Peter,
On Wednesday, February 25, 2004, at 10:38 AM, Peter W. Draper wrote:
> BTW, the FC solution worked
I discovered today that AC_PROG_FC takes two arguments, not the one I
thought.
AC_PROG_FC([COMPILERS...], [DIALECT])
[COMPILERS...] is a list of specific compilers you want it to try, so
you could give it [f77 g77] as an argument. That's if you have rather
specific demands or need to more-or-less customize this. More useful
is the DIALECT, which can be 77, 90, 95 or 2000. Thus if you give this
as
AC_PROG_FC([], 77) or AC_PROG_FC([], 1977)
then this will search only its internal list of Fortran 77 compilers
(which happens to be [g77 f77 xlf frt pgf77 fort77 fl32 af77]
currently).
I still need to sort out what the libtool tagname business is, but it
might be worth changing this invocation in those cases where the
dialect matters, in preference to changing the code to be acceptable to
a fussier compiler. Or where that would be completely horrific, such
as where there's any sort of jumping into loops going on.
And I'm really sorry I mentioned comments.
> and I've now got as far as not being able to
> create the SST monolith (same problem others have seen - "/usr/bin/ld:
> cannot find -lgcc_s").
I'll fix this after I've done the documentation.
Norman
[actually `on strike' today and tomorrow, in case you were wondering,
but since not producing SSN/78 isn't guaranteed to bring the University
management to its knees, there seems little point in not doing it,
hence a modestly productive day at home.]
--
----------------------------------------------------------------------
Norman Gray http://www.astro.gla.ac.uk/users/norman/
Physics and Astronomy, University of Glasgow [log in to unmask]
|