Re: 41-tests-on-defines

[Alexandre Oliva <oliva at lsd dot ic dot unicamp dot br>]

On Dec 22, 1999, Akim Demaille <akim@epita.fr> wrote:

> +	Check that the AC_DEFINEs from acspecific.m4 do have a template.

Ok, this is one check in.

> +	And after all, let's run the test on acgeneral macros too.  This
> +	allowed to catch a few things.

Ok.  I'll concede that the changes below the paragraph above are all
related, in a very lax interpretation of the term ``relationship'' :-)

> +	* doc/autoconf.texi: Documennt changes of AC_MSG_ERROR and
                                   ^^typo

> +	Clean up a bit the handling of the LANG stack.  More is needed.

Ok, separate check-in.

> +dnl AC_CHECK_FUNCS(FUNCTION..., [ACTION-IF-FOUND], [ACTION-IF-NOT-FOUND])
> +dnl ---------------------------------------------------------------------
> +dnl FIXME: Should we die if there are not enough arguments, or just
> +dnl ignore?

We'd better die at autoconf time than at configure time.

>  AC_DEFUN(AC_LINKER_OPTION,
[snip]
> +dnl I don't understand the point of having the test inside of the
> +dnl loop.
> +for ac_link_opt in $1; do
> +  if test x"$using_gnu_compiler" = xyes; then
> +    $2="[$]$2 -Xlinker $ac_link_opt"
> +  else
> +    $2="[$]$2 $ac_link_opt"
> +  fi
> +done])
 
It could well be changed into a if-then-for-else-for-fi.

BTW, shouldn't using_gnu_compiler be preceded by `ac_'?
 
> +ifelse([$1], , [AC_FATAL([$0]: 1st arg must be defined)])dnl
> +ifelse([$2], , [AC_FATAL([$0]: 2nd arg must be defined)])dnl

How about `missing argument [12]'?

-- 
Alexandre Oliva http://www.ic.unicamp.br/~oliva IC-Unicamp, Bra[sz]il
oliva@{lsd.ic.unicamp.br,guarana.{org,com}} aoliva@{acm,computer}.org
oliva@{gnu.org,kaffe.org,{egcs,sourceware}.cygnus.com,samba.org}
** I may forward mail about projects to mailing lists; please use them