Enable osx_arm64

[RaulPPelaez]

Checklist

• Used a personal fork of the feedstock to propose changes
• Bumped the build number (if the version is unchanged)
• Reset the build number to 0 (if the version changed)
• Re-rendered with the latest conda-smithy (Use the phrase @conda-forge-admin, please rerender in a comment in this PR for automated rerendering)
• Ensured the license file is being packaged.

Closes #25

[conda-forge-webservices]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

[RaulPPelaez]

Same error as in #25, swig does not like the CPU
/bin/sh: /Users/runner/miniforge3/conda-bld/openmm-plumed_1705661581832/_h_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_place/bin/swig: Bad CPU type in executable

[hadim]

Maybe try to add - swig # [build_platform != target_platform] in the build section?

[RaulPPelaez]

No luck @hadim! Also tried to add this to conda-forge.yaml:

test_on_native_only: true

[hadim]

I tested locally with an M3 and the build works. It means it's indeed related to cross-compilation. But not sure what is causing this...

[RaulPPelaez]

I reckon it is related with swig, a build tool, being downloaded for arm64, but we need it compiled for the build machine. Is there a way to download a package for the host architecture?

[peastman]

Here's the corresponding section from the openmm-torch recipe, which successfully cross-compiles.

requirements:
  build:
    - python                                 # [build_platform != target_platform]
    - cross-python_{{ target_platform }}     # [build_platform != target_platform]
    - numpy >=1.19                           # [build_platform != target_platform]
    - swig <4.1                              # [build_platform != target_platform]
    - pytorch                                # [build_platform != target_platform]
    - pytorch =*={{ torch_proc_type }}*      # [build_platform != target_platform]
    - openmm >=8.0.0                         # [build_platform != target_platform]
    - {{ compiler('c') }}
    - {{ compiler('cxx') }}
    - {{ compiler('cuda') }}  # [cuda_compiler_version not in (undefined, 'None')]
    - {{ cdt('mesa-libgl-devel') }}  # [linux]
    - cmake
    - make

The one obvious difference is that it also includes OpenMM. Try adding

    - openmm >=8.1.0                         # [build_platform != target_platform]

Note that OpenMM-Plumed now requires 8.1 or later

[RaulPPelaez]

I wonder why the same does not work here as in OpenMM-Torch....

[RaulPPelaez]

omg, I fixed arm but broke the other one -.-

[RaulPPelaez]

@conda-forge-admin, please rerender

[RaulPPelaez]

Looks like I pulled it off!
@hadim Would you please review?

[hadim]

I am not a maintainer but it looks good to me.

[RaulPPelaez]

Oh right -.-
Thanks!