We used to allocate enough buffer space to send one full-size frame.
Unfortunately, in large networks this causes our update coalescing
strategy to fail, just where it is the most useful.  Allocate enough
buffers to fit a full update.

Yeah, it's a hack.

Conflicts:
	message.c

Reported by Thomas McLure.

This avoids the quadratic behaviour of expire_sources.

Reported by Thomas McLure.

Only apply the normal amount of hysteresis.

Reported by Dave Taht.

This changes the route table to be a sorted table of linked lists of
routes to a given prefix, which makes most route operations behave in
O(log n).  Insertion and flushing of a prefix is still O(n), but these
are fairly rare operations.

A nice side-effect is that the route table is now private to route.c,
which should make it easy to switch to a different data structure in
the future.

This avoids warnings about being unable to determine things about
nonexistent interfaces.

Setting the IPv6 dscp field to CS6 allows for babel packets to
'jump' a bufferbloated queue and get to a destination sooner.

With the pfifo_fast qdisc (the default for ethernet), this
'does the right thing'.

With mac80211 wireless, this moves babel packets into the VO
queue, from which it is hopefully extracted and then multicast.

Thanks to Matthieu Boutier.

Conflicts:
	CHANGES

IANA has changed Babel's allocation to 6696/UDP.

Older Mac OS X doesn't have strnlen.

This is an incompatible babelz protocol change.

We used to recompute this in parse_packet, which is clearly the
wrong place.