(old) htmldiff from-(new)

SUDOT

Signed by unsigned integer indexed dot product.

The signed by unsigned integer indexed dot product instruction computes the dot product of a group of four signed 8-bit integer values held in each 32-bit element of the first source vector multiplied by a group of four unsigned 8-bit integer values in an indexed 32-bit element of the second source vector, and then destructively adds the widened dot product to the corresponding 32-bit element of the destination vector.

The groups within the second source vector are specified using an immediate index which selects the same group position within each 128-bit vector segment. The index range is from 0 to 3. This instruction is unpredicated.

ID_AA64ZFR0_EL1.I8MM indicates whether this instruction is implemented.

313029282726252423222120191817161514131211109876543210
01000100101i2Zm000111ZnZda

SVE

SUDOT <Zda>.S, <Zn>.B, <Zm>.B[<imm>]

if !HaveSVE() || !HaveInt8MatMulExt() then UNDEFINED; integer esize = 32; integer index = UInt(i2); integer n = UInt(Zn); integer m = UInt(Zm); integer da = UInt(Zda);

Assembler Symbols

<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.

<Zm>

Is the name of the second source scalable vector register Z0-Z7, encoded in the "Zm" field.

<imm>

Is the immediate index of a quadtuplet of four 8-bit elements within each 128-bit vector segment, in the range 0 to 3, encoded in the "i2" field.

Operation

CheckSVEEnabled(); integer elements = VL DIV esize; integer eltspersegment = 128 DIV esize; bits(VL) operand1 = Z[n]; bits(VL) operand2 = Z[m]; bits(VL) operand3 = Z[da]; bits(VL) result; for e = 0 to elements-1 integer segmentbase = e - e MOD eltspersegment; integer s = segmentbase + index; bits(esize) res = Elem[operand3, e, esize]; for i = 0 to 3 integer element1 = SInt(Elem[operand1, 4 * e + i, esize DIV 4]); integer element2 = UInt(Elem[operand2, 4 * s + i, esize DIV 4]); res = res + element1 * element2; Elem[result, e, esize] = res; Z[da] = result;

Operational information

ThisIf instructionPSTATE.DIT mightis be immediately preceded in program order by a1: MOVPRFX instruction. The MOVPRFX instruction must conform to all of the following requirements, otherwise the behavior of the MOVPRFX and this instruction is unpredictable:

This instruction might be immediately preceded in program order by a MOVPRFX instruction that conforms to all of the following requirements, otherwise the behavior of either or both instructions is unpredictable:

The MOVPRFX instructions that can be used with this instruction are as follows:


Internal version only: isa v31.05bv31.04, AdvSIMD v29.02, pseudocode v2019-12_rc3_1v2019-09_rc2_1, sve v2019-12_rc3v2019-09_rc3 ; Build timestamp: 2019-12-13T112019-09-27T17:0228

Copyright © 2010-2019 Arm Limited or its affiliates. All rights reserved. This document is Non-Confidential.

(old) htmldiff from-(new)