[mcstas-users] openmpi compilation problem

Fri Jun 15 13:24:34 CEST 2012

Hi Emmanuel and Peter!

thanks for the answers!
I did a fresh install of openmpi after removing mpich2.
I had to give the config...perl the location of the mpicc (mpirun was there already)
Runs without trouble.

I am now trying to install openvrml which apparently depends on a libjs which is not in repos...
any better alternative ? (there seems to be a way out which probably will require some google work, I will try that asap).

Best and thanks again, have a nice we (Sun is shining on this side of the Alps ;0) 
JF
-
Helmholtz Zentrum Geesthacht
Institut für Werkstoffforschung
Abteilung WPN, Instrument REFSANS
Lichtenbergstr. 1
85747 Garching FRM II
Tel.: +49 (0)89 289 10762
Internet: http://www.frm2.tum.de

-----mcstas-users-bounces at mcstas.org schrieb: -----
An: mcstas-users at mcstas.org
Von: Emmanuel FARHI 
Gesendet von: mcstas-users-bounces at mcstas.org
Datum: 06/14/2012 06:03PM
Betreff: Re: [mcstas-users] openmpi compilation problem

Yes, you probably have conflicts between the two MPI packages. I would 
recommend to suppress MPICH2, which requires a daemon to run 'mpd'. Just 
remoce the package, and any other mpich* libmpich* stuff. Only openmpi 
should remain (openmpi openmpi-dev libopenmpi).

Then, to make sure all is fine, try:
sudo update-alternatives --config mpicc
sudo update-alternatives --config mpirun

Emmanuel.

On 06/14/2012 01:49 PM, Jean-Francois.Moulin at hzg.de wrote:
>   Hi Emmanuel!
>
> thanks for the answer. Yes the build works but, this is what I get when I try to use it:
>
> (case A)
> ~/Lab/MCarlo/McStas$ mcgui refsans_simplified_0.instr
> ./refsans_simplified_0.out: error while loading shared libraries: libmpich.so.3: cannot open shared object file: No such file or directory
>
> And this is correct, the file is not on my system... why did the configure look for mpich in the first place?
>
> (caseB)
> Well it wants MPICH? I tried and installed mpich2 from ubuntu repos (including dev), ./configure make make install...
> mcgui now starts without problem and shows my instrument parameters (it did not in case A)
> but the simulation returns errors...
>
> mcrun: No MPI/grid machine list. Running locally.
>    Define /home/jfmoulin/.mcstas/hosts
>    or /usr/local/lib/mcstas/tools/perl/hosts
>    or use option --machines=<file>
> [jfm-lap:11887] [[INVALID],INVALID] ORTE_ERROR_LOG: A system-required executable either could not be found or was not executable by this user in file ../../../../../../orte/mca/ess/singleton/ess_singleton_module.c at line 357
> [jfm-lap:11887] [[INVALID],INVALID] ORTE_ERROR_LOG: A system-required executable either could not be found or was not executable by this user in file ../../../../../../orte/mca/ess/singleton/ess_singleton_module.c at line 230
> [jfm-lap:11887] [[INVALID],INVALID] ORTE_ERROR_LOG: A system-required executable either could not be found or was not executable by this user in file ../../../orte/runtime/orte_init.c at line 132
> --------------------------------------------------------------------------
> It looks like orte_init failed for some reason; your parallel process is
> likely to abort.  There are many reasons that a parallel process can
> fail during orte_init; some of which are due to configuration or
> environment problems.  This failure appears to be an internal failure;
> here's some additional information (which may only be relevant to an
> Open MPI developer):
>
>    orte_ess_set_name failed
>    -->  Returned value A system-required executable either could not be found or was not executable by this user (-127) instead of ORTE_SUCCESS
>
>
> Is there a reason to have conflicts between the two mpi (I have dependencies of other packages on openmpi) or do I forget a setup step?
> I had the box running before and did not change my home dir so I guess it should still work
> Anyway, thanks for your help!
>
> Ciao
> JF
> -
> Helmholtz Zentrum Geesthacht
> Institut für Werkstoffforschung
> Abteilung WPN, Instrument REFSANS
> Lichtenbergstr. 1
> 85747 Garching FRM II
> Tel.: +49 (0)89 289 10762
> Internet: http://www.frm2.tum.de
>
> -----Emmanuel FARHI<farhi at ill.eu>  schrieb: -----
> An: Jean-Francois.Moulin at hzg.de, "McStas users list at mcstas.org"<mcstas-users at mcstas.org>
> Von: Emmanuel FARHI<farhi at ill.eu>
> Datum: 06/14/2012 10:22AM
> Betreff: Re: [mcstas-users] openmpi compilation problem
>
> Hi Jean-Francois,
>
> to my knowledge, this is just a warning. The tar.gz configure; make;
> make install should proceed OK still, and McStas should be able to find
> MPI if 'mpicc' and 'mpirun' commands are available at the prompt.
>
> OpenMPI seems OK on recent systems. It was less reliable than MPICH
> until 2010-2011 (would stall for long simulations with e.g. ncount>  2e9).
>
> Emmanuel.
>
> On 06/13/2012 05:36 PM, Jean-Francois.Moulin at hzg.de wrote:
>> Hi to all!
>>
>> new Linux distro, new problem! On Linux Mint 13 I do not manage to install due to mpi problems. As already suggested some months ago I installed openmpi (packages from Ubuntu repos: libopenmpi 1.4.3.2 and its dev).
>> The ./configure tells me that :
>> checking mpi.h presence... no
>> checking for mpi.h... no
>> WARNING: /usr/bin/mpicc may fail to compile without mpi.h
>>
>> but I find this on my system:
>> ./usr/lib/openmpi/include/mpi.h
>> After reading the output of configure -h, I tried
>>    ./configure -I/usr/lib/openmpi/include/:
>> configure: error: unrecognized option: `-I'
>> I tried putting this path in the cppflags envirt variable. To no avail.
>> seems I miss the point.
>> Since I get a Makefile I looked for an include section where I could manually put the path but that does not look so easy... (for me)
>>
>> second question I am on a 64 bit machine, I should thus configure --with-pic right?
>>
>> Thanks in advance for your help!
>>
>> Best
>>
>> Jean-Francois
>>
>>
>> -
>> Helmholtz Zentrum Geesthacht
>> Institut für Werkstoffforschung
>> Abteilung WPN, Instrument REFSANS
>> Lichtenbergstr. 1
>> 85747 Garching FRM II
>> Tel.: +49 (0)89 289 10762
>> Internet: http://www.frm2.tum.deHelmholtz-Zentrum Geesthacht
>> Zentrum für Material- und Küstenforschung GmbH
>> Max-Planck-Straße 1 I 21502 Geesthacht I Deutschland/Germany
>>
>> Geschäftsführer/Board of Management: Prof. Dr. Wolfgang Kaysser, Dipl.-Ing. Michael Ganß
>> Vorsitzender des Aufsichtsrates/Chairman of the Supervisory Board: MinDirig Wilfried Kraus
>> Amtsgericht Lübeck HRB 285 GE (Register Court)
>> Internet: http://www.hzg.de
>> _______________________________________________
>> mcstas-users mailing list
>> mcstas-users at mcstas.org
>> http://mailman.mcstas.org/cgi-bin/mailman/listinfo/mcstas-users

-- 
Emmanuel FARHI,www.ill.eu/computing/people/emmanuel-farhi \|/ ____ \|/
CS-Group ILL4/221, Institut Laue-Langevin (ILL) Grenoble  ~@-/ oO \-@~
6 rue J. Horowitz, BP 156, 38042 Grenoble Cedex 9,France  /_( \__/ )_\
Work :Tel (33/0) 4 76 20 71 35. Fax (33/0) 4 76 48 39 06     \__U_/

_______________________________________________
mcstas-users mailing list
mcstas-users at mcstas.org
http://mailman.mcstas.org/cgi-bin/mailman/listinfo/mcstas-users
Helmholtz-Zentrum Geesthacht 
Zentrum für Material- und Küstenforschung GmbH 
Max-Planck-Straße 1 I 21502 Geesthacht I Deutschland/Germany 

Geschäftsführer/Board of Management: Prof. Dr. Wolfgang Kaysser, Dipl.-Ing. Michael Ganß 
Vorsitzender des Aufsichtsrates/Chairman of the Supervisory Board: MinDirig Wilfried Kraus 
Amtsgericht Lübeck HRB 285 GE (Register Court) 
Internet: http://www.hzg.de  