Build wheels for x86_64, intel and arm64 macOS

[andreas-el]

At this point i'm unsure if we should fix occasionally failing tests or mark them.
I have enabled building of wheels and testing according to the table provided. #7204 (comment)

ref.
https://github.com/equinor/komodo-releases/actions/runs/7932266339

---

• PR title captures the intent of the changes, and is fitting for release notes.
• Added appropriate release note label
• Commit history is consistent and clean, in line with the contribution guidelines.
• Make sure tests pass locally (after every commit!)

When applicable

• When there are user facing changes: Updated documentation
• New behavior or changes to existing untested code: Ensured that unit tests are added (See Ground Rules).
• Large PR: Prepare changes in small commits for more convenient review
• Bug fix: Add regression test for the bug
• Bug fix: Create Backport PR to latest release

[pinkwah]

Should we not test Intel macs?

[andreas-el]

Edit: This seems to be resolved somehow.

Currently fails due to precision.

/Users/runner/work/ert/ert/src/clib/tests/enkf/test_row_scaling.cpp:141: FAILED:
  CHECK( A(row, col) == Approx(expected).epsilon(1e-12) )
with expansion:
  -0.0000018113 == Approx( -0.0000018113 )

[andreas-el]

Should we not test Intel macs?

Good question.
My impression was that these workflows should match what the developers were using, but there might be some devs still using intel architecture still (for macOS).
Also, should we build wheels for all python versions, and all os + architecture combinations?
I think it would be nice to get some clarification regarding this.

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 85.26%. Comparing base (7f656ef) to head (9d3d09b).
Report is 4 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #7204      +/-   ##
==========================================
+ Coverage   84.55%   85.26%   +0.70%     
==========================================
  Files         382      382              
  Lines       22770    22770              
  Branches      886      891       +5     
==========================================
+ Hits        19253    19414     +161     
+ Misses       3404     3250     -154     
+ Partials      113      106       -7     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[sondreso]

We still have developers on macos intel, so we need to support this.

I borrowed this overview to illustrate which platforms we should build wheels for (this is the same as the latest release):

	macOS Intel	macOS Apple Silicon	Windows 64bit	Windows 32bit	Windows Arm64	manylinux musllinux x86_64	manylinux musllinux i686	manylinux musllinux aarch64	manylinux musllinux ppc64le	manylinux musllinux s390x
CPython 3.6	❌	N/A	❌	❌	N/A	❌	❌	❌	❌	❌
CPython 3.7	❌	N/A	❌	❌	N/A	❌	❌	❌	❌	❌
CPython 3.8	✅	❌	❌	❌	N/A	✅	❌	❌	❌	❌
CPython 3.9	✅	❌	❌	❌	❌	✅	❌	❌	❌	❌
CPython 3.10	✅	✅	❌	❌	❌	✅	❌	❌	❌	❌
CPython 3.11	✅	✅	❌	❌	❌	✅	❌	❌	❌	❌
CPython 3.12	✅	✅	❌	❌	❌	✅	❌	❌	❌	❌
PyPy 3.7 v7.3	❌	N/A	❌	N/A	N/A	❌	❌	❌	N/A	N/A
PyPy 3.8 v7.3	❌	❌	❌	N/A	N/A	❌	❌	❌	N/A	N/A
PyPy 3.9 v7.3	❌	❌	❌	N/A	N/A	❌	❌	❌	N/A	N/A
PyPy 3.10 v7.3	❌	❌	❌	N/A	N/A	❌	❌	❌	N/A	N/A

• Note 1: We are not building for musllinux.
• Note 2: We should not publish a source distribution.

In PRs, I think we should test against:

• Oldest Python version on all platforms (3.8 on macos intel, 3.8 on linux x86_64), but we skip Python 3.10 on Mac Arm
• Newest Python version on all platforms (3.12)
• Current and future Python versions for internal users on target platform linux x86_64 (Python 3.8 and 3.11)

Giving the following test matrix for PRs:

	macOS Intel	macOS Apple Silicon	Windows 64bit	Windows 32bit	Windows Arm64	manylinux musllinux x86_64	manylinux musllinux i686	manylinux musllinux aarch64	manylinux musllinux ppc64le	manylinux musllinux s390x
CPython 3.6	❌	N/A	❌	❌	N/A	❌	❌	❌	❌	❌
CPython 3.7	❌	N/A	❌	❌	N/A	❌	❌	❌	❌	❌
CPython 3.8	🔄	❌	❌	❌	N/A	🔄	❌	❌	❌	❌
CPython 3.9	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌
CPython 3.10	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌
CPython 3.11	❌	❌	❌	❌	❌	🔄	❌	❌	❌	❌
CPython 3.12	🔄	🔄	❌	❌	❌	🔄	❌	❌	❌	❌
PyPy 3.7 v7.3	❌	N/A	❌	N/A	N/A	❌	❌	❌	N/A	N/A
PyPy 3.8 v7.3	❌	❌	❌	N/A	N/A	❌	❌	❌	N/A	N/A
PyPy 3.9 v7.3	❌	❌	❌	N/A	N/A	❌	❌	❌	N/A	N/A
PyPy 3.10 v7.3	❌	❌	❌	N/A	N/A	❌	❌	❌	N/A	N/A

Sounds reasonable? 🙂

[andreas-el]

Lint GitHub Actions workflows / Check all workflow files fails

.github/workflows/build_and_test.yml:23:43: label "macos-latest-xlarge" is unknown. available labels are "windows-latest", "windows-2022", "windows-2019", "windows-2016", "ubuntu-latest", "ubuntu-22.04", "ubuntu-20.04", "ubuntu-18.04", "macos-latest", "macos-latest-xl", "macos-13-xl", "macos-13", "macos-13.0", "macos-12-xl", "macos-12", "macos-12.0", "macos-11", "macos-11.0", "macos-10.15", "self-hosted", "x64", "arm", "arm64", "linux", "macos", "windows". if it is a custom label for self-hosted runner, set list of labels in actionlint.yaml config file [runner-label]
   |
23 |         os: ['ubuntu-latest', 'macos-13', 'macos-latest-xlarge']

Additional macOS labels were added recently, only awaiting a new release for actionlint.
rhysd/actionlint#392

edit: This was resolved in the latest version of actionlint

[eivindjahren]

10/10