Standards
| Standards / Extensions |
C or C++ |
Dependencies |
|---|
C99
Single UNIX Specification, Version 3
C/C++ DFP
C++ TR1 C99
|
both
|
z/OS V1R8
|
Format
#define _ISOC99_SOURCE
#include <math.h>
int fpclassify (real-floating x);
#define __STDC_WANT_DEC_FP__
#include <math.h>
int fpclassify([real-floating|decimal-floating] x);
#define _TR1_C99
#include <math.h>
int fpclassify(real-floating x);
General description
This macro or function
template classifies its argument value as NaN, infinite, normal, subnormal
or zero based on the type of its argument. If the argument is represented
in a format wider than its semantic type, then it is converted to
its semantic type and then it is classified.
| Function |
Hex |
IEEE |
|---|
| fpclassify |
X |
X |
Notes: - To use IEEE decimal floating-point, the hardware must have the
Decimal Floating-Point Facility installed.
- This function works in IEEE decimal floating-point format. See
"IEEE Decimal Floating-Point" for more information.
Returned value
fpclassify() returns:
- FP_NAN if the argument is Not-a-Number.
- FP_INFINITE if the argument is plus or minus infinity.
- FP_ZERO if the argument is of value zero.
- FP_SUBNORMAL if the argument is too small to be represented in
the normal format.
- FP_NORMAL if none of the above.
Special behavior in hex: - FP_ZERO if the argument is of value zero.
- FP_NORMAL if the argument is a normalized number.
- FP_SUBNORMAL if the argument is an unnormalized number.