Deep Stream App SIGEGV after restart

Hi all,

Problem:
When I stop(NULL) an analytics pipeline , re-create the pipeline and then try to play. I get Process finished with exit code 139 (interrupted by signal 11: SIGSEGV). I want to be able to recreate the pipeline with new sources and new analytics config and then start it again. Ive recreated the issue with the sample app which ive hacked to simulate what im trying to do. Please just change the input URI and point nvinfer to your own classfier files and analytics to your own config.

Log just before crash:

gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream-5.1/lib/libnvds_nvdcf.so
gstnvtracker: Batch processing is ON
gstnvtracker: Past frame output is OFF
~~ CLOG[/dvs/git/dirty/git-master_linux/deepstream/sdk/src/utils/nvdcf/src/modules/NvDCF/NvDCF.cpp, NvDCF() @line 670]: !!![WARNING] Can’t open config file (tracker_config_nvDCF_orig.yml). Will go ahead with default values
~~ CLOG[/dvs/git/dirty/git-master_linux/deepstream/sdk/src/utils/nvdcf/src/modules/NvDCF/NvDCF.cpp, NvDCF() @line 682]: !!![WARNING] Invalid low-level config file is provided. Will go ahead with default values
[NvDCF] Initialized

Process finished with exit code 139 (interrupted by signal 11: SIGSEGV)

sample.txt (18.0 KB)

Ive attached the sample code i used to reproduce the issue
Deep Stream 5.1
classifier: yolov3

Regards

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU)

• DeepStream Version

• JetPack Version (valid for Jetson only)

• TensorRT Version

• NVIDIA GPU Driver Version (valid for GPU only)

• Issue Type( questions, new requirements, bugs)

• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)

• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)

Hardware Platform: Jetson Xavier AGX devkit
DS: 5.1
Jetpack: 4.5.1
TensorRt: 8.0.1.6-1+cuda10.2
Issue: SIGSEGV after deep stream pipeline re-create and start. The first pipeline play works perfectly, but after stopping the pipeline and re-creating it(new instance of pipeline with new instances of elements) and then starting again i get SIGSEGV just before deserializing yolo

first play log:
[NvDCF] Initialized
WARNING: [TRT]: Using an engine plan file across different models of devices is not recommended and is likely to affect performance or even cause errors.
Deserialize yoloLayerV3 plugin: yolo_83
Deserialize yoloLayerV3 plugin: yolo_95
Deserialize yoloLayerV3 plugin: yolo_107

second play log:
[NvDCF] Initialized
Process finished with exit code 139 (interrupted by signal 11: SIGSEGV)

so you can see it happens after [NvDCF] Initialized, this leads me to believe it is to do with running nvinfer on the second instance that causes the crash

How to reproduce the issue: run my adapted version of the analytics sample app which ive attached in the original post. Please insert your own URI and point the nvinfer to your own classifier files

  1. which deepstream sample is your code based on? can your modify code based on deepstream sample?
  2. after test your code, there is an error:
    python3 sample.py file:///opt/nvidia/deepstream/deepstream/samples/streams/sample_720p.mp4
    Traceback (most recent call last):
    File “sample.py”, line 449, in
    main(uri)
    File “sample.py”, line 227, in main
    Gst.debug_set_active(True)
    File “/usr/lib/python3/dist-packages/gi/overrides/Gst.py”, line 605, in fake_method
    raise NotInitialized(“Please call Gst.init(argv) before using GStreamer”)

There is no update from you for a period, assuming this is not an issue any more.
Hence we are closing this topic. If need further support, please open a new one.
Thanks

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.