Re: [PATCH] nl80211/cfg80211: add ssid filtering for sched_scan

[Luciano Coelho <coelho@ti.com>]

On Wed, 2011-08-31 at 09:39 +0200, Johannes Berg wrote: 
> On Wed, 2011-08-31 at 10:30 +0300, Luciano Coelho wrote:
> 
> > > Ok. I think the BSSID blacklist would just be a separate attribute, I
> > > see no reason to have a nesting for that.
> > 
> > Actually, we could make the filter and blacklisting very generic.  We
> > can use the same attributes for both.  The only difference would be that
> > the former would include anything that matches and the latter would
> > exclude anything that matches.
> 
> Hm, true, but does it really make sense to blacklist SSIDs when we have
> a whitelist for them? Or would you want to whitelist a wildcard and then
> blacklist a specific one?

Whitelisting and blacklisting the same SSID would be one of the
"illegal" cases that we may want to reject with -EINVAL.

But my idea was that you could do this:

MATCH={SSID=foo} and
NOT_MATCH={BSSID=00:01:02:03:04:05}

or

MATCH={BSSID=00:01:02:03:04:05} and
NOT_MATCH={SSID=foo}

or, making more sense:

MATCH={BSSID=00:01:02:03:04:05} and
NOT_MATCH={MIN_RSSI=-70}

My point is that many of these combinations might not make sense, but
generalizing the API allows for cases that *do* make sense but we have
not thought of yet.


> > I also prefer using match/not_match (couldn't find a better antonym),
> > because "filter" is ambiguous (let-through or leave-out?).
> 
> agree.

We could also use a single list like this:

FILTER={1={MATCH=true, SSID=foo}, 2={MATCH=false, BSSID=...}

But then we're becoming a bit too "netfiltery". :P What do you think?

-- 
Cheers,
Luca.