[ewg] Re: pkey change handling patch

[Roland Dreier]

    Michael> I looked at cache.c and you are right. Maybe we should
    Michael> either 1. report events after cache has been updated or
    Michael> 2. make cache queries error out (EBUSY?) if cache hs not
    Michael> updated yet.

    Michael> Option 1 requires core changes, option 2 - ULP changes

    Michael> I would be inclined to go for 2.  Roland?

Yes, I agree.  How about ESTALE as an error code?

 - R.