Delete lock interface implementation from UmfpackLU

[fredrikekre]

No description provided.

[codecov]

Codecov Report

Attention: Patch coverage is 84.61538% with 2 lines in your changes missing coverage. Please review.

Project coverage is 84.01%. Comparing base (4f8d075) to head (0b19204).
Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
src/solvers/umfpack.jl	84.61%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #617      +/-   ##
==========================================
+ Coverage   83.95%   84.01%   +0.06%     
==========================================
  Files          12       12              
  Lines        9265     9256       -9     
==========================================
- Hits         7778     7776       -2     
+ Misses       1487     1480       -7     

☔ 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.

[SobhanMP]

I'm not critical of the patch, but why? IMHO, it's fair to warn people that they are waiting for locks to do sparse operations.

[fredrikekre]

The motivation wasn't really to remove the warning but rather that it is kind of weird that a struct UmfpackLU is lockable, it isn't an AbstractLock for example.

Regarding the warning, since it isn't really actionable by the user, and safe to ignore, it seems kind of pointless?

[ViralBShah]

I think it would make sense to document it rather than give a warning.

[SobhanMP]

The motivation wasn't really to remove the warning but rather that it is kind of weird that a struct UmfpackLU is lockable, it isn't an AbstractLock for example.

Makes sense

Regarding the warning, since it isn't really actionable by the user, and safe to ignore, it seems kind of pointless?

It is actionable, you can pass a different workspace.

I think it would make sense to document it rather than give a warning.

Nobody expects solve to lock. IMHO, it's better to create a new workspace than to lock w/o a warning.

[fredrikekre]

Perhaps it would be better with a documentation section for multithreaded use then? The warning doesn't tell the user what to do or how to do it so that is what I mean with unactionable. Since JuliaLang/julia#52883 you can also see lock contention in @time which would help detecting the locking without logging a warning.

[KristofferC]

I agree with both things here. It is weird to implement these lock methods for UmfpackLU, if anything a user should probably use Lockable and for the warning message it isn't something that is typically done in Julia libraries.