From: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
To: Josh Stone <jistone@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>,
Hemant Kumar <hemant@linux.vnet.ibm.com>,
linux-kernel@vger.kernel.org, srikar@linux.vnet.ibm.com,
peterz@infradead.org, oleg@redhat.com,
hegdevasant@linux.vnet.ibm.com, mingo@redhat.com,
systemtap@sourceware.org, aravinda@linux.vnet.ibm.com,
penberg@iki.fi, Arnaldo Carvalho de Melo <acme@redhat.com>
Subject: Re: Re: [PATCH v4 5/5] perf/sdt: Add support to perf record to trace SDT events
Date: Thu, 06 Nov 2014 05:34:00 -0000 [thread overview]
Message-ID: <545B083D.2000205@hitachi.com> (raw)
In-Reply-To: <545AD9C7.50205@redhat.com>
(2014/11/06 11:15), Josh Stone wrote:
> On 11/05/2014 01:05 AM, Masami Hiramatsu wrote:
>> [Off topic] I really don't like that the current SDT's semaphore. If the user apps
>> see the instruction at the probe point, it is easy to check whether the event is
>> enabled or not. Thus I recommend to change its implementation and update version
>> instead of supporting current semaphore by perftools.
>
> You and I have banged heads on this before, but I don't think checking
> the instruction is a simple as you seem to think. I invite you to
> prototype this, and if you get it working we can discuss the tradeoffs.
Would you have the prototype? I'd like to look :)
> The good news is that other tools (stap and gdb) won't need to care. If
> the SDT semaphore goes automatic, then we can just set that note field
> to zero, unused from the tool's perspective.
>
> Another tactic is to just discourage developers from using the semaphore
> in the first place, as it's a completely optional feature. The marker
> is just a NOP, so adding some "if (enabled) {...}" around it is often a
> useless load and branch. It does make sense if the probe wants to
> provide some expensively-computed arguments though, like cpython does to
> prepare a function name string. So if you see a project testing the
> semaphore around simple arguments, I'd suggest they just probe directly
> instead.
I see, and we did that on qemu. I consider that someone maybe use
it in the future unless we remove it. If we can succeed to discourage
people using semaphore, we also should remove it.
Thank you,
--
Masami HIRAMATSU
Software Platform Research Dept. Linux Technology Research Center
Hitachi, Ltd., Yokohama Research Laboratory
E-mail: masami.hiramatsu.pt@hitachi.com
next prev parent reply other threads:[~2014-11-06 5:34 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-11-02 12:40 [PATCH v4 0/5] perf/sdt: SDT events listing/probing Hemant Kumar
2014-11-02 12:41 ` [PATCH v4 1/5] perf/sdt: ELF support for SDT Hemant Kumar
2014-11-02 12:42 ` [PATCH v4 3/5] perf/sdt: Show SDT cache contents Hemant Kumar
2014-11-02 12:42 ` [PATCH v4 4/5] perf/sdt: Delete SDT events from cache Hemant Kumar
2014-11-02 12:42 ` [PATCH v4 2/5] perf/sdt: Add SDT events into a cache Hemant Kumar
2014-11-02 12:43 ` [PATCH v4 5/5] perf/sdt: Add support to perf record to trace SDT events Hemant Kumar
2014-11-04 7:38 ` Namhyung Kim
2014-11-04 8:06 ` Hemant Kumar
2014-11-04 12:57 ` Masami Hiramatsu
[not found] ` <5459BD3E.7010804@linux.vnet.ibm.com>
2014-11-05 6:51 ` Hemant Kumar
2014-11-05 9:07 ` Masami Hiramatsu
2014-11-05 13:28 ` Arnaldo Carvalho de Melo
2014-11-05 7:06 ` Namhyung Kim
2014-11-05 9:05 ` Masami Hiramatsu
2014-11-06 2:16 ` Josh Stone
2014-11-06 5:34 ` Masami Hiramatsu [this message]
2014-11-06 7:06 ` Hemant Kumar
2014-11-06 14:56 ` Masami Hiramatsu
2014-11-07 8:21 ` [RFC] perf-cache command interface design Masami Hiramatsu
2014-11-07 8:42 ` Peter Zijlstra
2014-11-07 13:58 ` [PATCH RESEND 1/2] perf tools: Move disable_buildid_cache() to util/build-id.c Namhyung Kim
2014-11-07 13:58 ` [PATCH 2/2] perf tools: Add record.use-buildid-cache config option Namhyung Kim
2014-11-07 15:16 ` [RFC] perf-cache command interface design David Ahern
2014-11-07 15:33 ` Arnaldo Carvalho de Melo
2014-11-07 10:51 ` Hemant Kumar
2014-11-08 4:15 ` Masami Hiramatsu
2014-11-07 14:38 ` Arnaldo Carvalho de Melo
2014-11-08 4:26 ` Masami Hiramatsu
2014-11-07 14:43 ` Namhyung Kim
2014-11-08 4:38 ` Masami Hiramatsu
2014-11-10 10:59 ` Masami Hiramatsu
2014-11-10 12:23 ` Arnaldo Carvalho de Melo
2014-11-11 6:53 ` Masami Hiramatsu
2014-11-11 13:10 ` Arnaldo Carvalho de Melo
2014-11-12 15:26 ` Masami Hiramatsu
2014-11-17 3:09 ` Namhyung Kim
2014-11-17 3:17 ` Masami Hiramatsu
2014-11-17 22:09 ` Masami Hiramatsu
2014-11-18 4:51 ` Namhyung Kim
2014-11-18 11:16 ` Masami Hiramatsu
2014-11-18 4:41 ` Namhyung Kim
2014-11-18 10:32 ` Masami Hiramatsu
2014-11-17 18:59 ` Arnaldo Carvalho de Melo
2014-11-18 4:46 ` Namhyung Kim
2014-11-10 12:05 ` Hagen Paul Pfeifer
2014-11-10 12:31 ` Arnaldo Carvalho de Melo
2014-11-10 12:50 ` Peter Zijlstra
2014-11-10 13:37 ` Hagen Paul Pfeifer
2014-11-05 18:24 ` [PATCH v4 5/5] perf/sdt: Add support to perf record to trace SDT events Hemant Kumar
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=545B083D.2000205@hitachi.com \
--to=masami.hiramatsu.pt@hitachi.com \
--cc=acme@redhat.com \
--cc=aravinda@linux.vnet.ibm.com \
--cc=hegdevasant@linux.vnet.ibm.com \
--cc=hemant@linux.vnet.ibm.com \
--cc=jistone@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=namhyung@kernel.org \
--cc=oleg@redhat.com \
--cc=penberg@iki.fi \
--cc=peterz@infradead.org \
--cc=srikar@linux.vnet.ibm.com \
--cc=systemtap@sourceware.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 read-only IMAP folder(s) and NNTP newsgroup(s).