On Mon, 26 Jan 2015 12:11:18 +0100 Mikael Mortensen <[email protected]> wrote:
> > > On 26 Jan 2015, at 11:53, Jan Blechta <[email protected]> > > wrote: > > > > On Mon, 26 Jan 2015 11:42:42 +0100 > > Mikael Mortensen <[email protected] > > <mailto:[email protected]>> wrote: > > > >> Hi, > >> > >> With recent (today) hashdist installation I run the following > >> test.py script > >> > >> from dolfin import * > >> > >> mesh = UnitSquareMesh(4, 4) > >> V = VectorFunctionSpace(mesh, 'CG', 1) > >> u = interpolate(Expression(('x[0]', 'x[1]')), V) > >> print “Hi" > >> > >> > >> [mikael@ubuntu tests (master)]$ python test.py > >> Hi > >> > >> and then it just hangs there without exiting. Same thing in > >> ipython. I am currently using host python, but I got the same when > >> I compiled with hashdist python. Anyone else experiencing this > >> weirdness? Any idea what is going on? I’m on Ubuntu 14.04 and I’ve > >> done instant-clean. > > > > To debug it, you could try attaching debugger (if it is allowed on > > your system) > > > > gdb python <pid> > > > > hit ^C, and print stacktrace > > > > (gdb) bt > > > > Jan > > Thanks for the tip. Not really familiar with gdb so I tried pdb first > > python -m pdb test_GaussDivergence.py > > /home/mikael/MySoftware/fenicstools/tests/test.py(5)<module>() > -> from dolfin import * > (Pdb) continue > Hi > The program finished and will be restarted > > and then it just hangs there. Ctrl+C does nothing. > > > With gdb I am getting somewhere though: > > gdb python > Reading symbols from python...(no debugging symbols found)...done. > (gdb) run test.py > Starting > program: /home/mikael/.hashdist/bld/profile/6axink4nal3d/bin/python > test.py process 16999 is executing new > program: /home/mikael/.hashdist/bld/python/qlvjlzdbishm/bin/python2.7 > [Thread debugging using libthread_db enabled] Using host libthread_db > library "/lib/x86_64-linux-gnu/libthread_db.so.1". [New Thread > 0x7fffdc7e3700 (LWP 17005)] Hi ^C > Program received signal SIGINT, Interrupt. > __lll_lock_wait () > at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:135 > 135 ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S: No > such file or directory. (gdb) bt #0 __lll_lock_wait () > at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:135 #1 > 0x00007ffff7bc6657 in _L_lock_909 () > from /lib/x86_64-linux-gnu/libpthread.so.0 #2 0x00007ffff7bc6480 in > __GI___pthread_mutex_lock (mutex=0x7ffff063fa10 <attribute_lock+16>) > at ../nptl/pthread_mutex_lock.c:79 #3 0x00007ffff03a140c in > ompi_attr_get_c () > from /home/mikael/.hashdist/bld/mpi/vrn4awnke2cx/lib/libmpi.so.1 #4 > 0x00007ffff03cba27 in MPI_Attr_get () > from /home/mikael/.hashdist/bld/mpi/vrn4awnke2cx/lib/libmpi.so.1 #5 > 0x00007ffff07b2afb in Petsc_DelComm_Outer () > from /home/mikael/.hashdist/bld/petsc/dfw5fwbfp6hl/lib/libpetsc.so.3.5 > #6 0x00007ffff03a1e88 in ompi_attr_delete () > from /home/mikael/.hashdist/bld/mpi/vrn4awnke2cx/lib/libmpi.so.1 #7 > 0x00007ffff03cb93c in MPI_Attr_delete () > from /home/mikael/.hashdist/bld/mpi/vrn4awnke2cx/lib/libmpi.so.1 #8 > 0x00007ffff07a2baa in PetscCommDestroy () > from /home/mikael/.hashdist/bld/petsc/dfw5fwbfp6hl/lib/libpetsc.so.3.5 > #9 0x00007ffff07a52bc in PetscHeaderDestroy_Private () > from /home/mikael/.hashdist/bld/petsc/dfw5fwbfp6hl/lib/libpetsc.so.3.5 > #10 0x00007ffff081372c in ISLocalToGlobalMappingDestroy () > from /home/mikael/.hashdist/bld/petsc/dfw5fwbfp6hl/lib/libpetsc.so.3.5 > #11 0x00007ffff0819d08 in PetscLayoutDestroy () > from /home/mikael/.hashdist/bld/petsc/dfw5fwbfp6hl/lib/libpetsc.so.3.5 > #12 0x00007ffff0871718 in VecDestroy () > from /home/mikael/.hashdist/bld/petsc/dfw5fwbfp6hl/lib/libpetsc.so.3.5 > #13 0x00007ffff22bb15c in dolfin::PETScVector::~PETScVector() () > from /home/mikael/.hashdist/bld/dolfin/wt5zll4ojwwy/lib/libdolfin.so.1.5 > #14 0x00007ffff22bb229 in dolfin::PETScVector::~PETScVector() () > from /home/mikael/.hashdist/bld/dolfin/wt5zll4ojwwy/lib/libdolfin.so.1.5 > #15 0x00007ffff221c34a in dolfin::Function::~Function() () > from /home/mikael/.hashdist/bld/dolfin/wt5zll4ojwwy/lib/libdolfin.so.1.5 > #16 0x00007ffff221c4d9 in dolfin::Function::~Function() () > from /home/mikael/.hashdist/bld/dolfin/wt5zll4ojwwy/lib/libdolfin.so.1.5 > #17 0x00007ffff308f999 in > std::_Sp_counted_base<(__gnu_cxx::_Lock_policy)2>::_M_release > (this=0x19bbc60) at /usr/include/c++/4.8/bits/shared_ptr_base.h:144 > #18 0x00007fffdf8093cc in _wrap_delete_Function () > from > /home/mikael/Software/hashstack/default/lib/python2.7/site-packages/dolfin/cpp/_function.so > #19 0x00007fffdf7fc15e in SwigPyObject_dealloc () > from > /home/mikael/Software/hashstack/default/lib/python2.7/site-packages/dolfin/cpp/_function.so > #20 0x00000000005392ff in ?? () #21 0x00000000004d914b in ?? () #22 > 0x00000000004fdb96 in PyDict_SetItem () #23 0x000000000055a9e1 in > _PyModule_Clear () #24 0x00000000004f2ad4 in PyImport_Cleanup () #25 > 0x000000000042fa89 in Py_Finalize () #26 0x000000000046ac10 in > Py_Main () Can you reproduce it with PETSc example http://www.mcs.anl.gov/petsc/petsc-current/src/vec/is/examples/tests/ex1.c.html which also calls ISLocalToGlobalMappingDestroy? Just navigate to <petsc>/src/vec/is/examples/tests/, type make and execute. Jan > > So I’m trying to google this now. I think it’s because I’m using a > newer version of openmpi than in the default hashdist script. Default > is 1.6.5, but I have compiled with 1.8.2. 1.6.5 gives me problems > with h5py, which is why I upgraded it in the first place. So at the > moment h5py works, but not fenics:-( > > M > > > > >> > >> Mikael > >> > >> > >> _______________________________________________ > >> fenics-support mailing list > >> [email protected] > >> <mailto:[email protected]> > >> http://fenicsproject.org/mailman/listinfo/fenics-support > >> <http://fenicsproject.org/mailman/listinfo/fenics-support> _______________________________________________ fenics-support mailing list [email protected] http://fenicsproject.org/mailman/listinfo/fenics-support
