You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mxnet.apache.org by GitBox <gi...@apache.org> on 2020/01/16 03:18:38 UTC

[GitHub] [incubator-mxnet] cyrusbehr opened a new issue #17338: Unable to build mxnet with USE_OPENMP=OFF

cyrusbehr opened a new issue #17338: Unable to build mxnet with USE_OPENMP=OFF
URL: https://github.com/apache/incubator-mxnet/issues/17338
 
 
   I need to compile mxnet to run in single thread mode. From what I have found, using the environment variables `MXNET_ENGINE_TYPE=NaiveEngine` and `OMP_NUM_THREADS=1` and using `omp_set_num_threads(1);` in my c++ code gets the number of threads down to 2. To get the number of threads down to 1, I need to compile the library with `USE_OPENMP=OFF` (discussed more in [this](https://github.com/apache/incubator-mxnet/issues/15275) issue). In the past, this has worked successfully for me. However, when I tried doing this again, I get an error during compilation. I am using MKLDNN for the backend.
   
   Here are my steps:
   ```
   git clone --recursive https://github.com/apache/incubator-mxnet.git mxnet
   cd mxnet
   mkdir build 
   cd build
   
   cmake -DUSE_CPP_PACKAGE=1 -DUSE_CUDA=0 -DUSE_MKL_IF_AVAILABLE=1 -DUSE_OPENCV=0 -DUSE_LAPACK=0 -DUSE_OPENMP=0 \
     -DMKL_INCLUDE_DIR=/opt/intel/compilers_and_libraries/linux/mkl/include -DMKL_RT_LIBRARY=/opt/intel/compilers_and_libraries/linux/mkl/lib/intel64/libmkl_rt.so ..
   
   make -j16
   ```
   
   Here is the cmake output:
   ```
   -- The C compiler identification is GNU 7.4.0
   -- The CXX compiler identification is GNU 7.4.0
   -- Check for working C compiler: /usr/bin/cc
   -- Check for working C compiler: /usr/bin/cc -- works
   -- Detecting C compiler ABI info
   -- Detecting C compiler ABI info - done
   -- Detecting C compile features
   -- Detecting C compile features - done
   -- Check for working CXX compiler: /usr/bin/c++
   -- Check for working CXX compiler: /usr/bin/c++ -- works
   -- Detecting CXX compiler ABI info
   -- Detecting CXX compiler ABI info - done
   -- Detecting CXX compile features
   -- Detecting CXX compile features - done
   -- CMAKE_CROSSCOMPILING FALSE
   -- CMAKE_HOST_SYSTEM_PROCESSOR x86_64
   -- CMAKE_SYSTEM_PROCESSOR x86_64
   -- CMAKE_SYSTEM_NAME Linux
   -- CMake version '3.13.3' using generator 'Unix Makefiles'
   -- Performing Test SUPPORT_CXX11
   -- Performing Test SUPPORT_CXX11 - Success
   -- Performing Test SUPPORT_CXX0X
   -- Performing Test SUPPORT_CXX0X - Success
   -- Performing Test SUPPORT_MSSE3
   -- Performing Test SUPPORT_MSSE3 - Success
   -- Performing Test SUPPORT_MSSE2
   -- Performing Test SUPPORT_MSSE2 - Success
   -- Determining F16C support
   -- Performing Test COMPILER_SUPPORT_MF16C
   m-- Performing Test COMPILER_SUPPORT_MF16C - Success
   -- F16C enabled
   -- CMAKE_BUILD_TYPE is unset, defaulting to Release
   -- MKL-DNN compat: set DNNL_BUILD_EXAMPLES to MKLDNN_BUILD_EXAMPLES with value `OFF`
   -- MKL-DNN compat: set DNNL_BUILD_TESTS to MKLDNN_BUILD_TESTS with value `OFF`
   -- MKL-DNN compat: set DNNL_ENABLE_JIT_PROFILING to MKLDNN_ENABLE_JIT_PROFILING with value `OFF`
   -- MKL-DNN compat: set DNNL_LIBRARY_TYPE to MKLDNN_LIBRARY_TYPE with value `STATIC`
   -- MKL-DNN compat: set DNNL_ARCH_OPT_FLAGS to MKLDNN_ARCH_OPT_FLAGS with value ``
   -- Looking for pthread.h
   a-- Looking for pthread.h - found
   -- Looking for pthread_create
   ke-- Looking for pthread_create - not found
   -- Looking for pthread_create in pthreads
    -- Looking for pthread_create in pthreads - not found
   -- Looking for pthread_create in pthread
   --- Looking for pthread_create in pthread - found
   -- Found Threads: TRUE  
   j-- Found OpenMP_C: -fopenmp (found version "4.5") 
   -- Found OpenMP_CXX: -fopenmp (found version "4.5") 
   -- Found OpenMP: TRUE (found version "4.5")  
   -- GPU support is disabled
   -- Could NOT find Doxygen (missing: DOXYGEN_EXECUTABLE) 
   -- Found Git: /usr/bin/git (found version "2.17.1") 
   -- Intel(R) VTune(TM) Amplifier JIT profiling disabled
   -- Found MKL: /opt/intel/compilers_and_libraries/linux/mkl/include  
   -- Found MKL (include: /opt/intel/compilers_and_libraries/linux/mkl/include, lib: /opt/intel/compilers_and_libraries/linux/mkl/lib/intel64/libmkl_rt.so
   -- Could NOT find PkgConfig (missing: PKG_CONFIG_EXECUTABLE) 
   -- Could NOT find Jemalloc (missing: JEMALLOC_LIBRARY JEMALLOC_INCLUDE_DIR) 
   -- OpenCV Disabled
   -- Could NOT find PkgConfig (missing: PKG_CONFIG_EXECUTABLE) 
   -- Could NOT find Jemalloc (missing: JEMALLOC_LIBRARY JEMALLOC_INCLUDE_DIR) 
   CMake Warning at 3rdparty/googletest/googletest/CMakeLists.txt:47 (project):
     VERSION keyword not followed by a value or was followed by a value that
     expanded to nothing.
   
   
   -- Found PythonInterp: /usr/bin/python (found version "2.7.17") 
   -- Found GTest: gtest  
   -- Looking for clock_gettime in rt
   -- Looking for clock_gettime in rt - found
   -- Looking for fopen64
   -- Looking for fopen64 - not found
   -- Looking for C++ include cxxabi.h
   -- Looking for C++ include cxxabi.h - found
   -- Looking for nanosleep
   -- Looking for nanosleep - found
   -- Looking for backtrace
   -- Looking for backtrace - found
   -- backtrace facility detected in default set of libraries
   -- Found Backtrace: /usr/include  
   -- Check if the system is big endian
   -- Searching 16 bit integer
   -- Looking for sys/types.h
   -- Looking for sys/types.h - found
   -- Looking for stdint.h
   -- Looking for stdint.h - found
   -- Looking for stddef.h
   -- Looking for stddef.h - found
   -- Check size of unsigned short
   -- Check size of unsigned short - done
   -- Using unsigned short
   -- Check if the system is big endian - little endian
   -- /root/mxnet/3rdparty/dmlc-core/cmake/build_config.h.in -> include/dmlc/build_config.h
   CMake Warning at CMakeLists.txt:830 (message):
     OpenCV_VERSION_MAJOR: , version 3 with imgcodecs is required for im2rec,
     im2rec will not be available
   
   
   CMake Warning at example/image-classification/predict-cpp/CMakeLists.txt:3 (message):
     OpenCV version >= 3 should be enabled and found to build image
     classification example, skipping...
   
   
   -- Configuring done
   -- Generating done
   -- Build files have been written to: /root/mxnet/build_openmpoff
   
   ```
   
   The build reaches the `96%` point then prints out a bunch of errors (regarding openmp and mkldnn). Here is a preview of the errors:
   
   ```
   [ 96%] Building CXX object tests/CMakeFiles/mxnet_unit_tests.dir/cpp/test_main.cc.o
   In file included from /root/mxnet/3rdparty/mshadow/../../src/operator/nn/activation-inl.h:40:0,
                    from /root/mxnet/tests/cpp/operator/tune/operator_tune_test.cc:21:
   /root/mxnet/3rdparty/mshadow/../../src/operator/nn/../mshadow_op.h: In static member function ‘static bool mxnet::op::mshadow_op::mul::Map(bool, bool)’:
   /root/mxnet/3rdparty/mshadow/../../src/operator/nn/../mshadow_op.h:350:42: warning: ‘*’ in boolean context, suggest ‘&&’ instead [-Wint-in-bool-context]
   /root/mxnet/3rdparty/mshadow/../../src/operator/nn/../mshadow_op.h:111:14:
          return (expr); \
                 ~~~~~~                       
   /root/mxnet/3rdparty/mshadow/../../src/operator/nn/../mshadow_op.h:350:42:
    MXNET_BINARY_MATH_OP_NC_WITH_BOOL(mul, a * b);
   /root/mxnet/3rdparty/mshadow/../../src/operator/nn/../mshadow_op.h:111:15: note: in definition of macro ‘MXNET_BINARY_MATH_OP_NC_WITH_BOOL’
          return (expr); \
                  ^~~~
   In file included from /root/mxnet/tests/cpp/include/../../../src/ndarray/ndarray_function.h:34:0,
                    from /root/mxnet/tests/cpp/include/test_util.h:37,
                    from /root/mxnet/tests/cpp/storage/storage_test.cc:29:
   /root/mxnet/tests/cpp/include/../../../src/ndarray/../operator/mshadow_op.h: In static member function ‘static bool mxnet::op::mshadow_op::mul::Map(bool, bool)’:
   /root/mxnet/tests/cpp/include/../../../src/ndarray/../operator/mshadow_op.h:350:42: warning: ‘*’ in boolean context, suggest ‘&&’ instead [-Wint-in-bool-context]
   /root/mxnet/tests/cpp/include/../../../src/ndarray/../operator/mshadow_op.h:111:14:
          return (expr); \
                 ~~~~~~                       
   /root/mxnet/tests/cpp/include/../../../src/ndarray/../operator/mshadow_op.h:350:42:
    MXNET_BINARY_MATH_OP_NC_WITH_BOOL(mul, a * b);
   /root/mxnet/tests/cpp/include/../../../src/ndarray/../operator/mshadow_op.h:111:15: note: in definition of macro ‘MXNET_BINARY_MATH_OP_NC_WITH_BOOL’
          return (expr); \
                  ^~~~
   [ 96%] Built target mxnet
   Scanning dependencies of target cpp_package_op_h
   Running: OpWrapperGenerator.py
   Traceback (most recent call last):
     File "OpWrapperGenerator.py", line 433, in <module>
       raise(e)
   OSError: /root/mxnet/build_openmpoff/libmxnet.so: undefined symbol: omp_get_max_threads
   cpp-package/CMakeFiles/cpp_package_op_h.dir/build.make:58: recipe for target 'cpp-package/CMakeFiles/cpp_package_op_h' failed
   make[2]: *** [cpp-package/CMakeFiles/cpp_package_op_h] Error 1
   CMakeFiles/Makefile2:1730: recipe for target 'cpp-package/CMakeFiles/cpp_package_op_h.dir/all' failed
   make[1]: *** [cpp-package/CMakeFiles/cpp_package_op_h.dir/all] Error 2
   make[1]: *** Waiting for unfinished jobs....
   [ 96%] Linking CXX executable mxnet_unit_tests
   ../3rdparty/mkldnn/src/libdnnl.a(cpu_engine.cpp.o): In function `void dnnl::impl::parallel_nd<int const&, int const&, dnnl::impl::cpu::ref_shuffle_t<4>::ref_shuffle_t(dnnl::impl::cpu::ref_shuffle_t<4>::pd_t const*)::{lambda(int, int)#1}>(int const&, int const&, dnnl::impl::cpu::ref_shuffle_t<4>::ref_shuffle_t(dnnl::impl::cpu::ref_shuffle_t<4>::pd_t const*)::{lambda(int, int)#1}&&) [clone ._omp_fn.2]':
   cpu_engine.cpp:(.text._ZN4dnnl4impl11parallel_ndIJRKiS3_ZNS0_3cpu13ref_shuffle_tILi4EEC4EPKNS6_4pd_tEEUliiE_EEEvDpOT_._omp_fn.2[_ZNK4dnnl4impl3cpu13ref_shuffle_tILi4EE4pd_t16create_primitiveEPP14dnnl_primitive]+0x3e): undefined reference to `omp_get_num_threads'
   cpu_engine.cpp:(.text._ZN4dnnl4impl11parallel_ndIJRKiS3_ZNS0_3cpu13ref_shuffle_tILi4EEC4EPKNS6_4pd_tEEUliiE_EEEvDpOT_._omp_fn.2[_ZNK4dnnl4impl3cpu13ref_shuffle_tILi4EE4pd_t16create_primitiveEPP14dnnl_primitive]+0x45): undefined reference to `omp_get_thread_num'
   ../3rdparty/mkldnn/src/libdnnl.a(cpu_engine.cpp.o): In function `void dnnl::impl::parallel_nd<int const&, int const&, dnnl::impl::cpu::ref_shuffle_t<1>::ref_shuffle_t(dnnl::impl::cpu::ref_shuffle_t<1>::pd_t const*)::{lambda(int, int)#1}>(int const&, int const&, dnnl::impl::cpu::ref_shuffle_t<1>::ref_shuffle_t(dnnl::impl::cpu::ref_shuffle_t<1>::pd_t const*)::{lambda(int, int)#1}&&) [clone ._omp_fn.0]':
   cpu_engine.cpp:(.text._ZN4dnnl4impl11parallel_ndIJRKiS3_ZNS0_3cpu13ref_shuffle_tILi1EEC4EPKNS6_4pd_tEEUliiE_EEEvDpOT_._omp_fn.0[_ZNK4dnnl4impl3cpu13ref_shuffle_tILi1EE4pd_t16create_primitiveEPP14dnnl_primitive]+0x3e): undefined reference to `omp_get_num_threads'
   cpu_engine.cpp:(.text._ZN4dnnl4impl11parallel_ndIJRKiS3_ZNS0_3cpu13ref_shuffle_tILi1EEC4EPKNS6_4pd_tEEUliiE_EEEvDpOT_._omp_fn.0[_ZNK4dnnl4impl3cpu13ref_shuffle_tILi1EE4pd_t16create_primitiveEPP14dnnl_primitive]+0x45): undefined reference to `omp_get_thread_num'
   ../3rdparty/mkldnn/src/libdnnl.a(cpu_engine.cpp.o): In function `void dnnl::impl::parallel_nd<int const&, int const&, dnnl::impl::cpu::ref_shuffle_t<2>::ref_shuffle_t(dnnl::impl::cpu::ref_shuffle_t<2>::pd_t const*)::{lambda(int, int)#1}>(int const&, int const&, dnnl::impl::cpu::ref_shuffle_t<2>::ref_shuffle_t(dnnl::impl::cpu::ref_shuffle_t<2>::pd_t const*)::{lambda(int, int)#1}&&) [clone ._omp_fn.1]':
   cpu_engine.cpp:(.text._ZN4dnnl4impl11parallel_ndIJRKiS3_ZNS0_3cpu13ref_shuffle_tILi2EEC4EPKNS6_4pd_tEEUliiE_EEEvDpOT_._omp_fn.1[_ZNK4dnnl4impl3cpu13ref_shuffle_tILi2EE4pd_t16create_primitiveEPP14dnnl_primitive]+0x3e): undefined reference to `omp_get_num_threads'
   cpu_engine.cpp:(.text._ZN4dnnl4impl11parallel_ndIJRKiS3_ZNS0_3cpu13ref_shuffle_tILi2EEC4EPKNS6_4pd_tEEUliiE_EEEvDpOT_._omp_fn.1[_ZNK4dnnl4impl3cpu13ref_shuffle_tILi2EE4pd_t16create_primitiveEPP14dnnl_primitive]+0x45): undefined reference to `omp_get_thread_num'
   ../3rdparty/mkldnn/src/libdnnl.a(cpu_engine.cpp.o): In function `dnnl_status_t dnnl_primitive_desc::create<dnnl::impl::cpu::jit_avx512_core_bf16_convolution_fwd_t::pd_t>(dnnl_primitive_desc**, dnnl::impl::op_desc_t const*, dnnl_primitive_attr const*, dnnl_engine*, dnnl_primitive_desc const*)':
   cpu_engine.cpp:(.text._ZN19dnnl_primitive_desc6createIN4dnnl4impl3cpu38jit_avx512_core_bf16_convolution_fwd_t4pd_tEEE13dnnl_status_tPPS_PKNS2_9op_desc_tEPK19dnnl_primitive_attrP11dnnl_enginePKS_[_ZN19dnnl_primitive_desc6createIN4dnnl4impl3cpu38jit_avx512_core_bf16_convolution_fwd_t4pd_tEEE13dnnl_status_tPPS_PKNS2_9op_desc_tEPK19dnnl_primitive_attrP11dnnl_enginePKS_]+0x24a): undefined reference to `omp_get_max_threads'
   ../3rdparty/mkldnn/src/libdnnl.a(cpu_engine.cpp.o): In function `dnnl_status_t dnnl_primitive_desc::create<dnnl::impl::cpu::jit_avx512_common_convolution_fwd_t<(dnnl_data_type_t)3, (dnnl_data_type_t)3, (dnnl_data_type_t)3>::pd_t>(dnnl_primitive_desc**, dnnl::impl::op_desc_t const*, dnnl_primitive_attr const*, dnnl_engine*, dnnl_primitive_desc const*)':
   cpu_engine.cpp:(.text._ZN19dnnl_primitive_desc6createIN4dnnl4impl3cpu35jit_avx512_common_convolution_fwd_tIL16dnnl_data_type_t3ELS5_3ELS5_3EE4pd_tEEE13dnnl_status_tPPS_PKNS2_9op_desc_tEPK19dnnl_primitive_attrP11dnnl_enginePKS_[_ZN19dnnl_primitive_desc6createIN4dnnl4impl3cpu35jit_avx512_common_convolution_fwd_tIL16dnnl_data_type_t3ELS5_3ELS5_3EE4pd_tEEE13dnnl_status_tPPS_PKNS2_9op_desc_tEPK19dnnl_primitive_attrP11dnnl_enginePKS_]+0x11d): undefined reference to `omp_get_max_threads'
   ../3rdparty/mkldnn/src/libdnnl.a(cpu_engine.cpp.o): In function `dnnl_status_t dnnl_primitive_desc::create<dnnl::impl::cpu::jit_avx512_core_x8s8s32x_convolution_fwd_t<(dnnl_data_type_t)6, (dnnl_data_type_t)3>::pd_t>(dnnl_primitive_desc**, dnnl::impl::op_desc_t const*, dnnl_primitive_attr const*, dnnl_engine*, dnnl_primitive_desc const*)':
   cpu_engine.cpp:(.text._ZN19dnnl_primitive_desc6createIN4dnnl4impl3cpu42jit_avx512_core_x8s8s32x_convolution_fwd_tIL16dnnl_data_type_t6ELS5_3EE4pd_tEEE13dnnl_status_tPPS_PKNS2_9op_desc_tEPK19dnnl_primitive_attrP11dnnl_enginePKS_[_ZN19dnnl_primitive_desc6createIN4dnnl4impl3cpu42jit_avx512_core_x8s8s32x_convolution_fwd_tIL16dnnl_data_type_t6ELS5_3EE4pd_tEEE13dnnl_status_tPPS_PKNS2_9op_desc_tEPK19dnnl_primitive_attrP11dnnl_enginePKS_]+0x134): undefined reference to `omp_get_max_threads'
   ../3rdparty/mkldnn/src/libdnnl.a(cpu_engine.cpp.o): In function `dnnl_status_t dnnl_primitive_desc::create<dnnl::impl::cpu::jit_avx512_core_x8s8s32x_convolution_fwd_t<(dnnl_data_type_t)6, (dnnl_data_type_t)4>::pd_t>(dnnl_primitive_desc**, dnnl::impl::op_desc_t const*, dnnl_primitive_attr const*, dnnl_engine*, dnnl_primitive_desc const*)':
   cpu_engine.cpp:(.text._ZN19dnnl_primitive_desc6createIN4dnnl4impl3cpu42jit_avx512_core_x8s8s32x_convolution_fwd_tIL16dnnl_data_type_t6ELS5_4EE4pd_tEEE13dnnl_status_tPPS_PKNS2_9op_desc_tEPK19dnnl_primitive_attrP11dnnl_enginePKS_[_ZN19dnnl_primitive_desc6createIN4dnnl4impl3cpu42jit_avx512_core_x8s8s32x_convolution_fwd_tIL16dnnl_data_type_t6ELS5_4EE4pd_tEEE13dnnl_status_tPPS_PKNS2_9op_desc_tEPK19dnnl_primitive_attrP11dnnl_enginePKS_]+0x134): undefined reference to `omp_get_max_threads'
   ../3rdparty/mkldnn/src/libdnnl.a(cpu_engine.cpp.o): In function `dnnl_status_t dnnl_primitive_desc::create<dnnl::impl::cpu::jit_avx512_core_x8s8s32x_convolution_fwd_t<(dnnl_data_type_t)6, (dnnl_data_type_t)6>::pd_t>(dnnl_primitive_desc**, dnnl::impl::op_desc_t const*, dnnl_primitive_attr const*, dnnl_engine*, dnnl_primitive_desc const*)':
   cpu_engine.cpp:(.text._ZN19dnnl_primitive_desc6createIN4dnnl4impl3cpu42jit_avx512_core_x8s8s32x_convolution_fwd_tIL16dnnl_data_type_t6ELS5_6EE4pd_tEEE13dnnl_status_tPPS_PKNS2_9op_desc_tEPK19dnnl_primitive_attrP11dnnl_enginePKS_[_ZN19dnnl_primitive_desc6createIN4dnnl4impl3cpu42jit_avx512_core_x8s8s32x_convolution_fwd_tIL16dnnl_data_type_t6ELS5_6EE4pd_tEEE13dnnl_status_tPPS_PKNS2_9op_desc_tEPK19dnnl_primitive_attrP11dnnl_enginePKS_]+0x134): undefined reference to `omp_get_max_threads'
   
   ```

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services

[GitHub] [incubator-mxnet] leezu commented on issue #17338: Unable to build mxnet with USE_OPENMP=OFF with MKLDNN backend

Posted by GitBox <gi...@apache.org>.
leezu commented on issue #17338: Unable to build mxnet with USE_OPENMP=OFF with MKLDNN backend
URL: https://github.com/apache/incubator-mxnet/issues/17338#issuecomment-575145811
 
 
   @TaoLv this fixes the issue for me (though I didn't try linking with MKL). I'll reassign this issue to you, as it seems to be due to MKLDNN integration (ie above missing variable).

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services

[GitHub] [incubator-mxnet] leezu closed issue #17338: Unable to build mxnet with USE_OPENMP=OFF with MKLDNN backend

Posted by GitBox <gi...@apache.org>.
leezu closed issue #17338: Unable to build mxnet with USE_OPENMP=OFF with MKLDNN backend
URL: https://github.com/apache/incubator-mxnet/issues/17338
 
 
   

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services

[GitHub] [incubator-mxnet] leezu edited a comment on issue #17338: Unable to build mxnet with USE_OPENMP=OFF with MKLDNN backend

Posted by GitBox <gi...@apache.org>.
leezu edited a comment on issue #17338: Unable to build mxnet with USE_OPENMP=OFF with MKLDNN backend
URL: https://github.com/apache/incubator-mxnet/issues/17338#issuecomment-575145811
 
 
   @TaoLv this fixes the issue for me (though I didn't try linking with MKL). I'll reassign this issue to you, as it seems to be due to MKLDNN integration (ie above missing variable).
   
   Ie with your diff applied, `cmake -DUSE_CPP_PACKAGE=1 -DUSE_CUDA=0 -DUSE_MKLDNN=ON -DUSE_OPENCV=0 -DUSE_LAPACK=0 -DUSE_OPENMP=0 -GNinja ..; ninja` works

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services

[GitHub] [incubator-mxnet] TaoLv commented on issue #17338: Unable to build mxnet with USE_OPENMP=OFF with MKLDNN backend

Posted by GitBox <gi...@apache.org>.
TaoLv commented on issue #17338: Unable to build mxnet with USE_OPENMP=OFF with MKLDNN backend
URL: https://github.com/apache/incubator-mxnet/issues/17338#issuecomment-575073921
 
 
   @cyrusbehr Could you please try to apply below change to the CMakeList file?
   ```
   diff --git a/CMakeLists.txt b/CMakeLists.txt
   index 2612854..3068925 100644
   --- a/CMakeLists.txt
   +++ b/CMakeLists.txt
   @@ -288,6 +288,7 @@ if(USE_MKLDNN)
      set(MKLDNN_ARCH_OPT_FLAGS "" CACHE INTERNAL "" FORCE)
      set(MKLDNN_ENABLE_JIT_PROFILING OFF CACHE INTERNAL "" FORCE)
      set(MKLDNN_LIBRARY_TYPE STATIC CACHE INTERNAL "" FORCE)
   +  set(MKLDNN_CPU_RUNTIME SEQ CACHE INTERNAL "" FORCE)
   
      add_subdirectory(3rdparty/mkldnn)
   ```
   
   Let me know if it works for you. Thanks.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services

[GitHub] [incubator-mxnet] cyrusbehr commented on issue #17338: Unable to build mxnet with USE_OPENMP=OFF with MKLDNN backend

Posted by GitBox <gi...@apache.org>.
cyrusbehr commented on issue #17338: Unable to build mxnet with USE_OPENMP=OFF with MKLDNN backend
URL: https://github.com/apache/incubator-mxnet/issues/17338#issuecomment-575278640
 
 
   @leezu @TaoLv 
   Adding `set(MKLDNN_CPU_RUNTIME SEQ CACHE INTERNAL "" FORCE)` solved the problem. 
   Thanks!

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services

[GitHub] [incubator-mxnet] TaoLv commented on issue #17338: Unable to build mxnet with USE_OPENMP=OFF with MKLDNN backend

Posted by GitBox <gi...@apache.org>.
TaoLv commented on issue #17338: Unable to build mxnet with USE_OPENMP=OFF with MKLDNN backend
URL: https://github.com/apache/incubator-mxnet/issues/17338#issuecomment-575150538
 
 
   @leezu Thanks. We need check USE_OPENMP in cmake as what we did for makefile: https://github.com/apache/incubator-mxnet/blob/master/mkldnn.mk#L35. I will file a PR to fix the problem once the patch is verified by the reporter.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services