Apparatus and method for recording a floating point error pointer in zero cycles

A microprocessor for saving a floating point unit environment, comprising:

• translation/control logic, for generating a micro instruction sequence directing the microprocessor to execute an operation prescribed by a floating point macro instruction, said micro instruction sequence comprising:
• a record selectors micro instruction, directing the microprocessor to save selectors associated with said floating point macro instruction;
• and
• a record opcode/offset micro instruction extension, directing the microprocessor to store opcode/offset data associated with said floating point macro instruction;
• register/control logic, coupled to said translation/control logic, for storing said selectors, an opcode, and an instruction pointer;
• wherein the floating point unit environment provides data required by an exception handler, when a floating point unit exception occurs, to determine that said floating point unit exception occurred as a result of executing said floating point macro instruction.

A method for saving a floating point macro instruction exception pointer in a microprocessor, the method comprising:

• generating a floating point micro instruction sequence to perform an operation specified by a floating point macro instruction;
• producing a first micro instruction, directing the microprocessor to save a first part of the floating point macro instruction error pointer;
• providing a micro instruction extension, associated with a floating point micro instruction within said floating point micro instruction sequence, and directing the microprocessor to save a second part of the floating point macro instruction error pointer;
• and
• saving the floating point macro instruction exception pointer.