Message-ID: <44f956b493e33bf394b2947b77e7c86f394b28ed.camel@sipsolutions.net>
Subject: Re: [PATCH 1/2] wifi: mac80211: introduce
 ieee80211_refresh_tx_agg_session_timer()
From:   Johannes Berg <johannes@sipsolutions.net>
To:     Ryder Lee <ryder.lee@mediatek.com>
Cc:     Shayne Chen <shayne.chen@mediatek.com>,
        Evelyn Tsai <evelyn.tsai@mediatek.com>,
        linux-wireless@vger.kernel.org, linux-mediatek@lists.infradead.org
Date:   Fri, 17 Feb 2023 19:01:10 +0100
In-Reply-To: <7c3f72eac1c34921cd84a462e60d71e125862152.1676616450.git.ryder.lee@mediatek.com>
References: <7c3f72eac1c34921cd84a462e60d71e125862152.1676616450.git.ryder.lee@mediatek.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.46.3 (3.46.3-1.fc37) 
MIME-Version: 1.0
Precedence: bulk

On Sat, 2023-02-18 at 01:50 +0800, Ryder Lee wrote:
> This allows low level drivers to refresh the tx agg session timer, based =
on
> querying stats from the firmware usually. Especially for some mt76 device=
s
> support .net_fill_forward_path would bypass mac80211, which leads to tx B=
A
> session timeout for certain clients.
>=20

Does it even matter? We could just request sessions without a timeout in
the first place.

Or do you have a strong reason to need the timeout, such as limited
hardware resources for (TX) aggregation sessions?

But then maybe you should just time them out based on FW statistics
directly, rather than having to periodically refresh the timer in
mac80211?

I don't mind the patch, and I'll happily take it if it's needed, I'm
just wondering if that isn't a very roundabout way of achieving things.

johannes