Created attachment 453 [details] Macro file defining the view I am wondering if twisted solids are not thread safe. Iām getting this G4Exception with test202 with RayTracer, which uses multithreading. With RayTracerX, which stays on the master thread, all is OK. Try test202 in MT mode with /vis/open RayTracer /control/execute g4_01.g4view # the attached macro /vis/viewer/flush. -------- EEEE ------- G4Exception-START -------- EEEE ------- *** G4Exception : GeomSolids0003 issued by : G4TwistTubsSide::DistanceToPlane() Point p is behind the surfaces. *** Fatal Exception *** core dump *** -------- EEEE -------- G4Exception-END --------- EEEE ------- If I remove the twisted tubs - Tst202DetectorConstruction.cc line 106 becomes volumeSelection["Twisted_Tubs"] = false; I get this: -------- EEEE ------- G4Exception-START -------- EEEE ------- *** G4Exception : GeomNav0003 issued by : G4Navigator::GetLocalExitNormal() Parameters of solid: ----------------------------------------------------------- *** Dump for solid - aTwistedBox *** =================================================== Solid type: G4TwistedBox Parameters: pDx = 20 cm pDy = 30 cm pDz = 75 cm pPhiTwist = 50 deg ----------------------------------------------------------- Point for surface = (205.255,64.1194,-126.177) Surface Normal returned by Solid is not a Unit Vector. *** Fatal Exception *** core dump *** -------- EEEE -------- G4Exception-END --------- EEEE -------
I've fixed a potential race condition in G4VTwistedSurface, in tag germ-specific-V10-03-05.