You're right... the regexp is unneccessary, in that all that stuff is
on-path anyway for my (and probably all sane) installations of the intel
compiler suite.
>mpicc -show:
icc -ldl -ldl -ldl -ldl -I/opt/apps/easybuild/software/impi/
4.0.2.003/intel64/include -L/opt/apps/easybuild/software/impi/
4.0.2.003/intel64/lib ... etc etc etc
That doesn't mean it isn't broken (and breaking other things) as well
though.... so therefore the conclusion is just to get rid of that section
of the config file, or to replace it with a check to see if the libs are
on-path.
Josh
On 18 January 2013 13:46, Jason Swails <jason.swails.gmail.com> wrote:
> On Fri, Jan 18, 2013 at 5:27 AM, Josh Berryman <
> the.real.josh.berryman.gmail.com> wrote:
>
> > Hello, just a suggestion, I find that the regexp in
> > AmberTools/src/configure2:
> >
> > fl=`which ifort | sed -e 's.\(.*\)bin.\1lib.' -e 's.ifort..'`
> >
> > ...doesn't find the intel libraries on the cluster I'm using just now.
> The
> > admins have put ifort and the associated libs in:
> >
> > /opt/apps/easybuild/software/ifort/2011.6.233/bin/intel64/ifort
> > /opt/apps/easybuild/software/ifort/2011.6.233/compiler/lib/intel64/
> >
> > which seems reasonable enough, but breaks the regexp in two different
> > ways.
> >
> > This two-line approach works for me, does anyone with more experience of
> > this sort of thing see a reason not to commit it onto the main branch?
> >
> > ifhome=`which ifort | sed 's.\(.*\)bin\/.*.\1.'`
> > fl=`find $ifhome -name libifcore.a | sed 's.\/libifcore.a..'`
> >
>
> I'm not sure why this was all necessary in the first place... Have you
> sourced the scripts that set up your environment for the Intel compilers?
> Something like this:
>
> . /opt/apps/easybuild/software/ifort/2011.6.233/bin/compilervars.sh
> intel64
>
> or something to that effect. The problem if you _don't_ source those
> scripts is that unless you link statically, you'll need to set
> LD_LIBRARY_PATH to point to those libs anyway, yes? And what about icc in
> your system? Is that in a different place?
>
>
>
> >
> > Josh
> > _______________________________________________
> > AMBER-Developers mailing list
> > AMBER-Developers.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber-developers
> >
>
>
>
> --
> Jason M. Swails
> Quantum Theory Project,
> University of Florida
> Ph.D. Candidate
> 352-392-4032
> _______________________________________________
> AMBER-Developers mailing list
> AMBER-Developers.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber-developers
>
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Fri Jan 18 2013 - 06:00:05 PST
