Written by Michael Richardson <mcr@xelerance.com> on 2005-09-10.

1)	we are pretty certain that we have not consistently broken SMP
	support in KLIPS. It compiled, but it almost certainly locks up. This
	is probably an improvement, since before, it only sometimes locked up,
	and we didn't really know why.

	We do not have an ETA on a fix, although it is important to us.

2)	We are aware of problems where more than 10 initiations/second may
	cause the system to fail to negotiate ANY SAs. We are not certain
	of the cause, but it is not made better by setting nhelpers=0.
	Note that our previous record for SA setups/second was 11/second, so
	it may well be availability of entropy.
	
3)	We are aware that there may be other problems, as yet not clearly
	documented.
	We think that this release is miles above the 2.3.1 release, and
	it is high time to get something out.

