(old) | htmldiff from- | (new) |
Signed multiply-subtract long from accumulator (top, indexed).
Multiply the odd-numbered signed elements within each 128-bit segment of the first source vector by the specified signed element in the corresponding second source vector segment and destructively subtract from the overlapping double-width elements of the addend vector.
The elements within the second source vector are specified using an immediate index which 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, encoded in 2 or 3 bits depending on the size of the element.
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 | 1 | 0 | i3l | 1 | Zn | Zda | |||||||||||
size<1> | size<0> | S | U | T |
if !HaveSVE2() then UNDEFINED; integer esize = 16; integer index = UInt(i3h:i3l); integer n = UInt(Zn); integer m = UInt(Zm); integer da = UInt(Zda); integer sel = 1;
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 | 1 | 0 | i2l | 1 | Zn | Zda | |||||||||||
size<1> | size<0> | S | U | T |
if !HaveSVE2() then UNDEFINED; integer esize = 32; integer index = UInt(i2h:i2l); integer n = UInt(Zn); integer m = UInt(Zm); integer da = UInt(Zda); integer sel = 1;
<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();
integer elements =CheckSVEEnabled();
integer elements = VL DIV (2 * esize);
integer eltspersegment = 128 DIV (2 * esize);
bits(VL) operand1 = Z[n];
bits(VL) operand2 = Z[m];
bits(VL) result = Z[da];
for e = 0 to elements-1
integer s = e - e MOD eltspersegment;
integer element1 = VL DIV (2 * esize);
integer eltspersegment = 128 DIV (2 * esize);
bits(VL) operand1 = Z[n];
bits(VL) operand2 = Z[m];
bits(VL) result = Z[da];
for e = 0 to elements-1
integer s = e - (e MOD eltspersegment);
integer element1 = SInt(Elem[operand1, 2 * e + sel, esize]);
integer element2 = SInt(Elem[operand2, 2 * s + index, esize]);
bits(2*esize) product = (element1 * element2)<2*esize-1:0>;
Elem[result, e, 2*esize] = Elem[result, e, 2*esize] - product;
Z[da] = result;[result, e, 2*esize] - product;
Z[da] = result;
If PSTATE.DIT is 1:
This instruction might be immediately preceded in program order by a MOVPRFX instruction. The MOVPRFX instruction must conform to all of the following requirements, otherwise the behavior of the MOVPRFX and this instruction is unpredictable:
Internal version only: isa v32.03v31.05b, AdvSIMD v29.02, pseudocode v2020-03v2019-12_rc3_1, sve v2020-03_rc1v2019-12_rc3
; Build timestamp: 2020-04-15T132019-12-13T11:3302
Copyright © 2010-20202010-2019 Arm Limited or its affiliates. All rights reserved.
This document is Non-Confidential.
(old) | htmldiff from- | (new) |