bad TCP change (was Re: sf_buf patch for review)

[Jeffrey Hsu <hsu@xxxxxxxxxxx>]

  > Nobody port the network DOS attack patch that was just committed to
  > FreeBSD-5's ip_icmp.c.

I've already been asked privately about it and I replied that was a bad change.
That FreeBSD TCP patch is

	A.  A violation of the TCP protocol.
	B.  Confused about the context-switching cost of wakeups.
	C.  Protection for a non-existent problem.

								Jeffrey