Circuit for fixed point or floating point arithmetic operations

Claim 1

An operation circuit, comprising: decoding means for generating a control signal representative of a mode field in an instruction to be executed when said mode field is present in said instruction, said control signal specifying either a fixed point mode or a floating point mode; holding means, coupled to said decoding means, for storing and outputting an indication of said specified mode when said specified mode is indicated, and outputting an indication of a last specified mode when said mode field is not present in said instruction; an arithmetic logic unit, coupled to said decoding means and responsive to said control signal, for performing arithmetic operations on a set of input values, wherein said arithmetic logic unit manipulates said set of input values as fixed point numbers when said control signal specifies said fixed point mode and manipulates said set of input values as floating point numbers when said control signal specifies said floating point mode; a multiplier, coupled to said holding means, which accepts multiplier input values and outputs a multiplier output value at a multiplier output, said multiplier performing a fixed point multiplication on said multiplier input values and outputting a fixed point multiplier output value when said holding means specifies said fixed point mode, and said multiplier performing a floating point multiplication on said multiplier input values and outputting a floating point multiplier output value when said holding means specifies said floating point mode; and means for coupling said multiplier output to an input of said arithmetic logic unit.