[Pw_forum] /opt/intel/Compiler/11.0/069/lib/intel64/libimf.so: warning: warning: feupdateenv is not implemented and will always fail
akohlmey at cmm.chem.upenn.edu
Tue Nov 18 18:44:18 CET 2008
On Tue, 18 Nov 2008, Paolo Giannozzi wrote:
PG> Axel Kohlmeyer wrote:
PG> > my suspicion is more that you are running either out of stack memory or in
PG> > general out of memory, but that is something that
PG> > you can monitor easily
PG> well..."you can monitor", ok, but "easily" ... maybe easily
PG> for you, Axel! Anyway: I heard that MPI sometimes allocates
PG> sizable amounts of memory.
once has to carefully differentiate between reserved address space
and memory used. the first happens a lot on "RDMA" infrastructures
like infiniband or myrinet. a process may show several GB under VIRT
(when monitoring with top) but the more important columns are
RES and DATA (which may need to be explicitly enabled) which
show the resident set size and the data segment size.
in any case, if we would know what the exact error messages are
and what the exact input was, we could do a much better guess
at what was going on.
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 users