From: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
To: Ingo Molnar <mingo@elte.hu>
Cc: "Frank Ch. Eigler" <fche@redhat.com>,
Paul Mundt <lethal@linux-sh.org>,
linux-kernel <linux-kernel@vger.kernel.org>,
Jes Sorensen <jes@sgi.com>, Andrew Morton <akpm@osdl.org>,
Tom Zanussi <zanussi@us.ibm.com>,
Richard J Moore <richardj_moore@uk.ibm.com>,
Michel Dagenais <michel.dagenais@polymtl.ca>,
Christoph Hellwig <hch@infradead.org>,
Greg Kroah-Hartman <gregkh@suse.de>,
Thomas Gleixner <tglx@linutronix.de>,
William Cohen <wcohen@redhat.com>,
"Martin J. Bligh" <mbligh@mbligh.org>
Subject: Re: MARKER mechanism, try 2
Date: Mon, 18 Sep 2006 13:47:42 -0400 [thread overview]
Message-ID: <20060918174742.GA27398@Krystal> (raw)
In-Reply-To: <20060918162828.GA22910@elte.hu>
* Ingo Molnar (mingo@elte.hu) wrote:
>
> * Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> wrote:
>
> > It supports 5 modes :
> >
> > - marker becomes nothing
> > - marker calls printk
> > - marker calls a tracer
> > - marker puts a symbol (for kprobe)
> > - marker puts a symbol and 5 NOPS for a jump probe.
>
> just go for 'nothing' and the 5-NOP variant, and please implement
> support for it from within LTT, via a kprobe - if you want me to support
> this stuff for upstream inclusion. If we support any static tracer mode
> and LTT does not support the kprobe mode then we are back to square 1
> wrt. dependencies ...
>
I am open to make LTTng support kprobes as a commodity (in fact, this point has
been on the LTTng project roadmap for almost a year). But in no way does it
solve the entire tracing problem. As an example, LTTng traces the page fault
handler, when kprobes just can't instrument it.
I keep thinking that a complete marker mechanism must have the ability to be
turned into function calls or inline functions when necessary.
Going further, we could think of a marker mechanism that would be aware of the
"difficulty" level of the probe, so that even if CONFIG_KPROBELOG is selected,
it would use a direct call or inlined function for probing the page fault
handler.
i.e. :
"normal" (nothing, kprobe, jumpprobe, printk or tracer)
MARK(eventname, "%d %s", myint, mystring);
"cannot be probed dynamically" (used in kprobes itself, page fault handler)
(only nothing or tracer)
MARK_NOPROBE(eventname, "%d %s", myint, mystring);
"cannot use printk" (used in scheduler, NMIs, wakeup, printk itself)
(nothing, kprobe, jumpprobe or tracer)
MARK_NOPRINT(eventname, "%d %s", myint, mystring);
Using the following table to select the mechanism :
Config/probe declaration | normal | noprobe | noprint
------------------------------------------------------------------------------
nothing | nothing | nothing | nothing
kprobe | kprobe | tracer | kprobe
jumpprobe | jumpprobe | tracer | jumpprobe
printk | printk | tracer | kprobe
tracer | tracer | tracer | tracer
Therefore, selecting the "kprobe" configuration option would still let people
instrument the hardest paths while having mostly dynamic probes.
Mathieu
OpenPGP public key: http://krystal.dyndns.org:8080/key/compudj.gpg
Key fingerprint: 8CD5 52C3 8E3C 4140 715F BA06 3F25 A8FE 3BAE 9A68
next prev parent reply other threads:[~2006-09-18 17:47 UTC|newest]
Thread overview: 82+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-09-17 9:40 The emperor is naked: why *comprehensive* static markup belongs in mainline Karim Yaghmour
2006-09-17 11:21 ` Paul Mundt
2006-09-17 14:36 ` tracepoint maintainance models Ingo Molnar
2006-09-17 15:02 ` Roman Zippel
2006-09-17 15:09 ` Ingo Molnar
2006-09-17 17:18 ` Roman Zippel
2006-09-17 23:27 ` Ingo Molnar
2006-09-17 23:41 ` Ingo Molnar
2006-09-18 0:17 ` Roman Zippel
2006-09-18 9:01 ` Jes Sorensen
2006-09-17 20:37 ` Roman Zippel
2006-09-17 22:34 ` Ingo Molnar
2006-09-17 15:36 ` Mathieu Desnoyers
2006-09-18 0:07 ` Ingo Molnar
2006-09-18 1:12 ` Karim Yaghmour
2006-09-18 1:13 ` Ingo Molnar
2006-09-18 2:32 ` Karim Yaghmour
2006-09-18 2:57 ` Ingo Molnar
2006-09-18 3:54 ` Karim Yaghmour
2006-09-18 4:09 ` Ingo Molnar
2006-09-18 4:43 ` Karim Yaghmour
2006-09-18 2:43 ` Mathieu Desnoyers
2006-09-18 3:21 ` Ingo Molnar
2006-09-18 4:26 ` Mathieu Desnoyers
2006-09-18 5:08 ` Ingo Molnar
2006-09-18 12:25 ` Frank Ch. Eigler
2006-09-18 15:02 ` Ingo Molnar
2006-09-18 15:45 ` Mathieu Desnoyers
2006-09-18 15:48 ` Alan Cox
2006-09-18 15:22 ` Ingo Molnar
2006-09-18 16:19 ` Alan Cox
2006-09-18 16:15 ` Ingo Molnar
2006-09-18 17:02 ` Alan Cox
2006-09-18 16:15 ` Frank Ch. Eigler
2006-09-18 17:02 ` Alan Cox
2006-09-18 17:27 ` Frank Ch. Eigler
2006-09-18 18:04 ` Alan Cox
2006-09-18 17:54 ` Martin Bligh
2006-09-18 18:05 ` Frank Ch. Eigler
2006-09-18 19:10 ` Vara Prasad
2006-09-18 19:49 ` Alan Cox
2006-09-18 19:39 ` Frank Ch. Eigler
2006-09-18 20:28 ` Vara Prasad
2006-10-06 5:33 ` Steven Rostedt
2006-10-06 13:01 ` Frank Ch. Eigler
2006-10-06 14:23 ` Steven Rostedt
2006-10-06 23:17 ` Jeremy Fitzhardinge
2006-09-18 15:47 ` Frank Ch. Eigler
2006-09-18 15:42 ` Ingo Molnar
2006-09-18 16:30 ` MARKER mechanism, try 2 Mathieu Desnoyers
2006-09-18 16:28 ` Ingo Molnar
2006-09-18 17:47 ` Mathieu Desnoyers [this message]
2006-09-18 19:39 ` Alan Cox
2006-09-17 20:19 ` tracepoint maintainance models Nicholas Miell
2006-09-17 23:06 ` Ingo Molnar
2006-09-18 0:05 ` Roman Zippel
2006-09-18 1:52 ` Theodore Tso
2006-09-19 12:58 ` tracing - consensus building insteat of dogfights Christoph Hellwig
2006-09-19 13:25 ` Roman Zippel
2006-09-19 13:45 ` Karim Yaghmour
2006-09-19 14:25 ` Karim Yaghmour
2006-09-18 0:10 ` tracepoint maintainance models Nicholas Miell
2006-09-18 0:43 ` Roman Zippel
2006-09-18 0:56 ` Karim Yaghmour
2006-09-18 0:56 ` Ingo Molnar
2006-09-18 2:09 ` Karim Yaghmour
2006-09-18 3:30 ` Ingo Molnar
2006-09-18 3:52 ` Theodore Tso
2006-09-18 4:11 ` Ingo Molnar
2006-09-18 4:24 ` Karim Yaghmour
2006-09-18 4:32 ` Ingo Molnar
2006-09-18 5:03 ` LTTng and SystemTAP (Everyone who is scared to read this huge thread, skip to here) Mathieu Desnoyers
2006-09-18 15:11 ` Ingo Molnar
2006-09-23 15:50 ` Mathieu Desnoyers
2006-09-18 5:37 ` tracepoint maintainance models Karim Yaghmour
2006-09-18 20:12 ` Michel Dagenais
2006-09-18 4:14 ` Karim Yaghmour
2006-09-18 4:09 ` Ingo Molnar
2006-09-18 4:57 ` Karim Yaghmour
2006-09-18 1:03 ` The emperor is naked: why *comprehensive* static markup belongs in mainline Karim Yaghmour
2006-09-18 15:53 ` Jose R. Santos
2006-09-18 17:28 ` Karim Yaghmour
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=20060918174742.GA27398@Krystal \
--to=mathieu.desnoyers@polymtl.ca \
--cc=akpm@osdl.org \
--cc=fche@redhat.com \
--cc=gregkh@suse.de \
--cc=hch@infradead.org \
--cc=jes@sgi.com \
--cc=lethal@linux-sh.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mbligh@mbligh.org \
--cc=michel.dagenais@polymtl.ca \
--cc=mingo@elte.hu \
--cc=richardj_moore@uk.ibm.com \
--cc=tglx@linutronix.de \
--cc=wcohen@redhat.com \
--cc=zanussi@us.ibm.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