SQDMLSLBT

Signed saturating doubling multiply-subtract long from accumulator (bottom × top)

Multiply then double the corresponding even-numbered signed elements of the first and odd-numbered signed elements of the second source vector. Each intermediate value is saturated to the double-width N-bit value's signed integer range -2(N-1) to (2(N-1) )-1. Then destructively subtract from the overlapping double-width elements of the addend and destination vector. Each destination element is saturated to the double-width N-bit element's signed integer range -2(N-1) to (2(N-1) )-1. This instruction is unpredicated.

313029282726252423222120191817161514131211109876543210
01000100size0Zm000011ZnZda
S

SQDMLSLBT <Zda>.<T>, <Zn>.<Tb>, <Zm>.<Tb>

if !IsFeatureImplemented(FEAT_SVE2) && !IsFeatureImplemented(FEAT_SME) then UNDEFINED; if size == '00' then UNDEFINED; constant integer esize = 8 << UInt(size); constant integer n = UInt(Zn); constant integer m = UInt(Zm); constant integer da = UInt(Zda); constant integer sel1 = 0; constant integer sel2 = 1;

Assembler Symbols

<Zda>

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

<T>

Is the size specifier, encoded in size:

size <T>
00 RESERVED
01 H
10 S
11 D
<Zn>

Is the name of the first source scalable vector register, encoded in the "Zn" field.

<Tb>

Is the size specifier, encoded in size:

size <Tb>
00 RESERVED
01 B
10 H
11 S
<Zm>

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

Operation

CheckSVEEnabled(); constant integer VL = CurrentVL; constant integer PL = VL DIV 8; constant integer elements = VL DIV esize; constant bits(VL) operand1 = Z[n, VL]; constant bits(VL) operand2 = Z[m, VL]; bits(VL) result = Z[da, VL]; for e = 0 to elements-1 constant integer element1 = SInt(Elem[operand1, 2 * e + sel1, esize DIV 2]); constant integer element2 = SInt(Elem[operand2, 2 * e + sel2, esize DIV 2]); constant integer element3 = SInt(Elem[result, e, esize]); constant integer product = SInt(SignedSat(2 * element1 * element2, esize)); Elem[result, e, esize] = SignedSat(element3 - product, esize); Z[da, VL] = result;

Operational information

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


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.