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
|
/* ************************************************************************
* Copyright (C) 2018-2024 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 cop-
* ies 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 IM-
* PLIED, 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 CONNE-
* CTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
*
* ************************************************************************ */
#pragma once
#include "testing_common.hpp"
template <typename T>
void testing_set_get_matrix_async(const Arguments& arg)
{
auto rocblas_set_matrix_async_fn = rocblas_set_matrix_async;
auto rocblas_set_matrix_async_fn_64 = rocblas_set_matrix_async_64;
auto rocblas_get_matrix_async_fn = rocblas_get_matrix_async;
auto rocblas_get_matrix_async_fn_64 = rocblas_get_matrix_async_64;
int64_t rows = arg.M;
int64_t cols = arg.N;
int64_t lda = arg.lda;
int64_t ldb = arg.ldb;
int64_t ldd = arg.ldd;
rocblas_local_handle handle{arg};
hipStream_t stream;
rocblas_get_stream(handle, &stream);
// argument sanity check, quick return if input parameters are invalid before allocating invalid
// memory
bool invalidGPUMatrix = rows < 0 || cols < 0 || ldd <= 0 || ldd < rows;
bool invalidSet = invalidGPUMatrix || lda <= 0 || lda < rows;
bool invalidGet = invalidGPUMatrix || ldb <= 0 || ldb < rows;
if(invalidSet || invalidGet)
{
DAPI_EXPECT(invalidSet ? rocblas_status_invalid_size : rocblas_status_invalid_pointer,
rocblas_set_matrix_async,
(rows, cols, sizeof(T), nullptr, lda, nullptr, ldd, stream));
DAPI_EXPECT(invalidGet ? rocblas_status_invalid_size : rocblas_status_invalid_pointer,
rocblas_get_matrix_async,
(rows, cols, sizeof(T), nullptr, ldd, nullptr, ldb, stream));
return;
}
// Naming: dK is in GPU (device) memory. hK is in CPU (host) memory,
HOST_MEMCHECK(host_pinned_vector<T>,
hA,
(cols * size_t(lda))); // using vector layout to reuse pinned_vector
HOST_MEMCHECK(host_pinned_vector<T>, hB, (cols * size_t(ldb)));
HOST_MEMCHECK(host_vector<T>, hB_gold, (cols * size_t(ldb)));
double cpu_time_used;
double rocblas_error = 0.0;
// allocate memory on device
DEVICE_MEMCHECK(device_vector<T>, dD, (cols * size_t(ldd))); // vector layout
// Initial Data on CPU
rocblas_seedrand();
rocblas_init<T>(hA, rows, cols, lda);
rocblas_init<T>(hB, rows, cols, ldb);
if(arg.unit_check || arg.norm_check)
{
CHECK_HIP_ERROR(hipMemset(dD, 0, sizeof(T) * cols * ldd)); // vector layout
DAPI_CHECK(rocblas_set_matrix_async_fn, (rows, cols, sizeof(T), hA, lda, dD, ldd, stream));
DAPI_CHECK(rocblas_get_matrix_async_fn, (rows, cols, sizeof(T), dD, ldd, hB, ldb, stream));
// reference calculation
cpu_time_used = get_time_us_no_sync();
for(size_t i2 = 0; i2 < cols; i2++)
for(size_t i1 = 0; i1 < rows; i1++)
hB_gold[i1 + i2 * ldb] = hA[i1 + i2 * lda];
cpu_time_used = get_time_us_no_sync() - cpu_time_used;
hipStreamSynchronize(stream);
if(arg.unit_check)
{
unit_check_general<T>(rows, cols, ldb, hB, hB_gold);
}
if(arg.norm_check)
{
rocblas_error = norm_check_general<T>('F', rows, cols, ldb, hB, hB_gold);
}
}
if(arg.timing)
{
double gpu_time_used;
int number_cold_calls = arg.cold_iters;
int total_calls = number_cold_calls + arg.iters;
hipStream_t stream;
CHECK_ROCBLAS_ERROR(rocblas_get_stream(handle, &stream));
for(int iter = 0; iter < total_calls; iter++)
{
if(iter == number_cold_calls)
gpu_time_used = get_time_us_sync(stream);
DAPI_DISPATCH(rocblas_set_matrix_async,
(rows, cols, sizeof(T), hA, lda, dD, ldd, stream));
DAPI_DISPATCH(rocblas_get_matrix_async,
(rows, cols, sizeof(T), dD, ldd, hB, ldb, stream));
}
hipStreamSynchronize(stream);
gpu_time_used = get_time_us_sync(stream) - gpu_time_used;
ArgumentModel<e_M, e_N, e_lda, e_ldb, e_ldd>{}.log_args<T>(
rocblas_cout,
arg,
gpu_time_used,
ArgumentLogging::NA_value,
set_get_matrix_gbyte_count<T>(rows, cols),
cpu_time_used,
rocblas_error);
}
}
|
