[plt-scheme] FFI and pointer manipulations

From: Matthew Flatt (mflatt at cs.utah.edu)
Date: Sun Feb 11 17:02:55 EST 2007

At Sun, 11 Feb 2007 13:58:09 -0500, "Christopher Armstrong" wrote:
> On 2/11/07, Jim Blandy <jimb at red-bean.com> wrote:
> > Many systems with copying GC's represent "large" objects (big strings;
> > long vectors) as small header objects that hold a pointer to the large
> > body of the object.  The header is allocated in the ordinary heap, and
> > does get copied by collections; the body is allocated elsewhere, and
> > stays put.
> Yeah. The blame is clearly on the GC, and so it seems the GC should
> become more flexible to allow specifying different strategies for
> different objects (or arenas of objects). This FFI is only going to be
> useful for simple wrappers until this happens.

That's a dire prediction that seems inconsistent with our experience.
MrEd works with the same GC technology, for example, and I don't think
it qualifies as a simple wrapper.

Things could be better, but the current situation is not quite as bad
as that.


Posted on the users mailing list.