参考 :https://forums.developer.nvidia.com/t/no-cmake-cuda-compiler-could-be-found-cmake/353204
还是不行，直接把路径
/usr/local/cuda-11.4/bin/nvcc
添加到 CmakeGUI
CMAKE_CUDA_COMPILER 这一栏
配置编译成功。
TensorRT version: 8.5.2
Found TensorRT: /usr/include/aarch64-linux-gnu /usr/lib/aarch64-linux-gnu/libnvinfer.so;/usr/lib/aarch64-linux-gnu/libnvinfer_plugin.so;/usr/lib/aarch64-linux-gnu/libnvparsers.so;/usr/lib/aarch64-linux-gnu/libnvonnxparser.so /usr/src/tensorrt/samples
TensorRT version: 8.5.2
Configuring done
Generating done