>>> But, many SoCs do have a notion of device "number", which is relevant >>> for programming other general control registers in places. We need >>> to >>> encode it somehow, and it would be good to have a consistent way of >>> doing it. >> >> I'm not convinced there isn't a more direct way to represent >> the relevant relationships. >> >> Either way, we don't have enough experience with this stuff >> yet to know what works well and what doesn't (at least, I >> don't, and I haven't seen any evidence that others do); so >> I'd prefer to keep this in per-device bindings for now; it >> should be there anyhow, but once we do have experience with >> it we could do some recommendation. > > Well of course it will remain in the per-device bindings. But just > because these are in different per-device bindings doesn't mean we > can't *try* to use consistent property names for similar things...
Sure, if those properties do exactly the same thing, identical names are a good thing. Until we're reasonably certain that this is a good way to do things I'd rather people experiment a bit instead of all cluster around a random thing that would be labeled "de facto standard". :-) Segher _______________________________________________ Linuxppc-dev mailing list Linuxppc-dev@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-dev
