Describe the bug Hi all,

I was trying to build the onnxruntime from scoure on the Jetson devices, and I followed the steps from here: https://onnxruntime.ai/docs/how-to/build/eps.html#nvidia-jetson-tx1tx2nanoxavier

I can successfully build the whl file by this command:

git clone --recursive https://github.com/microsoft/onnxruntime
export CUDACXX="/usr/local/cuda/bin/nvcc"
export PATH="/usr/local/cuda/bin:${PATH}"
sudo apt install -y --no-install-recommends \
   build-essential software-properties-common libopenblas-dev \
   libpython3.8-dev python3-pip python3-dev python3-setuptools python3-wheel

./build.sh --config Release --update --build --parallel --build_wheel \
 --use_cuda --cuda_home /usr/local/cuda --cudnn_home /usr/lib/aarch64-linux-gnu

However, when I used it in python script, it encountered an issue.

Traceback (most recent call last):
  File "infererence_centernet_onnxruntime.py", line 10, in <module>
    sess = ort.InferenceSession(onnxpath)
  File "/usr/local/lib/python3.8/dist-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 324, in __init__
    self._create_inference_session(providers, provider_options, disabled_optimizers)
  File "/usr/local/lib/python3.8/dist-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 362, in _create_inference_session
    sess.initialize_session(providers, provider_options, disabled_optimizers)
RuntimeError: /workspace/onnxruntime/onnxruntime/core/providers/cuda/cuda_call.cc:122 bool onnxruntime::CudaCall(ERRTYPE, const char*, const char*, ERRTYPE, const char*) [with ERRTYPE = cudaError; bool THRW = true] /workspace/onnxruntime/onnxruntime/core/providers/cuda/cuda_call.cc:116 bool onnxruntime::CudaCall(ERRTYPE, const char*, const char*, ERRTYPE, const char*) [with ERRTYPE = cudaError; bool THRW = true] CUDA failure 35: CUDA driver version is insufficient for CUDA runtime version ; GPU=127 ; hostname=043f3cc5e143 ; expr=cudaSetDevice(info_.device_id); 

---

Hence, I was trying to build the onnxruntime with TensorRT.

Built from source by this command:

git clone --recursive https://github.com/microsoft/onnxruntime
export CUDACXX="/usr/local/cuda/bin/nvcc"
export PATH="/usr/local/cuda/bin:${PATH}"
sudo apt install -y --no-install-recommends \
   build-essential software-properties-common libopenblas-dev \
   libpython3.8-dev python3-pip python3-dev python3-setuptools python3-wheel

./build.sh --config Release --update --build --parallel --build_wheel \
 --use_tensorrt --cuda_home /usr/local/cuda --cudnn_home /usr/lib/aarch64-linux-gnu \
 --tensorrt_home /usr/lib/aarch64-linux-gnu

Error message during building stage:

[ 49%] Linking CXX static library libonnxruntime_optimizer.a
[ 49%] Built target onnxruntime_optimizer
make: *** [Makefile:166: all] Error 2
Traceback (most recent call last):
  File "/workspace/onnxruntime/tools/ci_build/build.py", line 2299, in <module>
    sys.exit(main())
  File "/workspace/onnxruntime/tools/ci_build/build.py", line 2220, in main
    build_targets(args, cmake_path, build_dir, configs, num_parallel_jobs, args.target)
  File "/workspace/onnxruntime/tools/ci_build/build.py", line 1136, in build_targets
    run_subprocess(cmd_args, env=env)
  File "/workspace/onnxruntime/tools/ci_build/build.py", line 612, in run_subprocess
    return run(*args, cwd=cwd, capture_stdout=capture_stdout, shell=shell, env=my_env)
  File "/workspace/onnxruntime/tools/python/util/run.py", line 42, in run
    completed_process = subprocess.run(
  File "/usr/lib/python3.8/subprocess.py", line 516, in run
    raise CalledProcessError(retcode, process.args,
subprocess.CalledProcessError: Command '['/usr/local/bin/cmake', '--build', '/workspace/onnxruntime/build/Linux/Release', '--config', 'Release', '--', '-j6']' returned non-zero exit status 2.
root@39246b46fb36:/workspace/onnxruntime# l
bash: l: command not found

System information

• OS Platform and Distribution : Linux 20.04 based on JetPack4.4.1
• ONNX Runtime installed from (source or binary): Source
• ONNX Runtime version: 1.8 (should be)
• Python version: Python3.8
• GCC/Compiler version (if compiling from source): g++ (Ubuntu/Linaro 8.4.0-3ubuntu2) 8.4.0
• CUDA/cuDNN version: 10.2
• Device : AGX Xavier
• Cmake version : 3.21.1

Goal

To build the whl file for python3.8