[m-rev.] for post-commit review: kind-specific statuses for insts and modes

[Zoltan Somogyi]

The next step towards kind-specific status types.

After this diff, the compiler will do all operations on the statuses
of insts and modes twice, on the old and the new representations,
and abort if the results don't agree. I bootchecked this diff in
three grades without such an abort, so I am willing to commit it,
but if anyone gets an abort, they can temporarily disable the
cross-check and just go with the results from the old representation
as a workaround until I find and fix the underlying problem for that
test case.

Zoltan.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Log.status2
Type: application/octet-stream
Size: 1248 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20150920/cd7b1373/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: DIFF.status2
Type: application/octet-stream
Size: 37401 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20150920/cd7b1373/attachment-0001.obj>