[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: interface groups
- To: tech_(_at_)_openbsd_(_dot_)_org
- Subject: Re: interface groups
- From: Can Erkin Acar <canacar_(_at_)_eee_(_dot_)_metu_(_dot_)_edu_(_dot_)_tr>
- Date: Wed, 23 Jun 2004 00:43:33 +0300
On Tue, Jun 22, 2004 at 03:26:33PM -0600, Henning Brauer wrote:
> * Peter Galbavy <peter_(_dot_)_galbavy_(_at_)_knowtion_(_dot_)_net> [2004-06-22 12:48]:
> > Philipp Buehler wrote:
> > > Or what did I miss (besides more coffee)?
> > OK, this is pretty un-thought-out
> indeed ;)
> > but how about a way to treat a group as a "virtual" interface for BPF ?
> no way. does not make sense, and canacar says it's impossible anyway,
> and I tend to believe him on claims about BPF ;)
To be more clear, in the general case your interface group might
consist of interfaces with completely different link layer types.
(pflog0 and fxp1 and tun0 for instance) they can not be combined
into the same bpf descriptor and trying to support it through
hacks in tcpdump itself would be quite ugly.