You copied the Doc URL to your clipboard.


Extract vector from pair of vectors.

Copy the indexed byte up to the last byte of the first source vector to the bottom of the result vector, then fill the remainder of the result starting from the first byte of the second source vector. The result is placed destructively in the first source vector. This instruction is unpredicated.

An index that is greater than or equal to the vector length in bytes is treated as zero, leaving the destination and first source vector unmodified.


EXT <Zdn>.B, <Zdn>.B, <Zm>.B, #<imm>

if !HaveSVE() then UNDEFINED;
integer esize = 8;
integer dn = UInt(Zdn);
integer m = UInt(Zm);
integer position = UInt(imm8h:imm8l);

Assembler Symbols


Is the name of the first source and destination scalable vector register, encoded in the "Zdn" field.


Is the name of the second source scalable vector register, encoded in the "Zm" field.


Is the unsigned immediate operand, in the range 0 to 255, encoded in the "imm8h:imm8l" fields.


integer elements = VL DIV esize;
bits(VL) operand1 = Z[dn];
bits(VL) operand2 = Z[m];
bits(VL) result;

if position >= elements then
    position = 0;

position = position << 3;
bits(VL*2) concat = operand2 : operand1;
result = concat<position+VL-1:position>;

Z[dn] = result;

Operational information

This instruction might be immediately preceded in program order by a MOVPRFX instruction. The MOVPRFX instruction must conform to all of the following requirements, otherwise the behavior of the MOVPRFX and this instruction is unpredictable:

  • The MOVPRFX instruction must be unpredicated.
  • The MOVPRFX instruction must specify the same destination register as this instruction.
  • The destination register must not refer to architectural register state referenced by any other source operand register of this instruction.