[RFC] Avoid passing LD_PRELOAD to rocmIsEnabled subprocess

[makortel]

PR description:

Should work around the problem in #47450

Resolves cms-sw/framework-team#1269

PR validation:

Added printouts around p.communicate() and ran cmsRun with LD_PRELOAD=libPerfToolsAllocMonitorPreload.so:libPerfToolsMaxMemoryPreload.so. With LD_PRELOAD included in the environment, saw the MaxMemoryPreload printouts from rocmIsEnabled, and with LD_PRELOAD excluded (as in this PR), the MaxMemoryPreload printouts disappeared.

[cmsbuild]

cms-bot internal usage

[cmsbuild]

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-47452/43874

[makortel]

enable gpu

[makortel]

@cmsbuild, please test

[cmsbuild]

+1

Size: This PR adds an extra 20KB to repository
Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-f9782d/44667/summary.html
COMMIT: 0e52f83
CMSSW: CMSSW_15_1_X_2025-02-25-1200/el8_amd64_gcc12
Additional Tests: GPU
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week0/cms-sw/cmssw/47452/44667/install.sh to create a dev area with all the needed externals and cmssw changes.

Comparison Summary

Summary:

• You potentially removed 1 lines from the logs
• Reco comparison results: 8 differences found in the comparisons
• DQMHistoTests: Total files compared: 49
• DQMHistoTests: Total histograms compared: 3920300
• DQMHistoTests: Total failures: 23
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 3920257
• DQMHistoTests: Total skipped: 20
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 0.0 KiB( 48 files compared)
• Checked 214 log files, 184 edm output root files, 49 DQM output files
• TriggerResults: no differences found

GPU Comparison Summary

Summary:

• No significant changes to the logs found
• Reco comparison results: 0 differences found in the comparisons
• DQMHistoTests: Total files compared: 7
• DQMHistoTests: Total histograms compared: 53071
• DQMHistoTests: Total failures: 39
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 53032
• DQMHistoTests: Total skipped: 0
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 0.0 KiB( 6 files compared)
• Checked 24 log files, 30 edm output root files, 7 DQM output files
• TriggerResults: no differences found

[fwyzard]

I understand that this is supposed to work around #47450, but it could also break the case where LD_PRELOAD is actually needed to run ROCm code - for example to preload the right AMD driver libraries.

Anyway, if this is needed, I would make the same changes for cudaIsEnabled.