Yes. It seems like there are two separate use-cases, which is leading
configure to have a bit of an inconsistent approach
1) ./configure -mpi intel:
Require that all the paths are set up: mpicc points to mpiicc, with all the
libs and paths set up correctly, and mpiicc pointing to icc. This is
reasonable for a cluster where someone has just typed "module load intel",
and not unreasonable for the sort of desktop user who can be bothered
installing the intel compilers at all.
Despite requiring all this prior setup, it still (currently, pulled from
dev tree about a week ago) has a kludgey look around for the intel library
path, sometimes breaking if it turns out that the path is not what was
expected.
2) ./configure intel:
Doesn't require scripts to be sourced, sets fc=ifort, etc, itself, and also
has a kludgey look around for the intel library path.
....I thought I was just submitting a bugfix, but it seems that this runs a
bit deep for a friday afternoon. My personal favourite approach is
actually (2), but with a better regexp, just to save users some trouble.
Installing the intel compilers can be a terrible faff, I have sometimes
even found it impossible when they made a buggy release, so it is probably
better to try everything we can to paper over the cracks of a
messy/incomplete environment setup. I am really feeling
uncharacteristically generous on this one.
Josh
On 18 January 2013 14:57, Jason Swails <jason.swails.gmail.com> wrote:
>
> Serial installations don't use mpicc, so if the libraries would still have
> to be exposed by LD_LIBRARY_PATH for a serial installation to work (which
> is where the C compiler will be most useful, since most C codes in Amber
> are serial-only). With ifort and icc installed in separate places, I'm
> guessing two sets of Intel scripts need to be sourced, yes?
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Fri Jan 18 2013 - 07:30:04 PST
