PDA

Bekijk Volledige Versie : Re: gcc 4.1 bug miscompiles pointer range checks, may place you



jat-public01@jaet.org
18/04/06, 21:40
Are you certain that should fail?

(unsigned long)-1 is a word with all bits set (on a twos-complement machine), so I believe the result should be undefined with regard to overflow adding a pointer.

It certainly seems reasonable for a compiler to optimize away a test for a pointer in the range of p to p+MAXINT-1, if p has the same number of bits as MAXINT.

If you really want to test for negative buffer sizes, you need to declare the length as long rather than unsigned long.

John Tamplin