cmssw: Replay failures after #41198 introduced

There are currently T0 replay failures that have begun in 13_0_2:

An exception of category 'FatalRootError' occurred while
   [0] Calling InputSource::getNextItemType
   Additional Info:
      [a] Fatal Root Error: @SUB=TBufferFile::CheckByteCount
object of class vector<l1t::MuonShower> read too many bytes: 119 instead of 118

This is likely caused by the introduction of https://github.com/cms-sw/cmssw/pull/41198

@dinyar please take a look, this is very urgent.

About this issue

Original URL
State: closed
Created a year ago
Comments: 56 (56 by maintainers)

Commits related to this issue

[Do not merge] For debugging cms-sw/cmssw#41246 — committed to cms-sw/cmsdist by smuzaffar a year ago

Most upvoted comments

@francescobrivio , if I understand correctly the https://github.com/cms-sw/cmssw/issues/41246#issuecomment-1494887051 comment then the problem was how the file was writing. I think it was written by root version without the fix. Including newer root means new files will be written using fixed version of root.

smuzaffar on Apr 4, 2023

For the record, the culprit PR in 13_0_X is https://github.com/cms-sw/cmssw/pull/41198

makortel on Mar 31, 2023

Looking at the blame log for the class in question https://github.com/cms-sw/cmssw/blame/master/DataFormats/L1Trigger/interface/MuonShower.h

it clearly shows that an additional member data of type bool was added to the class. That would account for why it tried to read 1 byte too many.

Dr15Jones on Mar 31, 2023