[PATCH] Constant fold -A - B as -B - A (take 2)
Fergus Henderson
fjh@cs.mu.OZ.AU
Mon Feb 10 16:27:00 GMT 2003
On 10-Feb-2003, Richard Kenner <kenner@vlsi1.ultra.nyu.edu> wrote:
> > There is *no* language, as far as I know, for which MAX_INT + 1 is
> > defined, let alone equal to MIN_INT.
>
> What about Java, C#, and Visual Basic?
>
> I stand corrected.
>
> This indicates that we really ought to add a bit to types to indicate
> whether or not overflow is defined. Right now that's the unsigned
> bit, but, from what you say, that's clearly wrong.
IMHO it would be more natural to add the bit to operations
(e.g. to PLUS_EXPR tree nodes) rather than to types.
Actually there are more than two possibilities:
1. overflow wraps
2. overflow leads to undefined behaviour
3. overflow is detected, and leads to a
trap/exception/signal/whatever.
--
Fergus Henderson <fjh@cs.mu.oz.au> | "I have always known that the pursuit
The University of Melbourne | of excellence is a lethal habit"
WWW: <http://www.cs.mu.oz.au/~fjh> | -- the last words of T. S. Garp.
More information about the Gcc-patches
mailing list