tools/perf/Documentation/perf-stat.txt

   1 perf-stat(1)
   2 ============
   3
   4 NAME
   5 ----
   6 perf-stat - Run a command and gather performance counter statistics
   7
   8 SYNOPSIS
   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
  16 DESCRIPTION
  17 -----------
  18 This command runs a command and gathers performance counter statistics
  19 from it.
  20
  21
  22 OPTIONS
  23 -------
  24 <command>...::
  25         Any command you can specify in a shell.
  26
  27 record::
  28         See STAT RECORD.
  29
  30 report::
  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         - a symbolically formed event like 'pmu/config=M,config1=N,config2=K/'
  47           where M, N, K are numbers (in decimal, hex, octal format).
  48           Acceptable values for each of 'config', 'config1' and 'config2'
  49           parameters are defined by corresponding entries in
  50           /sys/bus/event_source/devices/<pmu>/format/*
  51
  52         Note that the last two syntaxes support prefix and glob matching in
  53         the PMU name to simplify creation of events accross multiple instances
  54         of the same type of PMU in large systems (e.g. memory controller PMUs).
  55         Multiple PMU instances are typical for uncore PMUs, so the prefix
  56         'uncore_' is also ignored when performing this match.
  57
  58
  59 -i::
  60 --no-inherit::
  61         child tasks do not inherit counters
  62 -p::
  63 --pid=<pid>::
  64         stat events on existing process id (comma separated list)
  65
  66 -t::
  67 --tid=<tid>::
  68         stat events on existing thread id (comma separated list)
  69
  70
  71 -a::
  72 --all-cpus::
  73         system-wide collection from all CPUs (default if no target is specified)
  74
  75 -c::
  76 --scale::
  77         scale/normalize counter values
  78
  79 -d::
  80 --detailed::
  81         print more detailed statistics, can be specified up to 3 times
  82
  83            -d:          detailed events, L1 and LLC data cache
  84         -d -d:     more detailed events, dTLB and iTLB events
  85      -d -d -d:     very detailed events, adding prefetch events
  86
  87 -r::
  88 --repeat=<n>::
  89         repeat command and print average + stddev (max: 100). 0 means forever.
  90
  91 -B::
  92 --big-num::
  93         print large numbers with thousands' separators according to locale
  94
  95 -C::
  96 --cpu=::
  97 Count only on the list of CPUs provided. Multiple CPUs can be provided as a
  98 comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-2.
  99 In per-thread mode, this option is ignored. The -a option is still necessary
 100 to activate system-wide monitoring. Default is to count on all CPUs.
 101
 102 -A::
 103 --no-aggr::
 104 Do not aggregate counts across all monitored CPUs.
 105
 106 -n::
 107 --null::
 108         null run - don't start any counters
 109
 110 -v::
 111 --verbose::
 112         be more verbose (show counter open errors, etc)
 113
 114 -x SEP::
 115 --field-separator SEP::
 116 print counts using a CSV-style output to make it easy to import directly into
 117 spreadsheets. Columns are separated by the string specified in SEP.
 118
 119 -G name::
 120 --cgroup name::
 121 monitor only in the container (cgroup) called "name". This option is available only
 122 in per-cpu mode. The cgroup filesystem must be mounted. All threads belonging to
 123 container "name" are monitored when they run on the monitored CPUs. Multiple cgroups
 124 can be provided. Each cgroup is applied to the corresponding event, i.e., first cgroup
 125 to first event, second cgroup to second event and so on. It is possible to provide
 126 an empty cgroup (monitor all the time) using, e.g., -G foo,,bar. Cgroups must have
 127 corresponding events, i.e., they always refer to events defined earlier on the command
 128 line. If the user wants to track multiple events for a specific cgroup, the user can
 129 use '-e e1 -e e2 -G foo,foo' or just use '-e e1 -e e2 -G foo'.
 130
 131 If wanting to monitor, say, 'cycles' for a cgroup and also for system wide, this
 132 command line can be used: 'perf stat -e cycles -G cgroup_name -a -e cycles'.
 133
 134 -o file::
 135 --output file::
 136 Print the output into the designated file.
 137
 138 --append::
 139 Append to the output file designated with the -o option. Ignored if -o is not specified.
 140
 141 --log-fd::
 142
 143 Log output to fd, instead of stderr.  Complementary to --output, and mutually exclusive
 144 with it.  --append may be used here.  Examples:
 145      3>results  perf stat --log-fd 3          -- $cmd
 146      3>>results perf stat --log-fd 3 --append -- $cmd
 147
 148 --pre::
 149 --post::
 150         Pre and post measurement hooks, e.g.:
 151
 152 perf stat --repeat 10 --null --sync --pre 'make -s O=defconfig-build/clean' -- make -s -j64 O=defconfig-build/ bzImage
 153
 154 -I msecs::
 155 --interval-print msecs::
 156 Print count deltas every N milliseconds (minimum: 10ms)
 157 The overhead percentage could be high in some cases, for instance with small, sub 100ms intervals.  Use with caution.
 158         example: 'perf stat -I 1000 -e cycles -a sleep 5'
 159
 160 --interval-count times::
 161 Print count deltas for fixed number of times.
 162 This option should be used together with "-I" option.
 163         example: 'perf stat -I 1000 --interval-count 2 -e cycles -a'
 164
 165 --timeout msecs::
 166 Stop the 'perf stat' session and print count deltas after N milliseconds (minimum: 10 ms).
 167 This option is not supported with the "-I" option.
 168         example: 'perf stat --time 2000 -e cycles -a'
 169
 170 --metric-only::
 171 Only print computed metrics. Print them in a single line.
 172 Don't show any raw values. Not supported with --per-thread.
 173
 174 --per-socket::
 175 Aggregate counts per processor socket for system-wide mode measurements.  This
 176 is a useful mode to detect imbalance between sockets.  To enable this mode,
 177 use --per-socket in addition to -a. (system-wide).  The output includes the
 178 socket number and the number of online processors on that socket. This is
 179 useful to gauge the amount of aggregation.
 180
 181 --per-core::
 182 Aggregate counts per physical processor for system-wide mode measurements.  This
 183 is a useful mode to detect imbalance between physical cores.  To enable this mode,
 184 use --per-core in addition to -a. (system-wide).  The output includes the
 185 core number and the number of online logical processors on that physical processor.
 186
 187 --per-thread::
 188 Aggregate counts per monitored threads, when monitoring threads (-t option)
 189 or processes (-p option).
 190
 191 -D msecs::
 192 --delay msecs::
 193 After starting the program, wait msecs before measuring. This is useful to
 194 filter out the startup phase of the program, which is often very different.
 195
 196 -T::
 197 --transaction::
 198
 199 Print statistics of transactional execution if supported.
 200
 201 STAT RECORD
 202 -----------
 203 Stores stat data into perf data file.
 204
 205 -o file::
 206 --output file::
 207 Output file name.
 208
 209 STAT REPORT
 210 -----------
 211 Reads and reports stat data from perf data file.
 212
 213 -i file::
 214 --input file::
 215 Input file name.
 216
 217 --per-socket::
 218 Aggregate counts per processor socket for system-wide mode measurements.
 219
 220 --per-core::
 221 Aggregate counts per physical processor for system-wide mode measurements.
 222
 223 -M::
 224 --metrics::
 225 Print metrics or metricgroups specified in a comma separated list.
 226 For a group all metrics from the group are added.
 227 The events from the metrics are automatically measured.
 228 See perf list output for the possble metrics and metricgroups.
 229
 230 -A::
 231 --no-aggr::
 232 Do not aggregate counts across all monitored CPUs.
 233
 234 --topdown::
 235 Print top down level 1 metrics if supported by the CPU. This allows to
 236 determine bottle necks in the CPU pipeline for CPU bound workloads,
 237 by breaking the cycles consumed down into frontend bound, backend bound,
 238 bad speculation and retiring.
 239
 240 Frontend bound means that the CPU cannot fetch and decode instructions fast
 241 enough. Backend bound means that computation or memory access is the bottle
 242 neck. Bad Speculation means that the CPU wasted cycles due to branch
 243 mispredictions and similar issues. Retiring means that the CPU computed without
 244 an apparently bottleneck. The bottleneck is only the real bottleneck
 245 if the workload is actually bound by the CPU and not by something else.
 246
 247 For best results it is usually a good idea to use it with interval
 248 mode like -I 1000, as the bottleneck of workloads can change often.
 249
 250 The top down metrics are collected per core instead of per
 251 CPU thread. Per core mode is automatically enabled
 252 and -a (global monitoring) is needed, requiring root rights or
 253 perf.perf_event_paranoid=-1.
 254
 255 Topdown uses the full Performance Monitoring Unit, and needs
 256 disabling of the NMI watchdog (as root):
 257 echo 0 > /proc/sys/kernel/nmi_watchdog
 258 for best results. Otherwise the bottlenecks may be inconsistent
 259 on workload with changing phases.
 260
 261 This enables --metric-only, unless overriden with --no-metric-only.
 262
 263 To interpret the results it is usually needed to know on which
 264 CPUs the workload runs on. If needed the CPUs can be forced using
 265 taskset.
 266
 267 --no-merge::
 268 Do not merge results from same PMUs.
 269
 270 When multiple events are created from a single event specification,
 271 stat will, by default, aggregate the event counts and show the result
 272 in a single row. This option disables that behavior and shows
 273 the individual events and counts.
 274
 275 Multiple events are created from a single event specification when:
 276 1. Prefix or glob matching is used for the PMU name.
 277 2. Aliases, which are listed immediately after the Kernel PMU events
 278    by perf list, are used.
 279
 280 --smi-cost::
 281 Measure SMI cost if msr/aperf/ and msr/smi/ events are supported.
 282
 283 During the measurement, the /sys/device/cpu/freeze_on_smi will be set to
 284 freeze core counters on SMI.
 285 The aperf counter will not be effected by the setting.
 286 The cost of SMI can be measured by (aperf - unhalted core cycles).
 287
 288 In practice, the percentages of SMI cycles is very useful for performance
 289 oriented analysis. --metric_only will be applied by default.
 290 The output is SMI cycles%, equals to (aperf - unhalted core cycles) / aperf
 291
 292 Users who wants to get the actual value can apply --no-metric-only.
 293
 294 EXAMPLES
 295 --------
 296
 297 $ perf stat -- make -j
 298
 299  Performance counter stats for 'make -j':
 300
 301     8117.370256  task clock ticks     #      11.281 CPU utilization factor
 302             678  context switches     #       0.000 M/sec
 303             133  CPU migrations       #       0.000 M/sec
 304          235724  pagefaults           #       0.029 M/sec
 305     24821162526  CPU cycles           #    3057.784 M/sec
 306     18687303457  instructions         #    2302.138 M/sec
 307       172158895  cache references     #      21.209 M/sec
 308        27075259  cache misses         #       3.335 M/sec
 309
 310  Wall-clock time elapsed:   719.554352 msecs
 311
 312 CSV FORMAT
 313 ----------
 314
 315 With -x, perf stat is able to output a not-quite-CSV format output
 316 Commas in the output are not put into "". To make it easy to parse
 317 it is recommended to use a different character like -x \;
 318
 319 The fields are in this order:
 320
 321         - optional usec time stamp in fractions of second (with -I xxx)
 322         - optional CPU, core, or socket identifier
 323         - optional number of logical CPUs aggregated
 324         - counter value
 325         - unit of the counter value or empty
 326         - event name
 327         - run time of counter
 328         - percentage of measurement time the counter was running
 329         - optional variance if multiple values are collected with -r
 330         - optional metric value
 331         - optional unit of metric
 332
 333 Additional metrics may be printed with all earlier fields being empty.
 334
 335 SEE ALSO
 336 --------
 337 linkperf:perf-top[1], linkperf:perf-list[1]