Thread execution spaces through vector API

[Yurlungur]

PR Summary

This MR resolves issue #628 and pulls in lanl/ports-of-call#109 from ports-of-call. Here I expose the ability to select an execution space in the singularity-eos vector API:

• Each vector API call now optionally accepts an execution space argument. The default is PortsOfCall::Exec::Device() which is (when Kokkos is enabled) an alias to kokkos default execution space.
• To handle the huge number of overloads, I heavily leverage use macros to shrink the number of explicit declarations of vector methods #629 so that the changes are relatively minimal.
• I did need to heavily use SFINAE/concepts to ensure that all the different template specializations played nice.
• The python API now explicitly calls the PortsOfCall::Exec::Host() execution space.
• As usual, EOSPAC is a bit of an odd one out. As far as I know, EOSPAC does not have the ability to select an execution space like this. For now, I simply document that explicitly requesting an execution space with EOSPAC is undefined behavior. It will do whatever the backend decides to do. We should resolve this at some point down the road in collaboration with the EOSPAC devs if we need to.

This MR used generative AI, but agents weren't set loose, so there's no proposed change plan. I made changes by hand and then had the robots copy my changes by example.

I am still getting the CI to pass, but this is ready for review.

PR Checklist

• Adds a test for any bugs fixed. Adds tests for new features.
• Format your changes by using the make format command after configuring with cmake.
• Document any new features, update documentation for changes made.
• Make sure the copyright notice on any files you modified is up to date.
• After creating a pull request, note it in the CHANGELOG.md file.
• LANL employees: make sure tests pass both on the github CI and on the Darwin CI
• If ML was used, make sure to add a disclaimer at the top of a file indicating ML was used to assist in generating the file.
• If Agentic AI was used, have the AI generate a "proposed changes" markdown file and store it in the plan_histories folder, with a filename the same as the MR number.

If preparing for a new release, in addition please check the following:

• Update the version in cmake.
• Move the changes in the CHANGELOG.md file under a new header for the new release, and reset the categories.
• Maintainers: ensure spackages are up to date:
  • LANL-internal team, update XCAP spackages
  • Current maintainer of upstream spackages, submit MR to spack

[Yurlungur]

eliminates an unused variable warning

[Yurlungur]

Naming the SFINAE/concept type produced more useful error messages from the cuda compiler.

[Yurlungur]

The diffs for the modifiers are a bit hard to read for some reason, but what they're doing is adding Space to each vector call, and then a defaulted version that calls PortsOfCall::Exec::Device().

[Yurlungur]

Tests on re-git now passing. But I would like to clean this up a bit further and eliminate more of these redundant overloads.

[adamdempsey90]

Lots of boilerplate, so I didn't give it too close of a look

[adamdempsey90]

While I think s is a fine name, a more descriptive name like exec_space might be even better

[Yurlungur]

Do you think that's necessary even though the type is named Space?

[Yurlungur]

Lots of boilerplate, so I didn't give it too close of a look

Still basically all boilerplate, but significantly reduced by adding more macros a la #629 but for modifier overloads.

[Yurlungur]

Caused a failure on GPUs on the CI. Not sure why it only showed up now.

[Yurlungur]

@adamdempsey90 ping for re-review given the macros I added.

[adamdempsey90]

LGTM