1 perf-script(1) 1 perf-script(1)
2 ============= 2 =============
3 3
4 NAME 4 NAME
5 ---- 5 ----
6 perf-script - Read perf.data (created by perf 6 perf-script - Read perf.data (created by perf record) and display trace output
7 7
8 SYNOPSIS 8 SYNOPSIS
9 -------- 9 --------
10 [verse] 10 [verse]
11 'perf script' [<options>] 11 'perf script' [<options>]
12 'perf script' [<options>] record <script> [<re 12 'perf script' [<options>] record <script> [<record-options>] <command>
13 'perf script' [<options>] report <script> [scr 13 'perf script' [<options>] report <script> [script-args]
14 'perf script' [<options>] <script> <required-s 14 'perf script' [<options>] <script> <required-script-args> [<record-options>] <command>
15 'perf script' [<options>] <top-script> [script 15 'perf script' [<options>] <top-script> [script-args]
16 16
17 DESCRIPTION 17 DESCRIPTION
18 ----------- 18 -----------
19 This command reads the input file and displays 19 This command reads the input file and displays the trace recorded.
20 20
21 There are several variants of perf script: 21 There are several variants of perf script:
22 22
23 'perf script' to see a detailed trace of the 23 'perf script' to see a detailed trace of the workload that was
24 recorded. 24 recorded.
25 25
26 You can also run a set of pre-canned scripts 26 You can also run a set of pre-canned scripts that aggregate and
27 summarize the raw trace data in various ways 27 summarize the raw trace data in various ways (the list of scripts is
28 available via 'perf script -l'). The follow 28 available via 'perf script -l'). The following variants allow you to
29 record and run those scripts: 29 record and run those scripts:
30 30
31 'perf script record <script> <command>' to r 31 'perf script record <script> <command>' to record the events required
32 for 'perf script report'. <script> is the n 32 for 'perf script report'. <script> is the name displayed in the
33 output of 'perf script --list' i.e. the actu 33 output of 'perf script --list' i.e. the actual script name minus any
34 language extension. If <command> is not spe 34 language extension. If <command> is not specified, the events are
35 recorded using the -a (system-wide) 'perf re 35 recorded using the -a (system-wide) 'perf record' option.
36 36
37 'perf script report <script> [args]' to run 37 'perf script report <script> [args]' to run and display the results
38 of <script>. <script> is the name displayed 38 of <script>. <script> is the name displayed in the output of 'perf
39 script --list' i.e. the actual script name m 39 script --list' i.e. the actual script name minus any language
40 extension. The perf.data output from a prev 40 extension. The perf.data output from a previous run of 'perf script
41 record <script>' is used and should be prese 41 record <script>' is used and should be present for this command to
42 succeed. [args] refers to the (mainly optio 42 succeed. [args] refers to the (mainly optional) args expected by
43 the script. 43 the script.
44 44
45 'perf script <script> <required-script-args> 45 'perf script <script> <required-script-args> <command>' to both
46 record the events required for <script> and 46 record the events required for <script> and to run the <script>
47 using 'live-mode' i.e. without writing anyth 47 using 'live-mode' i.e. without writing anything to disk. <script>
48 is the name displayed in the output of 'perf 48 is the name displayed in the output of 'perf script --list' i.e. the
49 actual script name minus any language extens 49 actual script name minus any language extension. If <command> is
50 not specified, the events are recorded using 50 not specified, the events are recorded using the -a (system-wide)
51 'perf record' option. If <script> has any r 51 'perf record' option. If <script> has any required args, they
52 should be specified before <command>. This 52 should be specified before <command>. This mode doesn't allow for
53 optional script args to be specified; if opt 53 optional script args to be specified; if optional script args are
54 desired, they can be specified using separat 54 desired, they can be specified using separate 'perf script record'
55 and 'perf script report' commands, with the 55 and 'perf script report' commands, with the stdout of the record step
56 piped to the stdin of the report script, usi 56 piped to the stdin of the report script, using the '-o -' and '-i -'
57 options of the corresponding commands. 57 options of the corresponding commands.
58 58
59 'perf script <top-script>' to both record th 59 'perf script <top-script>' to both record the events required for
60 <top-script> and to run the <top-script> usi 60 <top-script> and to run the <top-script> using 'live-mode'
61 i.e. without writing anything to disk. <top 61 i.e. without writing anything to disk. <top-script> is the name
62 displayed in the output of 'perf script --li 62 displayed in the output of 'perf script --list' i.e. the actual
63 script name minus any language extension; a 63 script name minus any language extension; a <top-script> is defined
64 as any script name ending with the string 't 64 as any script name ending with the string 'top'.
65 65
66 [<record-options>] can be passed to the reco 66 [<record-options>] can be passed to the record steps of 'perf script
67 record' and 'live-mode' variants; this isn't 67 record' and 'live-mode' variants; this isn't possible however for
68 <top-script> 'live-mode' or 'perf script rep 68 <top-script> 'live-mode' or 'perf script report' variants.
69 69
70 See the 'SEE ALSO' section for links to lang 70 See the 'SEE ALSO' section for links to language-specific
71 information on how to write and run your own 71 information on how to write and run your own trace scripts.
72 72
73 OPTIONS 73 OPTIONS
74 ------- 74 -------
75 <command>...:: 75 <command>...::
76 Any command you can specify in a shell 76 Any command you can specify in a shell.
77 77
78 -D:: 78 -D::
79 --dump-raw-trace=:: 79 --dump-raw-trace=::
80 Display verbose dump of the trace data 80 Display verbose dump of the trace data.
81 81
82 --dump-unsorted-raw-trace=:: 82 --dump-unsorted-raw-trace=::
83 Same as --dump-raw-trace but not sorte 83 Same as --dump-raw-trace but not sorted in time order.
84 84
85 -L:: 85 -L::
86 --Latency=:: 86 --Latency=::
87 Show latency attributes (irqs/preempti 87 Show latency attributes (irqs/preemption disabled, etc).
88 88
89 -l:: 89 -l::
90 --list=:: 90 --list=::
91 Display a list of available trace scri 91 Display a list of available trace scripts.
92 92
93 -s ['lang']:: 93 -s ['lang']::
94 --script=:: 94 --script=::
95 Process trace data with the given scri 95 Process trace data with the given script ([lang]:script[.ext]).
96 If the string 'lang' is specified in p 96 If the string 'lang' is specified in place of a script name, a
97 list of supported languages will be di 97 list of supported languages will be displayed instead.
98 98
99 -g:: 99 -g::
100 --gen-script=:: 100 --gen-script=::
101 Generate perf-script.[ext] starter scr 101 Generate perf-script.[ext] starter script for given language,
102 using current perf.data. 102 using current perf.data.
103 103
104 --dlfilter=<file>:: 104 --dlfilter=<file>::
105 Filter sample events using the given s 105 Filter sample events using the given shared object file.
106 Refer linkperf:perf-dlfilter[1] 106 Refer linkperf:perf-dlfilter[1]
107 107
108 --dlarg=<arg>:: 108 --dlarg=<arg>::
109 Pass 'arg' as an argument to the dlfil 109 Pass 'arg' as an argument to the dlfilter. --dlarg may be repeated
110 to add more arguments. 110 to add more arguments.
111 111
112 --list-dlfilters:: 112 --list-dlfilters::
113 Display a list of available dlfilters. 113 Display a list of available dlfilters. Use with option -v (must come
114 before option --list-dlfilters) to sho 114 before option --list-dlfilters) to show long descriptions.
115 115
116 -a:: 116 -a::
117 Force system-wide collection. Scripts 117 Force system-wide collection. Scripts run without a <command>
118 normally use -a by default, while scri 118 normally use -a by default, while scripts run with a <command>
119 normally don't - this option allows th 119 normally don't - this option allows the latter to be run in
120 system-wide mode. 120 system-wide mode.
121 121
122 -i:: 122 -i::
123 --input=:: 123 --input=::
124 Input file name. (default: perf.data u 124 Input file name. (default: perf.data unless stdin is a fifo)
125 125
126 -d:: 126 -d::
127 --debug-mode:: 127 --debug-mode::
128 Do various checks like samples orderin 128 Do various checks like samples ordering and lost events.
129 129
130 -F:: 130 -F::
131 --fields:: 131 --fields::
132 Comma separated list of fields to prin 132 Comma separated list of fields to print. Options are:
133 comm, tid, pid, time, cpu, event, trac !! 133 comm, tid, pid, time, cpu, event, trace, ip, sym, dso, addr, symoff,
134 srcline, period, iregs, uregs, brstack 134 srcline, period, iregs, uregs, brstack, brstacksym, flags, bpf-output,
135 brstackinsn, brstackinsnlen, brstackdi !! 135 brstackinsn, brstackinsnlen, brstackoff, callindent, insn, insnlen, synth,
136 insnlen, synth, phys_addr, metric, mis !! 136 phys_addr, metric, misc, srccode, ipc, data_page_size, code_page_size, ins_lat,
137 code_page_size, ins_lat, machine_pid, !! 137 machine_pid, vcpu.
138 <<
139 Field list can be prepended with the t 138 Field list can be prepended with the type, trace, sw or hw,
140 to indicate to which event type the fi 139 to indicate to which event type the field list applies.
141 e.g., -F sw:comm,tid,time,ip,sym and 140 e.g., -F sw:comm,tid,time,ip,sym and -F trace:time,cpu,trace
142 141
143 perf script -F <fields> 142 perf script -F <fields>
144 143
145 is equivalent to: 144 is equivalent to:
146 145
147 perf script -F trace:<fields> 146 perf script -F trace:<fields> -F sw:<fields> -F hw:<fields>
148 147
149 i.e., the specified fields apply to al 148 i.e., the specified fields apply to all event types if the type string
150 is not given. 149 is not given.
151 150
152 In addition to overriding fields, it i 151 In addition to overriding fields, it is also possible to add or remove
153 fields from the defaults. For example 152 fields from the defaults. For example
154 153
155 -F -cpu,+insn 154 -F -cpu,+insn
156 155
157 removes the cpu field and adds the ins 156 removes the cpu field and adds the insn field. Adding/removing fields
158 cannot be mixed with normal overriding 157 cannot be mixed with normal overriding.
159 158
160 The arguments are processed in the ord 159 The arguments are processed in the order received. A later usage can
161 reset a prior request. e.g.: 160 reset a prior request. e.g.:
162 161
163 -F trace: -F comm,tid,time,ip, 162 -F trace: -F comm,tid,time,ip,sym
164 163
165 The first -F suppresses trace events ( 164 The first -F suppresses trace events (field list is ""), but then the
166 second invocation sets the fields to c 165 second invocation sets the fields to comm,tid,time,ip,sym. In this case a
167 warning is given to the user: 166 warning is given to the user:
168 167
169 "Overriding previous field req 168 "Overriding previous field request for all events."
170 169
171 Alternatively, consider the order: 170 Alternatively, consider the order:
172 171
173 -F comm,tid,time,ip,sym -F tra 172 -F comm,tid,time,ip,sym -F trace:
174 173
175 The first -F sets the fields for all e 174 The first -F sets the fields for all events and the second -F
176 suppresses trace events. The user is g 175 suppresses trace events. The user is given a warning message about
177 the override, and the result of the ab 176 the override, and the result of the above is that only S/W and H/W
178 events are displayed with the given fi 177 events are displayed with the given fields.
179 178
180 It's possible tp add/remove fields onl 179 It's possible tp add/remove fields only for specific event type:
181 180
182 -Fsw:-cpu,-period 181 -Fsw:-cpu,-period
183 182
184 removes cpu and period from software e 183 removes cpu and period from software events.
185 184
186 For the 'wildcard' option if a user se 185 For the 'wildcard' option if a user selected field is invalid for an
187 event type, a message is displayed to 186 event type, a message is displayed to the user that the option is
188 ignored for that type. For example: 187 ignored for that type. For example:
189 188
190 $ perf script -F comm,tid,trac 189 $ perf script -F comm,tid,trace
191 'trace' not valid for hardware 190 'trace' not valid for hardware events. Ignoring.
192 'trace' not valid for software 191 'trace' not valid for software events. Ignoring.
193 192
194 Alternatively, if the type is given an 193 Alternatively, if the type is given an invalid field is specified it
195 is an error. For example: 194 is an error. For example:
196 195
197 perf script -v -F sw:comm,tid,trace 196 perf script -v -F sw:comm,tid,trace
198 'trace' not valid for software events. 197 'trace' not valid for software events.
199 198
200 At this point usage is displayed, and 199 At this point usage is displayed, and perf-script exits.
201 200
202 The flags field is synthesized and may 201 The flags field is synthesized and may have a value when Instruction
203 Trace decoding. The flags are "bcrosyi 202 Trace decoding. The flags are "bcrosyiABExghDt" which stand for branch,
204 call, return, conditional, system, asy 203 call, return, conditional, system, asynchronous, interrupt,
205 transaction abort, trace begin, trace 204 transaction abort, trace begin, trace end, in transaction, VM-Entry,
206 VM-Exit, interrupt disabled and interr 205 VM-Exit, interrupt disabled and interrupt disable toggle respectively.
207 Known combinations of flags are printe 206 Known combinations of flags are printed more nicely e.g.
208 "call" for "bc", "return" for "br", "j 207 "call" for "bc", "return" for "br", "jcc" for "bo", "jmp" for "b",
209 "int" for "bci", "iret" for "bri", "sy 208 "int" for "bci", "iret" for "bri", "syscall" for "bcs", "sysret" for "brs",
210 "async" for "by", "hw int" for "bcyi", 209 "async" for "by", "hw int" for "bcyi", "tx abrt" for "bA", "tr strt" for "bB",
211 "tr end" for "bE", "vmentry" for "bcg" 210 "tr end" for "bE", "vmentry" for "bcg", "vmexit" for "bch".
212 However the "x", "D" and "t" flags wil 211 However the "x", "D" and "t" flags will be displayed separately in those
213 cases e.g. "jcc (xD)" for a condit 212 cases e.g. "jcc (xD)" for a condition branch within a transaction
214 with interrupts disabled. Note, interr 213 with interrupts disabled. Note, interrupts becoming disabled is "t",
215 whereas interrupts becoming enabled is 214 whereas interrupts becoming enabled is "Dt".
216 215
217 The callindent field is synthesized an 216 The callindent field is synthesized and may have a value when
218 Instruction Trace decoding. For calls 217 Instruction Trace decoding. For calls and returns, it will display the
219 name of the symbol indented with space 218 name of the symbol indented with spaces to reflect the stack depth.
220 219
221 When doing instruction trace decoding, !! 220 When doing instruction trace decoding insn and insnlen give the
222 instruction bytes, disassembled instru !! 221 instruction bytes and the instruction length of the current
223 and the instruction length of the curr !! 222 instruction.
224 223
225 The synth field is used by synthesized 224 The synth field is used by synthesized events which may be created when
226 Instruction Trace decoding. 225 Instruction Trace decoding.
227 226
228 The ipc (instructions per cycle) field 227 The ipc (instructions per cycle) field is synthesized and may have a value when
229 Instruction Trace decoding. 228 Instruction Trace decoding.
230 229
231 The machine_pid and vcpu fields are de 230 The machine_pid and vcpu fields are derived from data resulting from using
232 perf inject to insert a perf.data file 231 perf inject to insert a perf.data file recorded inside a virtual machine into
233 a perf.data file recorded on the host 232 a perf.data file recorded on the host at the same time.
234 233
235 The cgroup fields requires sample havi <<
236 when "--all-cgroups" option is passed <<
237 <<
238 Finally, a user may not set fields to 234 Finally, a user may not set fields to none for all event types.
239 i.e., -F "" is not allowed. 235 i.e., -F "" is not allowed.
240 236
241 The brstack output includes branch rel 237 The brstack output includes branch related information with raw addresses using the
242 /v/v/v/v/cycles syntax in the followin 238 /v/v/v/v/cycles syntax in the following order:
243 FROM: branch source instruction 239 FROM: branch source instruction
244 TO : branch target instruction 240 TO : branch target instruction
245 M/P/-: M=branch target mispredicted or 241 M/P/-: M=branch target mispredicted or branch direction was mispredicted, P=target predicted or direction predicted, -=not supported
246 X/- : X=branch inside a transactional 242 X/- : X=branch inside a transactional region, -=not in transaction region or not supported
247 A/- : A=TSX abort entry, -=not aborted 243 A/- : A=TSX abort entry, -=not aborted region or not supported
248 cycles 244 cycles
249 245
250 The brstacksym is identical to brstack 246 The brstacksym is identical to brstack, except that the FROM and TO addresses are printed in a symbolic form if possible.
251 247
252 When brstackinsn is specified the full 248 When brstackinsn is specified the full assembler sequences of branch sequences for each sample
253 is printed. This is the full execution 249 is printed. This is the full execution path leading to the sample. This is only supported when the
254 sample was recorded with perf record - 250 sample was recorded with perf record -b or -j any.
255 251
256 Use brstackinsnlen to print the brstac 252 Use brstackinsnlen to print the brstackinsn lenght. For example, you
257 can’t know the next sequential instr 253 can’t know the next sequential instruction after an unconditional branch unless
258 you calculate that based on its length 254 you calculate that based on its length.
259 255
260 brstackdisasm acts like brstackinsn, b <<
261 perf is built with the capstone librar <<
262 <<
263 The brstackoff field will print an off 256 The brstackoff field will print an offset into a specific dso/binary.
264 257
265 With the metric option perf script can 258 With the metric option perf script can compute metrics for
266 sampling periods, similar to perf stat 259 sampling periods, similar to perf stat. This requires
267 specifying a group with multiple event 260 specifying a group with multiple events defining metrics with the :S option
268 for perf record. perf will sample on t 261 for perf record. perf will sample on the first event, and
269 print computed metrics for all the eve 262 print computed metrics for all the events in the group. Please note
270 that the metric computed is averaged o 263 that the metric computed is averaged over the whole sampling
271 period (since the last sample), not ju 264 period (since the last sample), not just for the sample point.
272 265
273 For sample events it's possible to dis 266 For sample events it's possible to display misc field with -F +misc option,
274 following letters are displayed for ea 267 following letters are displayed for each bit:
275 268
276 PERF_RECORD_MISC_KERNEL 269 PERF_RECORD_MISC_KERNEL K
277 PERF_RECORD_MISC_USER 270 PERF_RECORD_MISC_USER U
278 PERF_RECORD_MISC_HYPERVISOR 271 PERF_RECORD_MISC_HYPERVISOR H
279 PERF_RECORD_MISC_GUEST_KERNEL 272 PERF_RECORD_MISC_GUEST_KERNEL G
280 PERF_RECORD_MISC_GUEST_USER 273 PERF_RECORD_MISC_GUEST_USER g
281 PERF_RECORD_MISC_MMAP_DATA* 274 PERF_RECORD_MISC_MMAP_DATA* M
282 PERF_RECORD_MISC_COMM_EXEC 275 PERF_RECORD_MISC_COMM_EXEC E
283 PERF_RECORD_MISC_SWITCH_OUT 276 PERF_RECORD_MISC_SWITCH_OUT S
284 PERF_RECORD_MISC_SWITCH_OUT_PREEMPT 277 PERF_RECORD_MISC_SWITCH_OUT_PREEMPT Sp
285 278
286 $ perf script -F +misc ... 279 $ perf script -F +misc ...
287 sched-messaging 1414 K 28690.6 280 sched-messaging 1414 K 28690.636582: 4590 cycles ...
288 sched-messaging 1407 U 28690.6 281 sched-messaging 1407 U 28690.636600: 325620 cycles ...
289 sched-messaging 1414 K 28690.6 282 sched-messaging 1414 K 28690.636608: 19473 cycles ...
290 misc field ___________/ 283 misc field ___________/
291 284
292 -k:: 285 -k::
293 --vmlinux=<file>:: 286 --vmlinux=<file>::
294 vmlinux pathname 287 vmlinux pathname
295 288
296 --kallsyms=<file>:: 289 --kallsyms=<file>::
297 kallsyms pathname 290 kallsyms pathname
298 291
299 --symfs=<directory>:: 292 --symfs=<directory>::
300 Look for files with symbols relative t 293 Look for files with symbols relative to this directory.
301 294
302 -G:: 295 -G::
303 --hide-call-graph:: 296 --hide-call-graph::
304 When printing symbols do not display c 297 When printing symbols do not display call chain.
305 298
306 --stop-bt:: 299 --stop-bt::
307 Stop display of callgraph at these sym 300 Stop display of callgraph at these symbols
308 301
309 -C:: 302 -C::
310 --cpu:: Only report samples for the list of CP 303 --cpu:: Only report samples for the list of CPUs provided. Multiple CPUs can
311 be provided as a comma-separated list 304 be provided as a comma-separated list with no space: 0,1. Ranges of
312 CPUs are specified with -: 0-2. Defaul 305 CPUs are specified with -: 0-2. Default is to report samples on all
313 CPUs. 306 CPUs.
314 307
315 -c:: 308 -c::
316 --comms=:: 309 --comms=::
317 Only display events for these comms. C 310 Only display events for these comms. CSV that understands
318 file://filename entries. 311 file://filename entries.
319 312
320 --pid=:: 313 --pid=::
321 Only show events for given process ID 314 Only show events for given process ID (comma separated list).
322 315
323 --tid=:: 316 --tid=::
324 Only show events for given thread ID ( 317 Only show events for given thread ID (comma separated list).
325 318
326 -I:: 319 -I::
327 --show-info:: 320 --show-info::
328 Display extended information about the 321 Display extended information about the perf.data file. This adds
329 information which may be very large an 322 information which may be very large and thus may clutter the display.
330 It currently includes: cpu and numa to 323 It currently includes: cpu and numa topology of the host system.
331 It can only be used with the perf scri 324 It can only be used with the perf script report mode.
332 325
333 --show-kernel-path:: 326 --show-kernel-path::
334 Try to resolve the path of [kernel.kal 327 Try to resolve the path of [kernel.kallsyms]
335 328
336 --show-task-events 329 --show-task-events
337 Display task related events (e.g. FORK 330 Display task related events (e.g. FORK, COMM, EXIT).
338 331
339 --show-mmap-events 332 --show-mmap-events
340 Display mmap related events (e.g. MMAP 333 Display mmap related events (e.g. MMAP, MMAP2).
341 334
342 --show-namespace-events 335 --show-namespace-events
343 Display namespace events i.e. events o 336 Display namespace events i.e. events of type PERF_RECORD_NAMESPACES.
344 337
345 --show-switch-events 338 --show-switch-events
346 Display context switch events i.e. eve 339 Display context switch events i.e. events of type PERF_RECORD_SWITCH or
347 PERF_RECORD_SWITCH_CPU_WIDE. 340 PERF_RECORD_SWITCH_CPU_WIDE.
348 341
349 --show-lost-events 342 --show-lost-events
350 Display lost events i.e. events of typ 343 Display lost events i.e. events of type PERF_RECORD_LOST.
351 344
352 --show-round-events 345 --show-round-events
353 Display finished round events i.e. eve 346 Display finished round events i.e. events of type PERF_RECORD_FINISHED_ROUND.
354 347
355 --show-bpf-events 348 --show-bpf-events
356 Display bpf events i.e. events of type 349 Display bpf events i.e. events of type PERF_RECORD_KSYMBOL and PERF_RECORD_BPF_EVENT.
357 350
358 --show-cgroup-events 351 --show-cgroup-events
359 Display cgroup events i.e. events of t 352 Display cgroup events i.e. events of type PERF_RECORD_CGROUP.
360 353
361 --show-text-poke-events 354 --show-text-poke-events
362 Display text poke events i.e. events o 355 Display text poke events i.e. events of type PERF_RECORD_TEXT_POKE and
363 PERF_RECORD_KSYMBOL. 356 PERF_RECORD_KSYMBOL.
364 357
365 --demangle:: 358 --demangle::
366 Demangle symbol names to human readabl 359 Demangle symbol names to human readable form. It's enabled by default,
367 disable with --no-demangle. 360 disable with --no-demangle.
368 361
369 --demangle-kernel:: 362 --demangle-kernel::
370 Demangle kernel symbol names to human 363 Demangle kernel symbol names to human readable form (for C++ kernels).
371 364
372 --addr2line=<path>:: <<
373 Path to addr2line binary. <<
374 <<
375 --header 365 --header
376 Show perf.data header. 366 Show perf.data header.
377 367
378 --header-only 368 --header-only
379 Show only perf.data header. 369 Show only perf.data header.
380 370
381 --itrace:: 371 --itrace::
382 Options for decoding instruction traci 372 Options for decoding instruction tracing data. The options are:
383 373
384 include::itrace.txt[] 374 include::itrace.txt[]
385 375
386 To disable decoding entirely, use --no 376 To disable decoding entirely, use --no-itrace.
387 377
388 --full-source-path:: 378 --full-source-path::
389 Show the full path for source files fo 379 Show the full path for source files for srcline output.
390 380
391 --max-stack:: 381 --max-stack::
392 Set the stack depth limit when parsing 382 Set the stack depth limit when parsing the callchain, anything
393 beyond the specified depth will be ign 383 beyond the specified depth will be ignored. This is a trade-off
394 between information loss and faster pr 384 between information loss and faster processing especially for
395 workloads that can have a very long ca 385 workloads that can have a very long callchain stack.
396 Note that when using the --itrace opti 386 Note that when using the --itrace option the synthesized callchain size
397 will override this value if the synthe 387 will override this value if the synthesized callchain size is bigger.
398 388
399 Default: 127 389 Default: 127
400 390
401 --ns:: 391 --ns::
402 Use 9 decimal places when displaying t 392 Use 9 decimal places when displaying time (i.e. show the nanoseconds)
403 393
404 -f:: 394 -f::
405 --force:: 395 --force::
406 Don't do ownership validation. 396 Don't do ownership validation.
407 397
408 --time:: 398 --time::
409 Only analyze samples within given time 399 Only analyze samples within given time window: <start>,<stop>. Times
410 have the format seconds.nanoseconds. I 400 have the format seconds.nanoseconds. If start is not given (i.e. time
411 string is ',x.y') then analysis starts 401 string is ',x.y') then analysis starts at the beginning of the file. If
412 stop time is not given (i.e. time stri 402 stop time is not given (i.e. time string is 'x.y,') then analysis goes
413 to end of file. Multiple ranges can be 403 to end of file. Multiple ranges can be separated by spaces, which
414 requires the argument to be quoted e.g 404 requires the argument to be quoted e.g. --time "1234.567,1234.789 1235,"
415 405
416 Also support time percent with multipl 406 Also support time percent with multiple time ranges. Time string is
417 'a%/n,b%/m,...' or 'a%-b%,c%-%d,...'. 407 'a%/n,b%/m,...' or 'a%-b%,c%-%d,...'.
418 408
419 For example: 409 For example:
420 Select the second 10% time slice: 410 Select the second 10% time slice:
421 perf script --time 10%/2 411 perf script --time 10%/2
422 412
423 Select from 0% to 10% time slice: 413 Select from 0% to 10% time slice:
424 perf script --time 0%-10% 414 perf script --time 0%-10%
425 415
426 Select the first and second 10% time s 416 Select the first and second 10% time slices:
427 perf script --time 10%/1,10%/2 417 perf script --time 10%/1,10%/2
428 418
429 Select from 0% to 10% and 30% to 40% s 419 Select from 0% to 10% and 30% to 40% slices:
430 perf script --time 0%-10%,30%-40% 420 perf script --time 0%-10%,30%-40%
431 421
432 --max-blocks:: 422 --max-blocks::
433 Set the maximum number of program bloc 423 Set the maximum number of program blocks to print with brstackinsn for
434 each sample. 424 each sample.
435 425
436 --reltime:: 426 --reltime::
437 Print time stamps relative to trace st 427 Print time stamps relative to trace start.
438 428
439 --deltatime:: 429 --deltatime::
440 Print time stamps relative to previous 430 Print time stamps relative to previous event.
441 431
442 --per-event-dump:: 432 --per-event-dump::
443 Create per event files with a "perf.da 433 Create per event files with a "perf.data.EVENT.dump" name instead of
444 printing to stdout, useful, for instan 434 printing to stdout, useful, for instance, for generating flamegraphs.
445 435
446 --inline:: 436 --inline::
447 If a callgraph address belongs to an i 437 If a callgraph address belongs to an inlined function, the inline stack
448 will be printed. Each entry has functi 438 will be printed. Each entry has function name and file/line. Enabled by
449 default, disable with --no-inline. 439 default, disable with --no-inline.
450 440
451 --insn-trace[=<raw|disasm>]:: 441 --insn-trace[=<raw|disasm>]::
452 Show instruction stream in bytes (raw) 442 Show instruction stream in bytes (raw) or disassembled (disasm)
453 for intel_pt traces. The default is 'r 443 for intel_pt traces. The default is 'raw'. To use xed, combine
454 'raw' with --xed to show disassembly d 444 'raw' with --xed to show disassembly done by xed.
455 445
456 --xed:: 446 --xed::
457 Run xed disassembler on output. Requir 447 Run xed disassembler on output. Requires installing the xed disassembler.
458 448
459 -S:: 449 -S::
460 --symbols=symbol[,symbol...]:: 450 --symbols=symbol[,symbol...]::
461 Only consider the listed symbols. Symb 451 Only consider the listed symbols. Symbols are typically a name
462 but they may also be hexadecimal addre 452 but they may also be hexadecimal address.
463 453
464 The hexadecimal address may be the sta 454 The hexadecimal address may be the start address of a symbol or
465 any other address to filter the trace 455 any other address to filter the trace records
466 456
467 For example, to select the symbol nopl 457 For example, to select the symbol noploop or the address 0x4007a0:
468 perf script --symbols=noploop,0x4007a0 458 perf script --symbols=noploop,0x4007a0
469 459
470 Support filtering trace records by sym 460 Support filtering trace records by symbol name, start address of
471 symbol, any hexadecimal address and ad 461 symbol, any hexadecimal address and address range.
472 462
473 The comparison order is: 463 The comparison order is:
474 464
475 1. symbol name comparison 465 1. symbol name comparison
476 2. symbol start address comparison. 466 2. symbol start address comparison.
477 3. any hexadecimal address comparison. 467 3. any hexadecimal address comparison.
478 4. address range comparison (see --add 468 4. address range comparison (see --addr-range).
479 469
480 --addr-range:: 470 --addr-range::
481 Use with -S or --symbols to list traced 471 Use with -S or --symbols to list traced records within address range.
482 472
483 For example, to list the traced records 473 For example, to list the traced records within the address range
484 [0x4007a0, 0x0x4007a9]: 474 [0x4007a0, 0x0x4007a9]:
485 perf script -S 0x4007a0 --addr-range 10 475 perf script -S 0x4007a0 --addr-range 10
486 476
487 --dsos=:: 477 --dsos=::
488 Only consider symbols in these DSOs. 478 Only consider symbols in these DSOs.
489 479
490 --call-trace:: 480 --call-trace::
491 Show call stream for intel_pt traces. 481 Show call stream for intel_pt traces. The CPUs are interleaved, but
492 can be filtered with -C. 482 can be filtered with -C.
493 483
494 --call-ret-trace:: 484 --call-ret-trace::
495 Show call and return stream for intel_ 485 Show call and return stream for intel_pt traces.
496 486
497 --graph-function:: 487 --graph-function::
498 For itrace only show specified functio 488 For itrace only show specified functions and their callees for
499 itrace. Multiple functions can be sepa 489 itrace. Multiple functions can be separated by comma.
500 490
501 --switch-on EVENT_NAME:: 491 --switch-on EVENT_NAME::
502 Only consider events after this event 492 Only consider events after this event is found.
503 493
504 --switch-off EVENT_NAME:: 494 --switch-off EVENT_NAME::
505 Stop considering events after this eve 495 Stop considering events after this event is found.
506 496
507 --show-on-off-events:: 497 --show-on-off-events::
508 Show the --switch-on/off events too. 498 Show the --switch-on/off events too.
509 499
510 --stitch-lbr:: 500 --stitch-lbr::
511 Show callgraph with stitched LBRs, whi 501 Show callgraph with stitched LBRs, which may have more complete
512 callgraph. The perf.data file must hav 502 callgraph. The perf.data file must have been obtained using
513 perf record --call-graph lbr. 503 perf record --call-graph lbr.
514 Disabled by default. In common cases w 504 Disabled by default. In common cases with call stack overflows,
515 it can recreate better call stacks tha 505 it can recreate better call stacks than the default lbr call stack
516 output. But this approach is not foolp !! 506 output. But this approach is not full proof. There can be cases
517 where it creates incorrect call stacks 507 where it creates incorrect call stacks from incorrect matches.
518 The known limitations include exceptio 508 The known limitations include exception handing such as
519 setjmp/longjmp will have calls/returns 509 setjmp/longjmp will have calls/returns not match.
520 510
521 :GMEXAMPLECMD: script 511 :GMEXAMPLECMD: script
522 :GMEXAMPLESUBCMD: 512 :GMEXAMPLESUBCMD:
523 include::guest-files.txt[] 513 include::guest-files.txt[]
524 514
525 SEE ALSO 515 SEE ALSO
526 -------- 516 --------
527 linkperf:perf-record[1], linkperf:perf-script- 517 linkperf:perf-record[1], linkperf:perf-script-perl[1],
528 linkperf:perf-script-python[1], linkperf:perf- 518 linkperf:perf-script-python[1], linkperf:perf-intel-pt[1],
529 linkperf:perf-dlfilter[1] 519 linkperf:perf-dlfilter[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.