Fix two spinlocks

[wddgit]

PR description:

Fix two spin locks that had bugs. One in LogErrorEventFilter and one in Path. Both could cause rare data races and crashes.

The one in LogErrorEventFilter was causing rare crashes in the IBs recently and that caused us to notice the problem. This is a minimal fix. We are going to discuss this more and possibly make further changes related to this in the future. See Issue #44413 for discussion.

The problem in Path has been there 4 years and no one has noticed or connected any problems to it. The problem in LogErrorEventFilter has been there 5 years and the problems were only recently noticed. I'm not sure if this is worth back porting or not. The problem should be occurring only very rarely...

In part the fix is a test. The problems are not reproducible so we are not 100% sure this fixes the problem. We plan to observe the IBs after this is merged and see if the recently noticed problems stop occurring.

PR validation:

Relies on existing tests.

[cmsbuild]

cms-bot internal usage

[cmsbuild]

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-44517/39640

• This PR adds an extra 28KB to repository

[wddgit]

enable threading

[cmsbuild]

A new Pull Request was created by @wddgit for master.

It involves the following packages:

• DPGAnalysis/Skims (pdmv)
• FWCore/Framework (core)

@Dr15Jones, @cmsbuild, @sunilUIET, @AdrianoDee, @smuzaffar, @miquork, @makortel can you please review it and eventually sign? Thanks.
@youyingli, @AnnikaStein, @makortel, @missirol this is something you requested to watch as well.
@rappoccio, @sextonkennedy, @antoniovilela you are the release manager for this.

cms-bot commands are listed here

[wddgit]

please test with #44447

[wddgit]

type bug

[cmsbuild]

+1

Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-3ea906/38353/summary.html
COMMIT: f25ce3b
CMSSW: CMSSW_14_1_X_2024-03-22-1100/el8_amd64_gcc12
Additional Tests: THREADING
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week1/cms-sw/cmssw/44517/38353/install.sh to create a dev area with all the needed externals and cmssw changes.

Comparison Summary

Summary:

• You potentially removed 86 lines from the logs
• Reco comparison results: 48 differences found in the comparisons
• DQMHistoTests: Total files compared: 48
• DQMHistoTests: Total histograms compared: 3297369
• DQMHistoTests: Total failures: 0
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 3297349
• DQMHistoTests: Total skipped: 20
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 0.0 KiB( 47 files compared)
• Checked 202 log files, 165 edm output root files, 48 DQM output files
• TriggerResults: no differences found

[makortel]

Comparisons show the usual

• Non reproducibility in wf 136.793 #43293
• Non-reproducibility in TriggerResults in 24834.0, 24834.911, and 25034.999 #43790

[makortel]

+core

[makortel]

The problem in Path has been there 4 years and no one has noticed or connected any problems to it. The problem in LogErrorEventFilter has been there 5 years and the problems were only recently noticed. I'm not sure if this is worth back porting or not. The problem should be occurring only very rarely...

My feeling is that #43522 somehow caused globalEndLumi transitions to be actually run concurrently. This (and #44447) could be worth of backporting nevertheless (or, doing the backports could be cheaper than trying to fully understand why this situation didn't happen before). The fixed code are wrong in any case.

[antoniovilela]

+1

[AdrianoDee]

+pdmv

[cmsbuild]

This pull request is fully signed and it will be integrated in one of the next master IBs (tests are also fine). This pull request will be automatically merged.