On Fri, Sep 14, 2012 at 02:28:19PM +0200, Marc Kleine-Budde wrote: > On 09/14/2012 01:58 PM, Kishon Vijay Abraham I wrote: > > The PHY framework provides a set of API's for the PHY drivers to > > create/remove a PHY and the PHY users to obtain a reference to the PHY > > using or without using phandle. If the PHY users has to obtain a reference to > > the PHY without using phandle, the platform specfic intialization code (say > > from board file) should have already called phy_bind with the binding > > information. The binding information consists of phy's device name, phy > > user device name and an index. The index is used when the same phy user > > binds to mulitple phys. > > > > PHY drivers should create the PHY by passing phy_descriptor that has > > information about the PHY and ops like init, exit, suspend, resume, > > poweron, shutdown. > > Some comments inside. > > While looking over the code, I was thinking why not abstract the phy > with a "bus" in the linux kernel. The ethernet phys are on the mdio_bus, > see /sys/bus/mdio_bus. This saves you hand crafting devices, drivers and > bindings, I don't think that's a good idea, actually. You can have USB PHYs which are memory mapped, or connected through i2c, or connected through any other bus. If the PHYs layer itself is a bus, it means we will need to register a device on two different buses, which doesn't sound very nice IMHO. -- balbi
Attachment:
signature.asc
Description: Digital signature
