Milter-greylist

On Sat, 2005-03-26 at 21:15 +0100, manu@... wrote:
> ranko_z <ranko@...> wrote:
> 
> > *** 210,219 ****
> >   fi
> >   if test $ipv6_cv_$1 = yes; then
> >     AC_DEFINE_UNQUOTED($ac_tr_lib)
> > -   ifelse([$2], , :, [$2])
> >   else
> >     ifelse([$3], , :, [$3])
> >   fi])
> > 
> >   IPv6_CHECK_FUNC(getaddrinfo,
> >       AC_DEFINE([HAVE_GETADDRINFO], [],
> > --- 210,219 ----
> >   fi
> >   if test $ipv6_cv_$1 = yes; then
> >     AC_DEFINE_UNQUOTED($ac_tr_lib)
> >   else
> >     ifelse([$3], , :, [$3])
> >   fi])
> > + #  ifelse([$2], , :, [$2])
> > 
> >   IPv6_CHECK_FUNC(getaddrinfo,
> >       AC_DEFINE([HAVE_GETADDRINFO], [],
> 
> What does this change do?

If you execute configure as it was you will notice in config.log that
HAVE_GETADDRINFO and HAVE_GETNAMEINFO do get declared twice in
confdefs.h as follows:
#define HAVE_GETADDRINFO
#define HAVE_GETADDRINFO 1
#define HAVE_GETNAMEINFO
#define HAVE_GETNAMEINFO 1

The above makes gcc to issue a warning message. Some checks, like the
one with localtime_r and snprintf, depend on gcc -Werror switch which
turns all warnings to errors and the above redefine makes the checks to
produce the wrong results.

What I have done above, I have removed the line that creates this
problem. I just did not remove it all the way. I've just shifted it out
of the IPv6_CHECK_FUNC declaration and commented it out.

Best regards,

Ranko

-- 
Ranko Zivojnovic,
IT Director/CTO          ranko@...

Spidernet Services Ltd., Tel: +357 22 844844
Nicosia, Cyprus          FAX: +357 22 669470