Fix a bug in the TRD cluster finder; the index of registering TRD hits to events was wrong, leading to one missing TRD hit. Refs #2112.

Fix missing registration of TofTracks to CbmEvent in littrack, needed for track length calculation. Refs #2112.

When reordering the include statements a name clash between a define statement
in L1 code and a function in Root's TF1.h was found which did not allow to
compile the code.
Rename the define statement such that the name clash is resolved. refs #2051

Pierre-Alain Loizeau authored 3 years ago and Florian Uhlig committed 3 years ago

- Headers without source files in fles/mcbm2018 and core/data/raw
- Fles IPC headers
- CPP_ZMQ main header

The shell script is extracted from CMakeLists.txt and runs the macros in
the same sequence for the sis100_electron setup only. The script can be
executed in the the installation.

* Add CbmEvent to RICH track extrapolation procedure.
* Add CbmEvent to the RICH track projection procedure.
* Add CbmEvent to ring reconstruction procedure.
* Add CbmEvent to ring fitting procedure.
* Add CbmEvent to ring-track matching procedure.
* Update run macros.
* Fix compilation error, include Logger.h
* Fix problem when CbmEvent::GetNofData return -1.
* Change some LOG outputs to debug mode.

The previous fix did not work for macosx 11.3. Somehow on this system the old
FindGEANT3.cmake file was used.
Probably this is not a problem with the macosx version but with the usage of
a case insensitive file system. To avoid the problem remove the old and
obsolete file.
Also do some fixes to properly initialise some cmake variables.

Administrator authored 3 years ago and Florian Uhlig committed 3 years ago

With the previous fix the ROOT_INCLUDE_PATH wasn't correct on all platforms.
This is now fixed. refs #2088

Administrator authored 3 years ago and Florian Uhlig committed 3 years ago

The generated config scripts did not contain the proper (DY)LD_LIBRARY_PATH
information when using FairRoot v18.6. refs #2088

Pascal Raisig authored 3 years ago and Florian Uhlig committed 3 years ago

The hit counter was resetted for each event, while
the number of clusters correponds to the full timeslice.
Now the printed hitCounterTs counts all hits in a TS without a reset.
No explicit hit counting per event is currently implemented.

For the continuous tests rerun failed tests a second time. This should reduce
the number of failing test pipelines.
In most cases of failing tests the Geant3 simulations crash due to the well
known memory error. If these tests are restarted they normally work without
problem.
Since the feature to restart tests is only available with cmake 3.17 change
the scripts of the CI pipeline to use at least this version.