1 perf-bench(1) 1 perf-bench(1)
2 ============= 2 =============
3 3
4 NAME 4 NAME
5 ---- 5 ----
6 perf-bench - General framework for benchmark s 6 perf-bench - General framework for benchmark suites
7 7
8 SYNOPSIS 8 SYNOPSIS
9 -------- 9 --------
10 [verse] 10 [verse]
11 'perf bench' [<common options>] <subsystem> <s 11 'perf bench' [<common options>] <subsystem> <suite> [<options>]
12 12
13 DESCRIPTION 13 DESCRIPTION
14 ----------- 14 -----------
15 This 'perf bench' command is a general framewo 15 This 'perf bench' command is a general framework for benchmark suites.
16 16
17 COMMON OPTIONS 17 COMMON OPTIONS
18 -------------- 18 --------------
19 -r:: 19 -r::
20 --repeat=:: 20 --repeat=::
21 Specify number of times to repeat the run (def !! 21 Specify amount of times to repeat the run (default 10).
22 22
23 -f:: 23 -f::
24 --format=:: 24 --format=::
25 Specify format style. 25 Specify format style.
26 Current available format styles are: 26 Current available format styles are:
27 27
28 'default':: 28 'default'::
29 Default style. This is mainly for human readin 29 Default style. This is mainly for human reading.
30 --------------------- 30 ---------------------
31 % perf bench sched pipe # 31 % perf bench sched pipe # with no style specified
32 (executing 1000000 pipe operations between two 32 (executing 1000000 pipe operations between two tasks)
33 Total time:5.855 sec 33 Total time:5.855 sec
34 5.855061 usecs/op 34 5.855061 usecs/op
35 170792 ops/sec 35 170792 ops/sec
36 --------------------- 36 ---------------------
37 37
38 'simple':: 38 'simple'::
39 This simple style is friendly for automated 39 This simple style is friendly for automated
40 processing by scripts. 40 processing by scripts.
41 --------------------- 41 ---------------------
42 % perf bench --format=simple sched pipe # 42 % perf bench --format=simple sched pipe # specified simple
43 5.988 43 5.988
44 --------------------- 44 ---------------------
45 45
46 SUBSYSTEM 46 SUBSYSTEM
47 --------- 47 ---------
48 48
49 'sched':: 49 'sched'::
50 Scheduler and IPC mechanisms. 50 Scheduler and IPC mechanisms.
51 51
52 'syscall':: <<
53 System call performance (throughput). <<
54 <<
55 'mem':: 52 'mem'::
56 Memory access performance. 53 Memory access performance.
57 54
58 'numa':: 55 'numa'::
59 NUMA scheduling and MM benchmarks. 56 NUMA scheduling and MM benchmarks.
60 57
61 'futex':: 58 'futex'::
62 Futex stressing benchmarks. 59 Futex stressing benchmarks.
63 60
64 'epoll':: 61 'epoll'::
65 Eventpoll (epoll) stressing benchmarks 62 Eventpoll (epoll) stressing benchmarks.
66 63
67 'internals':: <<
68 Benchmark internal perf functionality. <<
69 <<
70 'uprobe':: <<
71 Benchmark overhead of uprobe + BPF. <<
72 <<
73 'all':: 64 'all'::
74 All benchmark subsystems. 65 All benchmark subsystems.
75 66
76 SUITES FOR 'sched' 67 SUITES FOR 'sched'
77 ~~~~~~~~~~~~~~~~~~ 68 ~~~~~~~~~~~~~~~~~~
78 *messaging*:: 69 *messaging*::
79 Suite for evaluating performance of scheduler 70 Suite for evaluating performance of scheduler and IPC mechanisms.
80 Based on hackbench by Rusty Russell. 71 Based on hackbench by Rusty Russell.
81 72
82 Options of *messaging* 73 Options of *messaging*
83 ^^^^^^^^^^^^^^^^^^^^^^ 74 ^^^^^^^^^^^^^^^^^^^^^^
84 -p:: 75 -p::
85 --pipe:: 76 --pipe::
86 Use pipe() instead of socketpair() 77 Use pipe() instead of socketpair()
87 78
88 -t:: 79 -t::
89 --thread:: 80 --thread::
90 Be multi thread instead of multi process 81 Be multi thread instead of multi process
91 82
92 -g:: 83 -g::
93 --group=:: 84 --group=::
94 Specify number of groups 85 Specify number of groups
95 86
96 -l:: 87 -l::
97 --nr_loops=:: 88 --nr_loops=::
98 Specify number of loops 89 Specify number of loops
99 90
100 Example of *messaging* 91 Example of *messaging*
101 ^^^^^^^^^^^^^^^^^^^^^^ 92 ^^^^^^^^^^^^^^^^^^^^^^
102 93
103 --------------------- 94 ---------------------
104 % perf bench sched messaging # 95 % perf bench sched messaging # run with default
105 options (20 sender and receiver processes per 96 options (20 sender and receiver processes per group)
106 (10 groups == 400 processes run) 97 (10 groups == 400 processes run)
107 98
108 Total time:0.308 sec 99 Total time:0.308 sec
109 100
110 % perf bench sched messaging -t -g 20 # 101 % perf bench sched messaging -t -g 20 # be multi-thread, with 20 groups
111 (20 sender and receiver threads per group) 102 (20 sender and receiver threads per group)
112 (20 groups == 800 threads run) 103 (20 groups == 800 threads run)
113 104
114 Total time:0.582 sec 105 Total time:0.582 sec
115 --------------------- 106 ---------------------
116 107
117 *pipe*:: 108 *pipe*::
118 Suite for pipe() system call. 109 Suite for pipe() system call.
119 Based on pipe-test-1m.c by Ingo Molnar. 110 Based on pipe-test-1m.c by Ingo Molnar.
120 111
121 Options of *pipe* 112 Options of *pipe*
122 ^^^^^^^^^^^^^^^^^ 113 ^^^^^^^^^^^^^^^^^
123 -l:: 114 -l::
124 --loop=:: 115 --loop=::
125 Specify number of loops. 116 Specify number of loops.
126 117
127 -G:: <<
128 --cgroups=:: <<
129 Names of cgroups for sender and receiver, sepa <<
130 This is useful to check cgroup context switchi <<
131 Note that perf doesn't create nor delete the c <<
132 make sure that the cgroups exist and are acces <<
133 <<
134 <<
135 Example of *pipe* 118 Example of *pipe*
136 ^^^^^^^^^^^^^^^^^ 119 ^^^^^^^^^^^^^^^^^
137 120
138 --------------------- 121 ---------------------
139 % perf bench sched pipe 122 % perf bench sched pipe
140 (executing 1000000 pipe operations between two 123 (executing 1000000 pipe operations between two tasks)
141 124
142 Total time:8.091 sec 125 Total time:8.091 sec
143 8.091833 usecs/op 126 8.091833 usecs/op
144 123581 ops/sec 127 123581 ops/sec
145 128
146 % perf bench sched pipe -l 1000 # 129 % perf bench sched pipe -l 1000 # loop 1000
147 (executing 1000 pipe operations between two ta 130 (executing 1000 pipe operations between two tasks)
148 131
149 Total time:0.016 sec 132 Total time:0.016 sec
150 16.948000 usecs/op 133 16.948000 usecs/op
151 59004 ops/sec 134 59004 ops/sec
152 <<
153 % perf bench sched pipe -G AAA,BBB <<
154 (executing 1000000 pipe operations between cgr <<
155 # Running 'sched/pipe' benchmark: <<
156 # Executed 1000000 pipe operations between two <<
157 <<
158 Total time: 6.886 [sec] <<
159 <<
160 6.886208 usecs/op <<
161 145217 ops/sec <<
162 <<
163 --------------------- 135 ---------------------
164 136
165 SUITES FOR 'syscall' <<
166 ~~~~~~~~~~~~~~~~~~ <<
167 *basic*:: <<
168 Suite for evaluating performance of core syste <<
169 This uses a single thread simply doing getppid <<
170 cached by glibc. <<
171 <<
172 <<
173 SUITES FOR 'mem' 137 SUITES FOR 'mem'
174 ~~~~~~~~~~~~~~~~ 138 ~~~~~~~~~~~~~~~~
175 *memcpy*:: 139 *memcpy*::
176 Suite for evaluating performance of simple mem 140 Suite for evaluating performance of simple memory copy in various ways.
177 141
178 Options of *memcpy* 142 Options of *memcpy*
179 ^^^^^^^^^^^^^^^^^^^ 143 ^^^^^^^^^^^^^^^^^^^
180 -l:: 144 -l::
181 --size:: 145 --size::
182 Specify size of memory to copy (default: 1MB). 146 Specify size of memory to copy (default: 1MB).
183 Available units are B, KB, MB, GB and TB (case 147 Available units are B, KB, MB, GB and TB (case insensitive).
184 148
185 -f:: 149 -f::
186 --function:: 150 --function::
187 Specify function to copy (default: default). 151 Specify function to copy (default: default).
188 Available functions are depend on the architec 152 Available functions are depend on the architecture.
189 On x86-64, x86-64-unrolled, x86-64-movsq and x 153 On x86-64, x86-64-unrolled, x86-64-movsq and x86-64-movsb are supported.
190 154
191 -l:: 155 -l::
192 --nr_loops:: 156 --nr_loops::
193 Repeat memcpy invocation this number of times. 157 Repeat memcpy invocation this number of times.
194 158
195 -c:: 159 -c::
196 --cycles:: 160 --cycles::
197 Use perf's cpu-cycles event instead of gettime 161 Use perf's cpu-cycles event instead of gettimeofday syscall.
198 162
199 *memset*:: 163 *memset*::
200 Suite for evaluating performance of simple mem 164 Suite for evaluating performance of simple memory set in various ways.
201 165
202 Options of *memset* 166 Options of *memset*
203 ^^^^^^^^^^^^^^^^^^^ 167 ^^^^^^^^^^^^^^^^^^^
204 -l:: 168 -l::
205 --size:: 169 --size::
206 Specify size of memory to set (default: 1MB). 170 Specify size of memory to set (default: 1MB).
207 Available units are B, KB, MB, GB and TB (case 171 Available units are B, KB, MB, GB and TB (case insensitive).
208 172
209 -f:: 173 -f::
210 --function:: 174 --function::
211 Specify function to set (default: default). 175 Specify function to set (default: default).
212 Available functions are depend on the architec 176 Available functions are depend on the architecture.
213 On x86-64, x86-64-unrolled, x86-64-stosq and x 177 On x86-64, x86-64-unrolled, x86-64-stosq and x86-64-stosb are supported.
214 178
215 -l:: 179 -l::
216 --nr_loops:: 180 --nr_loops::
217 Repeat memset invocation this number of times. 181 Repeat memset invocation this number of times.
218 182
219 -c:: 183 -c::
220 --cycles:: 184 --cycles::
221 Use perf's cpu-cycles event instead of gettime 185 Use perf's cpu-cycles event instead of gettimeofday syscall.
222 186
223 SUITES FOR 'numa' 187 SUITES FOR 'numa'
224 ~~~~~~~~~~~~~~~~~ 188 ~~~~~~~~~~~~~~~~~
225 *mem*:: 189 *mem*::
226 Suite for evaluating NUMA workloads. 190 Suite for evaluating NUMA workloads.
227 191
228 SUITES FOR 'futex' 192 SUITES FOR 'futex'
229 ~~~~~~~~~~~~~~~~~~ 193 ~~~~~~~~~~~~~~~~~~
230 *hash*:: 194 *hash*::
231 Suite for evaluating hash tables. 195 Suite for evaluating hash tables.
232 196
233 *wake*:: 197 *wake*::
234 Suite for evaluating wake calls. 198 Suite for evaluating wake calls.
235 199
236 *wake-parallel*:: 200 *wake-parallel*::
237 Suite for evaluating parallel wake calls. 201 Suite for evaluating parallel wake calls.
238 202
239 *requeue*:: 203 *requeue*::
240 Suite for evaluating requeue calls. 204 Suite for evaluating requeue calls.
241 205
242 *lock-pi*:: 206 *lock-pi*::
243 Suite for evaluating futex lock_pi calls. 207 Suite for evaluating futex lock_pi calls.
244 208
245 SUITES FOR 'epoll' 209 SUITES FOR 'epoll'
246 ~~~~~~~~~~~~~~~~~~ 210 ~~~~~~~~~~~~~~~~~~
247 *wait*:: 211 *wait*::
248 Suite for evaluating concurrent epoll_wait cal 212 Suite for evaluating concurrent epoll_wait calls.
249 213
250 *ctl*:: 214 *ctl*::
251 Suite for evaluating multiple epoll_ctl calls. 215 Suite for evaluating multiple epoll_ctl calls.
252 <<
253 SUITES FOR 'internals' <<
254 ~~~~~~~~~~~~~~~~~~~~~~ <<
255 *synthesize*:: <<
256 Suite for evaluating perf's event synthesis pe <<
257 216
258 SEE ALSO 217 SEE ALSO
259 -------- 218 --------
260 linkperf:perf[1] 219 linkperf:perf[1]
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.