Unit test DRNTest: disabled setting LD_PRELOAD for xSAN IBs

[smuzaffar]

DRNTest unit test fails for ASAN and UBSAN IBs as it is not able to start the fallback server. This PR proposes to disabled setting LD_PRELOAD=libXsan.so for this test. Same change was done in #36874 to fix HeterogeneousCore/SonicTriton unit tests

[smuzaffar]

please test

[cmsbuild]

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-37763/29646

• This PR adds an extra 12KB to repository

[cmsbuild]

A new Pull Request was created by @smuzaffar (Malik Shahzad Muzaffar) for master.

It involves the following packages:

• RecoEcal/EgammaClusterProducers (reconstruction)

@jpata, @clacaputo, @slava77 can you please review it and eventually sign? Thanks.
@Sam-Harper, @lecriste, @rchatter, @jainshilpi, @valsdav, @rovere, @lgray, @sobhatta, @thomreis, @afiqaize, @simonepigazzini, @wrtabb, @argiro, @varuns23, @ram1123 this is something you requested to watch as well.
@perrotta, @dpiparo, @qliphy you are the release manager for this.

cms-bot commands are listed here

[clacaputo]

@cmsbuild please test

[cmsbuild]

+1

Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-5f7686/24382/summary.html
COMMIT: ae5ae11
CMSSW: CMSSW_12_4_X_2022-05-02-1100/slc7_amd64_gcc10
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week1/cms-sw/cmssw/37763/24382/install.sh to create a dev area with all the needed externals and cmssw changes.

Comparison Summary

Summary:

• No significant changes to the logs found
• Reco comparison results: 7 differences found in the comparisons
• DQMHistoTests: Total files compared: 49
• DQMHistoTests: Total histograms compared: 3703092
• DQMHistoTests: Total failures: 13
• DQMHistoTests: Total nulls: 1
• DQMHistoTests: Total successes: 3703056
• DQMHistoTests: Total skipped: 22
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: -0.004 KiB( 48 files compared)
• DQMHistoSizes: changed ( 312.0 ): -0.004 KiB MessageLogger/Warnings
• Checked 205 log files, 45 edm output root files, 49 DQM output files
• TriggerResults: no differences found

[clacaputo]

I saw a reco difference for _patTaus_slimmedTaus__reRECO_obj___tauIDs__21__second in workflow 136.793.
The same difference can be found also in the test results of other PRs, where CMSSW_12_4_X_2022-05-02-1100 has been used. I can't find what PR could have been introduced such a difference

[jpata]

This is probably coming from AVX512F+tensorflow: #32628

PR: 2022-05-02 12:52:58.512057: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  SSE4.1 SSE4.2 AVX AVX2 AVX512F FMA
vs.
baseline: 2022-05-02 14:03:30.196818: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  SSE4.1 SSE4.2 AVX AVX2 FMA

[clacaputo]

+reconstruction

• technical PR
• reco difference in _patTaus_slimmedTaus__reRECO_obj___tauIDs for workflow 136.793 not related to this PR (see Unit test DRNTest: disabled setting LD_PRELOAD for xSAN IBs #37763 (comment))

[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. @perrotta, @dpiparo, @qliphy (and backports should be raised in the release meeting by the corresponding L2)

[perrotta]

+1