> Wait a minute...
this is an interesting problem :))
the CPU integrates an overflow check internally, so you can tap into
this using assembler - but not so easially using C/C++ :) for an
obfuscated example, check out the following code:
in gcc (PRC-Tools), it would look like this:
{
UInt32 a, b;
Boolean result = false;
// push CPU regs a0, d0+d1 on stack
asm("movem.l %%a0/%%d0-%%d1, -(%%sp)" : : );
asm("move.l %0, %%a0" : : "g" (&result));
asm("move.l %0, %%d0" : : "g" (a));
asm("move.l %0, %%d1" : : "g" (b));
asm("add.l %%d0, %%d1 | do the addition
bvc NOOVERFLOW | check the overflow flag
ori.b #1, (%%a0) | set our overflow boolean
NOOVERFLOW:
" : : );
// pop CPU regs a0, d0+d1 from stack
asm("movem.l (%%sp)+, %%a0/%%d0-%%d1" : : );
// process :)
if (result) {
// do whatever you need here (if a+b > 0xffffffff)
}
}
> > > (a + b < a || a + b < b)?
well.. if a and b are both unsigned integers, (always positive),
then you would only need to check one condition:
if (a+b < a) {
}
if b is 0, condition remains false. if b is too high, the result
should wrap around and a will be larger.. hence, overflow.
cheers
// az "mr. obfuscation"
[EMAIL PROTECTED]
http://www.ardiri.com/ <--- free games!
--
For information on using the Palm Developer Forums, or to unsubscribe, please see
http://www.palmos.com/dev/tech/support/forums/
