Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ames!haven!ncifcrf!nlm-mcs!adm!smoke!gwyn
From: gwyn@smoke.BRL.MIL (Doug Gwyn )
Newsgroups: comp.std.c
Subject: Re: Sizes, alignments, and maxima
Message-ID: <9692@smoke.BRL.MIL>
Date: 23 Feb 89 05:41:12 GMT
References: <2508@ssc-vax.UUCP> <8943@alice.UUCP> <830@atanasoff.cs.iastate.edu> <11838@haddock.ima.isc.com>
Reply-To: gwyn@brl.arpa (Doug Gwyn (VLD/VMB) <gwyn>)
Organization: Ballistic Research Lab (BRL), APG, MD.
Lines: 22

In article <11838@haddock.ima.isc.com> karl@haddock.ima.isc.com (Karl Heuer) writes:
>Question for comp.std.c (to which I've redirected followups): I've been told
>that the Cray-2 has sizeof(int) == 8, yet INT_MAX == 0x7FFFFFFF (i.e. the
>arithmetic is only accurate to 4 bytes when using int).  Is this legal in a
>conforming implementation?  I think I can prove that UINT_MAX must be 2*^64-1,
>but I'm less sure about INT_MAX.  Section 3.1.2.5 has a restriction to binary
>architectures, which by the definition in the footnote seems to require every
>bit except the highest to represent a power of two; should this be interpreted
>as a requirement that 2*^63-1 must be representable in an 8-byte int?

I think an implementation such as you describe is legal.
What is required is that integers be represented in a binary
numeration system, that a nonnegative signed integer of a given size
have the same representation as the corresponding unsigned integer
with the same value, that all integers in the ranges given by INT_MAX
etc. be representable, and that unsigned arithmetic be performed
modulo UINT_MAX+1 (similarly for unsigned long, etc.).  I don't see
any way these requirements can be combined to "prove" that every bit
pattern contained in the "sizeof" soace has to be interpretable as a
valid integral value.  In fact I'm pretty sure we didn't want to
require that, since (as you note) some architectures really don't
support it.