Segmentation fault (core dumped) error on Jetson Nano

Robotics & Edge Computing Jetson & Embedded Systems Jetson Nano
1

While running ssd inception v2 coco model over a video with tensorrt engine, after running for 2 mins the process got stopped with Segmentation fault (core dumped) error on Jetson Nano

[TensorRT] INFO: Glob Size is 50485364 bytes.
[TensorRT] INFO: Added linear block of size 4292096
[TensorRT] INFO: Added linear block of size 1522176
[TensorRT] INFO: Added linear block of size 739328
[TensorRT] INFO: Added linear block of size 411648
[TensorRT] INFO: Added linear block of size 204800
[TensorRT] INFO: Added linear block of size 57344
[TensorRT] INFO: Added linear block of size 30720
[TensorRT] INFO: Added linear block of size 20992
[TensorRT] INFO: Added linear block of size 9728
[TensorRT] INFO: Added linear block of size 9216
[TensorRT] INFO: Added linear block of size 2560
[TensorRT] INFO: Added linear block of size 2560
[TensorRT] INFO: Added linear block of size 1024
[TensorRT] INFO: Added linear block of size 512
[TensorRT] INFO: Found Creator FlattenConcat_TRT
[TensorRT] INFO: Found Creator GridAnchor_TRT
[TensorRT] INFO: Found Creator FlattenConcat_TRT
[TensorRT] INFO: Found Creator NMS_TRT
[TensorRT] INFO: Deserialize required 3084695 microseconds.
Gtk-Message: 11:48:55.371: Failed to load module "canberra-gtk-module"
Segmentation fault (core dumped)
2

Hi,

May I know which sample do you use first?
Is it jetson_inference? https://github.com/dusty-nv/jetson-inference

Thanks.

3

No, it’s not a jetson_inference, rather I have converted my custom ssd inception v2 coco model to tensorrt engine. I was working fine but crashed all of a sudden after 3 mins with that error.

4

Hi,

Could you help to provide a simple reproducible source for us debugging?
Or could you try if this issue can be reproduced with our official sample?
Ex. /usr/src/tensorrt/samples/sampleUffSSD

Usually, this should be related to the data source type you use.
Do you run the inference code with live stream or camera?

Thanks.

5

Hi siddharthdas8,

Have you managed to get issue resolved?
Any result can be shared?

6

hi

i meet the same issue. error info is :
[TensorRT] ERROR: UffParser: Validator error: truediv_14/Cast: Unsupported operation _Cast
[TensorRT] ERROR: Network must have at least one output
Segmentation fault (core dumped)
I trained a yolo3 model with TF 1.15.0, and then convert it to uff. convertion log is below:
Loading output_frozen_model2.pb
NOTE: UFF has been tested with TensorFlow 1.12.0. Other versions are not guaranteed to work
WARNING: The version of TensorFlow installed on this system is not guaranteed to work with UFF.
UFF Version 0.6.3
=== Automatically deduced input nodes ===
[name: “input”
op: “Placeholder”
attr {
key: “dtype”
value {
type: DT_FLOAT
}
}
attr {
key: “shape”
value {
shape {
dim {
size: -1
}
dim {
size: -1
}
dim {
size: -1
}
dim {
size: 3
}
}
}
}
]

=== Automatically deduced output nodes ===
[name: “output/boxes”
op: “Identity”
input: “combined_non_max_suppression/CombinedNonMaxSuppression”
attr {
key: “T”
value {
type: DT_FLOAT
}
}
, name: “output/scores”
op: “Identity”
input: “combined_non_max_suppression/CombinedNonMaxSuppression:1”
attr {
key: “T”
value {
type: DT_FLOAT
}
}
, name: “output/labels”
op: “Identity”
input: “combined_non_max_suppression/CombinedNonMaxSuppression:2”
attr {
key: “T”
value {
type: DT_FLOAT
}
}
, name: “output/num_detections”
op: “Identity”
input: “combined_non_max_suppression/CombinedNonMaxSuppression:3”
attr {
key: “T”
value {
type: DT_INT32
}
}
]

1 Placeholder: “input”
2 Shape: “yolov3/Shape”
3 Const: “yolov3/strided_slice/stack”
4 Const: “yolov3/strided_slice/stack_1”
5 Const: “yolov3/strided_slice/stack_2”

7

I also traind a yolov3 with pytorch. also meet the same error Segmentation fault (core dumped). sample is as below.

def ONNX_build_engine(onnx_file_path):

    G_LOGGER = trt.Logger(trt.Logger.WARNING)
    EXPLICIT_BATCH = 1 << (int)(trt.NetworkDefinitionCreationFlag.EXPLICIT_BATCH)
    with trt.Builder(G_LOGGER) as builder, builder.create_network(EXPLICIT_BATCH) as network, trt.OnnxParser(network, G_LOGGER) as parser:
        print('1.1--Loading ONNX file from path {}...'.format(onnx_file_path))
    with open(onnx_file_path, 'rb') as model:
        print('1.2--Parsing ONNX model')
        parser.parse(model.read())
        builder.max_batch_size = 100
        builder.max_workspace_size = 1 << 30
        print('1.3--Completed parsing of ONNX model')
    with trt.Builder(TRT_LOGGER) as builder, builder.create_builder_config() as config, builder.build_cuda_engine(network, config) as engine:
        print('1.4--Building an engine from file {}; this may take a while...'.format(onnx_file_path))
        serialized_engine = engine.serialize()
        with trt.Runtime(TRT_LOGGER) as runtime:
            runtime.deserialize_cuda_engine(serialized_engine)
        print("1.5--Completed creating Engine")
8

Yup, got it running. Earlier I was using another module along with it, it might be that module was creating the issue, running this object detection task only worked for me. Thanks for the response @AastaLLL and @kayccc

9

Hi guo.feng,

Please open a new topic for your issue. Thanks