NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/lang/c++-part2/15642

Internet Message Format | 1996-08-05 | 2.1 KB

Path: mail2news.demon.co.uk!genesis.demon.co.uk
From: Lawrence Kirby <[e-mail]>
Newsgroups: comp.lang.c++,comp.lang.c
Subject: Re: HELP: no mapping at the fault address
Date: Sun, 07 Apr 96 19:17:55 GMT
Organization: none
Message-ID: <[e-mail]>
References: <[e-mail]> <4k3qn8$[e-mail]> <[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] "Tom Wheeley" writes:

>In article <[e-mail]>
>           [e-mail] "Lawrence Kirby" writes:
>
>> In article <4k3qn8$[e-mail]>
>>            [e-mail] "Jianguo Liu" writes:
>>
...
>> >The current line is
>> >
>> >new_ptr=(strucut tree *) malloc(sizeof(struct tree));
>>
>> Did you include stdlib.h? You must ensure that all functions that don't
>> return int or are variadic are declared (it is best that all functions are
>> declared). If that isn't the problem your program probably corrupted the
>
>Would I be right in saying that if he had omitted the cast for the return
>value of malloc() then this error would have arisen as a warning in an
>ISO compiler?  With (void *) being `equivalent' to any old pointer etc.

Without a declaration in scope the compiler must assume that malloc() returns
int. Trying to convert implicitly from an int to a pointer (other than
a constant expression with value 0) violates a constraint so, yes, the
compiler must diagnose. void * isn't equilalent to other pointer types but
may be converted implicitly to and from other pointer types (except pointers
to functions) so the cast here can be omitted and probably should be since
its main effect is to cover up a potential bug.

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