ROOT 6.36 results 510-660 MB higher memory usage in some workflows

[makortel]

Spinned off from cms-sw/cmsdist#10199

According to MaxMemoryPreload report of cms-sw/cmsdist#10199 (comment)

• Many jobs show < 10 MB variations
• Some jobs show variations up to 70 MB, seemingly compatible with MaxMemoryPreload shows variations of tens of MB in some cases #46966
• Following jobs show 510 - 660 MB increase
  • 2025.0000001_RunZeroBias2025B_10k steps 2 and 3
  • 2025.0010001_RunJetMET02025C_10k steps 2 and 3
  • 16834.0_TTbar_14TeV+2025 steps 2 and 3
  • 17034.0_TTbar_14TeV+2025PU steps 2 and 3

[makortel]

assign core

[makortel]

type root

[cmsbuild]

New categories assigned: core

@Dr15Jones,@makortel,@smuzaffar you have been requested to review this Pull request/Issue and eventually sign? Thanks

[makortel]

• 2025.0000001_RunZeroBias2025B_10k steps 2 and 3

Step 2 is L1REPACK:Full,HLT:@relval2025
Step 3 is RAW2DIGI,L1Reco,RECO,PAT,NANO,DQM:@standardDQM+@miniAODDQM+@nanoAODDQM

• 2025.0010001_RunJetMET02025C_10k steps 2 and 3

Step 2 is L1REPACK:Full,HLT:@relval2025
Step 3 is RECO,MINIAOD,NANOAOD,DQMIO

• 16834.0_TTbar_14TeV+2025 steps 2 and 3

Step 2 is DIGI:pdigi_valid,L1,DIGI2RAW,HLT:@relval2025
Step 3 is RAW2DIGI,L1Reco,RECO,RECOSIM,PAT,NANO,VALIDATION:@standardValidation+@miniAODValidation,DQM:@standardDQM+@ExtraHLT+@miniAODDQM+@nanoAODDQM

• 17034.0_TTbar_14TeV+2025PU steps 2 and 3

Step 2 is DIGI:pdigi_valid,L1,DIGI2RAW,HLT:@relval2025
Step 3 is RAW2DIGI,L1Reco,RECO,RECOSIM,PAT,NANO,VALIDATION:@standardValidation+@miniAODValidation,DQM:@standardDQM+@ExtraHLT+@miniAODDQM+@nanoAODDQM

[cmsbuild]

cms-bot internal usage

[cmsbuild]

A new Issue was created by @makortel.

@Dr15Jones, @ftenchini, @makortel, @mandrenguyen, @sextonkennedy, @smuzaffar can you please review it and eventually sign/assign? Thanks.

cms-bot commands are listed here

[makortel]

Just for more permanent record

On 2025.0000001_RunZeroBias2025B_10k

---

On 16834.0_TTbar_14TeV+2025

[makortel]

@pcanal @vepadulano

[makortel]

So about ~100 MB of the increase shows up as leaked memory (i.e. memory allocated during the program, but not released at the time when MaxMemoryAllocPreload gets destroyed at the end of the process)