On Wed, Nov 4, 2020 at 11:55 AM Andrew Lunn <and...@lunn.ch> wrote: > > > https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/drivers/net/ethernet/microchip/lan743x_main.c?h=v5.9.3&id=6f197fb63850b26ef8f70f1bfe5900e377910a5a > > If you look at that patch, you see: > > - ret = phy_connect_direct(netdev, phydev, > - lan743x_phy_link_status_change, > - PHY_INTERFACE_MODE_GMII); > - if (ret) > - goto return_error; > > > That was added as part of the first commit for the lan743x > driver. Changing that now seems dangerous.
My knowledge of netdev/phy is extremely limited, so bear with me. The code quoted above (the first commit for lan743x) has been reinstated in my patch. It's literally identical - in the case the kernel can't find any available/sensible devicetree phy description. In the case of devicetree phys, which have been added recently, the 'phy-connection-type' prop appeared to have been optional, defaulting to (G)MII. My patch removes that devicetree default. So I guess my question is this - is there really a need for a devicetree default for 'phy-connection-type'? If not, no code duplication or mdio refactoring is required.
