linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Andi Kleen <andi@firstfloor.org>
To: acme@kernel.org
Cc: jolsa@kernel.org, namhyung@kernel.org,
	linux-kernel@vger.kernel.org, linux-perf-users@vger.kernel.org
Subject: Support sample context in perf report
Date: Tue,  5 Mar 2019 06:47:43 -0800	[thread overview]
Message-ID: <20190305144758.12397-1-andi@firstfloor.org> (raw)

[Changes: 
v4:
Address review comments. 
Fix --cpu filtering.
Fix a sampling bug.
Add support for configuring custom script menu entries in perfconfig.
Fix display of more samples than fit on screen.
Fix some buffer overruns in legacy code.
Add more tips
]

We currently have two ways to look at sample data in perf:
either use perf report to aggregate everything, or use
perf script to look at all individual samples.

Both ways are useful. Of course aggregation is useful
to quickly find the most expensive part of the code.

But sometimes a single sample is not good enough to
determine the problem and we need to look at context, either
through branch contexts, or other previous samples (e.g. for
correlating different micro architecture events or computing
metrics)

This can be done through perf script today, but it can
be rather cumbersome to find the right samples to look
at.

Another problem with perf report is that it aggregates
the whole measurement period. But many real workloads
have phases where they behave quite differently, and it is
often not useful to combine them into a single histogram.

While this can be worked around with the --time option
to report, it can be quite cumbersome.

This patch kit attempts to address some of these
problems in perf report by making it time aware.

- It adds a new time sort key that allows perf report
to separate samples from different regions. The time
regions can be defined with a new --time-quantum option.

- Then it extends the perf script support in the
tui record browser to allow browsing samples for 
different time regions from within a perf report
session.

- Extends the report browser script display
to automatically select sensible defaults
based on what was recorded. For example it will
automatically show branch contexts with -b.

- Support browsing the context of individual samples.
perf report can save a limited number of random samples
per histogram entry with the new --samples option.
Then the browser allows directly jumping to any
of the saved samples and browsing the context on the current
thread or CPU.

There could be probably be done more to make
perf report even better for such use cases (e.g. a real
time line display), but this basic support is good
enough for many useful usages.

Also available in
git://git.kernel.org/pub/scm/linux/kernel/git/ak/linux-misc.git perf/streams-4



             reply	other threads:[~2019-03-05 14:49 UTC|newest]

Thread overview: 49+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-03-05 14:47 Andi Kleen [this message]
2019-03-05 14:47 ` [PATCH v4 01/15] perf tools: Add utility function to fetch executable Andi Kleen
2019-03-06 20:53   ` Arnaldo Carvalho de Melo
2019-03-06 21:08     ` Andi Kleen
2019-03-09 19:59   ` [tip:perf/urgent] perf thread: Generalize function to copy from thread addr space from intel-bts code tip-bot for Andi Kleen
2019-03-05 14:47 ` [PATCH v4 02/15] perf tools script: Support insn output for normal samples Andi Kleen
2019-03-08 13:37   ` Arnaldo Carvalho de Melo
2019-03-22 21:59   ` [tip:perf/urgent] perf " tip-bot for Andi Kleen
2019-03-05 14:47 ` [PATCH v4 03/15] perf tools script: Filter COMM/FORK/.. events by CPU Andi Kleen
2019-03-06 14:23   ` Adrian Hunter
2019-03-07 11:02     ` Jiri Olsa
2019-03-08 13:39       ` Arnaldo Carvalho de Melo
2019-03-08 15:08         ` Andi Kleen
2019-03-05 14:47 ` [PATCH v4 04/15] perf tools report: Support nano seconds Andi Kleen
2019-03-22 21:59   ` [tip:perf/urgent] perf report: Support output in nanoseconds tip-bot for Andi Kleen
2019-03-05 14:47 ` [PATCH v4 05/15] perf tools report: Parse time quantum Andi Kleen
2019-03-08 13:45   ` Arnaldo Carvalho de Melo
2019-03-08 13:52     ` Arnaldo Carvalho de Melo
2019-03-08 17:39       ` Andi Kleen
2019-03-22 22:01   ` [tip:perf/urgent] perf " tip-bot for Andi Kleen
2019-03-05 14:47 ` [PATCH v4 06/15] perf tools report: Support time sort key Andi Kleen
2019-03-05 14:47 ` [PATCH v4 07/15] perf tools report: Use less for scripts output Andi Kleen
2019-03-08 13:47   ` Arnaldo Carvalho de Melo
2019-03-09  7:52     ` Feng Tang
2019-03-05 14:47 ` [PATCH v4 08/15] perf tools report: Support running scripts for current time range Andi Kleen
2019-03-05 14:47 ` [PATCH v4 09/15] perf tools report: Support builtin perf script in scripts menu Andi Kleen
2019-03-05 14:47 ` [PATCH v4 10/15] perf tools: Add utility function to print ns time stamps Andi Kleen
2019-03-08 13:51   ` Arnaldo Carvalho de Melo
2019-03-22 22:00   ` [tip:perf/urgent] perf time-utils: Add utility function to print time stamps in nanoseconds tip-bot for Andi Kleen
2019-03-05 14:47 ` [PATCH v4 11/15] perf tools report: Implement browsing of individual samples Andi Kleen
2019-03-05 14:47 ` [PATCH v4 12/15] perf tools: Add some new tips describing the new options Andi Kleen
2019-03-05 14:47 ` [PATCH v4 13/15] perf tools report: Add custom scripts to script menu Andi Kleen
2019-03-05 14:47 ` [PATCH v4 14/15] perf tools script: Add array bound checking to list_scripts Andi Kleen
2019-03-05 14:47 ` [PATCH v4 15/15] perf tools ui: Fix ui popup browser for many entries Andi Kleen
2019-03-07 10:57 ` Support sample context in perf report Jiri Olsa
2019-03-07 16:57   ` Andi Kleen
  -- strict thread matches above, loose matches on Subject: below --
2019-03-11 14:44 Andi Kleen
2019-03-09  5:56 Andi Kleen
2019-02-28 18:35 Andi Kleen
2019-02-26  3:04 Andi Kleen
2019-02-26 22:33 ` Jiri Olsa
2019-02-26 22:55   ` Andi Kleen
2019-02-27 11:18 ` Jiri Olsa
2019-02-27 16:01   ` Andi Kleen
2019-02-27 16:16     ` Jiri Olsa
2019-02-27 17:29       ` Andi Kleen
2019-02-27 17:41         ` Arnaldo Carvalho de Melo
2019-02-27 17:56           ` Jiri Olsa
2019-02-24 15:37 Andi Kleen

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=20190305144758.12397-1-andi@firstfloor.org \
    --to=andi@firstfloor.org \
    --cc=acme@kernel.org \
    --cc=jolsa@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=namhyung@kernel.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;
as well as URLs for NNTP newsgroup(s).