Would negative 0 be an issue in this situation?

user_2474248 0
Viewed 112

I have a simple question that is motivated by safe coding.

I encountered a situation where I have to assign the negative value of the remainder of division or 0 to an int variable, so I used the following notation:

int a = -(b % c);

where both b and c are positive integers. The expression (b % c) will yield a positive number or 0 and a will be assigned the negative value of the result.

In the event that (b % c) yields 0, what does -0 evaluate to? 0 or negative 0? and what are the implications if the latter happened?

P.S I don't know what negative 0 is.

language-lawyer division c
1 Answers

-(b % c) cannot generate a negative 0 because C 2018 6.2.6.2 3 says:

If the implementation supports negative zeros, they shall be generated only by:

— the &, |, ^, ~ , << , and >> operators with operands that produce such a value;

— the +, - , *, /, and % operators where one operand is a negative zero and the result is zero;

— compound assignment operators based on the above cases.

Even if we interpret the above to include - as a unary operator, we have been given that b and c are positive integers, so b % c cannot produce a negative zero, so the operand to - in -(b % c) is not a negative zero.

If == is used to compare a negative zero and a non-negative zero, it produces true, as == operates based on the values of the operands, not their representations.

Negative zeros are extremely rare for integers in modern C implementations. Modern implementations overwhelmingly use two’s complement, which does not have a negative zero. Only specialized/historic/archaic implementations might use the alternative integer representations that have negative zeros (one’s complement and sign-and-magnitude). Floating-point representations commonly have negative zeros, but they are equal to positive zeros in comparisons for equality by value (the ==, !=, <, >, <=, and >= operators).

user_2982250
Related