| Summary: | Setting a neutron absorber with a Z > 92 and using EmStandardPhysics_option3 produces a segfault. | ||
|---|---|---|---|
| Product: | Examples/Extended | Reporter: | txenglish |
| Component: | hadronic/NeutronSource | Assignee: | Vladimir.Ivantchenko |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | CC: | Alberto.Ribon, michel.maire |
| Priority: | P4 | ||
| Version: | 11.2 | ||
| Hardware: | PC | ||
| OS: | Linux | ||
| Attachments: | gdb backtrace of segfault | ||
Hello, Thank you very much for this report! It is a real bug and the fix is under Geant4 testing. It will be available with the 1st patch to Geant4 11.2. This problem was not in the same form in previous G4 versions. The fix is different from your proposal, because your proposal would likely work for AmBe but not for all possible cases. VI Hello, is the problem fixed for you with the new version Geant4 11.2.1? VI Hello, the bug is fixed and is available in Geant4 11.2.1 and will be available in 11.3beta. VI |
Created attachment 847 [details] gdb backtrace of segfault When running the NeutronSource example, after making a modification of the neutron absorber and changing the EM Physics List, the simulation runs into a segfault. The resulting gdb backtrace can be seen in the attached photo. To reproduce the segfault: Change the neutron absorber to AmBe (Needs to be some neutron absorber with a Z > 92 but needs more than 1 element). Change EM Physics list to G4EmStandardPhysics_option3 Compile the build and run: ./NeutronSource run1.mac Given the information in the backtrace, it seems that the dE/dx information for Z materials greater than 92 are not loaded. But if you run just G4_Am the segfault is not produced meaning that G4IonICRU73Data::ReadElementData is missing an error checking step for a nullptr return. We believe that the following changes to G4IonICRU73Data.hh should fix the dE/dx issues: Change "const G4Int ZTARGMAX=92;" to "const G4Int ZTARGMAX=98;" And change “G4PhysicsLogVector* fElmData[81][93] = {{nullptr}};” to “G4PhysicsLogVector* fElmData[81][99] = {{nullptr}};”