Message: Core dump on unknown interaction | Not Logged In (login) |
|
|
Hello everyone. I have a simple geometry (particle gun pointed at a planar shield with a detector on the far side...any particle entering the detector is recorded and its track is fStopAndKilled in UserSteppingAction). I am using the Underground Physics list submitted by Alex Howard with a few tweaks: I increase the interaction length to 0.1mm for all particles, and I create all particles using the general particle constructors I am running Geant4.5.2.p02 with a fully-patched Red Hat 9.0 system and gcc3.2.2. I set the particle energy to 1 keV and the shield to 1mm of Lead. When I run with gammas, electrons, or protons, the simulation runs just fine. However, when I run with mu- or neutrons, I get a core dump. I set the randomization seed to time(NULL), and the simulation core dumps after a somewhat random number of primary vertices being generated (usually a couple dozen, but up to a couple hundred), which makes me believe the user code I'm running is ok, and the simulation only runs into a problem on specific kinds of particle interactions. I have set checkpoints, and the core dump seems to be occuring between SteppingAction::UserSteppingAction and EventAction::EndOfEventAction. My question is how do I go about determining what might be causing this core dump? I am unclear as to where to set checkpoints in the simulation code to track down the culprit. What functions/methods are called after UserSteppingAction but before EndOfEventAction? Thank you for any suggestions. Kareem Kazkaz |
Inline Depth: |
|
|
Outline Depth: |
|
|
|
Add message: ![]() |
|
to: |
|