1 perf-top(1) 1 perf-top(1)
2 =========== 2 ===========
3 3
4 NAME 4 NAME
5 ---- 5 ----
6 perf-top - System profiling tool. 6 perf-top - System profiling tool.
7 7
8 SYNOPSIS 8 SYNOPSIS
9 -------- 9 --------
10 [verse] 10 [verse]
11 'perf top' [-e <EVENT> | --event=EVENT] [<opti 11 'perf top' [-e <EVENT> | --event=EVENT] [<options>]
12 12
13 DESCRIPTION 13 DESCRIPTION
14 ----------- 14 -----------
15 This command generates and displays a performa 15 This command generates and displays a performance counter profile in real time.
16 16
17 17
18 OPTIONS 18 OPTIONS
19 ------- 19 -------
20 -a:: 20 -a::
21 --all-cpus:: 21 --all-cpus::
22 System-wide collection. (default) 22 System-wide collection. (default)
23 23
24 -c <count>:: 24 -c <count>::
25 --count=<count>:: 25 --count=<count>::
26 Event period to sample. 26 Event period to sample.
27 27
28 -C <cpu-list>:: 28 -C <cpu-list>::
29 --cpu=<cpu>:: 29 --cpu=<cpu>::
30 Monitor only on the list of CPUs provided. Mul 30 Monitor only on the list of CPUs provided. Multiple CPUs can be provided as a
31 comma-separated list with no space: 0,1. Range 31 comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-2.
32 Default is to monitor all CPUS. 32 Default is to monitor all CPUS.
33 33
34 -d <seconds>:: 34 -d <seconds>::
35 --delay=<seconds>:: 35 --delay=<seconds>::
36 Number of seconds to delay between ref 36 Number of seconds to delay between refreshes.
37 37
38 -e <event>:: 38 -e <event>::
39 --event=<event>:: 39 --event=<event>::
40 Select the PMU event. Selection can be 40 Select the PMU event. Selection can be a symbolic event name
41 (use 'perf list' to list all events) o !! 41 (use 'perf list' to list all events) or a raw PMU
42 of rN where N is a hexadecimal value t !! 42 event (eventsel+umask) in the form of rNNN where NNN is a
43 encoding with the layout of the event !! 43 hexadecimal event descriptor.
44 by entries in /sys/bus/event_source/de <<
45 <<
46 --filter=<filter>:: <<
47 Event filter. This option should foll <<
48 syntax see linkperf:perf-record[1]. <<
49 44
50 -E <entries>:: 45 -E <entries>::
51 --entries=<entries>:: 46 --entries=<entries>::
52 Display this many functions. 47 Display this many functions.
53 48
54 -f <count>:: 49 -f <count>::
55 --count-filter=<count>:: 50 --count-filter=<count>::
56 Only display functions with more event 51 Only display functions with more events than this.
57 52
>> 53 --group::
>> 54 Put the counters into a counter group.
>> 55
58 --group-sort-idx:: 56 --group-sort-idx::
59 Sort the output by the event at the in 57 Sort the output by the event at the index n in group. If n is invalid,
60 sort by the first event. It can suppor 58 sort by the first event. It can support multiple groups with different
61 amount of events. WARNING: This should 59 amount of events. WARNING: This should be used on grouped events.
62 60
63 -F <freq>:: 61 -F <freq>::
64 --freq=<freq>:: 62 --freq=<freq>::
65 Profile at this frequency. Use 'max' t 63 Profile at this frequency. Use 'max' to use the currently maximum
66 allowed frequency, i.e. the value in t 64 allowed frequency, i.e. the value in the kernel.perf_event_max_sample_rate
67 sysctl. 65 sysctl.
68 66
69 -i:: 67 -i::
70 --inherit:: 68 --inherit::
71 Child tasks do not inherit counters. 69 Child tasks do not inherit counters.
72 70
73 -k <path>:: 71 -k <path>::
74 --vmlinux=<path>:: 72 --vmlinux=<path>::
75 Path to vmlinux. Required for annotat 73 Path to vmlinux. Required for annotation functionality.
76 74
77 --ignore-vmlinux:: 75 --ignore-vmlinux::
78 Ignore vmlinux files. 76 Ignore vmlinux files.
79 77
80 --kallsyms=<file>:: 78 --kallsyms=<file>::
81 kallsyms pathname 79 kallsyms pathname
82 80
83 -m <pages>:: 81 -m <pages>::
84 --mmap-pages=<pages>:: 82 --mmap-pages=<pages>::
85 Number of mmap data pages (must be a p 83 Number of mmap data pages (must be a power of two) or size
86 specification in bytes with appended u !! 84 specification with appended unit character - B/K/M/G. The
87 The size is rounded up to the nearest !! 85 size is rounded up to have nearest pages power of two value.
88 86
89 -p <pid>:: 87 -p <pid>::
90 --pid=<pid>:: 88 --pid=<pid>::
91 Profile events on existing Process ID 89 Profile events on existing Process ID (comma separated list).
92 90
93 -t <tid>:: 91 -t <tid>::
94 --tid=<tid>:: 92 --tid=<tid>::
95 Profile events on existing thread ID ( 93 Profile events on existing thread ID (comma separated list).
96 94
97 -u:: 95 -u::
98 --uid=:: 96 --uid=::
99 Record events in threads owned by uid. 97 Record events in threads owned by uid. Name or number.
100 98
101 -r <priority>:: 99 -r <priority>::
102 --realtime=<priority>:: 100 --realtime=<priority>::
103 Collect data with this RT SCHED_FIFO p 101 Collect data with this RT SCHED_FIFO priority.
104 102
105 --sym-annotate=<symbol>:: 103 --sym-annotate=<symbol>::
106 Annotate this symbol. 104 Annotate this symbol.
107 105
108 -K:: 106 -K::
109 --hide_kernel_symbols:: 107 --hide_kernel_symbols::
110 Hide kernel symbols. 108 Hide kernel symbols.
111 109
112 -U:: 110 -U::
113 --hide_user_symbols:: 111 --hide_user_symbols::
114 Hide user symbols. 112 Hide user symbols.
115 113
116 --demangle-kernel:: 114 --demangle-kernel::
117 Demangle kernel symbols. 115 Demangle kernel symbols.
118 116
119 -D:: 117 -D::
120 --dump-symtab:: 118 --dump-symtab::
121 Dump the symbol table used for profili 119 Dump the symbol table used for profiling.
122 120
123 -v:: 121 -v::
124 --verbose:: 122 --verbose::
125 Be more verbose (show counter open err 123 Be more verbose (show counter open errors, etc).
126 124
127 -z:: 125 -z::
128 --zero:: 126 --zero::
129 Zero history across display updates. 127 Zero history across display updates.
130 128
131 -s:: 129 -s::
132 --sort:: 130 --sort::
133 Sort by key(s): pid, comm, dso, symbol 131 Sort by key(s): pid, comm, dso, symbol, parent, srcline, weight,
134 local_weight, abort, in_tx, transactio 132 local_weight, abort, in_tx, transaction, overhead, sample, period.
135 Please see description of --sort in th 133 Please see description of --sort in the perf-report man page.
136 134
137 --fields=:: 135 --fields=::
138 Specify output field - multiple keys c 136 Specify output field - multiple keys can be specified in CSV format.
139 Following fields are available: 137 Following fields are available:
140 overhead, overhead_sys, overhead_us, o 138 overhead, overhead_sys, overhead_us, overhead_children, sample and period.
141 Also it can contain any sort key(s). 139 Also it can contain any sort key(s).
142 140
143 By default, every sort keys not specif 141 By default, every sort keys not specified in --field will be appended
144 automatically. 142 automatically.
145 143
146 -n:: 144 -n::
147 --show-nr-samples:: 145 --show-nr-samples::
148 Show a column with the number of sampl 146 Show a column with the number of samples.
149 147
150 --show-total-period:: 148 --show-total-period::
151 Show a column with the sum of periods. 149 Show a column with the sum of periods.
152 150
153 --dsos:: 151 --dsos::
154 Only consider symbols in these dsos. 152 Only consider symbols in these dsos. This option will affect the
155 percentage of the overhead column. Se 153 percentage of the overhead column. See --percentage for more info.
156 154
157 --comms:: 155 --comms::
158 Only consider symbols in these comms. 156 Only consider symbols in these comms. This option will affect the
159 percentage of the overhead column. Se 157 percentage of the overhead column. See --percentage for more info.
160 158
161 --symbols:: 159 --symbols::
162 Only consider these symbols. This opt 160 Only consider these symbols. This option will affect the
163 percentage of the overhead column. Se 161 percentage of the overhead column. See --percentage for more info.
164 162
165 -M:: 163 -M::
166 --disassembler-style=:: Set disassembler style 164 --disassembler-style=:: Set disassembler style for objdump.
167 165
168 --addr2line=<path>:: <<
169 Path to addr2line binary. <<
170 <<
171 --objdump=<path>:: <<
172 Path to objdump binary. <<
173 <<
174 --prefix=PREFIX:: 166 --prefix=PREFIX::
175 --prefix-strip=N:: 167 --prefix-strip=N::
176 Remove first N entries from source fil 168 Remove first N entries from source file path names in executables
177 and add PREFIX. This allows to display 169 and add PREFIX. This allows to display source code compiled on systems
178 with different file system layout. 170 with different file system layout.
179 171
180 --source:: 172 --source::
181 Interleave source code with assembly c 173 Interleave source code with assembly code. Enabled by default,
182 disable with --no-source. 174 disable with --no-source.
183 175
184 --asm-raw:: 176 --asm-raw::
185 Show raw instruction encoding of assem 177 Show raw instruction encoding of assembly instructions.
186 178
187 -g:: 179 -g::
188 Enables call-graph (stack chain/backtr 180 Enables call-graph (stack chain/backtrace) recording.
189 181
190 --call-graph [mode,type,min[,limit],order[,key 182 --call-graph [mode,type,min[,limit],order[,key][,branch]]::
191 Setup and enable call-graph (stack cha 183 Setup and enable call-graph (stack chain/backtrace) recording,
192 implies -g. See `--call-graph` sectio 184 implies -g. See `--call-graph` section in perf-record and
193 perf-report man pages for details. 185 perf-report man pages for details.
194 186
195 --children:: 187 --children::
196 Accumulate callchain of children to pa 188 Accumulate callchain of children to parent entry so that then can
197 show up in the output. The output wil 189 show up in the output. The output will have a new "Children" column
198 and will be sorted on the data. It re 190 and will be sorted on the data. It requires -g/--call-graph option
199 enabled. See the `overhead calculatio 191 enabled. See the `overhead calculation' section for more details.
200 Enabled by default, disable with --no- 192 Enabled by default, disable with --no-children.
201 193
202 --max-stack:: 194 --max-stack::
203 Set the stack depth limit when parsing 195 Set the stack depth limit when parsing the callchain, anything
204 beyond the specified depth will be ign 196 beyond the specified depth will be ignored. This is a trade-off
205 between information loss and faster pr 197 between information loss and faster processing especially for
206 workloads that can have a very long ca 198 workloads that can have a very long callchain stack.
207 199
208 Default: /proc/sys/kernel/perf_event_m 200 Default: /proc/sys/kernel/perf_event_max_stack when present, 127 otherwise.
209 201
210 --ignore-callees=<regex>:: 202 --ignore-callees=<regex>::
211 Ignore callees of the function(s) matc 203 Ignore callees of the function(s) matching the given regex.
212 This has the effect of collecting the 204 This has the effect of collecting the callers of each such
213 function into one place in the call-gr 205 function into one place in the call-graph tree.
214 206
215 --percent-limit:: 207 --percent-limit::
216 Do not show entries which have an over 208 Do not show entries which have an overhead under that percent.
217 (Default: 0). 209 (Default: 0).
218 210
219 --percentage:: 211 --percentage::
220 Determine how to display the overhead 212 Determine how to display the overhead percentage of filtered entries.
221 Filters can be applied by --comms, --d 213 Filters can be applied by --comms, --dsos and/or --symbols options and
222 Zoom operations on the TUI (thread, ds 214 Zoom operations on the TUI (thread, dso, etc).
223 215
224 "relative" means it's relative to filt 216 "relative" means it's relative to filtered entries only so that the
225 sum of shown entries will be always 10 217 sum of shown entries will be always 100%. "absolute" means it retains
226 the original value before and after th 218 the original value before and after the filter is applied.
227 219
228 -w:: 220 -w::
229 --column-widths=<width[,width...]>:: 221 --column-widths=<width[,width...]>::
230 Force each column width to the provide 222 Force each column width to the provided list, for large terminal
231 readability. 0 means no limit (defaul 223 readability. 0 means no limit (default behavior).
232 224
233 --proc-map-timeout:: 225 --proc-map-timeout::
234 When processing pre-existing threads / 226 When processing pre-existing threads /proc/XXX/mmap, it may take
235 a long time, because the file may be h 227 a long time, because the file may be huge. A time out is needed
236 in such cases. 228 in such cases.
237 This option sets the time out limit. T 229 This option sets the time out limit. The default value is 500 ms.
238 230
239 231
240 -b:: 232 -b::
241 --branch-any:: 233 --branch-any::
242 Enable taken branch stack sampling. An 234 Enable taken branch stack sampling. Any type of taken branch may be sampled.
243 This is a shortcut for --branch-filter 235 This is a shortcut for --branch-filter any. See --branch-filter for more infos.
244 236
245 -j:: 237 -j::
246 --branch-filter:: 238 --branch-filter::
247 Enable taken branch stack sampling. Ea 239 Enable taken branch stack sampling. Each sample captures a series of consecutive
248 taken branches. The number of branches 240 taken branches. The number of branches captured with each sample depends on the
249 underlying hardware, the type of branc 241 underlying hardware, the type of branches of interest, and the executed code.
250 It is possible to select the types of 242 It is possible to select the types of branches captured by enabling filters.
251 For a full list of modifiers please se 243 For a full list of modifiers please see the perf record manpage.
252 244
253 The option requires at least one branc 245 The option requires at least one branch type among any, any_call, any_ret, ind_call, cond.
254 The privilege levels may be omitted, i 246 The privilege levels may be omitted, in which case, the privilege levels of the associated
255 event are applied to the branch filter 247 event are applied to the branch filter. Both kernel (k) and hypervisor (hv) privilege
256 levels are subject to permissions. Wh 248 levels are subject to permissions. When sampling on multiple events, branch stack sampling
257 is enabled for all the sampling events 249 is enabled for all the sampling events. The sampled branch type is the same for all events.
258 The various filters must be specified 250 The various filters must be specified as a comma separated list: --branch-filter any_ret,u,k
259 Note that this feature may not be avai 251 Note that this feature may not be available on all processors.
260 252
261 --branch-history:: <<
262 Add the addresses of sampled taken bra <<
263 This allows to examine the path the pr <<
264 <<
265 --raw-trace:: 253 --raw-trace::
266 When displaying traceevent output, do 254 When displaying traceevent output, do not use print fmt or plugins.
267 255
268 -H:: <<
269 --hierarchy:: 256 --hierarchy::
270 Enable hierarchical output. In the hi !! 257 Enable hierarchy output.
271 samples based on the criteria and then <<
272 level sort key. <<
273 <<
274 For example, in normal output: <<
275 <<
276 perf report -s dso,sym <<
277 # <<
278 # Overhead Shared Object Symbo <<
279 # ........ ................. ..... <<
280 50.00% [kernel.kallsyms] [k] k <<
281 20.00% perf [.] f <<
282 15.00% [kernel.kallsyms] [k] k <<
283 10.00% perf [.] b <<
284 5.00% libc.so [.] l <<
285 <<
286 In hierarchy output: <<
287 <<
288 perf report -s dso,sym --hierarchy <<
289 # <<
290 # Overhead Shared Object / Symbol <<
291 # .......... ...................... <<
292 65.00% [kernel.kallsyms] <<
293 50.00% [k] kfunc1 <<
294 15.00% [k] kfunc2 <<
295 30.00% perf <<
296 20.00% [.] foo <<
297 10.00% [.] bar <<
298 5.00% libc.so <<
299 5.00% [.] libcall <<
300 258
301 --overwrite:: 259 --overwrite::
302 Enable this to use just the most recen 260 Enable this to use just the most recent records, which helps in high core count
303 machines such as Knights Landing/Mill, 261 machines such as Knights Landing/Mill, but right now is disabled by default as
304 the pausing used in this technique is 262 the pausing used in this technique is leading to loss of metadata events such
305 as PERF_RECORD_MMAP which makes 'perf 263 as PERF_RECORD_MMAP which makes 'perf top' unable to resolve samples, leading
306 to lots of unknown samples appearing o 264 to lots of unknown samples appearing on the UI. Enable this if you are in such
307 machines and profiling a workload that 265 machines and profiling a workload that doesn't creates short lived threads and/or
308 doesn't uses many executable mmap oper 266 doesn't uses many executable mmap operations. Work is being planed to solve
309 this situation, till then, this will r 267 this situation, till then, this will remain disabled by default.
310 268
311 --force:: 269 --force::
312 Don't do ownership validation. 270 Don't do ownership validation.
313 271
314 --num-thread-synthesize:: 272 --num-thread-synthesize::
315 The number of threads to run when synt 273 The number of threads to run when synthesizing events for existing processes.
316 By default, the number of threads equa 274 By default, the number of threads equals to the number of online CPUs.
317 275
318 --namespaces:: 276 --namespaces::
319 Record events of type PERF_RECORD_NAME 277 Record events of type PERF_RECORD_NAMESPACES and display it with the
320 'cgroup_id' sort key. 278 'cgroup_id' sort key.
321 279
322 -G name:: <<
323 --cgroup name:: <<
324 monitor only in the container (cgroup) called <<
325 in per-cpu mode. The cgroup filesystem must be <<
326 container "name" are monitored when they run o <<
327 can be provided. Each cgroup is applied to the <<
328 to first event, second cgroup to second event <<
329 an empty cgroup (monitor all the time) using, <<
330 corresponding events, i.e., they always refer <<
331 line. If the user wants to track multiple even <<
332 use '-e e1 -e e2 -G foo,foo' or just use '-e e <<
333 <<
334 --all-cgroups:: 280 --all-cgroups::
335 Record events of type PERF_RECORD_CGRO 281 Record events of type PERF_RECORD_CGROUP and display it with the
336 'cgroup' sort key. 282 'cgroup' sort key.
337 283
338 --switch-on EVENT_NAME:: 284 --switch-on EVENT_NAME::
339 Only consider events after this event 285 Only consider events after this event is found.
340 286
341 E.g.: 287 E.g.:
342 288
343 Find out where broadcast packets ar 289 Find out where broadcast packets are handled
344 290
345 perf probe -L icmp_rcv 291 perf probe -L icmp_rcv
346 292
347 Insert a probe there: 293 Insert a probe there:
348 294
349 perf probe icmp_rcv:59 295 perf probe icmp_rcv:59
350 296
351 Start perf top and ask it to only c 297 Start perf top and ask it to only consider the cycles events when a
352 broadcast packet arrives This will 298 broadcast packet arrives This will show a menu with two entries and
353 will start counting when a broadcas 299 will start counting when a broadcast packet arrives:
354 300
355 perf top -e cycles,probe:icmp_ 301 perf top -e cycles,probe:icmp_rcv --switch-on=probe:icmp_rcv
356 302
357 Alternatively one can ask for a gro !! 303 Alternatively one can ask for --group and then two overhead columns
358 will appear, the first for cycles a 304 will appear, the first for cycles and the second for the switch-on event.
359 305
360 perf top -e '{cycles,probe:icm !! 306 perf top --group -e cycles,probe:icmp_rcv --switch-on=probe:icmp_rcv
361 307
362 This may be interesting to measure a w 308 This may be interesting to measure a workload only after some initialization
363 phase is over, i.e. insert a perf prob 309 phase is over, i.e. insert a perf probe at that point and use the above
364 examples replacing probe:icmp_rcv with 310 examples replacing probe:icmp_rcv with the just-after-init probe.
365 311
366 --switch-off EVENT_NAME:: 312 --switch-off EVENT_NAME::
367 Stop considering events after this eve 313 Stop considering events after this event is found.
368 314
369 --show-on-off-events:: 315 --show-on-off-events::
370 Show the --switch-on/off events too. T 316 Show the --switch-on/off events too. This has no effect in 'perf top' now
371 but probably we'll make the default no 317 but probably we'll make the default not to show the switch-on/off events
372 on the --group mode and if there is on 318 on the --group mode and if there is only one event besides the off/on ones,
373 go straight to the histogram browser, 319 go straight to the histogram browser, just like 'perf top' with no events
374 explicitly specified does. !! 320 explicitely specified does.
375 321
376 --stitch-lbr:: 322 --stitch-lbr::
377 Show callgraph with stitched LBRs, whi 323 Show callgraph with stitched LBRs, which may have more complete
378 callgraph. The option must be used wit 324 callgraph. The option must be used with --call-graph lbr recording.
379 Disabled by default. In common cases w 325 Disabled by default. In common cases with call stack overflows,
380 it can recreate better call stacks tha 326 it can recreate better call stacks than the default lbr call stack
381 output. But this approach is not foolp !! 327 output. But this approach is not full proof. There can be cases
382 where it creates incorrect call stacks 328 where it creates incorrect call stacks from incorrect matches.
383 The known limitations include exceptio 329 The known limitations include exception handing such as
384 setjmp/longjmp will have calls/returns 330 setjmp/longjmp will have calls/returns not match.
385 331
386 ifdef::HAVE_LIBPFM[] 332 ifdef::HAVE_LIBPFM[]
387 --pfm-events events:: 333 --pfm-events events::
388 Select a PMU event using libpfm4 syntax (see h 334 Select a PMU event using libpfm4 syntax (see http://perfmon2.sf.net)
389 including support for event filters. For examp 335 including support for event filters. For example '--pfm-events
390 inst_retired:any_p:u:c=1:i'. More than one eve 336 inst_retired:any_p:u:c=1:i'. More than one event can be passed to the
391 option using the comma separator. Hardware eve 337 option using the comma separator. Hardware events and generic hardware
392 events cannot be mixed together. The latter mu 338 events cannot be mixed together. The latter must be used with the -e
393 option. The -e option and this one can be mixe 339 option. The -e option and this one can be mixed and matched. Events
394 can be grouped using the {} notation. 340 can be grouped using the {} notation.
395 endif::HAVE_LIBPFM[] 341 endif::HAVE_LIBPFM[]
396 342
397 INTERACTIVE PROMPTING KEYS 343 INTERACTIVE PROMPTING KEYS
398 -------------------------- 344 --------------------------
399 345
400 [d]:: 346 [d]::
401 Display refresh delay. 347 Display refresh delay.
402 348
403 [e]:: 349 [e]::
404 Number of entries to display. 350 Number of entries to display.
405 351
406 [E]:: 352 [E]::
407 Event to display when multiple counter 353 Event to display when multiple counters are active.
408 354
409 [f]:: 355 [f]::
410 Profile display filter (>= hit count). 356 Profile display filter (>= hit count).
411 357
412 [F]:: 358 [F]::
413 Annotation display filter (>= % of tot 359 Annotation display filter (>= % of total).
414 360
415 [s]:: 361 [s]::
416 Annotate symbol. 362 Annotate symbol.
417 363
418 [S]:: 364 [S]::
419 Stop annotation, return to full profil 365 Stop annotation, return to full profile display.
420 366
421 [K]:: 367 [K]::
422 Hide kernel symbols. 368 Hide kernel symbols.
423 369
424 [U]:: 370 [U]::
425 Hide user symbols. 371 Hide user symbols.
426 372
427 [z]:: 373 [z]::
428 Toggle event count zeroing across disp 374 Toggle event count zeroing across display updates.
429 375
430 [qQ]:: 376 [qQ]::
431 Quit. 377 Quit.
432 378
433 Pressing any unmapped key displays a menu, and 379 Pressing any unmapped key displays a menu, and prompts for input.
434 380
435 include::callchain-overhead-calculation.txt[] 381 include::callchain-overhead-calculation.txt[]
436 382
437 SEE ALSO 383 SEE ALSO
438 -------- 384 --------
439 linkperf:perf-stat[1], linkperf:perf-list[1], 385 linkperf:perf-stat[1], linkperf:perf-list[1], linkperf:perf-report[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.