Loading...
1perf-stat(1)
2============
3
4NAME
5----
6perf-stat - Run a command and gather performance counter statistics
7
8SYNOPSIS
9--------
10[verse]
11'perf stat' [-e <EVENT> | --event=EVENT] [-a] <command>
12'perf stat' [-e <EVENT> | --event=EVENT] [-a] -- <command> [<options>]
13'perf stat' [-e <EVENT> | --event=EVENT] [-a] record [-o file] -- <command> [<options>]
14'perf stat' report [-i file]
15
16DESCRIPTION
17-----------
18This command runs a command and gathers performance counter statistics
19from it.
20
21
22OPTIONS
23-------
24<command>...::
25 Any command you can specify in a shell.
26
27record::
28 See STAT RECORD.
29
30report::
31 See STAT REPORT.
32
33-e::
34--event=::
35 Select the PMU event. Selection can be:
36
37 - a symbolic event name (use 'perf list' to list all events)
38
39 - a raw PMU event (eventsel+umask) in the form of rNNN where NNN is a
40 hexadecimal event descriptor.
41
42 - a symbolically formed event like 'pmu/param1=0x3,param2/' where
43 param1 and param2 are defined as formats for the PMU in
44 /sys/bus/event_source/devices/<pmu>/format/*
45
46 'percore' is a event qualifier that sums up the event counts for both
47 hardware threads in a core. For example:
48 perf stat -A -a -e cpu/event,percore=1/,otherevent ...
49
50 - a symbolically formed event like 'pmu/config=M,config1=N,config2=K/'
51 where M, N, K are numbers (in decimal, hex, octal format).
52 Acceptable values for each of 'config', 'config1' and 'config2'
53 parameters are defined by corresponding entries in
54 /sys/bus/event_source/devices/<pmu>/format/*
55
56 Note that the last two syntaxes support prefix and glob matching in
57 the PMU name to simplify creation of events across multiple instances
58 of the same type of PMU in large systems (e.g. memory controller PMUs).
59 Multiple PMU instances are typical for uncore PMUs, so the prefix
60 'uncore_' is also ignored when performing this match.
61
62
63-i::
64--no-inherit::
65 child tasks do not inherit counters
66-p::
67--pid=<pid>::
68 stat events on existing process id (comma separated list)
69
70-t::
71--tid=<tid>::
72 stat events on existing thread id (comma separated list)
73
74
75-a::
76--all-cpus::
77 system-wide collection from all CPUs (default if no target is specified)
78
79--no-scale::
80 Don't scale/normalize counter values
81
82-d::
83--detailed::
84 print more detailed statistics, can be specified up to 3 times
85
86 -d: detailed events, L1 and LLC data cache
87 -d -d: more detailed events, dTLB and iTLB events
88 -d -d -d: very detailed events, adding prefetch events
89
90-r::
91--repeat=<n>::
92 repeat command and print average + stddev (max: 100). 0 means forever.
93
94-B::
95--big-num::
96 print large numbers with thousands' separators according to locale
97
98-C::
99--cpu=::
100Count only on the list of CPUs provided. Multiple CPUs can be provided as a
101comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-2.
102In per-thread mode, this option is ignored. The -a option is still necessary
103to activate system-wide monitoring. Default is to count on all CPUs.
104
105-A::
106--no-aggr::
107Do not aggregate counts across all monitored CPUs.
108
109-n::
110--null::
111 null run - don't start any counters
112
113-v::
114--verbose::
115 be more verbose (show counter open errors, etc)
116
117-x SEP::
118--field-separator SEP::
119print counts using a CSV-style output to make it easy to import directly into
120spreadsheets. Columns are separated by the string specified in SEP.
121
122--table:: Display time for each run (-r option), in a table format, e.g.:
123
124 $ perf stat --null -r 5 --table perf bench sched pipe
125
126 Performance counter stats for 'perf bench sched pipe' (5 runs):
127
128 # Table of individual measurements:
129 5.189 (-0.293) #
130 5.189 (-0.294) #
131 5.186 (-0.296) #
132 5.663 (+0.181) ##
133 6.186 (+0.703) ####
134
135 # Final result:
136 5.483 +- 0.198 seconds time elapsed ( +- 3.62% )
137
138-G name::
139--cgroup name::
140monitor only in the container (cgroup) called "name". This option is available only
141in per-cpu mode. The cgroup filesystem must be mounted. All threads belonging to
142container "name" are monitored when they run on the monitored CPUs. Multiple cgroups
143can be provided. Each cgroup is applied to the corresponding event, i.e., first cgroup
144to first event, second cgroup to second event and so on. It is possible to provide
145an empty cgroup (monitor all the time) using, e.g., -G foo,,bar. Cgroups must have
146corresponding events, i.e., they always refer to events defined earlier on the command
147line. If the user wants to track multiple events for a specific cgroup, the user can
148use '-e e1 -e e2 -G foo,foo' or just use '-e e1 -e e2 -G foo'.
149
150If wanting to monitor, say, 'cycles' for a cgroup and also for system wide, this
151command line can be used: 'perf stat -e cycles -G cgroup_name -a -e cycles'.
152
153-o file::
154--output file::
155Print the output into the designated file.
156
157--append::
158Append to the output file designated with the -o option. Ignored if -o is not specified.
159
160--log-fd::
161
162Log output to fd, instead of stderr. Complementary to --output, and mutually exclusive
163with it. --append may be used here. Examples:
164 3>results perf stat --log-fd 3 -- $cmd
165 3>>results perf stat --log-fd 3 --append -- $cmd
166
167--pre::
168--post::
169 Pre and post measurement hooks, e.g.:
170
171perf stat --repeat 10 --null --sync --pre 'make -s O=defconfig-build/clean' -- make -s -j64 O=defconfig-build/ bzImage
172
173-I msecs::
174--interval-print msecs::
175Print count deltas every N milliseconds (minimum: 1ms)
176The overhead percentage could be high in some cases, for instance with small, sub 100ms intervals. Use with caution.
177 example: 'perf stat -I 1000 -e cycles -a sleep 5'
178
179--interval-count times::
180Print count deltas for fixed number of times.
181This option should be used together with "-I" option.
182 example: 'perf stat -I 1000 --interval-count 2 -e cycles -a'
183
184--interval-clear::
185Clear the screen before next interval.
186
187--timeout msecs::
188Stop the 'perf stat' session and print count deltas after N milliseconds (minimum: 10 ms).
189This option is not supported with the "-I" option.
190 example: 'perf stat --time 2000 -e cycles -a'
191
192--metric-only::
193Only print computed metrics. Print them in a single line.
194Don't show any raw values. Not supported with --per-thread.
195
196--per-socket::
197Aggregate counts per processor socket for system-wide mode measurements. This
198is a useful mode to detect imbalance between sockets. To enable this mode,
199use --per-socket in addition to -a. (system-wide). The output includes the
200socket number and the number of online processors on that socket. This is
201useful to gauge the amount of aggregation.
202
203--per-die::
204Aggregate counts per processor die for system-wide mode measurements. This
205is a useful mode to detect imbalance between dies. To enable this mode,
206use --per-die in addition to -a. (system-wide). The output includes the
207die number and the number of online processors on that die. This is
208useful to gauge the amount of aggregation.
209
210--per-core::
211Aggregate counts per physical processor for system-wide mode measurements. This
212is a useful mode to detect imbalance between physical cores. To enable this mode,
213use --per-core in addition to -a. (system-wide). The output includes the
214core number and the number of online logical processors on that physical processor.
215
216--per-thread::
217Aggregate counts per monitored threads, when monitoring threads (-t option)
218or processes (-p option).
219
220-D msecs::
221--delay msecs::
222After starting the program, wait msecs before measuring. This is useful to
223filter out the startup phase of the program, which is often very different.
224
225-T::
226--transaction::
227
228Print statistics of transactional execution if supported.
229
230STAT RECORD
231-----------
232Stores stat data into perf data file.
233
234-o file::
235--output file::
236Output file name.
237
238STAT REPORT
239-----------
240Reads and reports stat data from perf data file.
241
242-i file::
243--input file::
244Input file name.
245
246--per-socket::
247Aggregate counts per processor socket for system-wide mode measurements.
248
249--per-die::
250Aggregate counts per processor die for system-wide mode measurements.
251
252--per-core::
253Aggregate counts per physical processor for system-wide mode measurements.
254
255-M::
256--metrics::
257Print metrics or metricgroups specified in a comma separated list.
258For a group all metrics from the group are added.
259The events from the metrics are automatically measured.
260See perf list output for the possble metrics and metricgroups.
261
262-A::
263--no-aggr::
264Do not aggregate counts across all monitored CPUs.
265
266--topdown::
267Print top down level 1 metrics if supported by the CPU. This allows to
268determine bottle necks in the CPU pipeline for CPU bound workloads,
269by breaking the cycles consumed down into frontend bound, backend bound,
270bad speculation and retiring.
271
272Frontend bound means that the CPU cannot fetch and decode instructions fast
273enough. Backend bound means that computation or memory access is the bottle
274neck. Bad Speculation means that the CPU wasted cycles due to branch
275mispredictions and similar issues. Retiring means that the CPU computed without
276an apparently bottleneck. The bottleneck is only the real bottleneck
277if the workload is actually bound by the CPU and not by something else.
278
279For best results it is usually a good idea to use it with interval
280mode like -I 1000, as the bottleneck of workloads can change often.
281
282The top down metrics are collected per core instead of per
283CPU thread. Per core mode is automatically enabled
284and -a (global monitoring) is needed, requiring root rights or
285perf.perf_event_paranoid=-1.
286
287Topdown uses the full Performance Monitoring Unit, and needs
288disabling of the NMI watchdog (as root):
289echo 0 > /proc/sys/kernel/nmi_watchdog
290for best results. Otherwise the bottlenecks may be inconsistent
291on workload with changing phases.
292
293This enables --metric-only, unless overridden with --no-metric-only.
294
295To interpret the results it is usually needed to know on which
296CPUs the workload runs on. If needed the CPUs can be forced using
297taskset.
298
299--no-merge::
300Do not merge results from same PMUs.
301
302When multiple events are created from a single event specification,
303stat will, by default, aggregate the event counts and show the result
304in a single row. This option disables that behavior and shows
305the individual events and counts.
306
307Multiple events are created from a single event specification when:
3081. Prefix or glob matching is used for the PMU name.
3092. Aliases, which are listed immediately after the Kernel PMU events
310 by perf list, are used.
311
312--smi-cost::
313Measure SMI cost if msr/aperf/ and msr/smi/ events are supported.
314
315During the measurement, the /sys/device/cpu/freeze_on_smi will be set to
316freeze core counters on SMI.
317The aperf counter will not be effected by the setting.
318The cost of SMI can be measured by (aperf - unhalted core cycles).
319
320In practice, the percentages of SMI cycles is very useful for performance
321oriented analysis. --metric_only will be applied by default.
322The output is SMI cycles%, equals to (aperf - unhalted core cycles) / aperf
323
324Users who wants to get the actual value can apply --no-metric-only.
325
326EXAMPLES
327--------
328
329$ perf stat -- make
330
331 Performance counter stats for 'make':
332
333 83723.452481 task-clock:u (msec) # 1.004 CPUs utilized
334 0 context-switches:u # 0.000 K/sec
335 0 cpu-migrations:u # 0.000 K/sec
336 3,228,188 page-faults:u # 0.039 M/sec
337 229,570,665,834 cycles:u # 2.742 GHz
338 313,163,853,778 instructions:u # 1.36 insn per cycle
339 69,704,684,856 branches:u # 832.559 M/sec
340 2,078,861,393 branch-misses:u # 2.98% of all branches
341
342 83.409183620 seconds time elapsed
343
344 74.684747000 seconds user
345 8.739217000 seconds sys
346
347TIMINGS
348-------
349As displayed in the example above we can display 3 types of timings.
350We always display the time the counters were enabled/alive:
351
352 83.409183620 seconds time elapsed
353
354For workload sessions we also display time the workloads spent in
355user/system lands:
356
357 74.684747000 seconds user
358 8.739217000 seconds sys
359
360Those times are the very same as displayed by the 'time' tool.
361
362CSV FORMAT
363----------
364
365With -x, perf stat is able to output a not-quite-CSV format output
366Commas in the output are not put into "". To make it easy to parse
367it is recommended to use a different character like -x \;
368
369The fields are in this order:
370
371 - optional usec time stamp in fractions of second (with -I xxx)
372 - optional CPU, core, or socket identifier
373 - optional number of logical CPUs aggregated
374 - counter value
375 - unit of the counter value or empty
376 - event name
377 - run time of counter
378 - percentage of measurement time the counter was running
379 - optional variance if multiple values are collected with -r
380 - optional metric value
381 - optional unit of metric
382
383Additional metrics may be printed with all earlier fields being empty.
384
385SEE ALSO
386--------
387linkperf:perf-top[1], linkperf:perf-list[1]
1perf-stat(1)
2============
3
4NAME
5----
6perf-stat - Run a command and gather performance counter statistics
7
8SYNOPSIS
9--------
10[verse]
11'perf stat' [-e <EVENT> | --event=EVENT] [-a] <command>
12'perf stat' [-e <EVENT> | --event=EVENT] [-a] -- <command> [<options>]
13'perf stat' [-e <EVENT> | --event=EVENT] [-a] record [-o file] -- <command> [<options>]
14'perf stat' report [-i file]
15
16DESCRIPTION
17-----------
18This command runs a command and gathers performance counter statistics
19from it.
20
21
22OPTIONS
23-------
24<command>...::
25 Any command you can specify in a shell.
26
27record::
28 See STAT RECORD.
29
30report::
31 See STAT REPORT.
32
33-e::
34--event=::
35 Select the PMU event. Selection can be:
36
37 - a symbolic event name (use 'perf list' to list all events)
38
39 - a raw PMU event (eventsel+umask) in the form of rNNN where NNN is a
40 hexadecimal event descriptor.
41
42 - a symbolic or raw PMU event followed by an optional colon
43 and a list of event modifiers, e.g., cpu-cycles:p. See the
44 linkperf:perf-list[1] man page for details on event modifiers.
45
46 - a symbolically formed event like 'pmu/param1=0x3,param2/' where
47 param1 and param2 are defined as formats for the PMU in
48 /sys/bus/event_source/devices/<pmu>/format/*
49
50 'percore' is a event qualifier that sums up the event counts for both
51 hardware threads in a core. For example:
52 perf stat -A -a -e cpu/event,percore=1/,otherevent ...
53
54 - a symbolically formed event like 'pmu/config=M,config1=N,config2=K/'
55 where M, N, K are numbers (in decimal, hex, octal format).
56 Acceptable values for each of 'config', 'config1' and 'config2'
57 parameters are defined by corresponding entries in
58 /sys/bus/event_source/devices/<pmu>/format/*
59
60 Note that the last two syntaxes support prefix and glob matching in
61 the PMU name to simplify creation of events across multiple instances
62 of the same type of PMU in large systems (e.g. memory controller PMUs).
63 Multiple PMU instances are typical for uncore PMUs, so the prefix
64 'uncore_' is also ignored when performing this match.
65
66
67-i::
68--no-inherit::
69 child tasks do not inherit counters
70-p::
71--pid=<pid>::
72 stat events on existing process id (comma separated list)
73
74-t::
75--tid=<tid>::
76 stat events on existing thread id (comma separated list)
77
78ifdef::HAVE_LIBPFM[]
79--pfm-events events::
80Select a PMU event using libpfm4 syntax (see http://perfmon2.sf.net)
81including support for event filters. For example '--pfm-events
82inst_retired:any_p:u:c=1:i'. More than one event can be passed to the
83option using the comma separator. Hardware events and generic hardware
84events cannot be mixed together. The latter must be used with the -e
85option. The -e option and this one can be mixed and matched. Events
86can be grouped using the {} notation.
87endif::HAVE_LIBPFM[]
88
89-a::
90--all-cpus::
91 system-wide collection from all CPUs (default if no target is specified)
92
93--no-scale::
94 Don't scale/normalize counter values
95
96-d::
97--detailed::
98 print more detailed statistics, can be specified up to 3 times
99
100 -d: detailed events, L1 and LLC data cache
101 -d -d: more detailed events, dTLB and iTLB events
102 -d -d -d: very detailed events, adding prefetch events
103
104-r::
105--repeat=<n>::
106 repeat command and print average + stddev (max: 100). 0 means forever.
107
108-B::
109--big-num::
110 print large numbers with thousands' separators according to locale.
111 Enabled by default. Use "--no-big-num" to disable.
112 Default setting can be changed with "perf config stat.big-num=false".
113
114-C::
115--cpu=::
116Count only on the list of CPUs provided. Multiple CPUs can be provided as a
117comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-2.
118In per-thread mode, this option is ignored. The -a option is still necessary
119to activate system-wide monitoring. Default is to count on all CPUs.
120
121-A::
122--no-aggr::
123Do not aggregate counts across all monitored CPUs.
124
125-n::
126--null::
127 null run - don't start any counters
128
129-v::
130--verbose::
131 be more verbose (show counter open errors, etc)
132
133-x SEP::
134--field-separator SEP::
135print counts using a CSV-style output to make it easy to import directly into
136spreadsheets. Columns are separated by the string specified in SEP.
137
138--table:: Display time for each run (-r option), in a table format, e.g.:
139
140 $ perf stat --null -r 5 --table perf bench sched pipe
141
142 Performance counter stats for 'perf bench sched pipe' (5 runs):
143
144 # Table of individual measurements:
145 5.189 (-0.293) #
146 5.189 (-0.294) #
147 5.186 (-0.296) #
148 5.663 (+0.181) ##
149 6.186 (+0.703) ####
150
151 # Final result:
152 5.483 +- 0.198 seconds time elapsed ( +- 3.62% )
153
154-G name::
155--cgroup name::
156monitor only in the container (cgroup) called "name". This option is available only
157in per-cpu mode. The cgroup filesystem must be mounted. All threads belonging to
158container "name" are monitored when they run on the monitored CPUs. Multiple cgroups
159can be provided. Each cgroup is applied to the corresponding event, i.e., first cgroup
160to first event, second cgroup to second event and so on. It is possible to provide
161an empty cgroup (monitor all the time) using, e.g., -G foo,,bar. Cgroups must have
162corresponding events, i.e., they always refer to events defined earlier on the command
163line. If the user wants to track multiple events for a specific cgroup, the user can
164use '-e e1 -e e2 -G foo,foo' or just use '-e e1 -e e2 -G foo'.
165
166If wanting to monitor, say, 'cycles' for a cgroup and also for system wide, this
167command line can be used: 'perf stat -e cycles -G cgroup_name -a -e cycles'.
168
169-o file::
170--output file::
171Print the output into the designated file.
172
173--append::
174Append to the output file designated with the -o option. Ignored if -o is not specified.
175
176--log-fd::
177
178Log output to fd, instead of stderr. Complementary to --output, and mutually exclusive
179with it. --append may be used here. Examples:
180 3>results perf stat --log-fd 3 -- $cmd
181 3>>results perf stat --log-fd 3 --append -- $cmd
182
183--control fd:ctl-fd[,ack-fd]
184Listen on ctl-fd descriptor for command to control measurement ('enable': enable events,
185'disable': disable events). Measurements can be started with events disabled using
186--delay=-1 option. Optionally send control command completion ('ack\n') to ack-fd descriptor
187to synchronize with the controlling process. Example of bash shell script to enable and
188disable events during measurements:
189
190#!/bin/bash
191
192ctl_dir=/tmp/
193
194ctl_fifo=${ctl_dir}perf_ctl.fifo
195test -p ${ctl_fifo} && unlink ${ctl_fifo}
196mkfifo ${ctl_fifo}
197exec {ctl_fd}<>${ctl_fifo}
198
199ctl_ack_fifo=${ctl_dir}perf_ctl_ack.fifo
200test -p ${ctl_ack_fifo} && unlink ${ctl_ack_fifo}
201mkfifo ${ctl_ack_fifo}
202exec {ctl_fd_ack}<>${ctl_ack_fifo}
203
204perf stat -D -1 -e cpu-cycles -a -I 1000 \
205 --control fd:${ctl_fd},${ctl_fd_ack} \
206 -- sleep 30 &
207perf_pid=$!
208
209sleep 5 && echo 'enable' >&${ctl_fd} && read -u ${ctl_fd_ack} e1 && echo "enabled(${e1})"
210sleep 10 && echo 'disable' >&${ctl_fd} && read -u ${ctl_fd_ack} d1 && echo "disabled(${d1})"
211
212exec {ctl_fd_ack}>&-
213unlink ${ctl_ack_fifo}
214
215exec {ctl_fd}>&-
216unlink ${ctl_fifo}
217
218wait -n ${perf_pid}
219exit $?
220
221
222--pre::
223--post::
224 Pre and post measurement hooks, e.g.:
225
226perf stat --repeat 10 --null --sync --pre 'make -s O=defconfig-build/clean' -- make -s -j64 O=defconfig-build/ bzImage
227
228-I msecs::
229--interval-print msecs::
230Print count deltas every N milliseconds (minimum: 1ms)
231The overhead percentage could be high in some cases, for instance with small, sub 100ms intervals. Use with caution.
232 example: 'perf stat -I 1000 -e cycles -a sleep 5'
233
234If the metric exists, it is calculated by the counts generated in this interval and the metric is printed after #.
235
236--interval-count times::
237Print count deltas for fixed number of times.
238This option should be used together with "-I" option.
239 example: 'perf stat -I 1000 --interval-count 2 -e cycles -a'
240
241--interval-clear::
242Clear the screen before next interval.
243
244--timeout msecs::
245Stop the 'perf stat' session and print count deltas after N milliseconds (minimum: 10 ms).
246This option is not supported with the "-I" option.
247 example: 'perf stat --time 2000 -e cycles -a'
248
249--metric-only::
250Only print computed metrics. Print them in a single line.
251Don't show any raw values. Not supported with --per-thread.
252
253--per-socket::
254Aggregate counts per processor socket for system-wide mode measurements. This
255is a useful mode to detect imbalance between sockets. To enable this mode,
256use --per-socket in addition to -a. (system-wide). The output includes the
257socket number and the number of online processors on that socket. This is
258useful to gauge the amount of aggregation.
259
260--per-die::
261Aggregate counts per processor die for system-wide mode measurements. This
262is a useful mode to detect imbalance between dies. To enable this mode,
263use --per-die in addition to -a. (system-wide). The output includes the
264die number and the number of online processors on that die. This is
265useful to gauge the amount of aggregation.
266
267--per-core::
268Aggregate counts per physical processor for system-wide mode measurements. This
269is a useful mode to detect imbalance between physical cores. To enable this mode,
270use --per-core in addition to -a. (system-wide). The output includes the
271core number and the number of online logical processors on that physical processor.
272
273--per-thread::
274Aggregate counts per monitored threads, when monitoring threads (-t option)
275or processes (-p option).
276
277--per-node::
278Aggregate counts per NUMA nodes for system-wide mode measurements. This
279is a useful mode to detect imbalance between NUMA nodes. To enable this
280mode, use --per-node in addition to -a. (system-wide).
281
282-D msecs::
283--delay msecs::
284After starting the program, wait msecs before measuring (-1: start with events
285disabled). This is useful to filter out the startup phase of the program,
286which is often very different.
287
288-T::
289--transaction::
290
291Print statistics of transactional execution if supported.
292
293--metric-no-group::
294By default, events to compute a metric are placed in weak groups. The
295group tries to enforce scheduling all or none of the events. The
296--metric-no-group option places events outside of groups and may
297increase the chance of the event being scheduled - leading to more
298accuracy. However, as events may not be scheduled together accuracy
299for metrics like instructions per cycle can be lower - as both metrics
300may no longer be being measured at the same time.
301
302--metric-no-merge::
303By default metric events in different weak groups can be shared if one
304group contains all the events needed by another. In such cases one
305group will be eliminated reducing event multiplexing and making it so
306that certain groups of metrics sum to 100%. A downside to sharing a
307group is that the group may require multiplexing and so accuracy for a
308small group that need not have multiplexing is lowered. This option
309forbids the event merging logic from sharing events between groups and
310may be used to increase accuracy in this case.
311
312STAT RECORD
313-----------
314Stores stat data into perf data file.
315
316-o file::
317--output file::
318Output file name.
319
320STAT REPORT
321-----------
322Reads and reports stat data from perf data file.
323
324-i file::
325--input file::
326Input file name.
327
328--per-socket::
329Aggregate counts per processor socket for system-wide mode measurements.
330
331--per-die::
332Aggregate counts per processor die for system-wide mode measurements.
333
334--per-core::
335Aggregate counts per physical processor for system-wide mode measurements.
336
337-M::
338--metrics::
339Print metrics or metricgroups specified in a comma separated list.
340For a group all metrics from the group are added.
341The events from the metrics are automatically measured.
342See perf list output for the possble metrics and metricgroups.
343
344-A::
345--no-aggr::
346Do not aggregate counts across all monitored CPUs.
347
348--topdown::
349Print top down level 1 metrics if supported by the CPU. This allows to
350determine bottle necks in the CPU pipeline for CPU bound workloads,
351by breaking the cycles consumed down into frontend bound, backend bound,
352bad speculation and retiring.
353
354Frontend bound means that the CPU cannot fetch and decode instructions fast
355enough. Backend bound means that computation or memory access is the bottle
356neck. Bad Speculation means that the CPU wasted cycles due to branch
357mispredictions and similar issues. Retiring means that the CPU computed without
358an apparently bottleneck. The bottleneck is only the real bottleneck
359if the workload is actually bound by the CPU and not by something else.
360
361For best results it is usually a good idea to use it with interval
362mode like -I 1000, as the bottleneck of workloads can change often.
363
364The top down metrics are collected per core instead of per
365CPU thread. Per core mode is automatically enabled
366and -a (global monitoring) is needed, requiring root rights or
367perf.perf_event_paranoid=-1.
368
369Topdown uses the full Performance Monitoring Unit, and needs
370disabling of the NMI watchdog (as root):
371echo 0 > /proc/sys/kernel/nmi_watchdog
372for best results. Otherwise the bottlenecks may be inconsistent
373on workload with changing phases.
374
375This enables --metric-only, unless overridden with --no-metric-only.
376
377To interpret the results it is usually needed to know on which
378CPUs the workload runs on. If needed the CPUs can be forced using
379taskset.
380
381--no-merge::
382Do not merge results from same PMUs.
383
384When multiple events are created from a single event specification,
385stat will, by default, aggregate the event counts and show the result
386in a single row. This option disables that behavior and shows
387the individual events and counts.
388
389Multiple events are created from a single event specification when:
3901. Prefix or glob matching is used for the PMU name.
3912. Aliases, which are listed immediately after the Kernel PMU events
392 by perf list, are used.
393
394--smi-cost::
395Measure SMI cost if msr/aperf/ and msr/smi/ events are supported.
396
397During the measurement, the /sys/device/cpu/freeze_on_smi will be set to
398freeze core counters on SMI.
399The aperf counter will not be effected by the setting.
400The cost of SMI can be measured by (aperf - unhalted core cycles).
401
402In practice, the percentages of SMI cycles is very useful for performance
403oriented analysis. --metric_only will be applied by default.
404The output is SMI cycles%, equals to (aperf - unhalted core cycles) / aperf
405
406Users who wants to get the actual value can apply --no-metric-only.
407
408--all-kernel::
409Configure all used events to run in kernel space.
410
411--all-user::
412Configure all used events to run in user space.
413
414--percore-show-thread::
415The event modifier "percore" has supported to sum up the event counts
416for all hardware threads in a core and show the counts per core.
417
418This option with event modifier "percore" enabled also sums up the event
419counts for all hardware threads in a core but show the sum counts per
420hardware thread. This is essentially a replacement for the any bit and
421convenient for post processing.
422
423--summary::
424Print summary for interval mode (-I).
425
426EXAMPLES
427--------
428
429$ perf stat -- make
430
431 Performance counter stats for 'make':
432
433 83723.452481 task-clock:u (msec) # 1.004 CPUs utilized
434 0 context-switches:u # 0.000 K/sec
435 0 cpu-migrations:u # 0.000 K/sec
436 3,228,188 page-faults:u # 0.039 M/sec
437 229,570,665,834 cycles:u # 2.742 GHz
438 313,163,853,778 instructions:u # 1.36 insn per cycle
439 69,704,684,856 branches:u # 832.559 M/sec
440 2,078,861,393 branch-misses:u # 2.98% of all branches
441
442 83.409183620 seconds time elapsed
443
444 74.684747000 seconds user
445 8.739217000 seconds sys
446
447TIMINGS
448-------
449As displayed in the example above we can display 3 types of timings.
450We always display the time the counters were enabled/alive:
451
452 83.409183620 seconds time elapsed
453
454For workload sessions we also display time the workloads spent in
455user/system lands:
456
457 74.684747000 seconds user
458 8.739217000 seconds sys
459
460Those times are the very same as displayed by the 'time' tool.
461
462CSV FORMAT
463----------
464
465With -x, perf stat is able to output a not-quite-CSV format output
466Commas in the output are not put into "". To make it easy to parse
467it is recommended to use a different character like -x \;
468
469The fields are in this order:
470
471 - optional usec time stamp in fractions of second (with -I xxx)
472 - optional CPU, core, or socket identifier
473 - optional number of logical CPUs aggregated
474 - counter value
475 - unit of the counter value or empty
476 - event name
477 - run time of counter
478 - percentage of measurement time the counter was running
479 - optional variance if multiple values are collected with -r
480 - optional metric value
481 - optional unit of metric
482
483Additional metrics may be printed with all earlier fields being empty.
484
485SEE ALSO
486--------
487linkperf:perf-top[1], linkperf:perf-list[1]