std::abs(int) ambiguous?

[Helmut Jarausch]

Hi

I cannot find documentation whether

#include <cmath>
using std::abs;

int main() {
int N;
return abs(N);
}

should compile or not.

Currently, gcc-4.3.3 signals an ambiguous overload problem:
absT.C: In function 'int main()':
absT.C:6: error: call of overloaded 'abs(int&)' is ambiguous
/usr/lib/gcc/x86_64-pc-linux-gnu/4.3.3/include/g++-v4/cmath:107: note: candidates are: long double std::abs(long double)
/usr/lib/gcc/x86_64-pc-linux-gnu/4.3.3/include/g++-v4/cmath:103: note: float std::abs(float)
/usr/lib/gcc/x86_64-pc-linux-gnu/4.3.3/include/g++-v4/cmath:99: note: double std::abs(double)

Is this a bug in gcc-4.3.3 or is it conforming to the standard.
It the latter, how to use 'abs' for all numeric types?

Thanks for a hint,

Helmut Jarausch

Lehrstuhl fuer Numerische Mathematik
RWTH - Aachen University
D 52056 Aachen, Germany

[Bo Persson]

Post by Helmut Jarausch
Hi
I cannot find documentation whether
#include <cmath>
using std::abs;
int main() {
int N;
return abs(N);
}
should compile or not.
absT.C:6: error: call of overloaded 'abs(int&)' is ambiguous
note: candidates are: long double std::abs(long double)
note: float std::abs(float)
note: double std::abs(double)
Is this a bug in gcc-4.3.3 or is it conforming to the standard.
It the latter, how to use 'abs' for all numeric types?

It is a bit odd, but std::abs for integer types are found in
<cstdlib>, not in <cmath>.


Bo Persson

[litb]

Post by Helmut Jarausch
Hi

[...]

Post by Helmut Jarausch
Is this a bug in gcc-4.3.3 or is it conforming to the standard.
It the latter, how to use 'abs' for all numeric types?

Try including <stdlib> instead.

[Neil Butterworth]

Post by Helmut Jarausch
Hi
I cannot find documentation whether
#include <cmath>
using std::abs;
int main() {
int N;
return abs(N);
}
should compile or not.

Not without the ambiguity warnings you noted. The int version of abs()
actually lives in <cstdlib> per section 26.5 of C++ Standard.

Neil Butterworth