Containerize GCC & Intel CI workflows

[AlexanderRichert-NOAA]

DESCRIPTION OF CHANGES:

This PR containerizing the GCC and Intel CI workflows based on the containers generated by https://github.com/NOAA-EMC/ci-common-build-cache.

Note: I haven't looked yet at precaching test files, either in the UFS_UTILS container images or through the Actions cache.

TESTING

PR CI runs successfully with updated workflows (including CTest & installation).

DEPENDENCIES:

none

DOCUMENTATION:

Workflows are internally documented with comments, including links to https://github.com/NOAA-EMC/ci-common-build-cache documentation.

LINKED ISSUES:

• closes (replaces) Fix more CI issues by bringing in ESMF to spack in CI and adding PIO dep in UFS_UTILS #1156

[AlexanderRichert-NOAA]

@BrianCurtis-NOAA putting this here for discussion/proof of concept for containerizing workflows beyond NCEPLIBS.

[AlexanderRichert-NOAA]

In practice it looks like the repo-level caching for the oneAPI compilers only saves about 15-20 seconds, so, it may or may not be worth the cache space with respect to the 10GB limit.

[BrianCurtis-NOAA]

@AlexanderRichert-NOAA I assume this method allows for easy upgrading to newer versions as needed? Is there a way to tell the total space it's using in our 10GB limit?

[AlexanderRichert-NOAA]

That's the idea behind the commented block about modifying the Spack installation. We could try it out here if maybe you have an example of a package that gets updated frequently.

I don't think GitHub directly reports the total usage, but you can basically look under https://github.com/ufs-community/UFS_UTILS/actions/caches and add everything up (if this PR were merged, I think it would just be the Intel installation; conceivably in the future test files could get cached there as well, or in the container images).

[BrianCurtis-NOAA]

@AlexanderRichert-NOAA I think this is the way forward, do you need to check/do anything else before merging?

[AlexanderRichert-NOAA]

No, I think this is good. If I don't hear from you sooner I'll check in in a month or so just to see how it's going and if there are any lingering issues.