Remove old RICH framework + Future scalar algorithms

• Removes all the code related to the old RICH framework.
• Removes scalar implementations of various future algorithms, now replaced with SIMD versions.

mentioned in merge request Brunel!330 (merged)

@cattanem I see you have un WIP'ed !814 (merged), thanks. I am thinking though are you OK with testing this MR at the same time ? I do not expect any change in the nightlies tests this time, other than the removal of the 'oldrich' tests in Brunel!330 (merged).

@jgarrati @pnaik @asolomin Please take not of this MR, as once merged it will break (quite a lot) Panoptes built against master...

@jgarrati I think we need to now separate out the branches of Panoptes for the upgrade (master) from the branch we need for next year. To that end can you make a protected branch (lets call it 2018-patches to match those in Rec etc.), forked from master now, which we then use for the developments needed for next year. The master branch can then evolve at its own rate, once we start to know what we need to monitor for the upgrade. This will likely take a while.

@pnaik @asolomin This means you will need to think a little what branch you want merge requests to be against. Developments for next year will need to go to the new 2018-patches branch, and if you want them to persist into the upgrade you will need to push them to master as well.

unmarked as a Work In Progress

changed title from WIP : Remove old RICH framework to Remove old RICH framework

changed the description

marked as a Work In Progress

changed the description

mentioned in merge request Boole!105 (merged)

unmarked as a Work In Progress

marked as a Work In Progress

changed title from Remove old RICH framework to WIP : Remove old RICH framework

Sorry, that last WIP removal was unintentional... Something funny going on with gitlab or my web browser here...

added 2 commits

• 957c165e - Update photon reco test
• f45da7c9 - Merge branch 'Rich-RemoveOldFramework' of ssh://gitlab.cern.ch:7999/jonrob/Rec…

Compare with previous version

added 4 commits

• f45da7c9...047ebfec - 3 commits from branch lhcb:master
• d9cf05cf - Merge branch 'master' of ssh://gitlab.cern.ch:7999/lhcb/Rec into Rich-RemoveOldFramework

Compare with previous version

mentioned in merge request !821 (closed)

unmarked as a Work In Progress

changed title from WIP : Remove old RICH framework to Remove old RICH framework

changed the description

added 5 commits

• d9cf05cf...395c94e6 - 4 commits from branch lhcb:master
• b5bf82d6 - resolve conflict

Compare with previous version

added 1 commit

• 04f6cf4b - fix for clang

Compare with previous version

mentioned in issue #9 (closed)

• [2017-12-14 00:10] Validation started with lhcb-clang-test#800
• [2017-12-14 00:11] Validation started with lhcb-head#1691
• [2017-12-14 00:11] Validation started with lhcb-gaudi-head#1690
• [2017-12-14 00:37] Validation started with lhcb-gaudi-head-py3#3
• [2017-12-14 01:02] Validation started with lhcb-gaudi-head#1690
• [2017-12-15 00:13] Validation started with lhcb-gaudi-head#1691
• [2017-12-16 00:11] Validation started with lhcb-clang-test#802
• [2017-12-16 00:12] Validation started with lhcb-gaudi-head#1692
• [2017-12-17 00:09] Validation started with lhcb-clang-test#803
• [2017-12-17 00:13] Validation started with lhcb-gaudi-head#1693

@jonrob I didn't think this would be so intrusive. Moving this and related MRs to WIP plus lhcb-gaudi-head label until the tests are passing...

marked as a Work In Progress

Part of the problem is I think I accidentally committed some fixes needed here to !821 (closed). As these two MRs are not really that different in nature, I think the easiest fix is I merge them into one.

added 5 commits

• 0980be0b - Remove scalar implementations of algorithms.
• f8f7d6c0 - Merge branch 'Rich-RemoveOldFramework' of ssh://gitlab.cern.ch:7999/jonrob/Rec…
• 91d74782 - Merge branch 'Rich-RemoveOldFramework' of ssh://gitlab.cern.ch:7999/jonrob/Rec…
• 4d008bb0 - Clean up doxygen comments
• 3998d305 - Fix some bugs relating to the removal of the old RICH sequence

Compare with previous version

changed title from WIP Remove old RICH framework to WIP Remove old RICH framework + Future scalar algoruthms

changed the description

changed title from WIP Remove old RICH framework + Future scalar algoruthms to WIP Remove old RICH framework + Future scalar algorithms

mentioned in merge request Analysis!225 (merged)

added 1 commit

• 1b1f2bf5 - clean up

Compare with previous version

@jonrob There is still an issue in the Brunel Expert monitoring sequence, see brunel-mc11a-xdst test:

ExpertProtoMoni      INFO Member list: ChargedProtoParticleTupleAlg/ChargedProtoTuple, ANNGlobalPID::ChargedProtoANNPIDTrainingTuple/ChargedProtoPIDANNTuple, with context 'Offline'
ToolSvc             ERROR Cannot create tool Rich::Rec::MC::MCTruthTool (No factory found)
ChargedProtoTuple   FATAL ChargedProtoParticleTupleAlg:: Exception throw: tool():: Could not retrieve Tool 'Rich::Rec::MC::MCTruthTool'/'MCTruth' StatusCode=FAILURE
ChargedProtoTuple   FATAL  Exception with tag= is caught 
ChargedProtoTuple   ERROR 	ChargedProtoTuple:: tool():: Could not retrieve Tool 'Rich::Rec::MC::MCTruthTool'/'MCTruth'	 StatusCode=FAILURE
ExpertProtoMoni     ERROR GaudiSequencer:: Can not initialize ChargedProtoTuple StatusCode=FAILURE

added lhcb-clang-test label

Damn, I forgot about checking expert mode. Will take a look.

mentioned in merge request LHCb!1027 (merged)

added 1 commit

• 57b2cf62 - Update ChargedProtoANNPID and GlobalReco to adapt to removal of Rich Rec MCTruth tool

Compare with previous version

Fixed. The issue was whilst I had removed the implementation of an MC truth tool in Rec I had neglected to remove the interface for it in LHCb. Consequently a few (non RICH) components in which I had used the RICH tool (for convenience...) where still being built against it, but of course then failed to find the component at runtime.

LHCb!1027 (merged) and Lbcom!203 (merged) remove the interface, and port to the remaining lower level MC truth tool the one method they used, which was for Track to MCParticle associations. Eventually the components (ChargedProtoANNPID and GlobalReco) will anyway need to be modernised for the upgrade, but this keeps them going in the meantime.

mentioned in merge request Panoramix!13 (merged)

mentioned in commit LHCb@297c8094

unmarked as a Work In Progress

removed lhcb-clang-test ~1280 labels

resolved all discussions

merged

mentioned in commit ad233bf3

mentioned in merge request Brunel!335 (merged)

@jonrob This MR is breaking Panoptes (which I think you are aware of) but also Panoramix. Could you have a look please?

I already have MRs fixing both submitted.