#include <flang/Evaluate/integer-value.h>
Classes | |
| struct | PowerWithErrors |
| struct | Product |
| struct | QuotientWithRemainder |
| struct | ValueWithCarry |
| struct | ValueWithOverflow |
Public Member Functions | |
| IntegerValue (const IntegerValue &) | |
| IntegerValue (IntegerValue &&) | |
| IntegerValue & | operator= (const IntegerValue &) |
| IntegerValue & | operator= (IntegerValue &&) |
| IntegerValue (int kind, const IntegerValue &x) | |
| IntegerValue (int kind, IntegerValue &&x) | |
| template<typename INT, typename = std::enable_if_t<common::numeric_limits<INT>::is_integer>> | |
| IntegerValue (int kind, INT v) | |
| void | print (llvm::raw_ostream &os) const |
| LLVM_DUMP_METHOD void | dump () const |
| bool | IsNull () const |
| int | kind () const |
| The kind of the value currently stored. | |
| int | bits () const |
| std::size_t | bytesStored () const |
| Number of bytes accessed by FromRawBytes/StoreRawBytes. | |
| bool | operator< (const IntegerValue &y) const |
| bool | operator<= (const IntegerValue &y) const |
| bool | operator== (const IntegerValue &y) const |
| bool | operator!= (const IntegerValue &y) const |
| bool | operator>= (const IntegerValue &y) const |
| bool | operator> (const IntegerValue &y) const |
| std::string | UnsignedDecimal () const |
| std::string | SignedDecimal () const |
| std::string | Hexadecimal () const |
| Omits a leading "0x". | |
| bool | IsZero () const |
| bool | IsNegative () const |
| Ordering | CompareToZeroSigned () const |
| int | LEADZ () const |
| int | POPCNT () const |
| Count the number of bit positions that are set. | |
| bool | POPPAR () const |
| True when POPCNT is odd. | |
| int | TRAILZ () const |
| bool | BTEST (int pos) const |
| Ordering | CompareUnsigned (const IntegerValue &y) const |
| Ordering | CompareSigned (const IntegerValue &y) const |
| bool | BGE (const IntegerValue &y) const |
| bool | BGT (const IntegerValue &y) const |
| bool | BLE (const IntegerValue &y) const |
| bool | BLT (const IntegerValue &y) const |
| std::uint64_t | ToUInt64 () const |
| std::int64_t | ToInt64 () const |
| Fortran::common::uint128_t | ToUInt128 () const |
| Fortran::common::int128_t | ToInt128 () const |
| template<typename INT, typename = std::enable_if_t<std::is_signed_v<INT> || std::is_same_v<INT, Fortran::common::int128_t>>> | |
| INT | ToSInt () const |
| template<typename INT, typename = std::enable_if_t<std::is_unsigned_v<INT> || std::is_same_v<INT, Fortran::common::uint128_t>>> | |
| INT | ToUInt () const |
| IntegerValue | NOT () const |
| Ones'-complement (i.e., C's ~) | |
| ValueWithOverflow | Negate () const |
| ValueWithOverflow | ABS () const |
| IntegerValue | ISHFT (int count) const |
| IntegerValue | SHIFTL (int count) const |
| Left shift with zero fill. | |
| IntegerValue | ISHFTC (int count, int size) const |
| IntegerValue | ISHFTC (int count) const |
| IntegerValue | DSHIFTL (const IntegerValue &fill, int count) const |
| DSHIFTL(I,J) shifts I:J left; the second argument is the right fill. | |
| IntegerValue | DSHIFTR (const IntegerValue &v2, int count) const |
| DSHIFTR(I,J) shifts I:J right; the first argument is the left fill. | |
| IntegerValue | SHIFTR (int count) const |
| Vacated upper bits are filled with zeroes. | |
| IntegerValue | SHIFTA (int count) const |
| IntegerValue | IBCLR (int pos) const |
| Clears a single bit. | |
| IntegerValue | IBSET (int pos) const |
| Sets a single bit. | |
| IntegerValue | IBITS (int pos, int size) const |
| Extracts a field. | |
| IntegerValue | IAND (const IntegerValue &y) const |
| IntegerValue | IOR (const IntegerValue &y) const |
| IntegerValue | IEOR (const IntegerValue &y) const |
| IntegerValue | MERGE_BITS (const IntegerValue &y, const IntegerValue &mask) const |
| IntegerValue | MAX (const IntegerValue &y) const |
| IntegerValue | MIN (const IntegerValue &y) const |
| ValueWithCarry | AddUnsigned (const IntegerValue &y, bool carryIn=false) const |
| ValueWithOverflow | AddSigned (const IntegerValue &y) const |
| ValueWithOverflow | SubtractSigned (const IntegerValue &y) const |
| ValueWithOverflow | DIM (const IntegerValue &y) const |
| DIM(X,Y)=MAX(X-Y, 0) | |
| ValueWithOverflow | SIGN (const IntegerValue &sign) const |
| Product | MultiplyUnsigned (const IntegerValue &y) const |
| Product | MultiplySigned (const IntegerValue &y) const |
| QuotientWithRemainder | DivideUnsigned (const IntegerValue &y) const |
| QuotientWithRemainder | DivideSigned (const IntegerValue &y) const |
| ValueWithOverflow | MODULO (const IntegerValue &y) const |
| PowerWithErrors | Power (const IntegerValue &e) const |
| void | StoreRawBytes (void *dst, size_t size, bool *changed=nullptr) const |
Static Public Member Functions | |
| static IntegerValue | Zero (int kind) |
| static constexpr int | bits (int kind) |
| static constexpr std::size_t | bytesStored (int kind) |
| static IntegerValue | MASKL (int kind, int places) |
| Left-justified mask (e.g., MASKL(1) has only its sign bit set) | |
| static IntegerValue | MASKR (int kind, int places) |
| Right-justified mask (e.g., MASKR(1) == 1, MASKR(2) == 3, &c.) | |
| static ValueWithOverflow | Read (int kind, const char *&pp, int base, bool isSigned) |
| static ValueWithOverflow | ConvertUnsigned (int toKind, const IntegerValue &from) |
| ZExt or Trunc. | |
| static ValueWithOverflow | ConvertSigned (int toKind, const IntegerValue &from) |
| SExt or Trunc. | |
| static constexpr int | DIGITS (int kind) |
| static IntegerValue | HUGE (int kind) |
| static IntegerValue | Least (int kind) |
| static int | RANGE (int kind) |
| static int | UnsignedRANGE (int kind) |
| static IntegerValue | FromRawBytes (int kind, const void *raw, std::size_t expectedSize) |
Friends | |
| class | RealValueImpl |
Detailed Description
A two's-complement integer with dynamic bitwidth.
The bitwidth is dynamic, but only a predefined set of Fortran kinds are allowed. It is also kind-aware, i.e. knows which INTEGER kind it currently represents.
The implementation is hidden from this header using a pImpl-like idiom.
Member Function Documentation
◆ DivideSigned()
| IntegerValue::QuotientWithRemainder Fortran::evaluate::value::IntegerValue::DivideSigned | ( | const IntegerValue & | y | ) | const |
A nonzero remainder has the sign of the dividend, i.e., it computes the MOD intrinsic (X-INT(X/Y)*Y), not MODULO (which is below). 8/5 = 1r3; -8/5 = -1r-3; 8/-5 = -1r3; -8/-5 = 1r-3
◆ ISHFT()
|
inline |
Shifts the operand left when the count is positive, right when negative. Vacated bit positions are filled with zeroes.
◆ ISHFTC()
| IntegerValue Fortran::evaluate::value::IntegerValue::ISHFTC | ( | int | count, |
| int | size ) const |
Circular shift of a field of least-significant bits. The least-order "size" bits are shifted circularly in place by "count" positions; the shift is leftward if count is nonnegative, rightward otherwise. Higher-order bits are unchanged.
◆ IsNull()
| bool Fortran::evaluate::value::IntegerValue::IsNull | ( | ) | const |
Whether this object represents a default-initialized value (zero) of not-yet-known kind.
◆ LEADZ()
| int Fortran::evaluate::value::IntegerValue::LEADZ | ( | ) | const |
Count the number of contiguous most-significant bit positions that are clear.
◆ MODULO()
| IntegerValue::ValueWithOverflow Fortran::evaluate::value::IntegerValue::MODULO | ( | const IntegerValue & | y | ) | const |
Result has the sign of the divisor argument. 8 mod 5 = 3; -8 mod 5 = 2; 8 mod -5 = -2; -8 mod -5 = -3
◆ Negate()
| IntegerValue::ValueWithOverflow Fortran::evaluate::value::IntegerValue::Negate | ( | ) | const |
Two's-complement negation (-x = ~x + 1). An overflow flag accompanies the result, and will be true when the operand is the most negative signed number (MASKL(1)).
◆ SHIFTA()
| IntegerValue Fortran::evaluate::value::IntegerValue::SHIFTA | ( | int | count | ) | const |
Be advised, an arithmetic (sign-filling) right shift is not the same as a division by a power of two in all cases.
◆ Zero()
|
static |
Creates an integer with value 0 of a given kind. This is different from the default-ctor which creates a null state that represents 0 of unknown kind.
The documentation for this class was generated from the following files:
- include/flang/Evaluate/integer-value.h
- lib/Evaluate/integer-value.cpp
Generated on for FLANG by
1.14.0