[CP2K:4871] Re: problem in compile cp2k 2.4 for parallel code

Jörg Saßmannshausen j.sassma... at ucl.ac.uk
Thu Jan 16 23:31:34 UTC 2014


Hi,

you don't need to build OpenMPI with the Intel compiler. GCC is working here 
quite well. You only need to change a few (text) files to use the Intel 
compiler with OpenMPI which was build with GCC. 

Regarding the other error: in my experience that is a problem with the initial 
guess. Try out a few or change the number of cores you are using. I never 
really got to the bottom of that problem.

Still my 2 cents from a wet London

Jörg


On Mittwoch 01 Januar 2014 jian... at gmail.com wrote:
> > hi, thanks for the suggestion of Glen and Ari,
> 
> yes, the error is the lack of scalpack library.
> as I add the library, the cp2k.popt is compiled.
> 
> But there are some other problems,
> when I test the H2O-32.inp in QS.
> there is a error:
> *************************************************************************
>  *** 14:55:04 ERRORL2 in cp_dbcsr_cholesky:cp_dbcsr_cholesky_decompose ***
>  *** processor 1  :: err=-300 condition FAILED at line 123             ***
>  *************************************************************************
> as I had read the topic of "any known issues for cp2k/2.4 with Openmpi
> 1.6.5?" in cp2k google group,
> who has the same error.
> in that topic, the cp2k work well by changing the openmpi1.6.5 with 1.7.2?
> 
> but when I use openmpi 1.7.2 or 1.7.3 or 1.6.4 to compile cp2k, the error
> still exists.
> 
> 
> In addition, when I compile openmpi with intel compiler,
> there are some warning when configure is going.
> 
> ####################################
> 
===========================================================================
> = == Contributed software setup
> 
===========================================================================
> =
> 
> *** Configuring contributed software packages
> checking which contributed software packages should be disabled...
> 
> --- libompitrace (m4 configuration macro)
> checking if contributed component libompitrace can compile... yes
> 
> --- vt (m4 configuration macro)
> checking whether PMPI is enabled... yes
> checking whether C++ STL is suitable for vt... no
> configure: WARNING:
> **************************************************************
> configure: WARNING: *** VampirTrace cannot be built due to your STL appears
> to
> configure: WARNING: *** be broken.
> configure: WARNING: *** Pausing to give you time to read this message...
> configure: WARNING:
> **************************************************************
> checking if contributed component vt can compile... no
> ####################################
> 
> and after finishing the compiling progress, mpicc, mpifort, and mpirun work
> well,
> but mpicxx can not work.
> 
> $ mpicxx
> --------------------------------------------------------------------------
> The Open MPI wrapper compiler was unable to find the specified compiler
> no in your PATH.
> 
> Note that this compiler was either specified at configure time or in
> one of several possible environment variables.
> --------------------------------------------------------------------------
> 
> That is my Linux-x86-64-intel.popt
> 
> ###############################################
> CC       = gcc
> CPP      =
> FC       = mpif90
> LD       = mpif90
> AR       = ar -r
> INTEL_MKL=/opt/intel/mkl
> INTEL_INC = -I$(INTEL_MKL)/include -I$(INTEL_MKL)/include/fftw
> INTEL_LIB = $(INTEL_MKL)/lib/intel64
> DFLAGS   = -D__HAS_NO_ISO_C_BINDING -D__INTEL -D__FFTSG -D__parallel
> -D__BLACS -D__SCALAPACK -D__FFTW3  -D__LIBINT -D__LIBXC2
> CPPFLAGS =  -C -traditional $(DFLAGS) -I$(INTEL_INC)
> FCFLAGS  = $(DFLAGS) -I$(INTEL_INC) -O1 -msse2 -heap-arrays 64
> -funroll-loops -fpp -free
> FCFLAGS2 = $(DFLAGS) -I$(INTEL_INC) -O1 -msse2 -heap-arrays 64 -fpp -free
> LDFLAGS  = $(FCFLAGS) -I$(INTEL_INC)
> LIBS     = -Wl,--start-group -L${INTEL_LIB}  -lmkl_blacs_openmpi_lp64
> -lmkl_lapack95_lp64 \
> -lmkl_scalapack_lp64 -lmkl_sequential -lmkl_intel_lp64 -lmkl_core \
> $(INTEL_MKL)/interfaces/fftw3xf/libfftw3xf_intel.a -Wl,--end-group \
> /home/myu1310/work/cp2k/cp2k-2.4.0/tools/hfx_tools/libint_tools/libint_cpp_
> wrapper.o \
> /home/myu1310/work/cp2k/libint/libint114/lib/libderiv.a \
> /home/myu1310/work/cp2k/libint/libint114/lib/libint.a \
> /home/myu1310/work/cp2k/libxc/libxc202/lib/libxc.a \
> -lstdc++ -lm -lpthread  -liomp5
> 
> 
> OBJECTS_ARCHITECTURE = machine_intel.o
> 
> 
> graphcon.o: graphcon.F
>         $(FC) -c $(FCFLAGS2) $<
> 
> ###############################################
> 
> any suggestions?
> thanks for help!!!


-- 
*************************************************************
Jörg Saßmannshausen
University College London
Department of Chemistry
Gordon Street
London
WC1H 0AJ 

email: j.sassma... at ucl.ac.uk
web: http://sassy.formativ.net

Please avoid sending me Word or PowerPoint attachments.
See http://www.gnu.org/philosophy/no-word-attachments.html




More information about the CP2K-user mailing list