Configure patch for vfork et al on VMS

[Douglas B. Rupp]

VMS has the unique distinction of being considered a feature as well as an OS (at least in gcc), so it is not a violation of
protocol to test for it if there isn't a better way.

Also the test isn't really valid for VMS because it purports to test the maximum
length of the command line, but is actually testing the maximum length of a single argument. The maximum length of the command line
on VMS is unlimited AFAIK, so we can
hardcode it to a very large value and probably not have to ever change it.

There are no scripts already shipped to worry about, Gcc configure is the first.