[Openmcl-devel] Re: [Bug-openmcl] Apparent bug in the objc bridge

[Gary Byers]

On Wed, 10 Nov 2004, Dan Knapp wrote:

>    Hmm okay.  So it actually is a bug, then, and thus probably not
> something that needs to be documented.

Yes.  The Linux case seems to work (there are lots of little differences
in the C ABIs between Darwin/Linux), but the Darwin case is pretty
confused.

> But I can still document
> the implicit dynamic-extent stuff - do you know, offhand, exactly
> what situations it happens in?

I think that it's limited to :ADDRESS -> MACPTR (which is documented)
and to :SINGLE-FLOAT -> SINGLE-FLOAT and :DOUBLE-FLOAT -> DOUBLE-FLOAT
(which aren't.)  In theory, it -might- happen on coercion to 32- or
64-bit integers, but I don't think that this happens in practice.