Segmentation fault on Raspbian

[wu__junhao]

Hello,I downloaded the new release of OpenVINO for Raspberry PI but when I run the ssd demo the Segmentation fault happened 

pi@raspberrypi:~/inference_engine_samples_build/armv7l/Release $ ./object_detection_sample_ssd -m ~/Downloads/face-detection-adas-0001.xml -d MYRIAD -i /dev/video0
[ INFO ] InferenceEngine:
        API version ............ 1.6
        Build .................. 22443
Parsing input parameters
[ INFO ] Files were added: 1
[ INFO ] Loading plugin

        API version ............ 1.6
        Build .................. 22443
        Description ....... myriadPlugin
[ INFO ] Loading network files:
        /home/pi/Downloads/face-detection-adas-0001.xml
        /home/pi/Downloads/face-detection-adas-0001.bin
Segmentation fault

[Dmitry_K_Intel3]

Hi!

May I ask you to show how this demo is compiled? Have you added -march=armv7-a compilation flag as described in the guide https://docs.openvinotoolkit.org/latest/_docs_install_guides_installing_openvino_raspbian.html#run-sample? ;

[Shubha_R_Intel]

Dear wu, junhao

OpenVINO is compiled with –march=armv7-a. The problem is that Raspberry Pi compiler applies –march=armv6 by default (that was RPI team decision due to compatibility among different RPI boards starting from version 0 to 3). So mixed arch flags leads to crashes in standard libraries. A bug has been filed on this issue.

Sorry for the inconvenience !

Thanks !

Shubha

[Tolbert__Donnie]

Could you post a reference to the filed bug so we can track it?

[Shubha_R_Intel]

Dear Tolbert, Donnie,

Unless you are an Intel employee, I cannot. 

I was recently told that it won't be fixed since the documentation clearly tells you to use  –march=armv7-a. 

Thanks for using OpenVino,

Shubha

[Marcia_W_Intel]

junhao,

I see this is the directory -

~/inference_engine_samples_build/armv7l/Release $

This is usually created after running the build samples script.

Instead, follow these instructions:

Build and Run Object Detection Sample

Follow the next steps to run pre-trained Face Detection network using Inference Engine samples from the OpenVINO toolkit.

Navigate to a directory that you have write access to and create a samples build directory. This example uses a directory named build:

mkdir build && cd build

Build the Object Detection Sample:

cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_FLAGS="-march=armv7-a" /opt/intel/openvino/deployment_tools/inference_engine/samples

make -j2 object_detection_sample_ssd

Instructions --> https://docs.openvinotoolkit.org/latest/_docs_install_guides_installing_openvino_raspbian.html

Don't use these: ~/inference_engine_samples_build/armv7l/Release $

because if you do, you will keep getting seg faults