From mboxrd@z Thu Jan  1 00:00:00 1970
From: Eric Dumazet <eric.dumazet@gmail.com>
Subject: Re: [PATCH] net/ipv4:Remove two reports using kmemleak
Date: Mon, 16 Apr 2012 14:41:18 +0200
Message-ID: <1334580078.28012.68.camel@edumazet-glaptop>
References: <201204161940178124485@gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
Cc: davem <davem@davemloft.net>, netdev <netdev@vger.kernel.org>
To: majianpeng <majianpeng@gmail.com>
Return-path: <netdev-owner@vger.kernel.org>
Received: from mail-bk0-f46.google.com ([209.85.214.46]:38844 "EHLO
	mail-bk0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1753776Ab2DPMlY (ORCPT
	<rfc822;netdev@vger.kernel.org>); Mon, 16 Apr 2012 08:41:24 -0400
Received: by bkcik5 with SMTP id ik5so3591811bkc.19
        for <netdev@vger.kernel.org>; Mon, 16 Apr 2012 05:41:22 -0700 (PDT)
In-Reply-To: <201204161940178124485@gmail.com>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

On Mon, 2012-04-16 at 19:40 +0800, majianpeng wrote:
> From 8eb7c077dd3e746476a2a9fe059a676a6bc1a61f Mon Sep 17 00:00:00 2001
> From: majianpeng <majianpeng@gmail.com>
> Date: Mon, 16 Apr 2012 19:34:50 +0800
> Subject: [PATCH] net/ipv4:Remove two reports using kmemleak. Although those
>  two reports do not real kmemleak,I think can remove it  better.
> 
> Reported by kmemleak:
> unreferenced object 0xffff8800b8735090 (size 96):
>   comm "swapper/0", pid 1, jiffies 4294670057 (age 542.870s)
>   hex dump (first 32 bytes):
>     00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
>     00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
>   backtrace:
>     [<ffffffff81695df1>] kmemleak_alloc+0x21/0x50
>     [<ffffffff81117c3b>] __kmalloc+0x11b/0x1f0
>     [<ffffffff81188f57>] __register_sysctl_paths+0x127/0x1f0
>     [<ffffffff81189036>] register_sysctl_paths+0x16/0x20
>     [<ffffffff81cf1fc7>] ip_static_sysctl_init+0x17/0x19
>     [<ffffffff81cf2924>] inet_init+0xb4/0x2b2
>     [<ffffffff810001ca>] do_one_initcall+0x3a/0x160
>     [<ffffffff81cb9c7e>] kernel_init+0x12e/0x1b2
>     [<ffffffff816bdb94>] kernel_thread_helper+0x4/0x10
>     [<ffffffffffffffff>] 0xffffffffffffffff
> unreferenced object 0xffff8800b8735730 (size 96):
>   comm "swapper/0", pid 1, jiffies 4294670057 (age 542.870s)
>   hex dump (first 32 bytes):
>     80 80 e3 81 ff ff ff ff 00 00 00 00 01 00 00 00  ................
>     01 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
>   backtrace:
>     [<ffffffff81695df1>] kmemleak_alloc+0x21/0x50
>     [<ffffffff81117c3b>] __kmalloc+0x11b/0x1f0
>     [<ffffffff81188792>] __register_sysctl_table+0x62/0x4b0
>     [<ffffffff81188da5>] register_leaf_sysctl_tables+0x1c5/0x230
>     [<ffffffff81188ca0>] register_leaf_sysctl_tables+0xc0/0x230
>     [<ffffffff81188f89>] __register_sysctl_paths+0x159/0x1f0
>     [<ffffffff81189036>] register_sysctl_paths+0x16/0x20
>     [<ffffffff81cf1fc7>] ip_static_sysctl_init+0x17/0x19
>     [<ffffffff81cf2924>] inet_init+0xb4/0x2b2
>     [<ffffffff810001ca>] do_one_initcall+0x3a/0x160
>     [<ffffffff81cb9c7e>] kernel_init+0x12e/0x1b2
>     [<ffffffff816bdb94>] kernel_thread_helper+0x4/0x10
>     [<ffffffffffffffff>] 0xffffffffffffffff
> 
> 
> Signed-off-by: majianpeng <majianpeng@gmail.com>
> ---
>  net/ipv4/route.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> diff --git a/net/ipv4/route.c b/net/ipv4/route.c
> index 167ea10..817207b 100644
> --- a/net/ipv4/route.c
> +++ b/net/ipv4/route.c
> @@ -3499,12 +3499,13 @@ int __init ip_rt_init(void)
>  }
>  
>  #ifdef CONFIG_SYSCTL
> +static struct ctl_table_header *sysctl_header;
>  /*
>   * We really need to sanitize the damn ipv4 init order, then all
>   * this nonsense will go away.
>   */
>  void __init ip_static_sysctl_init(void)
>  {
> -	register_sysctl_paths(ipv4_path, ipv4_skeleton);
> +	sysctl_header = register_sysctl_paths(ipv4_path, ipv4_skeleton);
>  }
>  #endif

I dont think so.

Please check kmemleak_not_leak()