When running example of Objectdetector_SSD after creating engine segmentation faults occurs but video is seen

Please provide complete information as applicable to your setup.

**• Hardware Platform (Jetson / GPU)**dGPU T4
• DeepStream Version5.0
• JetPack Version (valid for Jetson only)
• TensorRT Version7.0
**• NVIDIA GPU Driver Version (valid for GPU only)**440
When running example of Objectdetector_SSD after creating engine segmentation faults occurs but video is seen

Hi,
Can you provide comlete log when error happened?

Any command to gove before running deepstream-app to get the log?

Kindly see the log here

deepstream-app -c deepstream_app_config_ssd.txt
Warn: ‘threshold’ parameter has been deprecated. Use ‘pre-cluster-threshold’ instead.
WARNING: …/nvdsinfer/nvdsinfer_func_utils.cpp:34 [TRT]: Current optimization profile is: 0. Please ensure there are no enqueued operations pending in this context prior to switching profiles
0:00:02.474194751 4563 0x55fabd01f360 INFO nvinfer gstnvinfer.cpp:602:gst_nvinfer_logger:<primary_gie> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1577> [UID = 1]: deserialized trt engine from :/opt/nvidia/deepstream/deepstream-5.0/sources/objectDetector_SSD/sample_ssd_relu6.uff_b1_gpu0_fp32.engine
INFO: …/nvdsinfer/nvdsinfer_model_builder.cpp:685 [Implicit Engine Info]: layers num: 3
0 INPUT kFLOAT Input 3x300x300
1 OUTPUT kFLOAT NMS 1x100x7
2 OUTPUT kFLOAT NMS_1 1x1x1

ERROR: …/nvdsinfer/nvdsinfer_func_utils.cpp:31 [TRT]: INVALID_ARGUMENT: Cannot find binding of given name: MarkOutput_0
0:00:02.474270324 4563 0x55fabd01f360 WARN nvinfer gstnvinfer.cpp:599:gst_nvinfer_logger:<primary_gie> NvDsInferContext[UID 1]: Warning from NvDsInferContextImpl::checkBackendParams() <nvdsinfer_context_impl.cpp:1545> [UID = 1]: Could not find output layer ‘MarkOutput_0’ in engine
0:00:02.474278026 4563 0x55fabd01f360 INFO nvinfer gstnvinfer.cpp:602:gst_nvinfer_logger:<primary_gie> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1681> [UID = 1]: Use deserialized engine model: /opt/nvidia/deepstream/deepstream-5.0/sources/objectDetector_SSD/sample_ssd_relu6.uff_b1_gpu0_fp32.engine
0:00:02.476900580 4563 0x55fabd01f360 INFO nvinfer gstnvinfer_impl.cpp:311:notifyLoadModelStatus:<primary_gie> [UID 1]: Load new model:/opt/nvidia/deepstream/deepstream-5.0/sources/objectDetector_SSD/config_infer_primary_ssd.txt sucessfully

Runtime commands:
h: Print this help
q: Quit

p: Pause
r: Resume

**PERF: FPS 0 (Avg)
**PERF: 0.00 (0.00)
** INFO: <bus_callback:181>: Pipeline ready

** INFO: <bus_callback:167>: Pipeline running

**PERF: 41.12 (40.34)
**PERF: 29.98 (34.72)
**PERF: 29.99 (33.05)
**PERF: 29.99 (32.25)
**PERF: 29.98 (31.78)
**PERF: 30.04 (31.48)
**PERF: 29.99 (31.26)
**PERF: 30.02 (31.09)
**PERF: 30.01 (30.97)
**PERF: 29.97 (30.87)
**PERF: 29.97 (30.79)
**PERF: 30.06 (30.72)
**PERF: 29.98 (30.66)
**PERF: 30.04 (30.61)
**PERF: 29.98 (30.57)
**PERF: 30.04 (30.53)
**PERF: 29.94 (30.50)
**PERF: 30.05 (30.47)
**PERF: 30.00 (30.44)

**PERF: FPS 0 (Avg)
**PERF: 29.99 (30.42)
**PERF: 29.93 (30.40)
**PERF: 30.02 (30.38)
**PERF: 30.04 (30.36)
**PERF: 30.00 (30.35)
**PERF: 29.98 (30.33)
**PERF: 29.76 (30.32)
**PERF: 30.24 (30.30)
**PERF: 30.01 (30.33)
**PERF: 29.95 (30.28)
**PERF: 30.02 (30.30)
**PERF: 30.03 (30.29)
**PERF: 30.00 (30.28)
**PERF: 29.96 (30.27)
**PERF: 29.95 (30.27)
**PERF: 30.07 (30.26)
**PERF: 30.02 (30.25)
**PERF: 30.01 (30.24)
**PERF: 29.96 (30.23)
**PERF: 29.98 (30.23)

**PERF: FPS 0 (Avg)
**PERF: 29.99 (30.22)
**PERF: 30.02 (30.22)
**PERF: 29.97 (30.21)
**PERF: 30.00 (30.21)
**PERF: 30.05 (30.20)
**PERF: 30.01 (30.20)
**PERF: 29.98 (30.19)
**PERF: 30.01 (30.19)
**PERF: 30.02 (30.12)
** INFO: <bus_callback:204>: Received EOS. Exiting …

Quitting
Segmentation fault (core dumped)

Did you use original model or customized model?

Hi @amycao,
This is same as specified in README

Can you provide below info:
• Hardware Platform (Jetson / GPU)
• DeepStream Version
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)

dGPU t4
Deepstream 5.0
Tensorrt 7.0
440

I see you using engine, can you run success with TensorRT sample trtexec?

yes I even get output video correctly in Deepstream but segmentation fault comes

How you build the engine?
the segmentation should be caused by “Cannot find binding of given name: MarkOutput_0”

Using the readme in the object detector SSD folder. Actually before creating the model engine files i.e running for the first time segmentation fault doesnt appear. But after creating engine file and running again the error appears

Can you remove the cache for gstreamer and try again?
rm ~/.cache/gstreamer-1.0/ -rf

(gst-plugin-scanner:8553): GStreamer-WARNING **: 13:04:30.412: Failed to load plugin ‘/usr/lib/x86_64-linux-gnu/gstreamer-1.0/deepstream/libnvdsgst_inferserver.so’: libtrtserver.so: cannot open shared object file: No such file or directory
Warn: ‘threshold’ parameter has been deprecated. Use ‘pre-cluster-threshold’ instead.
WARNING: …/nvdsinfer/nvdsinfer_func_utils.cpp:34 [TRT]: Current optimization profile is: 0. Please ensure there are no enqueued operations pending in this context prior to switching profiles
0:00:04.694936383 8552 0x55e233094160 INFO nvinfer gstnvinfer.cpp:602:gst_nvinfer_logger:<primary_gie> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1577> [UID = 1]: deserialized trt engine from :/opt/nvidia/deepstream/deepstream-5.0/sources/objectDetector_SSD/sample_ssd_relu6.uff_b1_gpu0_fp32.engine
INFO: …/nvdsinfer/nvdsinfer_model_builder.cpp:685 [Implicit Engine Info]: layers num: 3
0 INPUT kFLOAT Input 3x300x300
1 OUTPUT kFLOAT NMS 1x100x7
2 OUTPUT kFLOAT NMS_1 1x1x1

ERROR: …/nvdsinfer/nvdsinfer_func_utils.cpp:31 [TRT]: INVALID_ARGUMENT: Cannot find binding of given name: MarkOutput_0
0:00:04.695032331 8552 0x55e233094160 WARN nvinfer gstnvinfer.cpp:599:gst_nvinfer_logger:<primary_gie> NvDsInferContext[UID 1]: Warning from NvDsInferContextImpl::checkBackendParams() <nvdsinfer_context_impl.cpp:1545> [UID = 1]: Could not find output layer ‘MarkOutput_0’ in engine
0:00:04.695042338 8552 0x55e233094160 INFO nvinfer gstnvinfer.cpp:602:gst_nvinfer_logger:<primary_gie> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1681> [UID = 1]: Use deserialized engine model: /opt/nvidia/deepstream/deepstream-5.0/sources/objectDetector_SSD/sample_ssd_relu6.uff_b1_gpu0_fp32.engine
0:00:04.697615006 8552 0x55e233094160 INFO nvinfer gstnvinfer_impl.cpp:311:notifyLoadModelStatus:<primary_gie> [UID 1]: Load new model:/opt/nvidia/deepstream/deepstream-5.0/sources/objectDetector_SSD/config_infer_primary_ssd.txt sucessfully

Runtime commands:
h: Print this help
q: Quit

p: Pause
r: Resume

**PERF: FPS 0 (Avg)
**PERF: 0.00 (0.00)
** INFO: <bus_callback:181>: Pipeline ready

** INFO: <bus_callback:167>: Pipeline running

**PERF: 41.04 (40.30)
**PERF: 30.01 (34.71)
**PERF: 29.92 (33.05)
**PERF: 29.99 (32.25)
**PERF: 29.98 (31.79)
**PERF: 30.00 (31.48)
**PERF: 29.99 (31.26)
**PERF: 29.98 (31.10)
**PERF: 30.01 (30.97)
**PERF: 29.99 (30.87)
**PERF: 30.00 (30.79)
**PERF: 29.97 (30.72)
**PERF: 30.04 (30.66)
**PERF: 29.99 (30.62)
**PERF: 30.01 (30.57)
**PERF: 29.99 (30.54)
**PERF: 30.00 (30.50)
**PERF: 30.01 (30.47)
**PERF: 30.00 (30.45)

**PERF: FPS 0 (Avg)
**PERF: 29.99 (30.42)
**PERF: 30.00 (30.40)
**PERF: 30.00 (30.38)
**PERF: 30.01 (30.37)
**PERF: 30.01 (30.35)
**PERF: 29.98 (30.34)
**PERF: 30.01 (30.32)
**PERF: 30.00 (30.31)
**PERF: 30.00 (30.30)
**PERF: 30.00 (30.29)
**PERF: 30.05 (30.28)
**PERF: 29.93 (30.27)
**PERF: 30.00 (30.26)
**PERF: 30.00 (30.25)
**PERF: 30.00 (30.24)
**PERF: 30.00 (30.24)
**PERF: 30.00 (30.23)
**PERF: 30.01 (30.22)
**PERF: 30.00 (30.22)
**PERF: 30.00 (30.21)

**PERF: FPS 0 (Avg)
**PERF: 30.03 (30.20)
**PERF: 29.97 (30.20)
**PERF: 30.00 (30.19)
**PERF: 30.01 (30.19)
**PERF: 30.00 (30.18)
**PERF: 30.00 (30.18)
**PERF: 30.00 (30.18)
**PERF: 30.00 (30.17)
**PERF: 30.00 (30.13)
** INFO: <bus_callback:204>: Received EOS. Exiting …

Quitting
Cuda failure: 7
Aborted (core dumped)

This is the output

Can you move to DS 5.0GA to see if the issue still there, if does, we will see you there.

I am using the same DS 5.0

Please using 5.0 GA, the previous version is 5.0 DP