Pythons core dumped segmentation fault while reading the network using load_network()

[NSVR]

Hi,
I am currently using openvino 2020.4.287 and python 3.6 version on Intel Arria10 GX with PAC . While I am running inference on classification_sample.py I am getting this error
I have debugged and it's coming from some MKL package. Please find the error in the attachment

[Munesh_Intel]

Hi Saivenkataraju,

Thanks for contacting us.

What is the GCC version that you have installed currently? This error is a known issue that occurs due to GCC version. The suggested solution is to upgrade to GCC 4.9 or higher.

Please refer to the following GitHub thread and comment for more information:

https://github.com/facebookresearch/maskrcnn-benchmark/issues/21#issuecomment-433396994

https://gist.github.com/goldsborough/d466f43e8ffc948ff92de7486c5216d6

Regards,

Munesh

[NSVR]

I have verified GCC version but its 7.5.0.

[Munesh_Intel]

Hi Saivenkataraju,

Thanks for verifying GCC version.

We need to isolate the probable causes to narrow down on the actual cause of this issue. We suspect that this issue might be caused by HETERO plugin.

As such, please run inference by configuring another device to use with HETERO plugin (GPU or MYRIAD) to verify.

Please also share the hardware details, and environment details (versions of OS, OpenCV, CMake, etc.).

On another note, I am sharing a link here which contains a probable solution as well, for your information.

https://stackoverflow.com/questions/61951365/raspberry-pi4-openvino-open-cv-dnn-net-forward-segmentation-fault-ncs2

Regards,

Munesh

[Munesh_Intel]

Hi Saivenkataraju,

Thank you for your question. If you need any additional information from Intel, please submit a new question as this thread is no longer being monitored.

Regards,

Munesh