Add cuda_compiler to zip_keys

[jakirkham]

Adds cuda_compiler to zip_keys alongside cuda_compiler_version. This is needed as part of the move to using cuda-nvcc_{{ target_platform }} in the future (as opposed to nvcc_{{ target_platform }}). Though for now this still uses nvcc_{{ target_platform }} in all cases (merely providing a place to add cuda-nvcc_{{ target_platform }} later).

xref: #4400

---

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.

[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.

[jakirkham]

cc @bdice @leofang @adibbley @vyasr

[jakirkham]

cc @conda-forge/core

[isuruf]

You need to fix the cuda migrators too

[jakirkham]

Good point. Those should fixed now. Please take another look and let me know if anything else is needed

[isuruf]

Suggested change

	cuda_compiler: # [linux64 or win]
	- nvcc # [linux64 or win]
	cuda_compiler:
	- nvcc # [linux or win]

[jakirkham]

This was changed in response to this suggestion above ( #4407 (comment) )

Could you please share a bit more about why we want to revert that?

Should add the CUDA arch migrator now contains this key with appropriate length

[isuruf]

You are supposed to only change that for Line 46 and below.

[isuruf]

The selectors used should match that of cuda_compiler_version

[jakirkham]

Agreed. Had another question about that below

[jakirkham]

Sounds like we were ok with the suggestion below. So committed that

That said, if I'm still missing something here, please let me know

[isuruf]

Can you check that this works in an existing feedstock locally by passing -e <path/to/conda-forge-pinning-feedstock/recipe/conda_build_config.yaml to conda-smithy rerender?

[jakirkham]

Yep, submitted PR ( conda-forge/ucx-split-feedstock#121 ) with the result of that change

Should add the CUDA arch migrator changes were needed as well

[jakirkham]

Thanks Isuru! 🙏

[jakirkham]

Tested also in PR ( conda-forge/arrow-cpp-feedstock#1038 ) to see if cuda_compiler being None for CPU caused any issues, but it seems that CI job is working. So this looks good

[jakirkham]

If anything comes up, please feel free to ping me and I will take a look 🙂