Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!caen!hellgate.utah.edu!csn!pikes!aspen.craycos.com!jrbd
From: jrbd@craycos.com (James Davies)
Newsgroups: comp.lang.c
Subject: Re: casting void pointers
Message-ID: <1991Apr23.161345.4758@craycos.com>
Date: 23 Apr 91 16:13:45 GMT
References: <3121@enuxha.eas.asu.edu> <4256@rwthinf.UUCP>
Organization: Cray Computer Corporation
Lines: 14

In article <4256@rwthinf.UUCP> berg@marvin.e17.physik.tu-muenchen.de (Stephen R. van den Berg) writes:
>Roger A. Hurwitz writes:
>>Even K&R 2d states that an explicit cast should be used with malloc(),
>>in another part of the book (p. 142), but I am having trouble reconciling
>>that with the above language.
>
>malloc does not need the cast, any compiler complaining about that is
>braindamaged (or not ANSI compliant :-)

In fact, the cast can actually be damaging.  If you forget to declare
malloc on a 16-bit-int machine like a PC, but then cast the returned value,
most compilers will assume you know what you're doing and quietly strip off
the high 16 bits of your 32-bit pointer.  (I once spent two days finding
a bug like this...)