Re: [B.A.T.M.A.N.] [PATCH 5/6] batman-adv: Distributed ARP Table - add snooping functions for ARP messages

On Mon, Oct 31, 2011 at 01:12:21AM +0100, Simon Wunderlich wrote:

This patch fails checkpatch.pl --strict.

On Sun, Oct 30, 2011 at 09:56:01AM +0100, Antonio Quartulli wrote:

...

+bool arp_snoop_outgoing_request(struct bat_priv *bat_priv, struct sk_buff *skb) [...]

• arp_neigh_update(bat_priv, ip_src, hw_src);
• n = neigh_lookup(&arp_tbl, &ip_dst, primary_if->soft_iface);
• /* check if it is a valid neigh entry */
• if (n && (n->nud_state & NUD_CONNECTED)) {

• skb_new = arp_create(ARPOP_REPLY, ETH_P_ARP, ip_src,
  

• 		     primary_if->soft_iface, ip_dst, hw_src, n->ha,
  

• 		     hw_src);
  

• unicast_send_skb(skb_new, bat_priv);
  

• bat_dbg(DBG_ARP, bat_priv, "ARP request replied locally\n");
  

This function is hooked up in the tx patch, why do we send the unicast reply to the mesh? Shouldn't it be sent on the soft interface instead? I don't really understand this one ...

You are right. As we have already discussed on IRC, I should use netif_rx() directly here (to deliver the packet to the soft_iface).

...

if (is_my_mac(unicast_packet->dest)) { interface_rx(recv_if->soft_iface, skb, recv_if, hdr_size);

• return NET_RX_SUCCESS; }

This added newline is useless.

Yeah :D

Thanks,