speed up TrackExtenderWithMTDT

[slava77]

using D49 ttbar with PU200 (23434.21) I get 4.8 times speedup in TrackExtenderWithMTDT, or about 10% reco time (less I/O; 7% with I/O included).

There are several parts

• std::bind case is the worst offender and I'm mostly to blame for this, as it came out in review of MTD TDR Reconstruction + MTD Track Association BDT  #28815
• after the std::bind fix, other updates are just moving repeated computations out to a common place and add up to 30% additional speedup

I checked that the reco comparisons of trackExtenderWithMTDT are the same on 50 events

Profiler report as of 376a099 (before the last commit
9ac0a76 which sped up by about 10% fractionally) is at https://slava77sk.web.cern.ch/slava77sk/reco/cgi-bin/igprof-navigator/CMSSW_11_2_X_2020-07-21-2300-sign1106.376a0992c5d.23434.21.step3.50.pp/94

[cmsbuild]

The code-checks are being triggered in jenkins.

[cmsbuild]

-code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-30902/17295

• This PR adds an extra 28KB to repository

Code check has found code style and quality issues which could be resolved by applying following patch(s)

• code-format:
  https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-30902/17295/code-format.patch
  e.g. curl https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-30902/17295/code-format.patch | patch -p1
  You can also run scram build code-format to apply code format directly

[cmsbuild]

The code-checks are being triggered in jenkins.

[slava77]

@mtosi @VinInn
the TrackExtenderWithMTDT could benefit in speed if the track trajectories for generalTracks were already available in the event.
IIUC, they should already be available, but not sure how deep one would need to go to get them to generalTracks.
Can this be added for "free", at a cost of copying this data, or would it by itself be equivalent to calling TrackTransformer::transform(reco::Track that's done here anyways (in the latter case, it's not going to give an overall speedup)?

[cmsbuild]

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-30902/17296

• This PR adds an extra 28KB to repository

[cmsbuild]

Pull request #30902 was updated. @perrotta, @jpata, @cmsbuild, @kpedro88, @slava77 can you please check and sign again.

[slava77]

@cmsbuild please test

[cmsbuild]

The tests are being triggered in jenkins.

• CMSSW_11_2_X_2020-07-25-1100/slc7_amd64_gcc820: https://cmssdt.cern.ch/jenkins/job/ib-run-pr-tests/8290/console Started: 2020/07/25 18:18

[cmsbuild]

+1
Tested at: 6f27f6d
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-3822eb/8290/summary.html
CMSSW: CMSSW_11_2_X_2020-07-25-1100
SCRAM_ARCH: slc7_amd64_gcc820

[cmsbuild]

Comparison job queued.

[cmsbuild]

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-3822eb/8290/summary.html

Comparison Summary:

• No significant changes to the logs found
• Reco comparison results: 16 differences found in the comparisons
• DQMHistoTests: Total files compared: 34
• DQMHistoTests: Total histograms compared: 2526188
• DQMHistoTests: Total failures: 7
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 2526134
• DQMHistoTests: Total skipped: 47
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 0.0 KiB( 33 files compared)
• Checked 144 log files, 17 edm output root files, 34 DQM output files

[slava77]

Reco comparison results: 16 differences found in the comparisons

these are in slimmedTaus, and should be unrelated to this PR, I expect these to be random, related to #30341 .
After #30919 makes it to the IB/baseline, these should go away

[slava77]

@pmeridian @parbol
it would be nice if you could confirm/acknowledge that updates in this PR are OK for you.
Thank you.

[kpedro88]

+upgrade

[pmeridian]

@pmeridian @parbol
it would be nice if you could confirm/acknowledge that updates in this PR are OK for you.
Thank you.

Thanks for this nice technical improvement. Looking at some of your changes to the std::bind, I've assumed some of this were already implemented, maybe a missing push from my side...

[slava77]

+1

for #30902 6f27f6d

• code changes were also checked confirmed by MTD / @pmeridian speed up TrackExtenderWithMTDT #30902 (comment)
• jenkins tests pass and comparisons with the baseline show no relevant differences (the changes in slimmedTaus are unrelated reproducibility issue, already fixed in the IB)

@silviodonato
it would be nice to get this in pre3.

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

[silviodonato]

+1
(10% reco time, wow!)