VM performance

The performance of the VM code in late 2.4 (>=2.4.10) is very good.
However as said the most important reason for the rewrite was to move the 2.2 boxes to 2.4 without risking them to fall apart over the time.
Mike Galbraith did some nice benchmarking with a swap testcase `fresh boot -> time make -j30 bzImage`.
2.4.13-ac6
real    8m15.366s 
user    6m35.710s
sys     0m33.570s
2.4.17-pre1aa1
real    7m39.066s
user    6m38.400s
sys     0m29.140s
+7%