linux-trace-devel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Tzvetomir Stoyanov (VMware)" <tz.stoyanov@gmail.com>
To: rostedt@goodmis.org
Cc: linux-trace-devel@vger.kernel.org
Subject: [PATCH 0/5] Initial trace-cmd perf support
Date: Thu,  3 Dec 2020 08:02:21 +0200	[thread overview]
Message-ID: <20201203060226.476475-1-tz.stoyanov@gmail.com> (raw)

Initial implementation of trace-cmd perf subcommand. It collects call traces
of a given PID and resolves collected addresses to function names. Still
work in progress.
What it has till now:
 - Configures and runs perf to collect call traces of the given PID.
 - Runs per CPU thread to read perf buffers and store traces in temp files.
 - Resolves collected addresses to function names (well, at least tries to do
   it). This includes also dynamic libraries, loaded runtime during the
   trace.
What should be implemented:
 - Store the results in a trace.dat file. Should think about how to extend
   its format to hold these traces.
 - Visualise the call stack from trace.dat file with trace-cmd report and
   KernelShark.

Tzvetomir Stoyanov (VMware) (5):
  trace-cmd: Internal refactoring of pid address map logic
  trace-cmd: Make read_file_string() non static
  trace-cmd: New internal APIs for reading ELF header
  trace-cmd: Add a new option in trace.dat file for the address to
    function name mapping
  [WIP] trace-cmd: Add new subcomand "trace-cmd perf"

 Makefile                                      |  10 +
 .../include/private/trace-cmd-private.h       |   5 +
 lib/trace-cmd/trace-input.c                   | 144 +++
 tracecmd/Makefile                             |   2 +
 tracecmd/include/trace-local.h                |  46 +-
 tracecmd/trace-cmd.c                          |   1 +
 tracecmd/trace-dump.c                         |   9 +-
 tracecmd/trace-obj-debug.c                    | 907 ++++++++++++++++++
 tracecmd/trace-perf.c                         | 726 ++++++++++++++
 tracecmd/trace-record.c                       | 107 +--
 10 files changed, 1841 insertions(+), 116 deletions(-)
 create mode 100644 tracecmd/trace-obj-debug.c
 create mode 100644 tracecmd/trace-perf.c

-- 
2.28.0


             reply	other threads:[~2020-12-03  6:03 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-12-03  6:02 Tzvetomir Stoyanov (VMware) [this message]
2020-12-03  6:02 ` [PATCH 1/5] trace-cmd: Internal refactoring of pid address map logic Tzvetomir Stoyanov (VMware)
2020-12-03  6:02 ` [PATCH 2/5] trace-cmd: Make read_file_string() non static Tzvetomir Stoyanov (VMware)
2020-12-03  6:02 ` [PATCH 3/5] trace-cmd: New internal APIs for reading ELF header Tzvetomir Stoyanov (VMware)
2020-12-03  6:02 ` [PATCH 4/5] trace-cmd: Add a new option in trace.dat file for the address to function name mapping Tzvetomir Stoyanov (VMware)
2020-12-03  6:02 ` [PATCH 5/5] [WIP] trace-cmd: Add new subcomand "trace-cmd perf" Tzvetomir Stoyanov (VMware)
2021-02-19  2:03   ` Steven Rostedt
2021-02-19  3:53     ` Steven Rostedt
2021-02-19 17:51       ` Dario Faggioli
2021-02-19  7:16     ` Tzvetomir Stoyanov
2021-02-19 14:36       ` Steven Rostedt
2021-02-19 17:56         ` Tzvetomir Stoyanov
2021-02-19 19:11           ` Steven Rostedt

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=20201203060226.476475-1-tz.stoyanov@gmail.com \
    --to=tz.stoyanov@gmail.com \
    --cc=linux-trace-devel@vger.kernel.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;
as well as URLs for NNTP newsgroup(s).