GCC Undefined Behavior Sanitizer

[monarch_dodra via Digitalmars-d]

On Saturday, 18 October 2014 at 23:10:15 UTC, Ola Fosheim Grøstad 
wrote:
> On Saturday, 18 October 2014 at 08:22:25 UTC, monarch_dodra 
> wrote:
>> Besides, the code uses x + 1, so the code is already in 
>> undefined state. It's just as wrong as the "horrible code with 
>> UB" we wère trying to avoid in the first place.
>>
>> So much for convincing me that it's a good idea...
>
> Not sure if you are saying that modulo-arithmetic as a default 
> is a bad or good idea?

Op usually suggested that all overflows should be undefined 
behavior, and that you could "pre-emptivelly" check for overflow 
with the above code. The code provided itself overflowed, so was 
also undefined.

What I'm pointing out is that working with undefined behavior 
overflow is exceptionally difficult, see later.

> In D and (C++ for uint) it is modulo-arithmetic so it is 
> defined as a circular type with at discontinuity which makes 
> reasoning about integers harder.

What interesting is that overflow is only defined for unsigned 
integers. signed integer overflow is *undefined*, and GCC *will* 
optimize away any conditions that rely on it.

One thing I am certain of, is that making overflow *undefined* is 
*much* worst than simple having modulo arithmetic. In particular, 
implementing trivial overflow checks is much easier for the 
average developper. And worst case scenario, you can still have 
library defined checked integers.