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 14DESCRIPTION 15----------- 16This command runs a command and gathers performance counter statistics 17from it. 18 19 20OPTIONS 21------- 22<command>...:: 23 Any command you can specify in a shell. 24 25 26-e:: 27--event=:: 28 Select the PMU event. Selection can be a symbolic event name 29 (use 'perf list' to list all events) or a raw PMU 30 event (eventsel+umask) in the form of rNNN where NNN is a 31 hexadecimal event descriptor. 32 33-i:: 34--no-inherit:: 35 child tasks do not inherit counters 36-p:: 37--pid=<pid>:: 38 stat events on existing process id (comma separated list) 39 40-t:: 41--tid=<tid>:: 42 stat events on existing thread id (comma separated list) 43 44 45-a:: 46--all-cpus:: 47 system-wide collection from all CPUs 48 49-c:: 50--scale:: 51 scale/normalize counter values 52 53-d:: 54--detailed:: 55 print more detailed statistics, can be specified up to 3 times 56 57 -d: detailed events, L1 and LLC data cache 58 -d -d: more detailed events, dTLB and iTLB events 59 -d -d -d: very detailed events, adding prefetch events 60 61-r:: 62--repeat=<n>:: 63 repeat command and print average + stddev (max: 100). 0 means forever. 64 65-B:: 66--big-num:: 67 print large numbers with thousands' separators according to locale 68 69-C:: 70--cpu=:: 71Count only on the list of CPUs provided. Multiple CPUs can be provided as a 72comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-2. 73In per-thread mode, this option is ignored. The -a option is still necessary 74to activate system-wide monitoring. Default is to count on all CPUs. 75 76-A:: 77--no-aggr:: 78Do not aggregate counts across all monitored CPUs in system-wide mode (-a). 79This option is only valid in system-wide mode. 80 81-n:: 82--null:: 83 null run - don't start any counters 84 85-v:: 86--verbose:: 87 be more verbose (show counter open errors, etc) 88 89-x SEP:: 90--field-separator SEP:: 91print counts using a CSV-style output to make it easy to import directly into 92spreadsheets. Columns are separated by the string specified in SEP. 93 94-G name:: 95--cgroup name:: 96monitor only in the container (cgroup) called "name". This option is available only 97in per-cpu mode. The cgroup filesystem must be mounted. All threads belonging to 98container "name" are monitored when they run on the monitored CPUs. Multiple cgroups 99can be provided. Each cgroup is applied to the corresponding event, i.e., first cgroup 100to first event, second cgroup to second event and so on. It is possible to provide 101an empty cgroup (monitor all the time) using, e.g., -G foo,,bar. Cgroups must have 102corresponding events, i.e., they always refer to events defined earlier on the command 103line. 104 105-o file:: 106--output file:: 107Print the output into the designated file. 108 109--append:: 110Append to the output file designated with the -o option. Ignored if -o is not specified. 111 112--log-fd:: 113 114Log output to fd, instead of stderr. Complementary to --output, and mutually exclusive 115with it. --append may be used here. Examples: 116 3>results perf stat --log-fd 3 -- $cmd 117 3>>results perf stat --log-fd 3 --append -- $cmd 118 119--pre:: 120--post:: 121 Pre and post measurement hooks, e.g.: 122 123perf stat --repeat 10 --null --sync --pre 'make -s O=defconfig-build/clean' -- make -s -j64 O=defconfig-build/ bzImage 124 125-I msecs:: 126--interval-print msecs:: 127 Print count deltas every N milliseconds (minimum: 100ms) 128 example: perf stat -I 1000 -e cycles -a sleep 5 129 130--per-socket:: 131Aggregate counts per processor socket for system-wide mode measurements. This 132is a useful mode to detect imbalance between sockets. To enable this mode, 133use --per-socket in addition to -a. (system-wide). The output includes the 134socket number and the number of online processors on that socket. This is 135useful to gauge the amount of aggregation. 136 137--per-core:: 138Aggregate counts per physical processor for system-wide mode measurements. This 139is a useful mode to detect imbalance between physical cores. To enable this mode, 140use --per-core in addition to -a. (system-wide). The output includes the 141core number and the number of online logical processors on that physical processor. 142 143-D msecs:: 144--delay msecs:: 145After starting the program, wait msecs before measuring. This is useful to 146filter out the startup phase of the program, which is often very different. 147 148-T:: 149--transaction:: 150 151Print statistics of transactional execution if supported. 152 153EXAMPLES 154-------- 155 156$ perf stat -- make -j 157 158 Performance counter stats for 'make -j': 159 160 8117.370256 task clock ticks # 11.281 CPU utilization factor 161 678 context switches # 0.000 M/sec 162 133 CPU migrations # 0.000 M/sec 163 235724 pagefaults # 0.029 M/sec 164 24821162526 CPU cycles # 3057.784 M/sec 165 18687303457 instructions # 2302.138 M/sec 166 172158895 cache references # 21.209 M/sec 167 27075259 cache misses # 3.335 M/sec 168 169 Wall-clock time elapsed: 719.554352 msecs 170 171SEE ALSO 172-------- 173linkperf:perf-top[1], linkperf:perf-list[1] 174