[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <5e5c647b-f05d-d9d3-0554-68af99ee3f55@gmail.com>
Date: Thu, 19 Mar 2020 09:55:32 -0700
From: Florian Fainelli <f.fainelli@...il.com>
To: Heiner Kallweit <hkallweit1@...il.com>,
Russell King - ARM Linux admin <linux@...linux.org.uk>,
Andrew Lunn <andrew@...n.ch>
Cc: David Miller <davem@...emloft.net>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>
Subject: Re: [PATCH net-next 1/3] net: phy: add and use phy_check_downshift
Le 2020-03-19 à 09:36, Heiner Kallweit a écrit :
> *Downshift reporting/warning*
> In most cases downshift is caused by some problem with the cabling.
> Users like the typical Ubuntu user in most cases are not familiar
> with the concept of PHY downshift and what causes a downshift.
> Therefore it's not sufficient to just report a downshift, we have
> to provide the user with a hint what to do.
> Adding the "downshifted" info to phy_print_status() is a good idea,
> however I'd see it as an optional addition to the mentioned hint
> to the user what to do.
> The info "unknown downshift" IMO would just cause confusion. If we
> have nothing to say, then why say something. Also users may interpret
> "unknown" as "there's something wrong".
Ideally we would also have support for cable testing as well which would
allow users to know what to do next to figure out why their link speed
is not what they had hoped for. That does require a specific PHY driver
though as there is no standard way to obtain that information.
FWIW, a bunch of drivers like tg3, e1000e, igc and possibly others do
report when downshifting occurs, of course not all of those in that list
use the PHY library.
--
Florian
Powered by blists - more mailing lists