[Make-wifi-fast] [PATCH RFC/RFT 4/4] mac80211: Apply Airtime-based Queue Limit (AQL) on packet dequeue

[Yibo Zhao]

On 2019-09-25 16:11, Toke Høiland-Jørgensen wrote:
> Yibo Zhao <yiboz at codeaurora.org> writes:
> 
>> So if it is going to work together with virtual time based mechanism 
>> in
>> the future, the Tx criteria will be met both of below conditions,
>>         1. Lower than g_vt
>>         2. Lower than IEEE80211_AIRTIME_QUEUE_LIMIT
> 
>> Are we going to maintain two kinds of airtime that one is from
>> estimation and the other is basically from FW reporting?
> 
> Yes, that was my plan. For devices that don't have FW reporting of
> airtime, we can fall back to the estimation; but if we do have FW
> reporting that is most likely going to be more accurate, so better to
> use that for fairness...

Do you mean we will use airtime reported by FW to calculate 
local->airtime_queued in case we have FW reporting airtime?

> 
>> Meanwhile, airtime_queued will also limit the situation that we only
>> have a station for transmission. Not sure if the peak throughput will
>> be impacted. I believe it may work fine with 11ac in chromiumos, how
>> about 11n and 11a?
> 
> Well, we will need to test that, of course. But ath9k shows that it's
> quite possible to run with quite shallow buffers, so with a bit of
> tuning I think we should be fine. If anything, slower networks need
> *fewer* packets queued in the firmware, and it's *easier* for the host
> to keep up with transmission.
> 
>> Anyway, I think this approach will help to improve performance of the
>> virtual time based mechanism since it makes packets buffered in host
>> instead of FW's deep queue. We have observed 2-clients case with
>> different ratio in TCP fails to maintain the ratio because the packets
>> arriving at host get pushed to FW immediately and thus the airtime
>> weight sum is 0 in most of time meaning no TXQ in the rbtree. For UDP,
>> if we pump load more than the PHY rate, the ratio can be maintained
>> beacuse the FW queue is full and packtes begin to be buffered in host
>> making TXQs staying on the rbtree for most of time. However, TCP has 
>> its
>> own flow control that we can not push enough load like UDP.
> 
> Yes, fixing that is exactly the point of this series :)
> 
> -Toke

-- 
Yibo