Installing tensorrt via yay fails

General system Newbie
1

Hello,

When trying to install tensorrt from AUR, it fails, seemingly due to a CMake version issue.
The last few lines are the following:

[100%] Building CXX object parsers/onnx/CMakeFiles/nvonnxparser_static.dir/TensorOrWeights.cpp.o
[100%] Building CXX object parsers/onnx/CMakeFiles/nvonnxparser_static.dir/errorHelpers.cpp.o
[100%] Linking CXX static library ../../libnvonnxparser_static.a
[100%] Built target nvonnxparser_static
HEAD is now at 914c06fb2 chore: set to version 2.9.2
~/tensorrt/src/TensorRT/python/build ~/tensorrt/src/TensorRT/python
CMake Error at CMakeLists.txt:18 (cmake_minimum_required):
Compatibility with CMake < 3.5 has been removed from CMake.

Update the VERSION argument value. Or, use the … syntax
to tell CMake that the project requires at least but has been updated
to work with policies introduced by or earlier.

Or, add -DCMAKE_POLICY_VERSION_MINIMUM=3.5 to try configuring anyway.

– Configuring incomplete, errors occurred!
==> ERROR: A failure occurred in build().
Aborting…

I’m not sure if this is an endeavourOS, arch, or package problem, hence why I’m asking here first.

Is it possible to determine if it is a problem on my end or if it is universal?

2

The problem is that the archlinux cmake version got updated to a version (version 4.0.0 specifically) that removed compatability with cmake < 3.5. Nothing you can do about that specifically (it is not recommended to downgrade packages).

You can either report this issue upstream (i.e. the maintainer of that software needs to remove/ change this requirement in the CMakeLists.txt file of their source code), or manually build that package yourself and apply the compile flag suggested by the output you posted. (I highly doubt that this is an actual requirement, its just people put in their cmake files and forget about. If it isn’t a hard issue it may get updated soon enough anyways.)

Ultimately an upstream issue.

1 Like
3

Ah, I see. I will create a ticket to report the issue. Thank you very much for the help :smile: