Message-ID: <565DB363.6030702@iogearbox.net>
Date: Tue, 01 Dec 2015 15:49:07 +0100
From: Daniel Borkmann <daniel@iogearbox.net>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.4.0
MIME-Version: 1.0
To: Eric Dumazet <eric.dumazet@gmail.com>
CC: Dmitry Vyukov <dvyukov@google.com>,
        "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 <netdev@vger.kernel.org>,
        LKML <linux-kernel@vger.kernel.org>,
        Vlad Yasevich <vyasevich@gmail.com>,
        Neil Horman <nhorman@tuxdriver.com>, linux-sctp@vger.kernel.org,
        syzkaller <syzkaller@googlegroups.com>,
        Kostya Serebryany <kcc@google.com>,
        Alexander Potapenko <glider@google.com>,
        Sasha Levin <sasha.levin@oracle.com>,
        Eric Dumazet <edumazet@google.com>
Subject: Re: memory leak in do_ipv6_setsockopt
References: <CACT4Y+bv56b+opCd=5SHmN47MAHEFyZA18kSsu-Fa+XODj20PA@mail.gmail.com>	 <1448977016.25582.18.camel@edumazet-glaptop2.roam.corp.google.com>	 <565DA9BE.3060006@iogearbox.net>	 <1448979404.25582.23.camel@edumazet-glaptop2.roam.corp.google.com>	 <565DAD8D.9020800@iogearbox.net> <1448980723.25582.24.camel@edumazet-glaptop2.roam.corp.google.com>
In-Reply-To: <1448980723.25582.24.camel@edumazet-glaptop2.roam.corp.google.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1163
Lines: 32

On 12/01/2015 03:38 PM, Eric Dumazet wrote:
> On Tue, 2015-12-01 at 15:24 +0100, Daniel Borkmann wrote:
>> On 12/01/2015 03:16 PM, Eric Dumazet wrote:
>>> On Tue, 2015-12-01 at 15:07 +0100, Daniel Borkmann wrote:
>>>
>>>> Yeah, we miss inet6_destroy_sock() in SCTP. :-(
>>>>
>>>> Looks good to me.
>>>
>>> OK, I will send a formal (and tested ;) ) patch.
>>
>> I was shortly wondering whether there could be a use-after-free by
>> doing this after sctp_destroy_sock() due to the sctp_endpoint_destroy()
>> that would eventually drop a ref on the socket, but the endpoint holds
>> a separate ref, so we should be good.
>
> More generically ->destroy() caller must keep a reference on the socket.
>
> inet_csk_destroy_sock() for example uses sk after
>
> sk->sk_prot->destroy(sk);

Right, and later on, we might call into ->sk_destruct() when there are no
more refs (in SCTP case: sctp_destruct_sock()).

Thanks,
Daniel
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/