[CPMD-list] problems during compilation

[Axel Kohlmeyer]

On Thu, 9 Nov 2006, Zhang Bing wrote:

ZB> Hi, dear all:

dear zhang,

ZB> We have tried so many times to compile the lasted CPMD version on our
ZB> cluster, but got error reports every single time, for this time, we

it would be nice to have seen some of them. if you don't report a 
problem, we cannot help you.

ZB> tried with configuration of LINUX_IA64_INTEL-MPI, unfortunately, error
ZB> again! The Make file and make log are attached with this email. Any
ZB> advice or suggestions will be greatly appreciated from you.

first off, you don't need to send the whole makefile, only the first
part with the compiler flags (the rest is identical for all machines),
also in this case the error results from the link, so we don't need the
whole make log, in in case we might need it, it would be nice, if you
would compress it before posting. note, that any post to this mailing
list goes to over a thousand people, so i guess most of them would 
appreciate you not filling their inboxes with (unnessecarily) large 
files. thanks.

back to your problem. from the looks of it, it seems completely 
unrelated to cpmd. can you compile any other fortran code on that
machine? what version of ifort do you have (did you updated to the
latest patch from premier support??)? finally, it looks as if you
want to compile from MPI, but you do not include any mpi libraries
or are using an mpi compiler wrapper. what kind of network does 
your cluster have and what mpi library is installed (there has to
be one installed, or else compilation would have failed earlier
for my_para.o ).

ZB> 
ZB> By the way, I was wondering which old version could be compiled
ZB> successfully with the configuration of LINUX_IA64_INTEL-MPI? Maybe we
ZB> should try some other versions of CPMD.

i can only advise against it. if it is a problem in CPMD, it has to be
fixed and the major improvements of the recent releases of CPMD were
more in terms of bug fixes and speed improvements than in new features,
so may encounter _more_ problems with older versions. there are some
hard to find bugs on (some) 64-bit machines in older releases.

cheers,
  axel.


ZB> Thanks again!
ZB> 
ZB>  
ZB> 
ZB> Zhang Bing
ZB> 
ZB>  
ZB> 
ZB> Nanoscience and Nanotechnology Initiative,
ZB> 
ZB>  
ZB> 
ZB> NUS
ZB> 
ZB> 

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