[PATCH] libertas: monitor mode support for OLPC

[Luis Carlos Cobo]

On 7/4/07, Holger Schurig <hs4233 at mail.mn-solutions.de> wrote:
> Any technical reason why you went the echo-to-proc-sys way and
> not the "iwconfig eth1 mode monitor" ?  I like this latter
> approach better:

I use the echo-to-sys way because we are not setting any existing
interface into monitor mode but creating a new one. This is the way
Intel ipw2xxx drivers behave and I think it is a cleaner solution to
have separate interfaces, one for 802.3 and another one for
802.11+rtap; even if we don't use them at the same time. Additionally,
further development can lead to simultaneous operation of the ethX and
monitor interface. Anyway, I'm open to hear more pros/cons and
reconsider it.

> When you stop ethX and mshX anyway, then you won't really need an
> extra "monitormode" variable. You should re-use adapter->mode
> for this.

I'm not completely comfortable with that, in the future we could want
to get 802.11 frames via the rtap interface while associated. Also,
monitormode is not a flag but a mask indicating which kind of frames
(data, all management but beacons, beacons) we are receiving, so I
feel it's not elegant to put that on the mode variable.

> really adds a "rtap" device. So maybe you rename every
> occurence "prom" (Programmable Read Only Memory) into "rtap"

Agreed to change (prom is from promiscuous but is not that appropriate anyway).

> Kernel coding style says that there's no " " before the "(". But
> this is nitpicking, sorry :-)

No worries, thanks for pointing that out.

> Couldn't this be a table?

Of course, you are welcome to patch that :-) I was just fixing the values.

I will wait a day or so for more comments and resend the patch.

-- 
Luis Carlos Cobo Rus       GnuPG ID: 44019B60
cozybit Inc.