This is most likely just a thinko on my part. Easy to fix, just use APR_HAS_SO_ACCEPTFILTER
everywhere. :-)
Ryan
On Tuesday 24 July 2001 20:12, Cliff Woolley wrote:
> I'm trying to clear up a warning on FreeBSD related to
> apr_socket_accept_filter() (no prototype). The problem is that
> apr_network_io.h has an #ifdef APR_OS_ACCEPT_FILTER around the prototype
> for that function, and APR_OS_ACCEPT_FILTER doesn't exist anywhere else in
> the code. [The prototype is wrong anyway, but that's easy to fix.] It
> gets worse.
>
> I see about five variations of the name of this macro, as follows:
>
> 1) configure.in and apr.h.in check sys/socket.h for SO_ACCEPT_FILTER and
> set APR_HAS_SO_ACCEPT_FILTER and...
>
> 2) HAVE_SO_ACCEPT_FILTER [shouldn't #1 be APR_HAVE_SO_ACCEPT_FILTER?]
>
> 3) apr_network_io.h has an #ifdef APR_OS_ACCEPT_FILTER around the
> prototype as mentioned above
>
> 4) network_io/unix/sockopt.c uses #ifdef SO_ACCEPTFILTER around the
> function itself, pulling directly from the system header.
>
> 5) Apache's server/listen.c has a test for #ifdef APR_HAS_SO_ACCEPTFILTER.
> (Note that this is *not* the same as #1.)
>
> Sheesh!!
>
>
> FreeBSD is the only platform I've been able to find (checked Solaris 2.6,
> Linux, Irix, AIX, FreeBSD) that has either SO_ACCEPTFILTER or
> SO_ACCEPT_FILTER in any system header file. On FreeBSD, it's called
> SO_ACCEPTFILTER. Is there any platform that defines SO_ACCEPT_FILTER,
> which is what configure is currently checking for? If so, which one? If
> not, I'll change configure.in and apr.h.in appropriately and go from
> there.
>
> Thanks,
> Cliff
>
> --------------------------------------------------------------
> Cliff Woolley
> cliffwoolley@yahoo.com
> Charlottesville, VA
--
_____________________________________________________________________________
Ryan Bloom rbb@apache.org
Covalent Technologies rbb@covalent.net
-----------------------------------------------------------------------------
|