NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/lang/c-part1/3562

Internet Message Format | 1996-08-05 | 1.6 KB

Path: mail2news.demon.co.uk!genesis.demon.co.uk
From: Lawrence Kirby <[e-mail]>
Newsgroups: comp.lang.c
Subject: Re: Undefined behavior? on type conversion, was: Re: Hungarian notation
Date: Mon, 29 Jan 96 23:23:00 GMT
Organization: none
Message-ID: <[e-mail]>
References: <[e-mail]> <4d2ok0$[e-mail]> <[e-mail]>
Reply-To: [e-mail]
X-NNTP-Posting-Host: genesis.demon.co.uk
X-Newsreader: Demon Internet Simple News v1.27
X-Mail2News-Path: genesis.demon.co.uk

In article <[e-mail]>
           [e-mail] "James Kanze US/ESC 60/3/141 #40763" writes:

(Newsgroups trimmed)

>Does it?  I believe that it was the intent of the authors that the
>`result' of the conversion could be a signal (for example) or a core
>dump.

I disagree!

> (This is, of course, the only reasonable thing for an
>implementation to do.)

Possibly but existing practice won out here. The normal approach is for the
compiler to pick ot the relevant low order bits and ignore the higher order
ones. That also has the effect of preserving value in a
signed->unsigned->signed conversion although the standard gives no guarantees
about this which leaves a problem with char handling: can you portably assign
the return value of getchar() (non-EOF values can be represented in an
unsigned char) to a char which may be signed?

--
-----------------------------------------
Lawrence Kirby | [e-mail]
Wilts, England | [e-mail]
-----------------------------------------