🚀 go-pugleaf

In article <21624@pasteur.Berkeley.EDU> jac@modoc.berkeley.edu (Gordon Jacobs) writes:
>The program below demonstrates various uses of the "?"
>selection operator in c that seem to not work correctly on
>a DEC3100 running Ultrix3.1.  This has also been compiled
>on a unit running a test version of Ultrix4.0 with the 
>exact same results.  This seems like a fairly major bug and
>it could affect much old software.  Has anyone experienced
>this or confirmed it??  The test program shows version which
>work and don't work...
>
>Thanks,
>Gordie Jacobs,  UCB  
>------------------ cut here --------------------------------------
>
>#define MAX(A,B)  	((A)>(B))?(A):(B)
>
>#define BITS_PER_INT 	(sizeof(int)*8)
>#define INTS_PER_BUS 	((int)(20/BITS_PER_INT)+1)
>
>main () {
>
>	int	high, low;
>	int	bits_per_int, sizeofint;
>
>	printf("INTS_PER_BUS=%d BITS_PER_INT=%d\n",
>	       INTS_PER_BUS, BITS_PER_INT);
>
>	bits_per_int= BITS_PER_INT;
>	sizeofint= sizeof(int);
>
>	high= 19;
>	
>	/* All of these compute the wrong answer (-12 instead of 0) */ 
>	low = MAX(0, high-BITS_PER_INT+1);
>	low = (0>(high-BITS_PER_INT+1))?(0):(high-BITS_PER_INT+1);
>	low = MAX(0, high-sizeof(int)*8+1);
>	low = (0>(high-sizeof(int)*8+1))?(0):(high-sizeof(int)*8+1);
>
>	/* Works:
>	#define BITS_PER_INT 32
>	low = MAX(0, high-BITS_PER_INT+1);
>	*/
>
>	/* Works:
>	low = MAX(0, high-sizeofint*8+1);
>	*/
>
>	/* Works:
>	   low = MAX(0, low=high-BITS_PER_INT+1);
>	   */
>	
>	/* Works:
>	   low = (0>(high-bits_per_int+1))?(0):(high-bits_per_int+1);
>	   */
>	
>	/* Works:
>	   low = (0>(high-32+1))?(0):(high-32+1);
>	   */
>	
>	/* Works:
>	   low = high-BITS_PER_INT+1;
>	   low = MAX(0,low);
>	   */
>	
>	printf("high=%d low=%d MAX(0,-12)=%d\n", high, low, MAX(0,-12));
>}

The problem is the sizeof() operator is unsigned, therefore the expression
of the form
	(int) - (unsigned) + (int)
yields an unsigned result.  This result just happens to be a very large
number which cannot accurately be represented by an int.  This number is
assigned to an int, which causes the upper bit to be treated as a sign
bit, and the value appears to be -12.

A quick fix is to do

	#define SIZEOF(x) ((int)sizeof(x))

and to use SIZEOF instead of sizeof() in the program.
-- 
        "What is a DJ if he can't scratch?"  - Uncle Jamms Army
     David Lai (lai@mips.com || {ames,prls,pyramid,decwrl}!mips!lai)