[molpro-user] Molpro 2010.1.24 Intel compiler (+MKL) 12.1 MPP problem

Didier Rebeix didier.rebeix at u-bourgogne.fr
Wed Jun 27 12:38:12 BST 2012


Hi Molpro users,

I try to build Molpro 2010.1.24 using Intel compiler (+MKL) 12.1.
Serial version is fine but parallel version with openmpi 1.6
does not run properly.

My configure options are :

./configure -openmpi -noopenmp -mpp -ifort -icc -blaspath $MKLROOT/lib/intel64 -mppbase /opt/openmpi/1.6-2/intel-12.1/include

Using the given au2o_optdftecp2.test test job I get the following 
ouput :


 Primary working directories    : /tmp/uuuuuuuu
 Secondary working directories  : /tmp/uuuuuuuu
 Wavefunction directory         : /uuuuuuuu/uuuuuuuu/uuuuuuuu/wfu/
 Main file repository           : /tmp/uuuuuuuu/

 ARCHNAME  : Linux/x86_64
 FC        : /usr/local/intel/composer_xe_2011_sp1/bin/ifort
 FCVERSION : 12.1.3
 BLASLIB   : -L/usr/local/intel/composer_xe_2011_sp1/mkl/lib/intel64 -lmkl_intel_ilp64 -lmkl_sequential -lmkl_core
 id        : bou

 Nodes                  nprocs
 xxxxxxx.u-bourgogne.fr    3
 Number of processes for MPI-2 version of Molpro:   nprocs(total)=    4   nprocs(compute)=    3   nprocs(helper)=    1
 ga_uses_ma=false, calling ma_init with nominal heap.
 GA-space will be limited to   8.0 MW (determined by -G option)

 Using default tuning parameters: mindgm=1; mindgv=20; mindgc=4; mindgr=1; noblas=0; mincuda=1000; minvec=7
 default implementation of scratch files=sf  

  
 memory,2,m;
 geometry={
           au1
           o,au1,r
           q,o,r,au1,theta/2
           au2,o,r,au1,theta,q,0}
  
 ierr=0
  
 r=4.15
 theta=106
  
 geometry={
           au1
           o,au1,r
           au2,o,r,au1,theta}
  
 basis={
 ecp,Au,ECP1
 spd,Au,ECP1;C
 spd,o,6-31g*
 }
 {ks,b3lyp;wf,symmetry=1}
 optg,gradient=1.d-4
  
 ks,b3lyp
 eopt(1)=energy
 pop
 show,atcharge
 if(atcharge(1).ne.atcharge(3)) ierr=1
 chgmul_au(1)=atcharge(1)
 chgmul_o(1)=atcharge(2)
 dma
 show,atcharge
 if(atcharge(1).ne.atcharge(3)) ierr=1
 chgdma_au(1)=atcharge(1)
 chgdma_o(1)=atcharge(2)
 ropt(1)=r
 topt(1)=theta
  
 ecp=['ECP1']
 show,eopt,ropt,topt,chgmul_o,chgmul_au,chgdma_o,chgdma_au
  
 EOLD             =      -144.31176840 AU
 ROLD             =         4.14681183 BOHR
 TOLD             =       106.43371576 DEGREE
 CHGMUL_O_OLD         =        -0.51552030
 CHGMUL_AU_OLD        =         0.25776015
 CHGDMA_O_OLD         =        -0.72826760
 CHGDMA_AU_OLD        =         0.36413380
  
  
 de=max(abs(eopt-eold))
 dr=max(abs(ropt-rold))
 dt=max(abs(topt-told))
 dc=max(abs(chgmul_o-chgmul_o_old))+max(abs(chgmul_au-chgmul_au_old))
 dd=max(abs(chgdma_o-chgdma_o_old))+max(abs(chgdma_au-chgdma_au_old))
  
 if(de.gt.1.d-7.or.dr.gt.1.d-4.or.dt.gt.1.d-3.or.dc.gt.1.d-5.or.dd.gt.1.d-5) ierr=1
  
 if(ierr.eq.0) then
 table,ecp,eopt,ropt,topt,chgmul_o,chgmul_au,chgdma_o,chgdma_au
 save,test.log
 title,Results for job au2o_optdftecp2.test
 title,No errors detected.
 else
 table,ecp,eopt,eopt-eold,ropt,topt,chgmul_o,chgmul_au,chgdma_o,chgdma_au
 save,test.log
 title,Results for job au2o_optdftecp2.test
 title,ERRORS DETECTED. Error for r=$dr, error for theta=$dt, error for energy=$de
 title,error for Mulliken charges dc=$dc, error for dma charges dd=$dd
 endif

 Variables initialized (660), CPU time= 0.02 sec
 Commands  initialized (462), CPU time= 0.02 sec, 486 directives.
 Default parameters read. Elapsed time= 0.13 sec

 Checking input...
 Passed
1


                                         ***  PROGRAM SYSTEM MOLPRO  ***
                         Copyright, University College Cardiff Consultants Limited, 2008

                                    Version 2010.1 linked 26 jun 2012 17:29:34


 **********************************************************************************************************************************
 LABEL *                                                                                 
 Linux-2.6.18-274.3.1.el5/xxxxxxxx.u-bourgogne.fr(x86_64) 64 bit mpp version             DATE: 27-Jun-12          TIME: 13:30:20  
 **********************************************************************************************************************************

 SHA1:             c82fc9886d3ef081defdb91dc08d0bb97a8ad3e0
 **********************************************************************************************************************************
 SETTING IERR           =         0.00000000                                  
 SETTING R              =         4.15000000                                  
 SETTING THETA          =       106.00000000                                  

 Variable memory set to    2000000 words,  buffer space   230000 words

 SETTING BASIS          =    USERDEF


 Recomputing integrals since basis changed


 Using spherical harmonics

 Library entry Au   ECP ECP1                 selected for group  1
 Library entry AU     S ECP1                 selected for orbital group  1
 Library entry AU     P ECP1                 selected for orbital group  1
 Library entry AU     D ECP1                 selected for orbital group  1
 Library entry O      S 6-31G                selected for orbital group  2
 Library entry O      P 6-31G                selected for orbital group  2
 Library entry O      D 6-31G*               selected for orbital group  2

1PROGRAM * SEWARD (Integral evaluation for generally contracted gaussian basis sets)     Author: Roland Lindh, 1990

 Geometry written to block  1 of record 700

 Orientation using atomic masses  
 Molecule type: Asymmetric top,     Order of principal axis = 2
 Symmetry elements: X,Y
 Rotational constants:       0.4080178     18.8182340      0.4170606 GHz  (calculated with average atomic masses)

 Point group  C2v 



 ATOMIC COORDINATES

 NR  ATOM    CHARGE       X              Y              Z

   1  AU1    11.00    0.000000000   -3.314337367    0.097477094
   2  AU2    11.00    0.000000000    3.314337367    0.097477094
   3  O       8.00    0.000000000    0.000000000   -2.400055252

 Bond lengths in Bohr (Angstrom)

 1-2  6.628674733  1-3  4.150000000  2-3  4.150000000
     ( 3.507743595)     ( 2.196085417)     ( 2.196085417)

 Bond angles

  1-2-3   37.00000000   1-3-2  106.00000000   2-1-3   37.00000000

 NUCLEAR CHARGE:                   30
 NUMBER OF PRIMITIVE AOS:          88
 NUMBER OF SYMMETRY AOS:           81
 NUMBER OF CONTRACTIONS:           45   (  22A1  +   8B1  +  11B2  +   4A2  )
 NUMBER OF CORE ORBITALS:           1   (   1A1  +   0B1  +   0B2  +   0A2  )
 NUMBER OF VALENCE ORBITALS:       16   (   6A1  +   3B1  +   5B2  +   2A2  )


 NUCLEAR REPULSION ENERGY   60.66366441


 Eigenvalues of metric

         1 0.287E-04 0.356E-02 0.434E-01 0.928E-01 0.166E+00 0.209E+00 0.389E+00 0.497E+00
         2 0.168E+00 0.494E+00 0.931E+00 0.997E+00 0.103E+01 0.112E+01 0.151E+01 0.218E+01
         3 0.125E+00 0.223E+00 0.441E+00 0.632E+00 0.880E+00 0.958E+00 0.992E+00 0.106E+01
         4 0.577E+00 0.944E+00 0.993E+00 0.105E+01


 Contracted 2-electron integrals neglected if value below      1.0D-12
 iBas=                    81
 iBas_Aux=                     0
 iSO=                    27
 iSO_Aux=                    54
 iSO_Tot=                    27

 STACK TRACE:

 SEWARD                            0.00      0.00
  INPUT                            0.00      0.00
   SOCTL                           0.00      0.00
xquit called with rc= 128

 GLOBAL ERROR fehler on processor   0                                         
    0: fehler 307 (0x133).
    1: fehler 307 (0x133).
    1: In mpi_utils.c [MPIGA_Error]: now exiting...
    0: In mpi_utils.c [MPIGA_Error]: now exiting...
    2: fehler 307 (0x133).
    2: In mpi_utils.c [MPIGA_Error]: now exiting...
    1: fehler 307 (0x133).
    0: fehler 307 (0x133).
    2: fehler 307 (0x133).
--------------------------------------------------------------------------
MPI_ABORT was invoked on rank 0 in communicator MPI_COMM_WORLD 
with errorcode 307.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------
forrtl: error (78): process killed (SIGTERM)
Image              PC                Routine            Line        Source             
mca_btl_sm.so      00002AF7284C1949  Unknown               Unknown  Unknown
libmpi.so.1        00002AF72224DD66  Unknown               Unknown  Unknown
mca_pml_ob1.so     00002AF7277421D1  Unknown               Unknown  Unknown
libmpi.so.1        00002AF722198D2C  Unknown               Unknown  Unknown
molpro.exe         0000000002B62DB1  Unknown               Unknown  Unknown
molpro.exe         0000000002B62B27  Unknown               Unknown  Unknown
molpro.exe         0000000002B5A5C8  Unknown               Unknown  Unknown
molpro.exe         000000000174DA36  Unknown               Unknown  Unknown
molpro.exe         0000000001745A4B  Unknown               Unknown  Unknown
molpro.exe         000000000056F17F  Unknown               Unknown  Unknown
molpro.exe         00000000004D4E15  Unknown               Unknown  Unknown
molpro.exe         00000000004D4DBC  Unknown               Unknown  Unknown
libc.so.6          00000037B601D8B4  Unknown               Unknown  Unknown
molpro.exe         00000000004D4CC9  Unknown               Unknown  Unknown
--------------------------------------------------------------------------
mpirun has exited due to process rank 1 with PID 10263 on
node xxxxxxxx.u-bourgogne.fr exiting improperly. There are two reasons this could occur:

1. this process did not call "init" before exiting, but others in
the job did. This can cause a job to hang indefinitely while it waits
for all processes to call "init". By rule, if one process calls "init",
then ALL processes must call "init" prior to termination.

2. this process called "init", but exited without calling "finalize".
By rule, all processes that call "init" MUST call "finalize" prior to
exiting or it will be considered an "abnormal termination"

This may have caused other processes in the application to be
terminated by signals sent by mpirun (as reported here).
--------------------------------------------------------------------------
[xxxxxxxx.u-bourgogne.fr:10261] 2 more processes have sent help message help-mpi-api.txt / mpi-abort
[xxxxxxxx.u-bourgogne.fr:10261] Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error messages

Any clue greatly appreciated !

Thanks.

--
Didier REBEIX
Universite de Bourgogne




More information about the Molpro-user mailing list