[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <59bc0c2b-0ece-427e-80c5-5b6920132989@lunn.ch>
Date: Mon, 3 Feb 2025 19:09:38 +0100
From: Andrew Lunn <andrew@...n.ch>
To: Nikita Yushchenko <nikita.yoush@...entembedded.com>
Cc: Yoshihiro Shimoda <yoshihiro.shimoda.uh@...esas.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>,
Geert Uytterhoeven <geert+renesas@...der.be>,
netdev@...r.kernel.org, linux-renesas-soc@...r.kernel.org,
linux-kernel@...r.kernel.org,
Michael Dege <michael.dege@...esas.com>,
Christian Mardmoeller <christian.mardmoeller@...esas.com>,
Dennis Ostermann <dennis.ostermann@...esas.com>
Subject: Re: [PATCH net-next] net: renesas: rswitch: cleanup max_speed setting
On Mon, Feb 03, 2025 at 06:09:41PM +0100, Nikita Yushchenko wrote:
> It was observed on spider board that with upstream kernel, PHY
> auto-negotiation takes almost 1 second longer than with renesas BSP
> kernel. This was tracked down to upstream kernel allowing more PHY modes
> than renesas BSP kernel.
>
> To avoid that effect when possible, always set max_speed to not more
> than phy_interface allows.
Please could you provide more information about the hardware. Is the
PHY more capable than the MAC?
When you have multi Gigi PHYs, the phy-mode is often taken to be the
starting point. But when the PHY has negotiated a link, it will tell
the MAC what PHY mode it needs to swap to, e.g from SGMII to
2500BaseX, etc.
You should only need max-speed when you have a PHY which can do more
than the MAC.
Also, phylink handles this a lot better than phylib. So you might want
to change rswitch to phylink, especially if you have link speeds > 1G.
Andrew
Powered by blists - more mailing lists