NAME

ilogb, ilogbf, ilogbl — an unbiased exponent

LIBRARY

Math Library (libm, -lm)

SYNOPSIS

#include <math.h>

int
ilogb(double x);

int
ilogbf(float x);

int
ilogbl(long double x);

DESCRIPTION

The ilogb(), ilogbf(), and ilogbl() functions return the exponent of the non-zero real floating-point number x as a signed integer value. Formally the return value is the integral part of

where r is the radix of the machine's floating-point arithmetic defined by the FLT_RADIX constant in <float.h>.

RETURN VALUES

As described above, upon successful completion, the functions return the exponent. Functionally this is the same as calling the corresponding logb(3) function and casting the return value to int.

The following special cases may occur:

1. If x is zero, the value of FP_ILOGB0 is returned and a domain error occurs.
2. If x is infinite, a domain error occurs and the value of INT_MAX is returned.
3. If x is NaN, a domain error is raised and the value of FP_ILOGBNAN is returned.
4. If the correct value is outside the range of the return type, a domain error occurs but an unspecified value is returned.

SEE ALSO

ilog2(3), logb(3), math(3)

STANDARDS

The described functions conform to ISO/IEC 9899:1999 (“ISO C99”).

BUGS

Neither FP_ILOGB0 nor FP_ILOGBNAN is defined currently in NetBSD.