Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!seismo!rochester!rutgers!labrea!rocky!rokicki
From: rokicki@rocky.STANFORD.EDU (Tomas Rokicki)
Newsgroups: comp.sys.amiga
Subject: Manx (Pointer bug?)
Message-ID: <475@rocky.STANFORD.EDU>
Date: Fri, 7-Aug-87 11:22:07 EDT
Article-I.D.: rocky.475
Posted: Fri Aug  7 11:22:07 1987
Date-Received: Sun, 9-Aug-87 02:03:01 EDT
Organization: Stanford University Computer Science Department
Lines: 22

C'mon, folks, let's check our facts before we post!
In each of Manx 3.20a, 3.40a, and 3.40b, the code
fragment

char *p ;
s() {
   if (p==0) ;
   if (!p) ;
}

compiles correctly, with a tst.l _p, bne pair for
each.  There is no NULL bug in Manx, at least not
in 3.20a, 3.40a, or 3.40b; I've never had any
problems.  So let's not spread this any further.

The main reason that NULL might be set to 0L
instead of 0 is so you can pass NULL as a pointer
argument to a function.  That is why Manx has
NULL defined as 0L in each stdio.h file I have
received from them.

-tom