Hello,
This patchset is a rebased version of David's sched timehist work [1].
I plan to improve perf sched command more and think that having
timehist command before the work looks good. It seems David is busy
these days, so I'm retrying it by myself.
* changes in v2)
- change name 'b/n time' to 'wait time' (Ingo)
- show arrow between functions in the callchain (Ingo)
- fix a bug in calculating initial run time
This implements only basic feature and a few options. I just split
the patch to make it easier to review and did some cosmetic changes.
More patches will come later.
The below is from the David's original description (w/ slight change):
------------------------8<-------------------------
'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 time between
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[tid/pid] wait time sch delay run time
------------- ---- -------------------- --------- --------- ---------
79371.874569 [11] gcc[31949] 0.014 0.000 1.148
79371.874591 [10] gcc[31951] 0.000 0.000 0.024
79371.874603 [10] migration/10[59] 3.350 0.004 0.011
79371.874604 [11] <idle> 1.148 0.000 0.035
79371.874723 [05] <idle> 0.016 0.000 1.383
79371.874746 [05] gcc[31949] 0.153 0.078 0.022
...
Times are in msec.usec.
If callchains were recorded they are appended to the line with a default stack
depth of 5:
79371.874569 [11] gcc[31949] 0.014 0.000 1.148
wait_for_completion_killable <- do_fork <- sys_vfork <- stub_vfork <- __vfork
79371.874591 [10] gcc[31951] 0.000 0.000 0.024
__cond_resched <- _cond_resched <- wait_for_completion <- stop_one_cpu <-
sched_exec
79371.874603 [10] migration/10[59] 3.350 0.004 0.011
smpboot_thread_fn <- kthread <- ret_from_fork
79371.874604 [11] <idle> 1.148 0.000 0.035
cpu_startup_entry <- start_secondary
79371.874723 [05] <idle> 0.016 0.000 1.383
cpu_startup_entry <- start_secondary
79371.874746 [05] 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.
Similar to perf-trace -s and -S can be used to dump a statistical summary
without or with events (respectively). Statistics include min run time,
average run time and max run time. Stats are also shown for run time by
cpu.
The cpu-visual option provides a visual aid for sched switches by cpu:
...
79371.874569 [11] s gcc[31949] 0.014
0.000 1.148
79371.874591 [10] s gcc[31951] 0.000
0.000 0.024
79371.874603 [10] s migration/10[59] 3.350
0.004 0.011
79371.874604 [11] i <idle> 1.148
0.000 0.035
79371.874723 [05] i <idle> 0.016
0.000 1.383
79371.874746 [05] s gcc[31949] 0.153
0.078 0.022
...
------------------------8<-------------------------
This code is available at 'perf/timehist-v2' branch in my tree
git://git.kernel.org/pub/scm/linux/kernel/git/namhyung/linux-perf.git
Any feedback is welcomed.
Thanks,
Namhyung
[1] https://lkml.org/lkml/2013/12/1/129
David Ahern (6):
perf sched timehist: Introduce timehist command
perf sched timehist: Add summary options
perf sched timehist: Add -w/--wakeups option
perf sched timehist: Add call graph options
perf sched timehist: Add -V/--cpu-visual option
perf sched: Add documentation for timehist options
Namhyung Kim (2):
perf symbol: Print symbol offsets conditionally
perf tools: Support printing callchains with arrows
tools/perf/Documentation/perf-sched.txt | 46 +-
tools/perf/builtin-sched.c | 914 +++++++++++++++++++++++++++++++-
tools/perf/util/evsel.h | 1 +
tools/perf/util/evsel_fprintf.c | 12 +-
tools/perf/util/symbol.h | 3 +-
tools/perf/util/symbol_fprintf.c | 11 +-
6 files changed, 972 insertions(+), 15 deletions(-)
--
2.10.1
