Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!leah!itsgw!steinmetz!uunet!ficc!peter
From: peter@ficc.uu.net (Peter da Silva)
Newsgroups: comp.lang.c
Subject: Re: C optimizer
Keywords: C pure function optimization
Message-ID: <3121@ficc.uu.net>
Date: 16 Feb 89 22:16:00 GMT
References: <515@larry.UUCP> <9648@smoke.BRL.MIL> <36034@bbn.COM>
Organization: Xenix Support
Lines: 26

In article <36034@bbn.COM>, mesard@bbn.com (Wayne Mesard) writes:
>   A function which is entirely composed of known "pure" operations*
>   is, itself, pure.

> * Where pure operations is defined as functions that produce no side
> effects and return deterministic values, and most operators (excluding
> assignment to dereferenced pointers and globals).

Sounds good.

> Hint: Sleep(3) would,
> I believe, be labelled "pure" under this definition, so something's
> still missing.

No, since I don't think *any* system call is definable as a pure function.
In this case alarm() returns a nondeterministic value *and* produces the
side effect of setting a timer in your process table entry.

Can anyone think of any pure system calls? ioctl(fd, TCGETA, tbuf)?
No, I think a system call should be considered not only a global ref
but a volatile one!
-- 
Peter da Silva, Xenix Support, Ferranti International Controls Corporation.
Work: uunet.uu.net!ficc!peter, peter@ficc.uu.net, +1 713 274 5180.   `-_-'
Home: bigtex!texbell!sugar!peter, peter@sugar.uu.net.                 'U`
People have opinions. Companies have policy. And typos are my own business.