Update to HDF5 2.1.0, other updates from GMAO testing

[mathomp4]

Description

Updates in this PR:

1. Up the version of HDF5 to 2.1.0 by default
2. Add nccmp@1.11.0.0 as an any_of requirement. This is done as at one of the spack-stack telecons, someone from EPIC said they still need the older 1.9 version...I think
3. Updates for geos-gcm-env and the GMAO scripts

Dependencies

None

Issues addressed

Closes #1822

Applications affected

All apps that use HDF5, so probably all apps.

Systems affected

macos.gmao

Testing

• CI: Note whether the automatic tests (GitHub actions tests that run automatically for every commit) pass or not
  • GitHub actions CI tests pass
  • GitHub actions CI tests do not pass (provide explanation)
  • GitHub actions CI tests skipped (provide explanation if necessary)
• New tests added: List and describe any new tests added to GitHub actions
  • ...
• Additional testing: Add information on any additional tests conducted
  • ...

Checklist

• This PR addresses one issue/problem/enhancement or has a very good reason for not doing so.
• These changes have been tested on the affected systems and applications.
• All dependency PRs/issues have been resolved and this PR can be merged.
• All necessary updates to the documentation (spack-stack wiki) will be made when this PR is merged

[rickgrubin-noaa]

v1.9.0.1 still currently required by EPIC / UFS WM.

[mathomp4]

@rickgrubin-noaa This should allow it to be built. The question is: what version is "default" as it were. If that is:

      - any_of:
        - '@1.11.0.0'
        - '@1.9.0.1'

then anything requesting nccmp will get nccmp@1.11.0.0. But if we did:

      - any_of:
        - '@1.9.0.1'
        - '@1.11.0.0'

then nccmp → nccmp@1.9.0.1.

One of us will need to add a version pin somewhere.

So I guess a question for @climbfuji, et al, is what is the right way for spack-stack?

[rickgrubin-noaa]

1.9.0.1 can be set in a site-specific packages yaml file and any_of in either order would suffice, I think.

[climbfuji]

Is there a reason UFS is always pulling back? This makes no sense to me. It's just a file comparison tool. I would think (naively, perhaps) that when updating to spack-stack-2.2 with a new hdf5 and new nccmp, butterfly tests will need to be conducted to move on.

[mathomp4]

You'd think the order wouldn't matter but per Spack docs:

For any_of and one_of, the order of specs indicates a preference: items that appear earlier in the list are preferred (note that these preferences can be ignored in favor of others).

[mathomp4]

Aww nuts. Per the CI, I missed something with HDF5 requirements.

[mathomp4]

Crap. It's py-h5py. I didn't try building that.

@neutrinoceros does py-h5py support HDF5 v2? From the "what's new" for v3.16 I see:

Pre-built packages on PyPI are now built with HDF5 2.0 (PR 2750)

So I'd think so, but the spack package just says:

    depends_on("hdf5@1.10.7:1.14 +hl", when="@3.15:")

We could do:

    depends_on("hdf5@1.10.7: +hl", when="@3.16:")
    depends_on("hdf5@1.10.7:1.14 +hl", when="@3.15")

I guess now? not sure...

[mathomp4]

Oh boy. I'm testing the py-h5py edit and I think that pretty much triggers the "We need to move to Python 3.12/3.13" path. Maybe. Grah.

cf GMAO-SI-Team#1