From: Tim Wilkinson (tim@transvirtual.com)
Date: Thu Dec 03 1998 - 18:59:01 EST
I've read the paper some time back and am aware of the principle (essentially
the read-only pages act as a write barrier while the GC is running). The
solution is atractive because it's C transparent - but it's not terrible
portable - but since Kaffe is primarily a UNIX system currently it would at
least work on all UNIX platforms.
Tim
Alexandre Oliva wrote:
> On Dec 3, 1998, Archie Cobbs <archie@whistle.com> wrote:
>
> > They do present performance figures, but I haven't gotten that far yet :-)
> > The claim is that performance is no worse than normal "stop-the-world"
> > GC..
>
> Well, if they only mark as read-only pages that they have already
> scanned in a certain pass, I can kind of believe it. But I wonder if
> we could implement such a mechanism portably...
>
> --
> Alexandre Oliva http://www.dcc.unicamp.br/~oliva aoliva@{acm.org}
> oliva@{dcc.unicamp.br,gnu.org,egcs.cygnus.com,samba.org}
> Universidade Estadual de Campinas, SP, Brasil
-- Tim Wilkinson Tel: +1 510 704 1660 Transvirtual Technologies, Inc., Fax: +1 510 704 1893 Berkeley, CA, USA. Email: tim@transvirtual.com
This archive was generated by hypermail 2b29 : Sat Sep 23 2000 - 19:57:05 EDT