[CPMD-list] still cpmd2cube segmentation fault error occurring

Axel Kohlmeyer akohlmey at vitae.cmm.upenn.edu
Thu Dec 15 13:35:03 CET 2005 

On Thu, 15 Dec 2005, haider abbas wrote:

HA> Dear all cpmd user,
HA> 
HA> I am using cpmd-3.7.2. It was compiled in the mid of

you should seriously consider upgrading to the latest
version of cpmd. there were many bugs fixed with respect
to linux machines and the intel compiler. if the cpmd2cube
code is of the same age, than upgrade that one, too.
there is no point in debugging code that has been superseded
by newer releases.
   axel.

HA> last year with the help of Axel. I was totally new to
HA> this work and he had helped me a lot and when i got
HA> the executable, i saved it in a separate directory and
HA> to learn more i did different experiment with the 
HA> makefile so i am not sure that the below makefile
HA> which i have is the same as was used in compilation
HA> cpmd-3.7.2. It may be or may not be but i have only
HA> this one.
HA> it was compiled using libatlas library from Axel web
HA> page as i remember.
HA> #----------------------------------------------------------------------------
HA> # Makefile for cpmd.x (plane wave electronic
HA> calculation)
HA> # Configuration: PC-IFC
HA> # Creation of Makefile: May 24 2004
HA> # on Linux sarus 2.4.18-14 #1 Wed Sep 4 13:35:50 EDT
HA> 2002 i686 i686 i386 GNU/Linux
HA> # Author: haider
HA> #----------------------------------------------------------------------------
HA> #
HA> SHELL = /bin/sh
HA> #
HA> #--------------- Default Configuration for PC-IFC
HA> ---------------
HA> SRC  = .
HA> DEST = .
HA> BIN  = .
HA> #QMMM_FLAGS = -D__QMECHCOUPL
HA> #QMMM_LIBS  = -L. -lmm
HA> FFLAGS = -c -r8 -w90 -w95 -O3 -pc64 -axM -ip -tpp7 
HA> LFLAGS = -L. -latlas $(QMMM_LIBS)
HA> CFLAGS = -c -O2 -Wall
HA> CPP = /lib/cpp -P -C -traditional
HA> CPPFLAGS = -D__Linux -D__PGI -DLAPACK -DFFT_DEFAULT
HA> -DLINUX_IFC  
HA> NOOPT_FLAG = 
HA> CC = gcc 
HA> FC = ifc 
HA> LD = ifc 
HA> AR = 
HA> 
HA> and the makefile of cpmd2cube is 
HA> 
HA> #-----------------------------------------------------------------------------#
HA> # Makefile for cpmd2cube.x
HA> # Configuration: Linux-PC-IFC-P4
HA> # Creation of Makefile: Nov 23 2005
HA> # on Linux robin 2.4.21-15.EL #1 Sun May 16 02:45:06
HA> EDT 2004 i686 athlon i386 GNU/Linux
HA> # Author: haider
HA> #----------------------------------------------------------------------------
HA> #
HA> SHELL = /bin/sh
HA> #
HA> #--------------- Configuration
HA> FC	= ifc
HA> FFLAGS	=  -FR -D__IFC -DFFT_DEFAULT -O -tpp7 -w90 -w95
HA> -pc64
HA> LFLAGS	= -O -tpp7 -w90 -w95 -pc64
HA> LIBS	=  -Vaxlib 
HA> # uncomment and set only if needed
HA> #CC	= cc
HA> #CFLAGS	= 
HA> #--------------- End of Configuration
HA> ##############
HA> 
HA> 
HA> i am running cpmd2cube and cpmd on the same pc but the
HA> error is still there
HA> 
HA> 
HA> Dear Prof. Hutter and Dear Prof. Apsi,
HA> 
HA> below is the result of what you have suggested me. 
HA> 
HA> haider at robin:~> cd codes
HA> haider at robin:~/codes> cd cpmd2
HA> haider at robin:~/codes/cpmd2> gdb cpmd2cube.x
HA> GNU gdb Red Hat Linux (6.0post-0.20040223.20rh)
HA> Copyright 2004 Free Software Foundation, Inc.
HA> GDB is free software, covered by the GNU General
HA> Public License, and you are
HA> welcome to change it and/or distribute copies of it
HA> under certain conditions.
HA> Type "show copying" to see the conditions.
HA> There is absolutely no warranty for GDB.  Type "show
HA> warranty" for details.
HA> This GDB was configured as
HA> "i386-redhat-linux-gnu"...(no debugging symbols
HA> found)...Using host libthread_db library
HA> "/lib/tls/libthread_db.so.1".
HA>  
HA> (gdb) run -dens DENSITY
HA> Starting program: /home/haider/codes/cpmd2/cpmd2cube.x
HA> -dens DENSITY
HA> Any sufficiently advanced technology is
HA> indistinguishable from magic.
HA>                 -- Arthur C. Clarke
HA> _______________________________________________________________
HA> 
HA> (no debugging symbols found)...(no debugging symbols
HA> found)...(no debugging symbols found)...(no debugging
HA> symbols found)...[Thread debugging using libthread_db
HA> enabled]
HA> [New Thread -1220262496 (LWP 5561)]
HA> (no debugging symbols found)...(no debugging symbols
HA> found)...
HA> .read input file ... done
HA> I think you stored a density (sum^2 =  0.12582877E-03)
HA> .building g vectors ... done
HA>  
HA> Program received signal SIGSEGV, Segmentation fault.
HA> [Switching to Thread -1220262496 (LWP 5561)]
HA> 0x08055f58 in cpmd2cube_mp_do_geom_ ()
HA> (gdb)
HA> 
HA> 
HA> thanks 
HA> yours sincerely 
HA> Haider Abbas
HA> 
HA> 
HA> 
HA> 
HA> 
HA> --- Ari P Seitsonen <ari.p.seitsonen at iki.fi> wrote:
HA> 
HA> > 
HA> > Dear Haider,
HA> > 
HA> >    Adding to Juerg's comments: Which FFT are you
HA> > using during the 
HA> > compilation (pre-processor flag '-Dxxx')? Is it the
HA> > same as what you used 
HA> > for CPMD? What's the output from 'ulimit -a' (or
HA> > 'limit', if using 
HA> > 'tcsh')? You could also run cpmd2cube.x within a
HA> > debugger (preferrably 
HA> > compiled with debugging, like Juerg suggested) like
HA> > 'gdb' or 'dbx', you 
HA> > call
HA> > 
HA> > # > gdb cpmd2cube.x
HA> > # : run -dens DENSITY
HA> > and when the code crashes
HA> > # : where
HA> > 
HA> > and send us the output. This _might_ help (no
HA> > promises), otherwise we're 
HA> > just guessing.
HA> > 
HA> >      Greetings from Lausanne,
HA> > 
HA> >         apsi
HA> > 
HA> >
HA> -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
HA> >    Ari P Seitsonen / Ari.P.Seitsonen at iki.fi /
HA> > http://www.iki.fi/~apsi/
HA> >    CNRS & IMPMC, Université Pierre et Marie Curie
HA> >        4 place Jussieu, case 115 / F-75252 Paris
HA> >    Tel: +33-1-4427 7542, Fax: +33-1-4427 3785, GSM:
HA> > +33-6-6736 3820
HA> > 
HA> > On Wed, 14 Dec 2005, Juerg Hutter wrote:
HA> > 
HA> > > Hi
HA> > >
HA> > > maybe you could compile the code with the debug
HA> > flag on,
HA> > > so we would get a meaningful traceback.
HA> > >
HA> > > regards
HA> > >
HA> > > Juerg Hutter
HA> > >
HA> > >
HA> >
HA> ----------------------------------------------------------
HA> > > Juerg Hutter                   Phone : ++41 44 635
HA> > 4491
HA> > > Physical Chemistry Institute   FAX   : ++41 44 635
HA> > 6838
HA> > > University of Zurich           E-mail:
HA> > hutter at pci.unizh.ch
HA> > > Winterthurerstrasse 190
HA> > > CH-8057 Zurich, Switzerland
HA> > >
HA> >
HA> ----------------------------------------------------------
HA> > >
HA> > >
HA> > > On Wed, 14 Dec 2005, haider abbas wrote:
HA> > >
HA> > >> Dear Axel,
HA> > >> 
HA> > >> I have used the command
HA> > >> 
HA> > >> haider at robin:~/cpmd2cube> ulimit -s unlimited
HA> > >> haider at robin:~/cpmd2cube> cpmd2cube.x -dens
HA> > DENSITY
HA> > >> 
HA> > >> .read input file ... done
HA> > >> I think you stored a density (sum^2 = 
HA> > 0.12582877E-03)
HA> > >> .building g vectors ... done
HA> > >> forrtl: severe (174): SIGSEGV, segmentation fault
HA> > >> occurred
HA> > >> Image              PC        Routine           
HA> > Line
HA> > >>     Source
HA> > >> cpmd2cube.x        08055F58  Unknown
HA> > >> Unknown  Unknown
HA> > >> cpmd2cube.x        0805404E  Unknown
HA> > >> Unknown  Unknown
HA> > >> cpmd2cube.x        0805880F  Unknown
HA> > >> Unknown  Unknown
HA> > >> cpmd2cube.x        0806C9D8  Unknown
HA> > >> Unknown  Unknown
HA> > >> cpmd2cube.x        40099768  Unknown
HA> > >> Unknown  Unknown
HA> > >> cpmd2cube.x        08049CE1  Unknown
HA> > >> Unknown  Unknown
HA> > >> 
HA> > >> but it is still there, i am using intel_fc_80 and
HA> > >> White Box Enterprise Linux release 3.0
HA> > (Liberation
HA> > >> Respin 1)
HA> > >> Kernel 2.4.21-15.EL on an i686
HA> > >> 
HA> > >> now please guide me
HA> > >> Thanks
HA> > >> yours sincerely
HA> > >> Haider Abbas
HA> > >> 
HA> > >> 
HA> > >> 
HA> > >> --- Axel Kohlmeyer <akohlmey at vitae.cmm.upenn.edu>
HA> > >> wrote:
HA> > >> 
HA> > >>> 
HA> > >>> hi,
HA> > >>> 
HA> > >>> this is most likely not a problem of the code.
HA> > >>> please try raising the stacksize using the
HA> > ulimit
HA> > >>> command.
HA> > >>> the newer intel compilers allocate automatic
HA> > arrays
HA> > >>> from
HA> > >>> the stack.
HA> > >>> 
HA> > >>> axel.
HA> > >>> 
HA> > >>> 
HA> > >>
HA> >
HA> =======================================================================
HA> > >>> Axel Kohlmeyer    e-mail:
HA> > akohlmey at cmm.upenn.edu,
HA> > >>> tel: ++1-215-898-1582
HA> > >>>     Center for Molecular Modeling    --
HA> > >>> University of Pennsylvania
HA> > >>> Department of Chemistry, 231 S.34th Street,
HA> > >>> Philadelphia, PA 19104-6323
HA> > >>> 
HA> > >>
HA> >
HA> =======================================================================
HA> > >>> If you make something idiot-proof, the universe
HA> > >>> creates a better idiot.
HA> > >>> 
HA> > >>> On Mon, 12 Dec 2005, haider abbas wrote:
HA> > >>> 
HA> > >>>> Dear all cpmd user,
HA> > >>>> 
HA> > >>>> running cpmd2cube shows the following error
HA> > >>>> 
HA> > >>>> [haider at falcon1 cpmd2cube]$ cpmd2cube.x  -rho
HA> > >>>> -halfmesh ELPOT
HA> > >>>> 
HA> > >>>> .read input file ... done
HA> > >>>> I think you stored a density (sum^2 =
HA> > >>> 0.77965627E-03)
HA> > >>>> .building g vectors ... done
HA> > >>>> forrtl: severe (174): SIGSEGV, segmentation
HA> > fault
HA> > >>>> occurred
HA> > >>>> Image              PC        Routine
HA> > >>> Line
HA> > >>>>     Source
HA> > >>>> cpmd2cube.x        08055F58  Unknown
HA> > >>>> Unknown  Unknown
HA> > >>>> cpmd2cube.x        0805404E  Unknown
HA> > >>>> Unknown  Unknown
HA> > >>>> cpmd2cube.x        0805880F  Unknown
HA> > >>>> Unknown  Unknown
HA> > >>>> cpmd2cube.x        0806C9D8  Unknown
HA> > >>>> Unknown  Unknown
HA> > >>>> cpmd2cube.x        00763E33  Unknown
HA> > >>>> Unknown  Unknown
HA> > >>>> cpmd2cube.x        08049CE1  Unknown
HA> > >>>> Unknown  Unknown
HA> > >>>> 
HA> > >>>> 
HA> > >>>> the makefile of cpmd2cube.x is
HA> > >>>> 
HA> > >>>> 
HA> > >>> 
HA> > >>
HA> >
HA> #-----------------------------------------------------------------------------#
HA> > >>>> # Makefile for cpmd2cube.x
HA> > >>>> # Configuration: Linux-PC-IFC-P4
HA> > >>>> # Creation of Makefile: Nov 23 2005
HA> > >>>> # on Linux robin 2.4.21-15.EL #1 Sun May 16
HA> > >>> 02:45:06
HA> > >>>> EDT 2004 i686 athlon i386 GNU/Linux
HA> > >>>> # Author: haider
HA> > >>>> 
HA> > >>> 
HA> > >>
HA> >
HA> #----------------------------------------------------------------------------
HA> > >>>> #
HA> > >>>> SHELL = /bin/sh
HA> > >>>> #
HA> > >>>> #--------------- Configuration
HA> > >>>> FC	= ifc
HA> > >>>> FFLAGS	=  -FR -D__IFC -DFFT_DEFAULT -O -tpp7
HA> > -w90
HA> > >>> -w95
HA> > >>>> -pc64
HA> > >>>> LFLAGS	= -O -tpp7 -w90 -w95 -pc64
HA> > >>>> LIBS	=  -Vaxlib
HA> > 
HA> === message truncated ===>
HA> _______________________________________________
HA> > CPMD-list mailing list
HA> > CPMD-list at cpmd.org
HA> > http://cpmd.org/mailman/listinfo/cpmd-list
HA> > 
HA> 
HA> 
HA> 
HA> 		
HA> __________________________________________________________ 
HA> Yahoo! India Matrimony: Find your partner now. Go to http://yahoo.shaadi.com
HA> _______________________________________________
HA> CPMD-list mailing list
HA> CPMD-list at cpmd.org
HA> http://cpmd.org/mailman/listinfo/cpmd-list
HA> 

-- 
=======================================================================
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.

More information about the CPMD-list mailing list