1PERF-STAT(1) perf Manual PERF-STAT(1)
2
6 perf-stat - Run a command and gather performance counter statistics
7
9 perf stat [-e <EVENT> | --event=EVENT] [-a] <command>
10 perf stat [-e <EVENT> | --event=EVENT] [-a] -- <command> [<options>]
11 perf stat [-e <EVENT> | --event=EVENT] [-a] record [-o file] -- <command> [<options>]
12 perf stat report [-i file]
13
15 This command runs a command and gathers performance counter statistics
16 from it.
17
19 <command>...
20 Any command you can specify in a shell.
21 record
23 See STAT RECORD.
24 report
26 See STAT REPORT.
27 -e, --event=
29 Select the PMU event. Selection can be:
30 • a symbolic event name (use perf list to list all events)
32 • a raw PMU event in the form of rN where N is a hexadecimal
34 value that represents the raw register encoding with the layout
35 of the event control registers as described by entries in
36 /sys/bus/event_sources/devices/cpu/format/*.
37 • a symbolic or raw PMU event followed by an optional colon and a
39 list of event modifiers, e.g., cpu-cycles:p. See the perf-
40 list(1) man page for details on event modifiers.
41 • a symbolically formed event like pmu/param1=0x3,param2/ where
43 param1 and param2 are defined as formats for the PMU in
44 /sys/bus/event_source/devices/<pmu>/format/*
45 'percore' is a event qualifier that sums up the event counts for both
47 hardware threads in a core. For example:
48 perf stat -A -a -e cpu/event,percore=1/,otherevent ...
49 • a symbolically formed event like
51 pmu/config=M,config1=N,config2=K/ where M, N, K are numbers (in
52 decimal, hex, octal format). Acceptable values for each of
53 config, config1 and config2 parameters are defined by
54 corresponding entries in
55 /sys/bus/event_source/devices/<pmu>/format/*
56 Note that the last two syntaxes support prefix and glob matching in
58 the PMU name to simplify creation of events across multiple instances
59 of the same type of PMU in large systems (e.g. memory controller PMUs).
60 Multiple PMU instances are typical for uncore PMUs, so the prefix
61 'uncore_' is also ignored when performing this match.
62 -i, --no-inherit
64 child tasks do not inherit counters
65 -p, --pid=<pid>
67 stat events on existing process id (comma separated list)
68 -t, --tid=<tid>
70 stat events on existing thread id (comma separated list)
71 -b, --bpf-prog
73 stat events on existing bpf program id (comma separated list),
74 requiring root rights. bpftool-prog could be used to find program
75 id all bpf programs in the system. For example:
76 # bpftool prog | head -n 1
78 17247: tracepoint name sys_enter tag 192d548b9d754067 gpl
79 # perf stat -e cycles,instructions --bpf-prog 17247 --timeout 1000
81 Performance counter stats for 'BPF program(s) 17247':
83 85,967 cycles
85 28,982 instructions # 0.34 insn per cycle
86 1.102235068 seconds time elapsed
88 --bpf-counters
90 Use BPF programs to aggregate readings from perf_events. This
91 allows multiple perf-stat sessions that are counting the same
92 metric (cycles, instructions, etc.) to share hardware counters. To
93 use BPF programs on common events by default, use "perf config
94 stat.bpf-counter-events=<list_of_events>".
95 --bpf-attr-map
97 With option "--bpf-counters", different perf-stat sessions share
98 information about shared BPF programs and maps via a pinned
99 hashmap. Use "--bpf-attr-map" to specify the path of this pinned
100 hashmap. The default path is /sys/fs/bpf/perf_attr_map.
101 -a, --all-cpus
103 system-wide collection from all CPUs (default if no target is
104 specified)
105 --no-scale
107 Don’t scale/normalize counter values
108 -d, --detailed
110 print more detailed statistics, can be specified up to 3 times
111 -d: detailed events, L1 and LLC data cache
113 -d -d: more detailed events, dTLB and iTLB events
114 -d -d -d: very detailed events, adding prefetch events
115 -r, --repeat=<n>
117 repeat command and print average + stddev (max: 100). 0 means
118 forever.
119 -B, --big-num
121 print large numbers with thousands' separators according to locale.
122 Enabled by default. Use "--no-big-num" to disable. Default setting
123 can be changed with "perf config stat.big-num=false".
124 -C, --cpu=
126 Count only on the list of CPUs provided. Multiple CPUs can be
127 provided as a comma-separated list with no space: 0,1. Ranges of
128 CPUs are specified with -: 0-2. In per-thread mode, this option is
129 ignored. The -a option is still necessary to activate system-wide
130 monitoring. Default is to count on all CPUs.
131 -A, --no-aggr
133 Do not aggregate counts across all monitored CPUs.
134 -n, --null
136 null run - Don’t start any counters.
137 This can be useful to measure just elapsed wall-clock time - or to
139 assess the raw overhead of perf stat itself, without running any
140 counters.
141 -v, --verbose
143 be more verbose (show counter open errors, etc)
144 -x SEP, --field-separator SEP
146 print counts using a CSV-style output to make it easy to import
147 directly into spreadsheets. Columns are separated by the string
148 specified in SEP.
149 --table
151 Display time for each run (-r option), in a table format, e.g.:
152 $ perf stat --null -r 5 --table perf bench sched pipe
154 Performance counter stats for 'perf bench sched pipe' (5 runs):
156 # Table of individual measurements:
158 5.189 (-0.293) #
159 5.189 (-0.294) #
160 5.186 (-0.296) #
161 5.663 (+0.181) ##
162 6.186 (+0.703) ####
163 # Final result:
165 5.483 +- 0.198 seconds time elapsed ( +- 3.62% )
166 -G name, --cgroup name
168 monitor only in the container (cgroup) called "name". This option
169 is available only in per-cpu mode. The cgroup filesystem must be
170 mounted. All threads belonging to container "name" are monitored
171 when they run on the monitored CPUs. Multiple cgroups can be
172 provided. Each cgroup is applied to the corresponding event, i.e.,
173 first cgroup to first event, second cgroup to second event and so
174 on. It is possible to provide an empty cgroup (monitor all the
175 time) using, e.g., -G foo,,bar. Cgroups must have corresponding
176 events, i.e., they always refer to events defined earlier on the
177 command line. If the user wants to track multiple events for a
178 specific cgroup, the user can use -e e1 -e e2 -G foo,foo or just
179 use -e e1 -e e2 -G foo.
180 If wanting to monitor, say, cycles for a cgroup and also for system
182 wide, this command line can be used: perf stat -e cycles -G cgroup_name
183 -a -e cycles.
184 --for-each-cgroup name
186 Expand event list for each cgroup in "name" (allow multiple cgroups
187 separated by comma). It also support regex patterns to match
188 multiple groups. This has same effect that repeating -e option and
189 -G option for each event x name. This option cannot be used with
190 -G/--cgroup option.
191 -o file, --output file
193 Print the output into the designated file.
194 --append
196 Append to the output file designated with the -o option. Ignored if
197 -o is not specified.
198 --log-fd
200 Log output to fd, instead of stderr. Complementary to --output, and
201 mutually exclusive with it. --append may be used here. Examples:
202 3>results perf stat --log-fd 3 -- $cmd 3>>results perf stat
203 --log-fd 3 --append -- $cmd
204 --control=fifo:ctl-fifo[,ack-fifo], --control=fd:ctl-fd[,ack-fd]
206 ctl-fifo / ack-fifo are opened and used as ctl-fd / ack-fd as
207 follows. Listen on ctl-fd descriptor for command to control
208 measurement (enable: enable events, disable: disable events).
209 Measurements can be started with events disabled using --delay=-1
210 option. Optionally send control command completion (ack\n) to
211 ack-fd descriptor to synchronize with the controlling process.
212 Example of bash shell script to enable and disable events during
213 measurements:
214 #!/bin/bash
216 ctl_dir=/tmp/
218 ctl_fifo=${ctl_dir}perf_ctl.fifo
220 test -p ${ctl_fifo} && unlink ${ctl_fifo}
221 mkfifo ${ctl_fifo}
222 exec {ctl_fd}<>${ctl_fifo}
223 ctl_ack_fifo=${ctl_dir}perf_ctl_ack.fifo
225 test -p ${ctl_ack_fifo} && unlink ${ctl_ack_fifo}
226 mkfifo ${ctl_ack_fifo}
227 exec {ctl_fd_ack}<>${ctl_ack_fifo}
228 perf stat -D -1 -e cpu-cycles -a -I 1000 \
230 --control fd:${ctl_fd},${ctl_fd_ack} \
231 \-- sleep 30 &
232 perf_pid=$!
233 sleep 5 && echo 'enable' >&${ctl_fd} && read -u ${ctl_fd_ack} e1 && echo "enabled(${e1})"
235 sleep 10 && echo 'disable' >&${ctl_fd} && read -u ${ctl_fd_ack} d1 && echo "disabled(${d1})"
236 exec {ctl_fd_ack}>&-
238 unlink ${ctl_ack_fifo}
239 exec {ctl_fd}>&-
241 unlink ${ctl_fifo}
242 wait -n ${perf_pid}
244 exit $?
245 --pre, --post
247 Pre and post measurement hooks, e.g.:
248 perf stat --repeat 10 --null --sync --pre make -s
250 O=defconfig-build/clean -- make -s -j64 O=defconfig-build/ bzImage
251 -I msecs, --interval-print msecs
253 Print count deltas every N milliseconds (minimum: 1ms) The overhead
254 percentage could be high in some cases, for instance with small,
255 sub 100ms intervals. Use with caution. example: perf stat -I 1000
256 -e cycles -a sleep 5
257 If the metric exists, it is calculated by the counts generated in this
259 interval and the metric is printed after #.
260 --interval-count times
262 Print count deltas for fixed number of times. This option should be
263 used together with "-I" option. example: perf stat -I 1000
264 --interval-count 2 -e cycles -a
265 --interval-clear
267 Clear the screen before next interval.
268 --timeout msecs
270 Stop the perf stat session and print count deltas after N
271 milliseconds (minimum: 10 ms). This option is not supported with
272 the "-I" option. example: perf stat --time 2000 -e cycles -a
273 --metric-only
275 Only print computed metrics. Print them in a single line. Don’t
276 show any raw values. Not supported with --per-thread.
277 --per-socket
279 Aggregate counts per processor socket for system-wide mode
280 measurements. This is a useful mode to detect imbalance between
281 sockets. To enable this mode, use --per-socket in addition to -a.
282 (system-wide). The output includes the socket number and the number
283 of online processors on that socket. This is useful to gauge the
284 amount of aggregation.
285 --per-die
287 Aggregate counts per processor die for system-wide mode
288 measurements. This is a useful mode to detect imbalance between
289 dies. To enable this mode, use --per-die in addition to -a.
290 (system-wide). The output includes the die number and the number of
291 online processors on that die. This is useful to gauge the amount
292 of aggregation.
293 --per-core
295 Aggregate counts per physical processor for system-wide mode
296 measurements. This is a useful mode to detect imbalance between
297 physical cores. To enable this mode, use --per-core in addition to
298 -a. (system-wide). The output includes the core number and the
299 number of online logical processors on that physical processor.
300 --per-thread
302 Aggregate counts per monitored threads, when monitoring threads (-t
303 option) or processes (-p option).
304 --per-node
306 Aggregate counts per NUMA nodes for system-wide mode measurements.
307 This is a useful mode to detect imbalance between NUMA nodes. To
308 enable this mode, use --per-node in addition to -a. (system-wide).
309 -D msecs, --delay msecs
311 After starting the program, wait msecs before measuring (-1: start
312 with events disabled). This is useful to filter out the startup
313 phase of the program, which is often very different.
314 -T, --transaction
316 Print statistics of transactional execution if supported.
317 --metric-no-group
319 By default, events to compute a metric are placed in weak groups.
320 The group tries to enforce scheduling all or none of the events.
321 The --metric-no-group option places events outside of groups and
322 may increase the chance of the event being scheduled - leading to
323 more accuracy. However, as events may not be scheduled together
324 accuracy for metrics like instructions per cycle can be lower - as
325 both metrics may no longer be being measured at the same time.
326 --metric-no-merge
328 By default metric events in different weak groups can be shared if
329 one group contains all the events needed by another. In such cases
330 one group will be eliminated reducing event multiplexing and making
331 it so that certain groups of metrics sum to 100%. A downside to
332 sharing a group is that the group may require multiplexing and so
333 accuracy for a small group that need not have multiplexing is
334 lowered. This option forbids the event merging logic from sharing
335 events between groups and may be used to increase accuracy in this
336 case.
337 --quiet
339 Don’t print output. This is useful with perf stat record below to
340 only write data to the perf.data file.
341
343 Stores stat data into perf data file.
344 -o file, --output file
346 Output file name.
347
349 Reads and reports stat data from perf data file.
350 -i file, --input file
352 Input file name.
353 --per-socket
355 Aggregate counts per processor socket for system-wide mode
356 measurements.
357 --per-die
359 Aggregate counts per processor die for system-wide mode
360 measurements.
361 --per-core
363 Aggregate counts per physical processor for system-wide mode
364 measurements.
365 -M, --metrics
367 Print metrics or metricgroups specified in a comma separated list.
368 For a group all metrics from the group are added. The events from
369 the metrics are automatically measured. See perf list output for
370 the possible metrics and metricgroups.
371 -A, --no-aggr
373 Do not aggregate counts across all monitored CPUs.
374 --topdown
376 Print complete top-down metrics supported by the CPU. This allows
377 to determine bottle necks in the CPU pipeline for CPU bound
378 workloads, by breaking the cycles consumed down into frontend
379 bound, backend bound, bad speculation and retiring.
380 Frontend bound means that the CPU cannot fetch and decode instructions
382 fast enough. Backend bound means that computation or memory access is
383 the bottle neck. Bad Speculation means that the CPU wasted cycles due
384 to branch mispredictions and similar issues. Retiring means that the
385 CPU computed without an apparently bottleneck. The bottleneck is only
386 the real bottleneck if the workload is actually bound by the CPU and
387 not by something else.
388 For best results it is usually a good idea to use it with interval mode
390 like -I 1000, as the bottleneck of workloads can change often.
391 This enables --metric-only, unless overridden with --no-metric-only.
393 The following restrictions only apply to older Intel CPUs and Atom, on
395 newer CPUs (IceLake and later) TopDown can be collected for any thread:
396 The top down metrics are collected per core instead of per CPU thread.
398 Per core mode is automatically enabled and -a (global monitoring) is
399 needed, requiring root rights or perf.perf_event_paranoid=-1.
400 Topdown uses the full Performance Monitoring Unit, and needs disabling
402 of the NMI watchdog (as root): echo 0 > /proc/sys/kernel/nmi_watchdog
403 for best results. Otherwise the bottlenecks may be inconsistent on
404 workload with changing phases.
405 To interpret the results it is usually needed to know on which CPUs the
407 workload runs on. If needed the CPUs can be forced using taskset.
408 --td-level
410 Print the top-down statistics that equal to or lower than the input
411 level. It allows users to print the interested top-down metrics
412 level instead of the complete top-down metrics.
413 The availability of the top-down metrics level depends on the hardware.
415 For example, Ice Lake only supports L1 top-down metrics. The Sapphire
416 Rapids supports both L1 and L2 top-down metrics.
417 Default: 0 means the max level that the current hardware support. Error
419 out if the input is higher than the supported max level.
420 --no-merge
422 Do not merge results from same PMUs.
423 When multiple events are created from a single event specification,
425 stat will, by default, aggregate the event counts and show the result
426 in a single row. This option disables that behavior and shows the
427 individual events and counts.
428 Multiple events are created from a single event specification when: 1.
430 Prefix or glob matching is used for the PMU name. 2. Aliases, which are
431 listed immediately after the Kernel PMU events by perf list, are used.
432 --smi-cost
434 Measure SMI cost if msr/aperf/ and msr/smi/ events are supported.
435 During the measurement, the /sys/device/cpu/freeze_on_smi will be set
437 to freeze core counters on SMI. The aperf counter will not be effected
438 by the setting. The cost of SMI can be measured by (aperf - unhalted
439 core cycles).
440 In practice, the percentages of SMI cycles is very useful for
442 performance oriented analysis. --metric_only will be applied by
443 default. The output is SMI cycles%, equals to (aperf - unhalted core
444 cycles) / aperf
445 Users who wants to get the actual value can apply --no-metric-only.
447 --all-kernel
449 Configure all used events to run in kernel space.
450 --all-user
452 Configure all used events to run in user space.
453 --percore-show-thread
455 The event modifier "percore" has supported to sum up the event
456 counts for all hardware threads in a core and show the counts per
457 core.
458 This option with event modifier "percore" enabled also sums up the
460 event counts for all hardware threads in a core but show the sum counts
461 per hardware thread. This is essentially a replacement for the any bit
462 and convenient for post processing.
463 --summary
465 Print summary for interval mode (-I).
466 --no-csv-summary
468 Don’t print summary at the first column for CVS summary output.
469 This option must be used with -x and --summary.
470 This option can be enabled in perf config by setting the variable
472 stat.no-csv-summary.
473 $ perf config stat.no-csv-summary=true
475 --cputype
477 Only enable events on applying cpu with this type for hybrid
478 platform (e.g. core or atom)"
479
481 $ perf stat -- make
482 Performance counter stats for 'make':
484 83723.452481 task-clock:u (msec) # 1.004 CPUs utilized
486 0 context-switches:u # 0.000 K/sec
487 0 cpu-migrations:u # 0.000 K/sec
488 3,228,188 page-faults:u # 0.039 M/sec
489 229,570,665,834 cycles:u # 2.742 GHz
490 313,163,853,778 instructions:u # 1.36 insn per cycle
491 69,704,684,856 branches:u # 832.559 M/sec
492 2,078,861,393 branch-misses:u # 2.98% of all branches
493 83.409183620 seconds time elapsed
495 74.684747000 seconds user
497 8.739217000 seconds sys
498
500 As displayed in the example above we can display 3 types of timings. We
501 always display the time the counters were enabled/alive:
502 83.409183620 seconds time elapsed
504 For workload sessions we also display time the workloads spent in
506 user/system lands:
507 74.684747000 seconds user
509 8.739217000 seconds sys
510 Those times are the very same as displayed by the time tool.
512
514 With -x, perf stat is able to output a not-quite-CSV format output
515 Commas in the output are not put into "". To make it easy to parse it
516 is recommended to use a different character like -x \;
517 The fields are in this order:
519 • optional usec time stamp in fractions of second (with -I xxx)
521 • optional CPU, core, or socket identifier
523 • optional number of logical CPUs aggregated
525 • counter value
527 • unit of the counter value or empty
529 • event name
531 • run time of counter
533 • percentage of measurement time the counter was running
535 • optional variance if multiple values are collected with -r
537 • optional metric value
539 • optional unit of metric
541 Additional metrics may be printed with all earlier fields being empty.
543
545 Support for Intel hybrid events within perf tools.
546 For some Intel platforms, such as AlderLake, which is hybrid platform
548 and it consists of atom cpu and core cpu. Each cpu has dedicated event
549 list. Part of events are available on core cpu, part of events are
550 available on atom cpu and even part of events are available on both.
551 Kernel exports two new cpu pmus via sysfs: /sys/devices/cpu_core
553 /sys/devices/cpu_atom
554 The cpus files are created under the directories. For example,
556 cat /sys/devices/cpu_core/cpus 0-15
558 cat /sys/devices/cpu_atom/cpus 16-23
560 It indicates cpu0-cpu15 are core cpus and cpu16-cpu23 are atom cpus.
562 Quickstart
564
566 As before, use perf-list to list the symbolic event.
567 perf list
569 inst_retired.any [Fixed Counter: Counts the number of instructions
571 retired. Unit: cpu_atom] inst_retired.any [Number of instructions
572 retired. Fixed Counter - architectural event. Unit: cpu_core]
573 The Unit: xxx is added to brief description to indicate which pmu the
575 event is belong to. Same event name but with different pmu can be
576 supported.
577
579 To enable a core only event or atom only event, following syntax is
580 supported:
581 cpu_core/<event name>/
583 or
584 cpu_atom/<event name>/
585 For example, count the cycles event on core cpus.
587 perf stat -e cpu_core/cycles/
589
591 When creating one event and the event is available on both atom and
592 core, two events are created automatically. One is for atom, the other
593 is for core. Most of hardware events and cache events are available on
594 both cpu_core and cpu_atom.
595 For hardware events, they have pre-defined configs (e.g. 0 for cycles).
597 But on hybrid platform, kernel needs to know where the event comes from
598 (from atom or from core). The original perf event type
599 PERF_TYPE_HARDWARE can’t carry pmu information. So now this type is
600 extended to be PMU aware type. The PMU type ID is stored at
601 attr.config[63:32].
602 PMU type ID is retrieved from sysfs. /sys/devices/cpu_atom/type
604 /sys/devices/cpu_core/type
605 The new attr.config layout for PERF_TYPE_HARDWARE:
607 PERF_TYPE_HARDWARE: 0xEEEEEEEE000000AA AA: hardware event ID EEEEEEEE:
609 PMU type ID
610 Cache event is similar. The type PERF_TYPE_HW_CACHE is extended to be
612 PMU aware type. The PMU type ID is stored at attr.config[63:32].
613 The new attr.config layout for PERF_TYPE_HW_CACHE:
615 PERF_TYPE_HW_CACHE: 0xEEEEEEEE00DDCCBB BB: hardware cache ID CC:
617 hardware cache op ID DD: hardware cache op result ID EEEEEEEE: PMU type
618 ID
619 When enabling a hardware event without specified pmu, such as, perf
621 stat -e cycles -a (use system-wide in this example), two events are
622 created automatically.
623 ------------------------------------------------------------
625 perf_event_attr:
626 size 120
627 config 0x400000000
628 sample_type IDENTIFIER
629 read_format TOTAL_TIME_ENABLED|TOTAL_TIME_RUNNING
630 disabled 1
631 inherit 1
632 exclude_guest 1
633 ------------------------------------------------------------
634 and
636 ------------------------------------------------------------
638 perf_event_attr:
639 size 120
640 config 0x800000000
641 sample_type IDENTIFIER
642 read_format TOTAL_TIME_ENABLED|TOTAL_TIME_RUNNING
643 disabled 1
644 inherit 1
645 exclude_guest 1
646 ------------------------------------------------------------
647 type 0 is PERF_TYPE_HARDWARE. 0x4 in 0x400000000 indicates it’s
649 cpu_core pmu. 0x8 in 0x800000000 indicates it’s cpu_atom pmu (atom pmu
650 type id is random).
651 The kernel creates cycles (0x400000000) on cpu0-cpu15 (core cpus), and
653 create cycles (0x800000000) on cpu16-cpu23 (atom cpus).
654 For perf-stat result, it displays two events:
656 Performance counter stats for 'system wide':
658 6,744,979 cpu_core/cycles/
660 1,965,552 cpu_atom/cycles/
661 The first cycles is core event, the second cycles is atom event.
663
665 perf-stat reports the scaled counts for hybrid event and with a
666 percentage displayed. The percentage is the event’s running
667 time/enabling time.
668 One example, triad_loop runs on cpu16 (atom core), while we can see the
670 scaled value for core cycles is 160,444,092 and the percentage is
671 0.47%.
672 perf stat -e cycles -- taskset -c 16 ./triad_loop
674 As previous, two events are created.
676 .ft C
679 perf_event_attr:
680 size 120
681 config 0x400000000
682 sample_type IDENTIFIER
683 read_format TOTAL_TIME_ENABLED|TOTAL_TIME_RUNNING
684 disabled 1
685 inherit 1
686 enable_on_exec 1
687 exclude_guest 1
688 .ft
689 and
692 .ft C
695 perf_event_attr:
696 size 120
697 config 0x800000000
698 sample_type IDENTIFIER
699 read_format TOTAL_TIME_ENABLED|TOTAL_TIME_RUNNING
700 disabled 1
701 inherit 1
702 enable_on_exec 1
703 exclude_guest 1
704 .ft
705 Performance counter stats for 'taskset -c 16 ./triad_loop':
708 233,066,666 cpu_core/cycles/ (0.43%)
710 604,097,080 cpu_atom/cycles/ (99.57%)
711
713 If there is no -e specified in perf record, on hybrid platform, it
714 creates two default cycles and adds them to event list. One is for
715 core, the other is for atom.
716
718 If there is no -e specified in perf stat, on hybrid platform, besides
719 of software events, following events are created and added to event
720 list in order.
721 cpu_core/cycles/, cpu_atom/cycles/, cpu_core/instructions/,
723 cpu_atom/instructions/, cpu_core/branches/, cpu_atom/branches/,
724 cpu_core/branch-misses/, cpu_atom/branch-misses/
725 Of course, both perf-stat and perf-record support to enable hybrid
727 event with a specific pmu.
728 e.g. perf stat -e cpu_core/cycles/ perf stat -e cpu_atom/cycles/ perf
730 stat -e cpu_core/r1a/ perf stat -e cpu_atom/L1-icache-loads/ perf stat
731 -e cpu_core/cycles/,cpu_atom/instructions/ perf stat -e
732 {cpu_core/cycles/,cpu_core/instructions/}
733 But {cpu_core/cycles/,cpu_atom/instructions/} will return warning and
735 disable grouping, because the pmus in group are not matched (cpu_core
736 vs. cpu_atom).
737
739 perf-top(1), perf-list(1)
740perf 06/14/2022 PERF-STAT(1)