RFR(M): 8223051: support loops with long (64b) trip counts

[Roland Westrelin]

> Maybe someone else will have comments 0n that, but assuming
> that is more or less unchanged, this change set for 8223051 on
> top still looks good, taking into account renaming of min/max
> factories.
>
> After you post an updated webrev.02 the review should be easy.
> Tobias might wish to run some regression tests on the final changes.

Thanks for the review. Here is the webrev with the min/max renaming and
the hunk that was erroneously included in 8244504.

http://cr.openjdk.java.net/~roland/8223051/webrev.02/

Roland.