Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
        id S932938AbdCaJqA (ORCPT <rfc822;w@1wt.eu>);
        Fri, 31 Mar 2017 05:46:00 -0400
Received: from mail-oi0-f47.google.com ([209.85.218.47]:34235 "EHLO
        mail-oi0-f47.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
        with ESMTP id S932316AbdCaJp5 (ORCPT
        <rfc822;linux-kernel@vger.kernel.org>);
        Fri, 31 Mar 2017 05:45:57 -0400
MIME-Version: 1.0
In-Reply-To: <149076491191.24574.9187877160820798843.stgit@devbox>
References: <149076484118.24574.7083269903420611708.stgit@devbox> <149076491191.24574.9187877160820798843.stgit@devbox>
From: Alban Crequy <alban@kinvolk.io>
Date: Fri, 31 Mar 2017 11:45:55 +0200
Message-ID: <CADZs7q4k9YLgRuxccy7vuOns53GcPXG4D6HosRjfAeKm=RrL2A@mail.gmail.com>
Subject: Re: [RFC PATCH tip/master 1/3] trace: kprobes: Show sum of
 probe/retprobe nmissed count
To: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>, Ingo Molnar <mingo@redhat.com>,
        Alban Crequy <alban.crequy@gmail.com>,
        Alexei Starovoitov <ast@kernel.org>, Jonathan Corbet <corbet@lwn.net>,
        Arnaldo Carvalho de Melo <acme@redhat.com>,
        Omar Sandoval <osandov@fb.com>, linux-doc@vger.kernel.org,
        netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
        =?UTF-8?Q?Iago_L=C3=B3pez_Galeiras?= <iago@kinvolk.io>,
        Michael Schubert <michael@kinvolk.io>,
        Dorau Lukasz <lukasz.dorau@intel.com>, systemtap@sourceware.org
Content-Type: text/plain; charset=UTF-8
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org
Content-Length: 1285
Lines: 36

On Wed, Mar 29, 2017 at 7:22 AM, Masami Hiramatsu <mhiramat@kernel.org> wrote:
> Show sum of probe and retprobe nmissed count in
> kprobe_profile, since retprobe can be missed even
> if the kprobe itself succeeeded.
> This explains user why their return probe didn't hit
> sometimes.
>
> Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>

I tested this patch with my kretprobe on "inet_csk_accept" when there
are many processes waiting in the accept() syscall. I can now
successfully see the nmissed counter in
/sys/kernel/debug/tracing/kprobe_profile being incremented when the
kretprobe is missed.

Tested-by: Alban Crequy <alban@kinvolk.io>


> ---
>  kernel/trace/trace_kprobe.c |    2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c
> index 013f4e7..bbdc3de 100644
> --- a/kernel/trace/trace_kprobe.c
> +++ b/kernel/trace/trace_kprobe.c
> @@ -896,7 +896,7 @@ static int probes_profile_seq_show(struct seq_file *m, void *v)
>         seq_printf(m, "  %-44s %15lu %15lu\n",
>                    trace_event_name(&tk->tp.call),
>                    trace_kprobe_nhit(tk),
> -                  tk->rp.kp.nmissed);
> +                  tk->rp.kp.nmissed + tk->rp.nmissed);
>
>         return 0;
>  }
>