Skip to content
This repository has been archived by the owner on Aug 30, 2018. It is now read-only.

convert-caffe2-to-onnx get error : "Segmentation fault (core dumped)" #292

Open
icortana opened this issue Jul 4, 2018 · 3 comments
Open

convert-caffe2-to-onnx get error : "Segmentation fault (core dumped)" #292

icortana opened this issue Jul 4, 2018 · 3 comments

Comments

@icortana
Copy link

icortana commented Jul 4, 2018

Hi,
I'm trying to convert caffe2 model to onnx model, but simply type convert-caffe2-to-onnx will get this error : "Segmentation fault (core dumped)".

I have tried the following environments :

Ubuntu 16.04.4 LTS, Python 3.6 Caffe2 CPU;
Ubuntu 16.04.4 LTS, Python 3.6 Caffe2 GPU CUDA 9.0 CuDNN7;
Caffe2 docker Ubuntu 16.04.2 LTS, Python 2.7 Caffe2 CPU;

They all get the same result.
@ezyang
Copy link
Collaborator

ezyang commented Jul 4, 2018

Can you use gdb to get a backtrace and post it here?

@icortana
Copy link
Author

icortana commented Jul 5, 2018 

(gdb) r caffe2onnx.py
Starting program: /root/anaconda3/bin/python caffe2onnx.py
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".

Program received signal SIGSEGV, Segmentation fault.
0x00007fffef0b2d69 in pybind11::detail::make_new_python_type(pybind11::detail::type_record const&) ()
   from /root/anaconda3/lib/python3.6/site-packages/caffe2/python/caffe2_pybind11_state_gpu.cpython-36m-x86_64-linux-gnu.so
(gdb) bt
#0  0x00007fffef0b2d69 in pybind11::detail::make_new_python_type(pybind11::detail::type_record const&) ()
   from /root/anaconda3/lib/python3.6/site-packages/caffe2/python/caffe2_pybind11_state_gpu.cpython-36m-x86_64-linux-gnu.so
#1  0x00007fffef0b5968 in pybind11::detail::generic_type::initialize(pybind11::detail::type_record const&) ()
   from /root/anaconda3/lib/python3.6/site-packages/caffe2/python/caffe2_pybind11_state_gpu.cpython-36m-x86_64-linux-gnu.so
#2  0x00007fffef0b6061 in pybind11::class_<caffe2::NetBase>::class_<>(pybind11::handle, char const*) ()
   from /root/anaconda3/lib/python3.6/site-packages/caffe2/python/caffe2_pybind11_state_gpu.cpython-36m-x86_64-linux-gnu.so
#3  0x00007fffef084fa1 in caffe2::python::addObjectMethods(pybind11::module&) ()
   from /root/anaconda3/lib/python3.6/site-packages/caffe2/python/caffe2_pybind11_state_gpu.cpython-36m-x86_64-linux-gnu.so
#4  0x00007fffef0e43a4 in PyInit_caffe2_pybind11_state_gpu ()
   from /root/anaconda3/lib/python3.6/site-packages/caffe2/python/caffe2_pybind11_state_gpu.cpython-36m-x86_64-linux-gnu.so
#5  0x00000000082158e5 in _PyImport_LoadDynamicModuleWithSpec ()
#6  0x0000000008215ae5 in _imp_create_dynamic ()
#7  0x0000000008111a61 in PyCFunction_Call ()
#8  0x00000000081c5fdb in _PyEval_EvalFrameDefault ()
#9  0x0000000008197a94 in _PyEval_EvalCodeWithName ()
#10 0x0000000008198941 in fast_function ()
#11 0x000000000819e755 in call_function ()
#12 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#13 0x000000000819870b in fast_function ()
#14 0x000000000819e755 in call_function ()
#15 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#16 0x000000000819870b in fast_function ()
#17 0x000000000819e755 in call_function ()
#18 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
---Type <return> to continue, or q <return> to quit---
#19 0x000000000819870b in fast_function ()
#20 0x000000000819e755 in call_function ()
#21 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#22 0x000000000819870b in fast_function ()
#23 0x000000000819e755 in call_function ()
#24 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#25 0x0000000008198d7b in _PyFunction_FastCallDict ()
#26 0x000000000810ef5f in _PyObject_FastCallDict ()
#27 0x0000000008153670 in _PyObject_CallMethodIdObjArgs ()
#28 0x0000000008105a70 in PyImport_ImportModuleLevelObject ()
#29 0x00000000081c3033 in _PyEval_EvalFrameDefault ()
#30 0x0000000008199459 in PyEval_EvalCodeEx ()
#31 0x000000000819a1ec in PyEval_EvalCode ()
#32 0x00000000081be5bd in builtin_exec ()
#33 0x0000000008111a61 in PyCFunction_Call ()
#34 0x00000000081c5fdb in _PyEval_EvalFrameDefault ()
#35 0x0000000008197a94 in _PyEval_EvalCodeWithName ()
#36 0x0000000008198941 in fast_function ()
#37 0x000000000819e755 in call_function ()
#38 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#39 0x000000000819870b in fast_function ()
#40 0x000000000819e755 in call_function ()
#41 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#42 0x000000000819870b in fast_function ()
#43 0x000000000819e755 in call_function ()
#44 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#45 0x000000000819870b in fast_function ()
#46 0x000000000819e755 in call_function ()
#47 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#48 0x0000000008198d7b in _PyFunction_FastCallDict ()
#49 0x000000000810ef5f in _PyObject_FastCallDict ()
#50 0x0000000008153670 in _PyObject_CallMethodIdObjArgs ()
#51 0x0000000008105a70 in PyImport_ImportModuleLevelObject ()
#52 0x00000000081c3033 in _PyEval_EvalFrameDefault ()
#53 0x0000000008199459 in PyEval_EvalCodeEx ()
#54 0x000000000819a1ec in PyEval_EvalCode ()
#55 0x00000000081be5bd in builtin_exec ()
#56 0x0000000008111a61 in PyCFunction_Call ()
#57 0x00000000081c5fdb in _PyEval_EvalFrameDefault ()
#58 0x0000000008197a94 in _PyEval_EvalCodeWithName ()
#59 0x0000000008198941 in fast_function ()
#60 0x000000000819e755 in call_function ()
#61 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#62 0x000000000819870b in fast_function ()
#63 0x000000000819e755 in call_function ()
#64 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#65 0x000000000819870b in fast_function ()
#66 0x000000000819e755 in call_function ()
#67 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#68 0x000000000819870b in fast_function ()
#69 0x000000000819e755 in call_function ()
#70 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#71 0x0000000008198d7b in _PyFunction_FastCallDict ()
#72 0x000000000810ef5f in _PyObject_FastCallDict ()
#73 0x0000000008153670 in _PyObject_CallMethodIdObjArgs ()
#74 0x0000000008105a70 in PyImport_ImportModuleLevelObject ()
#75 0x00000000081a91da in builtin___import__ ()
#76 0x00000000081119f6 in PyCFunction_Call ()
#77 0x00000000081c5fdb in _PyEval_EvalFrameDefault ()
#78 0x0000000008197a94 in _PyEval_EvalCodeWithName ()
#79 0x0000000008198941 in fast_function ()
#80 0x000000000819e755 in call_function ()
---Type <return> to continue, or q <return> to quit---
#81 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#82 0x0000000008197f21 in _PyEval_EvalCodeWithName ()
#83 0x0000000008198e1b in _PyFunction_FastCallDict ()
#84 0x000000000810ef5f in _PyObject_FastCallDict ()
#85 0x0000000008153670 in _PyObject_CallMethodIdObjArgs ()
#86 0x0000000008105933 in PyImport_ImportModuleLevelObject ()
#87 0x00000000081c3033 in _PyEval_EvalFrameDefault ()
#88 0x0000000008199459 in PyEval_EvalCodeEx ()
#89 0x000000000819a1ec in PyEval_EvalCode ()
#90 0x00000000081be5bd in builtin_exec ()
#91 0x0000000008111a61 in PyCFunction_Call ()
#92 0x00000000081c5fdb in _PyEval_EvalFrameDefault ()
#93 0x0000000008197a94 in _PyEval_EvalCodeWithName ()
#94 0x0000000008198941 in fast_function ()
#95 0x000000000819e755 in call_function ()
#96 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#97 0x000000000819870b in fast_function ()
#98 0x000000000819e755 in call_function ()
#99 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#100 0x000000000819870b in fast_function ()
#101 0x000000000819e755 in call_function ()
#102 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#103 0x000000000819870b in fast_function ()
#104 0x000000000819e755 in call_function ()
#105 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#106 0x0000000008198d7b in _PyFunction_FastCallDict ()
#107 0x000000000810ef5f in _PyObject_FastCallDict ()
#108 0x0000000008153670 in _PyObject_CallMethodIdObjArgs ()
#109 0x0000000008105a70 in PyImport_ImportModuleLevelObject ()
#110 0x00000000081a91da in builtin___import__ ()
#111 0x00000000081119f6 in PyCFunction_Call ()
#112 0x00000000081c5fdb in _PyEval_EvalFrameDefault ()
#113 0x0000000008197a94 in _PyEval_EvalCodeWithName ()
#114 0x0000000008198941 in fast_function ()
#115 0x000000000819e755 in call_function ()
#116 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#117 0x0000000008197f21 in _PyEval_EvalCodeWithName ()
#118 0x0000000008198e1b in _PyFunction_FastCallDict ()
#119 0x000000000810ef5f in _PyObject_FastCallDict ()
#120 0x0000000008153670 in _PyObject_CallMethodIdObjArgs ()
#121 0x0000000008105933 in PyImport_ImportModuleLevelObject ()
#122 0x00000000081c3033 in _PyEval_EvalFrameDefault ()
#123 0x0000000008199459 in PyEval_EvalCodeEx ()
#124 0x000000000819a1ec in PyEval_EvalCode ()
#125 0x00000000081be5bd in builtin_exec ()
#126 0x0000000008111a61 in PyCFunction_Call ()
#127 0x00000000081c5fdb in _PyEval_EvalFrameDefault ()
#128 0x0000000008197a94 in _PyEval_EvalCodeWithName ()
#129 0x0000000008198941 in fast_function ()
#130 0x000000000819e755 in call_function ()
#131 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#132 0x000000000819870b in fast_function ()
#133 0x000000000819e755 in call_function ()
#134 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#135 0x000000000819870b in fast_function ()
#136 0x000000000819e755 in call_function ()
#137 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#138 0x000000000819870b in fast_function ()
#139 0x000000000819e755 in call_function ()
#140 0x00000000081c0cba in _PyEval_EvalFrameDefault ()
#141 0x0000000008198d7b in _PyFunction_FastCallDict ()
#142 0x000000000810ef5f in _PyObject_FastCallDict ()
---Type <return> to continue, or q <return> to quit---
#143 0x0000000008153670 in _PyObject_CallMethodIdObjArgs ()
#144 0x0000000008105a70 in PyImport_ImportModuleLevelObject ()
#145 0x00000000081c3033 in _PyEval_EvalFrameDefault ()
#146 0x0000000008199459 in PyEval_EvalCodeEx ()
#147 0x000000000819a1ec in PyEval_EvalCode ()
#148 0x00000000082149a4 in run_mod ()
#149 0x0000000008214da1 in PyRun_FileExFlags ()
#150 0x0000000008214fa4 in PyRun_SimpleFileExFlags ()
#151 0x0000000008218a9e in Py_Main ()
#152 0x00000000080e04be in main ()
(gdb)

caffe2onnx.py :

import onnx
import caffe2.python.onnx.frontend
from caffe2.proto import caffe2_pb2

# We need to provide type and shape of the model inputs,
# see above Note section for explanation
data_type = onnx.TensorProto.FLOAT
data_shape = (1, 3, 224, 224)
value_info = {
    'data': (data_type, data_shape)
}

predict_net = caffe2_pb2.NetDef()
with open('predict_net.pb', 'rb') as f:
    predict_net.ParseFromString(f.read())

init_net = caffe2_pb2.NetDef()
with open('init_net.pb', 'rb') as f:
    init_net.ParseFromString(f.read())

onnx_model = caffe2.python.onnx.frontend.caffe2_net_to_onnx_model(
    predict_net,
    init_net,
    value_info,
)

onnx.checker.check_model(onnx_model)

@muthiyanbhushan
Copy link

I am also having same issue. Any suggestions?

Thanks.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@ezyang @muthiyanbhushan @icortana