linux/tools/perf
Wang Nan 374ce938aa perf tools: Enable pre-event inherit setting by config terms
This patch allows perf record setting event's attr.inherit bit by
config terms like:

  # perf record -e cycles/no-inherit/ ...
  # perf record -e cycles/inherit/ ...

So user can control inherit bit for each event separately.

In following example, a.out fork()s in main then do some complex
CPU intensive computations in both of its children.

Basic result with and without inherit:

  # perf record -e cycles -e instructions ./a.out
  [ perf record: Woken up 9 times to write data ]
  [ perf record: Captured and wrote 2.205 MB perf.data (47920 samples) ]
  # perf report --stdio
  # ...
  # Samples: 23K of event 'cycles'
  # Event count (approx.): 23641752891
  ...
  # Samples: 24K of event 'instructions'
  # Event count (approx.): 30428312415

  # perf record -i -e cycles -e instructions ./a.out
  [ perf record: Woken up 5 times to write data ]
  [ perf record: Captured and wrote 1.111 MB perf.data (24019 samples) ]
  ...
  # Samples: 12K of event 'cycles'
  # Event count (approx.): 11699501775
  ...
  # Samples: 12K of event 'instructions'
  # Event count (approx.): 15058023559

Cancel inherit for one event when globally enable:

  # perf record -e cycles/no-inherit/ -e instructions ./a.out
  [ perf record: Woken up 7 times to write data ]
  [ perf record: Captured and wrote 1.660 MB perf.data (36004 samples) ]
  ...
  # Samples: 12K of event 'cycles/no-inherit/'
  # Event count (approx.): 11895759282
 ...
  # Samples: 24K of event 'instructions'
  # Event count (approx.): 30668000441

Enable inherit for one event when globally disable:

  # perf record -i -e cycles/inherit/ -e instructions ./a.out
  [ perf record: Woken up 7 times to write data ]
  [ perf record: Captured and wrote 1.654 MB perf.data (35868 samples) ]
  ...
  # Samples: 23K of event 'cycles/inherit/'
  # Event count (approx.): 23285400229
  ...
  # Samples: 11K of event 'instructions'
  # Event count (approx.): 14969050259

Committer note:

One can check if the bit was set, in addition to seeing the result in
the perf.data file size as above by doing one of:

  # perf record -e cycles -e instructions -a usleep 1
  [ perf record: Woken up 1 times to write data ]
  [ perf record: Captured and wrote 0.911 MB perf.data (63 samples) ]
  # perf evlist -v
  cycles: size: 112, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|ID|CPU|PERIOD, read_format: ID, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
  instructions: size: 112, config: 0x1, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|ID|CPU|PERIOD, read_format: ID, disabled: 1, inherit: 1, freq: 1, sample_id_all: 1, exclude_guest: 1
  #

So, the inherit bit was set in both, now, if we disable it globally using
--no-inherit:

  # perf record --no-inherit -e cycles -e instructions -a usleep 1
  [ perf record: Woken up 1 times to write data ]
  [ perf record: Captured and wrote 0.910 MB perf.data (56 samples) ]
  # perf evlist -v
  cycles: size: 112, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|ID|CPU|PERIOD, read_format: ID, disabled: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
  instructions: size: 112, config: 0x1, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|ID|CPU|PERIOD, read_format: ID, disabled: 1, freq: 1, sample_id_all: 1, exclude_guest: 1

No inherit bit set, then disabling it and setting just on the cycles event:

  # perf record --no-inherit -e cycles/inherit/ -e instructions -a usleep 1
  [ perf record: Woken up 1 times to write data ]
  [ perf record: Captured and wrote 0.909 MB perf.data (48 samples) ]
  # perf evlist -v
  cycles/inherit/: size: 112, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|ID|CPU|PERIOD, read_format: ID, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
  instructions: size: 112, config: 0x1, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|ID|CPU|PERIOD, read_format: ID, disabled: 1, freq: 1, sample_id_all: 1, exclude_guest: 1
  #

We can see it as well in by using a more verbose level of debug messages in
the tool that sets up the perf_event_attr, 'perf record' in this case:

  [root@zoo ~]# perf record -vv --no-inherit -e cycles/inherit/ -e instructions -a usleep 1
  ------------------------------------------------------------
  perf_event_attr:
    size                             112
    { sample_period, sample_freq }   4000
    sample_type                      IP|TID|TIME|ID|CPU|PERIOD
    read_format                      ID
    disabled                         1
    inherit                          1
    mmap                             1
    comm                             1
    freq                             1
    task                             1
    sample_id_all                    1
    exclude_guest                    1
    mmap2                            1
    comm_exec                        1
  ------------------------------------------------------------
  sys_perf_event_open: pid -1  cpu 0  group_fd -1  flags 0x8
  sys_perf_event_open: pid -1  cpu 1  group_fd -1  flags 0x8
  sys_perf_event_open: pid -1  cpu 2  group_fd -1  flags 0x8
  sys_perf_event_open: pid -1  cpu 3  group_fd -1  flags 0x8
  ------------------------------------------------------------
  perf_event_attr:
    size                             112
    config                           0x1
    { sample_period, sample_freq }   4000
    sample_type                      IP|TID|TIME|ID|CPU|PERIOD
    read_format                      ID
    disabled                         1
    freq                             1
    sample_id_all                    1
    exclude_guest                    1
  ------------------------------------------------------------
  sys_perf_event_open: pid -1  cpu 0  group_fd -1  flags 0x8

<SNIP>

Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@plumgrid.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1446029705-199659-2-git-send-email-wangnan0@huawei.com
[ s/u64/bool/ for the perf_evsel_config_term inherit field - jolsa]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-28 11:19:16 -03:00
..
arch perf tests: Add Intel CQM test 2015-10-05 16:56:07 -03:00
bench perf bench: Use named initializers in the trailer too 2015-10-19 18:17:25 -03:00
config Merge branch 'perf/urgent' into perf/core, to pick up fixes before applying new changes 2015-09-18 09:24:01 +02:00
Documentation perf report: Rename to --show-cpu-utilization 2015-10-26 14:06:04 -03:00
python perf python: Support the PERF_RECORD_SWITCH event 2015-10-07 19:41:50 -03:00
scripts perf tools: Add more documentation to export-to-postgresql.py script 2015-09-28 16:53:07 -03:00
tests perf cpu_map: Add data arg to cpu_map__build_map callback 2015-10-19 18:04:01 -03:00
trace/strace/groups perf trace: Add read/write to the file group 2015-09-04 13:22:06 -03:00
ui perf annotate: Don't die() when finding an invalid config option 2015-10-22 18:10:52 -03:00
util perf tools: Enable pre-event inherit setting by config terms 2015-10-28 11:19:16 -03:00
.gitignore perf tools: Add Intel PT instruction decoder 2015-08-17 11:11:36 -03:00
Build perf trace: Support 'strace' syscall event groups 2015-07-20 15:16:32 -03:00
builtin-annotate.c perf env: Rename some leftovers from rename to perf_env 2015-09-14 12:50:20 -03:00
builtin-bench.c perf bench: Run benchmarks, don't test them 2015-10-19 16:10:25 -03:00
builtin-buildid-cache.c perf buildid: Introduce sysfs/filename__sprintf_build_id 2015-08-28 14:53:50 -03:00
builtin-buildid-list.c perf buildid: Introduce sysfs/filename__sprintf_build_id 2015-08-28 14:53:50 -03:00
builtin-data.c perf data: Support using -f to override perf.data file ownership for 'convert' 2015-04-02 13:18:52 -03:00
builtin-diff.c perf callchain: Allow disabling call graphs per event 2015-08-12 13:20:28 -03:00
builtin-evlist.c perf tools: Introduce usage_with_options_msg() 2015-10-27 09:28:44 -03:00
builtin-help.c perf help: Change 'usage' to 'Usage' for consistency 2015-10-19 16:51:44 -03:00
builtin-inject.c perf inject: Add --strip option to strip out non-synthesized events 2015-09-28 17:19:51 -03:00
builtin-kmem.c perf machine: Add method for common kernel_map(FUNCTION) operation 2015-09-30 18:34:29 -03:00
builtin-kvm.c tools lib api fs: Replace debugfs/tracefs objects interface with fs.c 2015-09-14 12:50:15 -03:00
builtin-list.c perf list: Do event name substring search as last resort when no events found 2015-10-01 12:12:22 -03:00
builtin-lock.c perf machine: Protect the machine->threads with a rwlock 2015-05-08 16:19:27 -03:00
builtin-mem.c perf mem: Fill in the missing session freeing after an error occurs 2015-07-01 17:53:49 -03:00
builtin-probe.c perf tools: Introduce usage_with_options_msg() 2015-10-27 09:28:44 -03:00
builtin-record.c perf tools: Introduce usage_with_options_msg() 2015-10-27 09:28:44 -03:00
builtin-report.c perf report: Rename to --show-cpu-utilization 2015-10-26 14:06:04 -03:00
builtin-sched.c perf tools: Introduce usage_with_options_msg() 2015-10-27 09:28:44 -03:00
builtin-script.c perf tools: Introduce usage_with_options_msg() 2015-10-27 09:28:44 -03:00
builtin-stat.c perf stat: Cache aggregated map entries in extra cpumap 2015-10-27 15:08:07 -03:00
builtin-timechart.c perf tools: Elliminate alignment holes 2015-05-18 10:17:33 -03:00
builtin-top.c perf tools: Improve call graph documents and help messages 2015-10-22 16:23:19 -03:00
builtin-trace.c perf evsel: Propagate error info from tp_format 2015-09-15 09:48:33 -03:00
builtin.h perf tools: Add new 'perf data' command 2015-02-25 12:42:25 -03:00
command-list.txt perf tools: Add new 'perf data' command 2015-02-25 12:42:25 -03:00
CREDITS
design.txt perf tools: Update some code references in design.txt 2014-03-18 18:17:06 -03:00
Makefile perf build: Fix single target build dependency check 2015-06-22 11:27:09 -03:00
Makefile.perf tools build: Build fixdep helper from perf and basic libs 2015-09-28 15:50:55 -03:00
MANIFEST perf tools: Add include/err.h into MANIFEST 2015-09-22 10:47:05 -03:00
perf-archive.sh
perf-completion.sh perf tools: Avoid confusion with preloaded bash function for perf bash completion 2015-03-19 13:53:27 -03:00
perf-read-vdso.c perf tools: Build programs to copy 32-bit compatibility 2014-10-29 10:32:48 -02:00
perf-sys.h perf tools: Move generic barriers out of perf-sys.h 2015-05-08 16:05:08 -03:00
perf-with-kcore.sh perf tools: Fix perf-with-kcore handling of arguments containing spaces 2015-08-06 16:48:27 -03:00
perf.c perf tools: Handle -h and -v options 2015-10-05 16:36:18 -03:00
perf.h perf record: Add ability to name registers to record 2015-08-31 18:01:33 -03:00