build and test fails when Eigen Gitlab tests fail

[ncdorn]

Description

since the latest pull request #166 was merged, build and test is not passing on macOS-latest. There is an issue building the pysvzerod package. Perhaps I forgot to rebase my branch. will look into this

Reproduction

see macOS-latest workflow runs for #166

Expected behavior

pysvzerod should build without issue

Additional context

No response

Code of Conduct

• I agree to follow this project's Code of Conduct and Contributing Guidelines

[ncdorn]

After further investigation this appears to be caused by the latest Eigen commit yesterday. certain Build and Test workflow run will fail anytime the master branch of Eigen is not passing its own testing on that OS, which happens quite often. Maybe we should pin the Eigen Fetch to a good commit? Unless we don't care about the macOS-latest build passing at all times.

[mrp089]

It would be good if sv0D always works with the latest versions of external packages. Otherwise, we risk being unable to build it at some point (like SimVascular). @ncdorn, can you look into why the build fails? Maybe the changelog can give a hint. Sometimes it's just a small change that causes the build to fail.

[ncdorn]

So I think this might be from Eigen's end. It appears that more often than not people will push commits to the Eigen repository which do not pass their full testing pipeline, which tests the build on different architectures. In a commit last night, The tests for build:linux:cross:aarch64 failed, while the builds for all other architectures are passing (see their pipeline run here). The errors thrown in Eigen's testing and in our build and test pipeline are similar in nature. I assume that aarch64 in their case refers to the 64-bit version of ARM upon which silicon Macs are built. This would explain why only macOS-latest is failing for us. the good news is that often when people push these buggy commits, someone will fix the bug soon after. a few hours ago someone pushed a commit to the Eigen repository in order fix the bug in question, so we should find out soon if their tests all pass and subsequently if ours pass. I think if our tests pass whenever all Eigen's own tests pass we can probably close this issue, but let me know if you think otherwise @mrp089

[ncdorn]

so yeah now that the linux:cross:aarch64 tests are passing on Eigen's side, the macOS-latest test is passing for us. I think we can close this for now as our software is still very much up to date with a current functional Eigen master branch.