linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
To: Jovi Zhangwei <jovi.zhangwei@gmail.com>
Cc: Ingo Molnar <mingo@redhat.org>,
	Steven Rostedt <rostedt@goodmis.org>,
	linux-kernel@vger.kernel.org,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Frederic Weisbecker <fweisbec@gmail.com>
Subject: Re: [PATCH 14/28] ktap: add runtime/kp_events.[c|h]
Date: Mon, 31 Mar 2014 18:10:07 +0900	[thread overview]
Message-ID: <533930EF.8080703@hitachi.com> (raw)
In-Reply-To: <1396014469-5937-15-git-send-email-jovi.zhangwei@gmail.com>

(2014/03/28 22:47), Jovi Zhangwei wrote:
> kp_events.c handle ktap events management(registry, destroy, event callback)
> 
> This file is core event management interface between ktap and kernel.
> 
> Exposed functions:
> 1). kp_events_init/kp_events_exit
> 
> 2). kp_event_create_kprobe
>         create kprobe event, for example:
>                 kdebug.kprobe("SyS_futex", function () {})
> 
> 3). kp_event_create_tracepoint
>         create tracepoint event, for example"
>                 kdebug.tracepoint("sys_futex_enter", function () {})
> 
> 4). kp_event_create
>         create perf backend event, for example:
>                 trace sched:sched_switch { print(argstr) }
> 
>         It call kernel function 'perf_event_create_kernel_counter' to
>         register event(tracepoint/kprobe/uprobe)
> 
> 5). kp_event_getarg
>         get argument of event, from arg0 to arg9,
>         only can be called in probe context.
>                 trace sched:sched_switch { print(arg0, arg1) }
> 
> 6). kp_event_stringify/kp_event_tostr
>         stringify argstr, sometimes if store argstr as key to table,
>         then it need to stringify firstly, like below:
>                 var s={} trace sched:sched_switch { s[argstr] += 1 }
>         (This is quite rare usage, but ktap support it)
> 
> Note:
> Why ktap support 'kdebug.kprobe' and 'kdebug.tracepoint' when
> it already support perf backend event(trace xxx {})?
> 
> Because benchmark shows raw kprobe and tracpoint interface is faster
> than perf backed tracing, nearly 10+%, it's more fair to compare
> with Systemtap by raw tracing syntax, not perf backend tracing.
>

Do we really need it just for a +10% performance? I doubt that.
I think the benefit point of ktap is "dynamic & simple programmable
tracer in kernel", not the good performance at least at this point.
Thus I think we should start ktap only with perf backend.

Thank you,


-- 
Masami HIRAMATSU
IT Management Research Dept. Linux Technology Center
Hitachi, Ltd., Yokohama Research Laboratory
E-mail: masami.hiramatsu.pt@hitachi.com



  reply	other threads:[~2014-03-31  9:10 UTC|newest]

Thread overview: 68+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-03-28 13:47 [RFC PATCH 00/28] ktap: A lightweight dynamic tracing tool for Linux Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 01/28] ktap: add README file Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 02/28] ktap: add ktap tutorial Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 03/28] ktap: add sample scripts Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 04/28] ktap: add basic ktap types definition Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 05/28] ktap: add bytecode definition Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 06/28] ktap: add include/ktap_arch.h and error header file Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 07/28] ktap: add runtime/ktap.[c|h] Jovi Zhangwei
2014-03-28 18:38   ` Andi Kleen
2014-03-29  7:32     ` Jovi Zhangwei
2014-03-29 17:04       ` Greg Kroah-Hartman
2014-03-30  7:26         ` Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 08/28] ktap: add runtime/kp_bcread.[c|h] Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 09/28] ktap: add runtime/kp_vm.[c|h] Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 10/28] ktap: add runtime/kp_str.[c|h] and runtime/kp_mempool.[c|h] Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 11/28] ktap: add runtime/kp_tab.[c|h] Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 12/28] ktap: add runtime/kp_obj.[c|h] Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 13/28] ktap: add runtime/kp_transport.[c|h] Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 14/28] ktap: add runtime/kp_events.[c|h] Jovi Zhangwei
2014-03-31  9:10   ` Masami Hiramatsu [this message]
2014-03-31 10:14     ` Jovi Zhangwei
2014-04-01  6:59       ` Masami Hiramatsu
2014-04-01  7:28         ` Jovi Zhangwei
2014-04-01  8:05           ` Masami Hiramatsu
2014-03-28 13:47 ` [PATCH 15/28] ktap: add built-in functions and library (runtime/lib_*.c) Jovi Zhangwei
2014-03-28 18:51   ` Andi Kleen
2014-03-29  4:15     ` Jovi Zhangwei
2014-03-30  0:58       ` Andi Kleen
2014-03-31  2:01         ` Jovi Zhangwei
2014-03-31 13:13           ` Andi Kleen
2014-04-02  1:44             ` Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 16/28] ktap: add runtime/amalg.c Jovi Zhangwei
2014-03-28 18:52   ` Andi Kleen
2014-03-29  7:38     ` Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 17/28] ktap: add userspace/kp_main.c Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 18/28] ktap: add compiler(userspace/kp_lex.[c|h] and userspace/kp_parse.[c|h]) Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 19/28] ktap: add userspace/symbol.[c|h] Jovi Zhangwei
2014-04-01  7:28   ` Masami Hiramatsu
2014-03-28 13:47 ` [PATCH 20/28] ktap: add userspace/kp_parse_events.c Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 21/28] ktap: add userspace/kp_reader.c Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 22/28] ktap: add userspace/kp_bcwrite.c Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 23/28] ktap: add userspace/kp_util.c Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 24/28] ktap: add Makefile Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 25/28] ktap: add Kconfig Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 26/28] ktap: add testsuite Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 27/28] ktap: add vim syntax file Jovi Zhangwei
2014-03-28 13:47 ` [PATCH 28/28] ktap: add COPYRIGHT file Jovi Zhangwei
2014-03-28 16:08 ` [RFC PATCH 00/28] ktap: A lightweight dynamic tracing tool for Linux Greg Kroah-Hartman
2014-03-29  1:46   ` Jovi Zhangwei
2014-03-31  7:17 ` Ingo Molnar
2014-03-31 10:01   ` Jovi Zhangwei
2014-03-31 21:29     ` Alexei Starovoitov
2014-04-01  4:47       ` Jovi Zhangwei
2014-04-02  4:57         ` Alexei Starovoitov
2014-04-02  6:37           ` Jovi Zhangwei
2014-04-02  7:43             ` Ingo Molnar
2014-04-02  8:49               ` Jovi Zhangwei
2014-04-04  7:36                 ` Ingo Molnar
2014-04-08  6:50                   ` Jovi Zhangwei
2014-04-14 15:11                     ` Ingo Molnar
2014-04-14 15:28                       ` Daniel Borkmann
2014-04-02  7:42           ` Ingo Molnar
2014-04-07 13:55             ` Peter Zijlstra
2014-04-08  7:40               ` Masami Hiramatsu
2014-04-08  9:08                 ` Peter Zijlstra
2014-04-02  7:36         ` Ingo Molnar
2014-03-31 20:06   ` Daniel Borkmann
2014-03-31  9:18 ` Masami Hiramatsu

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=533930EF.8080703@hitachi.com \
    --to=masami.hiramatsu.pt@hitachi.com \
    --cc=fweisbec@gmail.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=jovi.zhangwei@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.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).