FYI Netcdf is going to be "extremely" hard to get to work on something like
Blue Gene because it uses an autoconfigure which will pick up all the wrong
machine specs etc. Someone may have managed to hack together a netcdf
library for Blue Gene that you can link sander against manually but I think
it will be a real pain.
All the best
Ross
/\
\/
|\oss Walker
| Assistant Research Professor |
| San Diego Supercomputer Center |
| Tel: +1 858 822 0854 | EMail:- ross.rosswalker.co.uk |
|
http://www.rosswalker.co.uk <
http://www.rosswalker.co.uk/> | PGP Key
available on request |
Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.
_____
From: owner-amber-developers.scripps.edu
[mailto:owner-amber-developers.scripps.edu] On Behalf Of Carlos Simmerling
Sent: Monday, April 14, 2008 11:06
To: amber-developers.scripps.edu
Cc: vbabin.ncsu.edu
Subject: Re: amber-developers: problems with NCSU in amber10
the netcdf part does show up in ncsu-umbrella.f unless you add
-DNCSU_NO_NETCDF,
which is not set up by config if you say nobintraj. when I set that things
are working ok,
thanks for the quick reply.
we're using nobintraj just because it was a hassle recently on blue gene and
I didn't want to deal
with it while fixing other problems.
carlos
On Mon, Apr 14, 2008 at 1:39 PM, David A. Case <case.scripps.edu> wrote:
On Mon, Apr 14, 2008, Carlos Simmerling wrote:
> I've got several problems with the NCSU code that I would put into
> bugzilla but want to confirm first that we've not done something wrong.
>
> ifort -c -w95 -vec_report0 -mp1 -ip -O3 -axNP -FR "-DDISABLE_NCSU" -o
> ncsu-umbrella.o _ncsu-umbrella.f
> fortcom: Error: _ncsu-umbrella.f, line 114: Error in opening the Library
> module file. [NETCDF]
> use netcdf, only : nf90_double, nf90_float
> ----^
>
> this was compiled with nobintraj so should not be looking for netcdf
I can't reproduce this here. The code in ncsu-umbrella.f looks OK to me
on first glance. The offending "use netcdf" line should not be in the
cpp-processed code (i.e. not in _ncsu-umbrella.f).
So, the problem looks like it is in the cpp step, and should not depend on
compiler, etc. Check what happened in the step right before the one you
quote above, and look at the _ncsu-umbrella.f file to see what you have.
[Just the obvious suggestions: make sure that you really have a clean
working
directory and the latest versions of the code -- although the ncsu stuff has
not changed in since last December or so.]
At a higher level: why are you turning off netcdf? If people are having
problems with this, we need to know about it.
...dac
Received on Fri Apr 18 2008 - 21:19:19 PDT
