public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* [Fwd: HDLC driver: Frame-relay Inverse ARP]
@ 2002-05-16 19:40 Livio Ceci
  2002-05-19 15:17 ` Francois Romieu
  0 siblings, 1 reply; 2+ messages in thread
From: Livio Ceci @ 2002-05-16 19:40 UTC (permalink / raw)
  To: khc, linux-kernel

[-- Attachment #1: Type: text/plain, Size: 1233 bytes --]

Hi Mr. Halasa.

I sent you a message some days ago asking for the inverse ARP
implementation status in the hdlc driver.

Once I haven't received any answer, I started implementing it by myself,
based on linux kernel version 2.4.17. The protocol message exchange,
ARPOP_InREQUEST and ARPOP_InREPLY, is done and working fine, both being
treated by arp_rcv function in arp.c file.

Now, I would like to know which action should be taken when arp_rcv
(arp.c) receives an inverse ARP reply message. I thought just to add a
new route to the host that sent the message, but I would prefer having
your opinion about that.

I added a new paramenter in sethdlc to enable inverse arp, and just by
now the pointopoint parameter in ifconfig should be set to 0.0.0.0.
Thus, to setup a pvc :

sethdlc hdlc0 v35 clock ext
sethdlc hdlc0 fr lmi ansi dce t391 10 t392 15 n391 10 n392 10 n393 10
ifconfig hdlc0 0 0.0.0.0 up

sethdlc hdlc0 create 16 inverse_arp
ifconfig pvc0 10.0.0.2 pointopoint 0.0.0.0 mtu 1500 up

I attached a diff file with the changes I did, and also the sethdlc.
Please analyse them and send me back your comments, and if possible,
something that help me to "fill the blanks" I left as ??? in the code.

Thanks in advance.

Livio.

[-- Attachment #2: invARP.tgz --]
[-- Type: application/octet-stream, Size: 10995 bytes --]

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [Fwd: HDLC driver: Frame-relay Inverse ARP]
  2002-05-16 19:40 [Fwd: HDLC driver: Frame-relay Inverse ARP] Livio Ceci
@ 2002-05-19 15:17 ` Francois Romieu
  0 siblings, 0 replies; 2+ messages in thread
From: Francois Romieu @ 2002-05-19 15:17 UTC (permalink / raw)
  To: livio; +Cc: khc, linux-kernel

Greetings,

Livio Ceci <livio@cyclades.com.br> :
[rfc2390 for linux fr]
> I attached a diff file with the changes I did, and also the sethdlc.
> Please analyse them and send me back your comments, and if possible,
> something that help me to "fill the blanks" I left as ??? in the code.

May be something like this:
net/ipv4/arp.c
/*
 *     Special case: We must set Frame Relay source Q.922 address
 */
        if (dev_type == ARPHRD_DLCI) {
                sha = dev->broadcast;
                if (arp->ar_op == __constant_htons(ARPOP_InREQUEST)) {
                        /*
                         * Use fib_lookup in order to comply with
                         * address selection in rfc2390 - 7.1 and
                         * set tip consequently
                         */
                        ...
                } else if (arp->ar_op == __constant_htons(ARPOP_InREPLY))
                        arp->ar_op = ARPOP_REPLY;
        }

Then one forces ARPOP_InREQUEST to follow the same path as ARPOP_REQUEST,
arp_send(ARPOP_InREPLY,...) being the sole difference.
               ^^
Remarks ?

-- 
Ueimor

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2002-05-19 15:17 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-05-16 19:40 [Fwd: HDLC driver: Frame-relay Inverse ARP] Livio Ceci
2002-05-19 15:17 ` Francois Romieu

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox