1perf-record(1)
2==============
3
4NAME
5----
6perf-record - Run a command and record its profile into perf.data
7
8SYNOPSIS
9--------
10[verse]
11'perf record' [-e <EVENT> | --event=EVENT] [-a] <command>
12'perf record' [-e <EVENT> | --event=EVENT] [-a] \-- <command> [<options>]
13
14DESCRIPTION
15-----------
16This command runs a command and gathers a performance counter profile
17from it, into perf.data - without displaying anything.
18
19This file can then be inspected later on, using 'perf report'.
20
21
22OPTIONS
23-------
24<command>...::
25	Any command you can specify in a shell.
26
27-e::
28--event=::
29	Select the PMU event. Selection can be:
30
31        - a symbolic event name	(use 'perf list' to list all events)
32
33        - a raw PMU event in the form of rN where N is a hexadecimal value
34          that represents the raw register encoding with the layout of the
35          event control registers as described by entries in
36          /sys/bus/event_source/devices/cpu/format/*.
37
38        - a symbolic or raw PMU event followed by an optional colon
39	  and a list of event modifiers, e.g., cpu-cycles:p.  See the
40	  linkperf:perf-list[1] man page for details on event modifiers.
41
42	- a symbolically formed PMU event like 'pmu/param1=0x3,param2/' where
43	  'param1', 'param2', etc are defined as formats for the PMU in
44	  /sys/bus/event_source/devices/<pmu>/format/*.
45
46	- a symbolically formed event like 'pmu/config=M,config1=N,config3=K/'
47
48          where M, N, K are numbers (in decimal, hex, octal format). Acceptable
49          values for each of 'config', 'config1' and 'config2' are defined by
50          corresponding entries in /sys/bus/event_source/devices/<pmu>/format/*
51          param1 and param2 are defined as formats for the PMU in:
52          /sys/bus/event_source/devices/<pmu>/format/*
53
54	  There are also some parameters which are not defined in .../<pmu>/format/*.
55	  These params can be used to overload default config values per event.
56	  Here are some common parameters:
57	  - 'period': Set event sampling period
58	  - 'freq': Set event sampling frequency
59	  - 'time': Disable/enable time stamping. Acceptable values are 1 for
60		    enabling time stamping. 0 for disabling time stamping.
61		    The default is 1.
62	  - 'call-graph': Disable/enable callgraph. Acceptable str are "fp" for
63			 FP mode, "dwarf" for DWARF mode, "lbr" for LBR mode and
64			 "no" for disable callgraph.
65	  - 'stack-size': user stack size for dwarf mode
66	  - 'name' : User defined event name. Single quotes (') may be used to
67		    escape symbols in the name from parsing by shell and tool
68		    like this: name=\'CPU_CLK_UNHALTED.THREAD:cmask=0x1\'.
69	  - 'aux-output': Generate AUX records instead of events. This requires
70			  that an AUX area event is also provided.
71	  - 'aux-sample-size': Set sample size for AUX area sampling. If the
72	  '--aux-sample' option has been used, set aux-sample-size=0 to disable
73	  AUX area sampling for the event.
74
75          See the linkperf:perf-list[1] man page for more parameters.
76
77	  Note: If user explicitly sets options which conflict with the params,
78	  the value set by the parameters will be overridden.
79
80	  Also not defined in .../<pmu>/format/* are PMU driver specific
81	  configuration parameters.  Any configuration parameter preceded by
82	  the letter '@' is not interpreted in user space and sent down directly
83	  to the PMU driver.  For example:
84
85	  perf record -e some_event/@cfg1,@cfg2=config/ ...
86
87	  will see 'cfg1' and 'cfg2=config' pushed to the PMU driver associated
88	  with the event for further processing.  There is no restriction on
89	  what the configuration parameters are, as long as their semantic is
90	  understood and supported by the PMU driver.
91
92        - a hardware breakpoint event in the form of '\mem:addr[/len][:access]'
93          where addr is the address in memory you want to break in.
94          Access is the memory access type (read, write, execute) it can
95          be passed as follows: '\mem:addr[:[r][w][x]]'. len is the range,
96          number of bytes from specified addr, which the breakpoint will cover.
97          If you want to profile read-write accesses in 0x1000, just set
98          'mem:0x1000:rw'.
99          If you want to profile write accesses in [0x1000~1008), just set
100          'mem:0x1000/8:w'.
101
102	- a group of events surrounded by a pair of brace ("{event1,event2,...}").
103	  Each event is separated by commas and the group should be quoted to
104	  prevent the shell interpretation.  You also need to use --group on
105	  "perf report" to view group events together.
106
107--filter=<filter>::
108	Event filter.  This option should follow an event selector (-e).
109	If the event is a tracepoint, the filter string will be parsed by
110	the kernel.  If the event is a hardware trace PMU (e.g. Intel PT
111	or CoreSight), it'll be processed as an address filter.  Otherwise
112	it means a general filter using BPF which can be applied for any
113	kind of event.
114
115	- tracepoint filters
116
117	In the case of tracepoints, multiple '--filter' options are combined
118	using '&&'.
119
120	- address filters
121
122	A hardware trace PMU advertises its ability to accept a number of
123	address filters	by specifying a non-zero value in
124	/sys/bus/event_source/devices/<pmu>/nr_addr_filters.
125
126	Address filters have the format:
127
128	filter|start|stop|tracestop <start> [/ <size>] [@<file name>]
129
130	Where:
131	- 'filter': defines a region that will be traced.
132	- 'start': defines an address at which tracing will begin.
133	- 'stop': defines an address at which tracing will stop.
134	- 'tracestop': defines a region in which tracing will stop.
135
136	<file name> is the name of the object file, <start> is the offset to the
137	code to trace in that file, and <size> is the size of the region to
138	trace. 'start' and 'stop' filters need not specify a <size>.
139
140	If no object file is specified then the kernel is assumed, in which case
141	the start address must be a current kernel memory address.
142
143	<start> can also be specified by providing the name of a symbol. If the
144	symbol name is not unique, it can be disambiguated by inserting #n where
145	'n' selects the n'th symbol in address order. Alternately #0, #g or #G
146	select only a global symbol. <size> can also be specified by providing
147	the name of a symbol, in which case the size is calculated to the end
148	of that symbol. For 'filter' and 'tracestop' filters, if <size> is
149	omitted and <start> is a symbol, then the size is calculated to the end
150	of that symbol.
151
152	If <size> is omitted and <start> is '*', then the start and size will
153	be calculated from the first and last symbols, i.e. to trace the whole
154	file.
155
156	If symbol names (or '*') are provided, they must be surrounded by white
157	space.
158
159	The filter passed to the kernel is not necessarily the same as entered.
160	To see the filter that is passed, use the -v option.
161
162	The kernel may not be able to configure a trace region if it is not
163	within a single mapping.  MMAP events (or /proc/<pid>/maps) can be
164	examined to determine if that is a possibility.
165
166	Multiple filters can be separated with space or comma.
167
168	- bpf filters
169
170	A BPF filter can access the sample data and make a decision based on the
171	data.  Users need to set an appropriate sample type to use the BPF
172	filter.  BPF filters need root privilege.
173
174	The sample data field can be specified in lower case letter.  Multiple
175	filters can be separated with comma.  For example,
176
177	  --filter 'period > 1000, cpu == 1'
178	or
179	  --filter 'mem_op == load || mem_op == store, mem_lvl > l1'
180
181	The former filter only accept samples with period greater than 1000 AND
182	CPU number is 1.  The latter one accepts either load and store memory
183	operations but it should have memory level above the L1.  Since the
184	mem_op and mem_lvl fields come from the (memory) data_source, it'd only
185	work with some events which set the data_source field.
186
187	Also user should request to collect that information (with -d option in
188	the above case).  Otherwise, the following message will be shown.
189
190	  $ sudo perf record -e cycles --filter 'mem_op == load'
191	  Error: cycles event does not have PERF_SAMPLE_DATA_SRC
192	   Hint: please add -d option to perf record.
193	  failed to set filter "BPF" on event cycles with 22 (Invalid argument)
194
195	Essentially the BPF filter expression is:
196
197	  <term> <operator> <value> (("," | "||") <term> <operator> <value>)*
198
199	The <term> can be one of:
200	  ip, id, tid, pid, cpu, time, addr, period, txn, weight, phys_addr,
201	  code_pgsz, data_pgsz, weight1, weight2, weight3, ins_lat, retire_lat,
202	  p_stage_cyc, mem_op, mem_lvl, mem_snoop, mem_remote, mem_lock,
203	  mem_dtlb, mem_blk, mem_hops
204
205	The <operator> can be one of:
206	  ==, !=, >, >=, <, <=, &
207
208	The <value> can be one of:
209	  <number> (for any term)
210	  na, load, store, pfetch, exec (for mem_op)
211	  l1, l2, l3, l4, cxl, io, any_cache, lfb, ram, pmem (for mem_lvl)
212	  na, none, hit, miss, hitm, fwd, peer (for mem_snoop)
213	  remote (for mem_remote)
214	  na, locked (for mem_locked)
215	  na, l1_hit, l1_miss, l2_hit, l2_miss, any_hit, any_miss, walk, fault (for mem_dtlb)
216	  na, by_data, by_addr (for mem_blk)
217	  hops0, hops1, hops2, hops3 (for mem_hops)
218
219--exclude-perf::
220	Don't record events issued by perf itself. This option should follow
221	an event selector (-e) which selects tracepoint event(s). It adds a
222	filter expression 'common_pid != $PERFPID' to filters. If other
223	'--filter' exists, the new filter expression will be combined with
224	them by '&&'.
225
226-a::
227--all-cpus::
228        System-wide collection from all CPUs (default if no target is specified).
229
230-p::
231--pid=::
232	Record events on existing process ID (comma separated list).
233
234-t::
235--tid=::
236        Record events on existing thread ID (comma separated list).
237        This option also disables inheritance by default.  Enable it by adding
238        --inherit.
239
240-u::
241--uid=::
242        Record events in threads owned by uid. Name or number.
243
244-r::
245--realtime=::
246	Collect data with this RT SCHED_FIFO priority.
247
248--no-buffering::
249	Collect data without buffering.
250
251-c::
252--count=::
253	Event period to sample.
254
255-o::
256--output=::
257	Output file name.
258
259-i::
260--no-inherit::
261	Child tasks do not inherit counters.
262
263-F::
264--freq=::
265	Profile at this frequency. Use 'max' to use the currently maximum
266	allowed frequency, i.e. the value in the kernel.perf_event_max_sample_rate
267	sysctl. Will throttle down to the currently maximum allowed frequency.
268	See --strict-freq.
269
270--strict-freq::
271	Fail if the specified frequency can't be used.
272
273-m::
274--mmap-pages=::
275	Number of mmap data pages (must be a power of two) or size
276	specification with appended unit character - B/K/M/G. The
277	size is rounded up to have nearest pages power of two value.
278	Also, by adding a comma, the number of mmap pages for AUX
279	area tracing can be specified.
280
281-g::
282	Enables call-graph (stack chain/backtrace) recording for both
283	kernel space and user space.
284
285--call-graph::
286	Setup and enable call-graph (stack chain/backtrace) recording,
287	implies -g.  Default is "fp" (for user space).
288
289	The unwinding method used for kernel space is dependent on the
290	unwinder used by the active kernel configuration, i.e
291	CONFIG_UNWINDER_FRAME_POINTER (fp) or CONFIG_UNWINDER_ORC (orc)
292
293	Any option specified here controls the method used for user space.
294
295	Valid options are "fp" (frame pointer), "dwarf" (DWARF's CFI -
296	Call Frame Information) or "lbr" (Hardware Last Branch Record
297	facility).
298
299	In some systems, where binaries are build with gcc
300	--fomit-frame-pointer, using the "fp" method will produce bogus
301	call graphs, using "dwarf", if available (perf tools linked to
302	the libunwind or libdw library) should be used instead.
303	Using the "lbr" method doesn't require any compiler options. It
304	will produce call graphs from the hardware LBR registers. The
305	main limitation is that it is only available on new Intel
306	platforms, such as Haswell. It can only get user call chain. It
307	doesn't work with branch stack sampling at the same time.
308
309	When "dwarf" recording is used, perf also records (user) stack dump
310	when sampled.  Default size of the stack dump is 8192 (bytes).
311	User can change the size by passing the size after comma like
312	"--call-graph dwarf,4096".
313
314	When "fp" recording is used, perf tries to save stack enties
315	up to the number specified in sysctl.kernel.perf_event_max_stack
316	by default.  User can change the number by passing it after comma
317	like "--call-graph fp,32".
318
319-q::
320--quiet::
321	Don't print any warnings or messages, useful for scripting.
322
323-v::
324--verbose::
325	Be more verbose (show counter open errors, etc).
326
327-s::
328--stat::
329	Record per-thread event counts.  Use it with 'perf report -T' to see
330	the values.
331
332-d::
333--data::
334	Record the sample virtual addresses.
335
336--phys-data::
337	Record the sample physical addresses.
338
339--data-page-size::
340	Record the sampled data address data page size.
341
342--code-page-size::
343	Record the sampled code address (ip) page size
344
345-T::
346--timestamp::
347	Record the sample timestamps. Use it with 'perf report -D' to see the
348	timestamps, for instance.
349
350-P::
351--period::
352	Record the sample period.
353
354--sample-cpu::
355	Record the sample cpu.
356
357--sample-identifier::
358	Record the sample identifier i.e. PERF_SAMPLE_IDENTIFIER bit set in
359	the sample_type member of the struct perf_event_attr argument to the
360	perf_event_open system call.
361
362-n::
363--no-samples::
364	Don't sample.
365
366-R::
367--raw-samples::
368Collect raw sample records from all opened counters (default for tracepoint counters).
369
370-C::
371--cpu::
372Collect samples only on the list of CPUs provided. Multiple CPUs can be provided as a
373comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-2.
374In per-thread mode with inheritance mode on (default), samples are captured only when
375the thread executes on the designated CPUs. Default is to monitor all CPUs.
376
377-B::
378--no-buildid::
379Do not save the build ids of binaries in the perf.data files. This skips
380post processing after recording, which sometimes makes the final step in
381the recording process to take a long time, as it needs to process all
382events looking for mmap records. The downside is that it can misresolve
383symbols if the workload binaries used when recording get locally rebuilt
384or upgraded, because the only key available in this case is the
385pathname. You can also set the "record.build-id" config variable to
386'skip to have this behaviour permanently.
387
388-N::
389--no-buildid-cache::
390Do not update the buildid cache. This saves some overhead in situations
391where the information in the perf.data file (which includes buildids)
392is sufficient.  You can also set the "record.build-id" config variable to
393'no-cache' to have the same effect.
394
395-G name,...::
396--cgroup name,...::
397monitor only in the container (cgroup) called "name". This option is available only
398in per-cpu mode. The cgroup filesystem must be mounted. All threads belonging to
399container "name" are monitored when they run on the monitored CPUs. Multiple cgroups
400can be provided. Each cgroup is applied to the corresponding event, i.e., first cgroup
401to first event, second cgroup to second event and so on. It is possible to provide
402an empty cgroup (monitor all the time) using, e.g., -G foo,,bar. Cgroups must have
403corresponding events, i.e., they always refer to events defined earlier on the command
404line. If the user wants to track multiple events for a specific cgroup, the user can
405use '-e e1 -e e2 -G foo,foo' or just use '-e e1 -e e2 -G foo'.
406
407If wanting to monitor, say, 'cycles' for a cgroup and also for system wide, this
408command line can be used: 'perf stat -e cycles -G cgroup_name -a -e cycles'.
409
410-b::
411--branch-any::
412Enable taken branch stack sampling. Any type of taken branch may be sampled.
413This is a shortcut for --branch-filter any. See --branch-filter for more infos.
414
415-j::
416--branch-filter::
417Enable taken branch stack sampling. Each sample captures a series of consecutive
418taken branches. The number of branches captured with each sample depends on the
419underlying hardware, the type of branches of interest, and the executed code.
420It is possible to select the types of branches captured by enabling filters. The
421following filters are defined:
422
423        - any:  any type of branches
424        - any_call: any function call or system call
425        - any_ret: any function return or system call return
426        - ind_call: any indirect branch
427        - ind_jmp: any indirect jump
428        - call: direct calls, including far (to/from kernel) calls
429        - u:  only when the branch target is at the user level
430        - k: only when the branch target is in the kernel
431        - hv: only when the target is at the hypervisor level
432	- in_tx: only when the target is in a hardware transaction
433	- no_tx: only when the target is not in a hardware transaction
434	- abort_tx: only when the target is a hardware transaction abort
435	- cond: conditional branches
436	- call_stack: save call stack
437	- no_flags: don't save branch flags e.g prediction, misprediction etc
438	- no_cycles: don't save branch cycles
439	- hw_index: save branch hardware index
440	- save_type: save branch type during sampling in case binary is not available later
441		     For the platforms with Intel Arch LBR support (12th-Gen+ client or
442		     4th-Gen Xeon+ server), the save branch type is unconditionally enabled
443		     when the taken branch stack sampling is enabled.
444	- priv: save privilege state during sampling in case binary is not available later
445
446+
447The option requires at least one branch type among any, any_call, any_ret, ind_call, cond.
448The privilege levels may be omitted, in which case, the privilege levels of the associated
449event are applied to the branch filter. Both kernel (k) and hypervisor (hv) privilege
450levels are subject to permissions.  When sampling on multiple events, branch stack sampling
451is enabled for all the sampling events. The sampled branch type is the same for all events.
452The various filters must be specified as a comma separated list: --branch-filter any_ret,u,k
453Note that this feature may not be available on all processors.
454
455-W::
456--weight::
457Enable weightened sampling. An additional weight is recorded per sample and can be
458displayed with the weight and local_weight sort keys.  This currently works for TSX
459abort events and some memory events in precise mode on modern Intel CPUs.
460
461--namespaces::
462Record events of type PERF_RECORD_NAMESPACES.  This enables 'cgroup_id' sort key.
463
464--all-cgroups::
465Record events of type PERF_RECORD_CGROUP.  This enables 'cgroup' sort key.
466
467--transaction::
468Record transaction flags for transaction related events.
469
470--per-thread::
471Use per-thread mmaps.  By default per-cpu mmaps are created.  This option
472overrides that and uses per-thread mmaps.  A side-effect of that is that
473inheritance is automatically disabled.  --per-thread is ignored with a warning
474if combined with -a or -C options.
475
476-D::
477--delay=::
478After starting the program, wait msecs before measuring (-1: start with events
479disabled), or enable events only for specified ranges of msecs (e.g.
480-D 10-20,30-40 means wait 10 msecs, enable for 10 msecs, wait 10 msecs, enable
481for 10 msecs, then stop). Note, delaying enabling of events is useful to filter
482out the startup phase of the program, which is often very different.
483
484-I::
485--intr-regs::
486Capture machine state (registers) at interrupt, i.e., on counter overflows for
487each sample. List of captured registers depends on the architecture. This option
488is off by default. It is possible to select the registers to sample using their
489symbolic names, e.g. on x86, ax, si. To list the available registers use
490--intr-regs=\?. To name registers, pass a comma separated list such as
491--intr-regs=ax,bx. The list of register is architecture dependent.
492
493--user-regs::
494Similar to -I, but capture user registers at sample time. To list the available
495user registers use --user-regs=\?.
496
497--running-time::
498Record running and enabled time for read events (:S)
499
500-k::
501--clockid::
502Sets the clock id to use for the various time fields in the perf_event_type
503records. See clock_gettime(). In particular CLOCK_MONOTONIC and
504CLOCK_MONOTONIC_RAW are supported, some events might also allow
505CLOCK_BOOTTIME, CLOCK_REALTIME and CLOCK_TAI.
506
507-S::
508--snapshot::
509Select AUX area tracing Snapshot Mode. This option is valid only with an
510AUX area tracing event. Optionally, certain snapshot capturing parameters
511can be specified in a string that follows this option:
512
513  - 'e': take one last snapshot on exit; guarantees that there is at least one
514       snapshot in the output file;
515  - <size>: if the PMU supports this, specify the desired snapshot size.
516
517In Snapshot Mode trace data is captured only when signal SIGUSR2 is received
518and on exit if the above 'e' option is given.
519
520--aux-sample[=OPTIONS]::
521Select AUX area sampling. At least one of the events selected by the -e option
522must be an AUX area event. Samples on other events will be created containing
523data from the AUX area. Optionally sample size may be specified, otherwise it
524defaults to 4KiB.
525
526--proc-map-timeout::
527When processing pre-existing threads /proc/XXX/mmap, it may take a long time,
528because the file may be huge. A time out is needed in such cases.
529This option sets the time out limit. The default value is 500 ms.
530
531--switch-events::
532Record context switch events i.e. events of type PERF_RECORD_SWITCH or
533PERF_RECORD_SWITCH_CPU_WIDE. In some cases (e.g. Intel PT, CoreSight or Arm SPE)
534switch events will be enabled automatically, which can be suppressed by
535by the option --no-switch-events.
536
537--vmlinux=PATH::
538Specify vmlinux path which has debuginfo.
539(enabled when BPF prologue is on)
540
541--buildid-all::
542Record build-id of all DSOs regardless whether it's actually hit or not.
543
544--buildid-mmap::
545Record build ids in mmap2 events, disables build id cache (implies --no-buildid).
546
547--aio[=n]::
548Use <n> control blocks in asynchronous (Posix AIO) trace writing mode (default: 1, max: 4).
549Asynchronous mode is supported only when linking Perf tool with libc library
550providing implementation for Posix AIO API.
551
552--affinity=mode::
553Set affinity mask of trace reading thread according to the policy defined by 'mode' value:
554
555  - node - thread affinity mask is set to NUMA node cpu mask of the processed mmap buffer
556  - cpu  - thread affinity mask is set to cpu of the processed mmap buffer
557
558--mmap-flush=number::
559
560Specify minimal number of bytes that is extracted from mmap data pages and
561processed for output. One can specify the number using B/K/M/G suffixes.
562
563The maximal allowed value is a quarter of the size of mmaped data pages.
564
565The default option value is 1 byte which means that every time that the output
566writing thread finds some new data in the mmaped buffer the data is extracted,
567possibly compressed (-z) and written to the output, perf.data or pipe.
568
569Larger data chunks are compressed more effectively in comparison to smaller
570chunks so extraction of larger chunks from the mmap data pages is preferable
571from the perspective of output size reduction.
572
573Also at some cases executing less output write syscalls with bigger data size
574can take less time than executing more output write syscalls with smaller data
575size thus lowering runtime profiling overhead.
576
577-z::
578--compression-level[=n]::
579Produce compressed trace using specified level n (default: 1 - fastest compression,
58022 - smallest trace)
581
582--all-kernel::
583Configure all used events to run in kernel space.
584
585--all-user::
586Configure all used events to run in user space.
587
588--kernel-callchains::
589Collect callchains only from kernel space. I.e. this option sets
590perf_event_attr.exclude_callchain_user to 1.
591
592--user-callchains::
593Collect callchains only from user space. I.e. this option sets
594perf_event_attr.exclude_callchain_kernel to 1.
595
596Don't use both --kernel-callchains and --user-callchains at the same time or no
597callchains will be collected.
598
599--timestamp-filename
600Append timestamp to output file name.
601
602--timestamp-boundary::
603Record timestamp boundary (time of first/last samples).
604
605--switch-output[=mode]::
606Generate multiple perf.data files, timestamp prefixed, switching to a new one
607based on 'mode' value:
608
609  - "signal" - when receiving a SIGUSR2 (default value) or
610  - <size>   - when reaching the size threshold, size is expected to
611               be a number with appended unit character - B/K/M/G
612  - <time>   - when reaching the time threshold, size is expected to
613               be a number with appended unit character - s/m/h/d
614
615               Note: the precision of  the size  threshold  hugely depends
616               on your configuration  - the number and size of  your  ring
617               buffers (-m). It is generally more precise for higher sizes
618               (like >5M), for lower values expect different sizes.
619
620A possible use case is to, given an external event, slice the perf.data file
621that gets then processed, possibly via a perf script, to decide if that
622particular perf.data snapshot should be kept or not.
623
624Implies --timestamp-filename, --no-buildid and --no-buildid-cache.
625The reason for the latter two is to reduce the data file switching
626overhead. You can still switch them on with:
627
628  --switch-output --no-no-buildid  --no-no-buildid-cache
629
630--switch-output-event::
631Events that will cause the switch of the perf.data file, auto-selecting
632--switch-output=signal, the results are similar as internally the side band
633thread will also send a SIGUSR2 to the main one.
634
635Uses the same syntax as --event, it will just not be recorded, serving only to
636switch the perf.data file as soon as the --switch-output event is processed by
637a separate sideband thread.
638
639This sideband thread is also used to other purposes, like processing the
640PERF_RECORD_BPF_EVENT records as they happen, asking the kernel for extra BPF
641information, etc.
642
643--switch-max-files=N::
644
645When rotating perf.data with --switch-output, only keep N files.
646
647--dry-run::
648Parse options then exit. --dry-run can be used to detect errors in cmdline
649options.
650
651'perf record --dry-run -e' can act as a BPF script compiler if llvm.dump-obj
652in config file is set to true.
653
654--synth=TYPE::
655Collect and synthesize given type of events (comma separated).  Note that
656this option controls the synthesis from the /proc filesystem which represent
657task status for pre-existing threads.
658
659Kernel (and some other) events are recorded regardless of the
660choice in this option.  For example, --synth=no would have MMAP events for
661kernel and modules.
662
663Available types are:
664
665  - 'task'    - synthesize FORK and COMM events for each task
666  - 'mmap'    - synthesize MMAP events for each process (implies 'task')
667  - 'cgroup'  - synthesize CGROUP events for each cgroup
668  - 'all'     - synthesize all events (default)
669  - 'no'      - do not synthesize any of the above events
670
671--tail-synthesize::
672Instead of collecting non-sample events (for example, fork, comm, mmap) at
673the beginning of record, collect them during finalizing an output file.
674The collected non-sample events reflects the status of the system when
675record is finished.
676
677--overwrite::
678Makes all events use an overwritable ring buffer. An overwritable ring
679buffer works like a flight recorder: when it gets full, the kernel will
680overwrite the oldest records, that thus will never make it to the
681perf.data file.
682
683When '--overwrite' and '--switch-output' are used perf records and drops
684events until it receives a signal, meaning that something unusual was
685detected that warrants taking a snapshot of the most current events,
686those fitting in the ring buffer at that moment.
687
688'overwrite' attribute can also be set or canceled for an event using
689config terms. For example: 'cycles/overwrite/' and 'instructions/no-overwrite/'.
690
691Implies --tail-synthesize.
692
693--kcore::
694Make a copy of /proc/kcore and place it into a directory with the perf data file.
695
696--max-size=<size>::
697Limit the sample data max size, <size> is expected to be a number with
698appended unit character - B/K/M/G
699
700--num-thread-synthesize::
701	The number of threads to run when synthesizing events for existing processes.
702	By default, the number of threads equals 1.
703
704ifdef::HAVE_LIBPFM[]
705--pfm-events events::
706Select a PMU event using libpfm4 syntax (see http://perfmon2.sf.net)
707including support for event filters. For example '--pfm-events
708inst_retired:any_p:u:c=1:i'. More than one event can be passed to the
709option using the comma separator. Hardware events and generic hardware
710events cannot be mixed together. The latter must be used with the -e
711option. The -e option and this one can be mixed and matched.  Events
712can be grouped using the {} notation.
713endif::HAVE_LIBPFM[]
714
715--control=fifo:ctl-fifo[,ack-fifo]::
716--control=fd:ctl-fd[,ack-fd]::
717ctl-fifo / ack-fifo are opened and used as ctl-fd / ack-fd as follows.
718Listen on ctl-fd descriptor for command to control measurement.
719
720Available commands:
721
722  - 'enable'           : enable events
723  - 'disable'          : disable events
724  - 'enable name'      : enable event 'name'
725  - 'disable name'     : disable event 'name'
726  - 'snapshot'         : AUX area tracing snapshot).
727  - 'stop'             : stop perf record
728  - 'ping'             : ping
729  - 'evlist [-v|-g|-F] : display all events
730
731                         -F  Show just the sample frequency used for each event.
732                         -v  Show all fields.
733                         -g  Show event group information.
734
735Measurements can be started with events disabled using --delay=-1 option. Optionally
736send control command completion ('ack\n') to ack-fd descriptor to synchronize with the
737controlling process.  Example of bash shell script to enable and disable events during
738measurements:
739
740 #!/bin/bash
741
742 ctl_dir=/tmp/
743
744 ctl_fifo=${ctl_dir}perf_ctl.fifo
745 test -p ${ctl_fifo} && unlink ${ctl_fifo}
746 mkfifo ${ctl_fifo}
747 exec {ctl_fd}<>${ctl_fifo}
748
749 ctl_ack_fifo=${ctl_dir}perf_ctl_ack.fifo
750 test -p ${ctl_ack_fifo} && unlink ${ctl_ack_fifo}
751 mkfifo ${ctl_ack_fifo}
752 exec {ctl_fd_ack}<>${ctl_ack_fifo}
753
754 perf record -D -1 -e cpu-cycles -a               \
755             --control fd:${ctl_fd},${ctl_fd_ack} \
756             -- sleep 30 &
757 perf_pid=$!
758
759 sleep 5  && echo 'enable' >&${ctl_fd} && read -u ${ctl_fd_ack} e1 && echo "enabled(${e1})"
760 sleep 10 && echo 'disable' >&${ctl_fd} && read -u ${ctl_fd_ack} d1 && echo "disabled(${d1})"
761
762 exec {ctl_fd_ack}>&-
763 unlink ${ctl_ack_fifo}
764
765 exec {ctl_fd}>&-
766 unlink ${ctl_fifo}
767
768 wait -n ${perf_pid}
769 exit $?
770
771--threads=<spec>::
772Write collected trace data into several data files using parallel threads.
773<spec> value can be user defined list of masks. Masks separated by colon
774define CPUs to be monitored by a thread and affinity mask of that thread
775is separated by slash:
776
777    <cpus mask 1>/<affinity mask 1>:<cpus mask 2>/<affinity mask 2>:...
778
779CPUs or affinity masks must not overlap with other corresponding masks.
780Invalid CPUs are ignored, but masks containing only invalid CPUs are not
781allowed.
782
783For example user specification like the following:
784
785    0,2-4/2-4:1,5-7/5-7
786
787specifies parallel threads layout that consists of two threads,
788the first thread monitors CPUs 0 and 2-4 with the affinity mask 2-4,
789the second monitors CPUs 1 and 5-7 with the affinity mask 5-7.
790
791<spec> value can also be a string meaning predefined parallel threads
792layout:
793
794    - cpu    - create new data streaming thread for every monitored cpu
795    - core   - create new thread to monitor CPUs grouped by a core
796    - package - create new thread to monitor CPUs grouped by a package
797    - numa   - create new threed to monitor CPUs grouped by a NUMA domain
798
799Predefined layouts can be used on systems with large number of CPUs in
800order not to spawn multiple per-cpu streaming threads but still avoid LOST
801events in data directory files. Option specified with no or empty value
802defaults to CPU layout. Masks defined or provided by the option value are
803filtered through the mask provided by -C option.
804
805--debuginfod[=URLs]::
806	Specify debuginfod URL to be used when cacheing perf.data binaries,
807	it follows the same syntax as the DEBUGINFOD_URLS variable, like:
808
809	  http://192.168.122.174:8002
810
811	If the URLs is not specified, the value of DEBUGINFOD_URLS
812	system environment variable is used.
813
814--off-cpu::
815	Enable off-cpu profiling with BPF.  The BPF program will collect
816	task scheduling information with (user) stacktrace and save them
817	as sample data of a software event named "offcpu-time".  The
818	sample period will have the time the task slept in nanoseconds.
819
820	Note that BPF can collect stack traces using frame pointer ("fp")
821	only, as of now.  So the applications built without the frame
822	pointer might see bogus addresses.
823
824include::intel-hybrid.txt[]
825
826SEE ALSO
827--------
828linkperf:perf-stat[1], linkperf:perf-list[1], linkperf:perf-intel-pt[1]
829