From: Alexandre Oliva <[EMAIL PROTECTED]>
Date: 28 Jul 2000 00:08:31 -0300
My only concern is that we won't be verifying whether funcname is
actually a function name.
Unfortunately I can't think of any way to do that reliably, without
having the macro caller specify a way to call the function, which is
more work. I don't think this extra work is needed.
we might as well just bite the bullet and use some broader-meaning
name that actually reflects what we do, such as AC_CHECK_SYMBOL.
That isn't an accurate name either, as you can't compare a structure
to zero. The technically correct name would be AC_CHECK_SCALAR.
But isn't that a bit obscure?
Or maybe just recommend the use of AC_CHECK_DECL instead?
Something like that might make more sense, yes. But my understanding
is that AC_CHECK_DECL(foo) does not define HAVE_FOO. This is
inconvenient, since much existing software uses HAVE_FOO to test
whether foo exists. It means that the autoconf user must invoke
AC_CHECK_FUNC(foo) as well as AC_CHECK_DECL(foo).
How about if we modify AC_CHECK_DECL(foo) so that it defines HAVE_FOO
as well as HAVE_DECL_FOO? Then we can ask people to use only
AC_CHECK_DECL and stop using AC_CHECK_FUNC.
