Gentoo Archives: gentoo-cluster

From: Bryan Green <bgreen@××××××××.gov>
To: gentoo-cluster@l.g.o
Subject: Re: [gentoo-cluster] using Portland Group fortran for lam-mpi
Date: Thu, 06 Apr 2006 00:56:10
Message-Id: 200604060055.k360tebP020982@ece06.nas.nasa.gov
1 Speaking of ebuilds and fortran, I've been nosing around trying to get
2 things working with lam-mpi (I'm actually just administering a machine that
3 someone else is trying to use mpi on).
4 >From looking at the various ebuilds for lam-mpi, lapack, and blas, it looks
5 like newer ebuilds use fortran.eclass, while older ones hardcode the
6 fortran compiler they want to use into the ebuild. There is no support
7 whatsoever for Portland Group (pgf).
8 To get everything working with pgf90, I had to setup some ebuilds in an
9 overlay which, a) avoided fortran.eclass and b) removed the FC/F77
10 hardcoding in the ebuilds for blas and lapack, so that my FC setting in
11 make.conf would be used.
12
13 fortran.eclass has to be avoided, because it simply punts if the compiler
14 you request is not in its hardcoded list of blessed compilers.
15 It looks like fortran.eclass is a work in progress. Anybody have more
16 information on that? I'm wondering if I can help so that I can pick a
17 fortran compiler to use on a system, without having to hack ebuilds in a
18 portage overlay.
19
20 -bryan
21
22 Donnie Berkholz writes:
23 > Bryan Green wrote:
24 > > Interesting question. Is there a difference?
25 > > I am under the impression that lam-mpi needs to be built with pgf for
26 > > programs that use lam-mpi to be built with pgf. Am I mistaken?
27 > > I any case, I found that I get get it to build by setting the FC environmen
28 > t
29 > > variable. My initial error was apparently caused because of the lack of a
30 > > license server setting.
31 >
32 > Good to hear it! If you ever come across an ebuild where FC/F77, CC, or
33 > CXX aren't respected, please file a bug.
34 >
35 > Thanks,
36 > Donnie
37 --
38 gentoo-cluster@g.o mailing list

Replies

Subject Author
Re: [gentoo-cluster] using Portland Group fortran for lam-mpi Donnie Berkholz <spyderous@g.o>