great. thank you for your help. oddly, that made it so that my
powerbook running Tiger worked swimmingly. unfortunately, it then
wouldn't work in the G5. it actually got just between one and two
lines further using any of my myriad fortran compilers. i do have
several fortran compilers (and i think all the libraries confuse each
other, which sucks), and got it to work fine using
./configure FC=gfortran LIBS=-lSystemStubs
which i found in the archives. when installing a different program
using g95, i'd had to add lSystemStubs to the LIBS as well. odd.
thanks again for all your help.
justin
On Jun 19, 2005, at 5:31 AM, Jeff Squyres wrote:
> On Jun 18, 2005, at 7:26 PM, justin litchfield wrote:
>
>
>> i'm having problems getting lam (7.1.1 or 7.0.6) to compile on my G5
>> running OSX 10.4.0. i've got g95 installed (among others) and need
>> fortran support for lam. when i run
>>
>> configure --with-fc (having previously exported FC as g95)
>>
>
> I think that this is your problem -- when you say "--with-fc", LAM's
> configure script is going to get the value "yes", which is actually an
> executable in OS X. That is, LAM expects --with-fc to have an
> argument; GNU Autoconf will supply the default argument "yes" if you
> don't. Try "--with-fc=g95", or just leave off "--with-fc" and let it
> find g95 by itself.
>
> Additionally, exporting $FC and using --with-fc is redundant; you
> really only need to do one or the other (they're meant to be two
> different mechanisms for essentially the same thing).
>
> Finally, if you're using Tiger, I would strongly recommend using
> one of
> the 7.1.2 betas (despite the name "beta", they're actually quite
> stable
> -- the biggest reason that it still has the "beta" label is because
> there is one IMPI bug left that I have not been able to find the time
> to fix yet. IMPI is a feature which needs to be specifically enabled
> at configure time; so unless you enable it, this bug does not affect
> you).
>
> --
> {+} Jeff Squyres
> {+} jsquyres_at_[hidden]
> {+} http://www.lam-mpi.org/
>
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>
|
