Separate thread wait time into 3 parts - sleep, iowait and preempt based
on the prev_state of the last event.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20170113104523.31212-1-namhyung@kernel.org
[ Fix the build on centos:5 where 'wait' shadows a global declaration ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Show length of analyzed sample time and rate of idle task running.
This also takes care of time range given by --time option.
$ perf sched timehist -sI | tail
Samples do not have callchains.
Idle stats:
CPU 0 idle for 930.316 msec ( 92.93%)
CPU 1 idle for 963.614 msec ( 96.25%)
CPU 2 idle for 885.482 msec ( 88.45%)
CPU 3 idle for 938.635 msec ( 93.76%)
Total number of unique tasks: 118
Total number of context switches: 2337
Total run time (msec): 3718.048
Total scheduling time (msec): 1001.131 (x 4)
Suggested-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20161222060350.17655-3-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When --time option is given with a value outside recorded time, the last
sample time (tprev) was set to that value and run time calculation might
be incorrect. This is a problem of the first samples for each cpus
since it would skip the runtime update when tprev is 0. But with --time
option it had non-zero (which is invalid) value so the calculation is
also incorrect.
For example, let's see the followging:
$ perf sched timehist
time cpu task name wait time sch delay run time
[tid/pid] (msec) (msec) (msec)
--------------- ------ ------------------------------ --------- --------- ---------
3195.968367 [0003] <idle> 0.000 0.000 0.000
3195.968386 [0002] Timer[4306/4277] 0.000 0.000 0.018
3195.968397 [0002] Web Content[4277] 0.000 0.000 0.000
3195.968595 [0001] JS Helper[4302/4277] 0.000 0.000 0.000
3195.969217 [0000] <idle> 0.000 0.000 0.621
3195.969251 [0001] kworker/1:1H[291] 0.000 0.000 0.033
The sample starts at 3195.968367 but when I gave a time interval from
3194 to 3196 (in sec) it will calculate the whole 2 second as runtime.
In below, 2 cpus accounted it as runtime, other 2 cpus accounted it as
idle time.
Before:
$ perf sched timehist --time 3194,3196 -s | tail
Idle stats:
CPU 0 idle for 1995.991 msec
CPU 1 idle for 20.793 msec
CPU 2 idle for 30.191 msec
CPU 3 idle for 1999.852 msec
Total number of unique tasks: 23
Total number of context switches: 128
Total run time (msec): 3724.940
After:
$ perf sched timehist --time 3194,3196 -s | tail
Idle stats:
CPU 0 idle for 10.811 msec
CPU 1 idle for 20.793 msec
CPU 2 idle for 30.191 msec
CPU 3 idle for 18.337 msec
Total number of unique tasks: 23
Total number of context switches: 128
Total run time (msec): 18.139
Committer notes:
Further testing:
Before:
Idle stats:
CPU 0 idle for 229.785 msec
CPU 1 idle for 937.944 msec
CPU 2 idle for 188.931 msec
CPU 3 idle for 986.185 msec
After:
# perf sched timehist --time 40602,40603 -s | tail
Idle stats:
CPU 0 idle for 229.785 msec
CPU 1 idle for 175.407 msec
CPU 2 idle for 188.931 msec
CPU 3 idle for 223.657 msec
Total number of unique tasks: 68
Total number of context switches: 814
Total run time (msec): 97.688
# for cpu in `seq 0 3` ; do echo -n "CPU $cpu idle for " ; perf sched timehist --time 40602,40603 | grep "\[000${cpu}\].*\<idle\>" | tr -s ' ' | cut -d' ' -f7 | awk '{entries++ ; s+=$1} END {print s " msec (entries: " entries ")"}' ; done
CPU 0 idle for 229.721 msec (entries: 123)
CPU 1 idle for 175.381 msec (entries: 65)
CPU 2 idle for 188.903 msec (entries: 56)
CPU 3 idle for 223.61 msec (entries: 102)
Difference due to the idle stats being accounted at nanoseconds precision while
the <idle> entries in 'perf sched timehist' are trucated at msec.usec.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Fixes: 853b740711 ("perf sched timehist: Add option to specify time window of interest")
Link: http://lkml.kernel.org/r/20161222060350.17655-2-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Now that the default 'comm_width' value is 30, no need to check that at
print_summary,
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20161222060350.17655-1-namhyung@kernel.org
[ Split from a larger patch ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Current default value is 20 but it's easily changed to a bigger value as
task has a long name and different tid and pid. And it makes the output
not aligned. So change it to have a large value as summary shows.
Committer notes:
Before:
# perf sched record
^C
# perf sched timehist
<SNIP>
40602.770537 [0001] rcuos/2[29] 7.970 0.002 0.020
40602.771512 [0003] <idle> 0.003 0.000 0.986
40602.771586 [0001] <idle> 0.020 0.000 1.049
40602.771606 [0001] qemu-system-x86[3593/3510] 0.000 0.002 0.020
40602.771629 [0003] qemu-system-x86[3510] 0.000 0.003 0.116
40602.771776 [0000] <idle> 0.001 0.000 1.892
<SNIP>
After:
# perf sched timehist
<SNIP>
40602.770537 [0001] rcuos/2[29] 7.970 0.002 0.020
40602.771512 [0003] <idle> 0.003 0.000 0.986
40602.771586 [0001] <idle> 0.020 0.000 1.049
40602.771606 [0001] qemu-system-x86[3593/3510] 0.000 0.002 0.020
40602.771629 [0003] qemu-system-x86[3510] 0.000 0.003 0.116
<SNIP>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20161222060350.17655-1-namhyung@kernel.org
[ Split from a larger patch ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Current default value is 20, but that may change in the future, so make
places where we have 20 hardcoded use 'comm_width'.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20161222060350.17655-1-namhyung@kernel.org
[ Split from a larger patch ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The --idle-hist option is to analyze system idle state so which process
makes cpu to go idle. If this option is specified, non-idle events will
be skipped and processes switching to/from idle will be shown.
This option is mostly useful when used with --summary(-only) option. In
the idle-time summary view, idle time is accounted to previous thread
which is run before idle task.
The example output looks like following:
Idle-time summary
comm parent sched-out idle-time min-idle avg-idle max-idle stddev migrations
(count) (msec) (msec) (msec) (msec) %
--------------------------------------------------------------------------------------------
rcu_preempt[7] 2 95 550.872 0.011 5.798 23.146 7.63 0
migration/1[16] 2 1 15.558 15.558 15.558 15.558 0.00 0
khugepaged[39] 2 1 3.062 3.062 3.062 3.062 0.00 0
kworker/0:1H[124] 2 2 4.728 0.611 2.364 4.116 74.12 0
systemd-journal[167] 1 1 4.510 4.510 4.510 4.510 0.00 0
kworker/u16:3[558] 2 13 74.737 0.080 5.749 12.960 21.96 0
irq/34-iwlwifi[628] 2 21 118.403 0.032 5.638 23.990 24.00 0
kworker/u17:0[673] 2 1 3.523 3.523 3.523 3.523 0.00 0
dbus-daemon[722] 1 1 6.743 6.743 6.743 6.743 0.00 0
ifplugd[741] 1 1 58.826 58.826 58.826 58.826 0.00 0
wpa_supplicant[1490] 1 1 13.302 13.302 13.302 13.302 0.00 0
wpa_actiond[1492] 1 2 4.064 0.168 2.032 3.896 91.72 0
dockerd[1500] 1 1 0.055 0.055 0.055 0.055 0.00 0
...
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Acked-by: David Ahern <dsahern@gmail.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20161208144755.16673-6-namhyung@kernel.org
Link: http://lkml.kernel.org/r/20161213080632.19099-2-namhyung@kernel.org
[ Merged fix sent by Namhyumg, as posted in the second Link: tag ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Sometimes it only focuses on idle-related events like upcoming idle-hist
feature. In this case we don't want to see other event to reduce noise.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: David Ahern <dsahern@gmail.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20161208144755.16673-5-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
In order to investigate the idleness reason, it is necessary to keep the
callchains when entering idle. This can be identified by the
sched:sched_switch event having the next_pid field as 0.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: David Ahern <dsahern@gmail.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20161208144755.16673-4-namhyung@kernel.org
Link: http://lkml.kernel.org/r/20161213080632.19099-1-namhyung@kernel.org
[ Merged fix from Namhyung, see second Link: tag ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The struct idle_time_data is to keep idle stats with callchains entering
to the idle task. The normal thread_runtime calculation is done
transparently since it extends the struct thread_runtime.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: David Ahern <dsahern@gmail.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20161208144755.16673-3-namhyung@kernel.org
[ Align struct field names ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The is_idle_sample() function actually does more than determining
whether sample come from idle task. Split the callchain part into
save_task_callchain() to make it clearer.
Also checking prev_pid from trace data looks preferred than just
checking sample->pid since it's possible, although rare, to have invalid
0 pid/tid on scheduling an exiting task.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: David Ahern <dsahern@gmail.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20161208144755.16673-2-namhyung@kernel.org
[ Remove some needless () in some return statements ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
It treats the idle_max_cpu little bit confusingly IMHO. Let's make it
more straight forward.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: David Ahern <dsahern@gmail.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20161206034010.6499-6-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Sometimes samples have tid of 0 but non-0 pid. It ends up having a new
thread of 0 tid/pid (instead of referring idle task) since tid is used
to search matching task. But I guess it's wrong to use 0 as a tid when
pid is set. This patch uses tid only if it has a non-zero value or same
as pid (of 0).
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20161206034010.6499-4-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The -D/--dump-raw-trace option is in the parent option so no need to
repeat it. Also move -f/--force option to parent as it's common to
handle data file.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: David Ahern <dsahern@gmail.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20161206034010.6499-2-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Arnaldo reported an unhelpful error message when running perf sched
timehist on a file that did not contain sched tracepoints:
[root@jouet ~]# perf sched timehist
No trace sample to read. Did you call 'perf record -R'?
[root@jouet ~]# perf evlist -v
cycles:ppp: size: 112, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|CALLCHAIN|CPU|PERIOD, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
Change the has_traces check to look for the sched_switch event. Analysis
for perf sched timehist requires at least this event.
Now when analyzing a file without sched tracepoints you get:
root@f21-vbox:/tmp$ perf sched timehist
No sched_switch events found. Have you run 'perf sched record'?
Signed-off-by: David Ahern <dsahern@gmail.com>
Reported-and-Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1480451988-43673-1-git-send-email-dsahern@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add handlers for sched:sched_migrate_task event. Total number of
migrations is added to summary display and -M/--migrations can be used
to show migration events.
Signed-off-by: David Ahern <dsahern@gmail.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: http://lkml.kernel.org/r/1480091321-35591-1-git-send-email-dsa@cumulusnetworks.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
For tracepoint events, callchains always contain certain functions.
Sometimes it'd be better to skip those functions as they have no value.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/20161124011114.7102-2-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
If callchains were recorded they are appended to the line with a default stack depth of 5:
1.874569 [0011] gcc[31949] 0.014 0.000 1.148 wait_for_completion_killable <- do_fork <- sys_vfork <- stub_vfork <- __vfork
1.874591 [0010] gcc[31951] 0.000 0.000 0.024 __cond_resched <- _cond_resched <- wait_for_completion <- stop_one_cpu <- sched_exec
1.874603 [0010] migration/10[59] 3.350 0.004 0.011 smpboot_thread_fn <- kthread <- ret_from_fork
1.874604 [0011] <idle> 1.148 0.000 0.035 cpu_startup_entry <- start_secondary
1.874723 [0005] <idle> 0.016 0.000 1.383 cpu_startup_entry <- start_secondary
1.874746 [0005] gcc[31949] 0.153 0.078 0.022 do_wait sys_wait4 <- system_call_fastpath <- __GI___waitpid
--no-call-graph can be used to not show the callchains. --max-stack is used
to control the number of frames shown (default of 5). -x/--excl options can
be used to collapse redundant callchains to get more relevant data on screen.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/20161116060634.28477-7-namhyung@kernel.org
[ Add documentation based on above commit message ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The -s/--summary option is to show process runtime statistics. And the
-S/--with-summary option is to show the stats with the normal output.
$ perf sched timehist -s
Runtime summary
comm parent sched-in run-time min-run avg-run max-run stddev
(count) (msec) (msec) (msec) (msec) %
---------------------------------------------------------------------------------------------------------
ksoftirqd/0[3] 2 2 0.011 0.004 0.005 0.006 14.87
rcu_preempt[7] 2 11 0.071 0.002 0.006 0.017 20.23
watchdog/0[11] 2 1 0.002 0.002 0.002 0.002 0.00
watchdog/1[12] 2 1 0.004 0.004 0.004 0.004 0.00
...
Terminated tasks:
sleep[7220] 7219 3 0.770 0.087 0.256 0.576 62.28
Idle stats:
CPU 0 idle for 2352.006 msec
CPU 1 idle for 2764.497 msec
CPU 2 idle for 2998.229 msec
CPU 3 idle for 2967.800 msec
Total number of unique tasks: 52
Total number of context switches: 2532
Total run time (msec): 218.036
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/20161116060634.28477-5-namhyung@kernel.org
[ Add documentation from last commit, so that docs comes with the cset that introduces the feature ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
'perf sched timehist' provides an analysis of scheduling events.
Example usage:
perf sched record -- sleep 1
perf sched timehist
By default it shows the individual schedule events, including the wait
time (time between sched-out and next sched-in events for the task), the
task scheduling delay (time between wakeup and actually running) and run
time for the task:
time cpu task name wait time sch delay run time
[tid/pid] (msec) (msec) (msec)
-------------- ------ -------------------- --------- --------- ---------
79371.874569 [0011] gcc[31949] 0.014 0.000 1.148
79371.874591 [0010] gcc[31951] 0.000 0.000 0.024
79371.874603 [0010] migration/10[59] 3.350 0.004 0.011
79371.874604 [0011] <idle> 1.148 0.000 0.035
79371.874723 [0005] <idle> 0.016 0.000 1.383
79371.874746 [0005] gcc[31949] 0.153 0.078 0.022
...
Times are in msec.usec.
Committer note:
Add above explanation as the 'perf sched timehist' entry for 'man
perf-sched'.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/20161116060634.28477-4-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Joonwoo reported that there's a mismatch between timestamps in script
and sched commands. This was because of difference in printing the
timestamp. Factor out the code and share it so that they can be in
sync. Also I found that sched map has similar problem, fix it too.
Committer notes:
Fixed the max_lat_at bug introduced by Namhyung's original patch, as
pointed out by Joonwoo, and made it a function following the scnprintf()
model, i.e. returning the number of bytes formatted, and receiving as
the first parameter the object from where the data to the formatting is
obtained, renaming it from:
char *timestamp_in_usec(char *bf, size_t size, u64 timestamp)
to
int timestamp__scnprintf_usec(u64 timestamp, char *bf, size_t size)
Reported-by: Joonwoo Park <joonwoop@codeaurora.org>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20161024020246.14928-3-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
I'd like to see the name of tasks with perf sched map, but it only shows
name of new tasks and then use short names after all. This is not good
for long running tasks since it's hard for users to track the short
names. This patch makes it show the names (except the idle task) when
-v option is used. Probably we may make it as default behavior.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20161024020246.14928-2-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Applying cpu color always doesn't help readability IMHO. Instead it
might be better to applying the color when there's an activity on those
CPUs.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20161024020246.14928-1-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The -i and -v options can be used in subcommands so enable cascading the
sched_options. This fixes the following inconvenience in 'perf sched':
$ perf sched -i perf.data.sched map
... (it works well) ...
$ perf sched map -i perf.data.sched
Error: unknown switch `i'
Usage: perf sched map [<options>]
--color-cpus <cpus>
highlight given CPUs in map
--color-pids <pids>
highlight given pids in map
--compact map output in compact mode
--cpus <cpus> display given CPUs in map
With this patch, the second command line works with the perf.data.sched
data file.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/20161024030003.28534-2-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Probably the next step is to introduce linux/time.h and use
timespec_to_ns(), etc.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-4nqhskn27fn93cz3ukbc8drf@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The tools so far have been using the strerror_r() GNU variant, that
returns a string, be it the buffer passed or something else.
But that, besides being tricky in cases where we expect that the
function using strerror_r() returns the error formatted in a provided
buffer (we have to check if it returned something else and copy that
instead), breaks the build on systems not using glibc, like Alpine
Linux, where musl libc is used.
So, introduce yet another wrapper, str_error_r(), that has the GNU
interface, but uses the portable XSI variant of strerror_r(), so that
users rest asured that the provided buffer is used and it is what is
returned.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-d4t42fnf48ytlk8rjxs822tf@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Introducing --cpus option that will display only given cpus. Could be
used together with color-cpus option.
$ perf sched map --cpus 0,1
*A0 309999.786924 secs A0 => rcu_sched:7
*. 309999.786930 secs
*B0 . 309999.786931 secs B0 => rcuos/2:25
B0 *A0 309999.786947 secs
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1460467771-26532-9-git-send-email-jolsa@kernel.org
[ Added entry to man page ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Adding --color-cpus option to display selected cpus with background
color (red by default). It helps on navigating through the perf sched
map output.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1460467771-26532-8-git-send-email-jolsa@kernel.org
[ Added entry to man page ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Adding --color-pids option to display selected pids in color (blue by
default). It helps on navigating through the 'perf sched map' output.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1460467771-26532-7-git-send-email-jolsa@kernel.org
[ Added entry to man page ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
As preparation for next patch.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1460467771-26532-5-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Move the subcommand-related files from perf to a new library named
libsubcmd.a.
Since we're moving files anyway, go ahead and rename 'exec_cmd.*' to
'exec-cmd.*' to be consistent with the naming of all the other files.
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/c0a838d4c878ab17fee50998811612b2281355c1.1450193761.git.jpoimboe@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The latency subcommand holds a tree of working atoms sorted by thread's
pid/tid. If there's new thread with same pid and tid, the old working atom is
found and assert bug condition is hit in search function:
thread_atoms_search: Assertion `!(thread != atoms->thread)' failed
Changing the sort function to use thread object pointers together with pid and
tid check. This way new thread will never find old one with same pid/tid.
Link: http://lkml.kernel.org/n/tip-o4doazhhv0zax5zshkg8hnys@git.kernel.org
Reported-by: Mohit Agrawal <moagrawa@redhat.com>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1446462625-15807-1-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Now usage_with_options() setup a pager before printing message so normal
printf() or pr_err() will not be shown. The usage_with_options_msg()
can be used to print some help message before usage strings.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1445701767-12731-4-git-send-email-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Sometimes when debugging large multi-threaded applications it is helpful
to collate all of the latency numbers into one bulk record to get an
idea of what is going on.
This patch does this by merging any entries that belong to the same comm
into one entry and then spits out those totals.
I've also slightly changed the output so you can see how many threads
were merged in the processing. Here is the new default output format
-----------------------------------------------------------------------------------------------------------
Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | Maximum delay at |
-----------------------------------------------------------------------------------------------------------
chrome:(23) | 740.878 ms | 2612 | avg: 0.022 ms | max: 0.845 ms | max at: 7935.254223 s
pulseaudio:1523 | 94.440 ms | 597 | avg: 0.027 ms | max: 0.110 ms | max at: 7934.668372 s
threaded-ml:6042 | 72.554 ms | 386 | avg: 0.035 ms | max: 1.186 ms | max at: 7935.330911 s
Chrome_IOThread:3832 | 52.388 ms | 456 | avg: 0.021 ms | max: 1.365 ms | max at: 7935.330602 s
Chrome_ChildIOT:(7) | 50.694 ms | 743 | avg: 0.021 ms | max: 1.448 ms | max at: 7935.256659 s
Compositor:5510 | 30.012 ms | 192 | avg: 0.019 ms | max: 0.131 ms | max at: 7936.636815 s
plugin_audio_th:6043 | 24.828 ms | 314 | avg: 0.018 ms | max: 0.143 ms | max at: 7936.205994 s
CompositorTileW:(2) | 14.099 ms | 45 | avg: 0.022 ms | max: 0.153 ms | max at: 7937.521800 s
the (#) after the task is the number of tasks merged, and then if there were
no tasks merged it just shows the pid. Here is the same trace file with the -p
option to print the per-pid latency numbers
-----------------------------------------------------------------------------------------------------------
Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | Maximum delay at |
-----------------------------------------------------------------------------------------------------------
chrome:5500 | 386.872 ms | 387 | avg: 0.023 ms | max: 0.241 ms | max at: 7936.001694 s
pulseaudio:1523 | 94.440 ms | 597 | avg: 0.027 ms | max: 0.110 ms | max at: 7934.668372 s
threaded-ml:6042 | 72.554 ms | 386 | avg: 0.035 ms | max: 1.186 ms | max at: 7935.330911 s
chrome:10226 | 69.710 ms | 251 | avg: 0.023 ms | max: 0.764 ms | max at: 7935.992305 s
chrome:4267 | 64.551 ms | 418 | avg: 0.021 ms | max: 0.294 ms | max at: 7937.862427 s
chrome:4827 | 62.268 ms | 54 | avg: 0.029 ms | max: 0.666 ms | max at: 7935.992813 s
Chrome_IOThread:3832 | 52.388 ms | 456 | avg: 0.021 ms | max: 1.365 ms | max at: 7935.330602 s
chrome:3776 | 46.150 ms | 349 | avg: 0.023 ms | max: 0.845 ms | max at: 7935.254223 s
Signed-off-by: Josef Bacik <jbacik@fb.com>
Acked-by: Ingo Molnar <mingo@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: kernel-team@fb.com
Link: http://lkml.kernel.org/r/1432300720-30478-1-git-send-email-jbacik@fb.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
In addition to using refcounts for the struct thread lifetime
management, we need to protect access to machine->threads from
concurrent access.
That happens in 'perf top', where a thread processes events, inserting
and deleting entries from that rb_tree while another thread decays
hist_entries, that end up dropping references and ultimately deleting
threads from the rb_tree and releasing its resources when no further
hist_entry (or other data structures, like in 'perf sched') references
it.
So the rule is the same for refcounts + protected trees in the kernel,
get the tree lock, find object, bump the refcount, drop the tree lock,
return, use object, drop the refcount if no more use of it is needed,
keep it if storing it in some other data structure, drop when releasing
that data structure.
I.e. pair "t = machine__find(new)_thread()" with a "thread__put(t)", and
"perf_event__preprocess_sample(&al)" with "addr_location__put(&al)".
The addr_location__put() one is because as we return references to
several data structures, we may end up adding more reference counting
for the other data structures and then we'll drop it at
addr_location__put() time.
Acked-by: David Ahern <dsahern@gmail.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-bs9rt4n0jw3hi9f3zxyy3xln@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Since sched->replay_repeat is set to 10 as default, the sched->run_avg,
sched->runavg_cpu_usage, and sched->runavg_parent_cpu_usage all use
10 to calculate their value.
However, the replay_repeat can be changed to other value by using -r
option, so the calculation above should use replay_repeat to achieve
more accurate results instead of the default value 10.
Signed-off-by: Yunlong Song <yunlong.song@huawei.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1427809596-29559-10-git-send-email-yunlong.song@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The soft maximum number of open files for a calling process is 1024,
which is defined as INR_OPEN_CUR in include/uapi/linux/fs.h, and the
hard maximum number of open files for a calling process is 4096, which
is defined as INR_OPEN_MAX in include/uapi/linux/fs.h.
Both INR_OPEN_CUR and INR_OPEN_MAX are used to limit the value of
RLIMIT_NOFILE in include/asm-generic/resource.h.
And the soft maximum number finally decides the limitation of the
maximum files which are allowed to be opened.
That is to say a process can use at most 1024 file descriptors for its
o pened files, or an EMFILE error will happen.
This error can be fixed by increasing the soft maximum number, under the
constraint that the soft maximum number can not exceed the hard maximum
number, or both soft and hard maximum number should be increased
simultaneously with privilege.
For perf sched replay, it uses sys_perf_event_open to create the file
descriptor for each of the tasks in order to handle information of perf
events.
That is to say each task needs a unique file descriptor. In x86_64,
there may be over 1024 or 4096 tasks correspoinding to the record in
perf.data, which causes that no enough file descriptors can be used.
As a result, EMFILE error happens and stops the replay process. To solve
this problem, we adaptively increase the soft and hard maximum number of
open files with a '-f' option.
Example:
Test environment: x86_64 with 160 cores
$ cat /proc/sys/kernel/pid_max
163840
$ cat /proc/sys/fs/file-max
6815744
$ ulimit -Sn
1024
$ ulimit -Hn
4096
Before this patch:
$ perf sched replay
...
task 1549 ( :163132: 163132), nr_events: 1
task 1550 ( :163540: 163540), nr_events: 1
task 1551 ( <unknown>: 0), nr_events: 10
Error: sys_perf_event_open() syscall returned with -1 (Too many open
files)
After this patch:
$ perf sched replay
...
task 1549 ( :163132: 163132), nr_events: 1
task 1550 ( :163540: 163540), nr_events: 1
task 1551 ( <unknown>: 0), nr_events: 10
Error: sys_perf_event_open() syscall returned with -1 (Too many open
files)
Have a try with -f option
$ perf sched replay -f
...
task 1549 ( :163132: 163132), nr_events: 1
task 1550 ( :163540: 163540), nr_events: 1
task 1551 ( <unknown>: 0), nr_events: 10
------------------------------------------------------------
#1 : 54.401, ravg: 54.40, cpu: 3285.21 / 3285.21
#2 : 199.548, ravg: 68.92, cpu: 4999.65 / 3456.66
#3 : 170.483, ravg: 79.07, cpu: 1349.94 / 3245.99
#4 : 192.034, ravg: 90.37, cpu: 1322.88 / 3053.67
#5 : 182.929, ravg: 99.62, cpu: 1406.51 / 2888.96
#6 : 152.974, ravg: 104.96, cpu: 1167.54 / 2716.82
#7 : 155.579, ravg: 110.02, cpu: 2992.53 / 2744.39
#8 : 130.557, ravg: 112.08, cpu: 1126.43 / 2582.59
#9 : 138.520, ravg: 114.72, cpu: 1253.22 / 2449.65
#10 : 134.328, ravg: 116.68, cpu: 1587.95 / 2363.48
Signed-off-by: Yunlong Song <yunlong.song@huawei.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1427809596-29559-8-git-send-email-yunlong.song@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Since there is sem_wait for each task in the wait_for_tasks(), e.g.
sem_wait(&task->work_done_sem).
The sem_wait can continue only when work_done_sem is greater than 0, or
it will be blocked.
For perf sched replay, one task may sem_post the work_done_sem of
another task, which causes the work_done_sem of that task processed in a
reasonable sequence, e.g. sem_post, sem_wait, sem_wait, sem_post...
This sequence simulates the sched process of the running tasks at the
time when perf sched record runs.
As a result, all the tasks are required and their threads must be
successfully created.
If any one (task A) of the tasks fails to create its thread, then
another task (task B), whose work_done_sem needs sem_post from that
failed task A, may likely block itself due to seg_wait.
And this is a dead halt, since task B's thread_func cannot continue at
all.
To solve this problem, perf sched replay should exit once any task fails
to create its thread.
Example:
Test environment: x86_64 with 160 cores
Before this patch:
$ perf sched replay
...
Error: sys_perf_event_open() syscall returned with -1 (Too many open
files)
------------------------------------------------------------ <- dead halt
After this patch:
$ perf sched replay
...
task 1551 ( <unknown>: 0), nr_events: 10
Error: sys_perf_event_open() syscall returned with -1 (Too many open
files)
$
As shown above, perf sched replay finishes the process after printing an
error message and does not block itself.
Signed-off-by: Yunlong Song <yunlong.song@huawei.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1427809596-29559-7-git-send-email-yunlong.song@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The pr_err in self_open_counters() prints error message to stderr.
Unlike stdout, stderr uses memory buffer on the stack of each calling
process.
The pr_err in self_open_counters() works in a thread called thread_func
created in function create_tasks, which concurrently creates
sched->nr_tasks threads.
If the error happens and pr_err prints the error message in each of
these threads, the stack size of the perf process (default is 8192
kbytes) will quickly run out and the segmentation fault will happen
then.
To solve this problem, pr_err with self_open_counters() should be moved
from newly created threads to the old main thread of the perf process.
Then the pr_err can work in a stable situation without the strange
segmentation fault problem.
Example:
Test environment: x86_64 with 160 cores
Before this patch:
$ perf sched replay
...
task 1549 ( :163132: 163132), nr_events: 1
task 1550 ( :163540: 163540), nr_events: 1
task 1551 ( <unknown>: 0), nr_events: 10
Segmentation fault
After this patch:
$ perf sched replay
...
task 1549 ( :163132: 163132), nr_events: 1
task 1550 ( :163540: 163540), nr_events: 1
task 1551 ( <unknown>: 0), nr_events: 10
...
As shown above, the result continues without any segmentation fault.
Signed-off-by: Yunlong Song <yunlong.song@huawei.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1427809596-29559-6-git-send-email-yunlong.song@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Although the memory of pid_to_task can be allocated via calloc according
to the value of /proc/sys/kernel/pid_max, it cannot handle the case when
pid_max is changed after 'perf sched record' has created its perf.data.
If the new pid_max configured in 'perf sched replay' is smaller than the
old pid_max configured in 'perf sched record', then it will cause the
assertion failure problem.
To solve this problem, we realloc the memory of pid_to_task stepwise
once the passed-in pid parameter in register_pid is larger than the
current pid_max.
Example:
Test environment: x86_64 with 160 cores
$ cat /proc/sys/kernel/pid_max
163840
$ perf sched record ls
$ echo 5000 > /proc/sys/kernel/pid_max
$ cat /proc/sys/kernel/pid_max
5000
Before this patch:
$ perf sched replay
run measurement overhead: 221 nsecs
sleep measurement overhead: 55356 nsecs
the run test took 1000011 nsecs
the sleep test took 1060940 nsecs
perf: builtin-sched.c:337: register_pid: Assertion `!(pid >= (unsigned
long)pid_max)' failed.
Aborted
After this patch:
$ perf sched replay
run measurement overhead: 221 nsecs
sleep measurement overhead: 55611 nsecs
the run test took 1000026 nsecs
the sleep test took 1060486 nsecs
nr_run_events: 10
nr_sleep_events: 1562
nr_wakeup_events: 5
task 0 ( :1: 1), nr_events: 1
task 1 ( :2: 2), nr_events: 1
task 2 ( :3: 3), nr_events: 1
task 3 ( :5: 5), nr_events: 1
...
Signed-off-by: Yunlong Song <yunlong.song@huawei.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1427809596-29559-5-git-send-email-yunlong.song@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The current memory allocation of struct task_desc *pid_to_task[MAX_PID]
is in a permanent and preset way, and it has two problems:
Problem 1: If the pid_max, which is the max number of pids in the
system, is much smaller than MAX_PID (1024*1000), then it causes a waste
of stack memory. This may happen in the case where the number of cpu
cores is much smaller than 1000.
Problem 2: If the pid_max is changed from the default value to a value
larger than MAX_PID, then it will cause assertion failure problem. The
maximum value of pid_max can be set to pid_max_max (see pidmap_init
defined in kernel/pid.c), which equals to PID_MAX_LIMIT. In x86_64,
PID_MAX_LIMIT is 4*1024*1024 (defined in include/linux/threads.h). This
value is much larger than MAX_PID, and will take up 32768 Kbytes
(4*1024*1024*8/1024) for memory allocation of pid_to_task, which is much
larger than the default 8192 Kbytes of the stack size of calling
process.
Due to these two problems, we use calloc to allocate the memory of
pid_to_task dynamically.
Example:
Test environment: x86_64 with 160 cores
$ cat /proc/sys/kernel/pid_max
163840
$ echo 1025000 > /proc/sys/kernel/pid_max
$ cat /proc/sys/kernel/pid_max
1025000
Run some applications until the pid of some process is greater than
the value of MAX_PID (1024*1000).
Before this patch:
$ perf sched replay
run measurement overhead: 221 nsecs
sleep measurement overhead: 55480 nsecs
the run test took 1000008 nsecs
the sleep test took 1063151 nsecs
perf: builtin-sched.c:330: register_pid: Assertion `!(pid >= 1024000)'
failed.
Aborted
After this patch:
$ perf sched replay
run measurement overhead: 221 nsecs
sleep measurement overhead: 55435 nsecs
the run test took 1000004 nsecs
the sleep test took 1059312 nsecs
nr_run_events: 10
nr_sleep_events: 1562
nr_wakeup_events: 5
task 0 ( :1: 1), nr_events: 1
task 1 ( :2: 2), nr_events: 1
task 2 ( :3: 3), nr_events: 1
task 3 ( :5: 5), nr_events: 1
...
Signed-off-by: Yunlong Song <yunlong.song@huawei.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1427809596-29559-4-git-send-email-yunlong.song@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Current MAX_PID is only 65536, which will cause assertion failure problem
when CPU cores are more than 64 in x86_64.
This is because the pid_max value in x86_64 is at least
PIDS_PER_CPU_DEFAULT * num_possible_cpus() (see function pidmap_init
defined in kernel/pid.c), where PIDS_PER_CPU_DEFAULT is 1024 (defined in
include/linux/threads.h).
Thus for MAX_PID = 65536, the correspoinding CPU cores are
65536/1024=64. This is obviously not enough at all for x86_64, and will
cause an assertion failure problem due to BUG_ON(pid >= MAX_PID) in the
codes.
We increase MAX_PID value from 65536 to 1024*1000, which can be used in
x86_64 with 1000 cores.
This number is finally decided according to the limitation of stack size
of calling process.
Use 'ulimit -a', the result shows the stack size of any process is 8192
Kbytes, which is defined in include/uapi/linux/resource.h (#define
_STK_LIM (8*1024*1024)).
Thus we choose a large enough value for MAX_PID, and make it satisfy to
the limitation of the stack size, i.e., making the perf process take up
a memory space just smaller than 8192 Kbytes.
We have calculated and tested that 1024*1000 is OK for MAX_PID.
This means perf sched replay can now be used with at most 1000 cores in
x86_64 without any assertion failure problem.
Example:
Test environment: x86_64 with 160 cores
$ cat /proc/sys/kernel/pid_max
163840
Before this patch:
$ perf sched replay
run measurement overhead: 240 nsecs
sleep measurement overhead: 55379 nsecs
the run test took 1000004 nsecs
the sleep test took 1059424 nsecs
perf: builtin-sched.c:330: register_pid: Assertion `!(pid >= 65536)'
failed.
Aborted
After this patch:
$ perf sched replay
run measurement overhead: 221 nsecs
sleep measurement overhead: 55397 nsecs
the run test took 999920 nsecs
the sleep test took 1053313 nsecs
nr_run_events: 10
nr_sleep_events: 1562
nr_wakeup_events: 5
task 0 ( :1: 1), nr_events: 1
task 1 ( :2: 2), nr_events: 1
task 2 ( :3: 3), nr_events: 1
task 3 ( :5: 5), nr_events: 1
...
Signed-off-by: Yunlong Song <yunlong.song@huawei.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1427809596-29559-3-git-send-email-yunlong.song@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
