Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

anaconda pytorch double free or corruption (fasttop) caused by glibc update #2205

Closed
fakaiwang opened this issue Jul 25, 2017 · 2 comments
Closed

Comments

@fakaiwang
Copy link

Hi,
My laboratory updated the glibc on our workstation, and after that my pytorch begin to fail from time to time. It seems that the error is caused by customed usage of glibc, which depends on a certain version probably. Does anyone encounter similar problem?
Thank you!
-----------------------------------update record ---------------
Jul 19 03:19:34 Updated: glibc-common-2.17-157.el7_3.5.x86_64
Jul 19 03:19:35 Updated: glibc-2.17-157.el7_3.5.x86_64
Jul 19 03:19:35 Updated: glibc-headers-2.17-157.el7_3.5.x86_64
Jul 19 03:19:35 Updated: glibc-devel-2.17-157.el7_3.5.x86_64
Jul 19 03:19:36 Updated: glibc-2.17-157.el7_3.5.i686
------------ldd info----------

ldd /vulcan/scratch/jackwang/anaconda3/lib/python3.6/site-packages/torch/lib/libcudnn.so

   linux-vdso.so.1 =>  (0x00007ffed0592000)
   librt.so.1 => /usr/lib64/librt.so.1 (0x00007f25fb472000)
   libpthread.so.0 => /usr/lib64/libpthread.so.0 (0x00007f25fb255000)
   libdl.so.2 => /usr/lib64/libdl.so.2 (0x00007f25fb051000)
   libstdc++.so.6 => /vulcan/scratch/jackwang/anaconda3/lib/python3.6/site-packages/torch/lib/../../../../libstdc++.so.6 (0x00007f25fad3b000)
   libm.so.6 => /usr/lib64/libm.so.6 (0x00007f25faa38000)
   libgcc_s.so.1 => /vulcan/scratch/jackwang/anaconda3/lib/python3.6/site-packages/torch/lib/../../../../libgcc_s.so.1 (0x00007f25fa822000)
   libc.so.6 => /usr/lib64/libc.so.6 (0x00007f25fa461000)
   /lib64/ld-linux-x86-64.so.2 (0x00007f2604c07000)

------------Below is the output:-------

*** Error in `python': double free or corruption (fasttop): 0x00007fe508103130 ***
======= Backtrace: =========
/usr/lib64/libc.so.6(+0x7c503)[0x7fe6a59e1503]
/usr/lib64/nvidia/libcuda.so.1(+0x1aa1df)[0x7fe674bf01df]
/usr/lib64/nvidia/libcuda.so.1(+0xd051b)[0x7fe674b1651b]
/usr/lib64/nvidia/libcuda.so.1(cuStreamCreate+0x5b)[0x7fe674c3629b]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/site-packages/torch/lib/libcudnn.so.6(+0x864236)[0x7fe690180236]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/site-packages/torch/lib/libcudnn.so.6(+0x8966a4)[0x7fe6901b26a4]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/site-packages/torch/lib/libcudnn.so.6(+0x7839bf)[0x7fe69009f9bf]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/site-packages/torch/lib/libcudnn.so.6(cudnnRNNBackwardData+0xf42)[0x7fe69009dd12]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/lib-dynload/_ctypes.cpython-36m-x86_64-linux-gnu.so(ffi_call_unix64+0x4c)[0x7fe69d1cf5b0]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/lib-dynload/_ctypes.cpython-36m-x86_64-linux-gnu.so(ffi_call+0x1f5)[0x7fe69d1ced55]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/lib-dynload/_ctypes.cpython-36m-x86_64-linux-gnu.so(ctypescallproc+0x3dc)[0x7fe69d1c689c]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/lib-dynload/_ctypes.cpython-36m-x86_64-linux-gnu.so(+0x9df3)[0x7fe69d1bedf3]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyObjectFastCallDict+0x9e)[0x7fe6a68bec1e]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(+0x14895b)[0x7fe6a699b95b]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyEvalEvalFrameDefault+0x2c40)[0x7fe6a699ed40]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(+0x146514)[0x7fe6a6999514]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(+0x148c88)[0x7fe6a699bc88]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyEvalEvalFrameDefault+0x2c40)[0x7fe6a699ed40]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(+0x146514)[0x7fe6a6999514]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyFunctionFastCallDict+0x285)[0x7fe6a699a515]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyObjectFastCallDict+0x166)[0x7fe6a68bece6]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyObjectCall_Prepend+0xcc)[0x7fe6a68bef3c]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyObject_Call+0x56)[0x7fe6a68befd6]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyEvalEvalFrameDefault+0x3ec9)[0x7fe6a699ffc9]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(+0x147100)[0x7fe6a699a100]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyFunctionFastCallDict+0x10c)[0x7fe6a699a39c]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyObjectFastCallDict+0x166)[0x7fe6a68bece6]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyObjectCall_Prepend+0xcc)[0x7fe6a68bef3c]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyObject_Call+0x56)[0x7fe6a68befd6]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/site-packages/torch/_C.cpython-36m-x86_64-linux-gnu.so(Z23THPFunctiondo_backwardP11THPFunctionP7_object+0x133)[0x7fe695b0d533]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyCFunctionFastCallDict+0x229)[0x7fe6a6916429]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(+0x148b8c)[0x7fe6a699bb8c]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyEvalEvalFrameDefault+0x2c40)[0x7fe6a699ed40]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(+0x147100)[0x7fe6a699a100]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyFunctionFastCallDict+0x10c)[0x7fe6a699a39c]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyObjectFastCallDict+0x166)[0x7fe6a68bece6]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyObjectCall_Prepend+0xcc)[0x7fe6a68bef3c]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyObject_Call+0x56)[0x7fe6a68befd6]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(+0x6c0bd)[0x7fe6a68bf0bd]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(+0x6c16d)[0x7fe6a68bf16d]
/vulcan/scratch/jackwang/anaconda3/envs/magma/bin/../lib/libpython3.6m.so.1.0(PyObject_CallMethod+0xe6)[0x7fe6a68c26e6]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/site-packages/torch/_C.cpython-36m-x86_64-linux-gnu.so(ZN5torch8autograd10PyFunction5applyERKSt6vectorISt10sharedptrINS0_8VariableEESaIS5_EE+0xe1)[0x7fe695b0dfe1]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/site-packages/torch/_C.cpython-36m-x86_64-linux-gnu.so(ZN5torch8autograd6Engine17evaluatefunctionERNS0_12FunctionTaskE+0x248)[0x7fe695b03be8]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/site-packages/torch/_C.cpython-36m-x86_64-linux-gnu.so(ZN5torch8autograd6Engine11threadmainESt10shared_ptrINS0_10ReadyQueueEE+0x3b)[0x7fe695b056eb]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/site-packages/torch/_C.cpython-36m-x86_64-linux-gnu.so(ZN12PythonEngine11threadmainESt10shared_ptrIN5torch8autograd10ReadyQueueEE+0x54)[0x7fe695b16284]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/site-packages/torch/_C.cpython-36m-x86_64-linux-gnu.so(ZNSt6thread5ImplISt12_Bind_simpleIFSt7_Mem_fnIMN5torch8autograd6EngineEFvSt10shared_ptrINS4_10ReadyQueueEEEEPS5_S8_EEE6_M_runEv+0x48)[0x7fe695b07f18]
/vulcan/scratch/jackwang/anaconda3/envs/magma/lib/python3.6/site-packages/torch/lib/../../../../libstdc++.so.6(+0xb7260)[0x7fe67f0eb260]
/usr/lib64/libpthread.so.0(+0x7dc5)[0x7fe6a663edc5]
/usr/lib64/libc.so.6(clone+0x6d)[0x7fe6a5a5c76d]
======= Memory map: ========

@coderwhisky
Copy link

I got the same error as you when predicting

@fakaiwang
Copy link
Author

My error has been solved, it's due to memory constraints of the system. The training process is very memory consuming and use 32 GB, which far exceeds the allocated 10 GB.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants