TensorRT engine execute on jetson TX2

AI & Data Science Deep Learning (Training & Inference) TensorRT
1

Hi, I was able to create tensorrt engine using c++ api, and NVCC compiler.
Now, I’ve tried to create code for inference.
I’ve tried to compile and run the executable file for every new line I added, in order to make kind of debugging on my jetson TX2.
My problem is with this line:
context->execute(1, bindings);
The compilation was successful but when I ran the output of the compilation I got this error:

cudnnEngine.cpp (420) - Cuda Error in execute: 4

Is anyone know this issue and know how to fix that?
thank you all!!!

2

Hello,

CUDA error 4 is cudaErrorLaunchFailure.
Usually due to dereferencing an invalid pointer or accessing out of bounds shared memory.

http://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html#group__CUDART__TYPES_1g3f51e3575c2178246db0a94a430e0038

Likely you have an out of resource issue.