Hi all,
It’s my first time to try build pytorch from source on windows. I followed the official build docs. Here is the cmake summaries
-- System : Windows
-- C++ compiler : C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.37.32822/bin/Hostx64/x64/cl.exe
-- C++ compiler version : 19.37.32825.0
-- CXX flags : /DWIN32 /D_WINDOWS /GR /EHsc /bigobj /FS -DUSE_PTHREADPOOL /std:c++17 /EHsc /wd26812
-- Build type : Release
-- Compile definitions : ONNX_ML=1;ONNXIFI_ENABLE_EXT=1;__STDC_FORMAT_MACROS
-- CMAKE_PREFIX_PATH : D:\workspace\envs\py38_torch1.13.1_cuda11.6\Lib\site-packages;C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6
-- CMAKE_INSTALL_PREFIX : D:/workspace/codes/pytorch/torch
-- CMAKE_MODULE_PATH : D:/workspace/codes/pytorch/cmake/Modules;D:/workspace/codes/pytorch/cmake/public/../Modules_CUDA_fix
--
-- ONNX version : 1.14.1
-- ONNX NAMESPACE : onnx_torch
-- ONNX_USE_LITE_PROTO : OFF
-- USE_PROTOBUF_SHARED_LIBS : OFF
-- Protobuf_USE_STATIC_LIBS : ON
-- ONNX_DISABLE_EXCEPTIONS : OFF
-- ONNX_WERROR : OFF
-- ONNX_BUILD_TESTS : OFF
-- ONNX_BUILD_BENCHMARKS : OFF
--
-- Protobuf compiler :
-- Protobuf includes :
-- Protobuf libraries :
-- BUILD_ONNX_PYTHON : OFF
--
-- ******** Summary ********
-- CMake version : 3.24.1
-- CMake command : D:/workspace/envs/py38_torch1.13.1_cuda11.6/Library/bin/cmake.exe
-- System : Windows
-- C++ compiler : C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.37.32822/bin/Hostx64/x64/cl.exe
-- C++ compiler version : 19.37.32825.0
-- CXX flags : /DWIN32 /D_WINDOWS /GR /EHsc /bigobj /FS -DUSE_PTHREADPOOL
-- Build type : Release
-- Compile definitions : ONNX_ML=1;ONNXIFI_ENABLE_EXT=1
-- CMAKE_PREFIX_PATH : D:\workspace\envs\py38_torch1.13.1_cuda11.6\Lib\site-packages;C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6
-- CMAKE_INSTALL_PREFIX : D:/workspace/codes/pytorch/torch
-- CMAKE_MODULE_PATH : D:/workspace/codes/pytorch/cmake/Modules;D:/workspace/codes/pytorch/cmake/public/../Modules_CUDA_fix
--
-- ONNX version : 1.4.1
-- ONNX NAMESPACE : onnx_torch
-- ONNX_BUILD_TESTS : OFF
-- ONNX_BUILD_BENCHMARKS : OFF
-- ONNX_USE_LITE_PROTO : OFF
-- ONNXIFI_DUMMY_BACKEND :
--
-- Protobuf compiler :
-- Protobuf includes :
-- Protobuf libraries :
-- BUILD_ONNX_PYTHON : OFF
-- Found CUDA with FP16 support, compiling with torch.cuda.HalfTensor
-- Adding -DNDEBUG to compile flags
-- MAGMA not found. Compiling without MAGMA support
-- Could not find hardware support for NEON on this machine.
-- No OMAP3 processor on this machine.
-- No OMAP4 processor on this machine.
-- MKL_THREADING = OMP
-- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core - libiomp5md]
-- Library mkl_intel_lp64: not found
-- Checking for [mkl_intel - mkl_intel_thread - mkl_core - libiomp5md]
-- Library mkl_intel: not found
-- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core]
-- Library mkl_intel_lp64: not found
-- Checking for [mkl_intel - mkl_intel_thread - mkl_core]
-- Library mkl_intel: not found
-- Checking for [mkl_intel_lp64 - mkl_sequential - mkl_core]
-- Library mkl_intel_lp64: not found
-- Checking for [mkl_intel - mkl_sequential - mkl_core]
-- Library mkl_intel: not found
-- Checking for [mkl_intel_lp64 - mkl_core - libiomp5md - pthread]
-- Library mkl_intel_lp64: not found
-- Checking for [mkl_intel - mkl_core - libiomp5md - pthread]
-- Library mkl_intel: not found
-- Checking for [mkl_intel_lp64 - mkl_core - pthread]
-- Library mkl_intel_lp64: not found
-- Checking for [mkl_intel - mkl_core - pthread]
-- Library mkl_intel: not found
-- Checking for [mkl - guide - pthread - m]
-- Library mkl: not found
-- MKL library not found
-- Checking for [blis]
-- Library blis: BLAS_blis_LIBRARY-NOTFOUND
-- Checking for [Accelerate]
-- Library Accelerate: BLAS_Accelerate_LIBRARY-NOTFOUND
-- Checking for [vecLib]
-- Library vecLib: BLAS_vecLib_LIBRARY-NOTFOUND
-- Checking for [flexiblas]
-- Library flexiblas: BLAS_flexiblas_LIBRARY-NOTFOUND
-- Checking for [openblas]
-- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND
-- Checking for [openblas - pthread - m]
-- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND
-- Checking for [openblas - pthread - m - gomp]
-- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND
-- Checking for [libopenblas]
-- Library libopenblas: BLAS_libopenblas_LIBRARY-NOTFOUND
-- Checking for [goto2 - gfortran]
-- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND
-- Checking for [goto2 - gfortran - pthread]
-- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND
-- Checking for [acml - gfortran]
-- Library acml: BLAS_acml_LIBRARY-NOTFOUND
-- Checking for [blis]
-- Library blis: BLAS_blis_LIBRARY-NOTFOUND
-- Could NOT find Atlas (missing: Atlas_CBLAS_INCLUDE_DIR Atlas_CLAPACK_INCLUDE_DIR Atlas_CBLAS_LIBRARY Atlas_BLAS_LIBRARY Atlas_LAPACK_LIBRARY)
-- Checking for [ptf77blas - atlas - gfortran]
-- Library ptf77blas: BLAS_ptf77blas_LIBRARY-NOTFOUND
-- Checking for []
-- Cannot find a library with BLAS API. Not using BLAS.
-- LAPACK requires BLAS
-- Cannot find a library with LAPACK API. Not using LAPACK.
disabling ROCM because NOT USE_ROCM is set
-- MIOpen not found. Compiling without MIOpen support
-- MKL_THREADING = OMP
-- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core - libiomp5md]
-- Library mkl_intel_lp64: not found
-- Checking for [mkl_intel - mkl_intel_thread - mkl_core - libiomp5md]
-- Library mkl_intel: not found
-- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core]
-- Library mkl_intel_lp64: not found
-- Checking for [mkl_intel - mkl_intel_thread - mkl_core]
-- Library mkl_intel: not found
-- Checking for [mkl_intel_lp64 - mkl_sequential - mkl_core]
-- Library mkl_intel_lp64: not found
-- Checking for [mkl_intel - mkl_sequential - mkl_core]
-- Library mkl_intel: not found
-- Checking for [mkl_intel_lp64 - mkl_core - libiomp5md - pthread]
-- Library mkl_intel_lp64: not found
-- Checking for [mkl_intel - mkl_core - libiomp5md - pthread]
-- Library mkl_intel: not found
-- Checking for [mkl_intel_lp64 - mkl_core - pthread]
-- Library mkl_intel_lp64: not found
-- Checking for [mkl_intel - mkl_core - pthread]
-- Library mkl_intel: not found
-- Checking for [mkl - guide - pthread - m]
-- Library mkl: not found
-- MKL library not found
-- Checking for [blis]
-- Library blis: BLAS_blis_LIBRARY-NOTFOUND
-- Checking for [Accelerate]
-- Library Accelerate: BLAS_Accelerate_LIBRARY-NOTFOUND
-- Checking for [vecLib]
-- Library vecLib: BLAS_vecLib_LIBRARY-NOTFOUND
-- Checking for [flexiblas]
-- Library flexiblas: BLAS_flexiblas_LIBRARY-NOTFOUND
-- Checking for [openblas]
-- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND
-- Checking for [openblas - pthread - m]
-- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND
-- Checking for [openblas - pthread - m - gomp]
-- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND
-- Checking for [libopenblas]
-- Library libopenblas: BLAS_libopenblas_LIBRARY-NOTFOUND
-- Checking for [goto2 - gfortran]
-- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND
-- Checking for [goto2 - gfortran - pthread]
-- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND
-- Checking for [acml - gfortran]
-- Library acml: BLAS_acml_LIBRARY-NOTFOUND
-- Checking for [blis]
-- Library blis: BLAS_blis_LIBRARY-NOTFOUND
-- Could NOT find Atlas (missing: Atlas_CBLAS_INCLUDE_DIR Atlas_CLAPACK_INCLUDE_DIR Atlas_CBLAS_LIBRARY Atlas_BLAS_LIBRARY Atlas_LAPACK_LIBRARY)
-- Checking for [ptf77blas - atlas - gfortran]
-- Library ptf77blas: BLAS_ptf77blas_LIBRARY-NOTFOUND
-- Checking for []
-- Cannot find a library with BLAS API. Not using BLAS.
-- MKLDNN_CPU_RUNTIME = OMP
-- DNNL_TARGET_ARCH: X64
-- DNNL_LIBRARY_NAME: dnnl
CMake Warning (dev) at D:/workspace/envs/py38_torch1.13.1_cuda11.6/Library/share/cmake-3.24/Modules/FindPackageHandleStandardArgs.cmake:438 (message):
The package name passed to `find_package_handle_standard_args` (OpenMP_C)
does not match the name of the calling package (OpenMP). This can lead to
problems in calling code that expects `find_package` result variables
(e.g., `_FOUND`) to follow a certain pattern.
Call Stack (most recent call first):
cmake/Modules/FindOpenMP.cmake:584 (find_package_handle_standard_args)
third_party/ideep/mkl-dnn/cmake/OpenMP.cmake:69 (find_package)
third_party/ideep/mkl-dnn/CMakeLists.txt:117 (include)
This warning is for project developers. Use -Wno-dev to suppress it.
CMake Warning (dev) at D:/workspace/envs/py38_torch1.13.1_cuda11.6/Library/share/cmake-3.24/Modules/FindPackageHandleStandardArgs.cmake:438 (message):
The package name passed to `find_package_handle_standard_args` (OpenMP_CXX)
does not match the name of the calling package (OpenMP). This can lead to
problems in calling code that expects `find_package` result variables
(e.g., `_FOUND`) to follow a certain pattern.
Call Stack (most recent call first):
cmake/Modules/FindOpenMP.cmake:584 (find_package_handle_standard_args)
third_party/ideep/mkl-dnn/cmake/OpenMP.cmake:69 (find_package)
third_party/ideep/mkl-dnn/CMakeLists.txt:117 (include)
This warning is for project developers. Use -Wno-dev to suppress it.
-- Could NOT find Doxyrest (missing: DOXYREST_EXECUTABLE)
-- Found PythonInterp: D:/workspace/envs/py38_torch1.13.1_cuda11.6/python.exe (found suitable version "3.8.18", minimum required is "2.7")
-- Could NOT find Sphinx (missing: SPHINX_EXECUTABLE)
-- Enabled workload: TRAINING
-- Enabled primitives: ALL
-- Enabled primitive CPU ISA: ALL
-- Enabled primitive GPU ISA: ALL
-- Primitive cache is enabled
-- Graph component is enabled
-- Found MKL-DNN: TRUE
-- Version: 10.1.0
-- Build type: Release
-- Using CPU-only version of Kineto
-- Configuring Kineto dependency:
-- KINETO_SOURCE_DIR = D:/workspace/codes/pytorch/third_party/kineto/libkineto
-- KINETO_BUILD_TESTS = OFF
-- KINETO_LIBRARY_TYPE = static
-- Found PythonInterp: D:/workspace/envs/py38_torch1.13.1_cuda11.6/python.exe (found version "3.8.18")
INFO CUDA_SOURCE_DIR =
INFO ROCM_SOURCE_DIR =
INFO CUPTI unavailable or disabled - not building GPU profilers
-- Kineto: FMT_SOURCE_DIR = D:/workspace/codes/pytorch/third_party/fmt
-- Kineto: FMT_INCLUDE_DIR = D:/workspace/codes/pytorch/third_party/fmt/include
INFO CUPTI_INCLUDE_DIR = /extras/CUPTI/include
INFO ROCTRACER_INCLUDE_DIR = /include/roctracer
INFO DYNOLOG_INCLUDE_DIR = D:/workspace/codes/pytorch/third_party/kineto/libkineto/third_party/dynolog/
INFO IPCFABRIC_INCLUDE_DIR = D:/workspace/codes/pytorch/third_party/kineto/libkineto/third_party/dynolog//dynolog/src/ipcfabric/
-- Configured Kineto (CPU)
CMake Warning (dev) at D:/workspace/envs/py38_torch1.13.1_cuda11.6/Library/share/cmake-3.24/Modules/CMakeDependentOption.cmake:89 (message):
Policy CMP0127 is not set: cmake_dependent_option() supports full Condition
Syntax. Run "cmake --help-policy CMP0127" for policy details. Use the
cmake_policy command to set the policy and suppress this warning.
Call Stack (most recent call first):
CMakeLists.txt:725 (cmake_dependent_option)
This warning is for project developers. Use -Wno-dev to suppress it.
--
-- Use the C++ compiler to compile (MI_USE_CXX=ON)
--
-- Library base name: mimalloc
-- Version : 1.8
-- Build type : release
-- C++ Compiler : C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.37.32822/bin/Hostx64/x64/cl.exe
-- Compiler flags : /Zc:__cplusplus
-- Compiler defines :
-- Link libraries : psapi;shell32;user32;advapi32;bcrypt
-- Build targets : static
--
-- Could NOT find Backtrace (missing: Backtrace_LIBRARY Backtrace_INCLUDE_DIR)
-- don't use NUMA
-- headers outputs:
-- sources outputs:
-- declarations_yaml outputs:
-- Using ATen parallel backend: OMP
AT_INSTALL_INCLUDE_DIR include/ATen/core
core header install: D:/workspace/codes/pytorch/build/aten/src/ATen/core/TensorBody.h
core header install: D:/workspace/codes/pytorch/build/aten/src/ATen/core/aten_interned_strings.h
core header install: D:/workspace/codes/pytorch/build/aten/src/ATen/core/enum_tag.h
-- Generating sources for unboxing kernels D:/workspace/envs/py38_torch1.13.1_cuda11.6/python.exe;-m;torchgen.gen_executorch;--source-path=D:/workspace/codes/pytorch/test/edge/../../test/edge;--install-dir=D:/workspace/codes/pytorch/build/out;--tags-path=D:/workspace/codes/pytorch/test/edge/../../aten/src/ATen/native/tags.yaml;--aten-yaml-path=D:/workspace/codes/pytorch/test/edge/../../aten/src/ATen/native/native_functions.yaml;--use-aten-lib;--op-selection-yaml-path=D:/workspace/codes/pytorch/test/edge/../../test/edge/selected_operators.yaml;--custom-ops-yaml-path=D:/workspace/codes/pytorch/test/edge/../../test/edge/custom_ops.yaml
CMake Warning (dev) at torch/CMakeLists.txt:389:
Syntax Warning in cmake code at column 107
Argument not separated from preceding token by whitespace.
This warning is for project developers. Use -Wno-dev to suppress it.
CMake Warning (dev) at torch/CMakeLists.txt:389:
Syntax Warning in cmake code at column 115
Argument not separated from preceding token by whitespace.
This warning is for project developers. Use -Wno-dev to suppress it.
-- Autodetected CUDA architecture(s): 8.6 6.1
-- Using Lib/site-packages as python relative installation path
CMake Warning at CMakeLists.txt:1113 (message):
Generated cmake files are only fully tested if one builds with system glog,
gflags, and protobuf. Other settings may generate files that are not well
tested.
-- somehow this is happening
--
-- ******** Summary ********
-- General:
-- CMake version : 3.24.1
-- CMake command : D:/workspace/envs/py38_torch1.13.1_cuda11.6/Library/bin/cmake.exe
-- System : Windows
-- C++ compiler : C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.37.32822/bin/Hostx64/x64/cl.exe
-- C++ compiler id : MSVC
-- C++ compiler version : 19.37.32825.0
-- Using ccache if found : OFF
-- CXX flags : /DWIN32 /D_WINDOWS /GR /EHsc /bigobj /FS -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOCUPTI -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE /utf-8 /wd4624 /wd4068 /wd4067 /wd4267 /wd4661 /wd4717 /wd4244 /wd4804 /wd4273
-- Build type : Release
-- Compile definitions : ONNX_ML=1;ONNXIFI_ENABLE_EXT=1;ONNX_NAMESPACE=onnx_torch;_CRT_SECURE_NO_DEPRECATE=1;USE_EXTERNAL_MZCRC;MINIZ_DISABLE_ZIP_READER_CRC32_CHECKS;WIN32_LEAN_AND_MEAN;_UCRT_LEGACY_INFINITY;NOMINMAX;USE_MIMALLOC;BUILD_NVFUSER
-- CMAKE_PREFIX_PATH : D:\workspace\envs\py38_torch1.13.1_cuda11.6\Lib\site-packages;C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6
-- CMAKE_INSTALL_PREFIX : D:/workspace/codes/pytorch/torch
-- USE_GOLD_LINKER : OFF
--
-- TORCH_VERSION : 2.1.0
-- BUILD_CAFFE2 : OFF
-- BUILD_CAFFE2_OPS : OFF
-- BUILD_STATIC_RUNTIME_BENCHMARK: OFF
-- BUILD_TENSOREXPR_BENCHMARK: OFF
-- BUILD_NVFUSER_BENCHMARK: OFF
-- BUILD_BINARY : OFF
-- BUILD_CUSTOM_PROTOBUF : ON
-- Link local protobuf : ON
-- BUILD_DOCS : OFF
-- BUILD_PYTHON : True
-- Python version : 3.8.18
-- Python executable : D:/workspace/envs/py38_torch1.13.1_cuda11.6/python.exe
-- Pythonlibs version : 3.8.18
-- Python library : D:/workspace/envs/py38_torch1.13.1_cuda11.6/libs/python38.lib
-- Python includes : D:/workspace/envs/py38_torch1.13.1_cuda11.6/include
-- Python site-packages: Lib/site-packages
-- BUILD_SHARED_LIBS : ON
-- CAFFE2_USE_MSVC_STATIC_RUNTIME : OFF
-- BUILD_TEST : True
-- BUILD_JNI : OFF
-- BUILD_MOBILE_AUTOGRAD : OFF
-- BUILD_LITE_INTERPRETER: OFF
-- INTERN_BUILD_MOBILE :
-- TRACING_BASED : OFF
-- USE_BLAS : 0
-- USE_LAPACK : 0
-- USE_ASAN : OFF
-- USE_TSAN : OFF
-- USE_CPP_CODE_COVERAGE : OFF
-- USE_CUDA : ON
-- Split CUDA :
-- CUDA static link : OFF
-- USE_CUDNN : ON
-- USE_EXPERIMENTAL_CUDNN_V8_API: ON
-- USE_CUSPARSELT : OFF
-- CUDA version : 11.6
-- USE_FLASH_ATTENTION : ON
-- cuDNN version : 8.9.5
-- CUDA root directory : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6
-- CUDA library : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6/lib/x64/cuda.lib
-- cudart library : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6/lib/x64/cudart.lib
-- cublas library : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6/lib/x64/cublas.lib
-- cufft library : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6/lib/x64/cufft.lib
-- curand library : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6/lib/x64/curand.lib
-- cusparse library : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6/lib/x64/cusparse.lib
-- cuDNN library : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6/lib/x64/cudnn.lib
-- nvrtc : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6/lib/x64/nvrtc.lib
-- CUDA include path : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6/include
-- NVCC executable : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6/bin/nvcc.exe
-- CUDA compiler : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6/bin/nvcc.exe
-- CUDA flags : -D_WINDOWS -Xcompiler=" /GR /EHsc" -Xcompiler /w -w -Xcompiler /FS -Xfatbin -compress-all -DONNX_NAMESPACE=onnx_torch --use-local-env -gencode arch=compute_86,code=sm_86 -gencode arch=compute_61,code=sm_61 -Xcudafe --diag_suppress=cc_clobber_ignored,--diag_suppress=set_but_not_used,--diag_suppress=field_without_dll_interface,--diag_suppress=base_class_has_different_dll_interface,--diag_suppress=dll_interface_conflict_none_assumed,--diag_suppress=dll_interface_conflict_dllexport_assumed,--diag_suppress=bad_friend_decl --Werror cross-execution-space-call --no-host-device-move-forward --expt-relaxed-constexpr --expt-extended-lambda -Xcompiler=/wd4819,/wd4503,/wd4190,/wd4244,/wd4251,/wd4275,/wd4522 -Wno-deprecated-gpu-targets --expt-extended-lambda -DCUB_WRAPPED_NAMESPACE=at_cuda_detail -DCUDA_HAS_FP16=1 -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__
-- CUDA host compiler :
-- CUDA --device-c : OFF
-- USE_TENSORRT : OFF
-- USE_ROCM : OFF
-- BUILD_NVFUSER : ON
-- USE_EIGEN_FOR_BLAS : ON
-- USE_FBGEMM : ON
-- USE_FAKELOWP : OFF
-- USE_KINETO : ON
-- USE_FFMPEG : OFF
-- USE_GFLAGS : OFF
-- USE_GLOG : OFF
-- USE_LEVELDB : OFF
-- USE_LITE_PROTO : OFF
-- USE_LMDB : OFF
-- USE_METAL : OFF
-- USE_PYTORCH_METAL : OFF
-- USE_PYTORCH_METAL_EXPORT : OFF
-- USE_MPS : OFF
-- USE_FFTW : OFF
-- USE_MKL : OFF
-- USE_MKLDNN : ON
-- USE_MKLDNN_ACL : OFF
-- USE_MKLDNN_CBLAS : OFF
-- USE_UCC : OFF
-- USE_ITT : ON
-- USE_NCCL : OFF
-- USE_NNPACK : OFF
-- USE_NUMPY : ON
-- USE_OBSERVERS : ON
-- USE_OPENCL : OFF
-- USE_OPENCV : OFF
-- USE_OPENMP : ON
-- USE_TBB : OFF
-- USE_MIMALLOC : ON
-- USE_VULKAN : OFF
-- USE_PROF : OFF
-- USE_QNNPACK : OFF
-- USE_PYTORCH_QNNPACK : OFF
-- USE_XNNPACK : ON
-- USE_REDIS : OFF
-- USE_ROCKSDB : OFF
-- USE_ZMQ : OFF
-- USE_DISTRIBUTED : ON
-- USE_MPI : OFF
-- USE_GLOO : ON
-- USE_GLOO_WITH_OPENSSL : OFF
-- USE_TENSORPIPE : OFF
-- Public Dependencies :
-- Private Dependencies : Threads::Threads;pthreadpool;cpuinfo;XNNPACK;fbgemm;ittnotify;fp16;caffe2::openmp;gloo;foxi_loader;fmt::fmt-header-only;kineto
-- Public CUDA Deps. : caffe2::cufft;caffe2::curand;caffe2::cublas
-- Private CUDA Deps. : torch::cudnn;gloo_cuda;C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.6/lib/x64/cudart_static.lib;CUDA::cusparse;CUDA::curand;CUDA::cufft;CUDA::cusolver;ATEN_CUDA_FILES_GEN_LIB
-- USE_COREML_DELEGATE : OFF
-- BUILD_LAZY_TS_BACKEND : ON
-- TORCH_DISABLE_GPU_ASSERTS : OFF
-- Configuring done
-- Generating done
The build goes well at the beginning. but when it comes caffe2 module, a fatal error occurs：
D:/workspace/codes/pytorch\c10/util/C++17.h(27): fatal error C1189: #error: You need C++17 to compile PyTorch
It seems that the cmake try to enable c++17 by adding -std=c++17 flag instead of /std:c++17:
FAILED: caffe2/CMakeFiles/torch_cuda.dir/__/aten/src/ATen/cuda/cub.cu.obj
C:\PROGRA~1\NVIDIA~2\CUDA\v11.6\bin\nvcc.exe -forward-unknown-to-host-compiler -DAT_PER_OPERATOR_HEADERS -DMINIZ_DISABLE_ZIP_READER_CRC32_CHECKS -DNOMINMAX -DONNXIFI_ENABLE_EXT=1 -DONNX_ML=1 -DONNX_NAMESPACE=onnx_torch -DTORCH_CUDA_BUILD_MAIN_LIB -DUSE_C10D_GLOO -DUSE_CUDA -DUSE_DISTRIBUTED -DUSE_EXPERIMENTAL_CUDNN_V8_API -DUSE_EXTERNAL_MZCRC -DUSE_FLASH_ATTENTION -DUSE_MIMALLOC -DWIN32_LEAN_AND_MEAN -D_CRT_SECURE_NO_DEPRECATE=1 -D_UCRT_LEGACY_INFINITY -Dtorch_cuda_EXPORTS -ID:\workspace\codes\pytorch\build\aten\src -ID:\workspace\codes\pytorch\aten\src -ID:\workspace\codes\pytorch\build -ID:\workspace\codes\pytorch -ID:\workspace\codes\pytorch\cmake\..\third_party\benchmark\include -ID:\workspace\codes\pytorch\third_party\onnx -ID:\workspace\codes\pytorch\build\third_party\onnx -ID:\workspace\codes\pytorch\third_party\foxi -ID:\workspace\codes\pytorch\build\third_party\foxi -ID:\workspace\codes\pytorch\third_party\mimalloc\include -ID:\workspace\codes\pytorch\aten\src\THC -ID:\workspace\codes\pytorch\aten\src\ATen\cuda -ID:\workspace\codes\pytorch\aten\src\ATen\..\..\..\third_party\cutlass\include -ID:\workspace\codes\pytorch\build\caffe2\aten\src -ID:\workspace\codes\pytorch\aten\src\ATen\.. -ID:\workspace\codes\pytorch\c10\cuda\..\.. -ID:\workspace\codes\pytorch\c10\.. -ID:\workspace\codes\pytorch\torch\csrc\api -ID:\workspace\codes\pytorch\torch\csrc\api\include -isystem=D:\workspace\codes\pytorch\build\third_party\gloo -isystem=D:\workspace\codes\pytorch\cmake\..\third_party\gloo -isystem=D:\workspace\codes\pytorch\cmake\..\third_party\googletest\googlemock\include -isystem=D:\workspace\codes\pytorch\cmake\..\third_party\googletest\googletest\include -isystem=D:\workspace\codes\pytorch\third_party\protobuf\src -isystem=D:\workspace\codes\pytorch\third_party\XNNPACK\include -isystem=D:\workspace\codes\pytorch\third_party\ittapi\include -isystem=D:\workspace\codes\pytorch\cmake\..\third_party\eigen -isystem="C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6\include" -isystem=D:\workspace\codes\pytorch\third_party\ideep\mkl-dnn\include\oneapi\dnnl -isystem=D:\workspace\codes\pytorch\third_party\ideep\include -isystem="C:\Program Files\NVIDIA Corporation\NvToolsExt\include" -isystem=D:\workspace\codes\pytorch\cmake\..\third_party\cudnn_frontend\include -D_WINDOWS -Xcompiler=" /GR /EHsc" -Xcompiler /w -w -Xcompiler /FS -Xfatbin -compress-all -DONNX_NAMESPACE=onnx_torch --use-local-env -gencode arch=compute_86,code=sm_86 -gencode arch=compute_61,code=sm_61 -Xcudafe --diag_suppress=cc_clobber_ignored,--diag_suppress=set_but_not_used,--diag_suppress=field_without_dll_interface,--diag_suppress=base_class_has_different_dll_interface,--diag_suppress=dll_interface_conflict_none_assumed,--diag_suppress=dll_interface_conflict_dllexport_assumed,--diag_suppress=bad_friend_decl --Werror cross-execution-space-call --no-host-device-move-forward --expt-relaxed-constexpr --expt-extended-lambda -Xcompiler=/wd4819,/wd4503,/wd4190,/wd4244,/wd4251,/wd4275,/wd4522 -Wno-deprecated-gpu-targets --expt-extended-lambda -DCUB_WRAPPED_NAMESPACE=at_cuda_detail -DCUDA_HAS_FP16=1 -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -Xcompiler="-O2 -Ob2" -DNDEBUG -Xcompiler /MD -Xcompiler=-MD -DCAFFE2_USE_GLOO -DTH_HAVE_THREAD -Xcompiler= -std=c++17 -MD -MT caffe2\CMakeFiles\torch_cuda.dir\__\aten\src\ATen\cuda\cub.cu.obj -MF caffe2\CMakeFiles\torch_cuda.dir\__\aten\src\ATen\cuda\cub.cu.obj.d -x cu -c D:\workspace\codes\pytorch\aten\src\ATen\cuda\cub.cu -o caffe2\CMakeFiles\torch_cuda.dir\__\aten\src\ATen\cuda\cub.cu.obj -Xcompiler=-Fdcaffe2\CMakeFiles\torch_cuda.dir\,-FS
cl: 命令行 warning D9002 :忽略未知选项“-std=c++17”
So I try add /std:c++17 to ${CMAKE_CXX_FLAGS} . But this seems not work.
Is there any configs are missed in this process? Hope someone can give me some help or tips.
Thanks.