ia_netbroadcast

[Andre Oppermann]

Pawel Jakub Dawidek wrote:
> 
> On Tue, May 18, 2004 at 03:47:56AM -0700, Luigi Rizzo wrote:
> +> On Tue, May 18, 2004 at 12:05:03PM +0200, Pawel Jakub Dawidek wrote:
> +> ...
> +> > +> > Do we still need ia_netboradcast field? It is calculated depending on
> +> > +> > old-fashion classes (A, B, C). Is it used still?
> +> > +> > I wonder if ia_broadaddr isn't sufficient today.
> +> > +>
> +> > +> not just that, it seems that ia_netboradcast contains bogus info.
> +> >
> +> > So?
> +>
> +> nuke it!
> +> Also if, as Andre said, it might be used by windows apps,
> +> one could certainly compute it on the fly from the IP address.
> 
> Ok, this patch:
> 
>         http://people.freebsd.org/~pjd/patches/no_old_style_nets.patch
> 
> removes ia_netbroadcast, ia_net, ia_netmask field from structure
> in_ifaddr and removes sysctl net.inet.ip.subnets_are_local as suggested
> in commit log of revision 1.24 of file netinet/in.c:
> 
> Set subnetsarelocal to false.  In a classless world, the other case
> is almost never useful.  (This is only a quick hack; someone should
> go back and delete the entire subnetsarelocal==1 code path.)
> 
> Tests and comments are welcome.

I have checked that at least Win2k doesn't use the classful broadcast
address anymore.  I don't know about older versions of Windows.

However I'm not yet sure we (better I) understand all implications of
removing the things you do in your patch.  Please hold off for a moment
until I've finished thinking and looking through the implications.

-- 
Andre