
|
|
||||||
# documentation and/or other materials provided with the distribution.
|
|
||||||
# * Neither the name of NVIDIA CORPORATION nor the names of its
|
|
||||||
# contributors may be used to endorse or promote products derived
|
|
||||||
# from this software without specific prior written permission.
|
|
||||||
#
|
|
||||||
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
|
||||||
# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
||||||
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
|
||||||
# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
|
||||||
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
|
||||||
# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
|
||||||
# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
|
||||||
# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
|
||||||
# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
|
||||||
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
|
||||||
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
||||||
#
|
|
||||||
################################################################################
|
|
||||||
#
|
|
||||||
# Makefile project only supported on Mac OS X and Linux Platforms)
|
|
||||||
#
|
|
||||||
################################################################################
|
|
||||||
|
|
||||||
# Location of the CUDA Toolkit
|
|
||||||
CUDA_PATH ?= /usr/local/cuda
|
|
||||||
|
|
||||||
##############################
|
|
||||||
# start deprecated interface #
|
|
||||||
##############################
|
|
||||||
ifeq ($(x86_64),1)
|
|
||||||
$(info WARNING - x86_64 variable has been deprecated)
|
|
||||||
$(info WARNING - please use TARGET_ARCH=x86_64 instead)
|
|
||||||
TARGET_ARCH ?= x86_64
|
|
||||||
endif
|
|
||||||
ifeq ($(ARMv7),1)
|
|
||||||
$(info WARNING - ARMv7 variable has been deprecated)
|
|
||||||
$(info WARNING - please use TARGET_ARCH=armv7l instead)
|
|
||||||
TARGET_ARCH ?= armv7l
|
|
||||||
endif
|
|
||||||
ifeq ($(aarch64),1)
|
|
||||||
$(info WARNING - aarch64 variable has been deprecated)
|
|
||||||
$(info WARNING - please use TARGET_ARCH=aarch64 instead)
|
|
||||||
TARGET_ARCH ?= aarch64
|
|
||||||
endif
|
|
||||||
ifeq ($(ppc64le),1)
|
|
||||||
$(info WARNING - ppc64le variable has been deprecated)
|
|
||||||
$(info WARNING - please use TARGET_ARCH=ppc64le instead)
|
|
||||||
TARGET_ARCH ?= ppc64le
|
|
||||||
endif
|
|
||||||
ifneq ($(GCC),)
|
|
||||||
$(info WARNING - GCC variable has been deprecated)
|
|
||||||
$(info WARNING - please use HOST_COMPILER=$(GCC) instead)
|
|
||||||
HOST_COMPILER ?= $(GCC)
|
|
||||||
endif
|
|
||||||
ifneq ($(abi),)
|
|
||||||
$(error ERROR - abi variable has been removed)
|
|
||||||
endif
|
|
||||||
############################
|
|
||||||
# end deprecated interface #
|
|
||||||
############################
|
|
||||||
|
|
||||||
# architecture
|
|
||||||
HOST_ARCH := $(shell uname -m)
|
|
||||||
TARGET_ARCH ?= $(HOST_ARCH)
|
|
||||||
ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 sbsa ppc64le armv7l))
|
|
||||||
ifneq ($(TARGET_ARCH),$(HOST_ARCH))
|
|
||||||
ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 sbsa ppc64le))
|
|
||||||
TARGET_SIZE := 64
|
|
||||||
else ifneq (,$(filter $(TARGET_ARCH),armv7l))
|
|
||||||
TARGET_SIZE := 32
|
|
||||||
endif
|
|
||||||
else
|
|
||||||
TARGET_SIZE := $(shell getconf LONG_BIT)
|
|
||||||
endif
|
|
||||||
else
|
|
||||||
$(error ERROR - unsupported value $(TARGET_ARCH) for TARGET_ARCH!)
|
|
||||||
endif
|
|
||||||
|
|
||||||
# sbsa and aarch64 systems look similar. Need to differentiate them at host level for now.
|
|
||||||
ifeq ($(HOST_ARCH),aarch64)
|
|
||||||
ifeq ($(CUDA_PATH)/targets/sbsa-linux,$(shell ls -1d $(CUDA_PATH)/targets/sbsa-linux 2>/dev/null))
|
|
||||||
HOST_ARCH := sbsa
|
|
||||||
TARGET_ARCH := sbsa
|
|
||||||
endif
|
|
||||||
endif
|
|
||||||
|
|
||||||
ifneq ($(TARGET_ARCH),$(HOST_ARCH))
|
|
||||||
ifeq (,$(filter $(HOST_ARCH)-$(TARGET_ARCH),aarch64-armv7l x86_64-armv7l x86_64-aarch64 x86_64-sbsa x86_64-ppc64le))
|
|
||||||
$(error ERROR - cross compiling from $(HOST_ARCH) to $(TARGET_ARCH) is not supported!)
|
|
||||||
endif
|
|
||||||
endif
|
|
||||||
|
|
||||||
# When on native aarch64 system with userspace of 32-bit, change TARGET_ARCH to armv7l
|
|
||||||
ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_SIZE),aarch64-aarch64-32)
|
|
||||||
TARGET_ARCH = armv7l
|
|
||||||
endif
|
|
||||||
|
|
||||||
# operating system
|
|
||||||
HOST_OS := $(shell uname -s 2>/dev/null | tr "[:upper:]" "[:lower:]")
|
|
||||||
TARGET_OS ?= $(HOST_OS)
|
|
||||||
ifeq (,$(filter $(TARGET_OS),linux darwin qnx android))
|
|
||||||
$(error ERROR - unsupported value $(TARGET_OS) for TARGET_OS!)
|
|
||||||
endif
|
|
||||||
|
|
||||||
# host compiler
|
|
||||||
ifdef HOST_COMPILER
|
|
||||||
CUSTOM_HOST_COMPILER = 1
|
|
||||||
endif
|
|
||||||
|
|
||||||
ifeq ($(TARGET_OS),darwin)
|
|
||||||
ifeq ($(shell expr `xcodebuild -version | grep -i xcode | awk '{print $$2}' | cut -d'.' -f1` \>= 5),1)
|
|
||||||
HOST_COMPILER ?= clang++
|
|
||||||
endif
|
|
||||||
else ifneq ($(TARGET_ARCH),$(HOST_ARCH))
|
|
||||||
ifeq ($(HOST_ARCH)-$(TARGET_ARCH),x86_64-armv7l)
|
|
||||||
ifeq ($(TARGET_OS),linux)
|
|
||||||
HOST_COMPILER ?= arm-linux-gnueabihf-g++
|
|
||||||
else ifeq ($(TARGET_OS),qnx)
|
|
||||||
ifeq ($(QNX_HOST),)
|
|
||||||
$(error ERROR - QNX_HOST must be passed to the QNX host toolchain)
|
|
||||||
endif
|
|
||||||
ifeq ($(QNX_TARGET),)
|
|
||||||
$(error ERROR - QNX_TARGET must be passed to the QNX target toolchain)
|
|
||||||
endif
|
|
||||||
export QNX_HOST
|
|
||||||
export QNX_TARGET
|
|
||||||
HOST_COMPILER ?= $(QNX_HOST)/usr/bin/arm-unknown-nto-qnx6.6.0eabi-g++
|
|
||||||
else ifeq ($(TARGET_OS),android)
|
|
||||||
HOST_COMPILER ?= arm-linux-androideabi-g++
|
|
||||||
endif
|
|
||||||
else ifeq ($(TARGET_ARCH),aarch64)
|
|
||||||
ifeq ($(TARGET_OS), linux)
|
|
||||||
HOST_COMPILER ?= aarch64-linux-gnu-g++
|
|
||||||
else ifeq ($(TARGET_OS),qnx)
|
|
||||||
ifeq ($(QNX_HOST),)
|
|
||||||
$(error ERROR - QNX_HOST must be passed to the QNX host toolchain)
|
|
||||||
endif
|
|
||||||
ifeq ($(QNX_TARGET),)
|
|
||||||
$(error ERROR - QNX_TARGET must be passed to the QNX target toolchain)
|
|
||||||
endif
|
|
||||||
export QNX_HOST
|
|
||||||
export QNX_TARGET
|
|
||||||
HOST_COMPILER ?= $(QNX_HOST)/usr/bin/q++
|
|
||||||
else ifeq ($(TARGET_OS), android)
|
|
||||||
HOST_COMPILER ?= aarch64-linux-android-clang++
|
|
||||||
endif
|
|
||||||
else ifeq ($(TARGET_ARCH),sbsa)
|
|
||||||
HOST_COMPILER ?= aarch64-linux-gnu-g++
|
|
||||||
else ifeq ($(TARGET_ARCH),ppc64le)
|
|
||||||
HOST_COMPILER ?= powerpc64le-linux-gnu-g++
|
|
||||||
endif
|
|
||||||
endif
|
|
||||||
HOST_COMPILER ?= g++
|
|
||||||
NVCC := $(CUDA_PATH)/bin/nvcc -ccbin $(HOST_COMPILER)
|
|
||||||
|
|
||||||
# internal flags
|
|
||||||
NVCCFLAGS := -m${TARGET_SIZE}
|
|
||||||
CCFLAGS :=
|
|
||||||
LDFLAGS :=
|
|
||||||
|
|
||||||
# build flags
|
|
||||||
|
|
||||||
# Link flag for customized HOST_COMPILER with gcc realpath
|
|
||||||
GCC_PATH := $(shell which gcc)
|
|
||||||
ifeq ($(CUSTOM_HOST_COMPILER),1)
|
|
||||||
ifneq ($(filter /%,$(HOST_COMPILER)),)
|
|
||||||
ifneq ($(findstring gcc,$(HOST_COMPILER)),)
|
|
||||||
ifneq ($(GCC_PATH),$(HOST_COMPILER))
|
|
||||||
LDFLAGS += -lstdc++
|
|
||||||
endif
|
|
||||||
endif
|
|
||||||
endif
|
|
||||||
endif
|
|
||||||
|
|
||||||
ifeq ($(TARGET_OS),darwin)
|
|
||||||
LDFLAGS += -rpath $(CUDA_PATH)/lib
|
|
||||||
CCFLAGS += -arch $(HOST_ARCH)
|
|
||||||
else ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_OS),x86_64-armv7l-linux)
|
|
||||||
LDFLAGS += --dynamic-linker=/lib/ld-linux-armhf.so.3
|
|
||||||
CCFLAGS += -mfloat-abi=hard
|
|
||||||
else ifeq ($(TARGET_OS),android)
|
|
||||||
LDFLAGS += -pie
|
|
||||||
CCFLAGS += -fpie -fpic -fexceptions
|
|
||||||
endif
|
|
||||||
|
|
||||||
ifneq ($(TARGET_ARCH),$(HOST_ARCH))
|
|
||||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-linux)
|
|
||||||
ifneq ($(TARGET_FS),)
|
|
||||||
GCCVERSIONLTEQ46 := $(shell expr `$(HOST_COMPILER) -dumpversion` \<= 4.6)
|
|
||||||
ifeq ($(GCCVERSIONLTEQ46),1)
|
|
||||||
CCFLAGS += --sysroot=$(TARGET_FS)
|
|
||||||
endif
|
|
||||||
LDFLAGS += --sysroot=$(TARGET_FS)
|
|
||||||
LDFLAGS += -rpath-link=$(TARGET_FS)/lib
|
|
||||||
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib
|
|
||||||
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib/arm-linux-gnueabihf
|
|
||||||
endif
|
|
||||||
endif
|
|
||||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-linux)
|
|
||||||
ifneq ($(TARGET_FS),)
|
|
||||||
GCCVERSIONLTEQ46 := $(shell expr `$(HOST_COMPILER) -dumpversion` \<= 4.6)
|
|
||||||
ifeq ($(GCCVERSIONLTEQ46),1)
|
|
||||||
CCFLAGS += --sysroot=$(TARGET_FS)
|
|
||||||
endif
|
|
||||||
LDFLAGS += --sysroot=$(TARGET_FS)
|
|
||||||
LDFLAGS += -rpath-link=$(TARGET_FS)/lib -L$(TARGET_FS)/lib
|
|
||||||
LDFLAGS += -rpath-link=$(TARGET_FS)/lib/aarch64-linux-gnu -L$(TARGET_FS)/lib/aarch64-linux-gnu
|
|
||||||
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib -L$(TARGET_FS)/usr/lib
|
|
||||||
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib/aarch64-linux-gnu -L$(TARGET_FS)/usr/lib/aarch64-linux-gnu
|
|
||||||
LDFLAGS += --unresolved-symbols=ignore-in-shared-libs
|
|
||||||
CCFLAGS += -isystem=$(TARGET_FS)/usr/include -I$(TARGET_FS)/usr/include -I$(TARGET_FS)/usr/include/libdrm
|
|
||||||
CCFLAGS += -isystem=$(TARGET_FS)/usr/include/aarch64-linux-gnu -I$(TARGET_FS)/usr/include/aarch64-linux-gnu
|
|
||||||
endif
|
|
||||||
endif
|
|
||||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-qnx)
|
|
||||||
NVCCFLAGS += -D_QNX_SOURCE
|
|
||||||
NVCCFLAGS += --qpp-config 8.3.0,gcc_ntoaarch64le
|
|
||||||
CCFLAGS += -DWIN_INTERFACE_CUSTOM -I/usr/include/aarch64-qnx-gnu
|
|
||||||
LDFLAGS += -lsocket
|
|
||||||
LDFLAGS += -L/usr/lib/aarch64-qnx-gnu
|
|
||||||
CCFLAGS += "-Wl\,-rpath-link\,/usr/lib/aarch64-qnx-gnu"
|
|
||||||
ifdef TARGET_OVERRIDE
|
|
||||||
LDFLAGS += -lslog2
|
|
||||||
endif
|
|
||||||
|
|
||||||
ifneq ($(TARGET_FS),)
|
|
||||||
LDFLAGS += -L$(TARGET_FS)/usr/lib
|
|
||||||
CCFLAGS += "-Wl\,-rpath-link\,$(TARGET_FS)/usr/lib"
|
|
||||||
LDFLAGS += -L$(TARGET_FS)/usr/libnvidia
|
|
||||||
CCFLAGS += "-Wl\,-rpath-link\,$(TARGET_FS)/usr/libnvidia"
|
|
||||||
CCFLAGS += -I$(TARGET_FS)/../include
|
|
||||||
endif
|
|
||||||
endif
|
|
||||||
endif
|
|
||||||
|
|
||||||
ifdef TARGET_OVERRIDE # cuda toolkit targets override
|
|
||||||
NVCCFLAGS += -target-dir $(TARGET_OVERRIDE)
|
|
||||||
endif
|
|
||||||
|
|
||||||
# Install directory of different arch
|
|
||||||
CUDA_INSTALL_TARGET_DIR :=
|
|
||||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-linux)
|
|
||||||
CUDA_INSTALL_TARGET_DIR = targets/armv7-linux-gnueabihf/
|
|
||||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-linux)
|
|
||||||
CUDA_INSTALL_TARGET_DIR = targets/aarch64-linux/
|
|
||||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),sbsa-linux)
|
|
||||||
CUDA_INSTALL_TARGET_DIR = targets/sbsa-linux/
|
|
||||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-android)
|
|
||||||
CUDA_INSTALL_TARGET_DIR = targets/armv7-linux-androideabi/
|
|
||||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-android)
|
|
||||||
CUDA_INSTALL_TARGET_DIR = targets/aarch64-linux-androideabi/
|
|
||||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-qnx)
|
|
||||||
CUDA_INSTALL_TARGET_DIR = targets/ARMv7-linux-QNX/
|
|
||||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-qnx)
|
|
||||||
CUDA_INSTALL_TARGET_DIR = targets/aarch64-qnx/
|
|
||||||
else ifeq ($(TARGET_ARCH),ppc64le)
|
|
||||||
CUDA_INSTALL_TARGET_DIR = targets/ppc64le-linux/
|
|
||||||
endif
|
|
||||||
|
|
||||||
# Debug build flags
|
|
||||||
ifeq ($(dbg),1)
|
|
||||||
NVCCFLAGS += -g -G
|
|
||||||
BUILD_TYPE := debug
|
|
||||||
else
|
|
||||||
BUILD_TYPE := release
|
|
||||||
endif
|
|
||||||
|
|
||||||
ALL_CCFLAGS :=
|
|
||||||
ALL_CCFLAGS += $(NVCCFLAGS)
|
|
||||||
ALL_CCFLAGS += $(EXTRA_NVCCFLAGS)
|
|
||||||
ALL_CCFLAGS += $(addprefix -Xcompiler ,$(CCFLAGS))
|
|
||||||
ALL_CCFLAGS += $(addprefix -Xcompiler ,$(EXTRA_CCFLAGS))
|
|
||||||
|
|
||||||
SAMPLE_ENABLED := 1
|
|
||||||
|
|
||||||
ALL_LDFLAGS :=
|
|
||||||
ALL_LDFLAGS += $(ALL_CCFLAGS)
|
|
||||||
ALL_LDFLAGS += $(addprefix -Xlinker ,$(LDFLAGS))
|
|
||||||
ALL_LDFLAGS += $(addprefix -Xlinker ,$(EXTRA_LDFLAGS))
|
|
||||||
|
|
||||||
# Common includes and paths for CUDA
|
|
||||||
INCLUDES := -I../../../Common
|
|
||||||
LIBRARIES :=
|
|
||||||
|
|
||||||
################################################################################
|
|
||||||
|
|
||||||
# Gencode arguments
|
|
||||||
ifeq ($(TARGET_ARCH),$(filter $(TARGET_ARCH),armv7l aarch64 sbsa))
|
|
||||||
SMS ?= 53 61 70 72 75 80 86 87 90
|
|
||||||
else
|
|
||||||
SMS ?= 50 52 60 61 70 75 80 86 89 90
|
|
||||||
endif
|
|
||||||
|
|
||||||
ifeq ($(SMS),)
|
|
||||||
$(info >>> WARNING - no SM architectures have been specified - waiving sample <<<)
|
|
||||||
SAMPLE_ENABLED := 0
|
|
||||||
endif
|
|
||||||
|
|
||||||
ifeq ($(GENCODE_FLAGS),)
|
|
||||||
# Generate SASS code for each SM architecture listed in $(SMS)
|
|
||||||
$(foreach sm,$(SMS),$(eval GENCODE_FLAGS += -gencode arch=compute_$(sm),code=sm_$(sm)))
|
|
||||||
|
|
||||||
# Generate PTX code from the highest SM architecture in $(SMS) to guarantee forward-compatibility
|
|
||||||
HIGHEST_SM := $(lastword $(sort $(SMS)))
|
|
||||||
ifneq ($(HIGHEST_SM),)
|
|
||||||
GENCODE_FLAGS += -gencode arch=compute_$(HIGHEST_SM),code=compute_$(HIGHEST_SM)
|
|
||||||
endif
|
|
||||||
endif
|
|
||||||
|
|
||||||
ALL_CCFLAGS += --threads 0 --std=c++11
|
|
||||||
|
|
||||||
ifeq ($(SAMPLE_ENABLED),0)
|
|
||||||
EXEC ?= @echo "[@]"
|
|
||||||
endif
|
|
||||||
|
|
||||||
################################################################################
|
|
||||||
|
|
||||||
# Target rules
|
|
||||||
all: build
|
|
||||||
|
|
||||||
build: simpleVoteIntrinsics
|
|
||||||
|
|
||||||
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
|
|
||||||
|
|
||||||
simpleVoteIntrinsics.o:simpleVoteIntrinsics.cu
|
|
||||||
$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $<
|
|
||||||
|
|
||||||
simpleVoteIntrinsics: simpleVoteIntrinsics.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) ./simpleVoteIntrinsics
|
|
||||||
|
|
||||||
testrun: build
|
|
||||||
|
|
||||||
clean:
|
|
||||||
rm -f simpleVoteIntrinsics simpleVoteIntrinsics.o
|
|
||||||
rm -rf ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)/simpleVoteIntrinsics
|
|
||||||
|
|
||||||
clobber: clean
|
|
||||||
@ -1,80 +0,0 @@
|
|||||||
<?xml version="1.0" encoding="UTF-8"?>
|
|
||||||
<!DOCTYPE entry SYSTEM "SamplesInfo.dtd">
|
|
||||||
<entry>
|
|
||||||
<name>simpleVoteIntrinsics</name>
|
|
||||||
<cuda_api_list>
|
|
||||||
<toolkit>cudaMemcpy</toolkit>
|
|
||||||
<toolkit>cudaFree</toolkit>
|
|
||||||
<toolkit>cudaDeviceSynchronize</toolkit>
|
|
||||||
<toolkit>cudaMalloc</toolkit>
|
|
||||||
<toolkit>cudaGetDeviceProperties</toolkit>
|
|
||||||
</cuda_api_list>
|
|
||||||
<description><![CDATA[Simple program which demonstrates how to use the Vote (__any_sync, __all_sync) intrinsic instruction in a CUDA kernel.]]></description>
|
|
||||||
<devicecompilation>whole</devicecompilation>
|
|
||||||
<includepaths>
|
|
||||||
<path>./</path>
|
|
||||||
<path>../</path>
|
|
||||||
<path>../../../Common</path>
|
|
||||||
</includepaths>
|
|
||||||
<keyconcepts>
|
|
||||||
<concept level="basic">Vote Intrinsics</concept>
|
|
||||||
</keyconcepts>
|
|
||||||
<keywords>
|
|
||||||
<keyword>CUDA</keyword>
|
|
||||||
<keyword>GPGPU</keyword>
|
|
||||||
<keyword>vote</keyword>
|
|
||||||
</keywords>
|
|
||||||
<libraries>
|
|
||||||
</libraries>
|
|
||||||
<librarypaths>
|
|
||||||
</librarypaths>
|
|
||||||
<nsight_eclipse>true</nsight_eclipse>
|
|
||||||
<primary_file>simpleVoteIntrinsics.cu</primary_file>
|
|
||||||
<scopes>
|
|
||||||
<scope>1:CUDA Basic Topics</scope>
|
|
||||||
</scopes>
|
|
||||||
<sm-arch>sm50</sm-arch>
|
|
||||||
<sm-arch>sm52</sm-arch>
|
|
||||||
<sm-arch>sm53</sm-arch>
|
|
||||||
<sm-arch>sm60</sm-arch>
|
|
||||||
<sm-arch>sm61</sm-arch>
|
|
||||||
<sm-arch>sm70</sm-arch>
|
|
||||||
<sm-arch>sm72</sm-arch>
|
|
||||||
<sm-arch>sm75</sm-arch>
|
|
||||||
<sm-arch>sm80</sm-arch>
|
|
||||||
<sm-arch>sm86</sm-arch>
|
|
||||||
<sm-arch>sm87</sm-arch>
|
|
||||||
<sm-arch>sm89</sm-arch>
|
|
||||||
<sm-arch>sm90</sm-arch>
|
|
||||||
<supported_envs>
|
|
||||||
<env>
|
|
||||||
<arch>x86_64</arch>
|
|
||||||
<platform>linux</platform>
|
|
||||||
</env>
|
|
||||||
<env>
|
|
||||||
<platform>windows7</platform>
|
|
||||||
</env>
|
|
||||||
<env>
|
|
||||||
<arch>x86_64</arch>
|
|
||||||
<platform>macosx</platform>
|
|
||||||
</env>
|
|
||||||
<env>
|
|
||||||
<arch>arm</arch>
|
|
||||||
</env>
|
|
||||||
<env>
|
|
||||||
<arch>aarch64</arch>
|
|
||||||
</env>
|
|
||||||
<env>
|
|
||||||
<arch>sbsa</arch>
|
|
||||||
</env>
|
|
||||||
<env>
|
|
||||||
<arch>ppc64le</arch>
|
|
||||||
<platform>linux</platform>
|
|
||||||
</env>
|
|
||||||
</supported_envs>
|
|
||||||
<supported_sm_architectures>
|
|
||||||
<include>all</include>
|
|
||||||
</supported_sm_architectures>
|
|
||||||
<title>Simple Vote Intrinsics</title>
|
|
||||||
<type>exe</type>
|
|
||||||
</entry>
|
|
||||||
@ -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}/simpleVoteIntrinsics_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 - simpleVoteIntrinsics_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: simpleVoteIntrinsics_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
|
|
||||||
|
|
||||||
simpleVoteIntrinsics.o:simpleVoteIntrinsics.cpp
|
|
||||||
$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $<
|
|
||||||
|
|
||||||
simpleVoteIntrinsics_nvrtc: simpleVoteIntrinsics.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
|
|
||||||
$(EXEC) ./simpleVoteIntrinsics_nvrtc
|
|
||||||
|
|
||||||
testrun: build
|
|
||||||
|
|
||||||
clean:
|
|
||||||
rm -f simpleVoteIntrinsics_nvrtc simpleVoteIntrinsics.o
|
|
||||||
rm -rf ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)/simpleVoteIntrinsics_nvrtc
|
|
||||||
|
|
||||||
clobber: clean
|
|
||||||
@ -1,74 +0,0 @@
|
|||||||
# simpleVoteIntrinsics_nvrtc - Simple Vote Intrinsics with libNVRTC
|
|
||||||
|
|
||||||
## Description
|
|
||||||
|
|
||||||
Simple program which demonstrates how to use the Vote (any, all) intrinsic instruction in a CUDA kernel with runtime compilation using NVRTC APIs. Requires Compute Capability 2.0 or higher.
|
|
||||||
|
|
||||||
## Key Concepts
|
|
||||||
|
|
||||||
Vote Intrinsics, CUDA Driver API, 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)
|
|
||||||
|
|
||||||
@ -1,341 +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.
|
|
||||||
*/
|
|
||||||
|
|
||||||
// System includes
|
|
||||||
#include <stdio.h>
|
|
||||||
#include <assert.h>
|
|
||||||
|
|
||||||
// CUDA runtime
|
|
||||||
#include <cuda_runtime.h>
|
|
||||||
#include "nvrtc_helper.h"
|
|
||||||
|
|
||||||
// helper functions and utilities to work with CUDA
|
|
||||||
#include <helper_functions.h>
|
|
||||||
|
|
||||||
#ifndef MAX
|
|
||||||
#define MAX(a, b) (a > b ? a : b)
|
|
||||||
#endif
|
|
||||||
|
|
||||||
static const char *sSDKsample = "[simpleVoteIntrinsics_nvrtc]\0";
|
|
||||||
|
|
||||||
////////////////////////////////////////////////////////////////////////////////
|
|
||||||
// Global types and parameters
|
|
||||||
////////////////////////////////////////////////////////////////////////////////
|
|
||||||
|
|
||||||
#define VOTE_DATA_GROUP 4
|
|
||||||
|
|
||||||
////////////////////////////////////////////////////////////////////////////////
|
|
||||||
// CUDA Voting Kernel functions
|
|
||||||
////////////////////////////////////////////////////////////////////////////////
|
|
||||||
|
|
||||||
// Generate the test pattern for Tests 1 and 2
|
|
||||||
void genVoteTestPattern(unsigned int *VOTE_PATTERN, int size) {
|
|
||||||
// For testing VOTE.Any (all of these threads will return 0)
|
|
||||||
for (int i = 0; i < size / 4; i++) {
|
|
||||||
VOTE_PATTERN[i] = 0x00000000;
|
|
||||||
}
|
|
||||||
|
|
||||||
// For testing VOTE.Any (1/2 these threads will return 1)
|
|
||||||
for (int i = 2 * size / 8; i < 4 * size / 8; i++) {
|
|
||||||
VOTE_PATTERN[i] = (i & 0x01) ? i : 0;
|
|
||||||
}
|
|
||||||
|
|
||||||
// For testing VOTE.all (1/2 of these threads will return 0)
|
|
||||||
for (int i = 2 * size / 4; i < 3 * size / 4; i++) {
|
|
||||||
VOTE_PATTERN[i] = (i & 0x01) ? 0 : i;
|
|
||||||
}
|
|
||||||
|
|
||||||
// For testing VOTE.all (all of these threads will return 1)
|
|
||||||
for (int i = 3 * size / 4; i < 4 * size / 4; i++) {
|
|
||||||
VOTE_PATTERN[i] = 0xffffffff;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
int checkErrors1(unsigned int *h_result, int start, int end, int warp_size,
|
|
||||||
const char *voteType) {
|
|
||||||
int i, sum = 0;
|
|
||||||
|
|
||||||
for (sum = 0, i = start; i < end; i++) {
|
|
||||||
sum += h_result[i];
|
|
||||||
}
|
|
||||||
|
|
||||||
if (sum > 0) {
|
|
||||||
printf("\t<%s>[%d - %d] = ", voteType, start, end - 1);
|
|
||||||
|
|
||||||
for (i = start; i < end; i++) {
|
|
||||||
printf("%d", h_result[i]);
|
|
||||||
}
|
|
||||||
|
|
||||||
printf("%d values FAILED\n", sum);
|
|
||||||
}
|
|
||||||
|
|
||||||
return (sum > 0);
|
|
||||||
}
|
|
||||||
|
|
||||||
int checkErrors2(unsigned int *h_result, int start, int end, int warp_size,
|
|
||||||
const char *voteType) {
|
|
||||||
int i, sum = 0;
|
|
||||||
|
|
||||||
for (sum = 0, i = start; i < end; i++) {
|
|
||||||
sum += h_result[i];
|
|
||||||
}
|
|
||||||
|
|
||||||
if (sum != warp_size) {
|
|
||||||
printf("\t<%s>[%d - %d] = ", voteType, start, end - 1);
|
|
||||||
|
|
||||||
for (i = start; i < end; i++) {
|
|
||||||
printf("%d", h_result[i]);
|
|
||||||
}
|
|
||||||
|
|
||||||
printf(" - FAILED\n");
|
|
||||||
}
|
|
||||||
|
|
||||||
return (sum != warp_size);
|
|
||||||
}
|
|
||||||
|
|
||||||
// Verification code for Kernel #1
|
|
||||||
int checkResultsVoteAnyKernel1(unsigned int *h_result, int size,
|
|
||||||
int warp_size) {
|
|
||||||
int error_count = 0;
|
|
||||||
|
|
||||||
error_count += checkErrors1(h_result, 0, VOTE_DATA_GROUP * warp_size / 4,
|
|
||||||
warp_size, "Vote.Any");
|
|
||||||
error_count +=
|
|
||||||
checkErrors2(h_result, VOTE_DATA_GROUP * warp_size / 4,
|
|
||||||
2 * VOTE_DATA_GROUP * warp_size / 4, warp_size, "Vote.Any");
|
|
||||||
error_count +=
|
|
||||||
checkErrors2(h_result, 2 * VOTE_DATA_GROUP * warp_size / 4,
|
|
||||||
3 * VOTE_DATA_GROUP * warp_size / 4, warp_size, "Vote.Any");
|
|
||||||
error_count +=
|
|
||||||
checkErrors2(h_result, 3 * VOTE_DATA_GROUP * warp_size / 4,
|
|
||||||
4 * VOTE_DATA_GROUP * warp_size / 4, warp_size, "Vote.Any");
|
|
||||||
|
|
||||||
printf((error_count == 0) ? "\tOK\n" : "\tERROR\n");
|
|
||||||
|
|
||||||
return error_count;
|
|
||||||
}
|
|
||||||
|
|
||||||
// Verification code for Kernel #2
|
|
||||||
int checkResultsVoteAllKernel2(unsigned int *h_result, int size,
|
|
||||||
int warp_size) {
|
|
||||||
int error_count = 0;
|
|
||||||
|
|
||||||
error_count += checkErrors1(h_result, 0, VOTE_DATA_GROUP * warp_size / 4,
|
|
||||||
warp_size, "Vote.All");
|
|
||||||
error_count +=
|
|
||||||
checkErrors1(h_result, VOTE_DATA_GROUP * warp_size / 4,
|
|
||||||
2 * VOTE_DATA_GROUP * warp_size / 4, warp_size, "Vote.All");
|
|
||||||
error_count +=
|
|
||||||
checkErrors1(h_result, 2 * VOTE_DATA_GROUP * warp_size / 4,
|
|
||||||
3 * VOTE_DATA_GROUP * warp_size / 4, warp_size, "Vote.All");
|
|
||||||
error_count +=
|
|
||||||
checkErrors2(h_result, 3 * VOTE_DATA_GROUP * warp_size / 4,
|
|
||||||
4 * VOTE_DATA_GROUP * warp_size / 4, warp_size, "Vote.All");
|
|
||||||
|
|
||||||
printf((error_count == 0) ? "\tOK\n" : "\tERROR\n");
|
|
||||||
|
|
||||||
return error_count;
|
|
||||||
}
|
|
||||||
|
|
||||||
// Verification code for Kernel #3
|
|
||||||
int checkResultsVoteAnyKernel3(bool *hinfo, int size) {
|
|
||||||
int i, error_count = 0;
|
|
||||||
|
|
||||||
for (i = 0; i < size * 3; i++) {
|
|
||||||
switch (i % 3) {
|
|
||||||
case 0:
|
|
||||||
// First warp should be all zeros.
|
|
||||||
if (hinfo[i] != (i >= size * 1)) {
|
|
||||||
error_count++;
|
|
||||||
}
|
|
||||||
break;
|
|
||||||
|
|
||||||
case 1:
|
|
||||||
// First warp and half of second should be all zeros.
|
|
||||||
if (hinfo[i] != (i >= size * 3 / 2)) {
|
|
||||||
error_count++;
|
|
||||||
}
|
|
||||||
break;
|
|
||||||
|
|
||||||
case 2:
|
|
||||||
// First two warps should be all zeros.
|
|
||||||
if (hinfo[i] != (i >= size * 2)) {
|
|
||||||
error_count++;
|
|
||||||
}
|
|
||||||
break;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
printf((error_count == 0) ? "\tOK\n" : "\tERROR\n");
|
|
||||||
|
|
||||||
return error_count;
|
|
||||||
}
|
|
||||||
|
|
||||||
int main(int argc, char **argv) {
|
|
||||||
unsigned int *h_input, *h_result;
|
|
||||||
CUdeviceptr d_input, d_result;
|
|
||||||
|
|
||||||
char *cubin, *kernel_file;
|
|
||||||
size_t cubinSize;
|
|
||||||
kernel_file = sdkFindFilePath("simpleVote_kernel.cuh", argv[0]);
|
|
||||||
compileFileToCUBIN(kernel_file, argc, argv, &cubin, &cubinSize, 0);
|
|
||||||
CUmodule module = loadCUBIN(cubin, argc, argv);
|
|
||||||
|
|
||||||
bool *hinfo = NULL;
|
|
||||||
CUdeviceptr dinfo;
|
|
||||||
|
|
||||||
int error_count[3] = {0, 0, 0};
|
|
||||||
int warp_size = 32;
|
|
||||||
|
|
||||||
printf("%s\n", sSDKsample);
|
|
||||||
|
|
||||||
h_input = (unsigned int *)malloc(VOTE_DATA_GROUP * warp_size *
|
|
||||||
sizeof(unsigned int));
|
|
||||||
h_result = (unsigned int *)malloc(VOTE_DATA_GROUP * warp_size *
|
|
||||||
sizeof(unsigned int));
|
|
||||||
|
|
||||||
checkCudaErrors(
|
|
||||||
cuMemAlloc(&d_input, VOTE_DATA_GROUP * warp_size * sizeof(unsigned int)));
|
|
||||||
checkCudaErrors(cuMemAlloc(
|
|
||||||
&d_result, VOTE_DATA_GROUP * warp_size * sizeof(unsigned int)));
|
|
||||||
|
|
||||||
genVoteTestPattern(h_input, VOTE_DATA_GROUP * warp_size);
|
|
||||||
|
|
||||||
checkCudaErrors(cuMemcpyHtoD(
|
|
||||||
d_input, h_input, VOTE_DATA_GROUP * warp_size * sizeof(unsigned int)));
|
|
||||||
|
|
||||||
// Start of Vote Any Test Kernel #1
|
|
||||||
printf("[VOTE Kernel Test 1/3]\n");
|
|
||||||
printf("\tRunning <<Vote.Any>> kernel1 ...\n");
|
|
||||||
{
|
|
||||||
dim3 gridBlock(1, 1);
|
|
||||||
dim3 threadBlock(VOTE_DATA_GROUP * warp_size, 1);
|
|
||||||
CUfunction kernel_addr;
|
|
||||||
checkCudaErrors(
|
|
||||||
cuModuleGetFunction(&kernel_addr, module, "VoteAnyKernel1"));
|
|
||||||
|
|
||||||
int size = VOTE_DATA_GROUP * warp_size;
|
|
||||||
void *arr[] = {(void *)&d_input, (void *)&d_result, (void *)&size};
|
|
||||||
|
|
||||||
checkCudaErrors(cuLaunchKernel(
|
|
||||||
kernel_addr, gridBlock.x, gridBlock.y, gridBlock.z, /* grid dim */
|
|
||||||
threadBlock.x, threadBlock.y, threadBlock.z, /* block dim */
|
|
||||||
0, 0, /* shared mem, stream */
|
|
||||||
&arr[0], /* arguments */
|
|
||||||
0));
|
|
||||||
|
|
||||||
checkCudaErrors(cuCtxSynchronize());
|
|
||||||
}
|
|
||||||
|
|
||||||
checkCudaErrors(cuMemcpyDtoH(
|
|
||||||
h_result, d_result, VOTE_DATA_GROUP * warp_size * sizeof(unsigned int)));
|
|
||||||
|
|
||||||
error_count[0] += checkResultsVoteAnyKernel1(
|
|
||||||
h_result, VOTE_DATA_GROUP * warp_size, warp_size);
|
|
||||||
|
|
||||||
// Start of Vote All Test Kernel #2
|
|
||||||
|
|
||||||
printf("\n[VOTE Kernel Test 2/3]\n");
|
|
||||||
|
|
||||||
printf("\tRunning <<Vote.All>> kernel2 ...\n");
|
|
||||||
{
|
|
||||||
dim3 gridBlock(1, 1);
|
|
||||||
dim3 threadBlock(VOTE_DATA_GROUP * warp_size, 1);
|
|
||||||
|
|
||||||
CUfunction kernel_addr;
|
|
||||||
checkCudaErrors(
|
|
||||||
cuModuleGetFunction(&kernel_addr, module, "VoteAllKernel2"));
|
|
||||||
|
|
||||||
int size = VOTE_DATA_GROUP * warp_size;
|
|
||||||
void *arr[] = {(void *)&d_input, (void *)&d_result, (void *)&size};
|
|
||||||
|
|
||||||
checkCudaErrors(cuLaunchKernel(
|
|
||||||
kernel_addr, gridBlock.x, gridBlock.y, gridBlock.z, /* grid dim */
|
|
||||||
threadBlock.x, threadBlock.y, threadBlock.z, /* block dim */
|
|
||||||
0, 0, /* shared mem, stream */
|
|
||||||
&arr[0], /* arguments */
|
|
||||||
0));
|
|
||||||
|
|
||||||
checkCudaErrors(cuCtxSynchronize());
|
|
||||||
}
|
|
||||||
|
|
||||||
checkCudaErrors(cuMemcpyDtoH(
|
|
||||||
h_result, d_result, VOTE_DATA_GROUP * warp_size * sizeof(unsigned int)));
|
|
||||||
|
|
||||||
error_count[1] += checkResultsVoteAllKernel2(
|
|
||||||
h_result, VOTE_DATA_GROUP * warp_size, warp_size);
|
|
||||||
|
|
||||||
// Second Vote Kernel Test #3 (both Any/All)
|
|
||||||
hinfo = (bool *)calloc(warp_size * 3 * 3, sizeof(bool));
|
|
||||||
|
|
||||||
checkCudaErrors(cuMemAlloc(&dinfo, warp_size * 3 * 3 * sizeof(bool)));
|
|
||||||
checkCudaErrors(cuMemcpyHtoD(dinfo, hinfo, warp_size * 3 * 3 * sizeof(bool)));
|
|
||||||
|
|
||||||
printf("\n[VOTE Kernel Test 3/3]\n");
|
|
||||||
printf("\tRunning <<Vote.Any>> kernel3 ...\n");
|
|
||||||
{
|
|
||||||
dim3 gridBlock(1, 1);
|
|
||||||
dim3 threadBlock(warp_size * 3, 1);
|
|
||||||
|
|
||||||
CUfunction kernel_addr;
|
|
||||||
|
|
||||||
checkCudaErrors(
|
|
||||||
cuModuleGetFunction(&kernel_addr, module, "VoteAnyKernel3"));
|
|
||||||
|
|
||||||
int size = warp_size;
|
|
||||||
void *arr[] = {(void *)&dinfo, (void *)&size};
|
|
||||||
|
|
||||||
checkCudaErrors(cuLaunchKernel(
|
|
||||||
kernel_addr, gridBlock.x, gridBlock.y, gridBlock.z, /* grid dim */
|
|
||||||
threadBlock.x, threadBlock.y, threadBlock.z, /* block dim */
|
|
||||||
0, 0, /* shared mem, stream */
|
|
||||||
&arr[0], /* arguments */
|
|
||||||
0));
|
|
||||||
|
|
||||||
checkCudaErrors(cuCtxSynchronize());
|
|
||||||
}
|
|
||||||
|
|
||||||
checkCudaErrors(cuMemcpyDtoH(hinfo, dinfo, warp_size * 3 * 3 * sizeof(bool)));
|
|
||||||
|
|
||||||
error_count[2] = checkResultsVoteAnyKernel3(hinfo, warp_size * 3);
|
|
||||||
|
|
||||||
// Now free these resources for Test #1,2
|
|
||||||
checkCudaErrors(cuMemFree(d_input));
|
|
||||||
checkCudaErrors(cuMemFree(d_result));
|
|
||||||
free(h_input);
|
|
||||||
free(h_result);
|
|
||||||
|
|
||||||
// Free resources from Test #3
|
|
||||||
free(hinfo);
|
|
||||||
checkCudaErrors(cuMemFree(dinfo));
|
|
||||||
|
|
||||||
printf("\tShutting down...\n");
|
|
||||||
|
|
||||||
return (error_count[0] == 0 && error_count[1] == 0 && error_count[2] == 0)
|
|
||||||
? EXIT_SUCCESS
|
|
||||||
: EXIT_FAILURE;
|
|
||||||
}
|
|
||||||
@ -1,80 +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 SIMPLEVOTE_KERNEL_CU
|
|
||||||
#define SIMPLEVOTE_KERNEL_CU
|
|
||||||
|
|
||||||
////////////////////////////////////////////////////////////////////////////////
|
|
||||||
// Vote Any/All intrinsic kernel function tests are supported only by CUDA
|
|
||||||
// capable devices that are CUDA hardware that has SM1.2 or later
|
|
||||||
// Vote Functions (refer to section 4.4.5 in the CUDA Programming Guide)
|
|
||||||
////////////////////////////////////////////////////////////////////////////////
|
|
||||||
|
|
||||||
// Kernel #1 tests the across-the-warp vote(any) intrinsic.
|
|
||||||
// If ANY one of the threads (within the warp) of the predicated condition
|
|
||||||
// returns a non-zero value, then all threads within this warp will return a
|
|
||||||
// non-zero value
|
|
||||||
extern "C" __global__ void VoteAnyKernel1(unsigned int *input,
|
|
||||||
unsigned int *result, int size) {
|
|
||||||
int tx = threadIdx.x;
|
|
||||||
|
|
||||||
int mask = 0xffffffff;
|
|
||||||
result[tx] = __any_sync(mask, input[tx]);
|
|
||||||
}
|
|
||||||
|
|
||||||
// Kernel #2 tests the across-the-warp vote(all) intrinsic.
|
|
||||||
// If ALL of the threads (within the warp) of the predicated condition returns
|
|
||||||
// a non-zero value, then all threads within this warp will return a non-zero
|
|
||||||
// value
|
|
||||||
extern "C" __global__ void VoteAllKernel2(unsigned int *input,
|
|
||||||
unsigned int *result, int size) {
|
|
||||||
int tx = threadIdx.x;
|
|
||||||
|
|
||||||
int mask = 0xffffffff;
|
|
||||||
result[tx] = __all_sync(mask, input[tx]);
|
|
||||||
}
|
|
||||||
|
|
||||||
// Kernel #3 is a directed test for the across-the-warp vote(all) intrinsic.
|
|
||||||
// This kernel will test for conditions across warps, and within half warps
|
|
||||||
extern "C" __global__ void VoteAnyKernel3(bool *info, int warp_size) {
|
|
||||||
int tx = threadIdx.x;
|
|
||||||
unsigned int mask = 0xffffffff;
|
|
||||||
bool *offs = info + (tx * 3);
|
|
||||||
|
|
||||||
// The following should hold true for the second and third warp
|
|
||||||
*offs = __any_sync(mask, (tx >= (warp_size * 3) / 2));
|
|
||||||
// The following should hold true for the "upper half" of the second warp,
|
|
||||||
// and all of the third warp
|
|
||||||
*(offs + 1) = (tx >= (warp_size * 3) / 2 ? true : false);
|
|
||||||
|
|
||||||
// The following should hold true for the third warp only
|
|
||||||
if (__all_sync(mask, (tx >= (warp_size * 3) / 2))) {
|
|
||||||
*(offs + 2) = true;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
#endif
|
|
||||||
@ -1,72 +0,0 @@
|
|||||||
<?xml version="1.0" encoding="UTF-8"?>
|
|
||||||
<!DOCTYPE entry SYSTEM "SamplesInfo.dtd">
|
|
||||||
<entry>
|
|
||||||
<name>template</name>
|
|
||||||
<cuda_api_list>
|
|
||||||
<toolkit>cudaMalloc</toolkit>
|
|
||||||
<toolkit>cudaMemcpy</toolkit>
|
|
||||||
<toolkit>cudaFree</toolkit>
|
|
||||||
</cuda_api_list>
|
|
||||||
<description><![CDATA[A trivial template project that can be used as a starting point to create new CUDA projects.]]></description>
|
|
||||||
<devicecompilation>whole</devicecompilation>
|
|
||||||
<includepaths>
|
|
||||||
<path>./</path>
|
|
||||||
<path>../</path>
|
|
||||||
<path>../../../Common</path>
|
|
||||||
</includepaths>
|
|
||||||
<keyconcepts>
|
|
||||||
<concept level="basic">Device Memory Allocation</concept>
|
|
||||||
</keyconcepts>
|
|
||||||
<keywords>
|
|
||||||
</keywords>
|
|
||||||
<libraries>
|
|
||||||
</libraries>
|
|
||||||
<librarypaths>
|
|
||||||
</librarypaths>
|
|
||||||
<nsight_eclipse>true</nsight_eclipse>
|
|
||||||
<primary_file>template.cu</primary_file>
|
|
||||||
<scopes>
|
|
||||||
<scope>1:CUDA Basic Topics</scope>
|
|
||||||
</scopes>
|
|
||||||
<sm-arch>sm50</sm-arch>
|
|
||||||
<sm-arch>sm52</sm-arch>
|
|
||||||
<sm-arch>sm53</sm-arch>
|
|
||||||
<sm-arch>sm60</sm-arch>
|
|
||||||
<sm-arch>sm61</sm-arch>
|
|
||||||
<sm-arch>sm70</sm-arch>
|
|
||||||
<sm-arch>sm72</sm-arch>
|
|
||||||
<sm-arch>sm75</sm-arch>
|
|
||||||
<sm-arch>sm80</sm-arch>
|
|
||||||
<sm-arch>sm86</sm-arch>
|
|
||||||
<sm-arch>sm87</sm-arch>
|
|
||||||
<sm-arch>sm89</sm-arch>
|
|
||||||
<sm-arch>sm90</sm-arch>
|
|
||||||
<supported_envs>
|
|
||||||
<env>
|
|
||||||
<arch>x86_64</arch>
|
|
||||||
<platform>linux</platform>
|
|
||||||
</env>
|
|
||||||
<env>
|
|
||||||
<platform>windows7</platform>
|
|
||||||
</env>
|
|
||||||
<env>
|
|
||||||
<arch>x86_64</arch>
|
|
||||||
<platform>macosx</platform>
|
|
||||||
</env>
|
|
||||||
<env>
|
|
||||||
<arch>arm</arch>
|
|
||||||
</env>
|
|
||||||
<env>
|
|
||||||
<arch>sbsa</arch>
|
|
||||||
</env>
|
|
||||||
<env>
|
|
||||||
<arch>ppc64le</arch>
|
|
||||||
<platform>linux</platform>
|
|
||||||
</env>
|
|
||||||
</supported_envs>
|
|
||||||
<supported_sm_architectures>
|
|
||||||
<include>all</include>
|
|
||||||
</supported_sm_architectures>
|
|
||||||
<title>Template</title>
|
|
||||||
<type>exe</type>
|
|
||||||
</entry>
|
|
||||||
Loading…
x
Reference in New Issue
Block a user