[CPMD-list] error while calculating the dipole moment
huawj
huawj at itcc.nju.edu.cn
Sat Oct 21 09:48:05 CEST 2006
Dear Axel,
======= 2006-10-20 23:58:20 you wrote£º=======
>On Fri, 20 Oct 2006, Alexandr Isayev wrote:
>
>AI> Dear Weijie,
>AI> Could you please try this keyword instead. I never had a problem like that.
>AI>
>AI> DIPOLE DYNAMICS WANNIER SAMPLE
>AI> 20
>AI>
>AI> Your keyword combination seems to be incorrect.
>
>don't think so. please look at the code in control.F
>both inputs should have the same result.
>
> ELSE IF (INDEX(LINE,'DIPO').NE.0 .AND.
> * INDEX(LINE,'DYNA').NE.0) THEN
>C..Store Dipole moments
> TDIPD=.TRUE.
> IF(INDEX(LINE,'WANNIER').NE.0) TWANN=.TRUE.
> IF(INDEX(LINE,'SAMPLE').NE.0) THEN
> LINEOLD=LINE
> READ(IUNIT,ERR=99,END=99,FMT=LINEFOR) LINE
> CALL READSI(LINE,1,IE,NPDIP,ERREAD)
> IF(ERREAD) GOTO 99
> ENDIF
> GOTO 10
> ELSE IF (INDEX(LINE,'WANNIER').NE.0 .AND.
Thanks for your reply.I now know why the two inputs are equivalent.
>i would suspect, that this is a miscompile
>that results in memory corruption (perhaps
> a missing -DPOINTER8 or so).
>
>weijie,
>can you please post the config section of your makefile.
The config section of the Makefile on the IBM eServer is as follows:
(the whole Makefile is also attached)
#----------------------------------------------------------------------------
# Makefile for cpmd.x (plane wave electronic calculation)
# Configuration: IFORT-AMD64-MPI
# Creation of Makefile: Oct 10 2006
# on Linux mnode 2.6.9-22.ELsmp #1 SMP Mon Sep 19 18:00:54 EDT 2005 x86_64 x86_64 x86_64 GNU/Linux
# Author: root
#----------------------------------------------------------------------------
#
SHELL = /bin/sh
#
#--------------- Default Configuration for IFORT-AMD64-MPI ---------------
SRC = .
DEST = .
BIN = .
FFLAGS = -pc64 -tpp7 -O2 -unroll
LFLAGS = -L/usr/local/soft/intel/mkl/8.0.2/lib/em64t -lmkl_lapack64 -lmkl_p4n -lguide -lpthread
CFLAGS = -O2 -Wall -m64
CPP = /lib/cpp -P -C -traditional
CPPFLAGS = -D__Linux -D__PGI -DFFT_DEFAULT -DPOINTER8 -DLINUX_IFC \
-DPARALLEL -DMYRINET
NOOPT_FLAG =
CC = /usr/local/topspin/mpi/mpich/bin/mpicc.i
FC = /usr/local/topspin/mpi/mpich/bin/mpif77.i -c
LD = /usr/local/topspin/mpi/mpich/bin/mpif77.i -i-static
AR = ar
#----------------------------------------------------------------------------
# Personal Configuration
#----------------------------------------------------------------------------
#----------------------------------------------------------------------------
# End of Personal Configuration
#----------------------------------------------------------------------------
CFGDEST = .
CFGMACH = IFORT-AMD64-MPI
CFGQMMM =
#
# LIST OF FILES
#
I also met with the same problem when I tried to do the same job on a
SGI platform. So I'd also like to paste the Makefile as an attachment.
(The platform is SGI Altix 330, and the compiler is
Intel(R) Fortran Itanium(R) Compiler for Itanium(R)-based applications
Version 9.1 Build 20060928 Package ID: l_fc_c_9.1.039 )
Thank you very much.
Sincerely yours,
Weijie
>cheers,
> axel.
>
>
>AI>
>AI> Sincerely,
>AI> Alexandr
>AI>
>AI>
>AI>
>AI>
>AI>
>AI> Thursday, October 19, 2006, 10:21:51 PM, you wrote:
>AI>
>AI> h> Dear CPMDers£¬
>AI>
>AI> h> I'm now doing a CP MD for a system of 31 H2O and 1 HCl molecules at 300K.
>AI> h> It runs quite well for a few picoseconds. When I remember to calculate the
>AI> h> dipole moment, so I add the following keywords in the input file and then restart.
>AI> h> DIPOLE DYNAMICS WANNIER
>AI> h> DIPOLE DYNAMICS SAMPLE
>AI> h> 20
>AI>
>AI> h> But the TOTAL INTEGRATED ELECTRONIC DENSITY in the output file becomes zero,
>AI> h> and the EKINEC also becomes zero,
>AI> h> ---------------------------------------------------------------------------------------
>AI> h> TOTAL INTEGRATED ELECTRONIC DENSITY
>AI> h> IN G-SPACE = 0.000000
>AI> h> IN R-SPACE = 0.000000
>AI>
>AI> h> (K+E1+L+N+X) TOTAL ENERGY = -243.35707846 A.U.
>AI> h> (K) KINETIC ENERGY = 0.00000000 A.U.
>AI> h> (E1=A-S+R) ELECTROSTATIC ENERGY = -243.35707846 A.U.
>AI> h> (S) ESELF = 408.25093361 A.U.
>AI> h> (R) ESR = 8.15805762 A.U.
>AI> h> (L) LOCAL PSEUDOPOTENTIAL ENERGY = 0.00000000 A.U.
>AI> h> (N) N-L PSEUDOPOTENTIAL ENERGY = 0.00000000 A.U.
>AI> h> (X) EXCHANGE-CORRELATION ENERGY = 0.00000000 A.U.
>AI> h> GRADIENT CORRECTION ENERGY = 0.00000000 A.U.
>AI> h>
>AI> h> NFI EKINC TEMPP EKS ECLASSIC EHAM DIS TCPU
>AI> h> 82001 0.00000 472.9 -243.35708 -242.51657 -242.51657 0.345E+02 14.87
>AI> h> 82002 0.00000 2124.3 -244.10715 -242.51947 -242.51947 0.345E+02 9.70
>AI> h> 82003 0.00000 5305.3 -245.56679 -242.52490 -242.52490 0.346E+02 9.65
>AI> h> 82004 0.00000 9665.9 -247.63808 -242.53218 -242.53218 0.346E+02 9.63
>AI> h> 82005 0.00000 14567.1 -250.17680 -242.54056 -242.54056 0.347E+02 9.69
>AI> h> 82006 0.00000 19123.5 -253.00362 -242.54942 -242.54942 0.348E+02 9.67
>AI> h> ------------------------------------------------------------------------------------------
>AI> h> finally the program stops with
>AI>
>AI> h> "PROGRAM STOPS IN SUBROUTINE DETERM| ILLEGAL RESULTS [PROC= 1]"
>AI>
>AI> h> Would you please help me to fix it ? The input and output files are also attached.
>AI> h> And the version of CPMD used is 3.11.1. I run it on the IBM eServer platform,
>AI> h> with Intel(R) Fortran Compiler for Intel(R) EM64T-based applications
>AI> h> (Version 9.1 ,Build 20060707 ,Package ID: l_fc_c_9.1.036).
>AI>
>AI> h> Dr. Axel Kohlmeyer once suggested that ddipo.F is one of the files that get
>AI> h> easily miscompiled by some compilers, so ddipo.F and wrener.F are preprocessored
>AI> h> with cpp and then compiled with -O0 option.
>AI>
>AI> h> Thanks in advance.
>AI>
>AI> h> Sincerely yours,
>AI>
>AI> h> Weijie Hua
>AI> h> Institute of Theoretical and Computational Chemistry,
>AI> h> Key Laboratory of Mesoscopic Chemistry,
>AI> h> School of Chemistry and Chemical Engineering,
>AI> h> Nanjing University,Nanjing 210093,
>AI> h> P.R.China
>AI> h> Email:huawj at itcc.nju.edu.cn
>AI> h> 2006-10-20
>AI>
>AI>
>AI>
>AI> -------------------------------------------------------
>AI> Alexandr Isayev,
>AI> Graduate Research Assistant, and System Administrator
>AI> @ Computational Center for Molecular Structure
>AI> and Interactions (CCMSI),
>AI> Jackson State University,
>AI> Jackson, MS USA
>AI> Tel: +(601) 979-1134
>AI> e-mail: alex(at)ccmsi.us
>AI> Web: http://www.ccmsi.us
>AI> --------------------------------------------------------
>AI>
>AI> _______________________________________________
>AI> CPMD-list mailing list
>AI> CPMD-list at cpmd.org
>AI> http://cpmd.org/mailman/listinfo/cpmd-list
>AI>
>
>--
>=======================================================================
>Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu
> Center for Molecular Modeling -- University of Pennsylvania
>Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
>tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425
>=======================================================================
>If you make something idiot-proof, the universe creates a better idiot.
>
>
-----------------------------------------------------
Weijie Hua
Institute of Theoretical and Computational Chemistry,
Key Laboratory of Mesoscopic Chemistry,
School of Chemistry and Chemical Engineering,
Nanjing University,Nanjing 210093,
P.R.China
Email:huawj at itcc.nju.edu.cn
2006-10-21
-----------------------------------------------------
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Makefile.sgiA330
Type: application/octet-stream
Size: 161322 bytes
Desc: not available
Url : http://cpmd.org/pipermail/cpmd-list/attachments/20061021/3e63cc18/attachment.obj
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Makefile.sgiA330
Type: application/octet-stream
Size: 161322 bytes
Desc: not available
Url : http://cpmd.org/pipermail/cpmd-list/attachments/20061021/3e63cc18/attachment-0001.obj
More information about the CPMD-list
mailing list