Re: definition of NULL

[Bruno Haible]

Hi,

I wrote:
> In m4/locale_h.m4 it's a waste of configure time to check for the include_next
> support of locale.h if it's not going to be replaced. I would keep doing
>   gl_CHECK_NEXT_HEADERS([locale.h])
> conditionally.
> 
> Likewise for m4/stddef_h.m4 and m4/wchar.m4.

The part for m4/wchar.m4 that I submitted was wrong: it breaks when
gl_REPLACE_WCHAR_H is invoked. For example, on mingw.

I'm undoing it:


2009-08-14  Bruno Haible  <address@hidden>

        * m4/wchar.m4 (gl_WCHAR_H): Undo invalid optimization in last commit.

--- m4/wchar.m4.orig    2009-08-15 02:37:11.000000000 +0200
+++ m4/wchar.m4 2009-08-15 02:34:12.000000000 +0200
@@ -7,7 +7,7 @@
 
 dnl Written by Eric Blake.
 
-# wchar.m4 serial 24
+# wchar.m4 serial 25
 
 AC_DEFUN([gl_WCHAR_H],
 [
@@ -45,9 +45,7 @@
     HAVE_WCHAR_H=0
   fi
   AC_SUBST([HAVE_WCHAR_H])
-  if test -n "$WCHAR_H"; then
-    gl_CHECK_NEXT_HEADERS([wchar.h])
-  fi
+  gl_CHECK_NEXT_HEADERS([wchar.h])
 ])
 
 dnl Unconditionally enables the replacement of <wchar.h>.