Xref: utzoo comp.lang.lisp:2676 comp.lang.misc:3935 comp.lang.smalltalk:1640
Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!tut.cis.ohio-state.edu!att!cbnewsc!lgm
From: lgm@cbnewsc.ATT.COM (lawrence.g.mayka)
Newsgroups: comp.lang.lisp,comp.lang.misc,comp.lang.smalltalk
Subject: Re: Garbage collection algorithms
Message-ID: <13022@cbnewsc.ATT.COM>
Date: 23 Jan 90 00:57:11 GMT
References: <1990Jan22.111348.5585@Neon.Stanford.EDU>
Reply-To: lgm@cbnewsc.ATT.COM (lawrence.g.mayka,ihp,)
Organization: AT&T Bell Laboratories
Lines: 22

In article <1990Jan22.111348.5585@Neon.Stanford.EDU> wilson@carcoar.Stanford.EDU (Paul Wilson) writes:
>  Lately several groups have been working on conservative mark-sweep
>  gc's that can deal with the kind of pointer ambiguities you get
>  in languages like C.  They even have generational versions.  People

Just a clarification:

My understanding is that such algorithms do not reclaim any
storage that seems to be "pointed at" by any integer or other
datum that happens to look like a pointer.  Similarly, such a
garbage collector cannot move storage contents in memory (e.g., to
avoid fragmentation) because it cannot tell the difference between
genuine pointers to the moved contents (which would have to be
updated) and false pointers (integers or other data that must be
left alone).


	Lawrence G. Mayka
	AT&T Bell Laboratories
	lgm@ihlpf.att.com

Standard disclaimer.