David Gibson wrote:
> I mean, obviously the MDIO bus is accessed via some of the
> board-control registers. What I'm questioning is whether it makes
> sense to have a distinct node to represent the mdio bus, or whether
> the phys should just hang straight of the bcsr node.
Ah, I see. I think it does make sense, because it's a bus; if there
were another bus-like thing on the bcsr, there'd be conflicts over what
the children mean.
>>>> + [EMAIL PROTECTED] {
>>>> + #address-cells = <1>;
>>>> + #size-cells = <1>;
>>>> + device_type = "soc";
>>> Ditch the device_type.
>> No, it's used by the bootwrapper. I'll get rid of it if you want to
>> write a find_node_by_compatible() function. :-)
>
> Well, now that libfdt is merged, there is one :-p.
OK, it just needs exporting via ops. I'll put it on my to-do list, but
I don't think it should hold up board support patches.
-Scott
_______________________________________________
Linuxppc-dev mailing list
[email protected]
https://ozlabs.org/mailman/listinfo/linuxppc-dev