[Freeciv-Dev] Re: build error under mac os x (PR#894)

[Brian Olson <locke@xxxxxxx>]

On Friday, August 17, 2001, at 08:08  PM, Gaute B Strokkenes wrote:

> > --- configure.in.orig   Fri Aug 17 01:54:12 2001
> > +++ configure.in        Fri Aug 17 01:58:17 2001
> > @@ -186,6 +186,11 @@
> >     AC_SUBST(ARFLAGS)
> >   fi
> >
> > +dnl MacOS X or Darwin specifics
> > +if test x`$UNAME -s` = xDarwin ; then
> > +  CFLAGS="$CFLAGS -traditional-cpp"
> > +fi
> > +
> >   dnl Windows-specific settings
> >   AC_MINGW32
> >   if test x"$MINGW32" = "xyes"; then
> I'm not sure this is the best way.  After all, it is possible to build
> a stock gcc on Darwin or OS X these days, so the check should examine
> the compiler, not the platform.  The question is, how do we do this?
> Does Apple's gcc define a preprocessor symbol of some sort that we
> could test, or do we have to grep the output of gcc -v or something?
Well, my (g)cc from apple does the following:
% cc --verbose -E - < /dev/null
Reading specs from /usr/libexec/gcc/darwin/ppc/2.95.2/specs
Apple Computer, Inc. version gcc-926, based on gcc version 2.95.2 
19991024 (release)
 /usr/libexec/gcc/darwin/ppc/2.95.2/cpp -lang-c -v -D__GNUC__=2 
-D__GNUC_MINOR__=95 -D__ppc__ -D__NATURAL_ALIGNMENT__ -D__MACH__ 
-D__BIG_ENDIAN__ -D__APPLE__ -D__ppc__ -D__NATURAL_ALIGNMENT__ 
-D__MACH__ -D__BIG_ENDIAN__ -D__APPLE__ -D__STDC__ -D__APPLE_CC__=926 
-D__DYNAMIC__ -
GNU CPP version 2.95.2 19991024 (release)
[excess verbosity snipped]

I would guess that a stock gcc wouldn't define __APPLE_CC__ so maybe 
that's what you'd need.
Alternately, construct a test out of a piece of code that fails the 
apple cpp. This is probably ideal so that things will Just Work someday 
when apple fixes their cpp.
But I don't yet know enough autoconf to do either.

[insert clever signoff here]
Brian Olson http://bolson.org/