The presumed bug where the value computed without the sign overflows
even though the negative result can be represented is not actually a
problem, because this only happens with signed results and the temporary
value is computed as unsigned (thus with extra range).
These macros cannot be tested properly until support for *printf and
*scanf is added (support for *printf currently exists but with a
temporary implementation), so this is LDEPS.
The size of the types has been checked though, so the definitions here
should be correct, the rest is printf/scanf logic.
