Path: ...!news.nobody.at!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: Lawrence D'Oliveiro <ldo@nz.invalid>
Newsgroups: comp.arch
Subject: Re: Continuations
Date: Mon, 15 Jul 2024 23:47:36 -0000 (UTC)
Organization: A noiseless patient Spider
Lines: 20
Message-ID: <v74ceo$tuk4$1@dont-email.me>
References: <v6tbki$3g9rg$1@dont-email.me>
	<memo.20240713174208.14216G@jgd.cix.co.uk> <v6v3fn$3povk$10@dont-email.me>
	<4j389jlb44kuhi3o5igf8ucnks3q9m3lc8@4ax.com> <v714in$1foh$2@gal.iecc.com>
	<lfims1Fl57mU1@mid.individual.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Injection-Date: Tue, 16 Jul 2024 01:47:36 +0200 (CEST)
Injection-Info: dont-email.me; posting-host="55f74b4ae2f116c131f41dc172ef4daf";
	logging-data="981636"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX19rQmjxzsLNPtCNLx9j2JIZ"
User-Agent: Pan/0.158 (Avdiivka; )
Cancel-Lock: sha1:g4wv9smrPsXTlIXoKISBBYB/lTc=
Bytes: 2168

On Sun, 14 Jul 2024 22:22:41 +0300, Niklas Holsti wrote:

> I think of continuation passing as always ending a subprogram by calling
> some other subprogram, passing it one or more "continuation" subprograms
> as parameters, and expecting that callee to do something and then to
> "continue" by further calling one of those "continuations" (with further
> "continuation" parameters) and NOT to return to the original caller,
> ever.

That’s the most general case. But it makes sense to have simpler 
conventions for common cases, like subroutine call/return. So a subroutine 
call implicitly passes a continuation that, when invoked, resumes the 
caller. And the subroutine block can end with an implicit invocation of 
this continuation. And such continuation objects are so common that it 
makes sense, as an optimization, to have a special place to allocate them 
in LIFO fashion--call it a “stack”.

You may have heard some people talk about “tree-like stacks”. This is 
starting to generalize from simple linear stacks to fully general 
continuations.