mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
To: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>,
	Arnaldo Carvalho de Melo <acme@kernel.org>,
	Ingo Molnar <mingo@kernel.org>,
	Peter Zijlstra <a.p.zijlstra@chello.nl>,
	LKML <linux-kernel@vger.kernel.org>,
	David Ahern <dsahern@gmail.com>
Subject: Re: Re: [PATCH 3/4] perf probe: Fix probing kretprobes
Date: Mon, 12 Jan 2015 21:08:53 +0900	[thread overview]
Message-ID: <54B3B955.80105@hitachi.com> (raw)
In-Reply-To: <20150112112604.GA4811@krava.brq.redhat.com>

(2015/01/12 20:26), Jiri Olsa wrote:
> On Sat, Jan 10, 2015 at 07:33:47PM +0900, Namhyung Kim wrote:
>> The commit dfef99cd0b2c ("perf probe: Use ref_reloc_sym based address
>> instead of the symbol name") converts kprobes to use ref_reloc_sym
>> (i.e. _stext) and offset instead of using symbol's name directly.  So
>> on my system, adding do_fork ends up with like below:
>>
>>   $ sudo perf probe -v --add do_fork%return
>>   probe-definition(0): do_fork%return
>>   symbol:do_fork file:(null) line:0 offset:0 return:1 lazy:(null)
>>   0 arguments
>>   Looking at the vmlinux_path (7 entries long)
>>   Using /lib/modules/3.17.6-1-ARCH/build/vmlinux for symbols
>>   Could not open debuginfo. Try to use symbols.
>>   Opening /sys/kernel/debug/tracing/kprobe_events write=1
>>   Added new event:
>>   Writing event: r:probe/do_fork _stext+456136
>>   Failed to write event: Invalid argument
>>   Error: Failed to add events. Reason: Operation not permitted (Code: -1)
>>
>> As you can see, the do_fork was translated to _stext+456136.  This was
>> because to support (local) symbols that have same name.  But the
>> problem is that kretprobe requires to be inserted at function start
>> point so it simply checks whether it's called with offset 0.  And if
>> not, it'll return with -EINVAL.  You can see it with dmesg.
>>
>>   $ dmesg | tail -1
>>     [125621.764103] Return probe must be used without offset.
>>
>> So we need to use the symbol name instead of ref_reloc_sym in case of
>> return probes.
> 
> with your patchset on top of current Arnaldo's perf/core,
> I'm still getting the same error:
> 
> 
> [root@krava perf]# ./perf probe -a fork_exit=do_fork%return
> Added new event:
> Failed to write event: Invalid argument
>   Error: Failed to add events.
> [root@krava perf]# ./perf probe -v --add do_fork%return
> probe-definition(0): do_fork%return
> symbol:do_fork file:(null) line:0 offset:0 return:1 lazy:(null)
> 0 arguments
> Using /root/.debug/.build-id/60/4c5fd8a6c9202acac3a9e82ecd34b22b8caef6 for symbols
> Open Debuginfo file: /root/.debug/.build-id/60/4c5fd8a6c9202acac3a9e82ecd34b22b8caef6
> Try to find probe point from debuginfo.
> Probe point found: do_fork+0
> Found 1 probe_trace_events.
> Opening /sys/kernel/debug/tracing/kprobe_events write=1
> Added new event:
> Writing event: r:probe/do_fork _text+597840
> Failed to write event: Invalid argument
>   Error: Failed to add events. Reason: Invalid argument (Code: -22)
> [root@krava perf]# dmesg | tail -1
> [113205.738833] Return probe must be used without offset.

Yes, it seems that Namhyung's patch fixes a half of the code path.
We should fix post_process_probe_trace_events() to skip symbol
conversion or we shouldn't use debuginfo when %return is given.
Even if we find inlined function for retprobe, we can't use it
for retprobe...

Thank you,

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



  reply	other threads:[~2015-01-12 12:09 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-01-10 10:33 [PATCH 1/4] perf tools: Allow use of an exclusive option more than once Namhyung Kim
2015-01-10 10:33 ` [PATCH 2/4] perf probe: Do not rely on map__load() filter to find symbols Namhyung Kim
2015-01-12 12:31   ` Masami Hiramatsu
2015-01-14  1:45     ` Namhyung Kim
2015-01-14  8:42       ` Masami Hiramatsu
2015-01-10 10:33 ` [PATCH 3/4] perf probe: Fix probing kretprobes Namhyung Kim
2015-01-12 11:26   ` Jiri Olsa
2015-01-12 12:08     ` Masami Hiramatsu [this message]
2015-01-12 12:22   ` Masami Hiramatsu
2015-01-14  1:59     ` Namhyung Kim
2015-01-10 10:33 ` [PATCH 4/4] perf probe: Propagate error code when write(2) failed Namhyung Kim
2015-01-12 11:17   ` Masami Hiramatsu
2015-01-12 14:03     ` Arnaldo Carvalho de Melo
2015-01-17 10:10   ` [tip:perf/urgent] " tip-bot for Namhyung Kim
2015-01-12 11:44 ` [PATCH 1/4] perf tools: Allow use of an exclusive option more than once Masami Hiramatsu
2015-01-12 14:02   ` Arnaldo Carvalho de Melo
2015-01-28 15:06 ` [tip:perf/core] " tip-bot for Namhyung Kim

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=54B3B955.80105@hitachi.com \
    --to=masami.hiramatsu.pt@hitachi.com \
    --cc=a.p.zijlstra@chello.nl \
    --cc=acme@kernel.org \
    --cc=dsahern@gmail.com \
    --cc=jolsa@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@kernel.org \
    --cc=namhyung@kernel.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

Powered by JetHome