> I've traced it some more: While mdiobus_register fails to find a PHY,
> creation of the "MDIO" bus is still successful and it returns successfully,
> having claimed the reset GPIO.
>
> of_phy_fixed_link_register tries to claim the same GPIO and fails.
I don't see where this is happening. It is looking for a gpio called
'link-gpios'.
> But regardless, there shouldn't have been an of_mdiobus_register and a MDIO
> bus probe
> before registering the fixed-link in the first place and my patch remedies
> that.
There are cases where you need both, e.g a switch controller over
MDIO. So we cannot make it one or the other.
However, there are currently no such boards. So far "net", lets go
with your partial revert patch. But we also need patches for
"net-next" which put it back again, allows both fixed-link and an
mdiobus inside a subnode, and not break backwards compatibility.
Andrew
