[TEST/REVIEW] ng_ipfw: node to glue together ipfw(4) andnetgraph(4)

Gleb Smirnoff glebius at freebsd.org
Wed Jan 19 05:01:48 PST 2005


On Wed, Jan 19, 2005 at 11:16:01AM +0100, Andre Oppermann wrote:
A> > On Wed, Jan 19, 2005 at 01:32:35AM -0800, Julian Elischer wrote:
A> > J> If each active divert socket number had a pointer to the module to which it
A> > J> was attached then  you could divert to either in-kernel netgraph targets or
A> > J> to userland socket based targets.  Currently of you divert to a divert
A> > J> 'port number' and nothing is attached to it, the packet is dropped.
A> > J> If a divert socket is attached to it, it is sent ot teh socket.
A> > J> I would just suggest that is not a great leap of imagination that
A> > J> attaching to a hook named 3245 would attach a netgrpah hook to the ipfw
A> > J> code in the sam enamespace as the divert portnumber, and that a
A> > J> subsequent attempt to attach a divert socket to that port number woild
A> > J> fail. The packets diverted there would simply go to the netgraph hook
A> > J> instead of going to a socket or being dropped.
A> > 
A> > I understand your idea now. I'll work in this direction.
A> 
A> I like Julian's idea.  And if you look at the mtag's the only thing that
A> is extracted is the rule number for divert, dummynet and netgraph (your
A> patch).  Ideally this should be merged into one tag if possible and not
A> an architectual hack.

When writing node, I was thinking about merging this into one tag. However, I
expected negative response to this idea, from other developers.

Anyone else agree that these tags should be merged?

-- 
Totus tuus, Glebius.
GLEBIUS-RIPN GLEB-RIPE


More information about the freebsd-net mailing list