Path: utzoo!attcan!uunet!lll-winken!ames!amdahl!pacbell!att!ihlpb!tainter
From: tainter@ihlpb.ATT.COM (Tainter)
Newsgroups: comp.lang.c
Subject: Re: malloc impossible?
Message-ID: <9408@ihlpb.ATT.COM>
Date: 14 Jan 89 21:44:05 GMT
References: <19@xenlink.UUCP> <225800106@uxe.cso.uiuc.edu> <310@twwells.uucp> <9342@smoke.BRL.MIL> <15427@mimsy.UUCP> <9351@smoke.BRL.MIL>
Reply-To: tainter@ihlpb.UUCP (55521-Tainter,J.A.)
Organization: AT&T Bell Laboratories - Naperville, Illinois
Lines: 29

In article <9351@smoke.BRL.MIL> gwyn@brl.arpa (Doug Gwyn (VLD/VMB) <gwyn>) writes:
>Assuming that integers could not also live in floating-point space,
>C would not be implementable on such a hypothetical architecture.

Sure it can.  Just not efficiently.

>Consider
>	union	{
>		double	d;
>		int	i;
>	}	u;
>	u.d = 123.0;
>	printf("%d\n", u.i);
>This is required to work, although the specific value printed of course
>depends on details of numeric representation on the specific system.
>All data object types in C must be able to live in the same kind of space.
>Whatever kind of space that is, is what one would parcel out via malloc().
>Therefore malloc() is implementable if C is (at the very least, along the
>lines I indicated in my previous example implementation).

This isn't even required to print anything useful.

To support this stuff you just give everything an indirection.

I haven't thought it completely through but I think this can be fleshed out
and even support malloc().


--johnathan.a.tainter