Improvements to L1 tracking code

[tomalin]

PR description:

This consists of a number of improvements from the L1 tracking group.

The biggest one is the implementation of a bit-accurate emulation of the L1 track duplicate removal algorithm (DR.cc) together with modifications of nearby code in the L1 tracking chain to incorperate it (e.g. DRin.cc, KFin.cc).

Other improvements include test code for study partially dead tracker scenarios (StubKiller.cc) and numerous small improvements to improve agreement with firwamre.

N.B. Despite the branch name, this has actually been rebased of CMSSW_14_0_0_pre0.

PR validation:

This has been checked by the detailed git CI of the L1 track group. The tracking performance remains good on 1000 ttbar + 0 pile-up events.

P.S. @skinnari shares responsibility for the code in this PR.

[cmsbuild]

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-42663/36729

• This PR adds an extra 392KB to repository

• Found files with invalid states:

  • L1Trigger/TrackTrigger/data/L1_TrackQuality_GBDT_emulation_digitized.json:
    • Added: 32b827f
    • Deleted: be9af48
  • L1Trigger/TrackTrigger/data/clf_GBDT_emulation_newKF_digitized.json:
    • Added: 3d02f8f
    • Deleted: 32b827f
  • .github/workflows/github_CI.yml:
    • Added: 087f536
    • Modified: 0e1ba25, 6609905
    • Deleted: 291a5af

[cmsbuild]

A new Pull Request was created by @tomalin (Ian Tomalin) for master.

It involves the following packages:

• L1Trigger/TrackFindingTMTT (l1)
• L1Trigger/TrackFindingTracklet (l1)
• L1Trigger/TrackTrigger (upgrade, l1)
• L1Trigger/TrackerDTC (upgrade, l1)
• L1Trigger/TrackerTFP (l1)
• SimTracker/TrackTriggerAssociation (l1, simulation)

@civanch, @epalencia, @mdhildreth, @cmsbuild, @AdrianoDee, @srimanob, @aloeliger can you please review it and eventually sign? Thanks.
@erikbutz, @VourMa, @mtosi, @GiacomoSguazzoni, @JanFSchulte, @rovere, @VinInn, @Martin-Grunewald, @missirol, @skinnari, @sviret, @mmusich, @threus, @dgulhan this is something you requested to watch as well.
@perrotta, @dpiparo, @antoniovilela, @rappoccio you are the release manager for this.

cms-bot commands are listed here

[tomalin]

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-42663/36729

• This PR adds an extra 392KB to repository

• Found files with invalid states:

  • L1Trigger/TrackTrigger/data/L1_TrackQuality_GBDT_emulation_digitized.json:

    • Added: 32b827f
    • Deleted: be9af48

  • L1Trigger/TrackTrigger/data/clf_GBDT_emulation_newKF_digitized.json:

    • Added: 3d02f8f
    • Deleted: 32b827f

  • .github/workflows/github_CI.yml:

    • Added: 087f536
    • Modified: 0e1ba25, 6609905
    • Deleted: 291a5af

We added all these files a few months ago, for the convenience of the L1 tracking group, but deleted them before making this PR, as they don't belong in central CMSSW. This seems to confuse git, which declares them to be in an "invalid state".

[aloeliger]

+code-checks
Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-42663/36729

• This PR adds an extra 392KB to repository

• Found files with invalid states:

  • L1Trigger/TrackTrigger/data/L1_TrackQuality_GBDT_emulation_digitized.json:

    • Added: 32b827f
    • Deleted: be9af48

  • L1Trigger/TrackTrigger/data/clf_GBDT_emulation_newKF_digitized.json:

    • Added: 3d02f8f
    • Deleted: 32b827f

  • .github/workflows/github_CI.yml:

    • Added: 087f536
    • Modified: 0e1ba25, 6609905
    • Deleted: 291a5af

We added all these files a few months ago, for the convenience of the L1 tracking group, but deleted them before making this PR, as they don't belong in central CMSSW. This seems to confuse git, which declares them to be in an "invalid state".

@tomalin This is typical for any PR which adds and deletes files in the same PR. If it is an issue for the PR history, these commits can be squashed together to remove it.

[aloeliger]

@tomalin Just wanted to ping you on the review comments.

[aloeliger]

@tomalin Pinging you again on the review comments

[cmsbuild]

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-42663/37086

• This PR adds an extra 396KB to repository

• Found files with invalid states:

  • L1Trigger/TrackTrigger/data/L1_TrackQuality_GBDT_emulation_digitized.json:
    • Added: 32b827f
    • Deleted: be9af48
  • L1Trigger/TrackTrigger/data/clf_GBDT_emulation_newKF_digitized.json:
    • Added: 3d02f8f
    • Deleted: 32b827f
  • .github/workflows/github_CI.yml:
    • Added: 087f536
    • Modified: 0e1ba25, 6609905
    • Deleted: 291a5af

[cmsbuild]

Pull request #42663 was updated. @aloeliger, @epalencia, @mdhildreth, @civanch, @srimanob, @AdrianoDee, @cmsbuild can you please check and sign again.

[srimanob]

@tomalin
Could you please confirm if I should trigger the test, or something still to come for reviewing? Thx.

[tomalin]

@tomalin Could you please confirm if I should trigger the test, or something still to come for reviewing? Thx.

@srimanob I've been away, so am just getting back to this. We must address the comments reviewers made above, which will mean small changes to the code. I do not expect any large changes to the current PR.

[aloeliger]

+l1

• @tomalin The MatchEngineUnit.h naming scheme really should be fixed for next time because the variable name + some number of underscores is one of those things that is easy to keep putting off, but could cause a lot of confusion..

[civanch]

+1

Nothing changed for simulation.

[tomalin]

I think this only misses approval from "trk". Is anyone able to check that?

[srimanob]

Hi @tomalin
Sorry, I missed this PR. Will trigger a test again after 2 weeks. If it is good, I will just sign.

[srimanob]

@cmsbuild please test

Re-trigger the test after 2 weeks.

[cmsbuild]

+1

Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-32d397/36369/summary.html
COMMIT: faa3fde
CMSSW: CMSSW_14_0_X_2023-12-07-1100/el8_amd64_gcc12
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week0/cms-sw/cmssw/42663/36369/install.sh to create a dev area with all the needed externals and cmssw changes.

Comparison Summary

Summary:

• You potentially added 278 lines to the logs
• Reco comparison results: 20 differences found in the comparisons
• DQMHistoTests: Total files compared: 50
• DQMHistoTests: Total histograms compared: 3430794
• DQMHistoTests: Total failures: 1553
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 3429219
• DQMHistoTests: Total skipped: 22
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 0.0 KiB( 49 files compared)
• Checked 214 log files, 167 edm output root files, 50 DQM output files
• TriggerResults: no differences found

[srimanob]

Hi @tomalin

I see that the failures in Phase-2 workflow comparison show up in L1T and SiOuterTracker. I assume L1T is expected. Do you check the SiOuterTracker? Thx.

[tomalin]

Hi @tomalin

I see that the failures in Phase-2 workflow comparison show up in L1T and SiOuterTracker. I assume L1T is expected. Do you check the SiOuterTracker? Thx.

Hi @srimanob , are you referring to https://cmssdt.cern.ch/SDT/jenkins-artifacts/baseLineComparisons/CMSSW_14_0_X_2023-12-07-1100+32d397/60200/25034.999_TTbar_14TeV+2026D98PU_PMXS1S2PR/SiOuterTracker_Tracks_HQ.html ? It claims that 100% of the 20 plots fail the comparison test, but by eye the plots seem in almost perfect agreement. This shows L1 tracks, so its normal that there are some changes.

We did not check if our PR changed offline track reco. It's hard to imagine how it could.

[srimanob]

+Upgrade

[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 now be reviewed by the release team before it's merged. @rappoccio, @antoniovilela, @sextonkennedy (and backports should be raised in the release meeting by the corresponding L2)

[rappoccio]

+1