All of lore.kernel.org
 help / color / mirror / Atom feed
From: Kyeyoon Park <kyeyoonp@qca.qualcomm.com>
To: netfilter@vger.kernel.org
Cc: "Malinen, Jouni" <jouni@qca.qualcomm.com>
Subject: IEEE 802.11 Proxy ARP using netfilter?
Date: Thu, 10 Jul 2014 15:23:19 -0700	[thread overview]
Message-ID: <53BF1257.80308@qca.qualcomm.com> (raw)

Hi guys,

I am trying to use netfilter infrastructure to implement Proxy ARP in 
IEEE 802.11. This is different from proxy_arp that can be enabled via 
/proc/sys/net/ipv4/conf/<iface>/proxy_arp.

I initially tried to implement this feature entirely in the mac80211 
stack, but was asked to look at alternative designs (i.e. netfilter). 
The relevant discussion can be found at:

http://comments.gmane.org/gmane.linux.kernel.wireless.general/123342

In a nutshell, I am trying to keep track of MAC <-> IP address mapping
for only the associated stations for the access point (AP). Of course, 
as the stations associate and disassociate from the AP, I will need to 
update whether to keep track of a particular MAC address or not. When an 
entity asks for the IP address of one of the associated stations, the AP 
will reply on behalf of the station and drop the broadcast ARP request 
that would go out to the wireless-medium.

I was thinking about implementing this using the target extension for 
ebtables (i.e. a target extension module for "ebtable_filter" something 
that would look similar to "ebt_arpreply.c").

Then I saw that I cannot dynamically change the data that is being used 
by my own target extension module. So when a new station associates or
when an associated station disassociates, I cannot dynamically alter 
database within my own target extension module; the dynamic data being 
the database of the MAC addresses to keep track of.

I COULD rewrite the whole table every time, but this is obviously not 
scalable (i.e. imagine rewriting the table of 100 MAC addresses from the 
userspace on every association/disassociation).

What do you think is the best way to approach this? I don't even have to 
use the target extension necessarily. I just thought it was the most 
suitable solution on a first glance.

Your comment on this would be greatly appreciated.

Best Regards,
Kyeyoon

                 reply	other threads:[~2014-07-10 22:23 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=53BF1257.80308@qca.qualcomm.com \
    --to=kyeyoonp@qca.qualcomm.com \
    --cc=jouni@qca.qualcomm.com \
    --cc=netfilter@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.