Skip to content
This repository was archived by the owner on Nov 29, 2024. It is now read-only.
This repository was archived by the owner on Nov 29, 2024. It is now read-only.

Segmentation Fault #1

@arun-kumark

Description

@arun-kumark

Hi Haanjack,
Thank you for this repo.
I was trying using your code for the inference openpose, it runs fine with the nvidia docker image 18.04, but when I am trying using the nvidia docker image 19.01, it gives the segmentation fault at the following location :

context->enqueue(gParams.batchSize, &buffers[0], stream, nullptr);

Building and running a GPU inference engine for OpenPose, N=1...
Input "image": 3x656x368
Output "net_output": 78x82x46
Run inference...
name=image, bindingIndex=0, buffers.size()=2
name=net_output, bindingIndex=1, buffers.size()=2
Segmentation fault (core dumped)

I tried investigating the issue, and found, the function:
int PReLUPlugin::enqueue(int batchSize, const void *const *inputs, void **outputs, void *workspace, cudaStream_t stream) is never called

I also checked the size of allocated buffer[0] and stream, which are successfully allocated, but no clue why segmentation fault?

Would you be giving me some debugging clues solving this issue.

Kind Regards
Arun

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions