Unsigned multiply-add long by indexed element (bottom)
This instruction multiplies the even-numbered unsigned elements within each 128-bit segment of the first source vector by the specified unsigned element in the corresponding second source vector segment, and destructively adds the results to the overlapping double-width elements of the addend vector.
The elements within the second source vector are specified using an immediate index that selects the same element position within each 128-bit vector segment. The index range is from 0 to one less than the number of elements per 128-bit segment.
It has encodings from 2 classes: 32-bit and 64-bit
| 31 | 30 | 29 | 28 | 27 | 26 | 25 | 24 | 23 | 22 | 21 | 20 | 19 | 18 | 17 | 16 | 15 | 14 | 13 | 12 | 11 | 10 | 9 | 8 | 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
| 0 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 1 | 0 | 1 | i3h | Zm | 1 | 0 | 0 | 1 | i3l | 0 | Zn | Zda | |||||||||||
| size | S | U | T | ||||||||||||||||||||||||||||
if !IsFeatureImplemented(FEAT_SVE2) && !IsFeatureImplemented(FEAT_SME) then EndOfDecode(Decode_UNDEF); end; let esize : integer{} = 16; let index : integer = UInt(i3h::i3l); let n : integer = UInt(Zn); let m : integer = UInt(Zm); let da : integer = UInt(Zda); let sel : integer = 0;
| 31 | 30 | 29 | 28 | 27 | 26 | 25 | 24 | 23 | 22 | 21 | 20 | 19 | 18 | 17 | 16 | 15 | 14 | 13 | 12 | 11 | 10 | 9 | 8 | 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
| 0 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | i2h | Zm | 1 | 0 | 0 | 1 | i2l | 0 | Zn | Zda | |||||||||||
| size | S | U | T | ||||||||||||||||||||||||||||
if !IsFeatureImplemented(FEAT_SVE2) && !IsFeatureImplemented(FEAT_SME) then EndOfDecode(Decode_UNDEF); end; let esize : integer{} = 32; let index : integer = UInt(i2h::i2l); let n : integer = UInt(Zn); let m : integer = UInt(Zm); let da : integer = UInt(Zda); let sel : integer = 0;
| <Zda> |
Is the name of the third source and destination scalable vector register, encoded in the "Zda" field. |
| <Zn> |
Is the name of the first source scalable vector register, encoded in the "Zn" field. |
CheckSVEEnabled(); let VL : integer{} = CurrentVL(); let elements : integer = VL DIV (2 * esize); let eltspersegment : integer = 128 DIV (2 * esize); let operand1 : bits(VL) = Z{}(n); let operand2 : bits(VL) = Z{}(m); var result : bits(VL) = Z{}(da); for e = 0 to elements-1 do let s : integer = e - (e MOD eltspersegment); let element1 : integer = UInt(operand1[(2 * e + sel)*:esize]); let element2 : integer = UInt(operand2[(2 * s + index)*:esize]); let product : bits(2*esize) = (element1 * element2)[2*esize-1:0]; result[e*:(2*esize)] = result[e*:(2*esize)] + product; end; Z{VL}(da) = result;
This instruction is a data-independent-time instruction as described in About PSTATE.DIT.
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:
2025-12_rel 2025-12-10 16:41:45
Copyright © 2010-2025 Arm Limited or its affiliates. All rights reserved. This document is Non-Confidential.