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
|
/*******************************************************************************
*
* MIT License
*
* Copyright (c) 2023 Advanced Micro Devices, Inc.
*
* 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.
*
*******************************************************************************/
#include <tuple>
#include <string_view>
#include "gtest_common.hpp"
#include "../conv2d.hpp"
namespace {
auto GetTestCases()
{
const auto env_fwd = std::tuple{
std::pair{MIOPEN_FIND_MODE, "normal"},
std::pair{MIOPEN_DEBUG_FIND_ONLY_SOLVER, "ConvAsmImplicitGemmGTCDynamicFwdDlopsNCHWC"}};
const std::string v = " --verbose";
const std::string dis_bk_data = " --disable-backward-data";
const std::string dis_bk_wei = " --disable-backward-weights";
const std::string in_nchw = " --in_layout NCHW";
const std::string fil_nchw = " --fil_layout NCHW";
const std::string fil_chwn = " --fil_layout CHWN";
const std::string out_nchw = " --out_layout NCHW";
const std::string tensor = " --tensor_vect 1";
const std::string vlen4 = " --vector_length 4";
const std::string vlen8 = " --vector_length 8";
const std::string common_base = " --cmode convfp16" + dis_bk_data + dis_bk_wei + in_nchw;
const std::string nchwc_nchwc_base = common_base + fil_nchw + out_nchw + tensor;
const std::string nchwc_nchwc_fwd_fp16x4 = nchwc_nchwc_base + vlen4;
const std::string nchwc_nchwc_fwd_fp16x8 = nchwc_nchwc_base + vlen8;
const std::string nchwc_chwnc_base = common_base + fil_chwn + out_nchw + tensor;
const std::string nchwc_chwnc_fwd_fp16x4 = nchwc_chwnc_base + vlen4;
const std::string nchwc_chwnc_fwd_fp16x8 = nchwc_chwnc_base + vlen8;
return std::vector{
// clang-format off
//nchwc_nchwc_fwd_fp16x4
std::pair{env_fwd, v + " --input 1 8 10 10 --weights 8 8 3 3 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 32 160 73 73 --weights 64 160 1 1 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 16 64 56 56 --weights 64 64 1 1 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 2 256 40 52 --weights 256 256 1 1 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 2 64 32 28 --weights 64 64 1 1 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 32 128 14 14 --weights 64 128 1 1 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 64 64 17 17 --weights 192 64 1 7 --pads_strides_dilations 0 3 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 64 64 17 17 --weights 192 64 7 1 --pads_strides_dilations 3 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 4 128 28 28 --weights 128 128 2 2 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 32 128 8 8 --weights 192 128 3 1 --pads_strides_dilations 1 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 64 192 17 17 --weights 160 192 3 3 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 64 32 73 73 --weights 64 32 3 3 --pads_strides_dilations 1 1 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 16 64 56 56 --weights 64 64 3 3 --pads_strides_dilations 1 1 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 16 16 25 25 --weights 64 16 3 3 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 4 32 79 141 --weights 64 32 5 10 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 400 256 7 7 --weights 1024 256 7 7 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 400 256 1 1 --weights 1024 256 1 1 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x4},
//nchwc_chwnc_fwd_fp16x4
std::pair{env_fwd, v + " --input 64 256 7 7 --weights 256 3 3 128 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 32 160 73 73 --weights 160 1 1 64 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 16 64 56 56 --weights 64 1 1 64 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 2 256 40 52 --weights 256 1 1 256 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 2 64 32 28 --weights 64 1 1 64 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 32 128 14 14 --weights 128 1 1 64 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 64 64 17 17 --weights 64 3 7 192 --pads_strides_dilations 0 3 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 64 64 17 17 --weights 64 7 1 192 --pads_strides_dilations 3 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 4 128 28 28 --weights 128 2 2 128 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 32 128 8 8 --weights 128 3 1 192 --pads_strides_dilations 1 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 64 192 17 17 --weights 192 3 3 160 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 64 32 73 73 --weights 32 3 3 64 --pads_strides_dilations 1 1 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 16 64 56 56 --weights 64 3 3 64 --pads_strides_dilations 1 1 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 16 16 25 25 --weights 16 3 3 64 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 4 32 79 141 --weights 32 5 10 64 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 400 256 7 7 --weights 256 7 7 1024 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
std::pair{env_fwd, v + " --input 400 256 1 1 --weights 256 1 1 1024 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x4},
//nchwc_nchwc_fwd_fp16x8
std::pair{env_fwd, v + " --input 1 8 10 10 --weights 8 8 3 3 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 32 160 73 73 --weights 64 160 1 1 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 16 64 56 56 --weights 64 64 1 1 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 2 256 40 52 --weights 256 256 1 1 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 2 64 32 28 --weights 64 64 1 1 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 32 128 14 14 --weights 64 128 1 1 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 64 64 17 17 --weights 192 64 1 7 --pads_strides_dilations 0 3 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 64 64 17 17 --weights 192 64 7 1 --pads_strides_dilations 3 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 4 128 28 28 --weights 128 128 2 2 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 32 128 8 8 --weights 192 128 3 1 --pads_strides_dilations 1 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 64 192 17 17 --weights 160 192 3 3 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 64 32 73 73 --weights 64 32 3 3 --pads_strides_dilations 1 1 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 16 64 56 56 --weights 64 64 3 3 --pads_strides_dilations 1 1 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 16 16 25 25 --weights 64 16 3 3 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 4 32 79 141 --weights 64 32 5 10 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 400 256 7 7 --weights 1024 256 7 7 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 400 256 1 1 --weights 1024 256 1 1 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_nchwc_fwd_fp16x8},
//nchwc_chwnc_fwd_fp16x8
std::pair{env_fwd, v + " --input 64 256 7 7 --weights 256 1 1 128 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 32 160 73 73 --weights 160 1 1 64 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 16 64 56 56 --weights 64 1 1 64 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 2 256 40 52 --weights 256 1 1 256 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 2 64 32 28 --weights 64 1 1 64 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 32 128 14 14 --weights 128 1 1 64 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 64 64 17 17 --weights 64 1 7 192 --pads_strides_dilations 0 3 1 1 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 64 64 17 17 --weights 64 7 1 192 --pads_strides_dilations 3 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 4 128 28 28 --weights 128 2 2 128 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 32 128 8 8 --weights 128 3 1 192 --pads_strides_dilations 1 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 64 192 17 17 --weights 192 3 3 160 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 64 32 73 73 --weights 32 3 3 64 --pads_strides_dilations 1 1 1 1 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 16 64 56 56 --weights 64 3 3 64 --pads_strides_dilations 1 1 1 1 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 16 16 25 25 --weights 16 3 3 64 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 4 32 79 141 --weights 32 5 10 64 --pads_strides_dilations 0 0 2 2 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 400 256 7 7 --weights 256 7 7 1024 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x8},
std::pair{env_fwd, v + " --input 400 256 1 1 --weights 256 1 1 1024 --pads_strides_dilations 0 0 1 1 1 1" + nchwc_chwnc_fwd_fp16x8}
// clang-format on
};
}
using TestCase = decltype(GetTestCases())::value_type;
bool SkipTest() { return get_handle_xnack(); }
bool IsTestSupportedForDevice()
{
using e_mask = enabled<Gpu::gfx103X>;
using d_mask = disabled<Gpu::gfx900, Gpu::gfx906, Gpu::gfx908, Gpu::gfx90A>;
return ::IsTestSupportedForDevMask<d_mask, e_mask>();
}
} // namespace
class GPU_Conv2dDefaultIGemmDynamicDLops_FP16 : public HalfTestCase<std::vector<TestCase>>
{
};
TEST_P(GPU_Conv2dDefaultIGemmDynamicDLops_FP16, HalfTest_conv_igemm_dynamic_dlops)
{
if(IsTestSupportedForDevice() && !SkipTest())
{
invoke_with_params<conv2d_driver, GPU_Conv2dDefaultIGemmDynamicDLops_FP16>(default_check);
}
else
{
GTEST_SKIP();
}
};
INSTANTIATE_TEST_SUITE_P(Full,
GPU_Conv2dDefaultIGemmDynamicDLops_FP16,
testing::Values(GetTestCases()));
|
