You copied the Doc URL to your clipboard.

SUB, SUBS (register)

Subtract (register) subtracts an optionally-shifted register value from a register value, and writes the result to the destination register.

If the destination register is not the PC, the SUBS variant of the instruction updates the condition flags based on the result.

The field descriptions for <Rd> identify the encodings where the PC is permitted as the destination register. However, when the destination register is the PC:

It has encodings from the following instruction sets: A32 ( A1 ) and T32 ( T1 and T2 ) .

A1

313029282726252423222120191817161514131211109876543210
!= 11110000010S!= 1101Rdimm5stype0Rm
condRn

SUB, rotate right with extend (S == 0 && imm5 == 00000 && stype == 11)

SUB{<c>}{<q>} {<Rd>,} <Rn>, <Rm>, RRX

SUB, shift or rotate by value (S == 0 && !(imm5 == 00000 && stype == 11))

SUB{<c>}{<q>} {<Rd>,} <Rn>, <Rm> {, <shift> #<amount>}

SUBS, rotate right with extend (S == 1 && imm5 == 00000 && stype == 11)

SUBS{<c>}{<q>} {<Rd>,} <Rn>, <Rm>, RRX

SUBS, shift or rotate by value (S == 1 && !(imm5 == 00000 && stype == 11))

SUBS{<c>}{<q>} {<Rd>,} <Rn>, <Rm> {, <shift> #<amount>}

if Rn == '1101' then SEE "SUB (SP minus register)";
d = UInt(Rd);  n = UInt(Rn);  m = UInt(Rm);  setflags = (S == '1');
(shift_t, shift_n) = DecodeImmShift(stype, imm5);

T1

1514131211109876543210
0001101RmRnRd

T1

SUB<c>{<q>} <Rd>, <Rn>, <Rm> // (Inside IT block)

SUBS{<q>} {<Rd>,} <Rn>, <Rm> // (Outside IT block) 

d = UInt(Rd);  n = UInt(Rn);  m = UInt(Rm);  setflags = !InITBlock();
(shift_t, shift_n) = (SRType_LSL, 0);

T2

15141312111098765432101514131211109876543210
11101011101S!= 1101(0)imm3Rdimm2stypeRm
Rn

SUB, rotate right with extend (S == 0 && imm3 == 000 && imm2 == 00 && stype == 11)

SUB{<c>}{<q>} {<Rd>,} <Rn>, <Rm>, RRX

SUB, shift or rotate by value (S == 0 && !(imm3 == 000 && imm2 == 00 && stype == 11))

SUB<c>.W {<Rd>,} <Rn>, <Rm> // (Inside IT block, and <Rd>, <Rn>, <Rm> can be represented in T1)

SUB{<c>}{<q>} {<Rd>,} <Rn>, <Rm> {, <shift> #<amount>}

SUBS, rotate right with extend (S == 1 && imm3 == 000 && Rd != 1111 && imm2 == 00 && stype == 11)

SUBS{<c>}{<q>} {<Rd>,} <Rn>, <Rm>, RRX

SUBS, shift or rotate by value (S == 1 && !(imm3 == 000 && imm2 == 00 && stype == 11) && Rd != 1111)

SUBS.W {<Rd>,} <Rn>, <Rm> // (Outside IT block, and <Rd>, <Rn>, <Rm> can be represented in T1)

SUBS{<c>}{<q>} {<Rd>,} <Rn>, <Rm> {, <shift> #<amount>}

if Rd == '1111' && S == '1' then SEE "CMP (register)";
if Rn == '1101' then SEE "SUB (SP minus register)";
d = UInt(Rd);  n = UInt(Rn);  m = UInt(Rm);  setflags = (S == '1');
(shift_t, shift_n) = DecodeImmShift(stype, imm3:imm2);
if (d == 15 && !setflags) || n == 15 || m == 15 then UNPREDICTABLE;
// Armv8-A removes UNPREDICTABLE for R13

For more information about the constrained unpredictable behavior of this instruction, see Architectural Constraints on UNPREDICTABLE behaviors.

Assembler Symbols

<c>

See Standard assembler syntax fields.

<q>

See Standard assembler syntax fields.

<Rd>

For encoding A1: is the general-purpose destination register, encoded in the "Rd" field. If omitted, this register is the same as <Rn>. Arm deprecates using the PC as the destination register, but if the PC is used:

For encoding T1 and T2: is the general-purpose destination register, encoded in the "Rd" field. If omitted, this register is the same as <Rn>.

<Rn>

For encoding A1: is the first general-purpose source register, encoded in the "Rn" field. The PC can be used, but this is deprecated. If the SP is used, see SUB (SP minus register).

For encoding T1: is the first general-purpose source register, encoded in the "Rn" field.

For encoding T2: is the first general-purpose source register, encoded in the "Rn" field. If the SP is used, see SUB (SP minus register).

<Rm>

For encoding A1: is the second general-purpose source register, encoded in the "Rm" field. The PC can be used, but this is deprecated.

For encoding T1 and T2: is the second general-purpose source register, encoded in the "Rm" field.

<shift> Is the type of shift to be applied to the second source register, encoded in stype:
stype <shift>
00 LSL
01 LSR
10 ASR
11 ROR
<amount>

For encoding A1: is the shift amount, in the range 1 to 31 (when <shift> = LSL or ROR) or 1 to 32 (when <shift> = LSR or ASR) encoded in the "imm5" field as <amount> modulo 32.

For encoding T2: is the shift amount, in the range 1 to 31 (when <shift> = LSL or ROR) or 1 to 32 (when <shift> = LSR or ASR), encoded in the "imm3:imm2" field as <amount> modulo 32.

Operation

if ConditionPassed() then
    EncodingSpecificOperations();
    shifted = Shift(R[m], shift_t, shift_n, PSTATE.C);
    (result, nzcv) = AddWithCarry(R[n], NOT(shifted), '1');
    if d == 15 then          // Can only occur for A32 encoding
        if setflags then
            ALUExceptionReturn(result);
        else
            ALUWritePC(result);
    else
        R[d] = result;
        if setflags then
            PSTATE.<N,Z,C,V> = nzcv;

Operational information

If CPSR.DIT is 1 and this instruction does not use R15 as either its source or destination:

  • The execution time of this instruction is independent of:
    • The values of the data supplied in any of its registers.
    • The values of the NZCV flags.
  • The response of this instruction to asynchronous exceptions does not vary based on:
    • The values of the data supplied in any of its registers.
    • The values of the NZCV flags.
Previous Next