Inside Macintosh: PowerPC Numerics / Part 2 - The PowerPC Numerics C Implementation

Chapter 10 - Transcendental Functions

Sign Manipulation Functions

MathLib provides two functions that manipulate the sign bit of the floating-point value:

$copysign(x,y)$	Copies the sign of y to x.
$fabs(x)$	Returns the absolute value (positive form) of x.

Because these functions only manipulate the sign bit of the value and do not try to compute the value at all, they raise no floating-point exceptions.