jlingema/cmssw

CMS Offline Software

cmssw's Issues

Move LUT to CondFormat

LUTs shuld be configurable from DB and also versionable. This needs some re-writing of the corresponding classes. Probably just need to touch MicroGMTLut.h...

Calo LUT object:
https://github.com/cms-sw/cmssw/blob/CMSSW_7_6_X/CondFormats/L1TObjects/interface/LUT.h

Move to BXVector

All L1 products should be contained in a BXVector. This needs also some re-writing of the TF producers.

Classes to adapt:

• GMTInternalMuon
• L1TRegionalMuonCandidate

Remove cout

The uGMT code should already be cout-free, but make sure this is the case.

Move global phi calculation

Currently global phi is calculated in GMTInternalMuon. This should be done somewhere else, as the DataFormats are not hooked up to the DB.

Firmware Versioning

Need a mechanism to take firmware version into account. Either implement factory or make simple ifs where anticipated?

see Calo Producer:
https://github.com/cms-sw/cmssw/blob/CMSSW_7_6_X/L1Trigger/L1TCalorimeter/plugins/L1TStage1Layer2Producer.cc#L388

Move sort logic out of the EDProducer

To ease versioning of the sorting logic, I assume it would be best to make a new class (e.g. MicroGMTSortingUnit) that takes care of what currently is done in the EDProducer.

Fix last naming convention issues

Some classes / configs are using the old uGMT naming convention. This should all be moved to MicroGMT. (Go through the code and make sure also other variations are gone (MicroGlobalMuon... etc.)

Processor Indexing

Make sure cancel-out and phi-assignment works with zero-indexed processor IDs.

Unpacker needed

We'll need an unpacker. We should be able to benefit from work in Calo (they have some code that decodes the AMC13 and MP7 headers). Also the uGMTScripts contain code that decodes the muon data from the 32 bit words.

Add sensibility checks for input muons

Currently the muons are accepted as they come in. This has led to some unfortunate bugs and debugging was a hassle. We should add some checks to make sure properties are in the correct interval.

Move to DB configs

To make configuring from DB possible need to implement an ESProducer and CondFormat object.

Calo ESProducer:
https://github.com/cms-sw/cmssw/blob/CMSSW_7_6_X/L1Trigger/L1TCalorimeter/plugins/L1TCaloParamsESProducer.cc
Calo CondFormat:
https://github.com/cms-sw/cmssw/blob/CMSSW_7_6_X/CondFormats/L1TObjects/interface/CaloParams.h

Rework track addresses

The track addresses should be encoded more explicitly in the RegionalMuCand to make it easier to access the "sub-addresses".

Migrate changes in dev-branch to 7_6_0_pre4.

The changes done over the last week should be moved to the new muon branch.

Trigger Primitive Masking

Currently the overlap between track-finders is much greater than we expected. E.g. the OMTF also sends candidates at eta > 1.8. This needs to be addressed (candidates should be filtered on chambers that are connected in the hardware).

Decide where masking should be done. Needs discussions with Artur.