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
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
|
/* This is part of the netCDF package.
Copyright 20057 University Corporation for Atmospheric Research/Unidata
See COPYRIGHT file for conditions of use.
Test HDF5 file code. These are not intended to be exhaustive tests,
but they use HDF5 the same way that netCDF-4 does, so if these
tests don't work, than netCDF-4 won't work either.
This files tests parallel I/O.
$Id: tst_h_par.c,v 1.15 2010/05/25 13:53:04 ed Exp $
*/
#include <nc_tests.h>
#include "err_macros.h"
#include <hdf5.h>
/* Defining USE_MPE causes the MPE trace library to be used (and you
* must also relink with -llmpe -lmpe). This causes clog2 output to be
* written, which can be converted to slog2 (by the program
* clog2TOslog2) and then used in the analysis program jumpshot. */
/*#define USE_MPE 1*/
#ifdef USE_MPE
#include <mpe.h>
#endif /* USE_MPE */
#define FILE_NAME "tst_h_par.h5"
#define VAR_NAME "HALs_memory"
#define NDIMS 1
#define MILLION 1000000
#define DIM2_LEN 16000000
#define SC1 100000 /* slice count. */
int
main(int argc, char **argv)
{
int p, my_rank;
#ifdef USE_MPE
int s_init, e_init, s_define, e_define, s_write, e_write, s_close, e_close;
#endif /* USE_MPE */
MPI_Init(&argc, &argv);
MPI_Comm_rank(MPI_COMM_WORLD, &my_rank);
MPI_Comm_size(MPI_COMM_WORLD, &p);
#ifdef USE_MPE
MPE_Init_log();
s_init = MPE_Log_get_event_number();
e_init = MPE_Log_get_event_number();
s_define = MPE_Log_get_event_number();
e_define = MPE_Log_get_event_number();
s_write = MPE_Log_get_event_number();
e_write = MPE_Log_get_event_number();
s_close = MPE_Log_get_event_number();
e_close = MPE_Log_get_event_number();
MPE_Describe_state(s_init, e_init, "Init", "red");
MPE_Describe_state(s_define, e_define, "Define", "yellow");
MPE_Describe_state(s_write, e_write, "Write", "green");
MPE_Describe_state(s_close, e_close, "Close", "purple");
MPE_Start_log();
MPE_Log_event(s_init, 0, "start init");
#endif /* USE_MPE */
if (!my_rank)
printf("*** Creating file for parallel I/O read, and rereading it...");
{
hid_t fapl_id, fileid, whole_spaceid, dsid, slice_spaceid, whole_spaceid1, xferid;
hsize_t start[NDIMS], count[NDIMS];
hsize_t dims[1];
int data[SC1], data_in[SC1];
int num_steps;
double ftime;
int write_us, read_us;
int max_write_us, max_read_us;
float write_rate, read_rate;
int i, s;
/* We will write the same slice of random data over and over to
* fill the file. */
for (i = 0; i < SC1; i++)
data[i] = rand();
#ifdef USE_MPE
MPE_Log_event(e_init, 0, "end init");
MPE_Log_event(s_define, 0, "start define file");
#endif /* USE_MPE */
/* Create file. */
if ((fapl_id = H5Pcreate(H5P_FILE_ACCESS)) < 0) ERR;
if (H5Pset_fapl_mpio(fapl_id, MPI_COMM_WORLD, MPI_INFO_NULL) < 0) ERR;
if ((fileid = H5Fcreate(FILE_NAME, H5F_ACC_TRUNC, H5P_DEFAULT,
fapl_id)) < 0) ERR;
/* Create a space to deal with one slice in memory. */
dims[0] = SC1;
if ((slice_spaceid = H5Screate_simple(NDIMS, dims, NULL)) < 0) ERR;
/* Create a space to write all slices. */
dims[0] = DIM2_LEN;
if ((whole_spaceid = H5Screate_simple(NDIMS, dims, NULL)) < 0) ERR;
/* Create dataset. */
if ((dsid = H5Dcreate1(fileid, VAR_NAME, H5T_NATIVE_INT,
whole_spaceid, H5P_DEFAULT)) < 0) ERR;
/* Use collective write operations. */
if ((xferid = H5Pcreate(H5P_DATASET_XFER)) < 0) ERR;
if (H5Pset_dxpl_mpio(xferid, H5FD_MPIO_COLLECTIVE) < 0) ERR;
#ifdef USE_MPE
MPE_Log_event(e_define, 0, "end define file");
if (my_rank)
sleep(my_rank);
#endif /* USE_MPE */
/* Write the data in num_step steps. */
ftime = MPI_Wtime();
num_steps = (DIM2_LEN/SC1) / p;
for (s = 0; s < num_steps; s++)
{
#ifdef USE_MPE
MPE_Log_event(s_write, 0, "start write slab");
#endif /* USE_MPE */
/* Select hyperslab for write of one slice. */
start[0] = s * SC1 * p + my_rank * SC1;
count[0] = SC1;
if (H5Sselect_hyperslab(whole_spaceid, H5S_SELECT_SET,
start, NULL, count, NULL) < 0) ERR;
if (H5Dwrite(dsid, H5T_NATIVE_INT, slice_spaceid, whole_spaceid,
xferid, data) < 0) ERR;
#ifdef USE_MPE
MPE_Log_event(e_write, 0, "end write file");
#endif /* USE_MPE */
}
write_us = (MPI_Wtime() - ftime) * MILLION;
MPI_Reduce(&write_us, &max_write_us, 1, MPI_INT, MPI_MAX, 0, MPI_COMM_WORLD);
if (!my_rank)
{
write_rate = (float)(DIM2_LEN * sizeof(int))/(float)max_write_us;
printf("\np=%d, write_rate=%g", p, write_rate);
}
#ifdef USE_MPE
MPE_Log_event(s_close, 0, "start close file");
#endif /* USE_MPE */
/* Close. These collective operations will allow every process
* to catch up. */
if (H5Dclose(dsid) < 0 ||
H5Sclose(whole_spaceid) < 0 ||
H5Sclose(slice_spaceid) < 0 ||
H5Pclose(fapl_id) < 0 ||
H5Fclose(fileid) < 0)
ERR;
#ifdef USE_MPE
MPE_Log_event(e_close, 0, "end close file");
#endif /* USE_MPE */
/* Open the file. */
if ((fapl_id = H5Pcreate(H5P_FILE_ACCESS)) < 0) ERR;
if (H5Pset_fapl_mpio(fapl_id, MPI_COMM_WORLD, MPI_INFO_NULL) < 0) ERR;
if (H5Pset_libver_bounds(fapl_id, H5F_LIBVER_LATEST, H5F_LIBVER_LATEST) < 0) ERR;
if ((fileid = H5Fopen(FILE_NAME, H5F_ACC_RDONLY, fapl_id)) < 0) ERR;
/* Create a space to deal with one slice in memory. */
dims[0] = SC1;
if ((slice_spaceid = H5Screate_simple(NDIMS, dims, NULL)) < 0) ERR;
/* Open the dataset. */
if ((dsid = H5Dopen(fileid, VAR_NAME)) < 0) ERR;
if ((whole_spaceid1 = H5Dget_space(dsid)) < 0) ERR;
ftime = MPI_Wtime();
/* Read the data, a slice at a time. */
for (s = 0; s < num_steps; s++)
{
/* Select hyperslab for read of one slice. */
start[0] = s * SC1 * p + my_rank * SC1;
count[0] = SC1;
if (H5Sselect_hyperslab(whole_spaceid1, H5S_SELECT_SET,
start, NULL, count, NULL) < 0)
{
ERR;
return 2;
}
if (H5Dread(dsid, H5T_NATIVE_INT, slice_spaceid, whole_spaceid1,
H5P_DEFAULT, data_in) < 0)
{
ERR;
return 2;
}
/* /\* Check the slice of data. *\/ */
/* for (i = 0; i < SC1; i++) */
/* if (data[i] != data_in[i]) */
/* { */
/* ERR; */
/* return 2; */
/* } */
}
read_us = (MPI_Wtime() - ftime) * MILLION;
MPI_Reduce(&read_us, &max_read_us, 1, MPI_INT, MPI_MAX, 0, MPI_COMM_WORLD);
if (!my_rank)
{
read_rate = (float)(DIM2_LEN * sizeof(int))/(float)max_read_us;
printf(", read_rate=%g\n", read_rate);
}
/* Close down. */
if (H5Dclose(dsid) < 0 ||
H5Sclose(slice_spaceid) < 0 ||
H5Sclose(whole_spaceid1) < 0 ||
H5Pclose(fapl_id) < 0 ||
H5Fclose(fileid) < 0)
ERR;
}
if (!my_rank)
SUMMARIZE_ERR;
MPI_Finalize();
if (!my_rank)
FINAL_RESULTS;
return 0;
}
|
