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
|
/**
* Copyright (C) Mellanox Technologies Ltd. 2001-2016. ALL RIGHTS RESERVED.
* Copyright (C) UT-Battelle, LLC. 2015. ALL RIGHTS RESERVED.
*
* See file LICENSE for terms.
*/
#include "ucp_test.h"
#include <common/test_helpers.h>
#if _OPENMP
#include "omp.h"
#endif
using namespace ucs; /* For vector<char> serialization */
class test_ucp_rma_mt : public ucp_test {
public:
void init()
{
ucp_test::init();
sender().connect(&receiver(), get_ep_params());
for (int i = 0; i < sender().get_num_workers(); i++) {
/* avoid deadlock for blocking rma */
flush_worker(sender(), i);
}
}
static void send_cb(void *req, ucs_status_t status)
{
}
static void get_test_variants(std::vector<ucp_test_variant>& variants) {
add_variant(variants, UCP_FEATURE_RMA, MULTI_THREAD_CONTEXT);
add_variant(variants, UCP_FEATURE_RMA, MULTI_THREAD_WORKER);
}
};
UCS_TEST_P(test_ucp_rma_mt, put_get) {
const unsigned num_threads = mt_num_threads();
ucs_status_t st;
uint64_t orig_data[num_threads] GTEST_ATTRIBUTE_UNUSED_;
uint64_t target_data[num_threads] GTEST_ATTRIBUTE_UNUSED_;
ucp_mem_map_params_t params;
ucp_mem_h memh;
void *memheap = target_data;
params.field_mask = UCP_MEM_MAP_PARAM_FIELD_ADDRESS |
UCP_MEM_MAP_PARAM_FIELD_LENGTH |
UCP_MEM_MAP_PARAM_FIELD_FLAGS;
params.address = memheap;
params.length = sizeof(uint64_t) * num_threads;
params.flags = get_variant_value();
st = ucp_mem_map(receiver().ucph(), ¶ms, &memh);
ASSERT_UCS_OK(st);
void *rkey_buffer;
size_t rkey_buffer_size;
st = ucp_rkey_pack(receiver().ucph(), memh, &rkey_buffer, &rkey_buffer_size);
ASSERT_UCS_OK(st);
std::vector<ucp_rkey_h> rkey;
rkey.resize(num_threads);
/* test parallel rkey unpack */
#if _OPENMP && ENABLE_MT
#pragma omp parallel for
for (int i = 0; i < num_threads; i++) {
int worker_index = 0;
if (get_variant_thread_type() == MULTI_THREAD_CONTEXT) {
worker_index = i;
}
ucs_status_t status = ucp_ep_rkey_unpack(sender().ep(worker_index),
rkey_buffer, &rkey[i]);
ASSERT_UCS_OK(status);
}
#endif
ucp_rkey_buffer_release(rkey_buffer);
/* test blocking PUT */
for (int i = 0; i < num_threads; i++) {
orig_data[i] = 0xdeadbeefdeadbeef + 10 * i;
target_data[i] = 0;
}
#if _OPENMP && ENABLE_MT
#pragma omp parallel for
for (int i = 0; i < num_threads; i++) {
int worker_index = 0;
if (get_variant_thread_type() == MULTI_THREAD_CONTEXT) {
worker_index = i;
}
void* req = ucp_put_nb(sender().ep(worker_index), &orig_data[i],
sizeof(uint64_t), (uintptr_t)((uint64_t*)memheap + i),
rkey[i], send_cb);
request_wait(req, worker_index);
flush_worker(sender(), worker_index);
EXPECT_EQ(orig_data[i], target_data[i]);
}
#endif
/* test nonblocking PUT */
for (int i = 0; i < num_threads; i++) {
orig_data[i] = 0xdeadbeefdeadbeef + 10 * i;
target_data[i] = 0;
}
#if _OPENMP && ENABLE_MT
#pragma omp parallel for
for (int i = 0; i < num_threads; i++) {
ucs_status_t status;
int worker_index = 0;
if (get_variant_thread_type() == MULTI_THREAD_CONTEXT)
worker_index = i;
status = ucp_put_nbi(sender().ep(worker_index), &orig_data[i], sizeof(uint64_t),
(uintptr_t)((uint64_t*)memheap + i), rkey[i]);
ASSERT_UCS_OK_OR_INPROGRESS(status);
flush_worker(sender(), worker_index);
EXPECT_EQ(orig_data[i], target_data[i]);
}
#endif
/* test blocking GET */
for (int i = 0; i < num_threads; i++) {
orig_data[i] = 0;
target_data[i] = 0xdeadbeefdeadbeef + 10 * i;
}
#if _OPENMP && ENABLE_MT
#pragma omp parallel for
for (int i = 0; i < num_threads; i++) {
int worker_index = 0;
if (get_variant_thread_type() == MULTI_THREAD_CONTEXT) {
worker_index = i;
}
void *req = ucp_get_nb(sender().ep(worker_index), &orig_data[i],
sizeof(uint64_t), (uintptr_t)((uint64_t*)memheap + i),
rkey[i], send_cb);
request_wait(req, worker_index);
flush_worker(sender(), worker_index);
EXPECT_EQ(orig_data[i], target_data[i]);
}
#endif
/* test nonblocking GET */
for (int i = 0; i < num_threads; i++) {
orig_data[i] = 0;
target_data[i] = 0xdeadbeefdeadbeef + 10 * i;
}
#if _OPENMP && ENABLE_MT
#pragma omp parallel for
for (int i = 0; i < num_threads; i++) {
ucs_status_t status;
int worker_index = 0;
if (get_variant_thread_type() == MULTI_THREAD_CONTEXT) {
worker_index = i;
}
status = ucp_get_nbi(sender().ep(worker_index), &orig_data[i], sizeof(uint64_t),
(uintptr_t)((uint64_t *)memheap + i), rkey[i]);
ASSERT_UCS_OK_OR_INPROGRESS(status);
flush_worker(sender(), worker_index);
EXPECT_EQ(orig_data[i], target_data[i]);
}
#endif
#if _OPENMP && ENABLE_MT
#pragma omp parallel for
for (int i = 0; i < num_threads; i++) {
ucp_rkey_destroy(rkey[i]);
}
#endif
st = ucp_mem_unmap(receiver().ucph(), memh);
ASSERT_UCS_OK(st);
}
UCP_INSTANTIATE_TEST_CASE(test_ucp_rma_mt)
|
