From mboxrd@z Thu Jan  1 00:00:00 1970
From: Eric Dumazet <eric.dumazet@gmail.com>
Subject: Re: ip6gre_tunnel_xmit() leaking an skb?
Date: Mon, 04 Feb 2013 07:04:45 -0800
Message-ID: <1359990285.30177.140.camel@edumazet-glaptop>
References: <CA+ydwtqabWALE=CtNWrDY=uiZqpHzUnami6OR-_OrK8TgNivQw@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
Cc: "David S. Miller" <davem@davemloft.net>,
	Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>,
	James Morris <jmorris@namei.org>,
	Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>,
	Patrick McHardy <kaber@trash.net>, netdev@vger.kernel.org,
	Dmitry Kozlov <xeb@mail.ru>
To: Tommi Rantala <tt.rantala@gmail.com>
Return-path: <netdev-owner@vger.kernel.org>
Received: from mail-pa0-f53.google.com ([209.85.220.53]:58302 "EHLO
	mail-pa0-f53.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1755685Ab3BDPEt (ORCPT
	<rfc822;netdev@vger.kernel.org>); Mon, 4 Feb 2013 10:04:49 -0500
Received: by mail-pa0-f53.google.com with SMTP id bg4so3361996pad.26
        for <netdev@vger.kernel.org>; Mon, 04 Feb 2013 07:04:48 -0800 (PST)
In-Reply-To: <CA+ydwtqabWALE=CtNWrDY=uiZqpHzUnami6OR-_OrK8TgNivQw@mail.gmail.com>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

On Mon, 2013-02-04 at 14:29 +0200, Tommi Rantala wrote:
> Hello,
> 
> kmemleak is giving me some complaints on boot:
> 
> # echo scan > /sys/kernel/debug/kmemleak
> # echo scan > /sys/kernel/debug/kmemleak
> [   13.084513] kmemleak: 2 new suspected memory leaks (see
> /sys/kernel/debug/kmemleak)
> # cat /sys/kernel/debug/kmemleak
> unreferenced object 0xffff88003b4e3c00 (size 256):
>   comm "dhcpcd", pid 1661, jiffies 4294670402 (age 15.097s)
>   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 80 4a 3b 00 88 ff ff  ..........J;....
>   backtrace:
>     [<ffffffff81b3d636>] kmemleak_alloc+0x26/0x50
>     [<ffffffff81193514>] kmem_cache_alloc_node+0xb4/0x180
>     [<ffffffff817c8fc0>] __alloc_skb+0x60/0x2a0
>     [<ffffffff817c57c1>] sock_alloc_send_pskb+0x1d1/0x340
>     [<ffffffff8190aa38>] packet_sendmsg+0xa88/0xd40
>     [<ffffffff817c0b60>] sock_sendmsg+0xb0/0xe0
>     [<ffffffff817c232e>] sys_sendto+0x11e/0x160
>     [<ffffffff81b62859>] system_call_fastpath+0x16/0x1b
>     [<ffffffffffffffff>] 0xffffffffffffffff
> unreferenced object 0xffff88003af07800 (size 1024):
>   comm "dhcpcd", pid 1661, jiffies 4294670402 (age 15.097s)
>   hex dump (first 32 bytes):
>     00 74 f0 3a 00 88 ff ff fe ed ca fe 28 00 00 00  .t.:........(...
>     28 00 00 00 ad 00 00 00 f0 03 1d b7 7b cb c5 ee  (...........{...
>   backtrace:
>     [<ffffffff81b3d636>] kmemleak_alloc+0x26/0x50
>     [<ffffffff81195ce0>] __kmalloc_node_track_caller+0xe0/0x1e0
>     [<ffffffff817c8efc>] __kmalloc_reserve+0x3c/0xa0
>     [<ffffffff817c8ff1>] __alloc_skb+0x91/0x2a0
>     [<ffffffff817c57c1>] sock_alloc_send_pskb+0x1d1/0x340
>     [<ffffffff8190aa38>] packet_sendmsg+0xa88/0xd40
>     [<ffffffff817c0b60>] sock_sendmsg+0xb0/0xe0
>     [<ffffffff817c232e>] sys_sendto+0x11e/0x160
>     [<ffffffff81b62859>] system_call_fastpath+0x16/0x1b
>     [<ffffffffffffffff>] 0xffffffffffffffff
> #
> 
> Patching ip6gre_tunnel_xmit() to not return -1 fixes the kmemleak
> complaint, but I have no clue if the patch is otherwise correct:
> 
> $ git diff net/ipv6/ip6_gre.c
> diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c
> index c727e47..131dd09 100644
> --- a/net/ipv6/ip6_gre.c
> +++ b/net/ipv6/ip6_gre.c
> @@ -960,7 +960,7 @@ static netdev_tx_t ip6gre_tunnel_xmit(struct sk_buff *skb,
>         int ret;
> 
>         if (!ip6_tnl_xmit_ctl(t))
> -               return -1;
> +               goto tx_err;
> 
>         switch (skb->protocol) {
>         case htons(ETH_P_IP):

Patch seems fine to me, please submit an official patch ;)