Eric Blake <ebb9 <at> byu.net> writes:
> > --- 66,79 ----
> > AC_DEFUN([gl_PREREQ_VASNPRINTF_DIRECTIVE_A],
> > [
> > AC_REQUIRE([gl_PRINTF_DIRECTIVE_A])
> > ! case "$gl_cv_func_printf_directive_a" in
> > ! *yes)
>
> ;;
> *) # You need this, to avoid flipping the sense of the test
>
> > ! AC_DEFINE([NEED_PRINTF_DIRECTIVE_A], 1,
In fact, without it, CVS M4 dumps core on cygwin. So I'm applying this:
2007-03-26 Eric Blake <[EMAIL PROTECTED]>
* m4/vasnprintf.m4: Fix regression in logic sense of previous
patch.
Index: m4/vasnprintf.m4
===================================================================
RCS file: /sources/gnulib/gnulib/m4/vasnprintf.m4,v
retrieving revision 1.15
diff -u -b -r1.15 vasnprintf.m4
--- m4/vasnprintf.m4 26 Mar 2007 10:24:24 -0000 1.15
+++ m4/vasnprintf.m4 26 Mar 2007 18:21:58 -0000
@@ -68,6 +68,8 @@
AC_REQUIRE([gl_PRINTF_DIRECTIVE_A])
case "$gl_cv_func_printf_directive_a" in
*yes)
+ ;;
+ *)
AC_DEFINE([NEED_PRINTF_DIRECTIVE_A], 1,
[Define if the vasnprintf implementation needs special code for
the 'a' and 'A' directives.])
