1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
|
//===----------------------------------------------------------------------===//
//
// 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.
//
//===----------------------------------------------------------------------===//
#ifndef _LIBCUDACXX___CMATH_FPCLASSIFY_H
#define _LIBCUDACXX___CMATH_FPCLASSIFY_H
#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/bit_cast.h>
#include <cuda/std/__cmath/isinf.h>
#include <cuda/std/__cmath/isnan.h>
#include <cuda/std/__concepts/concept_macros.h>
#include <cuda/std/__floating_point/fp.h>
#include <cuda/std/__type_traits/is_integral.h>
#include <cuda/std/limits>
#if !_CCCL_COMPILER(NVRTC)
# include <math.h>
#endif // !_CCCL_COMPILER(NVRTC)
#include <cuda/std/__cccl/prologue.h>
#if _CCCL_COMPILER(NVRTC)
# ifndef FP_NAN
# define FP_NAN 0
# endif // ! FP_NAN
# ifndef FP_INFINITE
# define FP_INFINITE 1
# endif // ! FP_INFINITE
# ifndef FP_ZERO
# define FP_ZERO 2
# endif // ! FP_ZERO
# ifndef FP_SUBNORMAL
# define FP_SUBNORMAL 3
# endif // ! FP_SUBNORMAL
# ifndef FP_NORMAL
# define FP_NORMAL 4
# endif // ! FP_NORMAL
#else // ^^^ _CCCL_COMPILER(NVRTC) ^^^ ^/ vvv !_CCCL_COMPILER(NVRTC) vvv
# include <math.h>
#endif // !_CCCL_COMPILER(NVRTC)
_LIBCUDACXX_BEGIN_NAMESPACE_STD
#if _CCCL_CHECK_BUILTIN(builtin_fpclassify) || _CCCL_COMPILER(GCC)
# define _CCCL_BUILTIN_FPCLASSIFY(...) __builtin_fpclassify(__VA_ARGS__)
#endif // _CCCL_CHECK_BUILTIN(builtin_fpclassify)
// nvcc does not implement __builtin_fpclassify
#if _CCCL_CUDA_COMPILER(NVCC)
# undef _CCCL_BUILTIN_FPCLASSIFY
#endif // _CCCL_CUDA_COMPILER(NVCC)
template <class _Tp>
[[nodiscard]] _CCCL_API constexpr int __fpclassify_impl(_Tp __x) noexcept
{
static_assert(numeric_limits<_Tp>::has_denorm, "The type must have denorm support");
if constexpr (numeric_limits<_Tp>::has_quiet_NaN || numeric_limits<_Tp>::has_signaling_NaN)
{
if (_CUDA_VSTD::isnan(__x))
{
return FP_NAN;
}
}
if constexpr (numeric_limits<_Tp>::has_infinity)
{
if (_CUDA_VSTD::isinf(__x))
{
return FP_INFINITE;
}
}
if constexpr (_CCCL_TRAIT(is_floating_point, _Tp))
{
if (__x > -numeric_limits<_Tp>::min() && __x < numeric_limits<_Tp>::min())
{
return (__x == _Tp{}) ? FP_ZERO : FP_SUBNORMAL;
}
return FP_NORMAL;
}
else
{
const auto __storage = _CUDA_VSTD::__fp_get_storage(__x);
if ((__storage & __fp_exp_mask_of_v<_Tp>) == 0)
{
return (__storage & __fp_mant_mask_of_v<_Tp>) ? FP_SUBNORMAL : FP_ZERO;
}
return FP_NORMAL;
}
}
[[nodiscard]] _CCCL_API constexpr int fpclassify(float __x) noexcept
{
#if defined(_CCCL_BUILTIN_FPCLASSIFY)
return _CCCL_BUILTIN_FPCLASSIFY(FP_NAN, FP_INFINITE, FP_NORMAL, FP_SUBNORMAL, FP_ZERO, __x);
#else // ^^^ _CCCL_BUILTIN_FPCLASSIFY ^^^ / vvv !_CCCL_BUILTIN_FPCLASSIFY vvv
if (!_CUDA_VSTD::__cccl_default_is_constant_evaluated())
{
NV_IF_TARGET(NV_IS_HOST, (return ::fpclassify(__x);))
}
return _CUDA_VSTD::__fpclassify_impl(__x);
#endif // !_CCCL_BUILTIN_FPCLASSIFY
}
[[nodiscard]] _CCCL_API constexpr int fpclassify(double __x) noexcept
{
#if defined(_CCCL_BUILTIN_FPCLASSIFY)
return _CCCL_BUILTIN_FPCLASSIFY(FP_NAN, FP_INFINITE, FP_NORMAL, FP_SUBNORMAL, FP_ZERO, __x);
#else // ^^^ _CCCL_BUILTIN_FPCLASSIFY ^^^ / vvv !_CCCL_BUILTIN_FPCLASSIFY vvv
if (!_CUDA_VSTD::__cccl_default_is_constant_evaluated())
{
NV_IF_TARGET(NV_IS_HOST, (return ::fpclassify(__x);))
}
return _CUDA_VSTD::__fpclassify_impl(__x);
#endif // !_CCCL_BUILTIN_FPCLASSIFY
}
#if _CCCL_HAS_LONG_DOUBLE()
[[nodiscard]] _CCCL_API constexpr int fpclassify(long double __x) noexcept
{
# if defined(_CCCL_BUILTIN_FPCLASSIFY)
return _CCCL_BUILTIN_FPCLASSIFY(FP_NAN, FP_INFINITE, FP_NORMAL, FP_SUBNORMAL, FP_ZERO, __x);
# else // ^^^ _CCCL_BUILTIN_SIGNBIT ^^^ / vvv !_CCCL_BUILTIN_SIGNBIT vvv
if (!_CUDA_VSTD::__cccl_default_is_constant_evaluated())
{
NV_IF_TARGET(NV_IS_HOST, (return ::fpclassify(__x);))
}
return _CUDA_VSTD::__fpclassify_impl(__x);
# endif // !_CCCL_BUILTIN_SIGNBIT
}
#endif // _CCCL_HAS_LONG_DOUBLE()
#if _CCCL_HAS_NVFP16()
[[nodiscard]] _CCCL_API constexpr int fpclassify(__half __x) noexcept
{
return _CUDA_VSTD::__fpclassify_impl(__x);
}
#endif // _CCCL_HAS_NVFP16()
#if _CCCL_HAS_NVBF16()
[[nodiscard]] _CCCL_API constexpr int fpclassify(__nv_bfloat16 __x) noexcept
{
return _CUDA_VSTD::__fpclassify_impl(__x);
}
#endif // _CCCL_HAS_NVBF16()
#if _CCCL_HAS_NVFP8_E4M3()
[[nodiscard]] _CCCL_API constexpr int fpclassify(__nv_fp8_e4m3 __x) noexcept
{
return _CUDA_VSTD::__fpclassify_impl(__x);
}
#endif // _CCCL_HAS_NVFP8_E4M3()
#if _CCCL_HAS_NVFP8_E5M2()
[[nodiscard]] _CCCL_API constexpr int fpclassify(__nv_fp8_e5m2 __x) noexcept
{
return _CUDA_VSTD::__fpclassify_impl(__x);
}
#endif // _CCCL_HAS_NVFP8_E5M2()
#if _CCCL_HAS_NVFP8_E8M0()
[[nodiscard]] _CCCL_API constexpr int fpclassify(__nv_fp8_e8m0 __x) noexcept
{
return ((__x.__x & __fp_exp_mask_of_v<__nv_fp8_e8m0>) == __fp_exp_mask_of_v<__nv_fp8_e8m0>) ? FP_NAN : FP_NORMAL;
}
#endif // _CCCL_HAS_NVFP8_E8M0()
#if _CCCL_HAS_NVFP6_E2M3()
[[nodiscard]] _CCCL_API constexpr int fpclassify(__nv_fp6_e2m3 __x) noexcept
{
return _CUDA_VSTD::__fpclassify_impl(__x);
}
#endif // _CCCL_HAS_NVFP6_E2M3()
#if _CCCL_HAS_NVFP6_E3M2()
[[nodiscard]] _CCCL_API constexpr int fpclassify(__nv_fp6_e3m2 __x) noexcept
{
return _CUDA_VSTD::__fpclassify_impl(__x);
}
#endif // _CCCL_HAS_NVFP6_E3M2()
#if _CCCL_HAS_NVFP4_E2M1()
[[nodiscard]] _CCCL_API constexpr int fpclassify(__nv_fp4_e2m1 __x) noexcept
{
return _CUDA_VSTD::__fpclassify_impl(__x);
}
#endif // _CCCL_HAS_NVFP4_E2M1()
_CCCL_TEMPLATE(class _Tp)
_CCCL_REQUIRES(_CCCL_TRAIT(is_integral, _Tp))
[[nodiscard]] _CCCL_API constexpr int fpclassify(_Tp __x) noexcept
{
return (__x == 0) ? FP_ZERO : FP_NORMAL;
}
_LIBCUDACXX_END_NAMESPACE_STD
#include <cuda/std/__cccl/epilogue.h>
#endif // _LIBCUDACXX___CMATH_FPCLASSIFY_H
|
