I'll install and run the testers, as I don't believe the sysadmin did indeed do that. thanks for the pointer, I'll post once I have a result (or working system!).<br><br><div><span class="gmail_quote">On 5/29/07,
<b class="gmail_sendername">Axel Kohlmeyer</b> <<a href="mailto:akoh...@gmail.com">akoh...@gmail.com</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>eric,<br><br>did your BLACS compilation pass all the BLACS testers?<br><br>the fact, that it compiles successfully, does not mean that<br>it executes correctly. with a complex package like BLACS/ScaLAPACK<br>you need to make sure that this works correct independently.
<br><br>i tried openmpi-1.1 and it did not work for me (x86_64, intel 9.1).<br>i upgraded to openmpi-1.2.1 and it worked.<br><br>cheers,<br> axel.<br><br>On 5/29/07, Eric Shamay <<a href="mailto:eric....@gmail.com">
eric....@gmail.com</a>> wrote:<br>><br>> I've managed to get the entire program to compile. Here are the<br>> pertinent stats of the system:<br>><br>> firstly, my arch file<br>> ---------------------------------------------
<br>> PERL = perl<br>> CC = mpicc<br>> CPP = cpp<br>> FC = mpif90 -FR<br>> FC_fixed = mpif90 -FI<br>> LD = mpif90<br>> AR = ar -r<br>> MKLPATH = $(MKL)/lib/64<br>> DFLAGS = -D__INTEL -D__FFTSG -D__FFTW3 -D__parallel -D__BLACS -
<br>> D__SCALAPACK -D__FFTMKL -I$(FFTW)/include -I$(MKL)/include<br>> CPPFLAGS = -C $(DFLAGS) -P -traditional<br>> FCFLAGS = -O2 -Vaxlib -tpp2<br>> LDFLAGS = $(FCFLAGS) -L$(HOME)/lib -L$(FFTW)/lib -L$(MPI)/lib -L$
<br>> (MKL)/lib/64<br>> #LIBS = -lmkl_scalapack -lmkl_blacsF77init -lmkl_blacs\<br>> # -lmkl_lapack -lmkl -lguide -lmpi -lfftw3<br>> LIBS = $(MKLPATH)/libmkl_blacs.a\<br>> $(MKLPATH)/libmkl_scalapack.a\
<br>> $(MKLPATH)/libmkl_blacsF77init.a\<br>> $(MKLPATH)/libmkl_lapack.a\<br>> $(MKLPATH)/libmkl_blacs_intelmpi.a\<br>> $(MKLPATH)/libguide.a\
<br>> -lpthread -lmpi -lmkl -lfftw3<br>><br>> OBJECTS_ARCHITECTURE = machine_intel.o<br>> ---------------------------------------------<br>><br>> intel/9.1 (for the compilers)<br>><br>
> > mpif90 -V<br>> Intel(R) Fortran Itanium(R) Compiler for Itanium(R)-based applications<br>> Version 9.1 Build 20070320 Package ID: <installpackageid><br>><br>> >mpicc -V<br>> Intel(R) C Itanium(R) Compiler for Itanium(R)-based applications
<br>> Version 9.1 Build 20070320 Package ID: <installpackageid><br>><br>> mkl/7.2cluster_64<br>><br>> fftw/3.1.1<br>><br>> mpi/openmpi-1.1.5_intel-9.1<br>><br>> This setup finishes through the entire compilation and exits without
<br>> error. I then attempt to run a test as per the INSTALL file but by<br>> running the command:<br>><br>> > mpirun -np 8 $PATH_TO_CP2KEXE H2O.inp > H2O.out<br>><br><br>> So far all the help was right on track and I've finally been able to
<br>> get it to compile, but now I'm stuck at this near-immediate crash<br>> after trying to run the program in parallel. Any further help would be<br>> greatly appreciated.<br>><br>> ~Eric<br>><br>>
<br>><br>><br>><br><br>--<br>=======================================================================<br>Axel Kohlmeyer <a href="mailto:akoh...@cmm.chem.upenn.edu">akoh...@cmm.chem.upenn.edu</a> <a href="http://www.cmm.upenn.edu">
http://www.cmm.upenn.edu</a><br> Center for Molecular Modeling -- University of Pennsylvania<br>Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323<br>tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425
<br>=======================================================================<br>If you make something idiot-proof, the universe creates a better idiot.<br><br></div><br><br clear="all"><br>-- <br>~E