Fix cudaMemcpyAsync for localCoordToHostAsync

[AdrianoDee]

PR description:

In TrackingRecHitSoADevice::localCoordToHostAsync used in SiPixelRecHitFromCUDA to fill the legacy hits, cudaMemcpyAsync is not taking into account the SoA layout buffer padding. So it's copying some wrong portions of memory. This was noted in #40604 and this solves it.

This fix is quick and dirty, given also this CUDA to legacy copy will be dropped soon.

PR validation:

Run 11634.59x.

[AdrianoDee]

type bugfix

[cmsbuild]

-code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-40869/34338

• This PR adds an extra 12KB to repository

Code check has found code style and quality issues which could be resolved by applying following patch(s)

• code-format:
  https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-40869/34338/code-format.patch
  e.g. curl -k https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-40869/34338/code-format.patch | patch -p1
  You can also run scram build code-format to apply code format directly

[AdrianoDee]

test parameters:

• workflows = 11634.596, 11634.592
• enable = gpu
• relvals_opt_gpu = --what cleanedupgrade,standard,highstats,pileup,generator,extendedgen,production,ged,machine,premix,nano

[cmsbuild]

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-40869/34339

• This PR adds an extra 12KB to repository

[cmsbuild]

A new Pull Request was created by @AdrianoDee (Adriano Di Florio) for master.

It involves the following packages:

• CUDADataFormats/TrackingRecHit (heterogeneous, reconstruction)

@cmsbuild, @makortel, @mandrenguyen, @clacaputo, @fwyzard can you please review it and eventually sign? Thanks.
@missirol, @rovere this is something you requested to watch as well.
@perrotta, @dpiparo, @rappoccio you are the release manager for this.

cms-bot commands are listed here

[AdrianoDee]

please test

[cmsbuild]

-1

Failed Tests: RelVals
Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-33341c/30872/summary.html
COMMIT: 56023e4
CMSSW: CMSSW_13_1_X_2023-02-23-2300/el8_amd64_gcc11
Additional Tests: GPU
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week1/cms-sw/cmssw/40869/30872/install.sh to create a dev area with all the needed externals and cmssw changes.

RelVals

ValueError: Undefined workflows: 11634.592, 11634.596

GPU Comparison Summary

Summary:

• You potentially removed 20 lines from the logs
• Reco comparison results: 0 differences found in the comparisons
• DQMHistoTests: Total files compared: 4
• DQMHistoTests: Total histograms compared: 19862
• DQMHistoTests: Total failures: 128
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 19734
• DQMHistoTests: Total skipped: 0
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 0.0 KiB( 3 files compared)
• Checked 12 log files, 9 edm output root files, 4 DQM output files
• TriggerResults: found differences in 3 / 3 workflows

[perrotta]

urgent

• the backport is wanted in 13_0_0
• @cms-sw/heterogeneous-l2 @cms-sw/reconstruction-l2 please have a look

[fwyzard]

+heterogeneous

[clacaputo]

please test

[perrotta]

@clacaputo please don't restart the tests uselessly, if you did so because of the relval-input errors, see #40889. If that was the reason, just review the PR as it is.

[clacaputo]

please abort

[clacaputo]

+reconstruction

[cmsbuild]

This pull request is fully signed and it will be integrated in one of the next master IBs after it passes the integration tests. This pull request will now be reviewed by the release team before it's merged. @perrotta, @dpiparo, @rappoccio (and backports should be raised in the release meeting by the corresponding L2)

[missirol]

please abort

Based on #40869 (comment).

(Signing a PR with the label 'tests-pending' starts the tests automatically.)

[perrotta]

merge