File size: 5,870 Bytes
8b7c501
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
/*
 * This implementation is extracted from Eigen:
 *   Repo: bitbucket.org/eigen/eigen
 *   File: Eigen/src/Core/arch/CUDA/Half.h
 *   Commit ID: 96e0f73a35de54f675d825bef5339b2f08e77eb4
 *
 * Removed a lot of redundant and cuda-specific code.
 */

#define EIGEN_STRONG_INLINE static inline
#define EIGEN_DEVICE_FUNC

// This file is part of Eigen, a lightweight C++ template library
// for linear algebra.
//
// This Source Code Form is subject to the terms of the Mozilla
// Public License v. 2.0. If a copy of the MPL was not distributed
// with this file, You can obtain one at http://mozilla.org/MPL/2.0/.
//
// The conversion routines are Copyright (c) Fabian Giesen, 2016.
// The original license follows:
//
// Copyright (c) Fabian Giesen, 2016
// All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted.
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// “AS IS” AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.


// Standard 16-bit float type, mostly useful for GPUs. Defines a new
// type Eigen::half (inheriting from CUDA's __half struct) with
// operator overloads such that it behaves basically as an arithmetic
// type. It will be quite slow on CPUs (so it is recommended to stay
// in fp32 for CPUs, except for simple parameter conversions, I/O
// to disk and the likes), but fast on GPUs.


#ifndef EIGEN_HALF_CUDA_H
#define EIGEN_HALF_CUDA_H

namespace Eigen {

namespace half_impl {

// Make our own __half definition that is similar to CUDA's.
struct __half {
  EIGEN_DEVICE_FUNC __half() : x(0) {}
  explicit EIGEN_DEVICE_FUNC __half(unsigned short raw) : x(raw) {}
  unsigned short x;
};

EIGEN_STRONG_INLINE EIGEN_DEVICE_FUNC __half raw_uint16_to_half(unsigned short x);
EIGEN_STRONG_INLINE EIGEN_DEVICE_FUNC __half float_to_half_rtne(float ff);
EIGEN_STRONG_INLINE EIGEN_DEVICE_FUNC float half_to_float(__half h);

// Conversion routines, including fallbacks for the host or older CUDA.
// Note that newer Intel CPUs (Haswell or newer) have vectorized versions of
// these in hardware. If we need more performance on older/other CPUs, they are
// also possible to vectorize directly.

EIGEN_STRONG_INLINE EIGEN_DEVICE_FUNC __half raw_uint16_to_half(unsigned short x) {
  __half h;
  h.x = x;
  return h;
}

union FP32 {
  unsigned int u;
  float f;
};

EIGEN_STRONG_INLINE EIGEN_DEVICE_FUNC __half float_to_half_rtne(float ff) {
#if defined(EIGEN_HAS_CUDA_FP16) && defined(__CUDA_ARCH__) && __CUDA_ARCH__ >= 300
  return __float2half(ff);

#elif defined(EIGEN_HAS_FP16_C)
  __half h;
  h.x = _cvtss_sh(ff, 0);
  return h;

#else
  FP32 f; f.f = ff;

  const FP32 f32infty = { 255 << 23 };
  const FP32 f16max = { (127 + 16) << 23 };
  const FP32 denorm_magic = { ((127 - 15) + (23 - 10) + 1) << 23 };
  unsigned int sign_mask = 0x80000000u;
  __half o;
  o.x = static_cast<unsigned short>(0x0u);

  unsigned int sign = f.u & sign_mask;
  f.u ^= sign;

  // NOTE all the integer compares in this function can be safely
  // compiled into signed compares since all operands are below
  // 0x80000000. Important if you want fast straight SSE2 code
  // (since there's no unsigned PCMPGTD).

  if (f.u >= f16max.u) {  // result is Inf or NaN (all exponent bits set)
    o.x = (f.u > f32infty.u) ? 0x7e00 : 0x7c00; // NaN->qNaN and Inf->Inf
  } else {  // (De)normalized number or zero
    if (f.u < (113 << 23)) {  // resulting FP16 is subnormal or zero
      // use a magic value to align our 10 mantissa bits at the bottom of
      // the float. as long as FP addition is round-to-nearest-even this
      // just works.
      f.f += denorm_magic.f;

      // and one integer subtract of the bias later, we have our final float!
      o.x = static_cast<unsigned short>(f.u - denorm_magic.u);
    } else {
      unsigned int mant_odd = (f.u >> 13) & 1; // resulting mantissa is odd

      // update exponent, rounding bias part 1
      f.u += ((unsigned int)(15 - 127) << 23) + 0xfff;
      // rounding bias part 2
      f.u += mant_odd;
      // take the bits!
      o.x = static_cast<unsigned short>(f.u >> 13);
    }
  }

  o.x |= static_cast<unsigned short>(sign >> 16);
  return o;
#endif
}

EIGEN_STRONG_INLINE EIGEN_DEVICE_FUNC float half_to_float(__half h) {
#if defined(EIGEN_HAS_CUDA_FP16) && defined(__CUDA_ARCH__) && __CUDA_ARCH__ >= 300
  return __half2float(h);

#elif defined(EIGEN_HAS_FP16_C)
  return _cvtsh_ss(h.x);

#else
  const FP32 magic = { 113 << 23 };
  const unsigned int shifted_exp = 0x7c00 << 13; // exponent mask after shift
  FP32 o;

  o.u = (h.x & 0x7fff) << 13;             // exponent/mantissa bits
  unsigned int exp = shifted_exp & o.u;   // just the exponent
  o.u += (127 - 15) << 23;                // exponent adjust

  // handle exponent special cases
  if (exp == shifted_exp) {     // Inf/NaN?
    o.u += (128 - 16) << 23;    // extra exp adjust
  } else if (exp == 0) {        // Zero/Denormal?
    o.u += 1 << 23;             // extra exp adjust
    o.f -= magic.f;             // renormalize
  }

  o.u |= (h.x & 0x8000) << 16;    // sign bit
  return o.f;
#endif
}

} // end namespace half_impl

} // end namespace Eigen

#endif // EIGEN_HALF_CUDA_H