NASP/cuda-samples
85
0
Fork 0
mirror of https://github.com/NVIDIA/cuda-samples.git synced 2026-10-23 21:28:17 +08:00
Code Issues Packages Projects Releases Wiki Activity

Add and update samples for cuda 11.0 support

Browse Source
This commit is contained in:
Mahesh Doijade
2020-05-18 22:22:06 +05:30
parent 6be514679b
commit 4f6e02970b
410 changed files with 21920 additions and 3864 deletions
Download Patch File Download Diff File Expand all files Collapse all files

51
Samples/conjugateGradientCudaGraphs/Makefile
View File

@@ -72,9 +72,9 @@ endif
# architecture
HOST_ARCH := $(shell uname -m)
TARGET_ARCH ?= $(HOST_ARCH)
ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 ppc64le armv7l))
ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 sbsa ppc64le armv7l))
ifneq ($(TARGET_ARCH),$(HOST_ARCH))
ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 ppc64le))
ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 sbsa ppc64le))
TARGET_SIZE := 64
else ifneq (,$(filter $(TARGET_ARCH),armv7l))
TARGET_SIZE := 32
@@ -85,8 +85,17 @@ ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 ppc64le armv7l))
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 $(CUDA_PATH)/targets/sbsa-linux))
HOST_ARCH := sbsa
TARGET_ARCH := sbsa
endif
endif
ifneq ($(TARGET_ARCH),$(HOST_ARCH))
ifeq (,$(filter $(HOST_ARCH)-$(TARGET_ARCH),aarch64-armv7l x86_64-armv7l x86_64-aarch64 x86_64-ppc64le))
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 $(TARGET_ARCH) is not supported!)
endif
endif
@@ -141,6 +150,8 @@ else ifneq ($(TARGET_ARCH),$(HOST_ARCH))
else ifeq ($(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
@@ -185,19 +196,27 @@ ifneq ($(TARGET_ARCH),$(HOST_ARCH))
CCFLAGS += --sysroot=$(TARGET_FS)
endif
LDFLAGS += --sysroot=$(TARGET_FS)
LDFLAGS += -rpath-link=$(TARGET_FS)/lib -L $(TARGET_FS)/lib
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib -L $(TARGET_FS)/usr/lib
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib/aarch64-linux-gnu -L $(TARGET_FS)/usr/lib/aarch64-linux-gnu
LDFLAGS += -rpath-link=$(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
CCFLAGS += -isystem=$(TARGET_FS)/usr/include/aarch64-linux-gnu
CCFLAGS += -isystem=$(TARGET_FS)/usr/include -I$(TARGET_FS)/usr/include
CCFLAGS += -isystem=$(TARGET_FS)/usr/include/aarch64-linux-gnu -I$(TARGET_FS)/usr/include/aarch64-linux-gnu
endif
endif
ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-qnx)
CCFLAGS += -DWIN_INTERFACE_CUSTOM -I/usr/include/aarch64-qnx-gnu
LDFLAGS += -lsocket
LDFLAGS += -rpath=/usr/lib/aarch64-qnx-gnu -L/usr/lib/aarch64-qnx-gnu
ifneq ($(TARGET_FS),)
LDFLAGS += -rpath=$(TARGET_FS)/usr/lib -L $(TARGET_FS)/usr/lib
LDFLAGS += -rpath=$(TARGET_FS)/usr/libnvidia -L $(TARGET_FS)/usr/libnvidia
endif
ifdef TARGET_OVERRIDE # cuda toolkit targets override
NVCCFLAGS += -target-dir $(TARGET_OVERRIDE)
endif
endif
endif
ifeq ($(TARGET_OS),qnx)
CCFLAGS += -DWIN_INTERFACE_CUSTOM
LDFLAGS += -lsocket
endif
# Install directory of different arch
@@ -206,6 +225,8 @@ 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)
@@ -247,9 +268,9 @@ LIBRARIES :=
# Gencode arguments
ifeq ($(TARGET_ARCH),$(filter $(TARGET_ARCH),armv7l aarch64))
SMS ?= 30 35 37 50 52 60 61 70 72 75
SMS ?= 35 37 50 52 60 61 70 72 75 80
else
SMS ?= 30 35 37 50 52 60 61 70 75
SMS ?= 35 37 50 52 60 61 70 75 80
endif
ifeq ($(SMS),)

2
Samples/conjugateGradientCudaGraphs/NsightEclipse.xml
View File

@@ -48,7 +48,6 @@
<scope>3:Linear Algebra</scope>
<scope>1:CUDA Graphs</scope>
</scopes>
<sm-arch>sm30</sm-arch>
<sm-arch>sm35</sm-arch>
<sm-arch>sm37</sm-arch>
<sm-arch>sm50</sm-arch>
@@ -58,6 +57,7 @@
<sm-arch>sm70</sm-arch>
<sm-arch>sm72</sm-arch>
<sm-arch>sm75</sm-arch>
<sm-arch>sm80</sm-arch>
<supported_envs>
<env>
<arch>x86_64</arch>

30
Samples/conjugateGradientCudaGraphs/README.md
View File

@@ -10,11 +10,11 @@ Linear Algebra, CUBLAS Library, CUSPARSE Library
## Supported SM Architectures
[SM 3.0 ](https://developer.nvidia.com/cuda-gpus) [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 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)
## Supported OSes
Linux, Windows, MacOSX
Linux, Windows
## Supported CPU Architecture
@@ -30,7 +30,7 @@ cudaStreamBeginCapture, cudaStreamEndCapture, cudaGraphCreate, cudaGraphLaunch,
## Prerequisites
Download and install the [CUDA Toolkit 10.2](https://developer.nvidia.com/cuda-downloads) for your corresponding platform.
Download and install the [CUDA Toolkit 11.0](https://developer.nvidia.com/cuda-downloads) for your corresponding platform.
Make sure the dependencies mentioned in [Dependencies]() section above are installed.
## Build and Run
@@ -70,29 +70,5 @@ The samples makefiles can take advantage of certain options:
$ make HOST_COMPILER=g++
```
### Mac
The Mac samples are built using makefiles. To use the makefiles, change directory into the sample directory you wish to build, and run make:
```
$ cd <sample_dir>
$ make
```
The samples makefiles can take advantage of certain options:
* **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="A B ..."
```
* **HOST_COMPILER=<host_compiler>** - override the default clang host compiler. See the [Mac Installation Guide](http://docs.nvidia.com/cuda/cuda-installation-guide-mac-os-x/index.html#system-requirements) for a list of supported host compilers.
```
$ make HOST_COMPILER=clang
```
## References (for more details)

52
Samples/conjugateGradientCudaGraphs/conjugateGradientCudaGraphs.cu
View File

@@ -42,13 +42,10 @@
#include <cuda_runtime.h>
#include <cusparse.h>
#include <cooperative_groups.h>
// Utilities and system includes
#include <helper_cuda.h> // helper function CUDA error checking and initialization
#include <helper_functions.h> // helper for shared functions common to CUDA Samples
namespace cg = cooperative_groups;
const char *sSDKname = "conjugateGradientCudaGraphs";
@@ -193,6 +190,26 @@ int main(int argc, char **argv) {
checkCudaErrors(cudaMalloc((void **)&d_na, sizeof(float)));
checkCudaErrors(cudaMalloc((void **)&d_b, sizeof(float)));
/* Wrap raw data into cuSPARSE generic API objects */
cusparseSpMatDescr_t matA = NULL;
checkCudaErrors(cusparseCreateCsr(
&matA, N, N, nz, d_row, d_col, d_val, CUSPARSE_INDEX_32I,
CUSPARSE_INDEX_32I, CUSPARSE_INDEX_BASE_ZERO, CUDA_R_32F));
cusparseDnVecDescr_t vecx = NULL;
checkCudaErrors(cusparseCreateDnVec(&vecx, N, d_x, CUDA_R_32F));
cusparseDnVecDescr_t vecp = NULL;
checkCudaErrors(cusparseCreateDnVec(&vecp, N, d_p, CUDA_R_32F));
cusparseDnVecDescr_t vecAx = NULL;
checkCudaErrors(cusparseCreateDnVec(&vecAx, N, d_Ax, CUDA_R_32F));
/* Allocate workspace for cuSPARSE */
size_t bufferSize = 0;
checkCudaErrors(cusparseSpMV_bufferSize(
cusparseHandle, CUSPARSE_OPERATION_NON_TRANSPOSE, &alpha, matA, vecx,
&beta, vecAx, CUDA_R_32F, CUSPARSE_MV_ALG_DEFAULT, &bufferSize));
void *buffer = NULL;
checkCudaErrors(cudaMalloc(&buffer, bufferSize));
cusparseMatDescr_t descr = 0;
checkCudaErrors(cusparseCreateMatDescr(&descr));
@@ -217,9 +234,9 @@ int main(int argc, char **argv) {
beta = 0.0;
checkCudaErrors(cusparseSetStream(cusparseHandle, stream1));
checkCudaErrors(
cusparseScsrmv(cusparseHandle, CUSPARSE_OPERATION_NON_TRANSPOSE, N, N, nz,
&alpha, descr, d_val, d_row, d_col, d_x, &beta, d_Ax));
checkCudaErrors(cusparseSpMV(
cusparseHandle, CUSPARSE_OPERATION_NON_TRANSPOSE, &alpha, matA, vecx,
&beta, vecAx, CUDA_R_32F, CUSPARSE_MV_ALG_DEFAULT, &buffer));
checkCudaErrors(cublasSetStream(cublasHandle, stream1));
checkCudaErrors(cublasSaxpy(cublasHandle, N, &alpham1, d_Ax, 1, d_r, 1));
@@ -231,9 +248,9 @@ int main(int argc, char **argv) {
k = 1;
// First Iteration when k=1 starts
checkCudaErrors(cublasScopy(cublasHandle, N, d_r, 1, d_p, 1));
checkCudaErrors(
cusparseScsrmv(cusparseHandle, CUSPARSE_OPERATION_NON_TRANSPOSE, N, N, nz,
&alpha, descr, d_val, d_row, d_col, d_p, &beta, d_Ax));
checkCudaErrors(cusparseSpMV(
cusparseHandle, CUSPARSE_OPERATION_NON_TRANSPOSE, &alpha, matA, vecp,
&beta, vecAx, CUDA_R_32F, CUSPARSE_MV_ALG_DEFAULT, &buffer));
checkCudaErrors(cublasSdot(cublasHandle, N, d_p, 1, d_Ax, 1, d_dot));
@@ -273,9 +290,9 @@ int main(int argc, char **argv) {
checkCudaErrors(
cusparseSetPointerMode(cusparseHandle, CUSPARSE_POINTER_MODE_HOST));
checkCudaErrors(
cusparseScsrmv(cusparseHandle, CUSPARSE_OPERATION_NON_TRANSPOSE, N, N, nz,
&alpha, descr, d_val, d_row, d_col, d_p, &beta, d_Ax));
checkCudaErrors(cusparseSpMV(
cusparseHandle, CUSPARSE_OPERATION_NON_TRANSPOSE, &alpha, matA, vecp,
&beta, vecAx, CUDA_R_32F, CUSPARSE_MV_ALG_DEFAULT, &buffer));
checkCudaErrors(cudaMemsetAsync(d_dot, 0, sizeof(float), stream1));
checkCudaErrors(cublasSdot(cublasHandle, N, d_p, 1, d_Ax, 1, d_dot));
@@ -317,9 +334,9 @@ int main(int argc, char **argv) {
cublasSetPointerMode(cublasHandle, CUBLAS_POINTER_MODE_HOST);
checkCudaErrors(cublasSaxpy(cublasHandle, N, &alpha, d_r, 1, d_p, 1));
checkCudaErrors(cusparseScsrmv(
cusparseHandle, CUSPARSE_OPERATION_NON_TRANSPOSE, N, N, nz, &alpha,
descr, d_val, d_row, d_col, d_p, &beta, d_Ax));
checkCudaErrors(cusparseSpMV(
cusparseHandle, CUSPARSE_OPERATION_NON_TRANSPOSE, &alpha, matA, vecp,
&beta, vecAx, CUDA_R_32F, CUSPARSE_MV_ALG_DEFAULT, &buffer));
cublasSetPointerMode(cublasHandle, CUBLAS_POINTER_MODE_DEVICE);
checkCudaErrors(cublasSdot(cublasHandle, N, d_p, 1, d_Ax, 1, d_dot));
@@ -378,6 +395,11 @@ int main(int argc, char **argv) {
cusparseDestroy(cusparseHandle);
cublasDestroy(cublasHandle);
if (matA ) { checkCudaErrors(cusparseDestroySpMat(matA)); }
if (vecx ) { checkCudaErrors(cusparseDestroyDnVec(vecx)); }
if (vecAx ) { checkCudaErrors(cusparseDestroyDnVec(vecAx)); }
if (vecp ) { checkCudaErrors(cusparseDestroyDnVec(vecp)); }
free(I);
free(J);
free(val);

2
Samples/conjugateGradientCudaGraphs/conjugateGradientCudaGraphs_vs2012.vcxproj
View File

@@ -62,7 +62,7 @@
<OutputFile>$(OutDir)/conjugateGradientCudaGraphs.exe</OutputFile>
</Link>
<CudaCompile>
<CodeGeneration>compute_30,sm_30;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;</CodeGeneration>
<CodeGeneration>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;</CodeGeneration>
<AdditionalOptions>-Xcompiler "/wd 4819" %(AdditionalOptions)</AdditionalOptions>
<Include>./;../../Common</Include>
<Defines>WIN32</Defines>

2
Samples/conjugateGradientCudaGraphs/conjugateGradientCudaGraphs_vs2013.vcxproj
View File

@@ -62,7 +62,7 @@
<OutputFile>$(OutDir)/conjugateGradientCudaGraphs.exe</OutputFile>
</Link>
<CudaCompile>
<CodeGeneration>compute_30,sm_30;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;</CodeGeneration>
<CodeGeneration>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;</CodeGeneration>
<AdditionalOptions>-Xcompiler "/wd 4819" %(AdditionalOptions)</AdditionalOptions>
<Include>./;../../Common</Include>
<Defines>WIN32</Defines>

2
Samples/conjugateGradientCudaGraphs/conjugateGradientCudaGraphs_vs2015.vcxproj
View File

@@ -62,7 +62,7 @@
<OutputFile>$(OutDir)/conjugateGradientCudaGraphs.exe</OutputFile>
</Link>
<CudaCompile>
<CodeGeneration>compute_30,sm_30;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;</CodeGeneration>
<CodeGeneration>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;</CodeGeneration>
<AdditionalOptions>-Xcompiler "/wd 4819" %(AdditionalOptions)</AdditionalOptions>
<Include>./;../../Common</Include>
<Defines>WIN32</Defines>

2
Samples/conjugateGradientCudaGraphs/conjugateGradientCudaGraphs_vs2017.vcxproj
View File

@@ -67,7 +67,7 @@
<OutputFile>$(OutDir)/conjugateGradientCudaGraphs.exe</OutputFile>
</Link>
<CudaCompile>
<CodeGeneration>compute_30,sm_30;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;</CodeGeneration>
<CodeGeneration>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;</CodeGeneration>
<AdditionalOptions>-Xcompiler "/wd 4819" %(AdditionalOptions)</AdditionalOptions>
<Include>./;../../Common</Include>
<Defines>WIN32</Defines>

2
Samples/conjugateGradientCudaGraphs/conjugateGradientCudaGraphs_vs2019.vcxproj
View File

@@ -63,7 +63,7 @@
<OutputFile>$(OutDir)/conjugateGradientCudaGraphs.exe</OutputFile>
</Link>
<CudaCompile>
<CodeGeneration>compute_30,sm_30;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;</CodeGeneration>
<CodeGeneration>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;</CodeGeneration>
<AdditionalOptions>-Xcompiler "/wd 4819" %(AdditionalOptions)</AdditionalOptions>
<Include>./;../../Common</Include>
<Defines>WIN32</Defines>