GBZ80(7) Miscellaneous Information Manual GBZ80(7) NAME gbz80 - Game Boy CPU instruction reference DESCRIPTION This is the list of instructions supported by rgbasm(1), including a short description, the number of bytes needed to encode them and the number of CPU cycles at 1MHz (or 2MHz in GBC double speed mode) needed to complete them. Instructons are documented according to the syntax accepted by rgbasm(1), which does not always match one-to-one with the way instructions are encoded: https://gbdev.io/gb-opcodes/optables/. Note: All arithmetic and logic instructions that use register A as a destination can omit the destination, since it is assumed to be register A by default. So the following two lines have the same effect: OR A,B OR B Furthermore, the CPL instruction can take an optional A destination, since it can only be register A. So the following two lines have the same effect: CPL CPL A LEGEND List of abbreviations used in this document. r8 Any of the 8-bit registers (A, B, C, D, E, H, L). r16 Any of the general-purpose 16-bit registers (BC, DE, HL). n8 8-bit integer constant (signed or unsigned, -128 to 255). n16 16-bit integer constant (signed or unsigned, -32768 to 65535). e8 8-bit signed offset (-128 to 127). u3 3-bit unsigned bit index (0 to 7, with 0 as the least significant bit). cc A condition code: Z Execute if Z is set. NZ Execute if Z is not set. C Execute if C is set. NC Execute if C is not set. vec An RST vector (0x00, 0x08, 0x10, 0x18, 0x20, 0x28, 0x30, and 0x38). INSTRUCTION OVERVIEW Load instructions LD r8,r8 LD r8,n8 LD r16,n16 LD [HL],r8 LD [HL],n8 LD r8,[HL] LD [r16],A LD [n16],A LDH [n16],A LDH [C],A LD A,[r16] LD A,[n16] LDH A,[n16] LDH A,[C] LD [HLI],A LD [HLD],A LD A,[HLI] LD A,[HLD] 8-bit arithmetic instructions ADC A,r8 ADC A,[HL] ADC A,n8 ADD A,r8 ADD A,[HL] ADD A,n8 CP A,r8 CP A,[HL] CP A,n8 DEC r8 DEC [HL] INC r8 INC [HL] SBC A,r8 SBC A,[HL] SBC A,n8 SUB A,r8 SUB A,[HL] SUB A,n8 16-bit arithmetic instructions ADD HL,r16 DEC r16 INC r16 Bitwise logic instructions AND A,r8 AND A,[HL] AND A,n8 CPL OR A,r8 OR A,[HL] OR A,n8 XOR A,r8 XOR A,[HL] XOR A,n8 Bit flag instructions BIT u3,r8 BIT u3,[HL] RES u3,r8 RES u3,[HL] SET u3,r8 SET u3,[HL] Bit shift instructions RL r8 RL [HL] RLA RLC r8 RLC [HL] RLCA RR r8 RR [HL] RRA RRC r8 RRC [HL] RRCA SLA r8 SLA [HL] SRA r8 SRA [HL] SRL r8 SRL [HL] SWAP r8 SWAP [HL] Jumps and subroutine instructions CALL n16 CALL cc,n16 JP HL JP n16 JP cc,n16 JR n16 JR cc,n16 RET cc RET RETI RST vec Carry flag instructions CCF SCF Stack manipulation instructions ADD HL,SP ADD SP,e8 DEC SP INC SP LD SP,n16 LD [n16],SP LD HL,SP+e8 LD SP,HL POP AF POP r16 PUSH AF PUSH r16 Interrupt-related instructions DI EI HALT Miscellaneous instructions DAA NOP STOP INSTRUCTION REFERENCE ADC A,r8 Add the value in r8 plus the carry flag to A. Cycles: 1 Bytes: 1 Flags: Z Set if result is 0. N 0 H Set if overflow from bit 3. C Set if overflow from bit 7. ADC A,[HL] Add the byte pointed to by HL plus the carry flag to A. Cycles: 2 Bytes: 1 Flags: See ADC A,r8 ADC A,n8 Add the value n8 plus the carry flag to A. Cycles: 2 Bytes: 2 Flags: See ADC A,r8 ADD A,r8 Add the value in r8 to A. Cycles: 1 Bytes: 1 Flags: Z Set if result is 0. N 0 H Set if overflow from bit 3. C Set if overflow from bit 7. ADD A,[HL] Add the byte pointed to by HL to A. Cycles: 2 Bytes: 1 Flags: See ADD A,r8 ADD A,n8 Add the value n8 to A. Cycles: 2 Bytes: 2 Flags: See ADD A,r8 ADD HL,r16 Add the value in r16 to HL. Cycles: 2 Bytes: 1 Flags: N 0 H Set if overflow from bit 11. C Set if overflow from bit 15. ADD HL,SP Add the value in SP to HL. Cycles: 2 Bytes: 1 Flags: See ADD HL,r16 ADD SP,e8 Add the signed value e8 to SP. Cycles: 4 Bytes: 2 Flags: Z 0 N 0 H Set if overflow from bit 3. C Set if overflow from bit 7. AND A,r8 Set A to the bitwise AND between the value in r8 and A. Cycles: 1 Bytes: 1 Flags: Z Set if result is 0. N 0 H 1 C 0 AND A,[HL] Set A to the bitwise AND between the byte pointed to by HL and A. Cycles: 2 Bytes: 1 Flags: See AND A,r8 AND A,n8 Set A to the bitwise AND between the value n8 and A. Cycles: 2 Bytes: 2 Flags: See AND A,r8 BIT u3,r8 Test bit u3 in register r8, set the zero flag if bit not set. Cycles: 2 Bytes: 2 Flags: Z Set if the selected bit is 0. N 0 H 1 BIT u3,[HL] Test bit u3 in the byte pointed by HL, set the zero flag if bit not set. Cycles: 3 Bytes: 2 Flags: See BIT u3,r8 CALL n16 Call address n16. This pushes the address of the instruction after the CALL on the stack, such that RET can pop it later; then, it executes an implicit JP n16. Cycles: 6 Bytes: 3 Flags: None affected. CALL cc,n16 Call address n16 if condition cc is met. Cycles: 6 taken / 3 untaken Bytes: 3 Flags: None affected. CCF Complement Carry Flag. Cycles: 1 Bytes: 1 Flags: N 0 H 0 C Inverted. CP A,r8 ComPare the value in A with the value in r8. This subtracts the value in r8 from A and sets flags accordingly, but discards the result. Cycles: 1 Bytes: 1 Flags: Z Set if result is 0. N 1 H Set if borrow from bit 4. C Set if borrow (i.e. if r8 > A). CP A,[HL] ComPare the value in A with the byte pointed to by HL. This subtracts the byte pointed to by HL from A and sets flags accordingly, but discards the result. Cycles: 2 Bytes: 1 Flags: See CP A,r8 CP A,n8 ComPare the value in A with the value n8. This subtracts the value n8 from A and sets flags accordingly, but discards the result. Cycles: 2 Bytes: 2 Flags: See CP A,r8 CPL ComPLement accumulator (A = ~A); also called bitwise NOT. Cycles: 1 Bytes: 1 Flags: N 1 H 1 DAA Decimal Adjust Accumulator. Designed to be used after performing an arithmetic instruction (ADD, ADC, SUB, SBC) whose inputs were in Binary-Coded Decimal (BCD), adjusting the result to likewise be in BCD. The exact behavior of this instruction depends on the state of the subtract flag N: If the subtract flag N is set: 1. Initialize the adjustment to 0. 2. If the half-carry flag H is set, then add $6 to the adjustment. 3. If the carry flag is set, then add $60 to the adjustment. 4. Subtract the adjustment from A. If the subtract flag N is not set: 1. Initialize the adjustment to 0. 2. If the half-carry flag H is set or A & $F > $9, then add $6 to the adjustment. 3. If the carry flag is set or A > $99, then add $60 to the adjustment and set the carry flag. 4. Add the adjustment to A. Cycles: 1 Bytes: 1 Flags: Z Set if result is 0. H 0 C Set or unaffected depending on the operation. DEC r8 Decrement the value in register r8 by 1. Cycles: 1 Bytes: 1 Flags: Z Set if result is 0. N 1 H Set if borrow from bit 4. DEC [HL] Decrement the byte pointed to by HL by 1. Cycles: 3 Bytes: 1 Flags: See DEC r8 DEC r16 Decrement the value in register r16 by 1. Cycles: 2 Bytes: 1 Flags: None affected. DEC SP Decrement the value in register SP by 1. Cycles: 2 Bytes: 1 Flags: None affected. DI Disable Interrupts by clearing the IME flag. Cycles: 1 Bytes: 1 Flags: None affected. EI Enable Interrupts by setting the IME flag. The flag is only set after the instruction following EI. Cycles: 1 Bytes: 1 Flags: None affected. HALT Enter CPU low-power consumption mode until an interrupt occurs. The exact behavior of this instruction depends on the state of the IME flag, and whether interrupts are pending (i.e. whether `[IE] & [IF]' is non-zero): If the IME flag is set: The CPU enters low-power mode until after an interrupt is about to be serviced. The handler is executed normally, and the CPU resumes execution after the HALT when that returns. If the IME flag is not set, and no interrupts are pending: As soon as an interrupt becomes pending, the CPU resumes execution. This is like the above, except that the handler is not called. If the IME flag is not set, and some interrupt is pending: The CPU continues execution after the HALT, but the byte after it is read twice in a row (PC is not incremented, due to a hardware bug). Cycles: - Bytes: 1 Flags: None affected. INC r8 Increment the value in register r8 by 1. Cycles: 1 Bytes: 1 Flags: Z Set if result is 0. N 0 H Set if overflow from bit 3. INC [HL] Increment the byte pointed to by HL by 1. Cycles: 3 Bytes: 1 Flags: See INC r8 INC r16 Increment the value in register r16 by 1. Cycles: 2 Bytes: 1 Flags: None affected. INC SP Increment the value in register SP by 1. Cycles: 2 Bytes: 1 Flags: None affected. JP n16 Jump to address n16; effectively, copy n16 into PC. Cycles: 4 Bytes: 3 Flags: None affected. JP cc,n16 Jump to address n16 if condition cc is met. Cycles: 4 taken / 3 untaken Bytes: 3 Flags: None affected. JP HL Jump to address in HL; effectively, copy the value in register HL into PC. Cycles: 1 Bytes: 1 Flags: None affected. JR n16 Relative Jump to address n16. The target address n16 is encoded as a signed 8-bit offset from the address immediately following the JR instruction, so it must be between -128 and 127 bytes away. For example: JR Label ; no-op; encoded offset of 0 Label: JR Label ; infinite loop; encoded offset of -2 Cycles: 3 Bytes: 2 Flags: None affected. JR cc,n16 Relative Jump to address n16 if condition cc is met. The target address n16 is encoded as a signed 8-bit offset from the address immediately following the JR instruction, so it must be between -128 and 127 bytes away. Cycles: 3 taken / 2 untaken Bytes: 2 Flags: None affected. LD r8,r8 Copy (aka Load) the value in register on the right into the register on the left. Storing a register into itself is a no-op; however, some Game Boy emulators interpret LD B,B as a breakpoint, or LD D,D as a debug message (such as BGB: https://bgb.bircd.org/manual.html#expressions). Cycles: 1 Bytes: 1 Flags: None affected. LD r8,n8 Copy the value n8 into register r8. Cycles: 2 Bytes: 2 Flags: None affected. LD r16,n16 Copy the value n16 into register r16. Cycles: 3 Bytes: 3 Flags: None affected. LD [HL],r8 Copy the value in register r8 into the byte pointed to by HL. Cycles: 2 Bytes: 1 Flags: None affected. LD [HL],n8 Copy the value n8 into the byte pointed to by HL. Cycles: 3 Bytes: 2 Flags: None affected. LD r8,[HL] Copy the value pointed to by HL into register r8. Cycles: 2 Bytes: 1 Flags: None affected. LD [r16],A Copy the value in register A into the byte pointed to by r16. Cycles: 2 Bytes: 1 Flags: None affected. LD [n16],A Copy the value in register A into the byte at address n16. Cycles: 4 Bytes: 3 Flags: None affected. LDH [n16],A Copy the value in register A into the byte at address n16. The destination address n16 is encoded as its 8-bit low byte and assumes a high byte of $FF, so it must be between $FF00 and $FFFF. Cycles: 3 Bytes: 2 Flags: None affected. LDH [C],A Copy the value in register A into the byte at address $FF00+C. Cycles: 2 Bytes: 1 Flags: None affected. This is sometimes written as `LD [$FF00+C],A'. LD A,[r16] Copy the byte pointed to by r16 into register A. Cycles: 2 Bytes: 1 Flags: None affected. LD A,[n16] Copy the byte at address n16 into register A. Cycles: 4 Bytes: 3 Flags: None affected. LDH A,[n16] Copy the byte at address n16 into register A. The source address n16 is encoded as its 8-bit low byte and assumes a high byte of $FF, so it must be between $FF00 and $FFFF. Cycles: 3 Bytes: 2 Flags: None affected. LDH A,[C] Copy the byte at address $FF00+C into register A. Cycles: 2 Bytes: 1 Flags: None affected. This is sometimes written as `LD A,[$FF00+C]'. LD [HLI],A Copy the value in register A into the byte pointed by HL and increment HL afterwards. Cycles: 2 Bytes: 1 Flags: None affected. This is sometimes written as `LD [HL+],A', or `LDI [HL],A'. LD [HLD],A Copy the value in register A into the byte pointed by HL and decrement HL afterwards. Cycles: 2 Bytes: 1 Flags: None affected. This is sometimes written as `LD [HL-],A', or `LDD [HL],A'. LD A,[HLD] Copy the byte pointed to by HL into register A, and decrement HL afterwards. Cycles: 2 Bytes: 1 Flags: None affected. This is sometimes written as `LD A,[HL-]', or `LDD A,[HL]'. LD A,[HLI] Copy the byte pointed to by HL into register A, and increment HL afterwards. Cycles: 2 Bytes: 1 Flags: None affected. This is sometimes written as `LD A,[HL+]', or `LDI A,[HL]'. LD SP,n16 Copy the value n16 into register SP. Cycles: 3 Bytes: 3 Flags: None affected. LD [n16],SP Copy SP & $FF at address n16 and SP >> 8 at address n16 + 1. Cycles: 5 Bytes: 3 Flags: None affected. LD HL,SP+e8 Add the signed value e8 to SP and copy the result in HL. Cycles: 3 Bytes: 2 Flags: Z 0 N 0 H Set if overflow from bit 3. C Set if overflow from bit 7. LD SP,HL Copy register HL into register SP. Cycles: 2 Bytes: 1 Flags: None affected. NOP No OPeration. Cycles: 1 Bytes: 1 Flags: None affected. OR A,r8 Set A to the bitwise OR between the value in r8 and A. Cycles: 1 Bytes: 1 Flags: Z Set if result is 0. N 0 H 0 C 0 OR A,[HL] Set A to the bitwise OR between the byte pointed to by HL and A. Cycles: 2 Bytes: 1 Flags: See OR A,r8 OR A,n8 Set A to the bitwise OR between the value n8 and A. Cycles: 2 Bytes: 2 Flags: See OR A,r8 POP AF Pop register AF from the stack. This is roughly equivalent to the following imaginary instructions: LD F, [SP] ; See below for individual flags INC SP LD A, [SP] INC SP Cycles: 3 Bytes: 1 Flags: Z Set from bit 7 of the popped low byte. N Set from bit 6 of the popped low byte. H Set from bit 5 of the popped low byte. C Set from bit 4 of the popped low byte. POP r16 Pop register r16 from the stack. This is roughly equivalent to the following imaginary instructions: LD LOW(r16), [SP] ; C, E or L INC SP LD HIGH(r16), [SP] ; B, D or H INC SP Cycles: 3 Bytes: 1 Flags: None affected. PUSH AF Push register AF into the stack. This is roughly equivalent to the following imaginary instructions: DEC SP LD [SP], A DEC SP LD [SP], F.Z << 7 | F.N << 6 | F.H << 5 | F.C << 4 Cycles: 4 Bytes: 1 Flags: None affected. PUSH r16 Push register r16 into the stack. This is roughly equivalent to the following imaginary instructions: DEC SP LD [SP], HIGH(r16) ; B, D or H DEC SP LD [SP], LOW(r16) ; C, E or L Cycles: 4 Bytes: 1 Flags: None affected. RES u3,r8 Set bit u3 in register r8 to 0. Bit 0 is the rightmost one, bit 7 the leftmost one. Cycles: 2 Bytes: 2 Flags: None affected. RES u3,[HL] Set bit u3 in the byte pointed by HL to 0. Bit 0 is the rightmost one, bit 7 the leftmost one. Cycles: 4 Bytes: 2 Flags: None affected. RET Return from subroutine. This is basically a POP PC (if such an instruction existed). See POP r16 for an explanation of how POP works. Cycles: 4 Bytes: 1 Flags: None affected. RET cc Return from subroutine if condition cc is met. Cycles: 5 taken / 2 untaken Bytes: 1 Flags: None affected. RETI Return from subroutine and enable interrupts. This is basically equivalent to executing EI then RET, meaning that IME is set right after this instruction. Cycles: 4 Bytes: 1 Flags: None affected. RL r8 Rotate bits in register r8 left, through the carry flag. Flags r8 C <- b7 <- ... <- b0 <- | | Cycles: 2 Bytes: 2 Flags: Z Set if result is 0. N 0 H 0 C Set according to result. RL [HL] Rotate the byte pointed to by HL left, through the carry flag. Flags [HL] C <- b7 <- ... <- b0 <- | | Cycles: 4 Bytes: 2 Flags: See RL r8 RLA Rotate register A left, through the carry flag. Flags A C <- b7 <- ... <- b0 <- | | Cycles: 1 Bytes: 1 Flags: Z 0 N 0 H 0 C Set according to result. RLC r8 Rotate register r8 left. Flags r8 C <- b7 <- ... <- b0 <- | | Cycles: 2 Bytes: 2 Flags: Z Set if result is 0. N 0 H 0 C Set according to result. RLC [HL] Rotate the byte pointed to by HL left. Flags [HL] C <- b7 <- ... <- b0 <- | | Cycles: 4 Bytes: 2 Flags: See RLC r8 RLCA Rotate register A left. Flags A C <- b7 <- ... <- b0 <- | | Cycles: 1 Bytes: 1 Flags: Z 0 N 0 H 0 C Set according to result. RR r8 Rotate register r8 right, through the carry flag. r8 Flags -> b7 -> ... -> b0 -> C | | Cycles: 2 Bytes: 2 Flags: Z Set if result is 0. N 0 H 0 C Set according to result. RR [HL] Rotate the byte pointed to by HL right, through the carry flag. [HL] Flags -> b7 -> ... -> b0 -> C | | Cycles: 4 Bytes: 2 Flags: See RR r8 RRA Rotate register A right, through the carry flag. A Flags -> b7 -> ... -> b0 -> C | | Cycles: 1 Bytes: 1 Flags: Z 0 N 0 H 0 C Set according to result. RRC r8 Rotate register r8 right. r8 Flags -> b7 -> ... -> b0 -> C | | Cycles: 2 Bytes: 2 Flags: Z Set if result is 0. N 0 H 0 C Set according to result. RRC [HL] Rotate the byte pointed to by HL right. [HL] Flags -> b7 -> ... -> b0 -> C | | Cycles: 4 Bytes: 2 Flags: See RRC r8 RRCA Rotate register A right. A Flags -> b7 -> ... -> b0 -> C | | Cycles: 1 Bytes: 1 Flags: Z 0 N 0 H 0 C Set according to result. RST vec Call address vec. This is a shorter and faster equivalent to CALL for suitable values of vec. Cycles: 4 Bytes: 1 Flags: None affected. SBC A,r8 Subtract the value in r8 and the carry flag from A. Cycles: 1 Bytes: 1 Flags: Z Set if result is 0. N 1 H Set if borrow from bit 4. C Set if borrow (i.e. if (r8 + carry) > A). SBC A,[HL] Subtract the byte pointed to by HL and the carry flag from A. Cycles: 2 Bytes: 1 Flags: See SBC A,r8 SBC A,n8 Subtract the value n8 and the carry flag from A. Cycles: 2 Bytes: 2 Flags: See SBC A,r8 SCF Set Carry Flag. Cycles: 1 Bytes: 1 Flags: N 0 H 0 C 1 SET u3,r8 Set bit u3 in register r8 to 1. Bit 0 is the rightmost one, bit 7 the leftmost one. Cycles: 2 Bytes: 2 Flags: None affected. SET u3,[HL] Set bit u3 in the byte pointed by HL to 1. Bit 0 is the rightmost one, bit 7 the leftmost one. Cycles: 4 Bytes: 2 Flags: None affected. SLA r8 Shift Left Arithmetically register r8. Flags r8 C <- b7 <- ... <- b0 <- 0 Cycles: 2 Bytes: 2 Flags: Z Set if result is 0. N 0 H 0 C Set according to result. SLA [HL] Shift Left Arithmetically the byte pointed to by HL. Flags [HL] C <- b7 <- ... <- b0 <- 0 Cycles: 4 Bytes: 2 Flags: See SLA r8 SRA r8 Shift Right Arithmetically register r8 (bit 7 of r8 is unchanged). r8 Flags b7 -> ... -> b0 -> C Cycles: 2 Bytes: 2 Flags: Z Set if result is 0. N 0 H 0 C Set according to result. SRA [HL] Shift Right Arithmetically the byte pointed to by HL (bit 7 of the byte pointed to by HL is unchanged). [HL] Flags b7 -> ... -> b0 -> C Cycles: 4 Bytes: 2 Flags: See SRA r8 SRL r8 Shift Right Logically register r8. r8 Flags 0 -> b7 -> ... -> b0 -> C Cycles: 2 Bytes: 2 Flags: Z Set if result is 0. N 0 H 0 C Set according to result. SRL [HL] Shift Right Logically the byte pointed to by HL. [HL] Flags 0 -> b7 -> ... -> b0 -> C Cycles: 4 Bytes: 2 Flags: See SRL r8 STOP Enter CPU very low power mode. Also used to switch between GBC double speed and normal speed CPU modes. The exact behavior of this instruction is fragile and may interpret its second byte as a separate instruction (see the Pan Docs: https://gbdev.io/pandocs/Reducing_Power_Consumption.html#using-the-stop-instruction), which is why rgbasm(1) allows explicitly specifying the second byte (STOP n8) to override the default of $00 (a NOP instruction). Cycles: - Bytes: 2 Flags: None affected. SUB A,r8 Subtract the value in r8 from A. Cycles: 1 Bytes: 1 Flags: Z Set if result is 0. N 1 H Set if borrow from bit 4. C Set if borrow (i.e. if r8 > A). SUB A,[HL] Subtract the byte pointed to by HL from A. Cycles: 2 Bytes: 1 Flags: See SUB A,r8 SUB A,n8 Subtract the value n8 from A. Cycles: 2 Bytes: 2 Flags: See SUB A,r8 SWAP r8 Swap the upper 4 bits in register r8 and the lower 4 ones. Cycles: 2 Bytes: 2 Flags: Z Set if result is 0. N 0 H 0 C 0 SWAP [HL] Swap the upper 4 bits in the byte pointed by HL and the lower 4 ones. Cycles: 4 Bytes: 2 Flags: See SWAP r8 XOR A,r8 Set A to the bitwise XOR between the value in r8 and A. Cycles: 1 Bytes: 1 Flags: Z Set if result is 0. N 0 H 0 C 0 XOR A,[HL] Set A to the bitwise XOR between the byte pointed to by HL and A. Cycles: 2 Bytes: 1 Flags: See XOR A,r8 XOR A,n8 Set A to the bitwise XOR between the value n8 and A. Cycles: 2 Bytes: 2 Flags: See XOR A,r8 SEE ALSO rgbasm(1), rgblink(1), rgbfix(1), rgbgfx(1), rgbasm-old(5), rgbds(7) HISTORY rgbasm(1) was originally written by Carsten Sorensen as part of the ASMotor package, and was later repackaged in RGBDS by Justin Lloyd. It is now maintained by a number of contributors at https://github.com/gbdev/rgbds. Linux 6.12.107+deb13-amd64 August 1, 2026 Linux 6.12.107+deb13-amd64