MooreThreads / torch_musa

torch_musa is an open source repository based on PyTorch, which can make full use of the super computing power of MooreThreads graphics cards.
Other
280 stars 17 forks source link

关于找不到模块,mthread-gmi找不到驱动,及编译报错问题 #10

Open fanjiaqi1995 opened 1 year ago

fanjiaqi1995 commented 1 year ago

在自己源码编译时,遇到了一些报错问题。同时发现在win10 22h2的wsl2 ubuntu2004下,会有一些驱动问题: 求问有没有相关的问题和解决文档

无法正常找到模块 [root@DESKTOP-6VT7GDB-LFS:torch_musa]# modprobe mtgpu modprobe: FATAL: Module mtgpu not found in directory /lib/modules/5.15.90.1-microsoft-standard-WSL2

mthreads-gmi找不到网卡: [root@DESKTOP-6VT7GDB-LFS:~]# mthreads-gmi Error: there no exist gpu device

但是能找到驱动: [root@DESKTOP-6VT7GDB-LFS:~]# dpkg --list musa Desired=Unknown/Install/Remove/Purge/Hold | Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend |/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad) ||/ Name Version Architecture Description +++-==============-============-============-================================= ii musa 2.1.0-Ubuntu amd64 Moore Threads MUSA driver

[root@DESKTOP-6VT7GDB-LFS:torch_musa]# dpkg -l | grep container-toolkit rc mt-container-toolkit 1.5.0-1 amd64 MT Container Toolkit

[root@DESKTOP-6VT7GDB-LFS:torch_musa]# dpkg -l | grep musa ii musa 2.1.0-Ubuntu amd64 Moore Threads MUSA driver

下面是编译报错代码: [root@DESKTOP-6VT7GDB-LFS:torch_musa]# python setup.py install cmake -DBUILD_TEST=True -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/root/torch_musa/torch_musa -DCMAKE_PREFIX_PATH=/usr/lib/python3.8/site-packages -DGENERATED_PORTING_DIR=generated_cuda_compatible -DPYTHON_INCLUDE_DIR=/usr/include/python3.8 -DUSE_PYTHON=True /root/torch_musa CMake Warning at CMakeLists.txt:181 (find_package): By not providing "FindMCCL.cmake" in CMAKE_MODULE_PATH this project has asked CMake to find a package configuration file provided by "MCCL", but CMake did not find one.

Could not find a package configuration file provided by "MCCL" with any of the following names:

MCCLConfig.cmake
mccl-config.cmake

Add the installation prefix of "MCCL" to CMAKE_PREFIX_PATH or set "MCCL_DIR" to a directory containing one of the above files. If "MCCL" provides a separate development package or SDK, be sure it has been installed.

CMake Warning at CMakeLists.txt:186 (message): NO MCCL FOUND?

-- CMake version : 3.25.2 -- CMake command : /usr/local/lib/python3.8/dist-packages/cmake/data/bin/cmake -- System : Linux -- C++ compiler : /usr/bin/c++ -- C++ compiler id : GNU -- C++ compiler version : 9.4.0 -- CXX flags : -O2 -fPIC -Wall -Wextra -Werror -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-sign-compare -Wno-missing-field-initializers -Wno-non-template-friend -Wno-comment -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -- CMAKE_PREFIX_PATH : /usr/lib/python3.8/site-packages -- CMAKE_INSTALL_PREFIX : /root/torch_musa/torch_musa -- USE_PYTHON : True -- BUILD_TEST : True -- BUILD_TYPE : Release -- MUSAARCH : mp -- PYTORCH_SOURCE_PATH : /root/pytorch -- PYTORCH_HEADERS_PATH : /root/torch_musa/build/generated_cuda_compatible/include -- MUDNN PATH : /usr/local/musa -- MUDNN_LIBRARIES : /usr/local/musa/lib/libmudnn.so -- MUSA TOOLKITS PATH : /usr/local/musa -- MUSAToolkits_LIBRARIES : /usr/local/musa/lib/libmusart.so -- Configuring done -- Generating done -- Build files have been written to: /root/torch_musa/build cmake --build . --target install --config Release -- -j 32 [ 1%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir/core/musa_kernels_generated_Sleep.mu.o [ 2%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/musa/musa_kernels_generated_cub.mu.o [ 3%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/musa/detail/musa_kernels_generated_IndexUtils.mu.o [ 3%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/musa/musa_kernels_generated_cub-RadixSortPairs.mu.o [ 4%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationGluKernel.mu.o [ 5%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationGeluKernel.mu.o [ 7%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationHardtanhKernel.mu.o [ 7%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationHardsigmoidKernel.mu.o [ 8%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationHardswishKernel.mu.o [ 8%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationLogSigmoidKernel.mu.o [ 9%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationPreluKernel.mu.o [ 10%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_BinaryBitwiseOpsKernels.mu.o [ 11%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationSoftplusKernel.mu.o [ 12%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_BinaryMiscOpsKernels.mu.o [ 13%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_BinaryMiscBackwardOpsKernels.mu.o [ 14%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_CompareKernels.mu.o [ 15%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_DilatedMaxPool3d.mu.o [ 15%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_DistributionExponentialKernel.mu.o [ 16%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_DistributionBernoulli.mu.o [ 17%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_DistributionUniform.mu.o [ 18%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_DistributionNormal.mu.o [ 19%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_EmbeddingBackwardKernel.mu.o [ 20%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_IndexKernel.mu.o [ 21%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_FillKernel.mu.o [ 22%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_GridSampler.mu.o [ 23%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_LegacyThrustHelpers.mu.o [ 24%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_Indexing.mu.o [ 25%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_PointwiseOpsKernel.mu.o [ 25%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_LossCTC.mu.o [ 26%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_RangeFactories.mu.o [ 27%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ReduceMaxValuesKernel.mu.o [ 28%] Building MCC (Device) object torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ReduceNormKernel.mu.o /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/ActivationPreluKernel.mu:8:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

/root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/ActivationGluKernel.mu:8:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

/root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/ActivationHardsigmoidKernel.mu:8:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

/root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/ActivationHardtanhKernel.mu:8:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

/root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/ActivationHardswishKernel.mu:8:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

/root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/ActivationSoftplusKernel.mu:8:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

/root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/ActivationGeluKernel.mu:8:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

/root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/ActivationLogSigmoidKernel.mu:8:10: fatal error: 'thrust/tuple.h' file not found In file included from In file included from #include <thrust/tuple.h> ^~~~/root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/musa/cub-RadixSortPairs.mu: 3/root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/musa/cub.mu:: 2: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/musa/cub.muh:11: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/musa/cub_definitions.muh:8/root/torch_musa/build/generated_cuda_compatible/include/ATen/musa/cub.muh::1011: : fatal error: 'cub/version.cuh' file not found /root/torch_musa/build/generated_cuda_compatible/include/ATen/musa/cub_definitions.muh#include <cub/version.cuh> ^~~~~ :8:10: fatal error: 'cub/version.cuh' file not found

include <cub/version.cuh>

     ^~~~~~~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/DistributionBernoulli.mu:3: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/musa/MUSA_PORT_ApplyUtils.muh:3: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/musa/ApplyGridUtils.muh:1: /root/torch_musa/torch_musa/csrc/aten/musa/MUSAContext.h:8:10: fatal error: 'musparse.h' file not found

include

     ^~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/FillKernel.mu:3: /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/Loops.muh:10:10: fatal error: 'thrust/tuple.h' file not found In file included from In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/CompareKernels.mu:6: /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/Loops.muh:10:10: fatal error: 'thrust/tuple.h' file not found In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/BinaryBitwiseOpsKernels.mu:4: /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/Loops.muh:10:10: fatal error: 'thrust/tuple.h' file not found In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/BinaryMiscOpsKernels.mu:4: /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/Loops.muh:10:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

/root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/BinaryMiscBackwardOpsKernels.mu:10: /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/Loops.muh:10:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/PointwiseOpsKernel.mu:5: /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/Loops.muh:10:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/IndexKernel.mu:9: /root/torch_musa/torch_musa/csrc/aten/musa/MUSAContext.h:8:10: fatal error: 'musparse.h' file not found

include

     ^~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/DistributionNormal.mu:4: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/DistributionTemplates.h:7: /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/Loops.muh:10:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/DistributionUniform.mu:4: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/DistributionTemplates.h:7: /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/Loops.muh:10:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/DistributionExponentialKernel.mu:4: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/DistributionTemplates.h:7: /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/Loops.muh:10:10: fatal error: 'thrust/tuple.h' file not found

include <thrust/tuple.h>

     ^~~~~~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/ReduceNormKernel.mu:4: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/Reduce.muh:5: /root/torch_musa/torch_musa/csrc/aten/musa/MUSAContext.h:8:10: fatal error: 'musparse.h' file not found

include

     ^~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/ReduceMaxValuesKernel.mu:12: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/Reduce.muh:5: /root/torch_musa/torch_musa/csrc/aten/musa/MUSAContext.h:8:10: fatal error: 'musparse.h' file not found

include

     ^~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/GridSampler.mu:7: /root/torch_musa/torch_musa/csrc/aten/musa/MUSAContext.h:8:10: fatal error: 'musparse.h' file not found

include

     ^~~~~~~~~~~~

1 error generated. 1 error generated. 1 error generated. 1 error generated. 1 error generated. 1 error generated. 1 error generated. 1 error generated. CMake Error at musa_kernels_generated_cub-RadixSortPairs.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/musa/./musa_kernels_generated_cub-RadixSortPairs.mu.o

1 error generated.

1 error generated. 1 error generated. CMake Error at musa_kernels_generated_cub.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/musa/./musa_kernels_generated_cub.mu.o

1 error generated. CMake Error at musa_kernels_generated_CompareKernels.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_CompareKernels.mu.o

CMake Error at musa_kernels_generated_BinaryMiscBackwardOpsKernels.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_BinaryMiscBackwardOpsKernels.mu.o

1 error generated. make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:77: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/musa/musa_kernels_generated_cub-RadixSortPairs.mu.o] Error 1 make[2]: Waiting for unfinished jobs.... 1make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:84: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/musa/musa_kernels_generated_cub.mu.o] Error 1 error generated. make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:566: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_CompareKernels.mu.o] Error 1 1 error generated. 1 error generated. 1 error generated. make[2]: *** [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:552: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_BinaryMiscBackwardOpsKernels.mu.o] Error 1 1 error generated. CMake Error at musa_kernels_generated_FillKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_FillKernel.mu.o

CMake Error at musa_kernels_generated_ActivationSoftplusKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_ActivationSoftplusKernel.mu.o

CMake Error at musa_kernels_generated_BinaryBitwiseOpsKernels.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_BinaryBitwiseOpsKernels.mu.o

CMake Error at musa_kernels_generated_GridSampler.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_GridSampler.mu.o

CMake Error at musa_kernels_generated_ActivationGeluKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_ActivationGeluKernel.mu.o

CMake Error at musa_kernels_generated_ActivationPreluKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_ActivationPreluKernel.mu.o

make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:545: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_BinaryBitwiseOpsKernels.mu.o] Error 1 make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:622: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_GridSampler.mu.o] Error 1 make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:538: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationSoftplusKernel.mu.o] Error 1 make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:615: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_FillKernel.mu.o] Error 1 CMake Error at musa_kernels_generated_ActivationGluKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_ActivationGluKernel.mu.o

1 error generated. make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:489: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationGeluKernel.mu.o] Error 1 make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:531: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationPreluKernel.mu.o] Error 1 CMake Error at musa_kernels_generated_ActivationHardtanhKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_ActivationHardtanhKernel.mu.o

make[2]: *** [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:496: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationGluKernel.mu.o] Error 1 CMake Error at musa_kernels_generated_ActivationHardsigmoidKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_ActivationHardsigmoidKernel.mu.o

make[2]: *** [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:517: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationHardtanhKernel.mu.o] Error 1 CMake Error at musa_kernels_generated_BinaryMiscOpsKernels.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_BinaryMiscOpsKernels.mu.o

CMake Error at musa_kernels_generated_ReduceMaxValuesKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_ReduceMaxValuesKernel.mu.o

CMake Error at musa_kernels_generated_PointwiseOpsKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_PointwiseOpsKernel.mu.o

CMake Error at musa_kernels_generated_ActivationLogSigmoidKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_ActivationLogSigmoidKernel.mu.o

make[2]: *** [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:503: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationHardsigmoidKernel.mu.o] Error 1 CMake Error at musa_kernels_generated_ActivationHardswishKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_ActivationHardswishKernel.mu.o

make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:559: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_BinaryMiscOpsKernels.mu.o] Error 1 make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:524: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationLogSigmoidKernel.mu.o] Error 1 make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:657: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_PointwiseOpsKernel.mu.o] Error 1 make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:671: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ReduceMaxValuesKernel.mu.o] Error 1 CMake Error at musa_kernels_generated_ReduceNormKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_ReduceNormKernel.mu.o

make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:510: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ActivationHardswishKernel.mu.o] Error 1 make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:678: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_ReduceNormKernel.mu.o] Error 1 1 error generated. 1 error generated. 1 error generated. 1 error generated. CMake Error at musa_kernels_generated_IndexKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_IndexKernel.mu.o

make[2]: *** [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:629: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_IndexKernel.mu.o] Error 1 CMake Error at musa_kernels_generated_DistributionUniform.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_DistributionUniform.mu.o

CMake Error at musa_kernels_generated_DistributionExponentialKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_DistributionExponentialKernel.mu.o

CMake Error at musa_kernels_generated_DistributionNormal.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_DistributionNormal.mu.o

make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:587: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_DistributionExponentialKernel.mu.o] Error 1 make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:601: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_DistributionUniform.mu.o] Error 1 make[2]: *** [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:594: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_DistributionNormal.mu.o] Error 1 1 error generated. CMake Error at musa_kernels_generated_DistributionBernoulli.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_DistributionBernoulli.mu.o

make[2]: *** [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:580: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_DistributionBernoulli.mu.o] Error 1 In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/LegacyThrustHelpers.mu:4: /root/torch_musa/build/generated_cuda_compatible/include/ATen/musa/cub_definitions.muh:8:10: fatal error: 'cub/version.cuh' file not found

include <cub/version.cuh>

     ^~~~~~~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/RangeFactories.mu:6: /root/torch_musa/torch_musa/csrc/aten/musa/MUSAContext.h:8:10: fatal error: 'musparse.h' file not found

include

     ^~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/EmbeddingBackwardKernel.mu:2: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/musa/EmbeddingBackwardKernel.muh:4: /root/torch_musa/torch_musa/csrc/aten/musa/MUSAContext.h:8:10: fatal error: 'musparse.h' file not found

include

     ^~~~~~~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/Indexing.mu:2: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/TensorAdvancedIndexing.h:9: /root/torch_musa/build/generated_cuda_compatible/include/ATen/native/cpu/radix_sort.h:25:10: fatal error: 'omp.h' file not found

include

     ^~~~~~~

In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/DilatedMaxPool3d.mu:10: /root/torch_musa/torch_musa/csrc/aten/musa/MUSAContext.h:8:10: fatal error: 'musparse.h' file not found

include

     ^~~~~~~~~~~~

1 error generated. CMake Error at musa_kernels_generated_EmbeddingBackwardKernel.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_EmbeddingBackwardKernel.mu.o

make[2]: *** [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:608: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_EmbeddingBackwardKernel.mu.o] Error 1 In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/native/musa/LossCTC.mu:18: /root/torch_musa/torch_musa/csrc/aten/musa/MUSAContext.h:8:10: fatal error: 'musparse.h' file not found

include

     ^~~~~~~~~~~~

1 error generated. CMake Error at musa_kernels_generated_LegacyThrustHelpers.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_LegacyThrustHelpers.mu.o

make[2]: *** [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:643: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_LegacyThrustHelpers.mu.o] Error 1 In file included from /root/torch_musa/build/generated_cuda_compatible/aten/src/ATen/musa/detail/IndexUtils.mu:1: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/musa/detail/IndexUtils.muh:3: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/core/TensorBase.h:6: In file included from /root/torch_musa/build/generated_cuda_compatible/include/c10/core/ScalarType.h:5: In file included from /root/torch_musa/build/generated_cuda_compatible/include/c10/util/Half.h:15: /root/torch_musa/build/generated_cuda_compatible/include/c10/util/complex.h:8:10: fatal error: 'thrust/complex.h' file not found

include <thrust/complex.h>

     ^~~~~~~~~~~~~~~~~~

In file included from /root/torch_musa/torch_musa/csrc/core/Sleep.mu:2: In file included from /root/torch_musa/torch_musa/csrc/core/MUSAStream.h:10: In file included from /root/torch_musa/torch_musa/csrc/aten/utils/Utils.h:4: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/Dispatch.h:3: In file included from /root/torch_musa/build/generated_cuda_compatible/include/ATen/core/DeprecatedTypeProperties.h:4: In file included from /root/torch_musa/build/generated_cuda_compatible/include/c10/core/ScalarType.h:5: In file included from /root/torch_musa/build/generated_cuda_compatible/include/c10/util/Half.h:15: /root/torch_musa/build/generated_cuda_compatible/include/c10/util/complex.h:8:10: fatal error: 'thrust/complex.h' file not found

include <thrust/complex.h>

     ^~~~~~~~~~~~~~~~~~

1 error generated when compiling for mp_10. CMake Error at musa_kernels_generated_IndexUtils.mu.o.Release.cmake:283 (message): Error generating file /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/musa/detail/./musa_kernels_generated_IndexUtils.mu.o

make[2]: *** [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:482: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/musa/detail/musa_kernels_generated_IndexUtils.mu.o] Error 1 1 error generated. 1 error generated. CMake Error at musa_kernels_generated_DilatedMaxPool3d.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_DilatedMaxPool3d.mu.o

CMake Error at musa_kernels_generated_LossCTC.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_LossCTC.mu.o

1 error generated. make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:573: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_DilatedMaxPool3d.mu.o] Error 1 make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:650: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_LossCTC.mu.o] Error 1 CMake Error at musa_kernels_generated_RangeFactories.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_RangeFactories.mu.o

make[2]: *** [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:664: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_RangeFactories.mu.o] Error 1 1 error generated. CMake Error at musa_kernels_generated_Indexing.mu.o.Release.cmake:222 (message): Error generating /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/./musa_kernels_generated_Indexing.mu.o

make[2]: *** [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:636: torch_musa/csrc/CMakeFiles/musa_kernels.dir///build/generated_cuda_compatible/aten/src/ATen/native/musa/musa_kernels_generated_Indexing.mu.o] Error 1 1 error generated when compiling for mp_10. CMake Error at musa_kernels_generated_Sleep.mu.o.Release.cmake:283 (message): Error generating file /root/torch_musa/build/torch_musa/csrc/CMakeFiles/musa_kernels.dir/core/./musa_kernels_generated_Sleep.mu.o

make[2]: [torch_musa/csrc/CMakeFiles/musa_kernels.dir/build.make:1317: torch_musa/csrc/CMakeFiles/musa_kernels.dir/core/musa_kernels_generated_Sleep.mu.o] Error 1 make[1]: [CMakeFiles/Makefile2:125: torch_musa/csrc/CMakeFiles/musa_kernels.dir/all] Error 2 make: *** [Makefile:136: all] Error 2

yaowang-mt commented 1 year ago

你好,现在源码编译的方式存在一些问题,有一些底层数学库尚未发布。你可以尝试一下通过镜像使用torch musa么?

fanjiaqi1995 commented 1 year ago

你好,现在docker运行它也有一定问题。问题内容如下:

2023-08-01 15:03:52 (py38) @.***:/home# python 2023-08-01 15:03:52 Python 3.8.17 (default, Jul 5 2023, 21:04:15) 2023-08-01 15:03:52 [GCC 11.2.0] :: Anaconda, Inc. on linux 2023-08-01 15:03:52 Type "help", "copyright", "credits" or "license" for more information. 2023-08-01 15:03:58 >>> import torch 2023-08-01 15:04:04 >>> import torch_musa 2023-08-01 15:04:04 Traceback (most recent call last): 2023-08-01 15:04:04 File "/opt/conda/envs/py38/lib/python3.8/site-packages/torch_musa-2.0.0-py3.8-linux-x86_64.egg/torch_musa/init.py", line 27, in 2023-08-01 15:04:04 import torch_musa._MUSAC 2023-08-01 15:04:04 ImportError: libsrv_um_MUSA.so: cannot open shared object file: No such file or directory 2023-08-01 15:04:04 2023-08-01 15:04:04 The above exception was the direct cause of the following exception: 2023-08-01 15:04:04 2023-08-01 15:04:04 Traceback (most recent call last): 2023-08-01 15:04:04 File "", line 1, in 2023-08-01 15:04:04 File "/opt/conda/envs/py38/lib/python3.8/site-packages/torch_musa-2.0.0-py3.8-linux-x86_64.egg/torch_musa/init.py", line 29, in 2023-08-01 15:04:04 raise ImportError("Please try running Python from a different directory!") from err 2023-08-01 15:04:04 ImportError: Please try running Python from a different directory! 2023-08-01 15:04:25 >>> ls 2023-08-01 15:04:25 Traceback (most recent call last): 2023-08-01 15:04:25 File "", line 1, in 2023-08-01 15:04:25 NameError: name 'ls' is not defined 2023-08-01 15:04:29 >>> exit()

fjq

@. | ---- 回复的原邮件 ---- | 发件人 | @.> | | 发送日期 | 2023年8月3日 10:01 | | 收件人 | @.> | | 抄送人 | @.> , @.***> | | 主题 | Re: [MooreThreads/torch_musa] 关于找不到模块,mthread-gmi找不到驱动,及编译报错问题 (Issue #10) |

你好,现在源码编译的方式存在一些问题,有一些底层数学库尚未发布。你可以尝试一下通过镜像使用torch musa么?

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

yaowang-mt commented 1 year ago

https://mcconline.mthreads.com/software/1?id=1#%E5%AE%89%E8%A3%85%E6%91%A9%E5%B0%94%E7%BA%BF%E7%A8%8B%E5%AE%B9%E5%99%A8%E8%BF%90%E8%A1%8C%E6%97%B6%E5%A5%97%E4%BB%B6

绑定摩尔线程容器运行时到 Docker,设置默认的容器运行时为 mthreads 并重启 Docker daemon:
$ (cd /usr/bin/musa && sudo ./docker setup $PWD)

试一下这个步骤?

fanjiaqi1995 commented 1 year ago

这个文档我看到了,但是有个问题,我的系统找不到摩尔线程的卡。目前暂时还不支持在wsl2中运行吗 我可能知道一个问题了。我没安装虚拟化插件。我先试试

@.***:~]# mthreads-gmi Error: there no exist gpu device

fjq

@. | ---- 回复的原邮件 ---- | 发件人 | @.> | | 发送日期 | 2023年8月3日 11:19 | | 收件人 | @.> | | 抄送人 | @.> , @.***> | | 主题 | Re: [MooreThreads/torch_musa] 关于找不到模块,mthread-gmi找不到驱动,及编译报错问题 (Issue #10) |

https://mcconline.mthreads.com/software/1?id=1#%E5%AE%89%E8%A3%85%E6%91%A9%E5%B0%94%E7%BA%BF%E7%A8%8B%E5%AE%B9%E5%99%A8%E8%BF%90%E8%A1%8C%E6%97%B6%E5%A5%97%E4%BB%B6

绑定摩尔线程容器运行时到 Docker,设置默认的容器运行时为 mthreads 并重启 Docker daemon: $ (cd /usr/bin/musa && sudo ./docker setup $PWD)

试一下这个步骤?

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

fanjiaqi1995 commented 1 year ago

在安装驱动时会遇到一些问题。在sgpu和musa_2.1.0-Ubuntu_amd64.deb中会遇到模块问题

这是安装sgpu:

Setting up sgpu-dkms (1.1.1) ... Loading new sgpu-1.1.1 DKMS files... Building for 5.4.0-42-generic Building for architecture x86_64 Module build for kernel 5.4.0-42-generic was skipped since the kernel headers for this kernel does not seem to be installed. update-initramfs: Generating /boot/initrd.img-5.4.0-42-generic W: missing /lib/modules/5.4.0-42-generic W: Ensure all necessary drivers are built into the linux image! depmod: ERROR: could not open directory /lib/modules/5.4.0-42-generic: No such file or directory depmod: FATAL: could not search modules: No such file or directory cat: /var/tmp/mkinitramfs_WIcj7Q/lib/modules/5.4.0-42-generic/modules.builtin: No such file or directory I: The initramfs will attempt to resume from /dev/sdb I: (UUID=48615a19-7af3-420a-b18e-5a0307903064) I: Set the RESUME variable to override this. depmod: WARNING: could not open modules.order at /var/tmp/mkinitramfs_WIcj7Q/lib/modules/5.4.0-42-generic: No such file or directory depmod: WARNING: could not open modules.builtin at /var/tmp/mkinitramfs_WIcj7Q/lib/modules/5.4.0-42-generic: No such file or directory the system needs to be restarted Processing triggers for libc-bin (2.31-0ubuntu9.9) ... /sbin/ldconfig.real: /usr/lib/wsl/lib/libcuda.so.1 is not a symbolic link /sbin/ldconfig.real: /usr/local/lib/libtbbbind_2_5.so.3 is not a symbolic link /sbin/ldconfig.real: /usr/local/lib/libtbbbind.so.3 is not a symbolic link /sbin/ldconfig.real: /usr/local/lib/libtbbmalloc.so.2 is not a symbolic link /sbin/ldconfig.real: /usr/local/lib/libtbb.so.12 is not a symbolic link /sbin/ldconfig.real: /usr/local/lib/libtbbmalloc_proxy.so.2 is not a symbolic link /sbin/ldconfig.real: /usr/local/lib/libtbbbind_2_0.so.3 is not a symbolic link @.***:~]# ls /lib/modules 5.4.0-153-generic 5.4.0-155-generic

这是安装驱动: Running the post_install script: depmod: ERROR: could not open directory /lib/modules/5.15.90.1-microsoft-standard-WSL2: No such file or directory depmod: FATAL: could not search modules: No such file or directory modprobe: FATAL: Module mtgpu not found in directory /lib/modules/5.15.90.1-microsoft-standard-WSL2

fjq

@. | ---- 回复的原邮件 ---- | 发件人 | @.> | | 发送日期 | 2023年8月3日 11:19 | | 收件人 | @.> | | 抄送人 | @.> , @.***> | | 主题 | Re: [MooreThreads/torch_musa] 关于找不到模块,mthread-gmi找不到驱动,及编译报错问题 (Issue #10) |

https://mcconline.mthreads.com/software/1?id=1#%E5%AE%89%E8%A3%85%E6%91%A9%E5%B0%94%E7%BA%BF%E7%A8%8B%E5%AE%B9%E5%99%A8%E8%BF%90%E8%A1%8C%E6%97%B6%E5%A5%97%E4%BB%B6

绑定摩尔线程容器运行时到 Docker,设置默认的容器运行时为 mthreads 并重启 Docker daemon: $ (cd /usr/bin/musa && sudo ./docker setup $PWD)

试一下这个步骤?

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

yaowang-mt commented 1 year ago

你好,目前整个软件栈对wsl2支持比较有限,建议使用baremetal的linux系统。

uniartisan commented 8 months ago

您好,我想询问目前s80对于wsl2的支持情况。