> On Mon, 06 Mar 2000 23:09:53 GMT, Brian Somers wrote:
> > Yes - with the TUNSIFMODE ioctl. This is different from how OpenBSD
> > does it (where our struct tuninfo has a ``dummy'' field, OpenBSDs has
> > a ``flags'' field where you put IFF_* bits).
> So then the following patch from NetBSD makes sense, yes?
No, not at the moment. IFF_POINTOPOINT and IFF_BROADCAST are
opposites, so TUNSIFMODE can be used to undo what it's already done.
In FreeBSD, IFF_MULTICAST is always there. In NetBSD there's
SIOCADDMULTI and SIOCDELMULTI and in OpenBSD there's struct tuninfo's
flags field (passed to TUNSIFINFO).
I'd condone supporting SIOCADDMULTI and SIOCDELMULTI, and also having
a flags field in struct tuninfo, but I don't know if this is a good
idea on a more general level.... ie, should there really be two ways
to do these things ?
For the moment though, I think the documentation is right.
> Index: tun.4
> RCS file: /home/ncvs/src/share/man/man4/tun.4,v
> retrieving revision 1.9
> diff -u -d -r1.9 tun.4
> --- tun.4 2000/03/06 07:08:14 1.9
> +++ tun.4 2000/03/07 12:37:25
> @@ -147,7 +147,10 @@
> its value must be either
> .Dv IFF_POINTOPOINT
> -.Dv IFF_BROADCAST .
> +.Dv IFF_BROADCAST
> +.Dv IFF_MULTICAST
> +may be or'ed into the value).
> The type of the corresponding
> .Em tun Ns Sy n
> interface is set to the supplied type. If the value is anything else,
Brian <brian@Awfulhak.org> <brian@[uk.]FreeBSD.org>
Don't _EVER_ lose your sense of humour !
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message