Re: [RFC] ath9k: Handle interface changes properly

[Ben Greear <greearb@candelatech.com>]

On 01/14/2011 11:24 AM, Rajkumar Manoharan wrote:
> On Sat, Jan 15, 2011 at 12:36:42AM +0530, Felix Fietkau wrote:
>> On 2011-01-14 7:53 PM, Rajkumar Manoharan wrote:
>>> On Fri, Jan 14, 2011 at 11:52:02PM +0530, Felix Fietkau wrote:
>>>>   On 2011-01-14 7:13 PM, Rajkumar Manoharan wrote:
>>>>   >   On Thu, Jan 13, 2011 at 10:19:38PM +0530, Felix Fietkau wrote:
>>>>   >>    On 2011-01-13 5:35 PM, Rajkumar Manoharan wrote:
>>>>   >>    >    Instead of setting opmde as AP for WDS, it is better to handle WDS
>>>>   >>    >    case in ath9k_hw.
>>>>   >>    Why? Right now I don't even see any NL80211_IFTYPE_WDS handling in
>>>>   >>    ath9k_hw, and I can't think of anything that should be handled
>>>>   >>    differently in ath9k_hw compared to the AP opmode.
>>>>   >   For WDS station, what should be the interface type? Forgive if I'm wrong.
>>>>   There is no WDS station opmode. 'WDS station' is a regular station
>>>>   interface with 4-addr mode enabled. It needs no special handling in ath9k.
>>> still not convinced. Then what is the point mac80211 is informing about
>>> WDS type to drivers. mac itself can pass it as AP type like what it
>>> is doing for p2p GO.
>> The WDS type is something else. If you have two APs, you can link them
>> together with a separate WDS vif on each side pointing at the remote MAC
>> address of the other node.
>> I think that when we use ah->opmode, we should only use it for very
>> generic operating modes:
>>
>> AP: no TSF sync, beacon tx can be enabled.
>> ADHOC: TSF sync against IBSS cell, beacon tx can be enabled
>> STATION: TSF sync against one AP, only station beacon timers for PS.
>>
>> Only the above distinctions are relevant for ath9k_hw, everything else
>> is handled by the driver/stack. There is no reason for adding extra
>> checks to ath9k_hw for mesh and WDS, since they work best with
>> ah->opmode set to AP, and there is nothing extra on the *hardware* side
>> that should be configured there via a different opmode. That's why I
>> think leaking the mac80211 interface types to ath9k_hw is a bad idea.
> Thanks for your detailed explanation. And meanwhile Ben is also
> fixing hw mode based on vif counter. It would be better if he make
> use this thread and address the same.
>
> Ben,
> Any comments.

If all I need to do is to treat WDS and MESH like an AP, that shouldn't
be too hard.

Does WDS need a beacon?

Is there any limitations to the number of WDS vifs allowed, including
combinations with other VIF types?

Thanks,
Ben

>
> --
> Rajkumar


-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com