LLVM: llvm::detail::IEEEFloat Class Reference (original) (raw)

#include "[llvm/ADT/APFloat.h](APFloat%5F8h%5Fsource.html)"

Public Member Functions
bool needsCleanup () const
Returns whether this instance allocated memory.
bool operator== (const IEEEFloat &) const =delete
The definition of equality is not straightforward for floating point, so we won't use operator==.
LLVM_ABI cmpResult compare (const IEEEFloat &) const
IEEE comparison with another floating point number (NaNs compare unordered, 0==-0).
LLVM_ABI bool bitwiseIsEqual (const IEEEFloat &) const
Bitwise comparison for equality (QNaNs compare equal, 0!=-0).
LLVM_ABI unsigned int convertToHexString (char *dst, unsigned int hexDigits, bool upperCase, roundingMode) const
Write out a hexadecimal representation of the floating point value to DST, which must be of sufficient size, in the C99 form [-]0xh.hhhhp[+-]d.
LLVM_ABI IEEEFloat & operator= (const IEEEFloat &)
LLVM_ABI IEEEFloat & operator= (IEEEFloat &&)
LLVM_ABI void toString (SmallVectorImpl< char > &Str, unsigned FormatPrecision=0, unsigned FormatMaxPadding=3, bool TruncateZero=true) const
Converts this value into a decimal string.
LLVM_ABI LLVM_READONLY int getExactLog2Abs () const
LLVM_ABI cmpResult compareAbsoluteValue (const IEEEFloat &) const
Constructors
LLVM_ABI IEEEFloat (const fltSemantics &)
LLVM_ABI IEEEFloat (const fltSemantics &, integerPart)
LLVM_ABI IEEEFloat (const fltSemantics &, uninitializedTag)
LLVM_ABI IEEEFloat (const fltSemantics &, const APInt &)
LLVM_ABI IEEEFloat (double d)
LLVM_ABI IEEEFloat (float f)
LLVM_ABI IEEEFloat (const IEEEFloat &)
LLVM_ABI IEEEFloat (IEEEFloat &&)
LLVM_ABI ~IEEEFloat ()
Arithmetic
LLVM_ABI opStatus add (const IEEEFloat &, roundingMode)
LLVM_ABI opStatus subtract (const IEEEFloat &, roundingMode)
LLVM_ABI opStatus multiply (const IEEEFloat &, roundingMode)
LLVM_ABI opStatus divide (const IEEEFloat &, roundingMode)
LLVM_ABI opStatus remainder (const IEEEFloat &)
IEEE remainder.
LLVM_ABI opStatus mod (const IEEEFloat &)
C fmod, or llvm frem.
LLVM_ABI opStatus fusedMultiplyAdd (const IEEEFloat &, const IEEEFloat &, roundingMode)
LLVM_ABI opStatus roundToIntegral (roundingMode)
LLVM_ABI opStatus next (bool nextDown)
IEEE-754R 5.3.1: nextUp/nextDown.
Sign operations.
LLVM_ABI void changeSign ()
Conversions
LLVM_ABI opStatus convert (const fltSemantics &, roundingMode, bool *)
IEEEFloat::convert - convert a value of one floating point type to another.
LLVM_ABI opStatus convertToInteger (MutableArrayRef< integerPart >, unsigned int, bool, roundingMode, bool *) const
LLVM_ABI opStatus convertFromAPInt (const APInt &, bool, roundingMode)
LLVM_ABI Expected< opStatus > convertFromString (StringRef, roundingMode)
LLVM_ABI APInt bitcastToAPInt () const
LLVM_ABI double convertToDouble () const
LLVM_ABI float convertToFloat () const
IEEE-754R 5.7.2 General operations.
bool isNegative () const
IEEE-754R isSignMinus: Returns true if and only if the current value is negative.
bool isNormal () const
IEEE-754R isNormal: Returns true if and only if the current value is normal.
bool isFinite () const
Returns true if and only if the current value is zero, subnormal, or normal.
bool isZero () const
Returns true if and only if the float is plus or minus zero.
LLVM_ABI bool isDenormal () const
IEEE-754R isSubnormal(): Returns true if and only if the float is a denormal.
bool isInfinity () const
IEEE-754R isInfinite(): Returns true if and only if the float is infinity.
bool isNaN () const
Returns true if and only if the float is a quiet or signaling NaN.
LLVM_ABI bool isSignaling () const
Returns true if and only if the float is a signaling NaN.
Special value setters.
LLVM_ABI void makeLargest (bool Neg=false)
Make this number the largest magnitude normal number in the given semantics.
LLVM_ABI void makeSmallest (bool Neg=false)
Make this number the smallest magnitude denormal number in the given semantics.
LLVM_ABI void makeNaN (bool SNaN=false, bool Neg=false, const APInt *fill=nullptr)
LLVM_ABI void makeInf (bool Neg=false)
LLVM_ABI void makeZero (bool Neg=false)
LLVM_ABI void makeQuiet ()
LLVM_ABI void makeSmallestNormalized (bool Negative=false)
Returns the smallest (by magnitude) normalized finite number in the given semantics.
Friends
class IEEEFloatUnitTestHelper
LLVM_ABI friend hash_code hash_value (const IEEEFloat &Arg)
Overload to compute a hash code for an APFloat value.
LLVM_ABI friend int ilogb (const IEEEFloat &Arg)
LLVM_ABI friend IEEEFloat scalbn (IEEEFloat X, int Exp, roundingMode RoundingMode)
LLVM_ABI friend IEEEFloat frexp (const IEEEFloat &X, int &Exp, roundingMode RM)
Simple Queries
fltCategory getCategory () const
const fltSemantics & getSemantics () const
bool isNonZero () const
bool isFiniteNonZero () const
bool isPosZero () const
bool isNegZero () const
LLVM_ABI bool isSmallest () const
Returns true if and only if the number has the smallest possible non-zero magnitude in the current semantics.
LLVM_ABI bool isSmallestNormalized () const
Returns true if this is the smallest (by magnitude) normalized finite number in the given semantics.
LLVM_ABI bool isLargest () const
Returns true if and only if the number has the largest possible finite magnitude in the current semantics.
LLVM_ABI bool isInteger () const
Returns true if and only if the number is an exact integer.

Definition at line 445 of file APFloat.h.

Definition at line 1195 of file APFloat.cpp.

References llvm::fltSemantics::hasZero, makeSmallestNormalized(), and makeZero().

Referenced by add(), bitwiseIsEqual(), compare(), compareAbsoluteValue(), divide(), frexp, fusedMultiplyAdd(), hash_value, IEEEFloat(), IEEEFloat(), IEEEFloat(), IEEEFloat(), IEEEFloat(), IEEEFloat(), ilogb, isInteger(), mod(), multiply(), operator=(), operator=(), operator==(), remainder(), roundToIntegral(), scalbn, and subtract().

IEEEFloat() [2/8]

IEEEFloat() [3/8]

IEEEFloat() [4/8]

IEEEFloat() [5/8]

llvm::detail::IEEEFloat::IEEEFloat ( double d) explicit

IEEEFloat() [6/8]

llvm::detail::IEEEFloat::IEEEFloat ( float f) explicit

IEEEFloat() [7/8]

IEEEFloat() [8/8]

llvm::detail::IEEEFloat::IEEEFloat ( IEEEFloat && rhs )

~IEEEFloat()

llvm::detail::IEEEFloat::~IEEEFloat ( )

add()

bitcastToAPInt()

APInt llvm::detail::IEEEFloat::bitcastToAPInt ( ) const

bitwiseIsEqual()

changeSign()

void llvm::detail::IEEEFloat::changeSign ( )

compare()

IEEE comparison with another floating point number (NaNs compare unordered, 0==-0).

Definition at line 2483 of file APFloat.cpp.

References assert(), llvm::detail::cmpEqual, llvm::detail::cmpGreaterThan, llvm::detail::cmpLessThan, llvm::detail::cmpUnordered, compareAbsoluteValue(), llvm::detail::fcInfinity, llvm::detail::fcNaN, llvm::detail::fcNormal, llvm::detail::fcZero, IEEEFloat(), llvm_unreachable, and PackCategoriesIntoKey.

Referenced by compareAbsoluteValue(), isInteger(), and remainder().

compareAbsoluteValue()

convert()

IEEEFloat::convert - convert a value of one floating point type to another.

The return value corresponds to the IEEE754 exceptions. *losesInfo records whether the transformation lost information, i.e. whether converting the result back to the original type will produce the original value (this is almost the same as return value==fsOK, but there are edge cases where this is not so).

Definition at line 2560 of file APFloat.cpp.

References llvm::detail::fcInfinity, llvm::detail::fcNaN, llvm::detail::fcZero, isFiniteNonZero(), isSignaling(), llvm::lfExactlyZero, makeNaN(), makeQuiet(), makeSmallestNormalized(), llvm::fltSemantics::minExponent, llvm::fltSemantics::nanEncoding, llvm::NanOnly, llvm::NegativeZero, llvm::fltSemantics::nonFiniteBehavior, llvm::detail::opInexact, llvm::detail::opInvalidOp, llvm::detail::opOK, llvm::partCountForBits(), llvm::fltSemantics::precision, llvm::shiftRight(), llvm::APInt::tcAssign(), llvm::APInt::tcSet(), llvm::APInt::tcSetBit(), and llvm::APInt::tcShiftLeft().

Referenced by remainder().

convertFromAPInt()

convertFromString()

convertToDouble()

double llvm::detail::IEEEFloat::convertToDouble ( ) const

convertToFloat()

float llvm::detail::IEEEFloat::convertToFloat ( ) const

convertToHexString()

Write out a hexadecimal representation of the floating point value to DST, which must be of sufficient size, in the C99 form [-]0xh.hhhhp[+-]d.

Return the number of characters written, excluding the terminating NUL.

Definition at line 3323 of file APFloat.cpp.

References llvm::detail::fcInfinity, llvm::detail::fcNaN, llvm::detail::fcNormal, llvm::detail::fcZero, llvm::infinityL, llvm::infinityU, llvm::NaNL, and llvm::NaNU.

convertToInteger()

divide()

fusedMultiplyAdd()

getCategory()

fltCategory llvm::detail::IEEEFloat::getCategory ( ) const inline

getExactLog2Abs()

int llvm::detail::IEEEFloat::getExactLog2Abs ( ) const

getSemantics()

isDenormal()

bool llvm::detail::IEEEFloat::isDenormal ( ) const

isFinite()

bool llvm::detail::IEEEFloat::isFinite ( ) const inline

isFiniteNonZero()

bool llvm::detail::IEEEFloat::isFiniteNonZero ( ) const inline

Definition at line 576 of file APFloat.h.

References isFinite(), and isZero().

Referenced by bitwiseIsEqual(), compareAbsoluteValue(), convert(), divide(), fusedMultiplyAdd(), hash_value, isDenormal(), isLargest(), isNormal(), isSmallest(), mod(), and multiply().

isInfinity()

bool llvm::detail::IEEEFloat::isInfinity ( ) const inline

isInteger()

bool llvm::detail::IEEEFloat::isInteger ( ) const

isLargest()

bool llvm::detail::IEEEFloat::isLargest ( ) const

isNaN()

bool llvm::detail::IEEEFloat::isNaN ( ) const inline

isNegative()

bool llvm::detail::IEEEFloat::isNegative ( ) const inline

isNegZero()

bool llvm::detail::IEEEFloat::isNegZero ( ) const inline

isNonZero()

bool llvm::detail::IEEEFloat::isNonZero ( ) const inline

isNormal()

bool llvm::detail::IEEEFloat::isNormal ( ) const inline

IEEE-754R isNormal: Returns true if and only if the current value is normal.

This implies that the current value of the float is not zero, subnormal, infinite, or NaN following the definition of normality from IEEE-754R.

Definition at line 544 of file APFloat.h.

References isDenormal(), and isFiniteNonZero().

isPosZero()

bool llvm::detail::IEEEFloat::isPosZero ( ) const inline

isSignaling()

bool llvm::detail::IEEEFloat::isSignaling ( ) const

isSmallest()

bool llvm::detail::IEEEFloat::isSmallest ( ) const

Returns true if and only if the number has the smallest possible non-zero magnitude in the current semantics.

Definition at line 1033 of file APFloat.cpp.

References isFiniteNonZero().

Referenced by next().

isSmallestNormalized()

bool llvm::detail::IEEEFloat::isSmallestNormalized ( ) const

isZero()

bool llvm::detail::IEEEFloat::isZero ( ) const inline

makeInf()

void llvm::detail::IEEEFloat::makeInf ( bool Neg = false )

makeLargest()

void llvm::detail::IEEEFloat::makeLargest ( bool Neg = false )

makeNaN()

void llvm::detail::IEEEFloat::makeNaN ( bool SNaN = false,
bool Neg = false,
const APInt * fill = nullptr )

Definition at line 930 of file APFloat.cpp.

References llvm::detail::fcNaN, llvm::fill(), llvm::FiniteOnly, llvm::APInt::getAllOnes(), llvm::APInt::getZero(), llvm_unreachable, llvm::NanOnly, llvm::NegativeZero, llvm::APInt::tcAssign(), llvm::APInt::tcClearBit(), llvm::APInt::tcIsZero(), llvm::APInt::tcSet(), and llvm::APInt::tcSetBit().

Referenced by convert(), makeInf(), and next().

makeQuiet()

void llvm::detail::IEEEFloat::makeQuiet ( )

makeSmallest()

void llvm::detail::IEEEFloat::makeSmallest ( bool Neg = false )

makeSmallestNormalized()

void llvm::detail::IEEEFloat::makeSmallestNormalized ( bool Negative = false )

makeZero()

void llvm::detail::IEEEFloat::makeZero ( bool Neg = false )

mod()

C fmod, or llvm frem.

Definition at line 2312 of file APFloat.cpp.

References assert(), llvm::detail::cmpLessThan, compareAbsoluteValue(), IEEEFloat(), ilogb, isFiniteNonZero(), isZero(), llvm::NegativeZero, llvm::detail::opOK, llvm::detail::rmNearestTiesToEven, scalbn, and subtract().

Referenced by remainder().

multiply()

needsCleanup()

bool llvm::detail::IEEEFloat::needsCleanup ( ) const inline

Returns whether this instance allocated memory.

Definition at line 463 of file APFloat.h.

next()

IEEE-754R 5.3.1: nextUp/nextDown.

IEEE-754R 2008 5.3.1: nextUp/nextDown.

NOTE since nextDown(x) = -nextUp(-x), we only implement nextUp with appropriate sign switching before/after the computation.

Definition at line 4549 of file APFloat.cpp.

References assert(), changeSign(), llvm::detail::fcInfinity, llvm::detail::fcNaN, llvm::detail::fcNormal, llvm::detail::fcZero, llvm::FiniteOnly, llvm::APFloat::hasSignificand(), isDenormal(), isLargest(), isNegative(), isSignaling(), isSmallest(), makeLargest(), makeNaN(), makeSmallest(), makeSmallestNormalized(), llvm::NanOnly, llvm::NegativeZero, next(), llvm::detail::opInvalidOp, llvm::detail::opOK, llvm::APInt::tcDecrement(), llvm::APInt::tcSet(), and llvm::APInt::tcSetBit().

Referenced by next().

operator=() [1/2]

operator=() [2/2]

operator==()

The definition of equality is not straightforward for floating point, so we won't use operator==.

Use one of the following, or write whatever it is you really mean.

References IEEEFloat(), and LLVM_ABI.

remainder()

IEEE remainder.

Definition at line 2202 of file APFloat.cpp.

References add(), assert(), llvm::detail::cmpEqual, llvm::detail::cmpGreaterThan, compare(), convert(), IEEEFloat(), isZero(), llvm::fltSemantics::maxExponent, llvm::fltSemantics::minExponent, mod(), llvm::NegativeZero, llvm::detail::opDivByZero, llvm::detail::opOK, P, llvm::fltSemantics::precision, llvm::detail::rmNearestTiesToEven, and subtract().

roundToIntegral()

Definition at line 2396 of file APFloat.cpp.

References add(), assert(), changeSign(), convertFromAPInt(), IEEEFloat(), isInfinity(), isNaN(), isNegative(), isSignaling(), isZero(), makeQuiet(), llvm::NextPowerOf2(), llvm::detail::opInvalidOp, llvm::detail::opOK, llvm::detail::rmNearestTiesToEven, llvm::APFloatBase::semanticsPrecision(), and subtract().

subtract()

toString()

Converts this value into a decimal string.

Parameters

FormatPrecision The maximum number of digits of precision to output. If there are fewer digits available, zero padding will not be used unless the value is integral and small enough to be expressed in FormatPrecision digits. 0 means to use the natural precision of the number.
FormatMaxPadding The maximum number of zeros to consider inserting before falling back to scientific notation. 0 means to always use scientific notation.
TruncateZero Indicate whether to remove the trailing zero in fraction part or not. Also setting this parameter to false forcing producing of output more similar to default printf behavior. Specifically the lower e is used as exponent delimiter and exponent always contains no less than two digits.

Number Precision MaxPadding Result


1.01E+4 5 2 10100 1.01E+4 4 2 1.01E+4 1.01E+4 5 1 1.01E+4 1.01E-2 5 2 0.0101 1.01E-2 4 2 0.0101 1.01E-2 4 1 1.01E-2

Definition at line 4460 of file APFloat.cpp.

References llvm::detail::fcInfinity, llvm::detail::fcNaN, llvm::detail::fcNormal, llvm::detail::fcZero, isNegative(), llvm::partCountForBits(), and toString().

Referenced by toString().

frexp

hash_value

Overload to compute a hash code for an APFloat value.

Note that the use of hash codes for floating point values is in general frought with peril. Equality is hard to define for these values. For example, should negative and positive zero hash to different codes? Are they equal or not? This hash value implementation specifically emphasizes producing different codes for different inputs in order to be used in canonicalization and memoization. As such, equality is bitwiseIsEqual, and 0 != -0.

Definition at line 3471 of file APFloat.cpp.

References llvm::hash_combine(), llvm::hash_combine_range(), IEEEFloat(), isFiniteNonZero(), isNaN(), and llvm::fltSemantics::precision.

IEEEFloatUnitTestHelper

friend class IEEEFloatUnitTestHelper friend

ilogb

Definition at line 4729 of file APFloat.cpp.

References getSemantics(), IEEEFloat(), llvm::APFloatBase::IEK_Inf, llvm::APFloatBase::IEK_NaN, llvm::APFloatBase::IEK_Zero, isDenormal(), isInfinity(), isNaN(), isZero(), llvm::lfExactlyZero, llvm::fltSemantics::precision, and llvm::APFloatBase::rmNearestTiesToEven.

Referenced by mod().

scalbn


The documentation for this class was generated from the following files: