Build issues in eOn v2.12 on Linux

[LEGION0515]

Dear eOn developers,

I’m really sorry to bother you. I noticed that eOn recently released a new version, so I tried to pull the latest code and build it. When building eOn version 2.12.0, I ran the following commands:

conda create -n eon python=3.14 numpy pyyaml eigen spdlog meson ninja pkg-config openmpi="4.1.*" libpng zlib gcc=14 gxx=14 gfortran=14 quill -c conda-forge -y

conda activate eon

git clone https://github.com/TheochemUI/eOn.git

cd eOn

meson setup bbdir
--prefix=$CONDA_PREFIX
--libdir=lib
--buildtype=release

meson compile -C bbdir -j2

After executing these commands, I encountered the following errors during the build process of eOn:

I tried to follow the hints and made some changes to client/LBFGS.cpp and related files, but unfortunately, that didn’t fix the problem. It seems like it might be an issue with some definitions.

My technical skills are quite limited, and I haven’t been able to solve this on my own. I’m also not sure if fixing this error might cause other errors to appear.

I honestly don’t know how to proceed and would be extremely grateful for any guidance you could give me.

Also, while compiling LAMMPS and MPI, I ran into some issues. I tried to resolve them, and the build did succeed, but I’m not entirely sure if everything is correct. I just wanted to share this with you.

[HaoZeke]

This is rather odd, since we have this building on conda-forge .. https://prefix.dev/channels/conda-forge/packages/eon / https://anaconda.org/channels/conda-forge/packages/eon/overview with the recipe here https://github.com/conda-forge/eon-feedstock

However, I'll be wrapping up #329 over the weekend which is a rather complete overhaul so maybe you could try to compile the new release next week and if that doesn't work we should look into this more..

[HaoZeke]

Actually maybe this should be as simple as installing sysroot_linux-64>=2.35 in the conda environment.. could you check?

[LEGION0515]

Thank you very, very much for your reply, I really appreciate it!

I understand, and I will wait for the new release and try again next week.

Regarding what I mentioned earlier about LAMMPS and MPI, I would also like to briefly share my experience. When I was compiling eOn 2.11 with these interfaces enabled, I encountered some errors.

For the LAMMPS interface, I found that after building LAMMPS, make install places lammps.h under $CONDA_PREFIX/include/lammps/. To make it compile, I modified client/potentials/LAMMPS/LAMMPSPot.cpp by changing

#include "LAMMPS.h"

to

#include "lammps/lammps.h"

and also added

#include "LAMMPSPot.h"

at the top.

I’m not sure if this is the correct way to handle it, but with these changes I was able to successfully build it.

As for the MPI interface, I also encountered some errors. It seems that this might be related to using a newer version of OpenMPI (version 4.x), which may not be fully compatible with the existing code.

I’m not sure whether these issues have already been addressed in the upcoming new release, especially since I haven’t yet been able to successfully build eOn 2.12 even without enabling these interfaces. I just wanted to share these observations in case they are helpful.

Thank you again for your time and all your efforts on eOn, I really appreciate your work.

[LEGION0515]

Thank you very much for your suggestion, I really appreciate it!

I tried installing sysroot_linux-64>=2.35 and rebuilding, but unfortunately the issue still persists.

I will wait for the new release next week and try again with the updated code.

Thanks again for your help, and I wish you a nice day! 😊😊😊

[HaoZeke]

Sorry this took way longer, had a bunch of new features, could you see if this is still an issue on 2.14? @LEGION0515

[HaoZeke]

Closing for now but do let me know if this needs to be reopened.