Segmentation fault when writing particle position to csv file

Recording, Visualizing and Persisting Data
1

Hello all,

So my simulation is a modification of the LXe example, and I am attempting to write to a csv file the position at which the primary particle enters/exits a particular physical volume. To do this, I have added the following code to the UserSteppingAction method in LXeSteppingAction.cc:

    if (theTrack->GetParentID() == 0)
    {
        if (thePostPV->GetName() == "scintillator1")
        {
            if (thePrePV->GetName() != "scintillator1")
            {
                CLHEP::Hep3Vector Det1 = theTrack->GetPosition();
                std::cout << Det1[0] << std::endl;
                analysisManager->FillNtupleDColumn(0, 14, Det1[0]);
            }
        }
    }

While this does successfully print the correct position, after the simulation finishes iterating through the rest of the particles but before the event is visualized, I get a segmentation fault and crash. This happens even if I delete either the print line or the csv write line, but if I remove both leaving only the line which gets the track position, I no longer receive the fault. I am new to C++ so this may be a basic programming error, but I haven’t had luck solving the problem on other forums. Any help would be much appreciated!

Geant4 11.1.1
Windows 10 enterprise
Visual Studio 2022
CMake 3.27.1

2

You should G4cout and G4endl. G4cout is thread safe.

1 Like
3

Thank you for these suggestions - I will implement them.

Unfortunately this will not solve my problem. I get the same fault even if I remove the cout line entirely.

4

Another thought…

Det1 is a CLHEP::Hep3Vector. What does Det1[0] mean? The notation implies it’s an array, but it’s not. I’m surprised it compiles. Look for some compilation errors or warnings.

Do you mean you want the x-component? That would be Det1.x().

John

1 Like
5

@allison Just a side note – Hep3Vector does have operator to access the coordinates by index.

1 Like
6

My first guess would be if your PostPV is a null pointer?

You could implement print statements after each if statement to see which one causes the fault

1 Like
7

I think ultimately you’ll have to run the program through a debugger to pinpoint the location and cause. There’s not really enough info here to diagnose properly.

1 Like
8

Thank you for the suggestion, this was exactly the problem. For reasons unclear to me, it seems like on the final loop of user stepping action the PostPV was a null pointer. Adding an if statement to check at the start fixed the problem. Thank you!

9

The final step for a lot of track is leaving the world volume. Since there’s nothing outside the world volume, the post-step pointer in that case must be null.

1 Like