Message: Re: Is RunManager C++ compilation optimization option dependent? Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

Feedback Re: Is RunManager C++ compilation optimization option dependent? 

Forum: Run Management
Re: Question Is RunManager C++ compilation optimization option dependent? (Abhijit Bhattacharyya)
Re: None Re: Is RunManager C++ compilation optimization option dependent? (Makoto Asai)
Re: None Re: Is RunManager C++ compilation optimization option dependent? (Abhijit Bhattacharyya)
Date: 30 Aug, 2018
From: Abhijit Bhattacharyya <Abhijit Bhattacharyya>

Hello!

Yes you are right. The "git log" shows 10.5.0.beta authored by Gabriele. best Abhijit

On Thu, 30 Aug 2018 15:52:35 GMT, Abhijit Bhattacharyya wrote:

> Hello,
> 
> I pulled from git master.
> 
> After installation, I see Geant4-10.5.0 in lib directory and also in share directory.
> 
> Best,
> Abhijit
> 
> On 30 Aug 2018 4:30 pm, "Asai, Makoto" <asai@slac.stanford.edu> wrote:
> Hello,
> Could you please confirm the version you use?
> We have not yet released version 10.5. Is it 10.5-beta?
> Makoto
> 
> On 8/30/18, 7:20 AM, "Abhijit Bhattacharyya"
> <abhijit.bhattacharyya@cern.ch> wrote:
> 
> >
> >*** Discussion title: Run Management
> >
> >I see one strange effect of optimization.
> >
> >If I compile my Geant4 10.5.0 (Multithreading) based code using
> >
> >cmake ../ -DCMAKE_BUILD_TYPE="Release" -DCMAKE_CXX_FLAGS="-std=c++11
> >-O3"
> >
> >and then run, I see core dump through following steps
> >
> >Code is started with 4 threads.
> > Define Blocks in ParallelWorld  initialised.
> >G4PhysListFactory::GetReferencePhysList <QGSP_BIC_HP>  EMoption= 0
> ><<< Geant4 Physics List simulation engine: QGSP_BIC_HP 2.0
> ><<< Reference Physics List QGSP_BIC_HP is built
> >BookHisto Ntuples created.
> >Root data is being stored.
> >BookHisto Ntuples created.
> >Root data is being stored.
> >***** Table : Nb of materials = 0 *****
> >./NeutVEGA.run: line 32: 29855 Segmentation fault      (core dumped)
> >./tarcVega Data/GeomData/tarc_geometry.gdml -m $1
> >real    0m0.428s
> >user    0m0.105s
> >sys     0m0.046s
> >
> >If I analyse the dumped core with the executable code in gdb then I see
> >that
> >
> >Reading symbols from ./NeutVega...(no debugging symbols found)...done.
> >[New LWP 29855]
> >[Thread debugging using libthread_db enabled]
> >Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
> >Core was generated by `./tarcVega Data/GeomData/tarc_geometry.gdml -m
> >run.mac'.
> >Program terminated with signal SIGSEGV, Segmentation fault.
> >#0  0x00007f5d61d92007 in
> >G4RunManagerKernel::DefineWorldVolume(G4VPhysicalVolume*, bool) ()
> >   from /opt/G4MT_Master/share/Geant4-10.5.0/../../lib/libG4run.so
> >(gdb)
> >(gdb)
> >(gdb) bt
> >#0  0x00007f5d61d92007 in
> >G4RunManagerKernel::DefineWorldVolume(G4VPhysicalVolume*, bool) ()
> >   from /opt/G4MT_Master/share/Geant4-10.5.0/../../lib/libG4run.so
> >#1  0x00007f5d61d7a0be in G4RunManager::InitializeGeometry() ()
> >   from /opt/G4MT_Master/share/Geant4-10.5.0/../../lib/libG4run.so
> >#2  0x00007f5d61d7d7a9 in G4RunManager::Initialize() ()
> >   from /opt/G4MT_Master/share/Geant4-10.5.0/../../lib/libG4run.so
> >#3  0x00007f5d61d85179 in G4MTRunManager::Initialize() ()
> >   from /opt/G4MT_Master/share/Geant4-10.5.0/../../lib/libG4run.so
> >#4  0x000056405b4ee3df in main ()
> >(gdb)
> >
> >However, if I compile the same code without any modification using
> >
> >cmake ../ -DCMAKE_BUILD_TYPE="Debug" -DCMAKE_CXX_FLAGS="-std=c++11 -O1
> >-g"
> >
> >then I do not see error and the code runs.
> >
> >Is it normal or I have problem in Geometry. By the by, I am using
> >imported GDML file.
> >
> >Thanks and Regards Abhijit
> >
> >-------------------------------------------------------------
> >Visit this GEANT4 at hypernews.slac.stanford.edu message (to reply or
> >unsubscribe) at:
> >http://hypernews.slac.stanford.edu/HyperNews/geant4/get/runmanage/449.html
> >
> 
>  [ MIME part of type text/html without a name stripped ]
> 

 Add Message Add Message
to: "Re: Is RunManager C++ compilation optimization option dependent?"

 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 ]