Re: [RFC] mac80211: Re-enable aggregation

[Sujith <Sujith.Manoharan@atheros.com>]

Johannes Berg wrote:
> > HT frames for aggregation enabled STAs are processed in the TX completion path, so
> > they are buffered.
> 
> But couldn't non-HT frames be buffered similarly? Maybe I'm missing one
> of the finer points of the 11n draft?

Probably because 11n aggregation have more rigorous timing requirements
between frames.

> > Makes sense, ath9k currently has a similar set of functions ( pausing/resuming TIDs).
> > How would the pending frames for each TID be managed ?
> > Will the driver be able to pull them as and when needed or would mac80211
> > push them down on checking if the TID is awake ?
> 
> I guess there's no clear answer here. How about "whichever you want"?
> Though I think I prefer pushing them down as that makes the model easier
> to understand. It probably also makes the Intel case easier to
> implement.

A way to pull down buffered frames for each TID (like ieee80211_get_buffered_bc() )
would be really useful for ath9k.

Sujith