VCVT (between floating-point and fixed-point, Advanced SIMD)

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

The vector elements are the same type, and are floating-point numbers or integers. Signed and unsigned integers are distinct.

The floating-point to fixed-point operation uses the Round towards Zero rounding mode. The fixed-point to floating-point operation uses the Round to Nearest rounding mode.

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
1111001U1Dimm6Vd11op0QM1Vm

64-bit SIMD vector (imm6 != 000xxx && Q == 0)

VCVT{<c>}{<q>}.<dt1>.<dt2> <Dd>, <Dm>, #<fbits>

128-bit SIMD vector (imm6 != 000xxx && Q == 1)

VCVT{<c>}{<q>}.<dt1>.<dt2> <Qd>, <Qm>, #<fbits>

if imm6 IN {'000xxx'} then SEE "Related encodings"; if op<1> == '0' && !HaveFP16Ext() then UNDEFINED; if op<1> == '0' && imm6 IN {'10xxxx'} then UNDEFINED; if imm6 IN {'0xxxxx'} then UNDEFINED; if Q == '1' && (Vd<0> == '1' || Vm<0> == '1') then UNDEFINED; to_fixed = (op<0> == '1'); frac_bits = 64 - UInt(imm6); unsigned = (U == '1'); integer esize; integer elements; case op<1> of when '0' esize = 16; elements = 4; when '1' esize = 32; elements = 2; d = UInt(D:Vd); m = UInt(M:Vm); regs = if Q == '0' then 1 else 2;

T1

15141312111098765432101514131211109876543210
111U11111Dimm6Vd11op0QM1Vm

64-bit SIMD vector (imm6 != 000xxx && Q == 0)

VCVT{<c>}{<q>}.<dt1>.<dt2> <Dd>, <Dm>, #<fbits>

128-bit SIMD vector (imm6 != 000xxx && Q == 1)

VCVT{<c>}{<q>}.<dt1>.<dt2> <Qd>, <Qm>, #<fbits>

if imm6 IN {'000xxx'} then SEE "Related encodings"; if op<1> == '0' && !HaveFP16Ext() then UNDEFINED; if op<1> == '0' && imm6 IN {'10xxxx'} then UNDEFINED; if imm6 IN {'0xxxxx'} then UNDEFINED; if Q == '1' && (Vd<0> == '1' || Vm<0> == '1') then UNDEFINED; to_fixed = (op<0> == '1'); frac_bits = 64 - UInt(imm6); unsigned = (U == '1'); integer esize; integer elements; case op<1> of when '0' esize = 16; elements = 4; when '1' esize = 32; elements = 2; d = UInt(D:Vd); m = UInt(M:Vm); regs = if Q == '0' then 1 else 2;

Related encodings: See Advanced SIMD one register and modified immediate for the T32 instruction set, or Advanced SIMD one register and modified immediate for the A32 instruction set.

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.

<dt1>

Is the data type for the elements of the destination vector, encoded in op:U:

op U <dt1>
00 x F16
01 0 S16
01 1 U16
10 x F32
11 0 S32
11 1 U32
<dt2>

Is the data type for the elements of the source vector, encoded in op:U:

op U <dt2>
00 0 S16
00 1 U16
01 x F16
10 0 S32
10 1 U32
11 x F32
<Qd>

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

<Qm>

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

<Dd>

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

<Dm>

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

<fbits>

The number of fraction bits in the fixed point number, in the range 1 to 32 for 32-bit elements, or in the range 1 to 16 for 16-bit elements:

  • (64 - <fbits>) is encoded in imm6.

An assembler can permit an <fbits> value of 0. This is encoded as floating-point to integer or integer to floating-point instruction, see VCVT (between floating-point and integer, Advanced SIMD).

Operation

if ConditionPassed() then EncodingSpecificOperations(); CheckAdvSIMDEnabled(); bits(esize) result; for r = 0 to regs-1 for e = 0 to elements-1 op1 = Elem[D[m+r],e,esize]; if to_fixed then result = FPToFixed(op1, frac_bits, unsigned, StandardFPSCRValue(), FPRounding_ZERO, esize); else result = FixedToFP(op1, frac_bits, unsigned, StandardFPSCRValue(), FPRounding_TIEEVEN, esize); Elem[D[d+r],e,esize] = result;


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.