Speeding up GC

[Richard Henderson]

On Mon, Jun 03, 2002 at 03:25:05PM +0200, Martin v. Löwis wrote:
> While it is, in general, not possible to determine whether an object
> has changed, I believe that for certain objects (in particular DECL
> nodes), there is a point in time when they are "frozen", meaning that
> they won't change anymore. I suggest to record this fact in the
> object (with the opportunity to unfreeze the object).

A better way to do this is to actually compact the GC arena
and mark the pages read-only.  If the objects are modified,
the SIGSEGV handler marks the page read-write again and flags
the page as having been modified.

This is called "generational GC".  It has been discussed on
the list off and on.

It's on-hold waiting for Geoff Keating's GC rewrite, since
at the moment we can't compact the GC arena.


r~