1perf-sched(1) 2============= 3 4NAME 5---- 6perf-sched - Tool to trace/measure scheduler properties (latencies) 7 8SYNOPSIS 9-------- 10[verse] 11'perf sched' {record|latency|map|replay|script|timehist} 12 13DESCRIPTION 14----------- 15There are several variants of 'perf sched': 16 17 'perf sched record <command>' to record the scheduling events 18 of an arbitrary workload. 19 20 'perf sched latency' to report the per task scheduling latencies 21 and other scheduling properties of the workload. 22 23 Example usage: 24 perf sched record -- sleep 1 25 perf sched latency 26 27 ------------------------------------------------------------------------------------------------------------------------------------------- 28 Task | Runtime ms | Count | Avg delay ms | Max delay ms | Max delay start | Max delay end | 29 ------------------------------------------------------------------------------------------------------------------------------------------- 30 perf:(2) | 2.804 ms | 66 | avg: 0.524 ms | max: 1.069 ms | max start: 254752.314960 s | max end: 254752.316029 s 31 NetworkManager:1343 | 0.372 ms | 13 | avg: 0.008 ms | max: 0.013 ms | max start: 254751.551153 s | max end: 254751.551166 s 32 kworker/1:2-xfs:4649 | 0.012 ms | 1 | avg: 0.008 ms | max: 0.008 ms | max start: 254751.519807 s | max end: 254751.519815 s 33 kworker/3:1-xfs:388 | 0.011 ms | 1 | avg: 0.006 ms | max: 0.006 ms | max start: 254751.519809 s | max end: 254751.519815 s 34 sleep:147736 | 0.938 ms | 3 | avg: 0.006 ms | max: 0.007 ms | max start: 254751.313817 s | max end: 254751.313824 s 35 36 It shows Runtime(time that a task spent actually running on the CPU), 37 Count(number of times a delay was calculated) and delay(time that a 38 task was ready to run but was kept waiting). 39 40 Tasks with the same command name are merged and the merge count is 41 given within (), However if -p option is used, pid is mentioned. 42 43 'perf sched script' to see a detailed trace of the workload that 44 was recorded (aliased to 'perf script' for now). 45 46 'perf sched replay' to simulate the workload that was recorded 47 via perf sched record. (this is done by starting up mockup threads 48 that mimic the workload based on the events in the trace. These 49 threads can then replay the timings (CPU runtime and sleep patterns) 50 of the workload as it occurred when it was recorded - and can repeat 51 it a number of times, measuring its performance.) 52 53 'perf sched map' to print a textual context-switching outline of 54 workload captured via perf sched record. Columns stand for 55 individual CPUs, and the two-letter shortcuts stand for tasks that 56 are running on a CPU. A '*' denotes the CPU that had the event, and 57 a dot signals an idle CPU. 58 59 'perf sched timehist' provides an analysis of scheduling events. 60 61 Example usage: 62 perf sched record -- sleep 1 63 perf sched timehist 64 65 By default it shows the individual schedule events, including the wait 66 time (time between sched-out and next sched-in events for the task), the 67 task scheduling delay (time between wakeup and actually running) and run 68 time for the task: 69 70 time cpu task name wait time sch delay run time 71 [tid/pid] (msec) (msec) (msec) 72 -------------- ------ -------------------- --------- --------- --------- 73 79371.874569 [0011] gcc[31949] 0.014 0.000 1.148 74 79371.874591 [0010] gcc[31951] 0.000 0.000 0.024 75 79371.874603 [0010] migration/10[59] 3.350 0.004 0.011 76 79371.874604 [0011] <idle> 1.148 0.000 0.035 77 79371.874723 [0005] <idle> 0.016 0.000 1.383 78 79371.874746 [0005] gcc[31949] 0.153 0.078 0.022 79 ... 80 81 Times are in msec.usec. 82 83OPTIONS 84------- 85-i:: 86--input=<file>:: 87 Input file name. (default: perf.data unless stdin is a fifo) 88 89-v:: 90--verbose:: 91 Be more verbose. (show symbol address, etc) 92 93-D:: 94--dump-raw-trace=:: 95 Display verbose dump of the sched data. 96 97-f:: 98--force:: 99 Don't complain, do it. 100 101OPTIONS for 'perf sched latency' 102------------------------------- 103 104-C:: 105--CPU <n>:: 106 CPU to profile on. 107 108-p:: 109--pids:: 110 latency stats per pid instead of per command name. 111 112-s:: 113--sort <key[,key2...]>:: 114 sort by key(s): runtime, switch, avg, max 115 by default it's sorted by "avg ,max ,switch ,runtime". 116 117OPTIONS for 'perf sched map' 118---------------------------- 119 120--compact:: 121 Show only CPUs with activity. Helps visualizing on high core 122 count systems. 123 124--cpus:: 125 Show just entries with activities for the given CPUs. 126 127--color-cpus:: 128 Highlight the given cpus. 129 130--color-pids:: 131 Highlight the given pids. 132 133OPTIONS for 'perf sched timehist' 134--------------------------------- 135-k:: 136--vmlinux=<file>:: 137 vmlinux pathname 138 139--kallsyms=<file>:: 140 kallsyms pathname 141 142-g:: 143--call-graph:: 144 Display call chains if present (default on). 145 146--max-stack:: 147 Maximum number of functions to display in backtrace, default 5. 148 149-C=:: 150--cpu=:: 151 Only show events for the given CPU(s) (comma separated list). 152 153-p=:: 154--pid=:: 155 Only show events for given process ID (comma separated list). 156 157-t=:: 158--tid=:: 159 Only show events for given thread ID (comma separated list). 160 161-s:: 162--summary:: 163 Show only a summary of scheduling by thread with min, max, and average 164 run times (in sec) and relative stddev. 165 166-S:: 167--with-summary:: 168 Show all scheduling events followed by a summary by thread with min, 169 max, and average run times (in sec) and relative stddev. 170 171--symfs=<directory>:: 172 Look for files with symbols relative to this directory. 173 174-V:: 175--cpu-visual:: 176 Show visual aid for sched switches by CPU: 'i' marks idle time, 177 's' are scheduler events. 178 179-w:: 180--wakeups:: 181 Show wakeup events. 182 183-M:: 184--migrations:: 185 Show migration events. 186 187-n:: 188--next:: 189 Show next task. 190 191-I:: 192--idle-hist:: 193 Show idle-related events only. 194 195--time:: 196 Only analyze samples within given time window: <start>,<stop>. Times 197 have the format seconds.microseconds. If start is not given (i.e., time 198 string is ',x.y') then analysis starts at the beginning of the file. If 199 stop time is not given (i.e, time string is 'x.y,') then analysis goes 200 to end of file. 201 202--state:: 203 Show task state when it switched out. 204 205SEE ALSO 206-------- 207linkperf:perf-record[1] 208