File: device_kernel.h

package info (click to toggle)
nvidia-cutlass 3.4.1%2Bds-2
  • links: PTS, VCS
  • area: contrib
  • in suites: forky, sid, trixie
  • size: 48,488 kB
  • sloc: cpp: 206,571; ansic: 69,215; python: 25,487; sh: 16; makefile: 15
file content (113 lines) | stat: -rw-r--r-- 4,321 bytes parent folder | download
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
/***************************************************************************************************
 * Copyright (c) 2017 - 2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 * SPDX-License-Identifier: BSD-3-Clause
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 * 1. Redistributions of source code must retain the above copyright notice, this
 * list of conditions and the following disclaimer.
 *
 * 2. Redistributions in binary form must reproduce the above copyright notice,
 * this list of conditions and the following disclaimer in the documentation
 * and/or other materials provided with the distribution.
 *
 * 3. Neither the name of the copyright holder nor the names of its
 * contributors may be used to endorse or promote products derived from
 * this software without specific prior written permission.
 *
 * 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.
 *
 **************************************************************************************************/
/*! \file
    \brief Template for generic CUTLASS kernel.
*/

#pragma once

// __grid_constant__ was introduced in CUDA 11.7.
#if ((__CUDACC_VER_MAJOR__ >= 12) || ((__CUDACC_VER_MAJOR__ == 11) && (__CUDACC_VER_MINOR__ >= 7)))
#  define CUTLASS_GRID_CONSTANT_SUPPORTED
#endif

// __grid_constant__ can be enabled only on SM70+
#if defined(CUTLASS_GRID_CONSTANT_SUPPORTED) && defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 700)
#  define CUTLASS_GRID_CONSTANT_ENABLED
#endif

#if ! defined(CUTLASS_GRID_CONSTANT)
#  if defined(CUTLASS_GRID_CONSTANT_ENABLED)
#    define CUTLASS_GRID_CONSTANT __grid_constant__
#  else
#    define CUTLASS_GRID_CONSTANT
#  endif
#endif

////////////////////////////////////////////////////////////////////////////////

namespace cutlass {

////////////////////////////////////////////////////////////////////////////////

/// Generic CUTLASS kernel template.
template <typename Operator>
CUTLASS_GLOBAL
void Kernel(typename Operator::Params params) {
  // Dynamic shared memory base pointer
  extern __shared__ int SharedStorageBase[];
  // Declare pointer to dynamic shared memory.
  typename Operator::SharedStorage *shared_storage =
      reinterpret_cast<typename Operator::SharedStorage *>(SharedStorageBase);

  Operator op;

  op(params, *shared_storage);
}


/// Generic CUTLASS kernel template.
template <typename Operator>
CUTLASS_GLOBAL
void Kernel2(typename Operator::Params params) {
  // Dynamic shared memory base pointer
  extern __shared__ int SharedStorageBase[];
  // Declare pointer to dynamic shared memory.
  typename Operator::SharedStorage *shared_storage =
      reinterpret_cast<typename Operator::SharedStorage *>(SharedStorageBase);

  Operator::invoke(params, *shared_storage);
}


////////////////////////////////////////////////////////////////////////////////
//
// 3.0 specific launch
//
////////////////////////////////////////////////////////////////////////////////

/// Generic CUTLASS kernel template.
template <typename Operator>
CUTLASS_GLOBAL
#ifdef __CUDACC__
// Enclosing this in __CUDACC__ suppresses MSVC warnings.
__launch_bounds__(Operator::MaxThreadsPerBlock, Operator::MinBlocksPerMultiprocessor)
#endif // __CUDACC__
void device_kernel(CUTLASS_GRID_CONSTANT typename Operator::Params const params)
{
  // Dynamic shared memory base pointer
  extern __shared__ char smem[];
  Operator op;
  op(params, smem);
}

////////////////////////////////////////////////////////////////////////////////
} /// namespace cutlass