Received-SPF: pass (google.com: domain of linux-crypto-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [RFCv2 1/9] tcp: authopt: Initial support and key management
From:   Leonard Crestez <cdleonard@gmail.com>
To:     Dmitry Safonov <0x7f454c46@gmail.com>
Cc:     Eric Dumazet <edumazet@google.com>,
        "David S. Miller" <davem@davemloft.net>,
        Herbert Xu <herbert@gondor.apana.org.au>,
        Kuniyuki Iwashima <kuniyu@amazon.co.jp>,
        David Ahern <dsahern@kernel.org>,
        Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>,
        Jakub Kicinski <kuba@kernel.org>,
        Yuchung Cheng <ycheng@google.com>,
        Francesco Ruggeri <fruggeri@arista.com>,
        Mat Martineau <mathew.j.martineau@linux.intel.com>,
        Christoph Paasch <cpaasch@apple.com>,
        Ivan Delalande <colona@arista.com>,
        Priyaranjan Jha <priyarjha@google.com>,
        Menglong Dong <dong.menglong@zte.com.cn>,
        open list <linux-kernel@vger.kernel.org>,
        linux-crypto@vger.kernel.org,
        Network Development <netdev@vger.kernel.org>,
        Dmitry Safonov <dima@arista.com>
References: <cover.1628544649.git.cdleonard@gmail.com>
 <67c1471683200188b96a3f712dd2e8def7978462.1628544649.git.cdleonard@gmail.com>
 <CAJwJo6aicw_KGQSM5U1=0X11QfuNf2dMATErSymytmpf75W=tA@mail.gmail.com>
 <1e2848fb-1538-94aa-0431-636fa314a36d@gmail.com>
Message-ID: <785d945e-c0d2-fee5-39d8-99dc06a074f1@gmail.com>
Date:   Thu, 12 Aug 2021 22:46:48 +0300
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.11.0
MIME-Version: 1.0
In-Reply-To: <1e2848fb-1538-94aa-0431-636fa314a36d@gmail.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Precedence: bulk

On 8/11/21 11:29 AM, Leonard Crestez wrote:
> On 8/10/21 11:41 PM, Dmitry Safonov wrote:
>> On Tue, 10 Aug 2021 at 02:50, Leonard Crestez <cdleonard@gmail.com> 
>>> +       /* If an old value exists for same local_id it is deleted */
>>> +       key_info = __tcp_authopt_key_info_lookup(sk, info, 
>>> opt.local_id);
>>> +       if (key_info)
>>> +               tcp_authopt_key_del(sk, info, key_info);
>>> +       key_info = sock_kmalloc(sk, sizeof(*key_info), GFP_KERNEL | 
>>> __GFP_ZERO);
>>> +       if (!key_info)
>>> +               return -ENOMEM;
>>
>> 1. You don't need sock_kmalloc() together with tcp_authopt_key_del().
>>      It just frees the memory and allocates it back straight away - no
>> sense in doing that.
> 
> The list is scanned in multiple places in later commits using nothing 
> but an rcu_read_lock, this means that keys can't be updated in-place.
> 
>> 2. I think RFC says you must not allow a user to change an existing key:
>>> MKT parameters are not changed. Instead, new MKTs can be installed, 
>>> and a connection
>>> can change which MKT it uses.
>>
>> IIUC, it means that one can't just change an existing MKT, but one can
>> remove and later
>> add MKT with the same (send_id, recv_id, src_addr/port, dst_addr/port).
>>
>> So, a reasonable thing to do:
>> if (key_info)
>>      return -EEXIST.
> 
> You're right, making the user delete keys explicitly is better.

On a second thought this might be required to mark keys as "send-only" 
and "recv-only" atomically.

Separately from RFC5925 some vendors implement a "keychain" model based 
on RFC8177 where each key has a distinct "accept-lifetime" and a 
"send-lifetime". This could be implemented by adding flags 
"expired_for_send" and "expired_for_recv" but requires the ability to 
set an expiration mark without the key ever being deleted.

--
Regards,
Leonard