new CARP implementation

[Gleb Smirnoff]

  Hello networkers,

  I'd like to present for review and early testing (for brave ones)
a new CARP implementation. The reason for this rewrite was that CARP
protocol actually doesn't bring a new interface, but is a property of
interface address. Rewriting it in this way helps to remove several
hacks from incoming packet processing[1], simplifies some code, makes
CARP addresses more sane from viewpoint of routing daemons such as
quagga/zebra. It also brings support for a single redundant address
on the subnet, the thing that is called "carpdev feature" in OpenBSD,
long awaited in FreeBSD.

More info and the patch itself is available here:

http://people.freebsd.org/~glebius/newcarp/README

I'm glad to here comments.

-- 
Totus tuus, Glebius.