You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Nov 17, 2023. It is now read-only.
I've recently started using the C++ custom operator framework. It's great and I like it a lot, but for some reason mxnet always segfaults on program exit when I've loaded a custom operator. This happens with all custom operators, even the tutorial operator gemm_lib.
This is on OSX Mojave, g++ version "Apple LLVM version 10.0.1 (clang-1001.0.46.4)". mxnet comes from pip, version 1.7.0.post2.
Error Message
Since it's a segfault there's not a backtrace as such, but here's the tail of a representative lldb session:
Welcome to Apache MXNet (incubating)! We are on a mission to democratize AI, and we are glad that you are contributing to it by opening this issue.
Please make sure to include all the relevant context, and one of the @apache/mxnet-committers will be here shortly.
If you are interested in contributing to our project, let us know! Also, be sure to check out our guide on contributing to MXNet and our development guides wiki.
Hi @david-seiler thanks for filing the issue. We have not done much testing of c++ custom ops on Mac (if at all). Mostly on linux and very slightly on windows. Im not aware of how OS X is different than linux kernel handling of dynamic libraries, so any help you can provide in that area would be great.
in 1.7.x we have some code in the LibraryInitializer::~LibraryInitializer() destructor that tries to close the custom libraries that have been opened:
Description
I've recently started using the C++ custom operator framework. It's great and I like it a lot, but for some reason mxnet always segfaults on program exit when I've loaded a custom operator. This happens with all custom operators, even the tutorial operator gemm_lib.
This is on OSX Mojave, g++ version "Apple LLVM version 10.0.1 (clang-1001.0.46.4)". mxnet comes from pip, version 1.7.0.post2.
Error Message
Since it's a segfault there's not a backtrace as such, but here's the tail of a representative lldb session:
To Reproduce
make gemm_lib && python test_gemm.py
from the custom op tutorial shows the error for me.Things That Didn't Work
Building my own mxnet 1.7.0 from source didn't help. The content of the custom op doesn't seem to matter either.
Environment
Environment Information
----------Python Info----------
Version : 3.6.5
Compiler : GCC 4.2.1 Compatible Clang 4.0.1 (tags/RELEASE_401/final)
Build : ('default', 'Apr 26 2018 08:42:37')
Arch : ('64bit', '')
------------Pip Info-----------
Version : 21.1.3
Directory : /Users/davseile/miniconda3/lib/python3.6/site-packages/pip
----------MXNet Info-----------
/Users/davseile/miniconda3/lib/python3.6/site-packages/sklearn/externals/joblib/externals/cloudpickle/cloudpickle.py:47: DeprecationWarning: the imp module is deprecated in favour of importlib; see the module's documentation for alternative uses
import imp
Version : 1.7.0
Directory : /Users/davseile/miniconda3/lib/python3.6/site-packages/mxnet
Commit Hash : 64f737c
Library : ['/Users/davseile/miniconda3/lib/python3.6/site-packages/mxnet/libmxnet.dylib']
Build features:
✖ CUDA
✖ CUDNN
✖ NCCL
✖ CUDA_RTC
✖ TENSORRT
✔ CPU_SSE
✔ CPU_SSE2
✔ CPU_SSE3
✔ CPU_SSE4_1
✖ CPU_SSE4_2
✖ CPU_SSE4A
✖ CPU_AVX
✖ CPU_AVX2
✖ OPENMP
✖ SSE
✖ F16C
✖ JEMALLOC
✖ BLAS_OPEN
✖ BLAS_ATLAS
✖ BLAS_MKL
✔ BLAS_APPLE
✔ LAPACK
✔ MKLDNN
✔ OPENCV
✖ CAFFE
✖ PROFILER
✖ DIST_KVSTORE
✖ CXX14
✖ INT64_TENSOR_SIZE
✔ SIGNAL_HANDLER
✖ DEBUG
✖ TVM_OP
----------System Info----------
Platform : Darwin-18.7.0-x86_64-i386-64bit
system : Darwin
node : 00e04c166570
release : 18.7.0
version : Darwin Kernel Version 18.7.0: Mon May 3 20:41:19 PDT 2021; root:xnu-4903.278.68~1/RELEASE_X86_64
----------Hardware Info----------
machine : x86_64
processor : i386
b'machdep.cpu.brand_string: Intel(R) Core(TM) i7-7660U CPU @ 2.50GHz'
b'machdep.cpu.features: FPU VME DE PSE TSC MSR PAE MCE CX8 APIC SEP MTRR PGE MCA CMOV PAT PSE36 CLFSH DS ACPI MMX FXSR SSE SSE2 SS HTT TM PBE SSE3 PCLMULQDQ DTES64 MON DSCPL VMX SMX EST TM2 SSSE3 FMA CX16 TPR PDCM SSE4.1 SSE4.2 x2APIC MOVBE POPCNT AES PCID XSAVE OSXSAVE SEGLIM64 TSCTMR AVX1.0 RDRAND F16C'
b'machdep.cpu.leaf7_features: RDWRFSGS TSC_THREAD_OFFSET SGX BMI1 HLE AVX2 SMEP BMI2 ERMS INVPCID RTM FPU_CSDS MPX RDSEED ADX SMAP CLFSOPT IPT MDCLEAR TSXFA IBRS STIBP L1DF SSBD'
b'machdep.cpu.extfeatures: SYSCALL XD 1GBPAGE EM64T LAHF LZCNT PREFETCHW RDTSCP TSCI'
----------Network Test----------
Setting timeout: 10
Timing for MXNet: https://github.com/apache/incubator-mxnet, DNS: 0.0430 sec, LOAD: 0.7044 sec.
Timing for Gluon Tutorial(en): http://gluon.mxnet.io, DNS: 0.0961 sec, LOAD: 0.5755 sec.
Error open Gluon Tutorial(cn): https://zh.gluon.ai, <urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed (_ssl.c:833)>, DNS finished in 0.08476710319519043 sec.
Timing for FashionMNIST: https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/dataset/fashion-mnist/train-labels-idx1-ubyte.gz, DNS: 0.0732 sec, LOAD: 0.3876 sec.
Timing for PYPI: https://pypi.python.org/pypi/pip, DNS: 0.0459 sec, LOAD: 1.0816 sec.
Error open Conda: https://repo.continuum.io/pkgs/free/, HTTP Error 403: Forbidden, DNS finished in 0.06121397018432617 sec.
----------Environment----------
The text was updated successfully, but these errors were encountered: