Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!gem.mps.ohio-state.edu!ginosko!uunet!ficc!peter
From: peter@ficc.uu.net (Peter da Silva)
Newsgroups: comp.lang.c
Subject: Re: Coroutines in C
Message-ID: <5899@ficc.uu.net>
Date: 28 Aug 89 13:09:08 GMT
References: <214@castle.ed.ac.uk>
Organization: Xenix Support, FICC
Lines: 34

In article <214@castle.ed.ac.uk>, sam@lfcs.ed.ac.uk (S. Manoharan) writes:
> In article <5773@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes:
> >In many C compilers I wouldn't even need to write any assembly code, since
> >setjmp/longjmp can be used to implement the context switch.

[ problem deleted to save space ]

> Intutively, it appears the co-routines could be applied in this
> context. Am I correct? 

I believe so.

> And then, is there an easy way to tackle this problem? 

Have a look at my article a couple of weeks back where I introduced  this
subject, then implement coroutines :->.

In existing, portable, C... no.

> Since there could be a number of event?()s, I think setjmp() 
> and longjmp() would be of no help.

The naive implementation of setjmp() is to save all the registers in jmp_buf.
The corresponding naive implementation of longjmp is to restore those registers
and return. If this implementation is used, it should be possible to create
a fake stack frame and register set in a jmp_buf and tranfer control to it
using longjmp. This is, of course, hideously non-portable. But if the calling
sequence for coroutines can be standardized then this non-portable code could
be hidden inside co_call() and friends.
-- 
Peter da Silva, *NIX support guy @ Ferranti International Controls Corporation.
Biz: peter@ficc.uu.net, +1 713 274 5180. Fun: peter@sugar.hackercorp.com. `-_-'
"Just once I'd like to meet an alien menace that isn't immune to bullets"  'U`
   -- The Brigadier, Dr Who.