From: rhg AT cps DOT com (Richard H. Gumpertz)
Subject: Re: strlen on a NULL
13 May 1998 00:44:32 -0700
Message-ID: <199805121632.LAA18276.cygnus.gnu-win32@Pooh.cps.com>
References: <3557E6C4 DOT EF2F27B3 AT kiwiplan DOT co DOT nz>
To: ian AT kiwiplan DOT co DOT nz (Ian Collins)
Cc: gnu-win32 AT cygnus DOT com

> If I do a strlen on a NULL pointer I get a coredump.
> I have the same code running on a few *other* Unix machines with
> exhibiting this behaviour.
> OK, so I could tidy it up, but I wondered is this an oversight or the
> correct behaviour?

Sounds like correct behavior to me.  The result of using strlen (or any other
function that indirects through a pointer and does not explicitly define
behavior for NULL) is undefined and therefore implementation dependent.  The
most reasonable behavior for a strlen implementation is probably to do
whatever makes the WELL-DEFINED uses of strlen as fast as possible.

-- 
 ==========================================================================
 | Richard H. Gumpertz, PhD        VOICE:(913) 642-1777 or (913) 534-5062 |
 | Computer Problem Solving        FAX:  (913) 752-6615 or (913) 642-8777 |
 | 4518 West 89th Street, Suite 220                                       |
 | Prairie Village, KS 66207-2293                     E-MAIL: rhg AT CPS DOT com |
 ==========================================================================
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request AT cygnus DOT com" with one line of text: "help".