Received-SPF: pass (google.com: best guess record for domain of linux-wireless-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Toke =?utf-8?Q?H=C3=B8iland-J=C3=B8rgensen?= <toke@redhat.com>
To:     Kan Yan <kyan@google.com>
Cc:     Johannes Berg <johannes@sipsolutions.net>,
        linux-wireless <linux-wireless@vger.kernel.org>,
        Make-Wifi-fast <make-wifi-fast@lists.bufferbloat.net>,
        ath10k@lists.infradead.org, John Crispin <john@phrozen.org>,
        Lorenzo Bianconi <lorenzo@kernel.org>,
        Felix Fietkau <nbd@nbd.name>,
        Rajkumar Manoharan <rmanohar@codeaurora.org>,
        Kevin Hayes <kevinhayes@google.com>
Subject: Re: [PATCH v9 4/4] mac80211: Use Airtime-based Queue Limits (AQL) on packet dequeue
In-Reply-To: <CA+iem5vVhwTsTvAgnC_HF0utMDSfgyLCe2M=gfUOFU2VjjCB8Q@mail.gmail.com>
References: <157382403672.580235.12525941420808058808.stgit@toke.dk> <157382404118.580235.14216392299709793599.stgit@toke.dk> <CA+iem5tF+TN-osfGxu=EU5Xt1Uq+PcKgBVaoAmZY3a3JzS5JzQ@mail.gmail.com> <87y2wgjas4.fsf@toke.dk> <CA+iem5vVhwTsTvAgnC_HF0utMDSfgyLCe2M=gfUOFU2VjjCB8Q@mail.gmail.com>
Date:   Mon, 18 Nov 2019 11:10:29 +0100
Message-ID: <87muctcx62.fsf@toke.dk>
MIME-Version: 1.0
Content-Type: text/plain; charset=WINDOWS-1252
Content-Transfer-Encoding: quoted-printable
Sender: linux-wireless-owner@vger.kernel.org
Precedence: bulk

Kan Yan <kyan@google.com> writes:

>> > Given the fact that AQL is only tested in very limited platforms,
>> > should we set the default to disabled by removing this change in the
>> > next update?
>> >
>> > -       local->airtime_flags =3D AIRTIME_USE_TX | AIRTIME_USE_RX;
>> > +
>> > +       local->airtime_flags =3D AIRTIME_USE_TX |
>> > +                              AIRTIME_USE_RX |
>> > +                              AIRTIME_USE_AQL;
>> > +       local->aql_threshold =3D IEEE80211_AQL_THRESHOLD;
>> > +       atomic_set(&local->aql_total_pending_airtime, 0);
>> Well, we have the whole -rc series to get more testing in if we merge it
>> as-is. It's up to the maintainers, of course, but I would be in favour
>> of merging as-is, then optionally backing out the default before the
>> final release if problems do turn up. But I would hope that the limits
>> are sufficiently conservative that it would not result in any problems :=
)
>
> Sounds good. The current default limits are reasonably conservative
> and are tunable via debugfs.
>
> I will give the v10 version of this patch serial a quick test and
> hopefully we can wrap it up soon.

Sounds good, thanks! Also, seems we got a 5.4-rc8 yesterday, so we have
another week before the merge window.

Johannes, any chance you'll get a chance to take a look at this sometime
this week? :)

-Toke