[m-rev.] for post-commit review: fix the calculation of max ho arg size

[Zoltan Somogyi]

For review by anyone.

This bug fix is effectively a change in the semantics of the
--higher-order-arg-limit option, back to the documented semantics
from something else. This *could* be worth a NEWS.md entry,
but the attached diff between the stage2 of the bootcheck of this diff
and the stage2 of my previous bootcheck shows minimal changes,
aside from the direct effects of this diff and the previous one (which was
not in the previous bootcheck). I think users don't need such an entry,
though I could be persuaded otherwise.

Zoltan.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Log.ho2
Type: application/octet-stream
Size: 515 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20240415/297d7ead/attachment-0002.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: DIFF.ho2
Type: application/octet-stream
Size: 3643 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20240415/297d7ead/attachment-0003.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: max_size_diff.gz
Type: application/x-gzip
Size: 131932 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20240415/297d7ead/attachment-0001.bin>