Re: Proposed SyncPPP layer modifications

From: Paul Fulghum (paulkf@microgate.com)
Date: Mon Feb 21 2000 - 12:31:10 EST


From: "Andrew Stanley-Jones" <asj@cban.com>
> From the basic HDLC information available and from my own testing,
> sending all packets with the multicast address set is incorrect and a bug
> in syncppp.
>
> The multicast address is only used when the system thinks the upper layer
> is sending a broadcast packet. Since we're generally not sending
> broadcast packets there's no reason to be using the multicast address.

I agree, but thought there would be less objection about
making it selectable in case something
(for whatever reason) required the existing behavior.

Do you think it should default to unicast or should be
hard coded to unicast?

> If you look closely btw, the 3600 is receiving the packet, recognizing it
> as an IP packet even. It presumably discards it before it get's to the
> forwarding stage.

Yes, you are right.

> On Mon, 21 Feb 2000, Paul Fulghum wrote:
>
> > I have encountered some Cisco routers (3600 series IOS 12.0 and 11.3)
that
> > refuse to recognize Cisco HDLC packets using the
> > Cisco multicast address (0xf9). They work fine with the
> > unicast address (0x09). A Cisco 1600 series (IOS 11.1) router accepts
> > both, but always uses the unicast for responses.
> >
> > syncppp is hard coded to always use the multicast address.
> > My suggestion (and how I am currently working) is to add a
> > flag PP_UNICAST to allow the user to select the address.
> > Default is not set so it behaves as before. If the flag is set then
> > the unicast address is used.

Paul Fulghum paulkf@microgate.com
Microgate Corporation www.microgate.com

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Wed Feb 23 2000 - 21:00:28 EST