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:     Johannes Berg <johannes@sipsolutions.net>
Cc:     linux-wireless@vger.kernel.org,
        make-wifi-fast@lists.bufferbloat.net,
        John Crispin <john@phrozen.org>,
        Lorenzo Bianconi <lorenzo@kernel.org>,
        Felix Fietkau <nbd@nbd.name>
Subject: Re: [PATCH RFC/RFT 1/4] mac80211: Rearrange ieee80211_tx_info to make room for tx_time_est
In-Reply-To: <fd0d467d352177ac20100239cb18f80310b43fa9.camel@sipsolutions.net>
References: <156889576422.191202.5906619710809654631.stgit@alrua-x1> <156889576534.191202.17686228416284995279.stgit@alrua-x1> <fd0d467d352177ac20100239cb18f80310b43fa9.camel@sipsolutions.net>
Date:   Tue, 01 Oct 2019 11:08:30 +0200
Message-ID: <87v9t8vnn5.fsf@toke.dk>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Sender: linux-wireless-owner@vger.kernel.org
Precedence: bulk

Johannes Berg <johannes@sipsolutions.net> writes:

> Hi,
>
> Sorry for the long time to review here ...
>
> On Thu, 2019-09-19 at 14:22 +0200, Toke H=C3=B8iland-J=C3=B8rgensen wrote=
:
>> From: Toke H=C3=B8iland-J=C3=B8rgensen <toke@redhat.com>
>>=20
>> To implement airtime queue limiting, we need to keep a running account o=
f
>> the estimated airtime of all skbs queued into the device. Do to this
>> correctly, we need to store the airtime estimate into the skb so we can
>> decrease the outstanding balance when the skb is freed. This means that =
the
>> time estimate must be stored somewhere that will survive for the lifetim=
e
>> of the skb.
>>=20
>> Fortunately, we had a couple of bytes left in the 'status' field in the
>> ieee80211_tx_info; and since we only plan to calculate the airtime estim=
ate
>> after the skb is dequeued from the FQ structure, on the control side we =
can
>> share the space with the codel enqueue time. And by rearranging the orde=
r
>> of elements it is possible to have the position of the new tx_time_est l=
ine
>> up between the control and status structs, so the value will survive fro=
m
>> when mac80211 hands the packet to the driver, and until the driver eithe=
r
>> frees it, or hands it back through TX status.
>
> Seems reasonable to me, if we end up needing it and don't have an out-
> of-band path (that you seem to have been discussing in this long
> thread too)

Awesome! Any idea for how to make it work on big-endian systems? I got a
splat from the kbuild robot that triggered the BUILD_BUG_ON when
building for m68k. I assume it's the union with codel_time_t that ends
up being aligned wrong...

-Toke