[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <6214363b-0242-481d-9b93-2db9e1ba5913@lunn.ch>
Date: Mon, 25 Aug 2025 14:08:31 +0200
From: Andrew Lunn <andrew@...n.ch>
To: Heiner Kallweit <hkallweit1@...il.com>
Cc: Alex Tran <alex.t.tran@...il.com>, Andrew Lunn <andrew+netdev@...n.ch>,
"David S . Miller" <davem@...emloft.net>,
Eric Dumazet <edumazet@...gle.com>,
Jakub Kicinski <kuba@...nel.org>, Paolo Abeni <pabeni@...hat.com>,
netdev@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH net v1] Fixes: xircom auto-negoation timer
On Mon, Aug 25, 2025 at 08:44:18AM +0200, Heiner Kallweit wrote:
> On 8/25/2025 3:28 AM, Alex Tran wrote:
> > Auto negoation for DP83840A takes ~3.5 seconds.
> > Removed sleeping in loop and replaced with timer based completion.
> >
> You state this is a fix. Which problem does it fix?
>
> IMO touching such legacy code makes only sense if you:
> - fix an actual bug
> - reduce complexity
> - avoid using deprecated API's
>
> Do you have this hardware for testing your patches?
>
> You might consider migrating this driver to use phylib.
> Provided this contributes to reducing complexity.
There is plenty to reduce. There is a full bit-banging MDIO
implementation which could be replaced with the core implementation.
The harder part for converting to phylib will be the ML6692 and
DP83840A. There are no Linux driver for these, but given the age,
there is a good chance genphy will work.
Andrew
Powered by blists - more mailing lists