The CPU module defines a protocol-based representation of a CPU's instruction set that can be used to translate between binary and semantic representations of instructions.
Every instruction in a CPU's instruction set has a unique binary footprint. This footprint represents one or more bytes in memory starting with the instruction's opcode and ending with the instruction's immediate, if the opcode requires one.
Each byte of the opcode corresponds to a location in a lookup table describing the intended shape and behavior of the instruction. See an example of a lookup table for the Gameboy CPU.
Many instructions have an immediate.
An immediate is one or more bytes of additional information that can augment the behavior of the instruction.
For example, the Gameboy CPU opcode $E0 corresponds to an instruction of the form LD [$FF00+(a8)], a which requires an 8-bit (one byte) immediate.
This means that $E0 $89 is represented as ld [$FF89], a in assembly.
Note: when an instruction is written in assembly form, each comma-separated value after the opcode is called an operand. Some operands are implicit in the opcode's instruction definition, while others come from the immediate.
The CPU module represents instruction sets using three important types:
InstructionSpec: describes a specific instruction's binary footprint.InstructionSet: translates a series of opcode bytes into anInstructionSpec.Instruction: represents a CPU instruction as anInstructionSpecand the value of its immediate, if it has one.