You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This probably needs a bit more work, because currently the PodioDataSvc is pretty ingrained in some places, and for some things I am not sure if they will work with the IOSvc.
The LcioEventAlgo currently puts an LCEventWrapper into the eventSvc
Does this still work with the IOSvc? @jmcarcell IIUC, this should still work, but the LCEvent will not be available for our functional algorithms, right?
For the converters it should be possible to put in place checks that check the IOSvc, resp. MetadataSvc and the PodioDataSvc.
As discussed during the EDM4hep meeting, we will not put too much effort into supporting IOSvc with the k4MarlinWrapper, because that is intended mainly for multithreaded usage, and only very few Marlin processors are actually thread safe. We will keep this issue open, so that it is visible for people who try to use the IOSvc with the wrapper.
If necessary we will come back to this in the future to add support. For now we will mainly try to avoid the segmentation fault and rather give users a proper error for that.
Trying to raise the priority of this issue: it is not really a matter of running in multi-threaded way, it is rather that we try to write all the new algorithms with Functionals and this issue currently prevents one to access the ILCSoft world.
The converters still partially refer to the
PodioDataSvc
explicitly and will fail with a segmentation violation when used with theIOSvc
.The text was updated successfully, but these errors were encountered: