The error always appears when init_value is completed executed
by project_solution. So I think _func is valid indeed.
In fact _func is assigned as follows in another function like
_func=std::tr1::shared_ptr<libMesh::FunctionBase<libMesh::Number> >
(new ParsedFunction<Number>(func_string));
Zhenyu
------------------ Original ------------------
From: "John Peterson";<[email protected]>;
Send time: Monday, Apr 20, 2015 10:37 PM
To: "grandrabbit"<[email protected]>;
Cc: "libmesh-users"<[email protected]>;
Subject: Re: [Libmesh-users] segmentation fault after projection solution
On Mon, Apr 20, 2015 at 8:04 AM, grandrabbit <[email protected]> wrote:
Hello,
I wrote a class for set up initial values and called it by
system.attach_init_function (init_sys);
Here is the part of init_sys
void init_sys(EquationSystems& es,const std::string& system_name)
{
System & system =
es.get_system<System>(system_name);
vector< string > formulas;
boost::split(formulas, system_name, boost::is_any_of("_"),
token_compress_on );
if(boost::iequals(formulas[0], string("ns")))
{
// TransientNonlinearImplicitSystem & system =
// es.get_system<TransientNonlinearImplicitSystem>(system_name);
system.project_solution(init_value, NULL, es.parameters); // run-time
error by petsc, segfault
cout<<"init_sys OK, "<<system_name<<endl; // this line never run due
to the following error
}
else
{
cout<<"unknown type from system_name yet"<<endl;
exit(-1);
}
}
Number init_value (const Point& p,
const Parameters& parameters,
const std::string& sysname,
const std::string& varname)
{
auto _func=_init_conditioner->get_func(sysname,varname);
if(!_func)
{
std::cerr<<"!!! init_value(...): _func is nullptr"<<endl;
exit(-1);
}
Real _time =parameters.get<Real> ("time");
Number result=(*_func)(p, _time);
ipoint++;
#if 1
cout<< sysname<<"."<<varname<<":";
cout<<"("<<p(0)<<","<<p(1)<<","<<p(2)<<"), "
<<"t= "<<_time
<<", value= "<<result
<<", ipoint="<<ipoint<<endl;
#endif
return result;
}
The whole code is cuessfully built and linked.
Now the question is when it runs, in the main() function
errors aborted the run when it reached
equation_systems.init ();
Then I located the error more precisely, say,
the above system.project_solution() in init_sys was partly executed.
I am sorry the original codes is long enough, but if any of you think it
necessary,
I can upload it afterwards,
Here are some screenshot of the errors:
ns_0.v:(5.98457,4.68673,0), t= 0, value= 0.069756, ipoint=783857
ns_0.v:(5.87916,4.9296,0), t= 0, value= 0.069756, ipoint=783858
ns_0.v:(5.93191,4.9295,0), t= 0, value= 0.069756, ipoint=783859
ns_0.v:(5.98465,4.92939,0), t= 0, value= 0.069756, ipoint=783860
ns_0.p:(5.86218,4.37566,0), t= 0, value= 0, ipoint=783861
ns_0.p:(6,4.37319,0), t= 0, value= 0, ipoint=783862
ns_0.p:(6,5,0), t= 0, value= 0, ipoint=783863
ns_0.p:(5.86402,5,0), t= 0, value= 0, ipoint=783864
[0]PETSC ERROR:
------------------------------------------------------------------------
[0]PETSC ERROR: Caught signal number 11 SEGV: Segmentation Violation, probably
memory access out of range
Looks like the segfault is in "_func" which I don't think you sent the code
for...
--
John
------------------------------------------------------------------------------
BPM Camp - Free Virtual Workshop May 6th at 10am PDT/1PM EDT
Develop your own process in accordance with the BPMN 2 standard
Learn Process modeling best practices with Bonita BPM through live exercises
http://www.bonitasoft.com/be-part-of-it/events/bpm-camp-virtual- event?utm_
source=Sourceforge_BPM_Camp_5_6_15&utm_medium=email&utm_campaign=VA_SF
_______________________________________________
Libmesh-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/libmesh-users
