Charm++ parallel programming system

[Eric Bohm <ebohm AT illinois.edu>]

Hard to say exactly without the Person code, but there is probably a bug in your Person constructor.

I created a minimalist shell for it in main.h (which you did not provide).  This version doesn't do anything interesting, but does run to completion.

#include "pgm.decl.h"
class Main : public CBase_Main {
 public:
  int numPersons;
  Main(CkArgMsg* m);
  Main(CkMigrateMessage* msg);
};

class Person: public CBase_Person
{
 public:
  Person(void){ CkExit();}
  Person(CkMigrateMessage *m){}
};



The output you provide is somewhat puzzling in that the charmrun command line is for launching ./pgm, whereas the internal output from charmrun is for running ./PseudoEpiFast.

On 03/26/2014 03:44 PM, Vijayasarathy Kannan wrote:

Hi Phil,

Compile: See attached Makefile

Run: > ./charmrun +p4 ./pgm

Full output: The below statements precede the backtrace.

=====

Running on 4 processors:  ./PseudoEpiFast 5 50 5

charmrun>  /usr/bin/setarch x86_64 -R  mpirun -np 4  ./PseudoEpiFast 5 50 5

Charm++> Running on MPI version: 2.2

Charm++> level of thread support used: MPI_THREAD_SINGLE (desired: MPI_THREAD_SINGLE)

Charm++> Running in non-SMP mode: numPes 4

Converse/Charm++ Commit ID: v6.6.0-rc1-29-g7e18651

CharmLB> Load balancer assumes all CPUs are same.

Charm++> Running on 1 unique compute nodes (12-way SMP).

Charm++> cpu topology info is gathered in 0.001 seconds.

=====

Thanks,

Vijay.

On Wed, Mar 26, 2014 at 4:38 PM, Phil Miller <mille121 AT illinois.edu> wrote:

Hi Vijay,

Could you provide a bit more detail, as follows:

- What command did you use to compile Charm++?

- What command did you use to run your test program?

- What is the full output from the runtime's startup preceding the backtrace shown? It should look something like
====
Charmrun> started all node programs in 1.674 seconds.
Converse/Charm++ Commit ID: v6.6.0-rc2-23-g590e140
Charm++> scheduler running in netpoll mode.
CharmLB> Load balancer assumes all CPUs are same.
Charm++> Running on 1 unique compute nodes (8-way SMP).
Charm++> cpu topology info is gathered in 0.005 seconds.
====

Please keep replies on the list, so that others can both offer help and benefit from the outcome.

Phil

On Wed, Mar 26, 2014 at 3:18 PM, Vijayasarathy Kannan <kvijay AT vt.edu> wrote:

Hi,

I get the following error when I try to run a charm++ program which initializes a chare array by the statement "personArray = CProxy_Person::ckNew(numPersons);" (line# 21 in main.C). I have attached my .C, .ci files.

pgm:29995 terminated with signal 11 at PC=4c0ba3 SP=7fffffffcaa0.  Backtrace:

./pgm(_ZN7CkArray13insertInitialERK12CkArrayIndexPvi+0x73)[0x4c0ba3]

./pgm(_ZN10CkArrayMap15populateInitialEiR12CkArrayIndexPvP8CkArrMgr+0x231)[0x4acbc1]

./pgm(_ZN7CkArrayC1ER14CkArrayOptionsR19CkMarshalledMessage10_ckGroupID+0x49c)[0x4c46ec]

./pgm(_ZN15CkIndex_CkArray23_call_CkArray_marshall1EPvS0_+0xbf)[0x4c5b7f]

./pgm(CkDeliverMessageFree+0x28)[0x497f48]

./pgm(_Z12_createGroup10_ckGroupIDP8envelope+0xcf)[0x49b9af]

./pgm(CkCreateGroup+0xaa)[0x49cf0a]

./pgm(_ZN14CProxy_CkArray5ckNewERK14CkArrayOptionsRK19CkMarshalledMessageRK10_ckGroupIDPK14CkEntryOptions+0x151)[0x4c3a71]

./pgm(_ZN16CProxy_ArrayBase13ckCreateArrayEP14CkArrayMessageiRK14CkArrayOptions+0x26e)[0x4c569e]

./pgm(_ZN13CProxy_Person5ckNewEi+0x46)[0x486126]

./pgm(_ZN4MainC1EP8CkArgMsg+0x260)[0x4864d0]

./pgm(_ZN12CkIndex_Main19_call_Main_CkArgMsgEPvS0_+0x12)[0x486612]

./pgm(_Z10_initCharmiPPc+0xc25)[0x494355]

./pgm(ConverseInit+0x1e6)[0x51a466]

./pgm(main+0x27)[0x48f807]

/lib64/libc.so.6(__libc_start_main+0xe6)[0x2aaaabad6bc6]

./pgm[0x4821a9]

I am not able to figure out why segmentation fault arises. Can you help me in finding out what I am doing wrong?

- Vijay.

_______________________________________________
charm mailing list
charm AT cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/charm

_______________________________________________
charm mailing list
charm AT cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/charm