2010/1/20 Randal L. Schwartz <[email protected]>:
>>>>>> "Randal" == Randal L Schwartz <[email protected]> writes:
>
> Randal> Parrot's GC suffers from having many different kinds of mixed objects.
>
> Randal> http://wknight8111.blogspot.com/2010/01/boehm-in-parrot.html
>
> And worse, having untyped data blobs on a stack... that's probably the worst
> thing there.
>
> If you can't tell what live pointers you have for sure, you can't GC, and if
> you can't GC reliably, you lose.

Boehm GC is conservative one. Not a good way to achieve good performance.
Mixing object pointers and non-object pointers on stack/object
contents is different story.
If you providing a good way to determine if given memory chunk
cointains oops, or just raw bytes (like compiled method or bytearray
etc),
then this is not a problem, since you still can use precise GC.


>
> --
> Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
> <[email protected]> <URL:http://www.stonehenge.com/merlyn/>
> Smalltalk/Perl/Unix consulting, Technical writing, Comedy, etc. etc.
> See http://methodsandmessages.vox.com/ for Smalltalk and Seaside discussion
>
> _______________________________________________
> Pharo-project mailing list
> [email protected]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>



-- 
Best regards,
Igor Stasenko AKA sig.

_______________________________________________
Pharo-project mailing list
[email protected]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Reply via email to