Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!rutgers!princeton!allegra!alice!bs
From: bs@alice.UUCP
Newsgroups: comp.lang.c++
Subject: bug re unsigned to signed conversion?
Message-ID: <6965@alice.UUCP>
Date: Mon, 8-Jun-87 14:20:29 EDT
Article-I.D.: alice.6965
Posted: Mon Jun  8 14:20:29 1987
Date-Received: Wed, 10-Jun-87 07:18:36 EDT
Organization: AT&T Bell Laboratories, Murray Hill NJ
Lines: 85
Keywords: bugs



 > From: eppstein@garfield.columbia.edu.UUCP (eppstein @ Columbia University CS Department) 
 > The following program prints both messages.  I would expect it to print
 > neither, but at least the second message shouldn't happen.  The
 > generated C code for the two lines is essentially identical; no doubt
 > the second line and probably the first should have generated a cast to int.
 > 
 > #include <stream.h>
 > 
 > int a = -1;
 > unsigned b = 2;
 > 
 > inline int conv() { return b; }
 > 
 > main()
 > {
 >     if (a > b) cout << "(int) -1 > (unsigned) 2 ???\n";
 >     if (a > conv()) cout << "(int) -1 > (int) (unsigned) 2 ????????\n";
 > }

It is clearly a cfront bug that ``conv()'' is expanded into ``b'' and not
``(int)b''. Sorry. For people with release 1.2 here is a fix (a new ck_cast()
for expand.c):

int ck_cast(Ptype t1, Ptype t2)
/*
	return a value of type t2 from a function returning a t1
	return 1 if cast is needed
*/
{
	while (t1->base == TYPE) t1 = Pbase(t1)->b_name->tp;
	while (t2->base == TYPE) t2 = Pbase(t2)->b_name->tp;

	if (t1 == t2) return 0;

	if (t1->base != t2->base) return 1;

	switch (t1->base) {
	case CHAR:
	case SHORT:
	case INT:
	case LONG:
		if (Pbase(t1)->b_unsigned != Pbase(t2)->b_unsigned) return 1;
	}

	if (t1->base == COBJ) {
		Pname nn = Pbase(t1)->b_name;

		if (Pclass(nn->tp)->csu==UNION ) return 0; 

		if (t2->base==COBJ && nn->tp==Pbase(t2)->b_name->tp) return 0;

		return 1;
	}

	return 0;
}

I'll leave the question of the correct value of ``a>b'' to your local C
compiler until the ANSI C standard finally appears. Interestingly enough
my C compiler causes this program

main()
{
	int a = -1;
	unsigned b = 2;

	if (-1>(unsigned)2)
		printf("true\n");
	else
		printf("false\n");

	if (a>b)
		printf("true\n");
	else
		printf("false\n");
}

so that it produces

false
true

which is clearly a bug.