TensorRT cross compile for Jetson.. nvcc fatal


[54%] Linking CUDA device code CMakeFiles/nvinfer_plugin.dir/cmake_device_link.o
nvcc fatal : ‘–device-link (-dlink)’ is not compatible with ‘–no-device-link (-nodlink)’
pluginin/CMakeFiles/nvinfer_plugin.dir/build.make:973: recipe for target ‘plugin/CMakeFiles/nvinfer_plugin.dir/cmake_device_link.o’ failed
make[2]: *** [plugin/CMakeFiles/nvinfer_plugin.dir/cmake_device_link.o] Error 1
CMakeFiles/Makefile2:308: recipe for target ‘plugin/CMakeFiles/nvinfer_plugin.dir/all’ failed
make[1]: *** [plugin/CMakeFiles/nvinfer_plugin.dir/all] Error 2
make[1]: *** Waiting for unfinished job…


TensorRT Version: 7.2.1
GPU Type: GeForce GTX 1660
Nvidia Driver Version: 460.32.03
CUDA Version: 10.2
CUDNN Version: 8.0
Operating System + Version: ubuntu 18.04
Python Version (if applicable):
TensorFlow Version (if applicable):
PyTorch Version (if applicable):
Baremetal or Container (if container which image + tag):

Relevant Files

Please attach or include links to any models, data, files, or scripts necessary to reproduce your issue. (Github repo, Google Drive, Dropbox, etc.)

Steps To Reproduce

I want to build TensorRT source code. https://github.com/NVIDIA/TensorRT
I followed the guidelines here.

Launch the TensorRT-OSS build container.

./docker/launch.sh --tag tensorrt-cross-jetpack --gpus all --release ~/Downloads/TensorRT- --source ~/TensorRT

Example: Ubuntu 18.04 Cross-Compile for Jetson (arm64) with cuda-10.2 (JetPack)

cd TensorRT
mkdir -p build && cd build
cmake … -DTRT_LIB_DIR=/pdk_files/tensorrt/lib -DTRT_OUT_DIR=‘pwd’/out -DCMKAE_TOOLCHAIN_FILE=/workspace/TensorRT/cmake/toolchains/cmake_aarch64.toolchain -DCUDA_VERSION=10.2 -DGPU_ARCHS=61 -DCUDNN_LIB=/pdk_files/cudnn/usr/lib/aarch64-linux-gnu/libcudnn.so -DCUBLAS_LIB=/usr/lib/aarch64-linux-gnu/libcublas.so -DCUBLASLT_LIB=/usr/lib/aarch64-linux-gnu/libcublas.so

make -j$(nproc)

If you tell me more information I need, I’ll let you know right away.

I referred to the advice in this link.


Thanks for your reporting.
We are going to reproduce this issue and get back to you later.


Sorry for the late update.
Since the latest TensorRT version for Jetson is still v7.1.3, please use release/7.1 branch instead.