Re: logic of HAVE_STAT_EMPTY_STRING_BUG is inverted?

[IIDA Yosiaki]

Hello, autoconf maintainers.

I wrote:
>I hope that the following patch will fix it.  Enjoy!
>-  return $1 ("", &sbuf) == 0;]])],
>+  return $1 ("", &sbuf);]])],

The above patch made mismatch with diag message.  Oops.

--- autoconf-2.60/lib/autoconf/functions.m4~    2006-06-23 15:14:13.000000000 
+0000
+++ autoconf-2.60/lib/autoconf/functions.m4     2006-09-20 05:47:28.875578000 
+0000
@@ -1447,3 +1447,4 @@
 [AC_REQUIRE([AC_FUNC_LSTAT_FOLLOWS_SLASHED_SYMLINK])dnl
-AC_CACHE_CHECK([whether $1 accepts an empty string],
+dnl Fixed by IIDA.Y 2006-09-20.
+AC_CACHE_CHECK([whether $1 fails an empty string],
               [ac_cv_func_$1_empty_string_bug],
@@ -1451,3 +1452,4 @@
 [[struct stat sbuf;
-  return $1 ("", &sbuf) == 0;]])],
+/* Fixed by IIDA.Y 2006-09-20. */
+  return $1 ("", &sbuf);]])],
            [ac_cv_func_$1_empty_string_bug=no],
--
  iida