This reverts merge request !27081

Setting up the late muon trigger to run on out of time RoIs (produced by TrigmuRoI). Also configuring the data decoding and ID tracking to run in these RoIs (this required some modification on the decoding configuration, and a new instnce of the fast track finder).
The chain now runs in a full scan RoI mode, then creates the out of time RoIs then runs the data decoding, ID tarcking, and late muon reconstruction in one step (ATR-17553).

Per Alexander Ekman authored 5 years ago and Adam Edward Barton committed 5 years ago

WIP: Prepared for implementation of beamspotMTChain. Reference file is different from master, this is probablu because I didn't pull this morning

Per Alexander Ekman authored 5 years ago and Adam Edward Barton committed 5 years ago

Changed Luminosity block dependence of costmonitor,LArCells, RPCSecondaryReadout to False. Cleaned up unused imports for costmonitor. Turned all the slices back on.

Adding a first pass at the late muon trigger. Currently seeded from and running on in time muon RoIs.
- Modified the inside out sequence to reconstruct slow muons for the late muon chain, and the normal in time inside out reco for standard muon chains.
- For late muons, added the segment finding and ID tracking reco to the inside out reco sequence, since we don't run this anywhere else in the chain
- Currently running L2 SA and L2 combined muons to run the decoding and fast ID tracking respectively. Both of these steps are in passthrough (after some reorganisation of the hypo configuration so that the passthrough option isn't being overwritten). This needs to be revisited in the future, but is fine for now.

Implemented a very simple serial merging in the muon menu generation (can be removed when serial merging is officially supported) in order to run the RoI based muon finding before the full scan muon finding (part of the migration documented in ATR-17553).

Some modifcations to get a dimuon chain with invariant mass cuts running:

- Setup the sequence for the invariant mass so that it is independent of the muon reco sequences
- Set the chains step with the invariant mass cut to have multiplicity of one to avoid it being treated as a combined hypo
- Some minor fixes for the hypo alg to allow for one-muon-per-RoI, and to not overwrite the feature link
- Added a test chain to the menu and updated the references.

Updating the muon MS track hypo to be configurable to require at least 3 precision layers in the endcap, and added the msonly_3leyersEC chain (ATR-20126, ATR-20049).

This allows the full scan muon finding to run unseeded (the intended use case) instead of from the L1 muon RoIs (in which case we were duplicating parts of the muon reconstruction depending on the multiplicity of L1 muon RoIs).

Carlo Varni authored 5 years ago and Vakhtang Tsulaia committed 5 years ago

New flags added into SimConfigFlags for running the G4AtlasAlgCfg
function with the new ConfigFlags.

Adding one roi based asymmetric dimuon chain to the menu (mu6_mu4) using the chain def merging (ATR-19636).