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) 2016, 2025, Oracle and/or its affiliates.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License, version 2.0,
as published by the Free Software Foundation.
This program is designed to work with certain software (including
but not limited to OpenSSL) that is licensed under separate terms,
as designated in a particular file or component or in included license
documentation. The authors of MySQL hereby grant you an additional
permission to link the program and your derivative works with the
separately licensed software that they have either included with
the program or referenced in the documentation.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License, version 2.0, for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */
#include <assert.h>
#include <math.h>
#include <stddef.h>
#include <stdio.h>
#include <algorithm>
#include <chrono>
using std::chrono::duration;
using std::chrono::nanoseconds;
using std::chrono::steady_clock;
static bool timer_running = false;
static double seconds_used;
static steady_clock::time_point timer_start;
static size_t bytes_processed = 0;
void StartBenchmarkTiming() {
assert(!timer_running);
timer_running = true;
timer_start = steady_clock::now();
}
void StopBenchmarkTiming() {
if (timer_running) {
auto used = steady_clock::now() - timer_start;
seconds_used += duration<double>(used).count();
timer_running = false;
}
}
void SetBytesProcessed(size_t bytes) { bytes_processed = bytes; }
void internal_do_microbenchmark(const char *name, void (*func)(size_t)) {
#if !defined(NDEBUG)
printf(
"WARNING: Running microbenchmark in debug mode. "
"Timings will be misleading.\n");
// There's no point in timing in debug mode, so just run 10 times
// so that we don't waste build time (this should give us enough runs
// to verify that we don't crash).
seconds_used = 0.0;
size_t num_iterations = 10;
StartBenchmarkTiming();
func(num_iterations);
StopBenchmarkTiming();
#else
// Do 100 iterations as rough calibration. (Often, this will over- or
// undershoot by as much as 50%, but that's fine.)
static constexpr size_t calibration_iterations = 100;
seconds_used = 0.0;
StartBenchmarkTiming();
func(calibration_iterations);
StopBenchmarkTiming();
double seconds_used_per_iteration = seconds_used / calibration_iterations;
size_t num_iterations;
// Do the actual run, unless we already took more than one second.
if (seconds_used < 1.0) {
// Scale so that we end up around one second per benchmark
// (but never less than 100).
num_iterations =
std::max<size_t>(lrint(1.0 / seconds_used_per_iteration), 100);
seconds_used = 0.0;
StartBenchmarkTiming();
func(num_iterations);
StopBenchmarkTiming();
} else {
// The calibration already took too long, so just reuse its results.
num_iterations = calibration_iterations;
}
#endif
printf("%-40s %10ld iterations %10.0f ns/iter", name,
static_cast<long>(num_iterations),
1e9 * seconds_used / double(num_iterations));
if (bytes_processed > 0) {
double bytes_per_second = bytes_processed / seconds_used;
if (bytes_per_second > (512 << 20)) // 0.5 GB/sec.
printf(" %8.2f GB/sec", bytes_per_second / (1 << 30));
else
printf(" %8.2f MB/sec", bytes_per_second / (1 << 20));
bytes_processed = 0; // Reset for next test.
}
printf("\n");
}
|
