Asynchronous OSCORE B.1.1 operation

[chrysn]

The WIP implementation of OSCORE B.1.1 (incrementing persisted sequence numbers in jumps of k) currently blocks while persisting sequence numbers. This only happens every k received packages, but should still not happen.

The planned solution is to start flushing at a high-water mark (say, 90%), note it as pending in the security context, and the network outruns the file system, raise ContextUnavailable into a 5.03 Max-Age:max(time-spent-in-fsync-so-far, 1).

Until then, blocking at the exhaustion point is probably good enough for practical use.