1 perf-sched(1) 1 perf-sched(1)
2 ============= 2 =============
3 3
4 NAME 4 NAME
5 ---- 5 ----
6 perf-sched - Tool to trace/measure scheduler p 6 perf-sched - Tool to trace/measure scheduler properties (latencies)
7 7
8 SYNOPSIS 8 SYNOPSIS
9 -------- 9 --------
10 [verse] 10 [verse]
11 'perf sched' {record|latency|map|replay|script 11 'perf sched' {record|latency|map|replay|script|timehist}
12 12
13 DESCRIPTION 13 DESCRIPTION
14 ----------- 14 -----------
15 There are several variants of 'perf sched': 15 There are several variants of 'perf sched':
16 16
17 'perf sched record <command>' to record the 17 'perf sched record <command>' to record the scheduling events
18 of an arbitrary workload. 18 of an arbitrary workload.
19 19
20 'perf sched latency' to report the per task 20 'perf sched latency' to report the per task scheduling latencies
21 and other scheduling properties of the workl 21 and other scheduling properties of the workload.
22 22
23 Example usage: 23 Example usage:
24 perf sched record -- sleep 1 24 perf sched record -- sleep 1
25 perf sched latency 25 perf sched latency
26 26
27 -------------------------------------------- 27 -------------------------------------------------------------------------------------------------------------------------------------------
28 Task | Runtime ms | Cou 28 Task | Runtime ms | Count | Avg delay ms | Max delay ms | Max delay start | Max delay end |
29 -------------------------------------------- 29 -------------------------------------------------------------------------------------------------------------------------------------------
30 perf:(2) | 2.804 ms | 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 | 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 | 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 | 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 | 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 35
36 It shows Runtime(time that a task spent actu 36 It shows Runtime(time that a task spent actually running on the CPU),
37 Count(number of times a delay was calculated 37 Count(number of times a delay was calculated) and delay(time that a
38 task was ready to run but was kept waiting). 38 task was ready to run but was kept waiting).
39 39
40 Tasks with the same command name are merged 40 Tasks with the same command name are merged and the merge count is
41 given within (), However if -p option is use 41 given within (), However if -p option is used, pid is mentioned.
42 42
43 'perf sched script' to see a detailed trace 43 'perf sched script' to see a detailed trace of the workload that
44 was recorded (aliased to 'perf script' for 44 was recorded (aliased to 'perf script' for now).
45 45
46 'perf sched replay' to simulate the workload 46 'perf sched replay' to simulate the workload that was recorded
47 via perf sched record. (this is done by star 47 via perf sched record. (this is done by starting up mockup threads
48 that mimic the workload based on the events 48 that mimic the workload based on the events in the trace. These
49 threads can then replay the timings (CPU run 49 threads can then replay the timings (CPU runtime and sleep patterns)
50 of the workload as it occurred when it was r 50 of the workload as it occurred when it was recorded - and can repeat
51 it a number of times, measuring its performa 51 it a number of times, measuring its performance.)
52 52
53 'perf sched map' to print a textual context- 53 'perf sched map' to print a textual context-switching outline of
54 workload captured via perf sched record. Co 54 workload captured via perf sched record. Columns stand for
55 individual CPUs, and the two-letter shortcut 55 individual CPUs, and the two-letter shortcuts stand for tasks that
56 are running on a CPU. A '*' denotes the CPU 56 are running on a CPU. A '*' denotes the CPU that had the event, and
57 a dot signals an idle CPU. 57 a dot signals an idle CPU.
58 58
59 'perf sched timehist' provides an analysis o 59 'perf sched timehist' provides an analysis of scheduling events.
60 60
61 Example usage: 61 Example usage:
62 perf sched record -- sleep 1 62 perf sched record -- sleep 1
63 perf sched timehist 63 perf sched timehist
64 64
65 By default it shows the individual schedule 65 By default it shows the individual schedule events, including the wait
66 time (time between sched-out and next sched 66 time (time between sched-out and next sched-in events for the task), the
67 task scheduling delay (time between runnabl 67 task scheduling delay (time between runnable and actually running) and
68 run time for the task: 68 run time for the task:
69 69
70 time cpu task name 70 time cpu task name wait time sch delay run time
71 [tid/pid] 71 [tid/pid] (msec) (msec) (msec)
72 -------------- ------ ----------------- 72 -------------- ------ -------------------- --------- --------- ---------
73 79371.874569 [0011] gcc[31949] 73 79371.874569 [0011] gcc[31949] 0.014 0.000 1.148
74 79371.874591 [0010] gcc[31951] 74 79371.874591 [0010] gcc[31951] 0.000 0.000 0.024
75 79371.874603 [0010] migration/10[59] 75 79371.874603 [0010] migration/10[59] 3.350 0.004 0.011
76 79371.874604 [0011] <idle> 76 79371.874604 [0011] <idle> 1.148 0.000 0.035
77 79371.874723 [0005] <idle> 77 79371.874723 [0005] <idle> 0.016 0.000 1.383
78 79371.874746 [0005] gcc[31949] 78 79371.874746 [0005] gcc[31949] 0.153 0.078 0.022
79 ... 79 ...
80 80
81 Times are in msec.usec. 81 Times are in msec.usec.
82 82
83 OPTIONS 83 OPTIONS
84 ------- 84 -------
85 -i:: 85 -i::
86 --input=<file>:: 86 --input=<file>::
87 Input file name. (default: perf.data u 87 Input file name. (default: perf.data unless stdin is a fifo)
88 88
89 -v:: 89 -v::
90 --verbose:: 90 --verbose::
91 Be more verbose. (show symbol address, 91 Be more verbose. (show symbol address, etc)
92 92
93 -D:: 93 -D::
94 --dump-raw-trace=:: 94 --dump-raw-trace=::
95 Display verbose dump of the sched data 95 Display verbose dump of the sched data.
96 96
97 -f:: 97 -f::
98 --force:: 98 --force::
99 Don't complain, do it. 99 Don't complain, do it.
100 100
101 OPTIONS for 'perf sched latency' 101 OPTIONS for 'perf sched latency'
102 ------------------------------- 102 -------------------------------
103 103
104 -C:: 104 -C::
105 --CPU <n>:: 105 --CPU <n>::
106 CPU to profile on. 106 CPU to profile on.
107 107
108 -p:: 108 -p::
109 --pids:: 109 --pids::
110 latency stats per pid instead of per c 110 latency stats per pid instead of per command name.
111 111
112 -s:: 112 -s::
113 --sort <key[,key2...]>:: 113 --sort <key[,key2...]>::
114 sort by key(s): runtime, switch, avg, 114 sort by key(s): runtime, switch, avg, max
115 by default it's sorted by "avg ,max ,s 115 by default it's sorted by "avg ,max ,switch ,runtime".
116 116
117 OPTIONS for 'perf sched map' 117 OPTIONS for 'perf sched map'
118 ---------------------------- 118 ----------------------------
119 119
120 --compact:: 120 --compact::
121 Show only CPUs with activity. Helps vi 121 Show only CPUs with activity. Helps visualizing on high core
122 count systems. 122 count systems.
123 123
124 --cpus:: 124 --cpus::
125 Show just entries with activities for 125 Show just entries with activities for the given CPUs.
126 126
127 --color-cpus:: 127 --color-cpus::
128 Highlight the given cpus. 128 Highlight the given cpus.
129 129
130 --color-pids:: 130 --color-pids::
131 Highlight the given pids. 131 Highlight the given pids.
132 132
133 --task-name <task>:: 133 --task-name <task>::
134 Map output only for the given task nam 134 Map output only for the given task name(s). Separate the
135 task names with a comma (without white 135 task names with a comma (without whitespace). The sched-out
136 time is printed and is represented by 136 time is printed and is represented by '*-' for the given
137 task name(s). 137 task name(s).
138 ('-' indicates other tasks while '.' i 138 ('-' indicates other tasks while '.' is idle).
139 139
140 --fuzzy-name:: 140 --fuzzy-name::
141 Given task name(s) can be partially ma 141 Given task name(s) can be partially matched (fuzzy matching).
142 142
143 OPTIONS for 'perf sched timehist' 143 OPTIONS for 'perf sched timehist'
144 --------------------------------- 144 ---------------------------------
145 -k:: 145 -k::
146 --vmlinux=<file>:: 146 --vmlinux=<file>::
147 vmlinux pathname 147 vmlinux pathname
148 148
149 --kallsyms=<file>:: 149 --kallsyms=<file>::
150 kallsyms pathname 150 kallsyms pathname
151 151
152 -g:: 152 -g::
153 --call-graph:: 153 --call-graph::
154 Display call chains if present (defaul 154 Display call chains if present (default on).
155 155
156 --max-stack:: 156 --max-stack::
157 Maximum number of functions to display 157 Maximum number of functions to display in backtrace, default 5.
158 158
159 -C=:: 159 -C=::
160 --cpu=:: 160 --cpu=::
161 Only show events for the given CPU(s) 161 Only show events for the given CPU(s) (comma separated list).
162 162
163 -p=:: 163 -p=::
164 --pid=:: 164 --pid=::
165 Only show events for given process ID 165 Only show events for given process ID (comma separated list).
166 166
167 -t=:: 167 -t=::
168 --tid=:: 168 --tid=::
169 Only show events for given thread ID ( 169 Only show events for given thread ID (comma separated list).
170 170
171 -s:: 171 -s::
172 --summary:: 172 --summary::
173 Show only a summary of scheduling by threa 173 Show only a summary of scheduling by thread with min, max, and average
174 run times (in sec) and relative stddev. 174 run times (in sec) and relative stddev.
175 175
176 -S:: 176 -S::
177 --with-summary:: 177 --with-summary::
178 Show all scheduling events followed by a s 178 Show all scheduling events followed by a summary by thread with min,
179 max, and average run times (in sec) and re 179 max, and average run times (in sec) and relative stddev.
180 180
181 --symfs=<directory>:: 181 --symfs=<directory>::
182 Look for files with symbols relative to th 182 Look for files with symbols relative to this directory.
183 183
184 -V:: 184 -V::
185 --cpu-visual:: 185 --cpu-visual::
186 Show visual aid for sched switches by 186 Show visual aid for sched switches by CPU: 'i' marks idle time,
187 's' are scheduler events. 187 's' are scheduler events.
188 188
189 -w:: 189 -w::
190 --wakeups:: 190 --wakeups::
191 Show wakeup events. 191 Show wakeup events.
192 192
193 -M:: 193 -M::
194 --migrations:: 194 --migrations::
195 Show migration events. 195 Show migration events.
196 196
197 -n:: 197 -n::
198 --next:: 198 --next::
199 Show next task. 199 Show next task.
200 200
201 -I:: 201 -I::
202 --idle-hist:: 202 --idle-hist::
203 Show idle-related events only. 203 Show idle-related events only.
204 204
205 --time:: 205 --time::
206 Only analyze samples within given time 206 Only analyze samples within given time window: <start>,<stop>. Times
207 have the format seconds.microseconds. 207 have the format seconds.microseconds. If start is not given (i.e., time
208 string is ',x.y') then analysis starts 208 string is ',x.y') then analysis starts at the beginning of the file. If
209 stop time is not given (i.e, time stri 209 stop time is not given (i.e, time string is 'x.y,') then analysis goes
210 to end of file. 210 to end of file.
211 211
212 --state:: 212 --state::
213 Show task state when it switched out. 213 Show task state when it switched out.
214 214
215 --show-prio:: 215 --show-prio::
216 Show task priority. 216 Show task priority.
217 217
218 --prio:: 218 --prio::
219 Only show events for given task priori 219 Only show events for given task priority(ies). Multiple priorities can be
220 provided as a comma-separated list wit 220 provided as a comma-separated list with no spaces: 0,120. Ranges of
221 priorities are specified with -: 120-1 221 priorities are specified with -: 120-129. A combination of both can also be
222 provided: 0,120-129. 222 provided: 0,120-129.
223 223
224 OPTIONS for 'perf sched replay' 224 OPTIONS for 'perf sched replay'
225 ------------------------------ 225 ------------------------------
226 226
227 -r:: 227 -r::
228 --repeat <n>:: 228 --repeat <n>::
229 repeat the workload n times (0: infini 229 repeat the workload n times (0: infinite). Default is 10.
230 230
231 SEE ALSO 231 SEE ALSO
232 -------- 232 --------
233 linkperf:perf-record[1] 233 linkperf:perf-record[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.