Iss1172 - Memory bugs and typos in DQM code

[EinarElen]

I am updating ldmx-sw, here are the details.

What are the issues that this addresses?

Hint: Use the phrase 'This resolves #< issue number >' so that they are linked automatically.
This resolves #1172 and one of the entries in #1166 (missing & in PN DQM).

I expect this to fail the validation step for the Hcal but not for the PN DQM, we have changed the histograms in the HcalDQM.

The following warning/static-analysis/cleanup was also done

• Add missing virtual destructors
• override final -> override
• Remove tautological check (pdgCode < 10000000000) )
• Fallthrough and default cases in switches
• Removing unused variables

[tomeichlersmith]

This looks good but I am curious about where histogram content is being changed - I don't see anything affecting fill or binning calls outside of white-space formatting.

[EinarElen]

This bugfix (it's subtle, thanks implicit conversion warnings!)

9305a03

[EinarElen]

The plots that are failing are the expected ones. The Hcal one is not failing because that is the only one where we arent running the particular analyzer :)

[tomeichlersmith]

Since only one plot fails for each of the three failing validation tests, I am posting it here.¹ @EinarElen do these make sense? It seems like a physical change (i.e. I don't see any screaming bugs), but I'm unsure how to interpret the change.

Signal

Inclusive

Ecal PN

Edit: Cropped PNGs so there's less ugly white-space.

Footnotes

1. I got this plot by downloading the artifacts for the failing PR Validation jobs. Unzipping and unpacking and then copying out the plot in the "fail" subdirectory of the unpacked directory. I then used pdftoppm to convert the PDF plots to PNG so GitHub will render them. ↩

[EinarElen]

It is a physical change in that the previous version was just straight up invalid. What the analyzer is trying to do is to check if any hit has been above threshold in the different hcal sections and if so, what was the earliest layer that had such a hit. By default we have each section set to a negative "failed to veto" value. The histogram is supposed to contain different categories like "the hit was vetoed by one of the side hcal modules" or "the hit was only vetoed by the back hcal"

  bool vetoedByBack{firstLayersHit[ldmx::HcalID::HcalSection::BACK] !=
                    failedVeto};
  bool vetoedByTop{firstLayersHit[ldmx::HcalID::HcalSection::TOP]};
  bool vetoedByBottom{firstLayersHit[ldmx::HcalID::HcalSection::BOTTOM]};
  bool vetoedByRight{firstLayersHit[ldmx::HcalID::HcalSection::RIGHT]};
  bool vetoedByLeft{firstLayersHit[ldmx::HcalID::HcalSection::LEFT]};
  bool vetoedBySide{vetoedByTop || vetoedByBottom || vetoedByRight ||
                    vetoedByLeft};

Problem is right here, only the first (vetoedByBack) condition is correct, all the others have missed the != failedVeto part.

[tomeichlersmith]

Thanks for the extra explanation! This looks good and I can make a tag release after merging.