Processing very central PbPb simulated event with Allen

Several problems occur when trying to run hlt1_PbPb_default.py over very central PbPb MC events available on eos; /eos/lhcb/wg/IonPhysics/Simulations/2023_PbPb_MB/model*/Gausstest_MB.xdigi.

Sometimes, the error triggers the SciFi decoding:

python3: ../device/SciFi/preprocessing/src/SciFiRawBankDecoder.cu:35: void make_cluster(int, const SciFi::SciFiGeometry&, uint32_t, uint8_t, uint8_t, SciFi::Hits&): Assertion `pseudoSize < 9 && "Pseudosize of cluster is > 8. Out of range."' failed.

Sometimes it is the UT algorithms:

python3: ../device/UT/compassUT/src/CompassUT.cu:340: void compass_ut::save_track(int, const float*, const MiniState&, const BestParams&, const int*, UT::ConstHits&, const float*, float, unsigned int*, UT::TrackHits*, int, float, float): Assertion `n_tracks < UT::Constants::max_num_tracks' failed.

or

python3: ../device/event_model/UT/include/UTEventModel.cuh:149: float UT::Hits_t<T>::yBegin(unsigned int) const [with T = const char]: Assertion `m_offset + index < m_total_number_of_hits' failed.

Likely, particles produced in such events are bigger than some of the hard-coded limits, but it is not clear if it is just a vector size problem.

Note that hlt1_PbPb_default.py will be used for the heavy-ion data-taking in October 2023.