Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site cyb-eng.UUCP
Path: utzoo!linus!philabs!cmcl2!seismo!ut-sally!cyb-eng!howard
From: howard@cyb-eng.UUCP (Howard Johnson)
Newsgroups: net.lang.c
Subject: Re: makenode() gotcha
Message-ID: <785@cyb-eng.UUCP>
Date: Thu, 14-Nov-85 18:26:37 EST
Article-I.D.: cyb-eng.785
Posted: Thu Nov 14 18:26:37 1985
Date-Received: Sat, 16-Nov-85 07:29:27 EST
References: <2482@brl-tgr.ARPA> <118@bsdpkh.UUCP>
Distribution: net
Organization: Cyb Systems, Austin, TX
Lines: 50

A fix to the following 'test' program still has a slight bug.
As a matter of design, adding a simple function removes all kinds
of nasty portability issues (i.e., sizes of pointers, paramter
passing, etc.).

>> struct node { int op; struct node *left, *right; double val; }
>> 	*lp, *rp, *p, *makenode();
>> double value;			/* constant value from lex */
>> ...
>> p = makenode( 'x', lp, rp );	/* create multiply node */
>> ...
>> p = makenode( 'k', value );	/* create real constant node */
>> ...
>> struct node *makenode( op, arga, argb )
>> int op; struct node *arga, *argb;
>> {	extern char *malloc();
>> ...
>> 	case 'k':		/* real constant */
>> 		new->val = *(double *)&arga;	/* XXX */
>> 		break;
>> 	}
>> 	return new;
>> }
>> 
>> The code failed at point "XXX".  The exercise for the
>> student is to figure out precisely WHY it fails ......

The attempted fix needs its indirection:

> 	p = makenode ('k', &value );
	...
	new->val = *((double *)arga); /* you forgot the indirection */

Better!

struct node *
makeconst(value)
double value;
{
	register struct node *new;

	if ((new = (struct node *) malloc(sizeof *new)) != NULL) {
		new->op = 'k';
		new->val = value;
	}
	return(new);
}
-- 
..!{seismo,topaz,mordor,harvard,gatech,nbires,ihnp4}!ut-sally!cyb-eng!howard
(ordered best to worst); also ..!ut-ngp!cyb-eng!howard  +1 512 458 6609