[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <cdf2e1e8-39ff-48b3-84b6-73c673ab1eb1@lunn.ch>
Date: Thu, 18 Jul 2024 05:43:28 +0200
From: Andrew Lunn <andrew@...n.ch>
To: Raju Lakkaraju <Raju.Lakkaraju@...rochip.com>
Cc: netdev@...r.kernel.org, davem@...emloft.net, kuba@...nel.org,
horms@...nel.org, hkallweit1@...il.com, richardcochran@...il.com,
rdunlap@...radead.org, linux@...linux.org.uk,
bryan.whitehead@...rochip.com, edumazet@...gle.com,
pabeni@...hat.com, linux-kernel@...r.kernel.org,
UNGLinuxDriver@...rochip.com
Subject: Re: [PATCH net-next V2 3/4] net: lan743x: Migrate phylib to phylink
On Tue, Jul 16, 2024 at 05:03:48PM +0530, Raju Lakkaraju wrote:
> Migrate phy support from phylib to phylink.
> Fixed phy support is still used together with phylink since we need to support
> dynamic fallback when a phy is not found over mdio. While phylink's FIXED mode
> supports fixed phys that, it's dynamic and requires device tree entries which
> are most of the time not present for LAN743x devices
> +static int lan743x_phylink_connect(struct lan743x_adapter *adapter)
> +{
> + struct device_node *dn = adapter->pdev->dev.of_node;
> + struct net_device *dev = adapter->netdev;
> + struct fixed_phy_status fphy_status = {
> + .link = 1,
> + .speed = SPEED_1000,
> + .duplex = DUPLEX_FULL,
> + };
So you are happy to limit it to 1G, even thought it can do more? That
is the problem with fixed PHY done this way. If you were to use
PHYLINK fixed PHY you can use the full bandwidth of the hardware.
You might want to look at what the wangxun drivers do for some ideas
how you can make use of PHYLINK fixed link without having DT.
Andrew
Powered by blists - more mailing lists