libmesh-users

[Libmesh-users] Runtime Libmesh issue

[Andrew D. <and...@gm...>]

Hi,

I am trying to solve a nonlinear PDE with libMesh.  I have installed
libmesh with petsc enabled and I have written a simple "hello world"
program that links with the library mesh_dbg, which complies fine.
 However, at runtime I get the message:

dyld: Symbol not found: _DM_CLASSID
  Referenced from: /usr/local/lib/libmesh_dbg.0.dylib
  Expected in: flat namespace
 in /usr/local/lib/libmesh_dbg.0.dylib
Trace/BPT trap: 5

Does anyone know what this means?  and/or how to fix this problem?

Thanks!
Andy

Re: [Libmesh-users] Runtime Libmesh issue

[John P. <jwp...@gm...>]

On Thu, Jun 6, 2013 at 8:44 AM, Andrew Davis <and...@gm...> wrote:
> Hi,
>
> I am trying to solve a nonlinear PDE with libMesh.  I have installed
> libmesh with petsc enabled and I have written a simple "hello world"
> program that links with the library mesh_dbg, which complies fine.
>  However, at runtime I get the message:
>
> dyld: Symbol not found: _DM_CLASSID
>   Referenced from: /usr/local/lib/libmesh_dbg.0.dylib
>   Expected in: flat namespace
>  in /usr/local/lib/libmesh_dbg.0.dylib
> Trace/BPT trap: 5

Might be a PETSc DM thing?

What does your Makefile look like?

--
John

Re: [Libmesh-users] Runtime Libmesh issue

[Andrew D. <and...@gm...>]

Hi,

I'm using cmake to generate the Makefile so it is difficult to interpret,
but you are quite right about the PETSc DM issue; I completely forgot to
tell cmake where the PETSc libraries live.  I added this to CMakeLists.txt
and re-complied my code and now I get this error (at runtime) when I
initialize libMesh:

[-1]PETSC ERROR: --------------------- Error Message
------------------------------------
[-1]PETSC ERROR: No support for this operation for this object type!
[-1]PETSC ERROR: You cannot set PETSC_COMM_WORLD if you have not
initialized MPI first!
[-1]PETSC ERROR:
------------------------------------------------------------------------
[-1]PETSC ERROR:
[-1]PETSC ERROR: See docs/changes/index.html for recent updates.
[-1]PETSC ERROR: See docs/faq.html for hints about trouble shooting.
[-1]PETSC ERROR: See docs/index.html for manual pages.
[-1]PETSC ERROR:
------------------------------------------------------------------------
[-1]PETSC ERROR: Libraries linked from
/Users/Andy/software/petsc/arch-darwin-c-debug/lib
[-1]PETSC ERROR: Configure run at Wed Jun  5 17:06:06 2013
[-1]PETSC ERROR: Configure options --with-cc=gcc --with-fc=gfortran
--download-f-blas-lapack --download-mpich
[-1]PETSC ERROR:
------------------------------------------------------------------------
[-1]PETSC ERROR: PetscInitialize() line 676 in
/Users/Andy/software/petsc/src/sys/objects/pinit.c
[-1]PETSC ERROR: LibMeshInit() line 457 in
"unknowndirectory/"src/base/libmesh.C

Could this be because I had to add the --disable-mpi flag to libMesh when I
configured?  If I don't disable MPI when I configure libMesh the compiler
complains (at compile time):

petsc/arch-darwin-c-debug/include/mpicxx.h:2723:34: error: declaration of C
function 'void Parmetis::MPI::Init(int&, char**&)' conflicts with
 extern void Init(int &, char **& );

Thanks for your help,
Andy


On Thu, Jun 6, 2013 at 11:19 AM, John Peterson <jwp...@gm...> wrote:

> On Thu, Jun 6, 2013 at 8:44 AM, Andrew Davis <and...@gm...> wrote:
> > Hi,
> >
> > I am trying to solve a nonlinear PDE with libMesh.  I have installed
> > libmesh with petsc enabled and I have written a simple "hello world"
> > program that links with the library mesh_dbg, which complies fine.
> >  However, at runtime I get the message:
> >
> > dyld: Symbol not found: _DM_CLASSID
> >   Referenced from: /usr/local/lib/libmesh_dbg.0.dylib
> >   Expected in: flat namespace
> >  in /usr/local/lib/libmesh_dbg.0.dylib
> > Trace/BPT trap: 5
>
> Might be a PETSc DM thing?
>
> What does your Makefile look like?
>
> --
> John
>

Re: [Libmesh-users] Runtime Libmesh issue

[Dmitry K. <ka...@mc...>]

It appears that --disable-mpi prevents MPI_Init() from being called by
libMesh, which causes libMesh::COMM_WORLD to be left uninitialized, as
far as I can determine. Since PETSc is being used, PETSC_COMM_WORLD is
set to libMesh::COMM_WORLD and PetscInitialize() is called.
Since MPI_Init() has not been called, at that stage PETSc expects
PETSC_COMM_WORLD to be MPI_COMM_NULL, not the uninitialized
libMesh::COMM_WORLD. Hence, the error you see.

It seems to me that --disable-mpi in libMesh is a bad option when
using PETSc :-)
I'm not sure why using MPI cases a conflict with ParMetis for you.
That issue seems to be closer to the root cause of you trouble.

Dmitry.

On Thu, Jun 6, 2013 at 11:09 AM, Andrew Davis <and...@gm...> wrote:
> Hi,
>
> I'm using cmake to generate the Makefile so it is difficult to interpret,
> but you are quite right about the PETSc DM issue; I completely forgot to
> tell cmake where the PETSc libraries live.  I added this to CMakeLists.txt
> and re-complied my code and now I get this error (at runtime) when I
> initialize libMesh:
>
> [-1]PETSC ERROR: --------------------- Error Message
> ------------------------------------
> [-1]PETSC ERROR: No support for this operation for this object type!
> [-1]PETSC ERROR: You cannot set PETSC_COMM_WORLD if you have not
> initialized MPI first!
> [-1]PETSC ERROR:
> ------------------------------------------------------------------------
> [-1]PETSC ERROR:
> [-1]PETSC ERROR: See docs/changes/index.html for recent updates.
> [-1]PETSC ERROR: See docs/faq.html for hints about trouble shooting.
> [-1]PETSC ERROR: See docs/index.html for manual pages.
> [-1]PETSC ERROR:
> ------------------------------------------------------------------------
> [-1]PETSC ERROR: Libraries linked from
> /Users/Andy/software/petsc/arch-darwin-c-debug/lib
> [-1]PETSC ERROR: Configure run at Wed Jun  5 17:06:06 2013
> [-1]PETSC ERROR: Configure options --with-cc=gcc --with-fc=gfortran
> --download-f-blas-lapack --download-mpich
> [-1]PETSC ERROR:
> ------------------------------------------------------------------------
> [-1]PETSC ERROR: PetscInitialize() line 676 in
> /Users/Andy/software/petsc/src/sys/objects/pinit.c
> [-1]PETSC ERROR: LibMeshInit() line 457 in
> "unknowndirectory/"src/base/libmesh.C
>
> Could this be because I had to add the --disable-mpi flag to libMesh when I
> configured?  If I don't disable MPI when I configure libMesh the compiler
> complains (at compile time):
>
> petsc/arch-darwin-c-debug/include/mpicxx.h:2723:34: error: declaration of C
> function 'void Parmetis::MPI::Init(int&, char**&)' conflicts with
>  extern void Init(int &, char **& );
>
> Thanks for your help,
> Andy
>
>
> On Thu, Jun 6, 2013 at 11:19 AM, John Peterson <jwp...@gm...> wrote:
>
>> On Thu, Jun 6, 2013 at 8:44 AM, Andrew Davis <and...@gm...> wrote:
>> > Hi,
>> >
>> > I am trying to solve a nonlinear PDE with libMesh.  I have installed
>> > libmesh with petsc enabled and I have written a simple "hello world"
>> > program that links with the library mesh_dbg, which complies fine.
>> >  However, at runtime I get the message:
>> >
>> > dyld: Symbol not found: _DM_CLASSID
>> >   Referenced from: /usr/local/lib/libmesh_dbg.0.dylib
>> >   Expected in: flat namespace
>> >  in /usr/local/lib/libmesh_dbg.0.dylib
>> > Trace/BPT trap: 5
>>
>> Might be a PETSc DM thing?
>>
>> What does your Makefile look like?
>>
>> --
>> John
>>
> ------------------------------------------------------------------------------
> How ServiceNow helps IT people transform IT departments:
> 1. A cloud service to automate IT design, transition and operations
> 2. Dashboards that offer high-level views of enterprise services
> 3. A single system of record for all IT processes
> http://p.sf.net/sfu/servicenow-d2d-j
> _______________________________________________
> Libmesh-users mailing list
> Lib...@li...
> https://lists.sourceforge.net/lists/listinfo/libmesh-users

Re: [Libmesh-users] Runtime Libmesh issue

[Roy S. <roy...@ic...>]

On Thu, 6 Jun 2013, Dmitry Karpeyev wrote:

> It seems to me that --disable-mpi in libMesh is a bad option when
> using PETSc :-)

Actually, our continuous integration tests --disable-mpi using PETSc.

*But*, you have to compile and link against a PETSc build which was
configured to be "uniprocessor" and avoid MPI as well.

If your PETSc is MPI-aware, your libMesh needs to be too.

> That issue seems to be closer to the root cause of you trouble.

Definitely.
---
Roy

Re: [Libmesh-users] Runtime Libmesh issue

[Andrew D. <and...@gm...>]

Thanks!  This makes sense.  I recompiled PETSc with the option --with-mpi=0
and then installed libMesh with --disable-mpi.  I'm no longer having
trouble initializing libMesh!  However, I am still not quite up and
running.

I create a libMesh::EquationSystems with a libMesh::NonlinearImplicitSystem
attached.  However, when I run EquationSystems::Init() I get the runtime
error

[0]PETSC ERROR: PetscCommDuplicate() line 140 in
petsc/src/sys/objects/tagm.c
[0]PETSC ERROR: PetscHeaderCreate_Private() line 55 in
petsc/src/sys/objects/inherit.c
[0]PETSC ERROR: VecCreate() line 39 in
petsc/src/vec/vec/interface/veccreate.c
[0]PETSC ERROR: VecCreateGhostWithArray() line 374 in
petsc/src/vec/vec/impls/mpi/pbvec.c
[0]PETSC ERROR: VecCreateGhost() line 447 in
petsc/src/vec/vec/impls/mpi/pbvec.c
[0]PETSC ERROR: init() line 894 in
"unknowndirectory/"./include/libmesh/petsc_vector.h
Abort trap: 6

Is this related to the MPI issue from I was having earlier?  Does libMesh
not work with the --disable-mpi option?  Am I turning off mpi correctly
when I reconfigure PETSc?

Thanks again for all your help,
Andy


On Thu, Jun 6, 2013 at 12:46 PM, Roy Stogner <roy...@ic...>wrote:

>
> On Thu, 6 Jun 2013, Dmitry Karpeyev wrote:
>
>  It seems to me that --disable-mpi in libMesh is a bad option when
>> using PETSc :-)
>>
>
> Actually, our continuous integration tests --disable-mpi using PETSc.
>
> *But*, you have to compile and link against a PETSc build which was
> configured to be "uniprocessor" and avoid MPI as well.
>
> If your PETSc is MPI-aware, your libMesh needs to be too.
>
>
>  That issue seems to be closer to the root cause of you trouble.
>>
>
> Definitely.
> ---
> Roy
>

Re: [Libmesh-users] Runtime Libmesh issue

[Roy S. <roy...@ic...>]

On Thu, 6 Jun 2013, Andrew Davis wrote:

> I create a libMesh::EquationSystems with a
> libMesh::NonlinearImplicitSystem attached.  However, when I run
> EquationSystems::Init() I get the runtime error 

Is there an error type?  This looks like just a stack trace, but
doesn't PETSc also let you know what assertion or signal killed you?

> [0]PETSC ERROR: PetscCommDuplicate() line 140 in petsc/src/sys/objects/tagm.c
> [0]PETSC ERROR: PetscHeaderCreate_Private() line 55 in petsc/src/sys/objects/inherit.c
> [0]PETSC ERROR: VecCreate() line 39 in petsc/src/vec/vec/interface/veccreate.c
> [0]PETSC ERROR: VecCreateGhostWithArray() line 374 in petsc/src/vec/vec/impls/mpi/pbvec.c
> [0]PETSC ERROR: VecCreateGhost() line 447 in petsc/src/vec/vec/impls/mpi/pbvec.c
> [0]PETSC ERROR: init() line 894 in "unknowndirectory/"./include/libmesh/petsc_vector.h
> Abort trap: 6
> 
> Is this related to the MPI issue from I was having earlier?  Does
> libMesh not work with the --disable-mpi option?  Am I turning off
> mpi correctly when I reconfigure PETSc?

We've got libMesh with --disable-mpi passing continuous integration
tests now, and all we do there is --with-mpi=0 for PETSc...  This
certainly *looks* like some kind of MPI issue, though!
---
Roy

Re: [Libmesh-users] Runtime Libmesh issue

[Dmitry K. <ka...@mc...>]

Is this the complete error message?
Can you run with -on_error_attach_debugger and send the backtrace
along with petsc's configure.log?

Dmitry.

On Thu, Jun 6, 2013 at 3:19 PM, Roy Stogner <roy...@ic...> wrote:
>
> On Thu, 6 Jun 2013, Andrew Davis wrote:
>
>> I create a libMesh::EquationSystems with a
>> libMesh::NonlinearImplicitSystem attached.  However, when I run
>> EquationSystems::Init() I get the runtime error
>
>
> Is there an error type?  This looks like just a stack trace, but
> doesn't PETSc also let you know what assertion or signal killed you?
>
>
>> [0]PETSC ERROR: PetscCommDuplicate() line 140 in
>> petsc/src/sys/objects/tagm.c
>> [0]PETSC ERROR: PetscHeaderCreate_Private() line 55 in
>> petsc/src/sys/objects/inherit.c
>> [0]PETSC ERROR: VecCreate() line 39 in
>> petsc/src/vec/vec/interface/veccreate.c
>> [0]PETSC ERROR: VecCreateGhostWithArray() line 374 in
>> petsc/src/vec/vec/impls/mpi/pbvec.c
>> [0]PETSC ERROR: VecCreateGhost() line 447 in
>> petsc/src/vec/vec/impls/mpi/pbvec.c
>> [0]PETSC ERROR: init() line 894 in
>> "unknowndirectory/"./include/libmesh/petsc_vector.h
>> Abort trap: 6
>>
>> Is this related to the MPI issue from I was having earlier?  Does
>> libMesh not work with the --disable-mpi option?  Am I turning off
>> mpi correctly when I reconfigure PETSc?
>
>
> We've got libMesh with --disable-mpi passing continuous integration
> tests now, and all we do there is --with-mpi=0 for PETSc...  This
> certainly *looks* like some kind of MPI issue, though!
> ---
> Roy

Re: [Libmesh-users] Runtime Libmesh issue

[Andrew D. <and...@gm...>]

Hi,

>Is there an error type?  This looks like just a stack trace, but
>doesn't PETSc also let you know what assertion or signal killed you?

Yes, I agree this is just a stack trace, but it is the only thing that
prints when the program crashes.

>We've got libMesh with --disable-mpi passing continuous integration
>tests now, and all we do there is --with-mpi=0 for PETSc...  This
>certainly *looks* like some kind of MPI issue, though!

This does look like some kind of MPI issue.  Just to be safe I tried
reinstalling MPI and then reconfigure/compiling PETSc and libMesh.  This
resulted in the same runtime error (the stack trace) as before.  I'm not
really sure what else to I could do to figure out what the issue is.

Thanks again,
Andy

On Thu, Jun 6, 2013 at 4:19 PM, Roy Stogner <roy...@ic...>wrote:

>
> On Thu, 6 Jun 2013, Andrew Davis wrote:
>
>  I create a libMesh::EquationSystems with a
>> libMesh::**NonlinearImplicitSystem attached.  However, when I run
>> EquationSystems::Init() I get the runtime error
>>
>
> Is there an error type?  This looks like just a stack trace, but
> doesn't PETSc also let you know what assertion or signal killed you?
>
>
>  [0]PETSC ERROR: PetscCommDuplicate() line 140 in
>> petsc/src/sys/objects/tagm.c
>> [0]PETSC ERROR: PetscHeaderCreate_Private() line 55 in
>> petsc/src/sys/objects/inherit.**c
>> [0]PETSC ERROR: VecCreate() line 39 in petsc/src/vec/vec/interface/**
>> veccreate.c
>> [0]PETSC ERROR: VecCreateGhostWithArray() line 374 in
>> petsc/src/vec/vec/impls/mpi/**pbvec.c
>> [0]PETSC ERROR: VecCreateGhost() line 447 in petsc/src/vec/vec/impls/mpi/
>> **pbvec.c
>> [0]PETSC ERROR: init() line 894 in "unknowndirectory/"./include/**
>> libmesh/petsc_vector.h
>> Abort trap: 6
>>
>> Is this related to the MPI issue from I was having earlier?  Does
>> libMesh not work with the --disable-mpi option?  Am I turning off
>> mpi correctly when I reconfigure PETSc?
>>
>
> We've got libMesh with --disable-mpi passing continuous integration
> tests now, and all we do there is --with-mpi=0 for PETSc...  This
> certainly *looks* like some kind of MPI issue, though!
> ---
> Roy

Re: [Libmesh-users] Runtime Libmesh issue

[Jed B. <je...@59...>]

Andrew Davis <and...@gm...> writes:

> This does look like some kind of MPI issue.  Just to be safe I tried
> reinstalling MPI and then reconfigure/compiling PETSc and libMesh.  

Exactly which options are you configuring with?  This message makes it
sound like you have an MPI-enabled PETSc and a libMesh configured
without MPI.

> This resulted in the same runtime error (the stack trace) as before.

Which one?  In MPI_Attr_get?  Can you attach a debugger and find out
what out what the value of comm was?  MPIUNI's implementation is simple:

int MPI_Attr_get(MPI_Comm comm,int keyval,void *attribute_val,int *flag)
{
  if (comm-1 < 0 || comm-1 > 3) return 1;
  if (!keyval) Keyval_setup();
  *flag                  = attr[comm-1][keyval].active;
  *(void**)attribute_val = attr[comm-1][keyval].attribute_val;
  return MPI_SUCCESS;
}

It appears that libMesh::COMM_WORLD is being set to 0 when not using
MPI?

Re: [Libmesh-users] Runtime Libmesh issue

[Andrew D. <and...@gm...>]

Hi,

Sorry for the delay getting back do you on this.  I think I managed to
solve some of my issues but haven't quite figured it out exactly.  I think
this is at least a different issue than before.

Previously, I had installed MPI with Homebrew which was apparently
installing it in some way libMesh/PETSc couldn't understand.  When I
installed it by hand I got both PETSc and libMesh to install without having
to disable MPI.  I also ran make check on libMesh and all of the tests have
passed.

However, I still get a runtime error when I initialize libMesh.  Here is
the code:

#include "libmesh/libmesh.h"

int main(int argc, char** argv) {
  libMesh::LibMeshInit init(argc,argv);
  return 0;
}

and I get the (runtime) error:

[Andrews-MacBook-Pro.local:10106] [[INVALID],INVALID] ORTE_ERROR_LOG:
Unknown error: -1 in file runtime/orte_globals.c at line 218
[Andrews-MacBook-Pro:10106] *** Process received signal ***
[Andrews-MacBook-Pro:10106] Signal: Segmentation fault: 11 (11)
[Andrews-MacBook-Pro:10106] Signal code: Address not mapped (1)
[Andrews-MacBook-Pro:10106] Failing at address: 0x0
[Andrews-MacBook-Pro:10106] [ 0] 2   libsystem_c.dylib
0x00007fff972f894a _sigtramp + 26
[Andrews-MacBook-Pro:10106] [ 1] 3   ???
0x0000000000000001 0x0 + 1
[Andrews-MacBook-Pro:10106] *** End of error message ***
Segmentation fault: 11

Do you know what could be causing this?  The error message means very
little to me.   I have attached the configure log for libMesh.

Thanks again!
Andy






On Thu, Jun 6, 2013 at 4:55 PM, Jed Brown <je...@59...> wrote:

> Andrew Davis <and...@gm...> writes:
>
> > This does look like some kind of MPI issue.  Just to be safe I tried
> > reinstalling MPI and then reconfigure/compiling PETSc and libMesh.
>
> Exactly which options are you configuring with?  This message makes it
> sound like you have an MPI-enabled PETSc and a libMesh configured
> without MPI.
>
> > This resulted in the same runtime error (the stack trace) as before.
>
> Which one?  In MPI_Attr_get?  Can you attach a debugger and find out
> what out what the value of comm was?  MPIUNI's implementation is simple:
>
> int MPI_Attr_get(MPI_Comm comm,int keyval,void *attribute_val,int *flag)
> {
>   if (comm-1 < 0 || comm-1 > 3) return 1;
>   if (!keyval) Keyval_setup();
>   *flag                  = attr[comm-1][keyval].active;
>   *(void**)attribute_val = attr[comm-1][keyval].attribute_val;
>   return MPI_SUCCESS;
> }
>
> It appears that libMesh::COMM_WORLD is being set to 0 when not using
> MPI?
>

Re: [Libmesh-users] Runtime Libmesh issue

[Paul T. B. <ptb...@gm...>]

ORTE is OpenMPI stuff I think. Make sure you're running the same MPI that
you used to build libMesh and PETSc.



On Thu, Jun 13, 2013 at 3:48 PM, Andrew Davis <and...@gm...> wrote:

> Hi,
>
> Sorry for the delay getting back do you on this.  I think I managed to
> solve some of my issues but haven't quite figured it out exactly.  I think
> this is at least a different issue than before.
>
> Previously, I had installed MPI with Homebrew which was apparently
> installing it in some way libMesh/PETSc couldn't understand.  When I
> installed it by hand I got both PETSc and libMesh to install without having
> to disable MPI.  I also ran make check on libMesh and all of the tests have
> passed.
>
> However, I still get a runtime error when I initialize libMesh.  Here is
> the code:
>
> #include "libmesh/libmesh.h"
>
> int main(int argc, char** argv) {
>   libMesh::LibMeshInit init(argc,argv);
>   return 0;
> }
>
> and I get the (runtime) error:
>
> [Andrews-MacBook-Pro.local:10106] [[INVALID],INVALID] ORTE_ERROR_LOG:
> Unknown error: -1 in file runtime/orte_globals.c at line 218
> [Andrews-MacBook-Pro:10106] *** Process received signal ***
> [Andrews-MacBook-Pro:10106] Signal: Segmentation fault: 11 (11)
> [Andrews-MacBook-Pro:10106] Signal code: Address not mapped (1)
> [Andrews-MacBook-Pro:10106] Failing at address: 0x0
> [Andrews-MacBook-Pro:10106] [ 0] 2   libsystem_c.dylib
> 0x00007fff972f894a _sigtramp + 26
> [Andrews-MacBook-Pro:10106] [ 1] 3   ???
> 0x0000000000000001 0x0 + 1
> [Andrews-MacBook-Pro:10106] *** End of error message ***
> Segmentation fault: 11
>
> Do you know what could be causing this?  The error message means very
> little to me.   I have attached the configure log for libMesh.
>
> Thanks again!
> Andy
>
>
>
>
>
>
> On Thu, Jun 6, 2013 at 4:55 PM, Jed Brown <je...@59...> wrote:
>
> > Andrew Davis <and...@gm...> writes:
> >
> > > This does look like some kind of MPI issue.  Just to be safe I tried
> > > reinstalling MPI and then reconfigure/compiling PETSc and libMesh.
> >
> > Exactly which options are you configuring with?  This message makes it
> > sound like you have an MPI-enabled PETSc and a libMesh configured
> > without MPI.
> >
> > > This resulted in the same runtime error (the stack trace) as before.
> >
> > Which one?  In MPI_Attr_get?  Can you attach a debugger and find out
> > what out what the value of comm was?  MPIUNI's implementation is simple:
> >
> > int MPI_Attr_get(MPI_Comm comm,int keyval,void *attribute_val,int *flag)
> > {
> >   if (comm-1 < 0 || comm-1 > 3) return 1;
> >   if (!keyval) Keyval_setup();
> >   *flag                  = attr[comm-1][keyval].active;
> >   *(void**)attribute_val = attr[comm-1][keyval].attribute_val;
> >   return MPI_SUCCESS;
> > }
> >
> > It appears that libMesh::COMM_WORLD is being set to 0 when not using
> > MPI?
> >
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Windows:
>
> Build for Windows Store.
>
> http://p.sf.net/sfu/windows-dev2dev
> _______________________________________________
> Libmesh-users mailing list
> Lib...@li...
> https://lists.sourceforge.net/lists/listinfo/libmesh-users
>
>

Re: [Libmesh-users] Runtime Libmesh issue

[Andrew D. <and...@gm...>]

Hi,

Yes! Sorry that was a silly mistake of mine in the cmake file; I was
pointing to the wrong MPI.  It looks like everything is working now.

Thanks for all your help!
Andrew


On Thu, Jun 13, 2013 at 2:20 PM, Paul T. Bauman <ptb...@gm...> wrote:

> ORTE is OpenMPI stuff I think. Make sure you're running the same MPI that
> you used to build libMesh and PETSc.
>
>
>
> On Thu, Jun 13, 2013 at 3:48 PM, Andrew Davis <and...@gm...>wrote:
>
>> Hi,
>>
>> Sorry for the delay getting back do you on this.  I think I managed to
>> solve some of my issues but haven't quite figured it out exactly.  I think
>> this is at least a different issue than before.
>>
>> Previously, I had installed MPI with Homebrew which was apparently
>> installing it in some way libMesh/PETSc couldn't understand.  When I
>> installed it by hand I got both PETSc and libMesh to install without
>> having
>> to disable MPI.  I also ran make check on libMesh and all of the tests
>> have
>> passed.
>>
>> However, I still get a runtime error when I initialize libMesh.  Here is
>> the code:
>>
>> #include "libmesh/libmesh.h"
>>
>> int main(int argc, char** argv) {
>>   libMesh::LibMeshInit init(argc,argv);
>>   return 0;
>> }
>>
>> and I get the (runtime) error:
>>
>> [Andrews-MacBook-Pro.local:10106] [[INVALID],INVALID] ORTE_ERROR_LOG:
>> Unknown error: -1 in file runtime/orte_globals.c at line 218
>> [Andrews-MacBook-Pro:10106] *** Process received signal ***
>> [Andrews-MacBook-Pro:10106] Signal: Segmentation fault: 11 (11)
>> [Andrews-MacBook-Pro:10106] Signal code: Address not mapped (1)
>> [Andrews-MacBook-Pro:10106] Failing at address: 0x0
>> [Andrews-MacBook-Pro:10106] [ 0] 2   libsystem_c.dylib
>> 0x00007fff972f894a _sigtramp + 26
>> [Andrews-MacBook-Pro:10106] [ 1] 3   ???
>> 0x0000000000000001 0x0 + 1
>> [Andrews-MacBook-Pro:10106] *** End of error message ***
>> Segmentation fault: 11
>>
>> Do you know what could be causing this?  The error message means very
>> little to me.   I have attached the configure log for libMesh.
>>
>> Thanks again!
>> Andy
>>
>>
>>
>>
>>
>>
>> On Thu, Jun 6, 2013 at 4:55 PM, Jed Brown <je...@59...> wrote:
>>
>> > Andrew Davis <and...@gm...> writes:
>> >
>> > > This does look like some kind of MPI issue.  Just to be safe I tried
>> > > reinstalling MPI and then reconfigure/compiling PETSc and libMesh.
>> >
>> > Exactly which options are you configuring with?  This message makes it
>> > sound like you have an MPI-enabled PETSc and a libMesh configured
>> > without MPI.
>> >
>> > > This resulted in the same runtime error (the stack trace) as before.
>> >
>> > Which one?  In MPI_Attr_get?  Can you attach a debugger and find out
>> > what out what the value of comm was?  MPIUNI's implementation is simple:
>> >
>> > int MPI_Attr_get(MPI_Comm comm,int keyval,void *attribute_val,int *flag)
>> > {
>> >   if (comm-1 < 0 || comm-1 > 3) return 1;
>> >   if (!keyval) Keyval_setup();
>> >   *flag                  = attr[comm-1][keyval].active;
>> >   *(void**)attribute_val = attr[comm-1][keyval].attribute_val;
>> >   return MPI_SUCCESS;
>> > }
>> >
>> > It appears that libMesh::COMM_WORLD is being set to 0 when not using
>> > MPI?
>> >
>>
>>
>> ------------------------------------------------------------------------------
>> This SF.net email is sponsored by Windows:
>>
>> Build for Windows Store.
>>
>> http://p.sf.net/sfu/windows-dev2dev
>>
>> _______________________________________________
>> Libmesh-users mailing list
>> Lib...@li...
>> https://lists.sourceforge.net/lists/listinfo/libmesh-users
>>
>>
>