Re: [RFC] design discussion: Collecting information for (non-peer) stations

From: Simon Wunderlich <simon.wunderlich@s2003.tu-chemnitz.de>
To: Simon Wunderlich <simon.wunderlich@s2003.tu-chemnitz.de>
Cc: Adrian Chadd <adrian@freebsd.org>,
	Felix Fietkau <nbd@openwrt.org>,
	Antonio Quartulli <antonio@open-mesh.com>,
	"linux-wireless@vger.kernel.org" <linux-wireless@vger.kernel.org>,
	Thomas Pedersen <thomas@cozybit.com>,
	"johannes@sipsolutions.net" <johannes@sipsolutions.net>,
	Marek Lindner <marek@open-mesh.com>,
	Mathias Kretschmer <mathias.kretschmer@fokus.fraunhofer.de>,
	Zefir Kurtisi <zefir.kurtisi@neratec.com>
Subject: Re: [RFC] design discussion: Collecting information for (non-peer) stations
Date: Fri, 8 Mar 2013 15:13:27 +0100	[thread overview]
Message-ID: <20130308141327.GA3363@pandem0nium> (raw)
In-Reply-To: <20130225102817.GA2069@pandem0nium>

[-- Attachment #1: Type: text/plain, Size: 1213 bytes --]

Hey,

On Mon, Feb 25, 2013 at 11:28:17AM +0100, Simon Wunderlich wrote:
> On Fri, Feb 22, 2013 at 09:42:43AM -0800, Adrian Chadd wrote:
> Anyway, I'll look at the rx_handler stuff what Felix suggested and see
> if this could work out. :)
> 

We've drafted some external module - the general approach seems to work quite
nicely, the only thing which appears needless it to add a radiotap header
and remove it later in the statistics module again ...

Anyway, here is some code to digest for anyone who is interested:

https://github.com/simonwunderlich/wifi_statistics

It hooks in an rx_handler onto mon0 (yes, that's currently static), and
throws out some info via debugfs. It parses the radiotap and wifi header
of every packet on the monitor interface.

@Antonio, Zefir: I'd be interested in your opinions. Would that be useful
for you too? Should we add all the statistics we are interested in, or better
use some kind of API if we are interested in "special statistics" which no one
else is?

If there is general interest, we can propose to include it in upstream Linux.
Otherwise, we can maintain it (together ?) as seperate module. It's not so
big anyway ...

Cheers,
	Simon

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 198 bytes --]