DUP (element)

Duplicate vector element to vector or scalar

This instruction duplicates the vector element at the specified element index in the source SIMD&FP register into a scalar or each element in a vector, and writes the result to the destination SIMD&FP register.

Depending on the settings in the CPACR_EL1, CPTR_EL2, and CPTR_EL3 registers, and the current Security state and Exception level, an attempt to execute the instruction might be trapped.

This instruction is used by the alias MOV (scalar).

It has encodings from 2 classes: Scalar and Vector

Scalar
(FEAT_AdvSIMD)

Decode for this encoding

if !IsFeatureImplemented(FEAT_AdvSIMD) then EndOfDecode(Decode_UNDEF); if imm5 IN 'x0000' then EndOfDecode(Decode_UNDEF); constant integer size = LowestSetBitNZ(imm5<3:0>); constant integer d = UInt(Rd); constant integer n = UInt(Rn); constant integer index = UInt(imm5<4:size+1>); constant integer idxdsize = 64 << UInt(imm5<4>); constant integer esize = 8 << size; constant integer datasize = esize; constant integer elements = 1;

Vector
(FEAT_AdvSIMD)

Decode for this encoding

if !IsFeatureImplemented(FEAT_AdvSIMD) then EndOfDecode(Decode_UNDEF); if imm5 IN 'x0000' then EndOfDecode(Decode_UNDEF); if imm5 IN 'x1000' && Q == '0' then EndOfDecode(Decode_UNDEF); constant integer size = LowestSetBitNZ(imm5<3:0>); constant integer d = UInt(Rd); constant integer n = UInt(Rn); constant integer index = UInt(imm5<4:size+1>); constant integer idxdsize = 64 << UInt(imm5<4>); constant integer esize = 8 << size; constant integer datasize = 64 << UInt(Q); constant integer elements = datasize DIV esize;