On Tue, 2007-01-09 at 22:57 +0100, Segher Boessenkool wrote:

quoted

Well, we use that method for most things,for example, the emac has
phandles to the zmii/rgmii/etc... for that specific case however, I
still prefer a cell index.

How about you put both in the device tree, and the kernel
can use whatever it wants -- best of both worlds :-)

I don't see the need of having some weird "soc" node that doesn't quite
mean anything (especially on axon) that has phandles to every sub device
in there :-)

I really don't see the interest in your approach but I may have missed
something.

On those ASICs, every device almost needs to know what is it's "cell
index" because of little details here or there. I really see that as an
attribute of the device and thus should be a property of the node.
 
Ben.