linux-trace-devel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Steven Rostedt <rostedt@goodmis.org>
To: Tzvetomir Stoyanov <tz.stoyanov@gmail.com>
Cc: Yordan Karadzhov <y.karadz@gmail.com>,
	Linux Trace Devel <linux-trace-devel@vger.kernel.org>
Subject: Re: [PATCH v4 08/10] libtracefs: Use the internal dynamic events API when creating synthetic events
Date: Fri, 5 Nov 2021 08:33:25 -0400	[thread overview]
Message-ID: <20211105083325.78b1de24@gandalf.local.home> (raw)
In-Reply-To: <CAPpZLN5bSdCKOvAMSefAaj0osq60qgcyRE7npMS6jPOcYoEAYg@mail.gmail.com>

On Fri, 5 Nov 2021 14:15:05 +0200
Tzvetomir Stoyanov <tz.stoyanov@gmail.com> wrote:

> I looked and tested more this code today, but as I understood the
> instance is mandatory in histograms creation and should not be removed
> from this API. That API appends the histogram config in the event's
> trigger files, which is for specific instances. The histogram is
> accumulated only in the context of that instance. I'm going to send
> next version of the patch set, but without changes to this API.

The histogram code is internal to the synthetic events. That's because it
uses the histogram to save the data across the start to end event to create
the synthetic event. The histogram is an implementation detail of the
synthetic event and should not be exposed via the API. And that is
*exactly* why instances should not be exposed to the synthetic event!

In fact, I don't wan the user to have control over the instances because
they are likely to screw it up if they don't know what they are doing.
That's why I suggested to convert the current API into a raw event, because
to use instances, you had better know what you are doing, otherwise you can
easily screw it up.

The histograms used by synthetic events are "special" and if not done
properly, will not do what people expect.

If you don't want to make this change, I will.

-- Steve

  reply	other threads:[~2021-11-05 12:33 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-04 11:10 [PATCH v4 00/10] libtracefs dynamic events support Tzvetomir Stoyanov (VMware)
2021-11-04 11:10 ` [PATCH v4 01/10] libtracefs: New APIs for dynamic events Tzvetomir Stoyanov (VMware)
2021-11-04 13:58   ` Steven Rostedt
2021-11-04 14:37   ` Steven Rostedt
2021-11-04 14:54   ` Steven Rostedt
2021-11-04 11:10 ` [PATCH v4 02/10] libtracefs: New APIs for kprobe allocation Tzvetomir Stoyanov (VMware)
2021-11-04 16:29   ` Steven Rostedt
2021-11-04 11:10 ` [PATCH v4 03/10] libtracefs: Remove redundant kprobes APIs Tzvetomir Stoyanov (VMware)
2021-11-04 11:10 ` [PATCH v4 04/10] libtracefs: Change tracefs_kprobe_info API Tzvetomir Stoyanov (VMware)
2021-11-04 16:33   ` Steven Rostedt
2021-11-04 17:28     ` Tzvetomir Stoyanov
2021-11-04 17:57       ` Steven Rostedt
2021-11-04 11:10 ` [PATCH v4 05/10] libtracefs: Reimplement kprobe raw APIs Tzvetomir Stoyanov (VMware)
2021-11-04 11:10 ` [PATCH v4 06/10] libtracefs: Extend kprobes unit test Tzvetomir Stoyanov (VMware)
2021-11-04 11:10 ` [PATCH v4 07/10] libtracefs: Rename tracefs_synth_init API Tzvetomir Stoyanov (VMware)
2021-11-04 11:10 ` [PATCH v4 08/10] libtracefs: Use the internal dynamic events API when creating synthetic events Tzvetomir Stoyanov (VMware)
2021-11-04 12:47   ` Yordan Karadzhov
2021-11-04 13:21     ` Steven Rostedt
2021-11-04 17:13   ` Steven Rostedt
2021-11-05 12:15     ` Tzvetomir Stoyanov
2021-11-05 12:33       ` Steven Rostedt [this message]
2021-11-04 11:10 ` [PATCH v4 09/10] libtracefs: Update kprobes man pages Tzvetomir Stoyanov (VMware)
2021-11-04 20:28   ` Steven Rostedt
2021-11-04 11:10 ` [PATCH v4 10/10] libtracefs: Document dynamic events APIs Tzvetomir Stoyanov (VMware)

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=20211105083325.78b1de24@gandalf.local.home \
    --to=rostedt@goodmis.org \
    --cc=linux-trace-devel@vger.kernel.org \
    --cc=tz.stoyanov@gmail.com \
    --cc=y.karadz@gmail.com \
    /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).