Re: PATCH: (official, i.e. thought-out) Re: FreeBSD FortranFailure

[Loren James Rittle <rittle at latour dot rsch dot comm dot mot dot com>]

[The list never saw this since I accidentally posted a round of the
conversation to gcc-patch@gcc.gnu.org.  I sent another copy to fix
that round but it reappeared.  I hereby assume that Mark intended to
make this response public and will quote it here.]

In article <20010515135512P.mitchell@codesourcery.com>,
Mark Mitchell <mark@codesourcery.com> writes:

> I'm sympathetic to your concerns about the cast to off_t, i.e., in
> that someone might accidentally remove it.  However, we don't want
> explicit __FreeBSD__ #ifdefs; we try to do #ifdef based on some kind
> of feature test.

Understood, I will do it.

> On the branch, please just install the cast to off_t.

Done (without a code comment, if I understood correctly).

> Add a comment above it explaining why.  Install that on the mainline
> too.  But, maybe we will figure out a better way of doing it.

Done (with the code comment).

> The autoconf issue is sort-of a red herring; autoconf usually checks
> that functions exist in the library, not that they are actually
> prototyped.

I remember now.

> Thanks for tracking this down.

Loren