UCVTF (scalar, fixed-point)

Unsigned fixed-point Convert to Floating-point (scalar). This instruction converts the unsigned value in the 32-bit or 64-bit general-purpose source register to a floating-point value using the rounding mode that is specified by the FPCR, and writes the result to the SIMD&FP destination register.

A floating-point exception can be generated by this instruction. Depending on the settings in FPCR, the exception results in either a flag being set in FPSR, or a synchronous exception being generated. For more information, see Floating-point exception traps.

Depending on the settings in the CPACR_EL1, CPTR_EL2, and CPTR_EL3 registers, and the Security state and Exception level in which the instruction is executed, an attempt to execute the instruction might be trapped.

313029282726252423222120191817161514131211109876543210
sf0011110ftype000011scaleRnRd
Srmodeopcode

32-bit to half-precision (sf == 0 && ftype == 11)
(FEAT_FP16)

UCVTF <Hd>, <Wn>, #<fbits>

64-bit to half-precision (sf == 1 && ftype == 11)
(FEAT_FP16)

UCVTF <Hd>, <Xn>, #<fbits>

32-bit to single-precision (sf == 0 && ftype == 00)

UCVTF <Sd>, <Wn>, #<fbits>

64-bit to single-precision (sf == 1 && ftype == 00)

UCVTF <Sd>, <Xn>, #<fbits>

32-bit to double-precision (sf == 0 && ftype == 01)

UCVTF <Dd>, <Wn>, #<fbits>

64-bit to double-precision (sf == 1 && ftype == 01)

UCVTF <Dd>, <Xn>, #<fbits>

if ftype == '10' || (ftype == '11' && !IsFeatureImplemented(FEAT_FP16)) then UNDEFINED; if sf == '0' && scale<5> == '0' then UNDEFINED; integer d = UInt(Rd); integer n = UInt(Rn); constant integer intsize = 32 << UInt(sf); constant integer decode_fltsize = 8 << UInt(ftype EOR '10'); integer fracbits = 64 - UInt(scale); boolean unsigned = TRUE;

Assembler Symbols

<Hd>

Is the 16-bit name of the SIMD&FP destination register, encoded in the "Rd" field.

<Wn>

Is the 32-bit name of the general-purpose source register, encoded in the "Rn" field.

<fbits>

For the 32-bit to double-precision, 32-bit to half-precision and 32-bit to single-precision variant: is the number of bits after the binary point in the fixed-point source, in the range 1 to 32, encoded as 64 minus "scale".

For the 64-bit to double-precision, 64-bit to half-precision and 64-bit to single-precision variant: is the number of bits after the binary point in the fixed-point source, in the range 1 to 64, encoded as 64 minus "scale".

<Xn>

Is the 64-bit name of the general-purpose source register, encoded in the "Rn" field.

<Sd>

Is the 32-bit name of the SIMD&FP destination register, encoded in the "Rd" field.

<Dd>

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

Operation

CheckFPEnabled64(); boolean merge = IsMerging(FPCR); constant integer fltsize = if merge then 128 else decode_fltsize; bits(fltsize) fltval = if merge then V[d, fltsize] else Zeros(fltsize); bits(intsize) intval = X[n, intsize]; FPRounding rounding = FPRoundingMode(FPCR); Elem[fltval, 0, decode_fltsize] = FixedToFP(intval, fracbits, unsigned, FPCR, rounding, decode_fltsize); V[d, fltsize] = fltval;


Internal version only: aarchmrs v2024-03_relA, pseudocode v2024-03_rel, sve v2024-03_rel ; Build timestamp: 2024-03-26T09:45

Copyright © 2010-2024 Arm Limited or its affiliates. All rights reserved. This document is Non-Confidential.