> to do so and worry about rcu grace periods etc. Has any driver model > expert looked at this struct device reuse and blessed it as ok? I am quite sure this did not happen. That's what I mean with "despite the docs": One shouldn't reuse struct device, some people did it and got this "fix" in place, and now people might rely on the "fix". As a result, we can't simply remove it by saying "Ah, that's cruft. It can just go." We'd need some deeper look into random drivers, like you sketched.
Attachment:
signature.asc
Description: PGP signature