Migration of L1 tracker primitives from SLHCDEV #13696
Conversation
at this stage, there is an error when building the dataformat library. In the L1Trigger code, some stub algos are commented out. In the L1Trigger tests, much is to be done (to be rewritten most probably).
Properly iterate over stacks and use the TrackerTopology
At this stage, it works with the window2013 algo. To make the tab2013 algo working, we need to pass the topology as well to the TTStubAlgo.
This implied to add a member to store the topology to the TTStubAlgorithm, and to adapt accordingly all derived classes. Comments and debug output were also removed. It runs and produces stubs when enabled in the sequence. We should see how to adapt the sequence (change removeTTI to keep the stubs).
A couple of methods changed namÃe (case)and the TrackerGeometry had to be changed: ModuleType has to be known outside the class.
Conflicts: Geometry/TrackerGeometryBuilder/interface/TrackerGeometry.h Geometry/TrackerGeometryBuilder/src/TrackerGeometry.cc solved by adopting the existing version (duplicate change)
|
A new Pull Request was created by @delaere (Christophe Delaere) for CMSSW_8_1_X. It involves the following packages: DataFormats/L1TrackTrigger @cmsbuild, @mulhearn, @davidlange6 can you please review it and eventually sign? Thanks. cms-bot commands are list here #13028 |
|
|
||
| private: | ||
| /// Data members | ||
| std::vector< T > theHits; |
There was a problem hiding this comment.
@delaere - is the maximum size of these vectors small and known?
[given the size of the PR, expect things little by little]
There was a problem hiding this comment.
Unfortunately not. In general the size is small but it can get very large in some cases. The only formal maximum is the number of strips in the module.
|
Pull request #13696 was updated. @cmsbuild, @rekovic, @mulhearn, @davidlange6 can you please check and sign again. |
|
please test |
|
The tests are being triggered in jenkins. |
Following a discussion with the L1TT group on hypernews, we converged on a minimal set to be kept in CMSSW_8_1_X. https://hypernews.cern.ch/HyperNews/CMS/get/l1-track-trigger/82/2.html TTClusterAlgorithm_2d2013.h --> KEEP: TTClusterAlgorithm_official.h TTClusterAlgorithm_2d.h --> REMOVE TTClusterAlgorithm_a.h --> REMOVE TTClusterAlgorithm_broadside.h --> REMOVE TTClusterAlgorithm_neighbor.h --> KEEP TTStubAlgorithm_a.h --> REMOVE TTStubAlgorithm_cbc3.h --> KEEP TTStubAlgorithm_globalgeometry.h --> REMOVE TTStubAlgorithm_pixelray.h --> REMOVE TTStubAlgorithm_tab2013.h --> KEEP: TTStubAlgorithm_official.h TTStubAlgorithm_window2012.h --> REMOVE TTStubAlgorithm_window2013.h --> REMOVE TTStubAlgorithm_window.h --> REMOVE
cmsbuild
added
tests-pending
comparison-pending
and removed
comparison-available
tests-approved
labels
|
Pull request #13696 was updated. @cmsbuild, @rekovic, @mulhearn, @davidlange6 can you please check and sign again. |
|
Dear all, With this last commit, the PR should be ok to proceed, now. All comments raised so far have been addressed or answered above. |
|
please test |
|
The tests are being triggered in jenkins. |
|
Dear all, Could somebody have a look and let me know if there are remaining issues? |
This PR introduces the tracker L1 primitives for phase2.
This is the first time that that code appears in a mainstream release, but it was developed and tested in SLHC and SLHCDEV releases for a long time.
Here, we introduce the core functionality: the TTCluster and TTStubs dataformats and related algorithms, which therefore only affects DataFormats and L1Trigger. The corresponding code in SimTracker, Validation and DQM is still being adapted to CMSSW_8_1_X.
For now, only the main algorithms are functional. Others are introduced here but commented out for now and will become effective in a future development. Still, it is important to have the classes there already for bookkeeping reasons.
This PR is the basis for the track trigger developments and studies needed for the tracker TDR.
@boudoul , @sudhaahuja FYI