[molpro-user] SVD failing on cray xc40 (UNCLASSIFIED)

Andy May MayAJ1 at cardiff.ac.uk
Tue Apr 14 16:05:18 BST 2015


James,

I just tried on swan.cray.com and found that there are error messages 
during configure.

The reason is that the -v option to ifort, on swan.cray.com at least, 
gives output with lots of extra garbage that meant configure didn't 
parse it correctly. If that happened on your system it will mean that 
configure didn't determine the ifort version correctly, so some files 
that should have been deoptimized were not.

I've made a commit to fix it, at least for that system, but the process 
of parsing that output is not robust so it could break again. In our 
master version we've changed to using the compiler CPP predefined macros 
to determine the version, so this problem will disappear one day. If you 
wait 24 hours then download the 'nightly' tarball it should contain the fix.

I then built with Intel 15.0.1 compilers with:

module swap PrgEnv-cray PrgEnv-intel
cd /lus/scratch/${LOGNAME}
tar -xzf molpro.tar.gz
cd Molpro
./configure -batch -mpp -mppbase ${MPICH_DIR}/include
make

For reference I attach the resulting CONFIG file. I then used the 
following job script to run:

#!/bin/sh
#PBS -q large
#PBS -l walltime=00:30:00
#PBS -j oe
#PBS -N Molpro

export TMPDIR="/lus/scratch/${LOGNAME}/Molpro/tmp"
wfudir="/lus/scratch/${LOGNAME}/Molpro/wfu"
mkdir ${wfudir} ${TMPDIR}

cd /lus/scratch/${LOGNAME}/Molpro
make -C lib MOLPRO_OPTIONS="-W ${wfudir}" init.log
make -C examples MOLPRO_OPTIONS="-n8 -W ${wfudir}" butane_opt_transition.out

which ran as expected.

Best wishes,

Andy

On 09/02/15 19:21, Ianni, James C CTR (US) wrote:
> Classification: UNCLASSIFIED
> Caveats: NONE
>
> I have Molpro compiled with Intel 15.0.1 compiler with cray-mpich/7.1.0 on a
> Cray XC40. When I run the butane_opt_transition.com example (or any other
> example) molpro on two nodes under PBS, molpro starts up, runs for a little
> bit, then I see SVD errors and molpro errors out. Here's my run line:
> TMPD=a directory on a lustre mount that all nodes can see
> LIBMOL points to ./lib/ that does exists with MOLPRO libs
> molpro -N 16 -n 8 -d ${TMPD} -L ${LIBMOL} ${input} >& molpro.out
>
> Here is the output:
>
> cat butane_opt_transition.out
>
>   Primary working directories    : /work1/ianni/187
>   Secondary working directories  : /work1/ianni/187
>   Wavefunction directory         : /usr/people/ianni/wfu/
>   Main file repository           : /work1/ianni/187/
>
>   SHA1      : xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
>   NAME      : 2012.1.20
>   ARCHNAME  : Linux/XT
>   FC        : ftn
>   FCVERSION : 15.0
>   BLASLIB   :
>   id        : xxxxxx
>
>   Nodes     nprocs
>   nid00016     7
>   Number of processes for MPI-2 version of Molpro:   nprocs(total)=    8
> nprocs(compute)=    7   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 customized tuning parameters: mindgm=1; mindgv=20; mindgc=4;
> mindgr=1; noblas=0; minvec=7
>   default implementation of scratch files=sf
>
>    ***, Bicyclo-[1.1.0]-butane Transition State
>    memory,1,m
>    basis=3-21G
>
>    L1=1.495  ang
> ! Define Active Variables
>    L2=1.418  ang
>    L3=1.463  ang
>    L4=1.093  ang
>    L5=1.111  ang
>    L6=1.098  ang
>    L7=1.097  ang
>    L8=1.110  ang
>    L9=1.106  ang
>    A1=92.1   degree
>    A2=62.1   degree
>    A3=136.0  degree
>    A4=123.5  degree
>    A5=122.4  degree
>    A6=124.7  degree
>    A7=126.7  degree
>    A8=117.9  degree
>    D1=-120.4 degree
>    D2=4.4    degree
>    D3=108.8  degree
>    D4=-107.5 degree
>    D5=84.2   degree
>    D6=109.3  degree
>    D7=-106.1 degree
>
>    geometry={C1
> ! Geometry Specification Z-Matrix
>              C2  1 L1
>              C3  2 L2 1 A1
>              C4  1 L3 2 A2 3 D1
>              H5  1 L4 2 A3 3 D2
>              H6  2 L5 1 A4 4 D3
>              H7  3 L6 2 A5 1 D4
>              H8  3 L7 2 A6 1 D5
>              H9  4 L8 1 A7 2 D6
>              H10 4 L9 1 A8 2 D7}
>    int
>    rhf
>    optg
>    root,2
> ! Transition State search
>    method,qsd
> ! Use Quadratic Steepest Descent Method
>
>   Variables initialized (782), CPU time= 0.00 sec
>   Commands  initialized (547), CPU time= 0.01 sec, 517 directives.
>   Default parameters read. Elapsed time= 0.14 sec
>
>   Checking input...
>   Passed
> 1
>
>
>                                           ***  PROGRAM SYSTEM MOLPRO  ***
>                           Copyright, University College Cardiff Consultants
> Limited, 2008
>
>                                      Version 2012.1 linked Feb  9 2015
> 11:53:33
>
>
>
> ****************************************************************************
> ******************************************************
>   LABEL *
>   Linux-3.0xxxxxx-cray_xxxxx/nid00016(x86_64) 64 bit mpp version
> DATE: 09-Feb-15          TIME: 12:56:10
>
> ****************************************************************************
> ******************************************************
>
>   SHA1:             xxxxxxxxxxxxxxxxxxxxxxxxxxx
>
> ****************************************************************************
> ******************************************************
>
>   Variable memory set to    1000000 words,  buffer space   230000 words
>
>   SETTING BASIS          =    3-21G
>   SETTING L1             =         1.49500000  ANG
>   SETTING L2             =         1.41800000  ANG
>   SETTING L3             =         1.46300000  ANG
>   SETTING L4             =         1.09300000  ANG
>   SETTING L5             =         1.11100000  ANG
>   SETTING L6             =         1.09800000  ANG
>   SETTING L7             =         1.09700000  ANG
>   SETTING L8             =         1.11000000  ANG
>   SETTING L9             =         1.10600000  ANG
>   SETTING A1             =        92.10000000  DEGREE
>   SETTING A2             =        62.10000000  DEGREE
>   SETTING A3             =       136.00000000  DEGREE
>   SETTING A4             =       123.50000000  DEGREE
>   SETTING A5             =       122.40000000  DEGREE
>   SETTING A6             =       124.70000000  DEGREE
>   SETTING A7             =       126.70000000  DEGREE
>   SETTING A8             =       117.90000000  DEGREE
>   SETTING D1             =      -120.40000000  DEGREE
>   SETTING D2             =         4.40000000  DEGREE
>   SETTING D3             =       108.80000000  DEGREE
>   SETTING D4             =      -107.50000000  DEGREE
>   SETTING D5             =        84.20000000  DEGREE
>   SETTING D6             =       109.30000000  DEGREE
>   SETTING D7             =      -106.10000000  DEGREE
>
>
>   Using spherical harmonics
>
>   Library entry C      S 3-21G                selected for orbital group  1
>   Library entry C      P 3-21G                selected for orbital group  1
>   Library entry H      S 3-21G                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
>   Rotational constants:       7.8042247      7.0211678     17.5553961 GHz
> (calculated with average atomic masses)
>
>   Point group  C1
>
>
>
>   ATOMIC COORDINATES
>
>   NR  ATOM    CHARGE       X              Y              Z
>
>     1  C1      6.00   -1.640579909    0.342559831   -0.904560907
>     2  C2      6.00    0.824024883    0.863916589    0.374239143
>     3  C3      6.00    0.084380924   -0.486625311    2.567271106
>     4  C4      6.00    0.668647928   -0.555867676   -2.130811827
>     5  H5      1.00   -3.419409756   -0.570397278   -0.386446301
>     6  H6      1.00    1.684650173    2.774997780    0.496440331
>     7  H7      1.00   -0.391230728    0.462658513    4.349949947
>     8  H8      1.00    0.249887640   -2.539815746    2.800655226
>     9  H9      1.00    1.295661019   -2.548968182   -2.316052099
>    10  H10     1.00    1.337443926    0.467435373   -3.826045695
>
>   Bond lengths in Bohr (Angstrom)
>
>   1-2  2.825140566  1-3  3.964423717  1-4  2.764669330  1-5  2.065470662  2-3
> 2.679631654
>       ( 1.495000000)     ( 2.097882678)     ( 1.463000000)     ( 1.093000000)
> ( 1.418000000)
>
>   2-4  2.883610535  2-6  2.099485732  3-7  2.074919292  3-8  2.073029566  4-9
> 2.097596006
>       ( 1.525940975)     ( 1.111000000)     ( 1.098000000)     ( 1.097000000)
> ( 1.110000000)
>
>    4-10  2.090037101
>         ( 1.106000000)
>
>   Bond angles
>
>    1-2-3   92.10000000   1-2-4   57.92057221   1-2-6  123.50000000   1-3-2
> 45.40967985
>
>    1-3-7  123.84702362   1-3-8  109.90647887   1-4-2   59.97942779   1-4-9
> 126.70000000
>
>    1- 4-10  117.90000000   2- 1- 3   42.49032015   2- 1- 4   62.10000000   2-
> 1- 5  136.00000000
>
>    2- 3- 7  122.40000000   2- 3- 8  124.70000000   2- 4- 9  121.90052635   2-
> 4-10  116.50736785
>
>    3-1-4   87.53730416   3-1-5   93.58906654   3-2-4  116.61172912   3-2-6
> 121.62027136
>
>    4- 1- 5  133.39075532   4- 2- 6  121.38803225   7- 3- 8  112.00585093   9-
> 4-10  107.33380036
>
>   NUCLEAR CHARGE:                   30
>   NUMBER OF PRIMITIVE AOS:          78
>   NUMBER OF SYMMETRY AOS:           78
>   NUMBER OF CONTRACTIONS:           48   (  48A   )
>   NUMBER OF CORE ORBITALS:           4   (   4A   )
>   NUMBER OF VALENCE ORBITALS:       22   (  22A   )
>
>
>   NUCLEAR REPULSION ENERGY  111.39771670
>
>
>   Eigenvalues of metric
>
>           1 0.118E-01 0.150E-01 0.249E-01 0.364E-01 0.476E-01 0.589E-01
> 0.607E-01 0.655E-01
>
>
>   Contracted 2-electron integrals neglected if value below      1.0D-11
>   AO integral compression algorithm  1   Integral accuracy      1.0D-11
>
>       4.719 MB (compressed) written to integral file ( 67.2%)
>
>       Node minimum: 0.524 MB, node maximum: 0.786 MB
>
>
>   NUMBER OF SORTED TWO-ELECTRON INTEGRALS:      99372.     BUFFER LENGTH:
> 32768
>   NUMBER OF SEGMENTS:   1  SEGMENT LENGTH:      99372      RECORD LENGTH:
> 262144
>
>   Memory used in sort:       0.39 MW
>
>   SORT1 READ      853217. AND WROTE       97357. INTEGRALS IN      1 RECORDS.
> CPU TIME:     0.01 SEC, REAL TIME:     0.03 SEC
>   SORT2 READ      672700. AND WROTE      692076. INTEGRALS IN     14 RECORDS.
> CPU TIME:     0.01 SEC, REAL TIME:     0.06 SEC
>
>   Node minimum:       98364.  Node maximum:       99372. integrals
>
>   OPERATOR DM      FOR CENTER  0  COORDINATES:    0.000000    0.000000
> 0.000000
>
>
>
> ****************************************************************************
> ******************************************************
>   DATASETS  * FILE   NREC   LENGTH (MB)   RECORD NAMES
>                1      18       14.43       500      610      700      900
> 950      970     1000      129      960     1100
>                                            VAR    BASINP    GEOM    SYMINP
> ZMAT    AOBASIS   BASIS     P2S    ABASIS      S
>                                           1400     1410     1200     1210
> 1080     1600     1650     1700
>                                             T        V       H0       H01
> AOSYM     SMH    MOLCAS    OPER
>
>   PROGRAMS   *        TOTAL       INT
>   CPU TIMES  *         0.13      0.04
>   REAL TIME  *         0.48 SEC
>   DISK USED  *        17.85 MB
>   GA USED    *         0.00 MB       (max)       0.00 MB       (current)
>
> ****************************************************************************
> ******************************************************
>
> 1PROGRAM * RHF-SCF (CLOSED SHELL)       Authors: W. Meyer, H.-J. Werner
>
>
>   NUMBER OF ELECTRONS:      15+   15-    SPACE SYMMETRY=1    SPIN SYMMETRY:
> Singlet
>   CONVERGENCE THRESHOLDS:    1.00E-06 (Density)    1.00E-07 (Energy)
>   MAX. NUMBER OF ITERATIONS:       60
>   INTERPOLATION TYPE:            DIIS
>   INTERPOLATION STEPS:              2 (START)      1 (STEP)
>   LEVEL SHIFTS:                  0.00 (CLOSED)  0.00 (OPEN)
>
>
>
>   Orbital guess generated from atomic densities. Full valence occupancy:   15
>
>   Molecular orbital dump at record        2100.2
>
>   ITERATION    DDIFF          GRAD             ENERGY        2-EL.EN.
> DIPOLE MOMENTS         DIIS   ORB.
>      1      0.000D+00      0.000D+00                NaN           NaN
> NaN        NaN        NaN    0    start
>      2      0.000D+00            NaN                NaN           NaN
> NaN        NaN        NaN    1    diag
>   ? Error
>   ? SVD solve failed
>   ? The problem occurs in rmp2f12/LinearSolveGenSvd
>
> ______________________________
>
>
>
> Classification: UNCLASSIFIED
> Caveats: NONE
>
>
>
>
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user
>
-------------- next part --------------
# MOLPRO CONFIG generated at Tue Apr 14 09:08:47 CDT 2015, for host swan, SHA1=89c4732a0b2b3cf0560d69a42b84eb4ee468be9e

CONFIGURE_OPTIONS= "-mpp" "-mppbase" "/opt/cray/mpt/7.2.0/gni/mpich2-intel/140/include"

AR=ar
ARCHNAME=Linux/XT
ARFLAGS=-rS
AWK=awk
BIBTEX=
BLASLIB=
BUILD=p
CAT=cat
CC=/opt/cray/craype/2.3.0/bin/cc
CCVERSION=15.0.1
CC_FRONT=/opt/intel/composer_xe_2015.1.133/bin/intel64/icc
CDEBUG=-g $(addprefix $(CDEFINE),_DEBUG)
CDEFINE=-D
CFLAGS=-ftz
CLDFLAGS=
CLEAN=echo 'target clean only available with git cloned versions, please unpack the tarball again'
CMPPINCLUDE=/opt/cray/mpt/7.2.0/gni/mpich2-intel/140/include
COPT=-O2
COPT0=-O0
COPT1=-O1
COPT2=-O2
COPT3=-O3
CP=cp -p
CPROFILE=-p
CUDACC=
CUDACCVERSION=
CUDACDEBUG=-g $(addprefix $(CUDACDEFINE),_DEBUG)
CUDACDEFINE=-D
CUDACFLAGS=
CUDACOPT=
CUDACOPT0=-O0
CUDACOPT1=-O1
CUDACOPT2=-O2
CUDACOPT3=-O3
CUDACPROFILE=-p
CXX=/opt/cray/craype/2.3.0/bin/CC
CXXFLAGS=$(CFLAGS)
DOXYGEN=/usr/bin/doxygen
ECHO=/bin/echo
EXPORT=export
F90FLAGS=-stand f03
FC=/opt/cray/craype/2.3.0/bin/ftn
FCVERSION=15.0.1
FDEBUG=-g $(addprefix $(FDEFINE),_DEBUG)
FDEFINE=-D
FFLAGS=-i8 -pc64 -auto -warn nousage -cxxlib
FLDFLAGS=
FOPT=-O3
FOPT0=-O0
FOPT1=-O1
FOPT2=-O2
FOPT3=-O3
FPROFILE=-p
FSTATIC=
HOSTFILE_FORMAT=
INSTALL_FILES_EXTRA=
INSTBIN=
INST_PL=0
INTEGER=8
LAPACKLIB=
LATEX2HTML=
LAUNCHER=/opt/cray/alps/5.2.1-2.0502.9041.11.6.ari/bin/aprun -n %n %x
LD_ENV=/opt/cray/craype/2.3.0/compiler/lib/intel64
LD_ENVNAME=LD_LIBRARY_PATH
LIBRARY_SUFFIX=a
LIBS=-lpthread /lus/scratch/${LOGNAME}/Molpro/src/boost-install/lib/libboost_system.a /lus/scratch/${LOGNAME}/Molpro/src/boost-install/lib/libboost_thread.a -lrt
LN=ln -s
MACROS=MOLPRO NDEBUG MOLPRO_f2003 MPI2 HAVE_BOOST_THREADS HAVE_SSE2 _I8_ MOLPRO_INT=8 BLAS_INT=4 LAPACK_INT=4 MOLPRO_AIMS MOLPRO_NECI _MOLCAS_MPP_ MOLPRO_BLAS MOLPRO_LAPACK
MAKEDEPEND_OPTIONS=
MAKEINDEX=
MAPLE=
MAX_INCREMENT_LIBRARY=0
MKDIR=mkdir -p
MODULE_FLAG=-I
MODULE_SUFFIX=mod
MPILIB=
MPPLIB=
OBJECT_SUFFIX=o
OPT0=B88.F copyc6.F
OPT1=explicit_util.F avcc.F koopro4.F dlaed4.F frequencies.F optg.F
OPT2=tstfnc.F dftgrid.F mrf12_singles.F90 basis_integral_shells.F integrals.F90 geminal.F surface.F gcc.F90
OPT3=
PAPER=a4paper
PARSE=parse-Linux-x86_64-i8.o
PDFLATEX=
PNAME=molprop_2012_1_Linux_XT_i8
PREFIX=/usr/local/molpro/molprop_2012_1_Linux_XT_i8
PTSIZE=11
PYTHON=/usr/bin/python
RANLIB=ranlib
RM=rm -rf
SHELL=/bin/sh
STRIP=strip
SUFFIXES=F F90 c cpp
TAR=tar -cf
UNTAR=tar -xf
VERBOSE=@
VERSION=2012.1
XSD=/usr/bin/xmllint --noout --schema
XSLT=/usr/bin/xsltproc
YACC=bison -b y

.SUFFIXES:
MAKEFLAGS+=-r
ifneq ($(LD_ENVNAME),)
$(LD_ENVNAME):=$(LD_ENV):$($(LD_ENVNAME))
endif


More information about the Molpro-user mailing list