Headers float.h(3HEAD)
NAME
float.h, float - floating types
SYNOPSIS
#include
DESCRIPTION
The characteristics of floating types are defined in terms
of a model that describes a representation of floating-point
numbers and values that provide information about an
implementation's floating-point arithmetic.
The following parameters are used to define the model for
each floating-point type:
s sign (]1)
b base or radix of exponent representation (an integer
>1)
e exponent (an integer between a minimum e(min) and a
maximum e(max))
p precision (the number of base-b digits in the signifi-
cand)
f(k) non-negative integers less than b (the significand
digits)
In addition to normalized floating-point numbers (f(1)>0 if
x=/0), floating types might be able to contain other kinds of
floating-point numbers, such as subnormal floating-point
numbers (x=/0, e=e(min), f(1)=0) and unnormalized floating-
point numbers (x=/0, e=e(min), f(1)=0), and values that are
not floating-point numbers, such as infinities and NaNs. A
NaN is an encoding signifying Not-a-Number. A quiet NaN pro-
pagates through almost every arithmetic operation without
raising a floating-point exception; a signaling NaN gen-
erally raises a floating-point exception when occurring as
an arithmetic operand.
The accuracy of the library functions in math.h(3HEAD) and
complex.h(3HEAD) that return floating-point results is
SunOS 5.11 Last change: 17 Dec 2003 1
Headers float.h(3HEAD)
defined on the libm(3LIB) manual page.
All integer values in the header, except
FLTROUNDS, are constant expressions suitable for use in #if
preprocessing directives; all floating values are constant
expressions. All except DECIMALDIG, FLTEVALMETHOD,
FLTRADIX, and FLTROUNDS have separate names for all three
floating-point types. The floating-point model representa-
tion is provided for all values except FLTEVALMETHOD and
FLTROUNDS.
The rounding mode for floating-point addition is character-
ized by the value of FLTROUNDS:
-1 Indeterminable.
0 Toward zero.
1 To nearest.
2 Toward positive infinity.
3 Toward negative infinity.
The values of operations with floating operands and values
subject to the usual arithmetic conversions and of floating
constants are evaluated to a format whose range and preci-
sion might be greater than required by the type. The use of
evaluation formats is characterized by the architecture-
dependent value of FLTEVALMETHOD:
-1 Indeterminable.
0 Evaluate all operations and constants just to the
range and precision of the type.
1 Evaluate operations and constants of type float and
double to the range and precision of the double type;
evaluate long double operations and constants to the
range and precision of the long double type.
SunOS 5.11 Last change: 17 Dec 2003 2
Headers float.h(3HEAD)
2 Evaluate all operations and constants to the range and
precision of the long double type.
The values given in the following list are defined as con-
stants.
o Radix of exponent representation, b.
FLTRADIX
o Number of base-FLTRADIX digits in the floating-
point significand, p.
FLTMANTDIG
DBLMANTDIG
LDBLMANTDIG
o Number of decimal digits, n, such that any
floating-point number in the widest supported
floating type with p(max) radix b digits can be
rounded to a floating-point number with n decimal
digits and back again without change to the value.
DECIMALDIG
o Number of decimal digits, q, such that any
floating-point number with q decimal digits can be
rounded into a floating-point number with p radix b
digits and back again without change to the q
decimal digits.
FLTDIG
DBLDIG
LDBLDIG
o Minimum negative integer such that FLTRADIX raised
to that power minus 1 is a normalized floating-
point number, e(min).
FLTMINEXP
DBLMINEXP
LDBLMINEXP
o Minimum negative integer such that 10 raised to
that power is in the range of normalized floating-
SunOS 5.11 Last change: 17 Dec 2003 3
Headers float.h(3HEAD)
point numbers.
FLTMIN10EXP
DBLMIN10EXP
LDBLMIN10EXP
o Maximum integer such that FLTRADIX raised to that
power minus 1 is a representable finite floating-
point number, e(max).
FLTMAXEXP
DBLMAXEXP
LDBLMAXEXP
o Maximum integer such that 10 raised to that power
is in the range of representable finite floating-
point numbers.
FLTMAX10EXP
DBLMAX10EXP
LDBLMAX10EXP
The values given in the following list are defined as con-
stant expressions with values that are greater than or equal
to those shown:
o Maximum representable finite floating-point number.
FLTMAX
DBLMAX
LDBLMAX
The values given in the following list are defined as con-
stant expressions with implementation-defined (positive)
values that are less than or equal to those shown:
o The difference between 1 and the least value
greater than 1 that is representable in the given
floating-point type, b^1 - p.
FLTEPSILON
DBLEPSILON
LDBLEPSILON
o Minimum normalized positive floating-point number,
SunOS 5.11 Last change: 17 Dec 2003 4
Headers float.h(3HEAD)
b^e(min)^-1.
FLTMIN
DBLMIN
LDBLMIN
ATRIBUTES
See attributes(5) for descriptions of the following attri-
butes:
ATRIBUTE TYPE ATRIBUTE VALUE
Interface Stability Standard
SEE ALSO
complex.h(3HEAD), math.h(3HEAD), attributes(5), standards(5)
SunOS 5.11 Last change: 17 Dec 2003 5
|
