[m-rev.] for review: optimizing mutually recursive tail calls in MLDS

[Zoltan Somogyi]

The logical reviewer for this diff would be Paul, but I don't think
I can ask him, so it is for review by anyone.

The diff passes bootcheck on my laptop, but I would like to ask Julien,
whether or not he reviews the change, to test it out in the Java and C#
grades. If they fail, I can still commit the diff, I would just need to turn off
the new algorithms via the new options.

Zoltan.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Log.mutual
Type: application/octet-stream
Size: 2902 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20170829/500324d3/attachment-0002.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: DIFF.mutual
Type: application/octet-stream
Size: 126913 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20170829/500324d3/attachment-0003.obj>