Turn plume lumi counters on by default

Currently the with_plume argument of lumi_reconstruction is the only one to default to False. This causes the plume lumi algorithm to be skipped (and all plume counters to come out at max value) unless with_plume is explicitly set to True in the configuration (which it hasn't been).

This MR changes the default value to True so that plume counters will be filled unless it's explicitly requested to exclude them. This is the same behaviour as we have for all other sub-detectors.

Depends on (and must be adapted for) !1326 (merged)

@mzdybal please can you start the CI tests for this

added RTA label

mentioned in issue Moore#623 (closed)

mentioned in issue Moore#627 (closed)

mentioned in issue Moore#634 (closed)

assigned to @dcraik

mentioned in merge request !1306 (closed)

added 27 commits

• 16efa627...a2415652 - 23 commits from branch master
• 8fa9ecf0 - Turn plume lumi counters on by default
• 9ded2768 - add empty return to plume decode for unknow bank version
• ae5ebb32 - increase default max plume avg adc to 7ffff
• a073d579 - Merge remote-tracking branch 'origin/sxian_lumi_plume_bugfix' into dcraik_lumi_plume_bugfix

Compare with previous version

added 2 commits

• 35b19425 - check if plume bank presents
• 92f4deaa - Merge remote-tracking branch 'origin/sxian_lumi_plume_bugfix' into dcraik_lumi_plume_bugfix

Compare with previous version

added 1 commit

• 6543f7db - Update PlumeDecode.cu to correct bank version check

Compare with previous version

added 1 commit

• fddfd002 - fix format

Compare with previous version

added 1 commit

• 6e7261c7 - Fix plume bank versions passed to the decoding

Compare with previous version

added 4 commits

• 6e7261c7...4d522f08 - 2 commits from branch master
• 921ccde8 - Merge remote-tracking branch 'origin/master' into dcraik_lumi_plume_bugfix
• 15fc0a29 - Fixed formatting

Compare with previous version

@dcraik @sxian This will need another rebase on master in order to fix Allen CI, for details see https://mattermost.web.cern.ch/lhcb/pl/kyw9abxhatd6bfjqq18bwdedhh

mentioned in issue Moore#652 (closed)

added 50 commits

• 15fc0a29...5b90acf1 - 48 commits from branch master
• 7c04485b - Remove plume bank version 2 support for now
• 6b46a9c2 - Merge remote-tracking branch 'origin/master' into dcraik_lumi_plume_bugfix

Compare with previous version

added 1 commit

• 9dd92c3b - Fixed formatting

Compare with previous version

resolved all threads

@jagoodin please can you start the CI tests

/ci-test

Started integration test build. Once done, check the results or the comparison to a reference build.

Throughput Test Moore_hlt2_fastest_reco: 482.0 Events/s -- change of 0.06% vs. reference

Throughput Test Moore_hlt2_pp_thor: 231.6 Events/s -- change of -0.29% vs. reference

Throughput Test Moore_spruce_all_lines: 337.1 Events/s -- change of -0.12% vs. reference . Total bandwidth 3162.1 MB/s -- change of -0.00% vs. reference

Throughput Test Moore_hlt1_pp_default: 26058.9 Events/s -- change of 0.07% vs. reference

I guess the majority of the failed tests come from the same DaVinci bug that affected some other CI tests. From a quick glance, I see one Allen test that needs a reference update.

added ci-test-triggered label

• [2023-09-15 12:08] Validation started with lhcb-master-mr#9223

mentioned in issue Moore#659 (closed)

mentioned in merge request !1327 (merged)

@fferrari @dcraik you should be aware that Allen assumes that all sources for a given subdetector produce the same bank version.

With @cagapopo we checked that the bank_version you get in the decoding is the version in the last MFP (source) in the MEP. So if the EB orders MFPs by source id (TBC), bank_version will be the version of source 0x5003 and not of 0x5001. This will mean that the v1 decoding will run, wrongly interpreting the v2 data as v1.

There might be other consequences of different sources giving different versions. One has to run on some data like this to see.

added WIP dependency label

This depends on !1326 (merged)

changed the description

mentioned in issue Moore#669 (closed)

mentioned in issue Moore#671 (closed)

added impactlow prioritylow labels

added RTA WP5 label

assigned to @rmatev

mentioned in issue Moore#688 (closed)

mentioned in issue Moore#700 (closed)

mentioned in issue Moore#708 (closed)

Superseded by !1326 (merged)

closed