Repository navigation
mcsmear crash because of ECAL geometry #415
Description
mcsmear for the latest release halld_sim 5.10.0 terminates with the following error message:
terminate called after throwing an instance of 'JException'
what(): GetSingle<DECALGeometry>: Factory has wrong number of items: 1 expected, 0 found.
It can be seen for example in the latest b1pi test:
https://halldweb.jlab.org/b1pi/2026-09-17/b1pi_ifarm2401.log
It can be reproduced with this command:
mcsmear /volatile/halld/gluex/b1pi/2026-09-17/Linux_Alma9-x86_64-gcc11.5.0/30480/hdgeant.hddm
Thanks to @imoewi01 for the heads up.
Metadata
Metadata
Assignees
Labels
No labels
Type
Fields
Priority
None yet
Activity
aaust commentedon Sep 17, 2026
It actually works for the runs above 130000 where the ECAL was installed, but crashes for earlier runs. I assume, it is cause by PR #414
aaust commentedon Sep 17, 2026
The pull request test already saw this error, but it did not correctly report the result:
https://pages.jlab.org/halld/-/pull-request-test/-/jobs/94790/artifacts/results/log/hdgeant4.txt
aaust commentedon Sep 25, 2026
Fixed by #416
b1pi test passed: https://halldweb.jlab.org/b1pi/2026-09-24/b1pi_ifarm2401.log