LD3B (scalar plus immediate)

Contiguous load three-byte structures to three vectors (immediate index)

Contiguous load three-byte structures, each to the same element number in three vector registers from the memory address generated by a 64-bit scalar base and an immediate index which is a multiple of 3 in the range -24 to 21 that is multiplied by the vector's in-memory size, irrespective of predication,

Each predicate element applies to the same element number in each of the three vector registers, or equivalently to the three consecutive bytes in memory which make up each structure. Inactive elements will not cause a read from Device memory or signal a fault, and the corresponding element is set to zero in each of the three destination vector registers.

313029282726252423222120191817161514131211109876543210
101001000100imm4111PgRnZt
msz<1>msz<0>

LD3B { <Zt1>.B, <Zt2>.B, <Zt3>.B }, <Pg>/Z, [<Xn|SP>{, #<imm>, MUL VL}]

if !IsFeatureImplemented(FEAT_SVE) && !IsFeatureImplemented(FEAT_SME) then UNDEFINED; constant integer t = UInt(Zt); constant integer n = UInt(Rn); constant integer g = UInt(Pg); constant integer esize = 8; constant integer offset = SInt(imm4); constant integer nreg = 3;

Assembler Symbols

<Zt1>

Is the name of the first scalable vector register to be transferred, encoded in the "Zt" field.

<Zt2>

Is the name of the second scalable vector register to be transferred, encoded as "Zt" plus 1 modulo 32.

<Zt3>

Is the name of the third scalable vector register to be transferred, encoded as "Zt" plus 2 modulo 32.

<Pg>

Is the name of the governing scalable predicate register P0-P7, encoded in the "Pg" field.

<Xn|SP>

Is the 64-bit name of the general-purpose base register or stack pointer, encoded in the "Rn" field.

<imm>

Is the optional signed immediate vector offset, a multiple of 3 in the range -24 to 21, defaulting to 0, encoded in the "imm4" field.

Operation

CheckSVEEnabled(); constant integer VL = CurrentVL; constant integer PL = VL DIV 8; constant integer elements = VL DIV esize; bits(64) base; constant bits(PL) mask = P[g, PL]; bits(64) addr; constant integer mbytes = esize DIV 8; array [0..2] of bits(VL) values; constant boolean contiguous = TRUE; constant boolean nontemporal = FALSE; constant boolean tagchecked = n != 31; constant AccessDescriptor accdesc = CreateAccDescSVE(MemOp_LOAD, nontemporal, contiguous, tagchecked); if !AnyActiveElement(mask, esize) then if n == 31 && ConstrainUnpredictableBool(Unpredictable_CHECKSPNONEACTIVE) then CheckSPAlignment(); else if n == 31 then CheckSPAlignment(); base = if n == 31 then SP[] else X[n, 64]; addr = AddressAdd(base, offset * elements * nreg * mbytes, accdesc); for e = 0 to elements-1 for r = 0 to nreg-1 if ActivePredicateElement(mask, e, esize) then Elem[values[r], e, esize] = Mem[addr, mbytes, accdesc]; else Elem[values[r], e, esize] = Zeros(esize); addr = AddressIncrement(addr, mbytes, accdesc); for r = 0 to nreg-1 Z[(t+r) MOD 32, VL] = values[r];

Operational information

If FEAT_SVE2 is implemented or FEAT_SME is implemented, then if PSTATE.DIT is 1, the timing of this instruction is insensitive to the value of the data being loaded or stored when its governing predicate register contains the same value for each execution.


Internal version only: aarchmrs v2024-03_relA, pseudocode v2024-03_rel, sve v2024-03_rel ; Build timestamp: 2024-03-26T09:45

Copyright © 2010-2024 Arm Limited or its affiliates. All rights reserved. This document is Non-Confidential.