|[16 earlier articles]|
|Re: Garbage collection firstname.lastname@example.org (Nick Roberts) (2004-09-03)|
|Re: Garbage collection email@example.com (Sebastian) (2004-09-07)|
|Re: Garbage collection firstname.lastname@example.org (2004-09-13)|
|Re: Garbage Collection email@example.com (1992-08-09)|
|Re: Garbage Collection firstname.lastname@example.org (1992-08-11)|
|Re: Garbage Collection email@example.com (1992-08-12)|
|Re: Garbage Collection David.Chase@Eng.Sun.COM (1992-08-13)|
|Re: Garbage Collection firstname.lastname@example.org (1992-08-14)|
|Garbage collection Olin.Shivers@cs.cmu.edu (1992-11-24)|
|From:||David.Chase@Eng.Sun.COM (David Chase)|
|Date:||Thu, 13 Aug 1992 21:42:09 GMT|
> Correct me if I'm wrong, but you seem to be saying that you're
> using the stack and the registers as roots of garbage collection. ...
> If we decide, in an attempt to be _really_conservative_,
> to assume everything is a pointer, then we pay a double penalty: since
> we're wasting time moving junk around, garbage collection takes longer,
> and since we're not reclaiming storage that is actually free, we collect
> more often.
> It seems clear that this naive approach won't be satisfactory, so
> what can we do?
I'd suggest reading "Garbage Collection in an Uncooperative
Environment" by Boehm and Weiser in _Software Practice and Experience_,
Their naive and clearly unsatisfactory approach works quite well, in
Return to the
Search the comp.compilers archives again.