Add bin/compiler to PATH for intel-oneapi-compilers

[climbfuji]

Description

In spack-ext/lib/jcsda-emc/spack-stack/stack/meta_modules.py: for intel-oneapi-compilers, attempt to add subdirectory compiler of bin to PATH, and test in CI.

This will have no effect on Cray systems using the Cray compiler wrappers, but that's fine. Cray systems have their own, Cray-generated compiler modules anyway.

Dependencies

None

Issues addressed

Closes #1903

Applications affected

None

Systems affected

Systems using Intel oneAPI compilers

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
  • Tested on Nautilus where this issue was identified first

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

[AlexanderRichert-NOAA]

Regarding #1903, shouldn't this be resolved in the system-provided compiler modules? It's not a hill I'll die on, but, could/should it be applied on a site-by-site basis?

[climbfuji]

Regarding #1903, shouldn't this be resolved in the system-provided compiler modules? It's not a hill I'll die on, but, could/should it be applied on a site-by-site basis?

The system-provided modules are often straight from Intel, and Intel doesn't set the path for compiler (c.f. spack-stack discussion earlier today).

[AlexanderRichert-NOAA]

That still strikes me as something for SAs to address (either by manually adding it to the modules or by engaging Intel and asking them to fix the modules), but like I said, not a hill I intend to die on.

[climbfuji]

That still strikes me as something for SAs to address (either by manually adding it to the modules or by engaging Intel and asking them to fix the modules), but like I said, not a hill I intend to die on.

I know, we are covering up for an issue in the Intel-generated module files. Fortunately, the fix is easy enough to remove.