//===----------------------------------------------------------------------===//
//
// Part of libcu++, the C++ Standard Library for your entire system,
// 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
// SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES.
//
//===----------------------------------------------------------------------===//

// MIT License
//
// Modifications Copyright (C) 2026 Advanced Micro Devices, Inc. All rights reserved.
//
// Permission is hereby granted, free of charge, to any person obtaining a copy
// of this software and associated documentation files (the "Software"), to deal
// in the Software without restriction, including without limitation the rights
// to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
// copies of the Software, and to permit persons to whom the Software is
// furnished to do so, subject to the following conditions:
//
// The above copyright notice and this permission notice shall be included in all
// copies or substantial portions of the Software.
//
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
// SOFTWARE.

#ifndef _LIBCUDACXX___CMATH_ILOG
#define _LIBCUDACXX___CMATH_ILOG

#include <cuda/std/detail/__config>

#if defined(_CCCL_IMPLICIT_SYSTEM_HEADER_GCC)
#  pragma GCC system_header
#elif defined(_CCCL_IMPLICIT_SYSTEM_HEADER_CLANG)
#  pragma clang system_header
#elif defined(_CCCL_IMPLICIT_SYSTEM_HEADER_MSVC)
#  pragma system_header
#endif // no system header

#include <cuda/std/__bit/integral.h>
#include <cuda/std/__cmath/rounding_functions.h>
#include <cuda/std/__concepts/concept_macros.h>
#include <cuda/std/__limits/numeric_limits.h>
#include <cuda/std/__type_traits/is_constant_evaluated.h>
#include <cuda/std/__type_traits/is_integer.h>
#include <cuda/std/__type_traits/is_same.h>
#include <cuda/std/__type_traits/make_unsigned.h>
#include <cuda/std/array>
#include <cuda/std/cstdint>

_LIBCUDACXX_BEGIN_NAMESPACE_CUDA

_CCCL_TEMPLATE(typename _Tp)
_CCCL_REQUIRES(_CCCL_TRAIT(_CUDA_VSTD::__cccl_is_integer, _Tp))
_LIBCUDACXX_HIDE_FROM_ABI constexpr int ilog2(_Tp __t) noexcept
{
  using _Up = _CUDA_VSTD::make_unsigned_t<_Tp>;
  _CCCL_ASSERT(__t > 0, "ilog2() argument must be strictly positive");
  auto __log10_approx = _CUDA_VSTD::__bit_log2(static_cast<_Up>(__t));
  _CCCL_ASSUME(__log10_approx <= _CUDA_VSTD::numeric_limits<_Tp>::digits);
  return __log10_approx;
}

_CCCL_NODISCARD _LIBCUDACXX_HIDE_FROM_ABI constexpr _CUDA_VSTD::array<uint32_t, 10> __power_of_10_32bit() noexcept
{
  return {10,
          100,
          1'000,
          10'000,
          100'000,
          1'000'000,
          10'000'000,
          100'000'000,
          1'000'000'000,
          _CUDA_VSTD::numeric_limits<uint32_t>::max()};
}

_CCCL_NODISCARD _LIBCUDACXX_HIDE_FROM_ABI constexpr _CUDA_VSTD::array<uint64_t, 20> __power_of_10_64bit() noexcept
{
  return {
    10,
    100,
    1'000,
    10'000,
    100'000,
    1'000'000,
    10'000'000,
    100'000'000,
    1'000'000'000,
    10'000'000'000,
    100'000'000'000,
    1'000'000'000'000,
    10'000'000'000'000,
    100'000'000'000'000,
    1'000'000'000'000'000,
    10'000'000'000'000'000,
    100'000'000'000'000'000,
    1'000'000'000'000'000'000,
    10'000'000'000'000'000'000ull,
    _CUDA_VSTD::numeric_limits<uint64_t>::max()};
}

#if _CCCL_HAS_INT128()

_CCCL_NODISCARD _LIBCUDACXX_HIDE_FROM_ABI constexpr _CUDA_VSTD::array<__uint128_t, 39> __power_of_10_128bit() noexcept
{
  return {
    10,
    100,
    1'000,
    10'000,
    100'000,
    1'000'000,
    10'000'000,
    100'000'000,
    1'000'000'000,
    10'000'000'000,
    100'000'000'000,
    1'000'000'000'000,
    10'000'000'000'000,
    100'000'000'000'000,
    1'000'000'000'000'000,
    10'000'000'000'000'000,
    100'000'000'000'000'000,
    1'000'000'000'000'000'000,
    10'000'000'000'000'000'000ull,
    __uint128_t{10'000'000'000'000'000'000ull} * 10,
    __uint128_t{10'000'000'000'000'000'000ull} * 100,
    __uint128_t{10'000'000'000'000'000'000ull} * 1'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 10'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 100'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 1'000'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 10'000'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 100'000'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 1'000'000'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 10'000'000'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 100'000'000'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 1'000'000'000'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 10'000'000'000'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 100'000'000'000'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 1'000'000'000'000'000,
    __uint128_t{10'000'000'000'000'000'000ull} * 1'000'000'000'000'0000,
    __uint128_t{10'000'000'000'000'000'000ull} * 10'000'000'000'000'0000,
    __uint128_t{10'000'000'000'000'000'000ull} * 100'000'000'000'000'0000,
    __uint128_t{10'000'000'000'000'000'000ull} * 1'000'000'000'000'000'0000ull,
    _CUDA_VSTD::numeric_limits<__uint128_t>::max()};
}
#endif // _CCCL_HAS_INT128()

_CCCL_TEMPLATE(typename _Tp)
_CCCL_REQUIRES(_CCCL_TRAIT(_CUDA_VSTD::__cccl_is_integer, _Tp))
_LIBCUDACXX_HIDE_FROM_ABI constexpr int ilog10(_Tp __t) noexcept
{
  _CCCL_ASSERT(__t > 0, "ilog10() argument must be strictly positive");
  constexpr auto __reciprocal_log2_10 = 0.301029995663f; // 1 / log2(10)
  auto __log2                         = ::cuda::ilog2(__t) * __reciprocal_log2_10;
  auto __log10_approx                 = static_cast<int>(__log2);
  if constexpr (sizeof(_Tp) <= sizeof(uint32_t))
  {
    _CCCL_ASSERT(__log10_approx < static_cast<int>(::cuda::__power_of_10_32bit().size()), "out of bounds");
    // NOTE(HIP/AMD): on Windows unsigned long is 32 bits, but on other platforms it's 64 bits,
    // so we need to check for both types here
    if constexpr (_CUDA_VSTD::is_same_v<_Tp, uint32_t> or _CUDA_VSTD::is_same_v<_Tp, unsigned long>)
    {
      // don't replace +1 with >= because wraparound behavior is needed here
      __log10_approx += static_cast<uint32_t>(__t) + 1 > __power_of_10_32bit()[__log10_approx];
    }
    else
    {
      __log10_approx += static_cast<uint32_t>(__t) >= __power_of_10_32bit()[__log10_approx];
    }
  }
  else if constexpr (sizeof(_Tp) == sizeof(uint64_t))
  {
    _CCCL_ASSERT(__log10_approx < static_cast<int>(::cuda::__power_of_10_64bit().size()), "out of bounds");
    // +1 is not needed here
    __log10_approx += static_cast<uint64_t>(__t) >= __power_of_10_64bit()[__log10_approx];
  }
#if _CCCL_HAS_INT128()
  else
  {
    _CCCL_ASSERT(__log10_approx < static_cast<int>(::cuda::__power_of_10_128bit().size()), "out of bounds");
    if constexpr (_CUDA_VSTD::is_same_v<_Tp, __uint128_t>)
    {
      // don't replace +1 with >= because wraparound behavior is needed here
      __log10_approx += static_cast<__uint128_t>(__t) + 1 > __power_of_10_128bit()[__log10_approx];
    }
    else
    {
      __log10_approx += static_cast<__uint128_t>(__t) >= __power_of_10_128bit()[__log10_approx];
    }
  }
#endif // _CCCL_HAS_INT128()
  _CCCL_ASSUME(__log10_approx <= _CUDA_VSTD::numeric_limits<_Tp>::digits / 3); // 2^X < 10^(x/3) -> 8^X < 10^x
  return __log10_approx;
}

_LIBCUDACXX_END_NAMESPACE_CUDA

#endif // _LIBCUDACXX___CMATH_ILOG