Follow up to the Alpaka integration in CMSSW [14.0.x]

[fwyzard]

PR description:

HeterogeneousCore/AlpakaInterface

Rewrite the work division utilities:

• rewrite the uniform element kernel loops.
• rewrite the independent element kernel loops.
• remove the obsolete alpakatools utilities.

Rewrite the zeroAndInit kernel using the alpakatools utilities.
Rename elements_with_stride to uniform_elements in user code.

RecoLocalTracker/SiPixelClusterizer

Implement fixes that were lost during the CUDA-to-Alpaka migration.

Collect the updates from #39711, #42618, #42977 and #43052, that were missing from the Alpaka implementation:

• skip invalid or corrupted ROCs: make the Alpaka implementation of the pixel unpacker skip spurious ROCs, similar to the legacy and CUDA versions of the unpacker, and store the
  invalid ROC number error (errorType=36);
• fix the decoding of channels with timeout errors (errorType=27);
• synchronise the treatment of pixel errorType=26, 27, 30 with the legacy code;
• disable printf statements at compile time.

RecoTracker/PixelSeeding

Do not call the fishbone for events with pixel hits only in BPIX1 (reimplement #35638).

RecoTracker/MeasurementDet

Add a check on the pixel ROC range, to avoid overflows in the (unsigned) range differences.

DataFormats/TrackingRecHitSoA

Extend TrackingRecHits collections for testing and their unit tests.

CondFormats/SiPixelObjects

Clean up includes and dependencies, and avoid including framework headers in the device compilation.

Geometry/CommonTopologies

Update comments to pixel topologies.

Other changes

Fix include guards after files were moved to different packages.
Clean up include files.
Simplify nested namespace declarations.

PR validation:

Unit tests pass.

Validated running a recent HLT menu over 20k events without issues, both on GPU and on CPU, on data and MC: the alpaka implementation running on CPU or GPU gives the same results as the CUDA implementation, within the same non-reproducibility margins (1 event or up to 1% for high-rate triggers).

Backport status:

Backport of #43853 to 14.0.x (same branch) for 2024 data taking.

[fwyzard]

backport #43853

[fwyzard]

type bugfix

[cmsbuild]

A new Pull Request was created by @fwyzard (Andrea Bocci) for CMSSW_14_0_X.

It involves the following packages:

• CondFormats/SiPixelObjects (alca, db)
• Geometry/CommonTopologies (geometry)
• RecoLocalTracker/SiPixelClusterizer (reconstruction)
• RecoTracker/MeasurementDet (reconstruction)

@cmsbuild, @mandrenguyen, @jfernan2, @bsunanda, @saumyaphor4252, @civanch, @francescobrivio, @perrotta, @consuegs, @makortel, @Dr15Jones, @mdhildreth can you please review it and eventually sign? Thanks.
@mroguljic, @JanFSchulte, @dkotlins, @rovere, @fabiocos, @threus, @yuanchao, @VinInn, @PonIlya, @rsreds, @gpetruc, @missirol, @seemasharmafnal, @dgulhan, @tvami, @GiacomoSguazzoni, @felicepantaleo, @mmusich, @bsunanda, @ferencek, @tocheng, @VourMa, @ebrondol, @mtosi this is something you requested to watch as well.
@sextonkennedy, @rappoccio, @antoniovilela you are the release manager for this.

cms-bot commands are listed here

• Backported from Follow up to the Alpaka integration in CMSSW #43853

[cmsbuild]

cms-bot internal usage

[mandrenguyen]

+1

[perrotta]

+1

• Verbatim backport of Follow up to the Alpaka integration in CMSSW #43853, which was successully merged since CMSSW_14_1_X_2024-02-12-2300
• Current errors in the master IBs were already present even before the merge of Follow up to the Alpaka integration in CMSSW #43853 (and should be hopefuly fixed at some point...)

[civanch]

+1

[srimanob]

+Upgrade

[AdrianoDee]

+pdmv

[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. @rappoccio, @sextonkennedy, @antoniovilela (and backports should be raised in the release meeting by the corresponding L2)

[antoniovilela]

+1