Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site utcsri.UUCP
Path: utzoo!utcsri!flaps
From: flaps@utcsri.UUCP (Alan J Rosenthal)
Newsgroups: comp.sys.amiga
Subject: Re: Is this a Manx Bug?
Message-ID: <4050@utcsri.UUCP>
Date: Mon, 2-Feb-87 17:52:03 EST
Article-I.D.: utcsri.4050
Posted: Mon Feb  2 17:52:03 1987
Date-Received: Wed, 4-Feb-87 05:40:25 EST
References: <958@tekred.TEK.COM>
Reply-To: flaps@utcsri.UUCP (Alan J Rosenthal)
Organization: University of Toronto
Lines: 31
Keywords: auto variables
Summary: 


In article <958@tekred.TEK.COM> joels@tekred.UUCP writes:
>later I was reading my C manual and it said that all 'auto' variables
>should be automaticly initialized to zero at every entry to the subroutine.
>The UNIX C does this, but Manx does not. Is this a Manx bug?

As Matt pointed out, the initial value in this case is undefined (see the
middle of page 198 of K&R).

UNIX C does not do this.  However, when memory is allocated to a process in
the first place it is zeroed.  Therefore many things which are supposed to
contain "garbage" (such as uninitialized auto variables) in fact will contain
zero SOMETIMES.  Counting on this is stupid!  In addition, this is generally
not the case with micros, as the reason for this zeroing of memory is to
protect users from having other users see their memory contents just
deallocated.

(An example of a devious use possible on some mainframes if they didn't zero
memory (but not applicable to unix) is allocating a large amount of memory
right after someone logs in & looking through it to see if their password is
there.)

-- 

Alan J Rosenthal

UUCP: {backbone}!seismo!mnetor!utcs!flaps, ubc-vision!utai!utcs!flaps,
      or utzoo!utcs!flaps (among other possibilities)
ARPA: flaps@csri.toronto.edu
CSNET: flaps@toronto
BITNET: flaps at utorgpu