significand.3m (2010 09)

s
significand(3M) significand(3M)
(HP Integiryt Server Only)
NAME
significand(), significandf(), significandl(), significandw(), significandq() - normalized significand functions
SYNOPSIS
#include <math.h>
double significand(double
x);
float significandf(float
x);
long double significandl(long double
x);
extended significandw(extended
x);
quad significandq(quad
x);
DESCRIPTION
The significand()
functions compute the significand of a floating point value x,where:
x = significand *2ˆexponent and 1 <= significand <2
Thus, if x is subnormal, it is treated as though it were normalized before the significand is determined.
significandf()
is a float version of significand()
; it takes a float argument and returns a
float result.
significandl() is a long double
version of significand(); it takes a long double argu-
ment and returns a
long double result.
significandw() is an
extended version of significand(); it takes an extended argument and
returns an
extended result.
significandq() is equivalent to
significandl() on HP-UX systems.
USAGE
To use these functions, compile either with the default
-Ae option or with the
-Aa and
-D_HPUX_SOURCE options.
To use
significandw()
or significandq(), compile also with -fpwidetypes.
Make sure your program includes
<math.h>. Link in the math library by specifying -lm
on the com-
piler or linker command line. For more information, see the HP-UX floating-point guide for HP Integrity
servers at the following site:
http://www.hp.com/go/fp
.
RETURN VALUE
Upon successful completion,
significand()
returns the normalized significand of x.
If x is zero, infinite, or NaN,
significand()
returns x.
ERRORS
No errors are defined.
SEE ALSO
frexp(3M), ilogb(3M), logb(3M), math(5).
STANDARDS CONFORMANCE
These functions are not specified by any standard.
HP-UX 11i Version 3: September 2010 − 1 − Hewlett-Packard Company 1