On Wed, Oct 08, 2003 at 12:49:27PM -0700, cgd@broadcom.com wrote:
> At Wed, 8 Oct 2003 19:31:12 +0000 (UTC), "Andrey Petrov" wrote:
> > I used 3.3.1 and this example is all right on sparc64 too.
> > here is my test
> > 
> > int test2(int *p)
> > {
> > 
> > 	if (((*p >> 16) & 0xffff) == 123)
> > 		return 1;
> > 
> > 	return 2;
> > }
> > [ ... ]
> > Interesting that gcc indeed does access it as a halfword but still wants
> > to apply mask. It certainly looks like gcc optimizes this, just not
> > always successfully. Seems that I got to take back my performance argument:-)
> 
> it accessed it as a signed halfword.  if you change your source to:
> 
> int test2(int *p)
> {
> 
>       if ((unsigned short)((*p >> 16) & 0xffff) == 123)
>               return 1;
> 
>       return 2;
> }
> 

Yeah, it's signedness, I just changed argument
int test2(unsigned int *p)

and got it optimised

	lduh	[%o0], %g1
	cmp	%g1, 123
	mov	1, %o0
	retl
	movne	%icc, 2, %o0

Quite cool, I should take back my gcc blame.

	Andrey