Message: Re: Infinite Loop in G4SubtractionSolid::DistanceToIn() Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

Feedback Re: Infinite Loop in G4SubtractionSolid::DistanceToIn() 

Forum: Geometry
Re: Question Infinite Loop in G4SubtractionSolid::DistanceToIn() (Stefan E. Mueller)
Date: 10 Sep, 2014
From: Gabriele Cosmo <Gabriele Cosmo>

Hi Stefan,

I would suggest you open a bug report in Bugzilla:
  http://bugzilla-geant4.kek.jp/enter_bug.cgi?product=Geant4
assign it to "processes/scoring" and provide all the information (cut&paste your
findings from this thread), so we can properly track the issue.
At the same time, if you can provide a test-case which reproduces the problem
it would be very helpful!
The infinite loop is definitely the effect of a corrupted state your application has
come through (enters in a loop in G4SubtractionSolid because the points
provided are not valid...).
It would be also interesting if you can run your test through Valgrind, now that
you can quicky get to the offending event, to see if any memory error gets
detected by the tool and where exactly in the code!

Thanks,
Gabriele

-----------------
On Fri, 05 Sep 2014 13:45:38 GMT, Stefan E. Mueller wrote:
> Dear all,
> 
> when I try to run my application, at a certain point, an infinite loop
> is created with a call to G4SubtractionSolid::DistanceToIn(), and I
> obtain the message:
> 
> -------- WWWW ------- G4Exception-START -------- WWWW -------
> *** G4Exception : GeomSolids1001
>       issued by : G4SubtractionSolid::DistanceToIn(p,v)
> Illegal condition caused by solids: B48 and B43
> Looping detected in point (nan,nan,nan), from original point (nan,nan,nan) and direction (0.7903103663410459,-0.4908280694712698,0.3667387777055873)
> Computed candidate distance: 0*mm. Returning candidate distance.
> 
> *** This is just a warning message. ***
> 
> I made sure that the two bodies do not have touching surfaces,
> nevertheless the error did not go away. Also strange that the starting
> point is (nan,nan,nan). DistanceToIn is called from
> G4VoxelNavigation::ComputeStep in my case.
> 
> In problem 1304 on bugzilla-geant4.kek.jp, it was mentioned that there
> is still an issue with "Navigator". Since I am using GEANT4 Primitive
> Scorers in my code, I gave it a try and commented out the call to the
> G4ScoringManager in my main, as well as all scoring related commands in
> my input macro.
> 
> I rerun the application, using the same random seeds as before, and now
> the runs finish without problems (e.g. no infinite loops).
> 
> However, I don't have the scoring any more.
> 
> Anyone has any ideas why the Warning goes away when I switch of scoring?
> And how to get scoring without the warning (and the infinite loop)?
> 
> I am using Geant4.10p02 plus some additional patches provided by Geant4
> developers to account for other problems.
> 
> Stefan
> 




On Fri, 05 Sep 2014 13:45:38 GMT, Stefan E. Mueller wrote:
> Dear all,
> 
> when I try to run my application, at a certain point, an infinite loop
> is created with a call to G4SubtractionSolid::DistanceToIn(), and I
> obtain the message:
> 
> -------- WWWW ------- G4Exception-START -------- WWWW -------
> *** G4Exception : GeomSolids1001
>       issued by : G4SubtractionSolid::DistanceToIn(p,v)
> Illegal condition caused by solids: B48 and B43
> Looping detected in point (nan,nan,nan), from original point (nan,nan,nan) and direction (0.7903103663410459,-0.4908280694712698,0.3667387777055873)
> Computed candidate distance: 0*mm. Returning candidate distance.
> 
> *** This is just a warning message. ***
> 
> I made sure that the two bodies do not have touching surfaces,
> nevertheless the error did not go away. Also strange that the starting
> point is (nan,nan,nan). DistanceToIn is called from
> G4VoxelNavigation::ComputeStep in my case.
> 
> In problem 1304 on bugzilla-geant4.kek.jp, it was mentioned that there
> is still an issue with "Navigator". Since I am using GEANT4 Primitive
> Scorers in my code, I gave it a try and commented out the call to the
> G4ScoringManager in my main, as well as all scoring related commands in
> my input macro.
> 
> I rerun the application, using the same random seeds as before, and now
> the runs finish without problems (e.g. no infinite loops).
> 
> However, I don't have the scoring any more.
> 
> Anyone has any ideas why the Warning goes away when I switch of scoring?
> And how to get scoring without the warning (and the infinite loop)?
> 
> I am using Geant4.10p02 plus some additional patches provided by Geant4
> developers to account for other problems.
> 
> Stefan
> 

 Add Message Add Message
to: "Re: Infinite Loop in G4SubtractionSolid::DistanceToIn()"

 Subscribe Subscribe

This site runs SLAC HyperNews version 1.11-slac-98, derived from the original HyperNews


[ Geant 4 Home | Geant 4 HyperNews | Search | Request New Forum | Feedback ]