fmaf(3m) - osol9

1fma(3M)                 Mathematical Library Functions                 fma(3M)
2
3
4

NAME

6       fma, fmaf, fmal - floating-point multiply-add
7

SYNOPSIS

9       c99 [ flag... ] file... -lm [ library... ]
10       #include <math.h>
11
12       double fma(double x, double y, double z);
13
14
15       float fmaf(float x, float y, float z);
16
17
18       long double fmal(long double x, long double y, long double z);
19
20

DESCRIPTION

22       These  functions compute (x * y) + z, rounded as one ternary operation.
23       They compute the value (as if) to infinite precision and round once  to
24       the  result format, according to the rounding mode characterized by the
25       value of FLT_ROUNDS.
26

RETURN VALUES

28       Upon successful completion, these functions return (x * y) + z, rounded
29       as one ternary operation.
30
31
32       If x or y are NaN, a NaN is returned.
33
34
35       If x multiplied by y is an exact infinity and z is also an infinity but
36       with the opposite sign, a domain error occurs and a NaN is returned.
37
38
39       If one of x and y is infinite, the other is 0, and z is not  a  NaN,  a
40       domain error occurs and a NaN is returned.
41
42
43       If x*y is not 0*Inf nor Inf*0 and z is a NaN, a NaN is returned.
44

ERRORS

46       These functions will fail if:
47
48       Domain Error    The  value  of  x*y+z  is  invalid  or the value x*y is
49                       invalid.
50
51                       If   the   integer   expression   (math_errhandling   &
52                       MATH_ERREXCEPT) is non-zero, the invalid floating-point
53                       exception will be raised.
54
55
56       Range Error     The result overflows.
57
58                       If   the   integer   expression   (math_errhandling   &
59                       MATH_ERREXCEPT)  is  non-zero,  the  overflow floating-
60                       point exception will be raised.
61
62

USAGE

64       An application wanting to check for exceptions should  call  feclearex‐
65       cept(FE_ALL_EXCEPT)  before  calling  these  functions.  On  return, if
66       fetestexcept(FE_INVALID | FE_DIVBYZERO | FE_OVERFLOW | FE_UNDERFLOW) is
67       non-zero,  an  exception  has been raised. An application should either
68       examine the return value or check the floating point exception flags to
69       detect exceptions.
70

ATTRIBUTES

72       See attributes(5) for descriptions of the following attributes:
73
74
75
76
77       ┌─────────────────────────────┬─────────────────────────────┐
78       │      ATTRIBUTE TYPE         │      ATTRIBUTE VALUE        │
79       ├─────────────────────────────┼─────────────────────────────┤
80       │Interface Stability          │Standard                     │
81       ├─────────────────────────────┼─────────────────────────────┤
82       │MT-Level                     │MT-Safe                      │
83       └─────────────────────────────┴─────────────────────────────┘
84