NAME
recipf4_fast
- return approximate reciprocals of float elements
SYNOPSIS
Procedure
call syntax:
#include <simdmath.h> |
vector float recipf4_fast(vector float x); |
Link with -lsimdmath |
Inline call syntax:
#include <simdmath.h> |
#include <recipf4.h> |
vector float _recipf4_fast(vector float x); |
Parameters |
|
x |
input vector |
DESCRIPTION
The recipf4_fast function
returns a vector of the reciprocals of the corresponding elements of x.
The recipf4 function provides a slower but more accurate version of recipf4_fast.
RETURN VALUE
The function
recipf4_fast returns
a float vector in which each element is defined as:
- the reciprocals of the corresponding element of x.
- When an element of x is Inf the result is 0 with
the sign of x.
- When an element of x is 0:
- on the PPU the result is Inf with the sign of x,
- on the SPU the result is HUGE_VAL with the sign of x.
- When an element of x is NaN the result is NaN.
ENVIRONMENT
recipf4_fast:
SPU only
CONFORMING TO
SIMD Math
library specification for the Cell Broadband Engineā¢ Architecture.
NOTES
Basis
ISO9899
(C99) recip functions.
SEE ALSO
recip(3),
recipd2(3),
recipd2(3),
div(3),
divf4(3),
divi4(3),
divu4(3),
divd2(3),
lldivi2(3),
lldivu2(3),
fma(3),
fmaf4(3),
fmad2(3),
modf(3),
modff4(3),
modfd2(3), fmod(3),
fmodf4(3),
fmodd2(3), remainder(3),
remainderf4(3),
remainderd2(3), remquo(3),
remquof4(3),
remquod2(3),
rsqrt(3),
rsqrtf4(3),
rsqrtd2(3)