Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:5 as permitted sender) client-ip=2620:137:e000::3:5;
Message-ID: <30fe685f-d09b-48b7-840d-9d19d6c183db@arista.com>
Date:   Wed, 29 Nov 2023 15:42:02 +0000
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH v3 5/7] net/tcp: Don't add key with non-matching VRF on
 connected sockets
Content-Language: en-US
To:     David Ahern <dsahern@kernel.org>
Cc:     linux-kernel@vger.kernel.org,
        Dmitry Safonov <0x7f454c46@gmail.com>,
        Francesco Ruggeri <fruggeri05@gmail.com>,
        Salam Noureddine <noureddine@arista.com>,
        Simon Horman <horms@kernel.org>, netdev@vger.kernel.org,
        Eric Dumazet <edumazet@google.com>,
        Paolo Abeni <pabeni@redhat.com>,
        Jakub Kicinski <kuba@kernel.org>,
        "David S. Miller" <davem@davemloft.net>
References: <20231128205749.312759-1-dima@arista.com>
 <20231128205749.312759-6-dima@arista.com>
 <eb9a46a5-d074-445a-9e18-514ef78395d7@kernel.org>
From:   Dmitry Safonov <dima@arista.com>
In-Reply-To: <eb9a46a5-d074-445a-9e18-514ef78395d7@kernel.org>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

Hi David,

On 11/29/23 01:34, David Ahern wrote:
> On 11/28/23 1:57 PM, Dmitry Safonov wrote:
>> If the connection was established, don't allow adding TCP-AO keys that
>> don't match the peer. Currently, there are checks for ip-address
>> matching, but L3 index check is missing. Add it to restrict userspace
> 
> you say L3 index check is missing - add it. yet ...
> 
>> shooting itself somewhere.
>>
>> Fixes: 248411b8cb89 ("net/tcp: Wire up l3index to TCP-AO")
>> Signed-off-by: Dmitry Safonov <dima@arista.com>
>> ---
>>  net/ipv4/tcp_ao.c | 3 +++
>>  1 file changed, 3 insertions(+)
>>
>> diff --git a/net/ipv4/tcp_ao.c b/net/ipv4/tcp_ao.c
>> index bf41be6d4721..2d000e275ce7 100644
>> --- a/net/ipv4/tcp_ao.c
>> +++ b/net/ipv4/tcp_ao.c
>> @@ -1608,6 +1608,9 @@ static int tcp_ao_add_cmd(struct sock *sk, unsigned short int family,
>>  		if (!dev || !l3index)
>>  			return -EINVAL;
>>  
>> +		if (!((1 << sk->sk_state) & (TCPF_LISTEN | TCPF_CLOSE)))
>> +			return -EINVAL;
> 
> ... this is checking socket state.

Right you are, it should have been under check for
: if (bound_dev_if != cmd.ifindex)

Currently it's warning for all sockets (which can be re-bound), but for
sockets in the connected state it doesn't make sense as the key lookup
is not expecting non peer-matching key post connect()/accept().

In this patch version the check will restrict adding a key on a
connected socket with VRF regardless if it's matching the bound VRF.
Will fix!

Thanks for spotting this,
             Dmitry