Seems like another example of an
xyproblem.
Methinks someone doesn't understand pseudo filesystems.
If one is capable of writing kernel code to expose kobjects, that same someone had better be able to manage updates to the sysfs representation of it.
