[galabovaa]

Description

Fixes TSAN error that popped up in an unrelated PR:

https://github.com/ERGO-Code/HiGHS/actions/runs/40368145015/job/121731331326?pr=3409

Checklist

• I have read the contributing guidelines
• This PR targets the latest branch
• Tests are passing
• Documentation was updated where relevant
• This PR is not primarily AI-generated (per the AI contributions policy in CONTRIBUTING.md)

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 75.19%. Comparing base (410e953) to head (942463a).
⚠️ Report is 9 commits behind head on latest.

Additional details and impacted files

@@            Coverage Diff             @@
##           latest    #3475      +/-   ##
==========================================
- Coverage   75.19%   75.19%   -0.01%     
==========================================
  Files         460      460              
  Lines      112413   112413              
  Branches    17825    17827       +2     
==========================================
- Hits        84532    84524       -8     
- Misses      27599    27607       +8     
  Partials      282      282              

☔ View full report in Codecov by Harness.
📢 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.

[filikat]

Thanks! This makes sense to me.

If anyone's interested:

• the first load doesn't need to be tagged acquire, because it calls try_acquire() anyway. Since the loop runs many times, it's better to keep it relaxed.
• the exchange must be tagged acquire, or it doesn't synchronise with the release in release().
• the last load needs to be tagged acquire to guard against the possibility (quite remote) of a spurious wakeup of the condition variable, while another thread is pre-empted during release() just before acquiring the mutex.