Analog Devices ADSP-2106x SHARC User Manual page 538

ALU Floating-Point
B Compute Operations
Rn = LOGB Fx
Syntax:
Rn = LOGB Fx
Function:
Converts the exponent of the floating-point operand in register Fx to an
unbiased twos-complement fixed-point integer. The result is placed in the
fixed-point field in register Rn. Unbiasing is done by subtracting 127 from
the floating-point exponent in Fx. If saturation mode is not set, a ±Infinity
input returns a floating-point +Infinity and a ±Zero input returns a
floating-point –Infinity. If saturation mode is set, a ±Infinity input returns
the maximum positive value (0x7FFF FFFF) and a ±Zero input returns the
maximum negative value (0x8000 0000). Denormal inputs are flushed to
±Zero. A NAN input returns an all 1s result.
Status flags:
AZ Is set if the fixed-point result is zero, otherwise cleared
AU Is cleared
AN Is set if the result is negative, otherwise cleared
AV Is set if the input operand is an Infinity or a Zero, otherwise cleared
AC Is cleared
AS Is cleared
AI Is set if the input is a NAN, otherwise cleared
B – 36
www.BDTIC.com/ADI