2016-09-25 17:23:39 -07:00
|
|
|
//===-- llvm/Support/MathExtras.h - Useful math functions -------*- C++ -*-===//
|
|
|
|
|
//
|
2022-05-20 18:59:53 -04:00
|
|
|
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
|
|
|
|
|
// See https://llvm.org/LICENSE.txt for license information.
|
|
|
|
|
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
|
2016-09-25 17:23:39 -07:00
|
|
|
//
|
|
|
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
|
//
|
|
|
|
|
// This file contains some functions that are useful for math stuff.
|
|
|
|
|
//
|
|
|
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
#ifndef WPIUTIL_WPI_MATHEXTRAS_H
|
|
|
|
|
#define WPIUTIL_WPI_MATHEXTRAS_H
|
2016-09-25 17:23:39 -07:00
|
|
|
|
2023-07-12 22:50:13 -07:00
|
|
|
#include "wpi/bit.h"
|
2018-04-29 23:33:19 -07:00
|
|
|
#include "wpi/Compiler.h"
|
2023-07-12 22:50:13 -07:00
|
|
|
#include <bit>
|
2016-09-25 17:49:01 -07:00
|
|
|
#include <cassert>
|
2018-05-22 23:31:08 -07:00
|
|
|
#include <climits>
|
2022-05-20 18:59:53 -04:00
|
|
|
#include <cstdint>
|
2016-09-25 17:49:01 -07:00
|
|
|
#include <cstring>
|
2016-09-25 17:23:39 -07:00
|
|
|
#include <limits>
|
2018-05-22 23:31:08 -07:00
|
|
|
#include <type_traits>
|
2016-09-25 17:23:39 -07:00
|
|
|
|
2018-04-29 23:33:19 -07:00
|
|
|
namespace wpi {
|
2022-05-20 18:59:53 -04:00
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Create a bitmask with the N right-most bits set to 1, and all other
|
|
|
|
|
/// bits set to 0. Only unsigned types are allowed.
|
|
|
|
|
template <typename T> T maskTrailingOnes(unsigned N) {
|
2023-09-21 19:54:33 -07:00
|
|
|
static_assert(std::is_unsigned_v<T>, "Invalid type!");
|
2018-05-22 23:31:08 -07:00
|
|
|
const unsigned Bits = CHAR_BIT * sizeof(T);
|
|
|
|
|
assert(N <= Bits && "Invalid bit index");
|
|
|
|
|
return N == 0 ? 0 : (T(-1) >> (Bits - N));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Create a bitmask with the N left-most bits set to 1, and all other
|
|
|
|
|
/// bits set to 0. Only unsigned types are allowed.
|
|
|
|
|
template <typename T> T maskLeadingOnes(unsigned N) {
|
|
|
|
|
return ~maskTrailingOnes<T>(CHAR_BIT * sizeof(T) - N);
|
2016-09-25 17:23:39 -07:00
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Create a bitmask with the N right-most bits set to 0, and all other
|
|
|
|
|
/// bits set to 1. Only unsigned types are allowed.
|
|
|
|
|
template <typename T> T maskTrailingZeros(unsigned N) {
|
|
|
|
|
return maskLeadingOnes<T>(CHAR_BIT * sizeof(T) - N);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Create a bitmask with the N left-most bits set to 0, and all other
|
|
|
|
|
/// bits set to 1. Only unsigned types are allowed.
|
|
|
|
|
template <typename T> T maskLeadingZeros(unsigned N) {
|
|
|
|
|
return maskTrailingOnes<T>(CHAR_BIT * sizeof(T) - N);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Macro compressed bit reversal table for 256 bits.
|
2016-09-25 17:49:01 -07:00
|
|
|
///
|
|
|
|
|
/// http://graphics.stanford.edu/~seander/bithacks.html#BitReverseTable
|
|
|
|
|
static const unsigned char BitReverseTable256[256] = {
|
|
|
|
|
#define R2(n) n, n + 2 * 64, n + 1 * 64, n + 3 * 64
|
|
|
|
|
#define R4(n) R2(n), R2(n + 2 * 16), R2(n + 1 * 16), R2(n + 3 * 16)
|
|
|
|
|
#define R6(n) R4(n), R4(n + 2 * 4), R4(n + 1 * 4), R4(n + 3 * 4)
|
|
|
|
|
R6(0), R6(2), R6(1), R6(3)
|
|
|
|
|
#undef R2
|
|
|
|
|
#undef R4
|
|
|
|
|
#undef R6
|
|
|
|
|
};
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Reverse the bits in \p Val.
|
2023-07-12 22:50:13 -07:00
|
|
|
template <typename T> T reverseBits(T Val) {
|
2022-05-20 18:59:53 -04:00
|
|
|
#if __has_builtin(__builtin_bitreverse8)
|
2023-07-12 22:50:13 -07:00
|
|
|
if constexpr (std::is_same_v<T, uint8_t>)
|
|
|
|
|
return __builtin_bitreverse8(Val);
|
2022-05-20 18:59:53 -04:00
|
|
|
#endif
|
|
|
|
|
#if __has_builtin(__builtin_bitreverse16)
|
2023-07-12 22:50:13 -07:00
|
|
|
if constexpr (std::is_same_v<T, uint16_t>)
|
|
|
|
|
return __builtin_bitreverse16(Val);
|
2022-05-20 18:59:53 -04:00
|
|
|
#endif
|
|
|
|
|
#if __has_builtin(__builtin_bitreverse32)
|
2023-07-12 22:50:13 -07:00
|
|
|
if constexpr (std::is_same_v<T, uint32_t>)
|
|
|
|
|
return __builtin_bitreverse32(Val);
|
2022-05-20 18:59:53 -04:00
|
|
|
#endif
|
|
|
|
|
#if __has_builtin(__builtin_bitreverse64)
|
2023-07-12 22:50:13 -07:00
|
|
|
if constexpr (std::is_same_v<T, uint64_t>)
|
|
|
|
|
return __builtin_bitreverse64(Val);
|
2022-05-20 18:59:53 -04:00
|
|
|
#endif
|
|
|
|
|
|
2023-07-12 22:50:13 -07:00
|
|
|
unsigned char in[sizeof(Val)];
|
|
|
|
|
unsigned char out[sizeof(Val)];
|
|
|
|
|
std::memcpy(in, &Val, sizeof(Val));
|
|
|
|
|
for (unsigned i = 0; i < sizeof(Val); ++i)
|
|
|
|
|
out[(sizeof(Val) - i) - 1] = BitReverseTable256[in[i]];
|
|
|
|
|
std::memcpy(&Val, out, sizeof(Val));
|
|
|
|
|
return Val;
|
|
|
|
|
}
|
|
|
|
|
|
2016-09-25 17:49:01 -07:00
|
|
|
// NOTE: The following support functions use the _32/_64 extensions instead of
|
|
|
|
|
// type overloading so that signed and unsigned integers can be used without
|
|
|
|
|
// ambiguity.
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Return the high 32 bits of a 64 bit value.
|
|
|
|
|
constexpr inline uint32_t Hi_32(uint64_t Value) {
|
2016-09-25 17:49:01 -07:00
|
|
|
return static_cast<uint32_t>(Value >> 32);
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Return the low 32 bits of a 64 bit value.
|
|
|
|
|
constexpr inline uint32_t Lo_32(uint64_t Value) {
|
2016-09-25 17:49:01 -07:00
|
|
|
return static_cast<uint32_t>(Value);
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Make a 64-bit integer from a high / low pair of 32-bit integers.
|
|
|
|
|
constexpr inline uint64_t Make_64(uint32_t High, uint32_t Low) {
|
2016-09-25 17:49:01 -07:00
|
|
|
return ((uint64_t)High << 32) | (uint64_t)Low;
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Checks if an integer fits into the given bit width.
|
|
|
|
|
template <unsigned N> constexpr inline bool isInt(int64_t x) {
|
2023-07-12 22:50:13 -07:00
|
|
|
if constexpr (N == 8)
|
|
|
|
|
return static_cast<int8_t>(x) == x;
|
|
|
|
|
if constexpr (N == 16)
|
|
|
|
|
return static_cast<int16_t>(x) == x;
|
|
|
|
|
if constexpr (N == 32)
|
|
|
|
|
return static_cast<int32_t>(x) == x;
|
|
|
|
|
if constexpr (N < 64)
|
|
|
|
|
return -(INT64_C(1) << (N - 1)) <= x && x < (INT64_C(1) << (N - 1));
|
|
|
|
|
(void)x; // MSVC v19.25 warns that x is unused.
|
|
|
|
|
return true;
|
2016-09-25 17:49:01 -07:00
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Checks if a signed integer is an N bit number shifted left by S.
|
|
|
|
|
template <unsigned N, unsigned S>
|
|
|
|
|
constexpr inline bool isShiftedInt(int64_t x) {
|
|
|
|
|
static_assert(
|
|
|
|
|
N > 0, "isShiftedInt<0> doesn't make sense (refers to a 0-bit number.");
|
|
|
|
|
static_assert(N + S <= 64, "isShiftedInt<N, S> with N + S > 64 is too wide.");
|
|
|
|
|
return isInt<N + S>(x) && (x % (UINT64_C(1) << S) == 0);
|
2016-09-25 17:49:01 -07:00
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Checks if an unsigned integer fits into the given bit width.
|
2023-07-12 22:50:13 -07:00
|
|
|
template <unsigned N> constexpr inline bool isUInt(uint64_t x) {
|
2018-05-22 23:31:08 -07:00
|
|
|
static_assert(N > 0, "isUInt<0> doesn't make sense");
|
2023-07-12 22:50:13 -07:00
|
|
|
if constexpr (N == 8)
|
|
|
|
|
return static_cast<uint8_t>(x) == x;
|
|
|
|
|
if constexpr (N == 16)
|
|
|
|
|
return static_cast<uint16_t>(x) == x;
|
|
|
|
|
if constexpr (N == 32)
|
|
|
|
|
return static_cast<uint32_t>(x) == x;
|
|
|
|
|
if constexpr (N < 64)
|
|
|
|
|
return x < (UINT64_C(1) << (N));
|
|
|
|
|
(void)x; // MSVC v19.25 warns that x is unused.
|
2018-05-22 23:31:08 -07:00
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Checks if a unsigned integer is an N bit number shifted left by S.
|
|
|
|
|
template <unsigned N, unsigned S>
|
|
|
|
|
constexpr inline bool isShiftedUInt(uint64_t x) {
|
|
|
|
|
static_assert(
|
|
|
|
|
N > 0, "isShiftedUInt<0> doesn't make sense (refers to a 0-bit number)");
|
|
|
|
|
static_assert(N + S <= 64,
|
|
|
|
|
"isShiftedUInt<N, S> with N + S > 64 is too wide.");
|
|
|
|
|
// Per the two static_asserts above, S must be strictly less than 64. So
|
|
|
|
|
// 1 << S is not undefined behavior.
|
|
|
|
|
return isUInt<N + S>(x) && (x % (UINT64_C(1) << S) == 0);
|
2016-09-25 17:49:01 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Gets the maximum value for a N-bit unsigned integer.
|
|
|
|
|
inline uint64_t maxUIntN(uint64_t N) {
|
|
|
|
|
assert(N > 0 && N <= 64 && "integer width out of range");
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
// uint64_t(1) << 64 is undefined behavior, so we can't do
|
|
|
|
|
// (uint64_t(1) << N) - 1
|
|
|
|
|
// without checking first that N != 64. But this works and doesn't have a
|
|
|
|
|
// branch.
|
|
|
|
|
return UINT64_MAX >> (64 - N);
|
2016-09-25 17:49:01 -07:00
|
|
|
}
|
|
|
|
|
|
2019-05-31 13:43:32 -07:00
|
|
|
#ifdef _WIN32
|
|
|
|
|
#pragma warning(push)
|
|
|
|
|
#pragma warning(disable : 4146)
|
|
|
|
|
#endif
|
|
|
|
|
|
2016-09-25 17:49:01 -07:00
|
|
|
/// Gets the minimum value for a N-bit signed integer.
|
|
|
|
|
inline int64_t minIntN(int64_t N) {
|
|
|
|
|
assert(N > 0 && N <= 64 && "integer width out of range");
|
|
|
|
|
|
2022-05-20 18:59:53 -04:00
|
|
|
return UINT64_C(1) + ~(UINT64_C(1) << (N - 1));
|
2016-09-25 17:49:01 -07:00
|
|
|
}
|
|
|
|
|
|
2019-05-31 13:43:32 -07:00
|
|
|
#ifdef _WIN32
|
|
|
|
|
#pragma warning(pop)
|
|
|
|
|
#endif
|
|
|
|
|
|
2016-09-25 17:49:01 -07:00
|
|
|
/// Gets the maximum value for a N-bit signed integer.
|
|
|
|
|
inline int64_t maxIntN(int64_t N) {
|
|
|
|
|
assert(N > 0 && N <= 64 && "integer width out of range");
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
// This relies on two's complement wraparound when N == 64, so we convert to
|
|
|
|
|
// int64_t only at the very end to avoid UB.
|
|
|
|
|
return (UINT64_C(1) << (N - 1)) - 1;
|
2016-09-25 17:49:01 -07:00
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Checks if an unsigned integer fits into the given (dynamic) bit width.
|
2016-09-25 17:49:01 -07:00
|
|
|
inline bool isUIntN(unsigned N, uint64_t x) {
|
|
|
|
|
return N >= 64 || x <= maxUIntN(N);
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Checks if an signed integer fits into the given (dynamic) bit width.
|
2016-09-25 17:49:01 -07:00
|
|
|
inline bool isIntN(unsigned N, int64_t x) {
|
|
|
|
|
return N >= 64 || (minIntN(N) <= x && x <= maxIntN(N));
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Return true if the argument is a non-empty sequence of ones starting at the
|
|
|
|
|
/// least significant bit with the remainder zero (32 bit version).
|
|
|
|
|
/// Ex. isMask_32(0x0000FFFFU) == true.
|
|
|
|
|
constexpr inline bool isMask_32(uint32_t Value) {
|
2016-09-25 17:49:01 -07:00
|
|
|
return Value && ((Value + 1) & Value) == 0;
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Return true if the argument is a non-empty sequence of ones starting at the
|
|
|
|
|
/// least significant bit with the remainder zero (64 bit version).
|
|
|
|
|
constexpr inline bool isMask_64(uint64_t Value) {
|
2016-09-25 17:49:01 -07:00
|
|
|
return Value && ((Value + 1) & Value) == 0;
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Return true if the argument contains a non-empty sequence of ones with the
|
|
|
|
|
/// remainder zero (32 bit version.) Ex. isShiftedMask_32(0x0000FF00U) == true.
|
|
|
|
|
constexpr inline bool isShiftedMask_32(uint32_t Value) {
|
2016-09-25 17:49:01 -07:00
|
|
|
return Value && isMask_32((Value - 1) | Value);
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Return true if the argument contains a non-empty sequence of ones with the
|
|
|
|
|
/// remainder zero (64 bit version.)
|
|
|
|
|
constexpr inline bool isShiftedMask_64(uint64_t Value) {
|
2016-09-25 17:49:01 -07:00
|
|
|
return Value && isMask_64((Value - 1) | Value);
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Return true if the argument is a power of two > 0.
|
|
|
|
|
/// Ex. isPowerOf2_32(0x00100000U) == true (32 bit edition.)
|
|
|
|
|
constexpr inline bool isPowerOf2_32(uint32_t Value) {
|
2023-07-12 22:50:13 -07:00
|
|
|
return std::has_single_bit(Value);
|
2016-09-25 17:49:01 -07:00
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Return true if the argument is a power of two > 0 (64 bit edition.)
|
|
|
|
|
constexpr inline bool isPowerOf2_64(uint64_t Value) {
|
2023-07-12 22:50:13 -07:00
|
|
|
return std::has_single_bit(Value);
|
2016-09-25 17:49:01 -07:00
|
|
|
}
|
|
|
|
|
|
2023-07-12 22:50:13 -07:00
|
|
|
/// Return true if the argument contains a non-empty sequence of ones with the
|
|
|
|
|
/// remainder zero (32 bit version.) Ex. isShiftedMask_32(0x0000FF00U) == true.
|
|
|
|
|
/// If true, \p MaskIdx will specify the index of the lowest set bit and \p
|
|
|
|
|
/// MaskLen is updated to specify the length of the mask, else neither are
|
|
|
|
|
/// updated.
|
|
|
|
|
inline bool isShiftedMask_32(uint32_t Value, unsigned &MaskIdx,
|
|
|
|
|
unsigned &MaskLen) {
|
|
|
|
|
if (!isShiftedMask_32(Value))
|
|
|
|
|
return false;
|
|
|
|
|
MaskIdx = std::countr_zero(Value);
|
|
|
|
|
MaskLen = std::popcount(Value);
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Return true if the argument contains a non-empty sequence of ones with the
|
|
|
|
|
/// remainder zero (64 bit version.) If true, \p MaskIdx will specify the index
|
|
|
|
|
/// of the lowest set bit and \p MaskLen is updated to specify the length of the
|
|
|
|
|
/// mask, else neither are updated.
|
|
|
|
|
inline bool isShiftedMask_64(uint64_t Value, unsigned &MaskIdx,
|
|
|
|
|
unsigned &MaskLen) {
|
|
|
|
|
if (!isShiftedMask_64(Value))
|
|
|
|
|
return false;
|
|
|
|
|
MaskIdx = std::countr_zero(Value);
|
|
|
|
|
MaskLen = std::popcount(Value);
|
|
|
|
|
return true;
|
2016-09-25 17:49:01 -07:00
|
|
|
}
|
|
|
|
|
|
2022-05-20 18:59:53 -04:00
|
|
|
/// Compile time Log2.
|
|
|
|
|
/// Valid only for positive powers of two.
|
|
|
|
|
template <size_t kValue> constexpr inline size_t CTLog2() {
|
|
|
|
|
static_assert(kValue > 0 && wpi::isPowerOf2_64(kValue),
|
|
|
|
|
"Value is not a valid power of 2");
|
|
|
|
|
return 1 + CTLog2<kValue / 2>();
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
template <> constexpr inline size_t CTLog2<1>() { return 0; }
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Return the floor log base 2 of the specified value, -1 if the value is zero.
|
|
|
|
|
/// (32 bit edition.)
|
2016-09-25 17:23:39 -07:00
|
|
|
/// Ex. Log2_32(32) == 5, Log2_32(1) == 0, Log2_32(0) == -1, Log2_32(6) == 2
|
|
|
|
|
inline unsigned Log2_32(uint32_t Value) {
|
2023-07-12 22:50:13 -07:00
|
|
|
return static_cast<unsigned>(31 - std::countl_zero(Value));
|
2016-09-25 17:23:39 -07:00
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Return the floor log base 2 of the specified value, -1 if the value is zero.
|
|
|
|
|
/// (64 bit edition.)
|
2016-09-25 17:23:39 -07:00
|
|
|
inline unsigned Log2_64(uint64_t Value) {
|
2023-07-12 22:50:13 -07:00
|
|
|
return static_cast<unsigned>(63 - std::countl_zero(Value));
|
2016-09-25 17:23:39 -07:00
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Return the ceil log base 2 of the specified value, 32 if the value is zero.
|
|
|
|
|
/// (32 bit edition).
|
2016-09-25 17:23:39 -07:00
|
|
|
/// Ex. Log2_32_Ceil(32) == 5, Log2_32_Ceil(1) == 0, Log2_32_Ceil(6) == 3
|
|
|
|
|
inline unsigned Log2_32_Ceil(uint32_t Value) {
|
2023-07-12 22:50:13 -07:00
|
|
|
return static_cast<unsigned>(32 - std::countl_zero(Value - 1));
|
2016-09-25 17:23:39 -07:00
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Return the ceil log base 2 of the specified value, 64 if the value is zero.
|
|
|
|
|
/// (64 bit edition.)
|
2016-09-25 17:23:39 -07:00
|
|
|
inline unsigned Log2_64_Ceil(uint64_t Value) {
|
2023-07-12 22:50:13 -07:00
|
|
|
return static_cast<unsigned>(64 - std::countl_zero(Value - 1));
|
2022-05-20 18:59:53 -04:00
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// A and B are either alignments or offsets. Return the minimum alignment that
|
|
|
|
|
/// may be assumed after adding the two together.
|
|
|
|
|
constexpr inline uint64_t MinAlign(uint64_t A, uint64_t B) {
|
2016-09-25 17:49:01 -07:00
|
|
|
// The largest power of 2 that divides both A and B.
|
|
|
|
|
//
|
|
|
|
|
// Replace "-Value" by "1+~Value" in the following commented code to avoid
|
|
|
|
|
// MSVC warning C4146
|
|
|
|
|
// return (A | B) & -(A | B);
|
|
|
|
|
return (A | B) & (1 + ~(A | B));
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Returns the next power of two (in 64-bits) that is strictly greater than A.
|
|
|
|
|
/// Returns zero on overflow.
|
2023-07-12 22:50:13 -07:00
|
|
|
constexpr inline uint64_t NextPowerOf2(uint64_t A) {
|
2016-09-25 17:23:39 -07:00
|
|
|
A |= (A >> 1);
|
|
|
|
|
A |= (A >> 2);
|
|
|
|
|
A |= (A >> 4);
|
|
|
|
|
A |= (A >> 8);
|
|
|
|
|
A |= (A >> 16);
|
|
|
|
|
A |= (A >> 32);
|
|
|
|
|
return A + 1;
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Returns the power of two which is greater than or equal to the given value.
|
|
|
|
|
/// Essentially, it is a ceil operation across the domain of powers of two.
|
|
|
|
|
inline uint64_t PowerOf2Ceil(uint64_t A) {
|
|
|
|
|
if (!A)
|
|
|
|
|
return 0;
|
|
|
|
|
return NextPowerOf2(A - 1);
|
|
|
|
|
}
|
|
|
|
|
|
2016-09-25 17:49:01 -07:00
|
|
|
/// Returns the next integer (mod 2**64) that is greater than or equal to
|
|
|
|
|
/// \p Value and is a multiple of \p Align. \p Align must be non-zero.
|
|
|
|
|
///
|
|
|
|
|
/// Examples:
|
|
|
|
|
/// \code
|
|
|
|
|
/// alignTo(5, 8) = 8
|
|
|
|
|
/// alignTo(17, 8) = 24
|
|
|
|
|
/// alignTo(~0LL, 8) = 0
|
|
|
|
|
/// alignTo(321, 255) = 510
|
2023-07-12 22:50:13 -07:00
|
|
|
/// \endcode
|
|
|
|
|
inline uint64_t alignTo(uint64_t Value, uint64_t Align) {
|
|
|
|
|
assert(Align != 0u && "Align can't be 0.");
|
|
|
|
|
return (Value + Align - 1) / Align * Align;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
inline uint64_t alignToPowerOf2(uint64_t Value, uint64_t Align) {
|
|
|
|
|
assert(Align != 0 && (Align & (Align - 1)) == 0 &&
|
|
|
|
|
"Align must be a power of 2");
|
2023-10-24 00:35:13 -04:00
|
|
|
// Replace unary minus to avoid compilation error on Windows:
|
|
|
|
|
// "unary minus operator applied to unsigned type, result still unsigned"
|
|
|
|
|
uint64_t negAlign = (~Align) + 1;
|
|
|
|
|
return (Value + Align - 1) & negAlign;
|
2023-07-12 22:50:13 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// If non-zero \p Skew is specified, the return value will be a minimal integer
|
|
|
|
|
/// that is greater than or equal to \p Size and equal to \p A * N + \p Skew for
|
|
|
|
|
/// some integer N. If \p Skew is larger than \p A, its value is adjusted to '\p
|
|
|
|
|
/// Skew mod \p A'. \p Align must be non-zero.
|
2016-09-25 17:49:01 -07:00
|
|
|
///
|
2023-07-12 22:50:13 -07:00
|
|
|
/// Examples:
|
|
|
|
|
/// \code
|
2016-09-25 17:49:01 -07:00
|
|
|
/// alignTo(5, 8, 7) = 7
|
|
|
|
|
/// alignTo(17, 8, 1) = 17
|
|
|
|
|
/// alignTo(~0LL, 8, 3) = 3
|
|
|
|
|
/// alignTo(321, 255, 42) = 552
|
|
|
|
|
/// \endcode
|
2023-07-12 22:50:13 -07:00
|
|
|
inline uint64_t alignTo(uint64_t Value, uint64_t Align, uint64_t Skew) {
|
2018-05-22 23:31:08 -07:00
|
|
|
assert(Align != 0u && "Align can't be 0.");
|
2016-09-25 17:49:01 -07:00
|
|
|
Skew %= Align;
|
2023-07-12 22:50:13 -07:00
|
|
|
return alignTo(Value - Skew, Align) + Skew;
|
2016-09-25 17:49:01 -07:00
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Returns the next integer (mod 2**64) that is greater than or equal to
|
|
|
|
|
/// \p Value and is a multiple of \c Align. \c Align must be non-zero.
|
|
|
|
|
template <uint64_t Align> constexpr inline uint64_t alignTo(uint64_t Value) {
|
|
|
|
|
static_assert(Align != 0u, "Align must be non-zero");
|
|
|
|
|
return (Value + Align - 1) / Align * Align;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Returns the integer ceil(Numerator / Denominator).
|
|
|
|
|
inline uint64_t divideCeil(uint64_t Numerator, uint64_t Denominator) {
|
|
|
|
|
return alignTo(Numerator, Denominator) / Denominator;
|
|
|
|
|
}
|
|
|
|
|
|
2022-05-20 18:59:53 -04:00
|
|
|
/// Returns the integer nearest(Numerator / Denominator).
|
|
|
|
|
inline uint64_t divideNearest(uint64_t Numerator, uint64_t Denominator) {
|
|
|
|
|
return (Numerator + (Denominator / 2)) / Denominator;
|
|
|
|
|
}
|
2018-05-22 23:31:08 -07:00
|
|
|
|
2016-09-25 17:49:01 -07:00
|
|
|
/// Returns the largest uint64_t less than or equal to \p Value and is
|
|
|
|
|
/// \p Skew mod \p Align. \p Align must be non-zero
|
|
|
|
|
inline uint64_t alignDown(uint64_t Value, uint64_t Align, uint64_t Skew = 0) {
|
2018-05-22 23:31:08 -07:00
|
|
|
assert(Align != 0u && "Align can't be 0.");
|
2016-09-25 17:49:01 -07:00
|
|
|
Skew %= Align;
|
|
|
|
|
return (Value - Skew) / Align * Align + Skew;
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Sign-extend the number in the bottom B bits of X to a 32-bit integer.
|
|
|
|
|
/// Requires 0 < B <= 32.
|
|
|
|
|
template <unsigned B> constexpr inline int32_t SignExtend32(uint32_t X) {
|
|
|
|
|
static_assert(B > 0, "Bit width can't be 0.");
|
|
|
|
|
static_assert(B <= 32, "Bit width out of range.");
|
|
|
|
|
return int32_t(X << (32 - B)) >> (32 - B);
|
2016-09-25 17:49:01 -07:00
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Sign-extend the number in the bottom B bits of X to a 32-bit integer.
|
2022-05-20 18:59:53 -04:00
|
|
|
/// Requires 0 < B <= 32.
|
2016-09-25 17:49:01 -07:00
|
|
|
inline int32_t SignExtend32(uint32_t X, unsigned B) {
|
2018-05-22 23:31:08 -07:00
|
|
|
assert(B > 0 && "Bit width can't be 0.");
|
|
|
|
|
assert(B <= 32 && "Bit width out of range.");
|
2016-09-25 17:49:01 -07:00
|
|
|
return int32_t(X << (32 - B)) >> (32 - B);
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Sign-extend the number in the bottom B bits of X to a 64-bit integer.
|
2022-05-20 18:59:53 -04:00
|
|
|
/// Requires 0 < B <= 64.
|
2018-05-22 23:31:08 -07:00
|
|
|
template <unsigned B> constexpr inline int64_t SignExtend64(uint64_t x) {
|
|
|
|
|
static_assert(B > 0, "Bit width can't be 0.");
|
|
|
|
|
static_assert(B <= 64, "Bit width out of range.");
|
2016-09-25 17:49:01 -07:00
|
|
|
return int64_t(x << (64 - B)) >> (64 - B);
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Sign-extend the number in the bottom B bits of X to a 64-bit integer.
|
2022-05-20 18:59:53 -04:00
|
|
|
/// Requires 0 < B <= 64.
|
2016-09-25 17:49:01 -07:00
|
|
|
inline int64_t SignExtend64(uint64_t X, unsigned B) {
|
2018-05-22 23:31:08 -07:00
|
|
|
assert(B > 0 && "Bit width can't be 0.");
|
|
|
|
|
assert(B <= 64 && "Bit width out of range.");
|
2016-09-25 17:49:01 -07:00
|
|
|
return int64_t(X << (64 - B)) >> (64 - B);
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Subtract two unsigned integers, X and Y, of type T and return the absolute
|
|
|
|
|
/// value of the result.
|
2016-09-25 17:49:01 -07:00
|
|
|
template <typename T>
|
2023-09-21 19:54:33 -07:00
|
|
|
std::enable_if_t<std::is_unsigned_v<T>, T> AbsoluteDifference(T X, T Y) {
|
2022-05-20 18:59:53 -04:00
|
|
|
return X > Y ? (X - Y) : (Y - X);
|
2016-09-25 17:49:01 -07:00
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Add two unsigned integers, X and Y, of type T. Clamp the result to the
|
|
|
|
|
/// maximum representable value of T on overflow. ResultOverflowed indicates if
|
|
|
|
|
/// the result is larger than the maximum representable value of type T.
|
2016-09-25 17:49:01 -07:00
|
|
|
template <typename T>
|
2023-09-21 19:54:33 -07:00
|
|
|
std::enable_if_t<std::is_unsigned_v<T>, T>
|
2016-09-25 17:49:01 -07:00
|
|
|
SaturatingAdd(T X, T Y, bool *ResultOverflowed = nullptr) {
|
|
|
|
|
bool Dummy;
|
|
|
|
|
bool &Overflowed = ResultOverflowed ? *ResultOverflowed : Dummy;
|
|
|
|
|
// Hacker's Delight, p. 29
|
|
|
|
|
T Z = X + Y;
|
|
|
|
|
Overflowed = (Z < X || Z < Y);
|
|
|
|
|
if (Overflowed)
|
2019-05-31 13:43:32 -07:00
|
|
|
return (std::numeric_limits<T>::max)();
|
2016-09-25 17:49:01 -07:00
|
|
|
else
|
|
|
|
|
return Z;
|
|
|
|
|
}
|
|
|
|
|
|
2023-07-12 22:50:13 -07:00
|
|
|
/// Add multiple unsigned integers of type T. Clamp the result to the
|
|
|
|
|
/// maximum representable value of T on overflow.
|
|
|
|
|
template <class T, class... Ts>
|
|
|
|
|
std::enable_if_t<std::is_unsigned_v<T>, T> SaturatingAdd(T X, T Y, T Z,
|
|
|
|
|
Ts... Args) {
|
|
|
|
|
bool Overflowed = false;
|
|
|
|
|
T XY = SaturatingAdd(X, Y, &Overflowed);
|
|
|
|
|
if (Overflowed)
|
|
|
|
|
return SaturatingAdd((std::numeric_limits<T>::max)(), T(1), Args...);
|
|
|
|
|
return SaturatingAdd(XY, Z, Args...);
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Multiply two unsigned integers, X and Y, of type T. Clamp the result to the
|
|
|
|
|
/// maximum representable value of T on overflow. ResultOverflowed indicates if
|
|
|
|
|
/// the result is larger than the maximum representable value of type T.
|
2016-09-25 17:49:01 -07:00
|
|
|
template <typename T>
|
2023-09-21 19:54:33 -07:00
|
|
|
std::enable_if_t<std::is_unsigned_v<T>, T>
|
2016-09-25 17:49:01 -07:00
|
|
|
SaturatingMultiply(T X, T Y, bool *ResultOverflowed = nullptr) {
|
|
|
|
|
bool Dummy;
|
|
|
|
|
bool &Overflowed = ResultOverflowed ? *ResultOverflowed : Dummy;
|
|
|
|
|
|
|
|
|
|
// Hacker's Delight, p. 30 has a different algorithm, but we don't use that
|
|
|
|
|
// because it fails for uint16_t (where multiplication can have undefined
|
|
|
|
|
// behavior due to promotion to int), and requires a division in addition
|
|
|
|
|
// to the multiplication.
|
|
|
|
|
|
|
|
|
|
Overflowed = false;
|
|
|
|
|
|
|
|
|
|
// Log2(Z) would be either Log2Z or Log2Z + 1.
|
|
|
|
|
// Special case: if X or Y is 0, Log2_64 gives -1, and Log2Z
|
|
|
|
|
// will necessarily be less than Log2Max as desired.
|
|
|
|
|
int Log2Z = Log2_64(X) + Log2_64(Y);
|
2019-05-31 13:43:32 -07:00
|
|
|
const T Max = (std::numeric_limits<T>::max)();
|
2016-09-25 17:49:01 -07:00
|
|
|
int Log2Max = Log2_64(Max);
|
|
|
|
|
if (Log2Z < Log2Max) {
|
|
|
|
|
return X * Y;
|
|
|
|
|
}
|
|
|
|
|
if (Log2Z > Log2Max) {
|
|
|
|
|
Overflowed = true;
|
|
|
|
|
return Max;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// We're going to use the top bit, and maybe overflow one
|
|
|
|
|
// bit past it. Multiply all but the bottom bit then add
|
|
|
|
|
// that on at the end.
|
|
|
|
|
T Z = (X >> 1) * Y;
|
|
|
|
|
if (Z & ~(Max >> 1)) {
|
|
|
|
|
Overflowed = true;
|
|
|
|
|
return Max;
|
|
|
|
|
}
|
|
|
|
|
Z <<= 1;
|
|
|
|
|
if (X & 1)
|
|
|
|
|
return SaturatingAdd(Z, Y, ResultOverflowed);
|
|
|
|
|
|
|
|
|
|
return Z;
|
|
|
|
|
}
|
|
|
|
|
|
2018-05-22 23:31:08 -07:00
|
|
|
/// Multiply two unsigned integers, X and Y, and add the unsigned integer, A to
|
|
|
|
|
/// the product. Clamp the result to the maximum representable value of T on
|
|
|
|
|
/// overflow. ResultOverflowed indicates if the result is larger than the
|
|
|
|
|
/// maximum representable value of type T.
|
2016-09-25 17:49:01 -07:00
|
|
|
template <typename T>
|
2023-09-21 19:54:33 -07:00
|
|
|
std::enable_if_t<std::is_unsigned_v<T>, T>
|
2016-09-25 17:49:01 -07:00
|
|
|
SaturatingMultiplyAdd(T X, T Y, T A, bool *ResultOverflowed = nullptr) {
|
|
|
|
|
bool Dummy;
|
|
|
|
|
bool &Overflowed = ResultOverflowed ? *ResultOverflowed : Dummy;
|
|
|
|
|
|
|
|
|
|
T Product = SaturatingMultiply(X, Y, &Overflowed);
|
|
|
|
|
if (Overflowed)
|
|
|
|
|
return Product;
|
|
|
|
|
|
|
|
|
|
return SaturatingAdd(A, Product, &Overflowed);
|
|
|
|
|
}
|
|
|
|
|
|
2022-05-20 18:59:53 -04:00
|
|
|
/// Use this rather than HUGE_VALF; the latter causes warnings on MSVC.
|
|
|
|
|
extern const float huge_valf;
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/// Add two signed integers, computing the two's complement truncated result,
|
2023-07-12 22:50:13 -07:00
|
|
|
/// returning true if overflow occurred.
|
2022-05-20 18:59:53 -04:00
|
|
|
template <typename T>
|
2023-09-21 19:54:33 -07:00
|
|
|
std::enable_if_t<std::is_signed_v<T>, T> AddOverflow(T X, T Y, T &Result) {
|
2022-05-20 18:59:53 -04:00
|
|
|
#if __has_builtin(__builtin_add_overflow)
|
|
|
|
|
return __builtin_add_overflow(X, Y, &Result);
|
|
|
|
|
#else
|
|
|
|
|
// Perform the unsigned addition.
|
|
|
|
|
using U = std::make_unsigned_t<T>;
|
|
|
|
|
const U UX = static_cast<U>(X);
|
|
|
|
|
const U UY = static_cast<U>(Y);
|
|
|
|
|
const U UResult = UX + UY;
|
|
|
|
|
|
|
|
|
|
// Convert to signed.
|
|
|
|
|
Result = static_cast<T>(UResult);
|
|
|
|
|
|
|
|
|
|
// Adding two positive numbers should result in a positive number.
|
|
|
|
|
if (X > 0 && Y > 0)
|
|
|
|
|
return Result <= 0;
|
|
|
|
|
// Adding two negatives should result in a negative number.
|
|
|
|
|
if (X < 0 && Y < 0)
|
|
|
|
|
return Result >= 0;
|
|
|
|
|
return false;
|
|
|
|
|
#endif
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Subtract two signed integers, computing the two's complement truncated
|
|
|
|
|
/// result, returning true if an overflow ocurred.
|
|
|
|
|
template <typename T>
|
2023-09-21 19:54:33 -07:00
|
|
|
std::enable_if_t<std::is_signed_v<T>, T> SubOverflow(T X, T Y, T &Result) {
|
2022-05-20 18:59:53 -04:00
|
|
|
#if __has_builtin(__builtin_sub_overflow)
|
|
|
|
|
return __builtin_sub_overflow(X, Y, &Result);
|
|
|
|
|
#else
|
|
|
|
|
// Perform the unsigned addition.
|
|
|
|
|
using U = std::make_unsigned_t<T>;
|
|
|
|
|
const U UX = static_cast<U>(X);
|
|
|
|
|
const U UY = static_cast<U>(Y);
|
|
|
|
|
const U UResult = UX - UY;
|
|
|
|
|
|
|
|
|
|
// Convert to signed.
|
|
|
|
|
Result = static_cast<T>(UResult);
|
|
|
|
|
|
|
|
|
|
// Subtracting a positive number from a negative results in a negative number.
|
|
|
|
|
if (X <= 0 && Y > 0)
|
|
|
|
|
return Result >= 0;
|
|
|
|
|
// Subtracting a negative number from a positive results in a positive number.
|
|
|
|
|
if (X >= 0 && Y < 0)
|
|
|
|
|
return Result <= 0;
|
|
|
|
|
return false;
|
|
|
|
|
#endif
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Multiply two signed integers, computing the two's complement truncated
|
|
|
|
|
/// result, returning true if an overflow ocurred.
|
|
|
|
|
template <typename T>
|
2023-09-21 19:54:33 -07:00
|
|
|
std::enable_if_t<std::is_signed_v<T>, T> MulOverflow(T X, T Y, T &Result) {
|
2022-05-20 18:59:53 -04:00
|
|
|
// Perform the unsigned multiplication on absolute values.
|
|
|
|
|
using U = std::make_unsigned_t<T>;
|
|
|
|
|
const U UX = X < 0 ? (0 - static_cast<U>(X)) : static_cast<U>(X);
|
|
|
|
|
const U UY = Y < 0 ? (0 - static_cast<U>(Y)) : static_cast<U>(Y);
|
|
|
|
|
const U UResult = UX * UY;
|
|
|
|
|
|
|
|
|
|
// Convert to signed.
|
|
|
|
|
const bool IsNegative = (X < 0) ^ (Y < 0);
|
|
|
|
|
Result = IsNegative ? (0 - UResult) : UResult;
|
|
|
|
|
|
|
|
|
|
// If any of the args was 0, result is 0 and no overflow occurs.
|
|
|
|
|
if (UX == 0 || UY == 0)
|
|
|
|
|
return false;
|
|
|
|
|
|
|
|
|
|
// UX and UY are in [1, 2^n], where n is the number of digits.
|
|
|
|
|
// Check how the max allowed absolute value (2^n for negative, 2^(n-1) for
|
|
|
|
|
// positive) divided by an argument compares to the other.
|
|
|
|
|
if (IsNegative)
|
|
|
|
|
return UX > (static_cast<U>((std::numeric_limits<T>::max)()) + U(1)) / UY;
|
|
|
|
|
else
|
|
|
|
|
return UX > (static_cast<U>((std::numeric_limits<T>::max)())) / UY;
|
|
|
|
|
}
|
|
|
|
|
|
2019-11-19 06:47:59 -08:00
|
|
|
// Typesafe implementation of the signum function.
|
|
|
|
|
// Returns -1 if negative, 1 if positive, 0 if 0.
|
2019-11-09 23:16:42 -05:00
|
|
|
template <typename T>
|
2019-11-19 06:47:59 -08:00
|
|
|
constexpr int sgn(T val) {
|
2019-11-09 23:16:42 -05:00
|
|
|
return (T(0) < val) - (val < T(0));
|
|
|
|
|
}
|
|
|
|
|
|
2021-12-30 20:08:05 -07:00
|
|
|
/**
|
|
|
|
|
* Linearly interpolates between two values.
|
|
|
|
|
*
|
|
|
|
|
* @param startValue The start value.
|
|
|
|
|
* @param endValue The end value.
|
|
|
|
|
* @param t The fraction for interpolation.
|
|
|
|
|
*
|
|
|
|
|
* @return The interpolated value.
|
|
|
|
|
*/
|
|
|
|
|
template <typename T>
|
|
|
|
|
constexpr T Lerp(const T& startValue, const T& endValue, double t) {
|
|
|
|
|
return startValue + (endValue - startValue) * t;
|
|
|
|
|
}
|
2022-05-20 18:59:53 -04:00
|
|
|
} // End wpi namespace
|
2016-09-25 17:23:39 -07:00
|
|
|
|
|
|
|
|
#endif
|