[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <aMD7sFlcQJpOK51B@pengutronix.de>
Date: Wed, 10 Sep 2025 06:16:48 +0200
From: Oleksij Rempel <o.rempel@...gutronix.de>
To: Jakub Kicinski <kuba@...nel.org>
Cc: Hubert Wiśniewski <hubert.wisniewski.25632@...il.com>,
Andrew Lunn <andrew+netdev@...n.ch>,
"David S. Miller" <davem@...emloft.net>,
Eric Dumazet <edumazet@...gle.com>, Paolo Abeni <pabeni@...hat.com>,
stable@...r.kernel.org, kernel@...gutronix.de,
linux-kernel@...r.kernel.org, netdev@...r.kernel.org,
Lukas Wunner <lukas@...ner.de>,
Russell King <linux@...linux.org.uk>, Xu Yang <xu.yang_2@....com>,
linux-usb@...r.kernel.org
Subject: Re: [PATCH net v1 1/1] net: usb: asix: ax88772: drop phylink use in
PM to avoid MDIO runtime PM wakeups
On Tue, Sep 09, 2025 at 04:56:45PM -0700, Jakub Kicinski wrote:
> On Tue, 9 Sep 2025 09:17:17 +0200 Oleksij Rempel wrote:
> > > > Fixes: e0bffe3e6894 ("net: asix: ax88772: migrate to phylink")
> > >
> > > It does, but v5.15 (including v5.15.191 LTS) is affected as well, from
> > > 4a2c7217cd5a ("net: usb: asix: ax88772: manage PHY PM from MAC"). I think
> > > it could also use a patch, but I won't insist.
> >
> > Ack, I'll try do address it later.
>
> Any idea what the problem is there? Deadlocking on a different lock?
Yes, it is PM lock taken in MDIO access inside of PM routine. We can't
reliably detect a PM context from MDIO bus read/write calls. Taking
local counters or setting flags would partially work, but they all are
not safe against race conditions.
--
Pengutronix e.K. | |
Steuerwalder Str. 21 | http://www.pengutronix.de/ |
31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
Powered by blists - more mailing lists