STL/stl/inc/bit

// bit standard header (core)

// Copyright (c) Microsoft Corporation.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception

#pragma once
#ifndef _BIT_
#define _BIT_
#include <yvals_core.h>
#if _STL_COMPILER_PREPROCESSOR
#if !_HAS_CXX20
_EMIT_STL_WARNING(STL4038, "The contents of <bit> are available only with C++20 or later.");
#else // ^^^ !_HAS_CXX20 / _HAS_CXX20 vvv

#include <isa_availability.h>
#include <limits>
#include <type_traits>

#include _STL_INTRIN_HEADER

#pragma pack(push, _CRT_PACKING)
#pragma warning(push, _STL_WARNING_LEVEL)
#pragma warning(disable : _STL_DISABLED_WARNINGS)
_STL_DISABLE_CLANG_WARNINGS
#pragma push_macro("new")
#undef new

_STD_BEGIN

_EXPORT_STD template <class _To, class _From,
    enable_if_t<conjunction_v<bool_constant<sizeof(_To) == sizeof(_From)>, is_trivially_copyable<_To>,
                    is_trivially_copyable<_From>>,
        int> = 0>
_NODISCARD constexpr _To bit_cast(const _From& _Val) noexcept {
    return __builtin_bit_cast(_To, _Val);
}

#if _HAS_CXX23
_NODISCARD constexpr unsigned short _Byteswap_ushort(const unsigned short _Val) noexcept {
    if (_STD is_constant_evaluated()) {
        return static_cast<unsigned short>((_Val << 8) | (_Val >> 8));
    } else {
        return _byteswap_ushort(_Val);
    }
}

_NODISCARD constexpr unsigned long _Byteswap_ulong(const unsigned long _Val) noexcept {
    if (_STD is_constant_evaluated()) {
        return (_Val << 24) | ((_Val << 8) & 0x00FF'0000) | ((_Val >> 8) & 0x0000'FF00) | (_Val >> 24);
    } else {
        return _byteswap_ulong(_Val);
    }
}

_NODISCARD constexpr unsigned long long _Byteswap_uint64(const unsigned long long _Val) noexcept {
    if (_STD is_constant_evaluated()) {
        return (_Val << 56) | ((_Val << 40) & 0x00FF'0000'0000'0000) | ((_Val << 24) & 0x0000'FF00'0000'0000)
             | ((_Val << 8) & 0x0000'00FF'0000'0000) | ((_Val >> 8) & 0x0000'0000'FF00'0000)
             | ((_Val >> 24) & 0x0000'0000'00FF'0000) | ((_Val >> 40) & 0x0000'0000'0000'FF00) | (_Val >> 56);
    } else {
        return _byteswap_uint64(_Val);
    }
}

_EXPORT_STD template <class _Ty, enable_if_t<is_integral_v<_Ty>, int> = 0>
_NODISCARD constexpr _Ty byteswap(const _Ty _Val) noexcept {
    if constexpr (sizeof(_Ty) == 1) {
        return _Val;
    } else if constexpr (sizeof(_Ty) == 2) {
        return static_cast<_Ty>(_Byteswap_ushort(static_cast<unsigned short>(_Val)));
    } else if constexpr (sizeof(_Ty) == 4) {
        return static_cast<_Ty>(_Byteswap_ulong(static_cast<unsigned long>(_Val)));
    } else if constexpr (sizeof(_Ty) == 8) {
        return static_cast<_Ty>(_Byteswap_uint64(static_cast<unsigned long long>(_Val)));
    } else {
        static_assert(_Always_false<_Ty>, "Unexpected integer size");
    }
}
#endif // _HAS_CXX23

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> = 0>
_NODISCARD constexpr int countl_zero(_Ty _Val) noexcept;

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> = 0>
_NODISCARD constexpr bool has_single_bit(const _Ty _Val) noexcept {
    return _Val != 0 && (_Val & (_Val - 1)) == 0;
}

inline void _Precondition_violation_in_bit_ceil() noexcept {}

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> = 0>
_NODISCARD constexpr _Ty bit_ceil(const _Ty _Val) noexcept /* strengthened */ {
    if (_Val <= 1u) {
        return _Ty{1};
    }

    const int _Num = numeric_limits<_Ty>::digits - _STD countl_zero(static_cast<_Ty>(_Val - 1));

    if constexpr (sizeof(_Ty) < sizeof(unsigned int)) { // for types subject to integral promotion
        if (_STD is_constant_evaluated()) {
            // Because N4892 [expr.shift]/1 "integral promotions are performed"
            // the compiler will not generate a compile time error for
            // uint8_t{1} << 8
            // or
            // uint16_t{1} << 16
            // so we must manually enforce N4892 [bit.pow.two]/5, 8:
            // "Preconditions: N is representable as a value of type T."
            // "Remarks: A function call expression that violates the precondition in the Preconditions: element
            // is not a core constant expression (7.7)."
            if (_Num == numeric_limits<_Ty>::digits) {
                _Precondition_violation_in_bit_ceil();
            }
        }
    }

    return static_cast<_Ty>(_Ty{1} << _Num);
}

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> = 0>
_NODISCARD constexpr _Ty bit_floor(const _Ty _Val) noexcept {
    if (_Val == 0) {
        return 0;
    }

    return static_cast<_Ty>(_Ty{1} << (numeric_limits<_Ty>::digits - 1 - _STD countl_zero(_Val)));
}

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> = 0>
_NODISCARD constexpr int bit_width(const _Ty _Val) noexcept {
    return numeric_limits<_Ty>::digits - _STD countl_zero(_Val);
}

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> = 0>
_NODISCARD constexpr _Ty rotr(_Ty _Val, int _Rotation) noexcept;

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> = 0>
_NODISCARD constexpr _Ty rotl(const _Ty _Val, const int _Rotation) noexcept {
    constexpr auto _Digits = numeric_limits<_Ty>::digits;

    if (!_STD is_constant_evaluated()) {
        if constexpr (_Digits == 64) {
            return _rotl64(_Val, _Rotation);
        } else if constexpr (_Digits == 32) {
            return _rotl(_Val, _Rotation);
        } else if constexpr (_Digits == 16) {
            return _rotl16(_Val, static_cast<unsigned char>(_Rotation));
        } else {
            _STL_INTERNAL_STATIC_ASSERT(_Digits == 8);
            return _rotl8(_Val, static_cast<unsigned char>(_Rotation));
        }
    }

    const auto _Remainder = _Rotation % _Digits;
    if (_Remainder > 0) {
        return static_cast<_Ty>(
            static_cast<_Ty>(_Val << _Remainder) | static_cast<_Ty>(_Val >> (_Digits - _Remainder)));
    } else if (_Remainder == 0) {
        return _Val;
    } else { // _Remainder < 0
        return _STD rotr(_Val, -_Remainder);
    }
}

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> /* = 0 */>
_NODISCARD constexpr _Ty rotr(const _Ty _Val, const int _Rotation) noexcept {
    constexpr auto _Digits = numeric_limits<_Ty>::digits;

    if (!_STD is_constant_evaluated()) {
        if constexpr (_Digits == 64) {
            return _rotr64(_Val, _Rotation);
        } else if constexpr (_Digits == 32) {
            return _rotr(_Val, _Rotation);
        } else if constexpr (_Digits == 16) {
            return _rotr16(_Val, static_cast<unsigned char>(_Rotation));
        } else {
            _STL_INTERNAL_STATIC_ASSERT(_Digits == 8);
            return _rotr8(_Val, static_cast<unsigned char>(_Rotation));
        }
    }

    const auto _Remainder = _Rotation % _Digits;
    if (_Remainder > 0) {
        return static_cast<_Ty>(
            static_cast<_Ty>(_Val >> _Remainder) | static_cast<_Ty>(_Val << (_Digits - _Remainder)));
    } else if (_Remainder == 0) {
        return _Val;
    } else { // _Remainder < 0
        return _STD rotl(_Val, -_Remainder);
    }
}

#if defined(_M_IX86) || (defined(_M_X64) && !defined(_M_ARM64EC))

extern "C" {
extern int __isa_available;
}

template <class _Ty>
_NODISCARD int _Countl_zero_lzcnt(const _Ty _Val) noexcept {
    constexpr int _Digits = numeric_limits<_Ty>::digits;

    if constexpr (_Digits <= 16) {
        return static_cast<int>(__lzcnt16(_Val) - (16 - _Digits));
    } else if constexpr (_Digits == 32) {
        return static_cast<int>(__lzcnt(_Val));
    } else {
#ifdef _M_IX86
        const unsigned int _High = _Val >> 32;
        const auto _Low          = static_cast<unsigned int>(_Val);
        if (_High == 0) {
            return 32 + _Countl_zero_lzcnt(_Low);
        } else {
            return _Countl_zero_lzcnt(_High);
        }
#else // ^^^ _M_IX86 / !_M_IX86 vvv
        return static_cast<int>(__lzcnt64(_Val));
#endif // _M_IX86
    }
}

template <class _Ty>
_NODISCARD int _Countl_zero_bsr(const _Ty _Val) noexcept {
    constexpr int _Digits = numeric_limits<_Ty>::digits;

    unsigned long _Result;
    if constexpr (_Digits <= 32) {
        if (!_BitScanReverse(&_Result, _Val)) {
            return _Digits;
        }
    } else {
#ifdef _M_IX86
        const unsigned int _High = _Val >> 32;
        if (_BitScanReverse(&_Result, _High)) {
            return static_cast<int>(31 - _Result);
        }

        const auto _Low = static_cast<unsigned int>(_Val);
        if (!_BitScanReverse(&_Result, _Low)) {
            return _Digits;
        }
#else // ^^^ _M_IX86 / !_M_IX86 vvv
        if (!_BitScanReverse64(&_Result, _Val)) {
            return _Digits;
        }
#endif // _M_IX86
    }
    return static_cast<int>(_Digits - 1 - _Result);
}

template <class _Ty>
_NODISCARD int _Checked_x86_x64_countl_zero(const _Ty _Val) noexcept {
#ifdef __AVX2__
    return _Countl_zero_lzcnt(_Val);
#else // __AVX2__
    const bool _Definitely_have_lzcnt = __isa_available >= __ISA_AVAILABLE_AVX2;
    if (_Definitely_have_lzcnt) {
        return _Countl_zero_lzcnt(_Val);
    } else {
        return _Countl_zero_bsr(_Val);
    }
#endif // __AVX2__
}
#endif // defined(_M_IX86) || (defined(_M_X64) && !defined(_M_ARM64EC))

#if defined(_M_ARM) || defined(_M_ARM64)
#ifdef __clang__ // TRANSITION, GH-1586
_NODISCARD constexpr int _Clang_arm_arm64_countl_zero(const unsigned short _Val) {
    return __builtin_clzs(_Val);
}

_NODISCARD constexpr int _Clang_arm_arm64_countl_zero(const unsigned int _Val) {
    return __builtin_clz(_Val);
}

_NODISCARD constexpr int _Clang_arm_arm64_countl_zero(const unsigned long _Val) {
    return __builtin_clzl(_Val);
}

_NODISCARD constexpr int _Clang_arm_arm64_countl_zero(const unsigned long long _Val) {
    return __builtin_clzll(_Val);
}
#endif // TRANSITION, GH-1586

template <class _Ty>
_NODISCARD int _Checked_arm_arm64_countl_zero(const _Ty _Val) noexcept {
    constexpr int _Digits = numeric_limits<_Ty>::digits;
    if (_Val == 0) {
        return _Digits;
    }

#ifdef __clang__ // TRANSITION, GH-1586
    if constexpr (is_same_v<remove_cv_t<_Ty>, unsigned char>) {
        return _Clang_arm_arm64_countl_zero(static_cast<unsigned short>(_Val))
             - (numeric_limits<unsigned short>::digits - _Digits);
    } else {
        return _Clang_arm_arm64_countl_zero(_Val);
    }
#else // ^^^ workaround / no workaround vvv
    if constexpr (_Digits <= 32) {
        return static_cast<int>(_CountLeadingZeros(_Val)) - (numeric_limits<unsigned long>::digits - _Digits);
    } else {
        return static_cast<int>(_CountLeadingZeros64(_Val));
    }
#endif // TRANSITION, GH-1586
}
#endif // defined(_M_ARM) || defined(_M_ARM64)

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> /* = 0 */>
_NODISCARD constexpr int countl_zero(const _Ty _Val) noexcept {
#if defined(_M_IX86) || (defined(_M_X64) && !defined(_M_ARM64EC))
    if (!_STD is_constant_evaluated()) {
        return _Checked_x86_x64_countl_zero(_Val);
    }
#elif defined(_M_ARM) || defined(_M_ARM64)
    if (!_STD is_constant_evaluated()) {
        return _Checked_arm_arm64_countl_zero(_Val);
    }
#endif // defined(_M_ARM) || defined(_M_ARM64)

    return _Countl_zero_fallback(_Val);
}

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> = 0>
_NODISCARD constexpr int countl_one(const _Ty _Val) noexcept {
    return _STD countl_zero(static_cast<_Ty>(~_Val));
}

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> = 0>
_NODISCARD constexpr int countr_zero(const _Ty _Val) noexcept {
    return _Countr_zero(_Val);
}

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> = 0>
_NODISCARD constexpr int countr_one(const _Ty _Val) noexcept {
    return _Countr_zero(static_cast<_Ty>(~_Val));
}

_EXPORT_STD template <class _Ty, enable_if_t<_Is_standard_unsigned_integer<_Ty>, int> = 0>
_NODISCARD constexpr int popcount(const _Ty _Val) noexcept {
    return _Popcount(_Val);
}

_EXPORT_STD enum class endian { little = 0, big = 1, native = little };

_STD_END
#pragma pop_macro("new")
_STL_RESTORE_CLANG_WARNINGS
#pragma warning(pop)
#pragma pack(pop)
#endif // _HAS_CXX20
#endif // _STL_COMPILER_PREPROCESSOR
#endif // _BIT_