[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <f11b6e5d-380c-4f13-af91-3672e2c120d6@lunn.ch>
Date: Fri, 15 Nov 2024 19:08:19 +0100
From: Andrew Lunn <andrew@...n.ch>
To: Tristram.Ha@...rochip.com
Cc: Woojung.Huh@...rochip.com, olteanv@...il.com, robh@...nel.org,
krzk+dt@...nel.org, conor+dt@...nel.org, davem@...emloft.net,
edumazet@...gle.com, kuba@...nel.org, pabeni@...hat.com,
marex@...x.de, UNGLinuxDriver@...rochip.com,
devicetree@...r.kernel.org, netdev@...r.kernel.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH net-next 2/2] net: dsa: microchip: Add SGMII port support
to KSZ9477 switch
> That leaves the one situation where the SGMII port is connected directly
> to a MAC or each other. A customer once tried to do that and the SGMII
> register write was changed to support that, but I do not know if that
> project became a real product.
This is often done to cascade switches. In this setup, going down to
100Mbps or 10Mbps makes no sense, so 1000BaseX is used, not
SGMII. Today, fixed-link is used in this situation, combined with
setting phy-mode to 1000basex. Russell King has said in the past that
phylink could probably support this without fixed-link.
> The SGMII port in another chip can use 2.5G. The driver uses fixed PHY
> to get the MAC running. But the fixed PHY driver can only support speed
> up to 1000. There is no issue to adding higher speeds to that driver, but
> I think that is not advised?
Well, 2.5G is obviously not SGMII. It is 2500BaseX. There are also
many broken 2500BaseX implementations out there, which are SGMII cores
overclocked, and the signalling disabled, because SGMII signalling at
2.5G makes no sense, you want 2500BaseX signalling.
phylink fixed link also is not limited to 1G, it can do any speed.
Andrew
Powered by blists - more mailing lists