← Home

TRN1, TRN2 (vectors)

Interleave even or odd elements from two vectors

This instruction interleaves alternating even-numbered or odd-numbered elements from the first and second source vectors, and places the values in the destination vector. This instruction is unpredicated.

The Even (quadwords) and Odd (quadwords) variants require an Effective SVE vector length of at least 256 bits. ID_AA64ZFR0_EL1.F64MM or ID_AA64ZFR0_EL1.EltPerm indicates whether the Even (quadwords) and Odd (quadwords) variants are implemented. The Even (quadwords) and Odd (quadwords) variants are legal when executed in Streaming SVE mode if FEAT_SVE_PERM256 is implemented.

It has encodings from 4 classes: Even , Even (quadwords) , Odd and Odd (quadwords)

Even class

(FEAT_SVE || FEAT_SME)

313029282726252423222120191817161514131211109876543210
00000101size1Zm011100ZnZd
H

Encoding

TRN1 <Zd>.<T>, <Zn>.<T>, <Zm>.<T>

Decode

if !IsFeatureImplemented(FEAT_SVE) && !IsFeatureImplemented(FEAT_SME) then EndOfDecode(Decode_UNDEF); end; let esize : integer{} = 8 << UInt(size); let n : integer = UInt(Zn); let m : integer = UInt(Zm); let d : integer = UInt(Zd); let part : integer = 0;

Even (quadwords) class

(FEAT_F64MM || FEAT_SVE_PERM256)

313029282726252423222120191817161514131211109876543210
00000101101Zm000110ZnZd
opcH

Encoding

TRN1 <Zd>.Q, <Zn>.Q, <Zm>.Q

Decode

if !IsFeatureImplemented(FEAT_F64MM) && !IsFeatureImplemented(FEAT_SVE_PERM256) then EndOfDecode(Decode_UNDEF); end; if MaxImplementedVL() < 256 then EndOfDecode(Decode_UNDEF); end; let esize : integer{} = 128; let n : integer = UInt(Zn); let m : integer = UInt(Zm); let d : integer = UInt(Zd); let part : integer = 0;

Odd class

(FEAT_SVE || FEAT_SME)

313029282726252423222120191817161514131211109876543210
00000101size1Zm011101ZnZd
H

Encoding

TRN2 <Zd>.<T>, <Zn>.<T>, <Zm>.<T>

Decode

if !IsFeatureImplemented(FEAT_SVE) && !IsFeatureImplemented(FEAT_SME) then EndOfDecode(Decode_UNDEF); end; let esize : integer{} = 8 << UInt(size); let n : integer = UInt(Zn); let m : integer = UInt(Zm); let d : integer = UInt(Zd); let part : integer = 1;

Odd (quadwords) class

(FEAT_F64MM || FEAT_SVE_PERM256)

313029282726252423222120191817161514131211109876543210
00000101101Zm000111ZnZd
opcH

Encoding

TRN2 <Zd>.Q, <Zn>.Q, <Zm>.Q

Decode

if !IsFeatureImplemented(FEAT_F64MM) && !IsFeatureImplemented(FEAT_SVE_PERM256) then EndOfDecode(Decode_UNDEF); end; if MaxImplementedVL() < 256 then EndOfDecode(Decode_UNDEF); end; let esize : integer{} = 128; let n : integer = UInt(Zn); let m : integer = UInt(Zm); let d : integer = UInt(Zd); let part : integer = 1;

Assembler Symbols

<Zd>

Is the name of the destination scalable vector register, encoded in the "Zd" field.

<T>

Is the size specifier, encoded in size:

size <T>
00 B
01 H
10 S
11 D
<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, encoded in the "Zm" field.

Operation

if esize < 128 || IsFeatureImplemented(FEAT_SVE_PERM256) then CheckSVEEnabled(); else CheckNonStreamingSVEEnabled(); end; let VL : integer{} = CurrentVL(); if VL < esize * 2 then EndOfDecode(Decode_UNDEF); end; let pairs : integer = VL DIV (esize * 2); let operand1 : bits(VL) = Z{}(n); let operand2 : bits(VL) = Z{}(m); var result : bits(VL) = Zeros{}; for p = 0 to pairs-1 do result[(2*p+0)*:esize] = operand1[(2*p+part)*:esize]; result[(2*p+1)*:esize] = operand2[(2*p+part)*:esize]; end; Z{VL}(d) = result;

Operational information

This instruction is a data-independent-time instruction as described in About PSTATE.DIT.


Version 2026.09 — Copyright © 2010-2026 Arm Limited or its affiliates.

This site is provided as a community resource and is NOT affiliated with nor endorsed by Arm Limited.