public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Namhyung Kim <namhyung@kernel.org>
To: Jiri Olsa <jolsa@redhat.com>
Cc: linux-kernel@vger.kernel.org,
	Corey Ashford <cjashfor@linux.vnet.ibm.com>,
	Frederic Weisbecker <fweisbec@gmail.com>,
	Ingo Molnar <mingo@elte.hu>, Paul Mackerras <paulus@samba.org>,
	Peter Zijlstra <a.p.zijlstra@chello.nl>,
	Arnaldo Carvalho de Melo <acme@redhat.com>,
	Steven Rostedt <rostedt@goodmis.org>,
	David Ahern <dsahern@gmail.com>
Subject: Re: [RFC 00/22] perf tools: Display tracepoint enahncements
Date: Wed, 05 Feb 2014 10:58:37 +0900	[thread overview]
Message-ID: <87eh3i1fsi.fsf@sejong.aot.lge.com> (raw)
In-Reply-To: <1391377150-23920-1-git-send-email-jolsa@redhat.com> (Jiri Olsa's message of "Sun, 2 Feb 2014 22:38:48 +0100")

On Sun,  2 Feb 2014 22:38:48 +0100, Jiri Olsa wrote:
> hi,
> sending out tracepoint (mostly) events display enahncements.
>
> * adding the '--tp' option for report command to show
>   tracepoint related info. Use can specify following switches:
>     fields: shows separated tracepoint fields
>     format: shows tracepoints 'print fmt' in single column
>              (This is default if no switch is given.)

Looks like an useful feature.. but unfortunately it seems not to work
well with --children and (especially) --group options.  Do you have an
idea?

Also as Ingo requested I plan to extend controlling output column with
-F/--field option.  I think I need to more consideration..

Thanks,
Namhyung


>
>   $ perf report --tp --no-children
>   Samples: 4K of event 'sched:sched_switch', Event count (approx.): 4788
>    Overhead          Command      Shared Object          Symbol  Print fmt
>   +  26.27%          swapper  [kernel.kallsyms]  [k] __schedule  swapper/2:0 [120] R ==> offlineimap:22134 [120]
>   +  26.27%      offlineimap  [kernel.kallsyms]  [k] __schedule  offlineimap:22134 [120] S ==> swapper/2:0 [120]
>   -   8.15%          swapper  [kernel.kallsyms]  [k] __schedule  swapper/3:0 [120] R ==> offlineimap:22134 [120]
>        __schedule
>        schedule_preempt_disabled
>        cpu_startup_entry
>        start_secondary
>   +   8.15%      offlineimap  [kernel.kallsyms]  [k] __schedule  offlineimap:22134 [120] S ==> swapper/3:0 [120]
>
>   $ perf report --tp=fields --no-children
>   Samples: 4K of event 'sched:sched_switch', Event count (approx.): 4788
>     Overhead          Command      Shared Object          Symbol                       prev_comm    prev_pid   prev_prio          prev_state                       next_comm    next_pid   next_prio
>   +   26.27%          swapper  [kernel.kallsyms]  [k] __schedule                       swapper/2           0         120                   0                     offlineimap       22134         120
>   +   26.27%      offlineimap  [kernel.kallsyms]  [k] __schedule                     offlineimap       22134         120                   1                       swapper/2           0         120
>   -    8.15%          swapper  [kernel.kallsyms]  [k] __schedule                       swapper/3           0         120                   0                     offlineimap       22134         120
>        __schedule
>        schedule_preempt_disabled
>        cpu_startup_entry
>        start_secondary
>   +    8.15%      offlineimap  [kernel.kallsyms]  [k] __schedule                     offlineimap       22134         120                   1                       swapper/3           0         120
>
>
> * adding '--list' report option to display entries sequentialy:
>
>   Samples: 2K of event 'sched:sched_switch', Event count (approx.): 2450
>         Self  Children                              Time          Command                Shared Object                                                       Symbol  Print fmt
>   +    0.04%     0.04%        9582.431783 +000000.000000          swapper  [kernel.kallsyms]            [k] __schedule                                               swapper/1:0 [120] R ==> perf:11168 [120]
>   -    0.00%     0.04%        9582.431783 +000000.000000          swapper  [kernel.kallsyms]            [k] __schedule                                               swapper/1:0 [120] R ==> perf:11168 [120]
>        __schedule
>        schedule
>        cpu_idle
>        start_secondary
>   +    0.00%     0.04%        9582.431783 +000000.000000          swapper  [kernel.kallsyms]            [k] schedule                                                 swapper/1:0 [120] R ==> perf:11168 [120]
>   +    0.00%     0.04%        9582.431783 +000000.000000          swapper  [kernel.kallsyms]            [k] cpu_idle                                                 swapper/1:0 [120] R ==> perf:11168 [120]
>
> * adding 'H' key handler to togle header columns
>
>
> thanks for comments,
> jirka
>
>
> Signed-off-by: Jiri Olsa <jolsa@redhat.com>
> Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
> Cc: Frederic Weisbecker <fweisbec@gmail.com>
> Cc: Ingo Molnar <mingo@elte.hu>
> Cc: Namhyung Kim <namhyung@kernel.org>
> Cc: Paul Mackerras <paulus@samba.org>
> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
> Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
> Cc: Steven Rostedt <rostedt@goodmis.org>
> Cc: David Ahern <dsahern@gmail.com>
> ---
> Jiri Olsa (22):
>       perf tools: Fix memory leak in event_format__print function
>       perf tools: Add time sort entry
>       perf tools: Add idx sort entry
>       perf tools: Add --list report option
>       perf tools: Add sort_entry struct into sort entries callbacks
>       perf tools: Add selected bool into se_snprintf sort entries callback
>       perf tools: Implement selected bool se_snprintf callback logic
>       perf tools: Implement selected logic for time sort entry
>       perf tools: Factor ui_browser ops out of ui_browser struct
>       perf tools: Add header callback into ui_browser_ops struct
>       perf tools: Remove ev_name argument from perf_evsel__hists_browse
>       perf tools: Add header callback to hist browser
>       perf tools: Factor hpp_arg struct to carry hist_browser
>       perf tools tui: Display columns header text on 'H' press
>       tools lib traceevent: Factor print_event_fields function
>       tools lib traceevent: Make the name output optional in pevent_field_info
>       tools lib traceevent: Add pevent_field_cmp function
>       perf tools: Factor sort entries loops
>       perf tools: Add local hists sort entry list
>       perf tools: Make hists col_len dynamicaly alocated
>       perf tools: Add raw info into hist entry
>       perf tools: Add support for tracepoint fields
>
>  tools/lib/traceevent/event-parse.c       | 181 +++++++++++++++++++++----------
>  tools/lib/traceevent/event-parse.h       |  10 ++
>  tools/perf/Documentation/perf-report.txt |   6 ++
>  tools/perf/Makefile.perf                 |   1 +
>  tools/perf/builtin-annotate.c            |   4 +-
>  tools/perf/builtin-diff.c                |   4 +-
>  tools/perf/builtin-report.c              |  52 ++++++++-
>  tools/perf/tests/hists_link.c            |   6 +-
>  tools/perf/ui/browser.c                  | 101 ++++++++++++-----
>  tools/perf/ui/browser.h                  |  31 ++++--
>  tools/perf/ui/browsers/annotate.c        |  19 ++--
>  tools/perf/ui/browsers/header.c          |   8 +-
>  tools/perf/ui/browsers/hists.c           | 168 ++++++++++++++++++++++-------
>  tools/perf/ui/browsers/map.c             |   8 +-
>  tools/perf/ui/browsers/scripts.c         |   2 +-
>  tools/perf/ui/gtk/hists.c                |   8 +-
>  tools/perf/ui/hist.c                     |  10 +-
>  tools/perf/ui/stdio/hist.c               |   6 +-
>  tools/perf/ui/tui/util.c                 |   8 +-
>  tools/perf/util/evsel.c                  |  24 +++--
>  tools/perf/util/evsel.h                  |   6 +-
>  tools/perf/util/hist.c                   |  65 ++++++++---
>  tools/perf/util/hist.h                   |  20 +++-
>  tools/perf/util/python.c                 |   3 +-
>  tools/perf/util/report-tp.c              | 244 +++++++++++++++++++++++++++++++++++++++++
>  tools/perf/util/report-tp.h              |  18 ++++
>  tools/perf/util/sort.c                   | 370 +++++++++++++++++++++++++++++++++++++++++++++++++--------------
>  tools/perf/util/sort.h                   |  29 ++++-
>  tools/perf/util/symbol.h                 |   3 +-
>  tools/perf/util/trace-event-parse.c      |   1 +
>  tools/perf/util/util.h                   |  17 +++
>  31 files changed, 1144 insertions(+), 289 deletions(-)
>  create mode 100644 tools/perf/util/report-tp.c
>  create mode 100644 tools/perf/util/report-tp.h

      parent reply	other threads:[~2014-02-05  1:58 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-02-02 21:38 [RFC 00/22] perf tools: Display tracepoint enahncements Jiri Olsa
2014-02-02 21:38 ` [PATCH 01/22] perf tools: Fix memory leak in event_format__print function Jiri Olsa
2014-02-05  1:41   ` Namhyung Kim
2014-02-22 17:57   ` [tip:perf/core] " tip-bot for Jiri Olsa
2014-02-02 21:38 ` [PATCH 02/22] perf tools: Add time sort entry Jiri Olsa
2014-02-02 21:38 ` [PATCH 03/22] perf tools: Add idx " Jiri Olsa
2014-02-02 21:38 ` [PATCH 04/22] perf tools: Add --list report option Jiri Olsa
2014-02-05  1:50   ` Namhyung Kim
2014-02-05  9:14     ` Jiri Olsa
2014-02-02 21:38 ` [PATCH 05/22] perf tools: Add sort_entry struct into sort entries callbacks Jiri Olsa
2014-02-02 21:38 ` [PATCH 06/22] perf tools: Add selected bool into se_snprintf sort entries callback Jiri Olsa
2014-02-02 21:38 ` [PATCH 07/22] perf tools: Implement selected bool se_snprintf callback logic Jiri Olsa
2014-02-02 21:38 ` [PATCH 08/22] perf tools: Implement selected logic for time sort entry Jiri Olsa
2014-02-02 21:38 ` [PATCH 09/22] perf tools: Factor ui_browser ops out of ui_browser struct Jiri Olsa
2014-02-02 21:38 ` [PATCH 10/22] perf tools: Add header callback into ui_browser_ops struct Jiri Olsa
2014-02-02 21:38 ` [PATCH 11/22] perf tools: Remove ev_name argument from perf_evsel__hists_browse Jiri Olsa
2014-02-02 21:39 ` [PATCH 12/22] perf tools: Add header callback to hist browser Jiri Olsa
2014-02-02 21:39 ` [PATCH 13/22] perf tools: Factor hpp_arg struct to carry hist_browser Jiri Olsa
2014-02-02 21:39 ` [PATCH 14/22] perf tools tui: Display columns header text on 'H' press Jiri Olsa
2014-02-02 21:39 ` [PATCH 15/22] tools lib traceevent: Factor print_event_fields function Jiri Olsa
2014-02-02 21:39 ` [PATCH 16/22] tools lib traceevent: Make the name output optional in pevent_field_info Jiri Olsa
2014-02-02 21:39 ` [PATCH 17/22] tools lib traceevent: Add pevent_field_cmp function Jiri Olsa
2014-02-02 21:39 ` [PATCH 18/22] perf tools: Factor sort entries loops Jiri Olsa
2014-02-02 21:39 ` [PATCH 19/22] perf tools: Add local hists sort entry list Jiri Olsa
2014-02-02 21:39 ` [PATCH 20/22] perf tools: Make hists col_len dynamicaly alocated Jiri Olsa
2014-02-02 21:39 ` [PATCH 21/22] perf tools: Add raw info into hist entry Jiri Olsa
2014-02-02 21:39 ` [PATCH 22/22] perf tools: Add support for tracepoint fields Jiri Olsa
2014-02-02 21:45 ` [RFC 00/22] perf tools: Display tracepoint enahncements Jiri Olsa
2014-02-05  1:40   ` Namhyung Kim
2014-02-05  9:15     ` Jiri Olsa
2014-02-05  1:58 ` Namhyung Kim [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87eh3i1fsi.fsf@sejong.aot.lge.com \
    --to=namhyung@kernel.org \
    --cc=a.p.zijlstra@chello.nl \
    --cc=acme@redhat.com \
    --cc=cjashfor@linux.vnet.ibm.com \
    --cc=dsahern@gmail.com \
    --cc=fweisbec@gmail.com \
    --cc=jolsa@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=paulus@samba.org \
    --cc=rostedt@goodmis.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox