The current test time was slightly on the low side for run4. When running the
test locally on an empty node the execution time was below the maximum test
time. When running the test in parallel with other tests the test hits the
time limit. Fixing this by increasing the test time.

Eoin Clerkin authored 4 months ago and Pierre-Alain Loizeau committed 4 months ago

Introduce cfv as test

- Update geometry and parameters hashes to most recent ones (bring tof 21k + also previously stacked mvd geo creation macros)
- Force the specific filename for run 2391 in the 2022 reco macros and the mTOF clusterizer initialization script (and higher runs in most 2022 macros)
- Remove switch to custom alignment file in 2022 mcbm_event_reco_l1 macro for TOF 21k

users

Timeout for mcbm_2024_rec_3105 is temporarily increased to 95s, timeout for mcbm_2024_qa_3105 is increased to 135s

Extra time is needed to initialize the KF-setup instance in CbmL1::Init(), which
is done aside to the old geometry scheme intialization for CA. Keeping
of the two setup versions is temporary until the track-fitting utilities will
be fully moved to the KF framework.

CA: replacing the hard-coded disabling of detector subsystems with respect to the tracking mode (STS, mCBM, global) to the corresponding CA-configs; adding geometry layer index to ca::Station

CA: replaced the TrackingSetupBuilder class from KfCbm to L1; disable detectors, which are not used in a particular TrackingMode, directly in corresponding yaml-configs

Added macro run_unpack_tsa_algo.C with exemplary execution.

See !1872 (comment 45195)

- Store logs of each step in temp file + check if success pattern found
- Stop early if a step fails
- Fix some warnings for printf usage from ShellCheck online checker
- Fix some variables names (typos?) found by ShellCheck
- Avoid overwriting reco output with symbolic link
- In L1 macro, avoir error if output folder already existing (mkdir option)
- In L1 macro, set verbosity one level up to LOW to see severity