[14_0_X] Fixes for Alpaka Phase2 Pixel Reco

[AdrianoDee]

Plain backport of #44874.

[cmsbuild]

A new Pull Request was created by @AdrianoDee for CMSSW_14_0_X.

It involves the following packages:

• DataFormats/SiPixelDigiSoA (heterogeneous, reconstruction)
• Geometry/CommonTopologies (geometry)
• RecoLocalTracker/SiPixelClusterizer (reconstruction)

@jfernan2, @fwyzard, @Dr15Jones, @mdhildreth, @civanch, @makortel, @mandrenguyen, @cmsbuild, @bsunanda can you please review it and eventually sign? Thanks.
@missirol, @threus, @felicepantaleo, @mmusich, @GiacomoSguazzoni, @mtosi, @VourMa, @tsusa, @ferencek, @VinInn, @fabiocos, @JanFSchulte, @tvami, @bsunanda, @mroguljic, @dkotlins, @rovere, @gpetruc this is something you requested to watch as well.
@sextonkennedy, @antoniovilela, @rappoccio you are the release manager for this.

cms-bot commands are listed here

• Backported from Fixes for Alpaka Phase2 Pixel Reco #44874

[cmsbuild]

cms-bot internal usage

[fwyzard]

Looks good to me, as it's identical to #44874.
I'll check the impact on the HLT and sign.

[AdrianoDee]

please test

[AdrianoDee]

Looks good to me, as it's identical to #44874. I'll check the impact on the HLT and sign.

Note: I've preferred back porting all of it since it was simpler and the changes to Phase2 only were trivial and minimal fixes.

[AdrianoDee]

backport of #44874

[fwyzard]

Looks good from the HLT performance point of view.

before

Running 4 times over 10300 events with 8 jobs, each with 32 threads, 24 streams and 1 GPUs
   683.6 ±   0.2 ev/s (10000 events, 98.0% overlap)
   687.3 ±   0.2 ev/s (10000 events, 99.1% overlap)
   684.3 ±   0.2 ev/s (10000 events, 99.0% overlap)
   690.9 ±   0.2 ev/s (10000 events, 98.3% overlap)
 --------------------
   686.5 ±   3.3 ev/s

after

Running 4 times over 10300 events with 8 jobs, each with 32 threads, 24 streams and 1 GPUs
   683.8 ±   0.2 ev/s (10000 events, 98.9% overlap)
   688.4 ±   0.2 ev/s (10000 events, 98.6% overlap)
   689.3 ±   0.2 ev/s (10000 events, 98.7% overlap)
   686.6 ±   0.2 ev/s (10000 events, 98.7% overlap)
 --------------------
   687.0 ±   2.5 ev/s

[fwyzard]

+heterogeneous

[mmusich]

type bug-fix

[cmsbuild]

+1

Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-eefebf/39274/summary.html
COMMIT: 4bd3572
CMSSW: CMSSW_14_0_X_2024-05-07-1100/el8_amd64_gcc12
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week0/cms-sw/cmssw/44915/39274/install.sh to create a dev area with all the needed externals and cmssw changes.

Comparison Summary

Summary:

• You potentially added 95 lines to the logs
• Reco comparison results: 43 differences found in the comparisons
• DQMHistoTests: Total files compared: 48
• DQMHistoTests: Total histograms compared: 3326485
• DQMHistoTests: Total failures: 6
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 3326459
• 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

[civanch]

+1

[mandrenguyen]

+1

[cmsbuild]

This pull request is fully signed and it will be integrated in one of the next CMSSW_14_0_X IBs (tests are also fine) and once validation in the development release cycle CMSSW_14_1_X is complete. This pull request will now be reviewed by the release team before it's merged. @antoniovilela, @rappoccio, @sextonkennedy (and backports should be raised in the release meeting by the corresponding L2)

[rappoccio]

+1