[Dune] error running code with MPI

[Vogelbacher Florian]

Dear Dune,

the reason for a new installation (see last mail on mailinglist about 
dune-grid-howto build problem) was a problem running our code using MPI 
on our machine.

The code is running in parallel on a Mac, but not on my machine. Has 
anyone seen something similar? ALUGrid problem?

Best wishes,
Florian

1.) the system:
gcc (GCC) 4.3.1

Linux felsim01 2.6.18-194.8.1.el5 #1 SMP Thu Jul 1 16:05:53 EDT 2010 
x86_64 x86_64 x86_64 GNU/Linux

mpirun (Open MPI) 1.3.3


2.) the error:

[...]
Created parallel ALUSimplexGrid<3,3> from macro grid file 
'../../../grids/spherical-capacitor.dgf'.

process[1] : #[total leaf element]=0
2010-Sep-03 10:11:07.563938 ::: hades3delectrostatic.cc:   320 ::: 
PRODUCTION  #[total leaf element]=7863
2010-Sep-03 10:11:07.564055 ::: hades3delectrostatic.cc:   323 ::: 
PRODUCTION  #[process in this run]=2
process[0] : #[total leaf element]=7863
2010-Sep-03 10:11:07.564110 ::: 
../globaldatainterface/globaldatainterface.hh:    75 ::: PRODUCTION  [ 
Get element tags from macro grid ...
2010-Sep-03 10:11:07.567824 ::: 
../globaldatainterface/globaldatainterface.hh:   134 ::: PRODUCTION  all 
element tags from macro grid are now available on cpu 0. ]
2010-Sep-03 10:11:07.567973 ::: hades3delectrostatic.cc:   353 ::: 
PRODUCTION  [parallel - #[global refinement]=0
2010-Sep-03 10:11:07.568019 ::: hades3delectrostatic.cc:   374 ::: 
PRODUCTION  [parallel - initiate load balance
2010-Sep-03 10:11:07.568065 ::: hades3delectrostatic: 
parallel/gitter_pll_sti.h:732: void 
ALUGridSpace::LinkedObject::Identifier::read(__gnu_cxx::__normal_iterator<const 
int*, std::vector<int, std::allocator<int> > >&, const 
__gnu_cxx::__normal_iterator<const int*, std::vector<int, 
std::allocator<int> > >&): Assertion `pos != end' failed.
[felsim01:02544] *** Process received signal ***
hades3delectrostatic: parallel/gitter_pll_sti.h:732: void 
ALUGridSpace::LinkedObject::Identifier::read(__gnu_cxx::__normal_iterator<const 
int*, std::vector<int, std::allocator<int> > >&, const 
__gnu_cxx::__normal_iterator<const int*, std::vector<int, 
std::allocator<int> > >&): Assertion `pos != end' failed.
[felsim01:02544] Signal: Aborted (6)
[felsim01:02544] Signal code:  (-6)
[felsim01:02543] *** Process received signal ***
[felsim01:02543] Signal: Aborted (6)
[felsim01:02543] Signal code:  (-6)
[felsim01:02543] [ 0] /lib64/libpthread.so.0 [0x35f8a0de60]
[felsim01:02543] [ 1] /lib64/libc.so.6(gsignal+0x35) [0x35f7e30045]
[felsim01:02543] [ 2] /lib64/libc.so.6(abort+0x110) [0x35f7e31ae0]
[felsim01:02543] [ 3] /lib64/libc.so.6(__assert_fail+0xf6) [0x35f7e29756]
[felsim01:02543] [ 4] 
../../../hades3delectrostatic/hades3delectrostatic(_ZN12ALUGridSpace8identifyINS_6Gitter5hedgeEEEvNS_14AccessIteratorIT_E6HandleERSt6vectorISt4pairISt4listIS6_SaIS6_EESB_ESaISC_EERKNS_13MpAccessLocalE+0x1645) 
[0x898de5]
[felsim01:02543] [ 5] 
../../../hades3delectrostatic/hades3delectrostatic(_ZN12ALUGridSpace9GitterPll14MacroGitterPll14identificationERNS_13MpAccessLocalE+0x386) 
[0x82bd16]
[felsim01:02543] [ 6] 
../../../hades3delectrostatic/hades3delectrostatic(_ZN12ALUGridSpace9GitterPll22notifyMacroGridChangesEv+0x75) 
[0x80eba5]
[felsim01:02543] [ 7] 
../../../hades3delectrostatic/hades3delectrostatic(_ZN12ALUGridSpace9GitterPll29loadBalancerGridChangesNotifyEv+0x448) 
[0x8129a8]
[felsim01:02543] [ 8] 
../../../hades3delectrostatic/hades3delectrostatic(_ZN12ALUGridSpace13GitterDunePll15duneLoadBalanceEv+0xd) 
[0x80c25d]
[felsim01:02543] [ 9] 
../../../hades3delectrostatic/hades3delectrostatic(main+0x199e) [0x6369ae]
[felsim01:02543] [10] /lib64/libc.so.6(__libc_start_main+0xf4) 
[0x35f7e1d8a4]
[felsim01:02543] [11] 
../../../hades3delectrostatic/hades3delectrostatic(_ZNSt8ios_base4InitD1Ev+0x51) 
[0x632b39]
[felsim01:02543] *** End of error message ***
[felsim01:02544] [ 0] /lib64/libpthread.so.0 [0x35f8a0de60]
[felsim01:02544] [ 1] /lib64/libc.so.6(gsignal+0x35) [0x35f7e30045]
[felsim01:02544] [ 2] /lib64/libc.so.6(abort+0x110) [0x35f7e31ae0]
[felsim01:02544] [ 3] /lib64/libc.so.6(__assert_fail+0xf6) [0x35f7e29756]
[felsim01:02544] [ 4] 
../../../hades3delectrostatic/hades3delectrostatic(_ZN12ALUGridSpace8identifyINS_6Gitter5hedgeEEEvNS_14AccessIteratorIT_E6HandleERSt6vectorISt4pairISt4listIS6_SaIS6_EESB_ESaISC_EERKNS_13MpAccessLocalE+0x1645) 
[0x898de5]
[felsim01:02544] [ 5] 
../../../hades3delectrostatic/hades3delectrostatic(_ZN12ALUGridSpace9GitterPll14MacroGitterPll14identificationERNS_13MpAccessLocalE+0x386) 
[0x82bd16]
[felsim01:02544] [ 6] 
../../../hades3delectrostatic/hades3delectrostatic(_ZN12ALUGridSpace9GitterPll22notifyMacroGridChangesEv+0x75) 
[0x80eba5]
[felsim01:02544] [ 7] 
../../../hades3delectrostatic/hades3delectrostatic(_ZN12ALUGridSpace9GitterPll29loadBalancerGridChangesNotifyEv+0x448) 
[0x8129a8]
[felsim01:02544] [ 8] 
../../../hades3delectrostatic/hades3delectrostatic(_ZN12ALUGridSpace13GitterDunePll15duneLoadBalanceEv+0xd) 
[0x80c25d]
[felsim01:02544] [ 9] 
../../../hades3delectrostatic/hades3delectrostatic(main+0x199e) [0x6369ae]
[felsim01:02544] [10] /lib64/libc.so.6(__libc_start_main+0xf4) 
[0x35f7e1d8a4]
[felsim01:02544] [11] 
../../../hades3delectrostatic/hades3delectrostatic(_ZNSt8ios_base4InitD1Ev+0x51) 
[0x632b39]
[felsim01:02544] *** End of error message ***
--------------------------------------------------------------------------
mpirun noticed that process rank 1 with PID 2544 on node felsim01 exited 
on signal 6 (Aborted).
--------------------------------------------------------------------------


-- 
---------------------------
Paul Scherrer Institut
Florian Vogelbacher
WBCA/004
5232 Villigen PSI
Switzerland

Mail: 	Florian.Vogelbacher at psi.ch
Phone:	+41 (0) 563105019
Web:	www.psi.ch
---------------------------