Flags affected:

        O D I T S Z A P C  OF: Overflow flag  TF: Trap flag AF: Aux carry
        F F F F F F F F F  DF: Direction flag SF: Sign flag PF: Parity flag
            <none>         IF: Interrupt flag ZF: Zero flag CF: Carry flag

Legal forms:

        LOOP d8

Examples:

        LOOP PokeValue

Notes:

LOOP is a combination decrement counter, test, and jump instruction. It uses the CX register in 16-bit modes, and ECX in 32-bit modes. The operation of LOOP is logistically identical in both modes, and I use 16-bit coding as an example here.

LOOP simplifies code by acting as a DEC CX instruction, a CMP CX,0 instruction, and JZ instruction, all at once. A repeat count must be initially loaded into CX. When the LOOP instruction is executed, it first decrements CX. Then it tests to see if CX = 0. If CX is not 0, LOOP transfers control to the displacement specified as its operand:

             MOV CX,17
        DoIt:     CALL CrunchIt
             CALL StuffIt
             LOOP DoIt

Here, the two procedure CALLs will be made 17 times. The first 16 times through, CX will still be nonzero and LOOP will transfer control to DoIt. On the 17th pass, however, LOOP will decrement CX to 0, and then fall through to the next instruction in sequence when it tests CX.

LOOP does not alter any flags, even when CX is decremented to 0. Warning: Watch your initial conditions! If CX is initially 0, LOOP will decrement it to 65,535 (0FFFFH) and then perform the loop 65,535 times. Worse, if you're working in 32-bit protected mode and enter a loop with ECX = 0, the loop will be performed over 2 billion times, which might be long enough to look like a system lockup.

        r8 = AL AH BL BH CL CH DL DH        r16 = AX BX CX DX BP SP SI DI
        sr = CS DS SS ES FS GS              r32 = EAX EBX ECX EDX EBP ESP ESI EDI
        m8 = 8-bit memory data              m16 = 16-bit memory data
        m32 = 32-bit memory data            i8 = 8-bit immediate data
        i16 = 16-bit immediate data         i32 = 32-bit immediate data
        d8 = 8-bit signed displacement      d16 = 16-bit signed displacement
        d32 = 32-bit unsigned displacement