[m-rev.] for post-commit review: replace cons/3

[Zoltan Somogyi]

For review post commit because the concept has been reviewed,
and the implementation is *mostly* straightforward. The exceptions
are all in modules that have their own entries in the log message.
The diff is with -b.

I ended up going with Julien's suggestion of du_data_ctor/1
instead of heap_ctor/1, because in the earliest compiler stages,
that function symbol is used not just for du type's constructors
and the tuple constructor, but also for char constants, for which
heap_ctor would be misleading. (Even more so than for enums,
which at least are a special case of something that does use
the heap in the general case.)

Zoltan.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Log.cons
Type: application/octet-stream
Size: 5656 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20240630/e5f9c277/attachment-0002.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: DIFF.cons
Type: application/octet-stream
Size: 252326 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20240630/e5f9c277/attachment-0003.obj>