VCVT (between half-precision and single-precision, Advanced SIMD)

Vector Convert between half-precision and single-precision converts each element in a vector from single-precision to half-precision floating-point, or from half-precision to single-precision, and places the results in a second vector.

Depending on settings in the CPACR, NSACR, and HCPTR registers, and the Security state and PE mode in which the instruction is executed, an attempt to execute the instruction might be undefined, or trapped to Hyp mode. For more information see Enabling Advanced SIMD and floating-point support.

It has encodings from the following instruction sets: A32 ( A1 ) and T32 ( T1 ) .

A1

313029282726252423222120191817161514131211109876543210
111100111D11size10Vd011op00M0Vm

Half-precision to single-precision (op == 1)

VCVT{<c>}{<q>}.F32.F16 <Qd>, <Dm> // (Encoded as op = 1)

Single-precision to half-precision (op == 0)

VCVT{<c>}{<q>}.F16.F32 <Dd>, <Qm> // (Encoded as op = 0)

if size != '01' then UNDEFINED; half_to_single = (op == '1'); if half_to_single && Vd<0> == '1' then UNDEFINED; if !half_to_single && Vm<0> == '1' then UNDEFINED; esize = 16; elements = 4; m = UInt(M:Vm); d = UInt(D:Vd);

T1

15141312111098765432101514131211109876543210
111111111D11size10Vd011op00M0Vm

Half-precision to single-precision (op == 1)

VCVT{<c>}{<q>}.F32.F16 <Qd>, <Dm> // (Encoded as op = 1)

Single-precision to half-precision (op == 0)

VCVT{<c>}{<q>}.F16.F32 <Dd>, <Qm> // (Encoded as op = 0)

if size != '01' then UNDEFINED; half_to_single = (op == '1'); if half_to_single && Vd<0> == '1' then UNDEFINED; if !half_to_single && Vm<0> == '1' then UNDEFINED; esize = 16; elements = 4; m = UInt(M:Vm); d = UInt(D:Vd);

Assembler Symbols

<c>

For encoding A1: see Standard assembler syntax fields. This encoding must be unconditional.

For encoding T1: see Standard assembler syntax fields.

<q>

See Standard assembler syntax fields.

<Qd>

Is the 128-bit name of the SIMD&FP destination register, encoded in the "D:Vd" field as <Qd>*2.

<Dm>

Is the 64-bit name of the SIMD&FP source register, encoded in the "M:Vm" field.

<Dd>

Is the 64-bit name of the SIMD&FP destination register, encoded in the "D:Vd" field.

<Qm>

Is the 128-bit name of the SIMD&FP source register, encoded in the "M:Vm" field as <Qm>*2.

Operation

if ConditionPassed() then EncodingSpecificOperations(); CheckAdvSIMDEnabled(); for e = 0 to elements-1 if half_to_single then Elem[Q[d>>1],e,32] = FPConvert(Elem[Din[m],e,16], StandardFPSCRValue(), 32); else Elem[D[d],e,16] = FPConvert(Elem[Qin[m>>1],e,32], StandardFPSCRValue(), 16);


Internal version only: isa v01_31, pseudocode v2023-06_rel, sve v2023-06_rel ; Build timestamp: 2023-07-04T18:06

Copyright © 2010-2023 Arm Limited or its affiliates. All rights reserved. This document is Non-Confidential.