Skip to content

Conversation

@odow
Copy link
Collaborator

@odow odow commented Dec 16, 2025

Follow-up to #2699

We need to remove the manually installed versions of metis and gklib in CI to ensure we're actually using the vendored copies.

@odow odow requested review from filikat and galabovaa December 16, 2025 20:50
@codecov
Copy link

codecov bot commented Dec 16, 2025

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 81.29%. Comparing base (cd2bd0b) to head (07bcb09).
⚠️ Report is 39 commits behind head on latest.

Additional details and impacted files
@@           Coverage Diff           @@
##           latest    #2702   +/-   ##
=======================================
  Coverage   81.29%   81.29%           
=======================================
  Files         349      349           
  Lines       85786    85786           
=======================================
+ Hits        69736    69738    +2     
+ Misses      16050    16048    -2     

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

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

@odow
Copy link
Collaborator Author

odow commented Dec 16, 2025

There's a segfault in the unit tests which seems suspicious.

@filikat
Copy link
Collaborator

filikat commented Dec 16, 2025

There's a segfault in the unit tests which seems suspicious.

That test is not compiled with HiPO, so it's a bit weird.
However, it made me realise two things that we are doing wrong @galabovaa:
1 - TestHipo.cpp should only be included in the unit tests if HIPO=ON.
2 - if solver=hipo is used in a build that does not include HiPO, at the moment IPX is selected, because of how lp_data/HighsSolve.cpp:useHipo is set up (line 724). It would be better to give the error in lp_data/HighsSolve.cpp:93, which I don't think is reachable at the moment.

@galabovaa
Copy link
Contributor

Noted, @filikat

I've got updates to the workflows on a WIP branch. I have updated build_tarballs there too, so I'm closing this, @odow

@galabovaa galabovaa closed this Dec 16, 2025
@odow
Copy link
Collaborator Author

odow commented Dec 16, 2025

Can we please consider moving to a PR-based workflow over branches? It's hard to tell what people are working on.

@odow odow deleted the od/rm-metis branch December 16, 2025 21:56
@odow
Copy link
Collaborator Author

odow commented Dec 16, 2025

I think this is also a good motivation to simplify some of the CI workflows with a build matrix. We shouldn't need to copy-paste code in so many places. (e.g., where I was headed with #2551)

@galabovaa
Copy link
Contributor

Thanks for the suggestion, but we will retain them for now

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants