netdev - Re: [PATCH] tipc: ensure skb->lock is initialised

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 9 Jul 2019 15:45:32 +0200
From:   Eric Dumazet <eric.dumazet@...il.com>
To:     Jon Maloy <jon.maloy@...csson.com>,
        Eric Dumazet <eric.dumazet@...il.com>,
        Chris Packham <Chris.Packham@...iedtelesis.co.nz>,
        "ying.xue@...driver.com" <ying.xue@...driver.com>,
        "davem@...emloft.net" <davem@...emloft.net>
Cc:     "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        "tipc-discussion@...ts.sourceforge.net" 
        <tipc-discussion@...ts.sourceforge.net>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] tipc: ensure skb->lock is initialised



On 7/9/19 3:25 PM, Jon Maloy wrote:
> 
> 
>> -----Original Message-----
>> From: Eric Dumazet <eric.dumazet@...il.com>
>> Sent: 9-Jul-19 03:31
>> To: Chris Packham <Chris.Packham@...iedtelesis.co.nz>; Eric Dumazet
>> <eric.dumazet@...il.com>; Jon Maloy <jon.maloy@...csson.com>;
>> ying.xue@...driver.com; davem@...emloft.net
>> Cc: netdev@...r.kernel.org; tipc-discussion@...ts.sourceforge.net; linux-
>> kernel@...r.kernel.org
>> Subject: Re: [PATCH] tipc: ensure skb->lock is initialised
>>
>>
>>
>> On 7/8/19 11:13 PM, Chris Packham wrote:
>>> On 9/07/19 8:43 AM, Chris Packham wrote:
>>>> On 8/07/19 8:18 PM, Eric Dumazet wrote:
>>>>>
>>>>>
>>>>> On 7/8/19 12:53 AM, Chris Packham wrote:
>>>>>> tipc_named_node_up() creates a skb list. It passes the list to
>>>>>> tipc_node_xmit() which has some code paths that can call
>>>>>> skb_queue_purge() which relies on the list->lock being initialised.
>>>>>> Ensure tipc_named_node_up() uses skb_queue_head_init() so that the
>>>>>> lock is explicitly initialised.
>>>>>>
>>>>>> Signed-off-by: Chris Packham <chris.packham@...iedtelesis.co.nz>
>>>>>
>>>>> I would rather change the faulty skb_queue_purge() to
>>>>> __skb_queue_purge()
>>>>>
>>>>
>>>> Makes sense. I'll look at that for v2.
>>>>
>>>
>>> Actually maybe not. tipc_rcast_xmit(), tipc_node_xmit_skb(),
>>> tipc_send_group_msg(), __tipc_sendmsg(), __tipc_sendstream(), and
>>> tipc_sk_timeout() all use skb_queue_head_init(). So my original change
>>> brings tipc_named_node_up() into line with them.
>>>
>>> I think it should be safe for tipc_node_xmit() to use
>>> __skb_queue_purge() since all the callers seem to have exclusive
>>> access to the list of skbs. It still seems that the callers should all
>>> use
>>> skb_queue_head_init() for consistency.
> 
> I agree with that.
> 
>>>
>>
>> No, tipc does not use the list lock (it relies on the socket lock)  and therefore
>> should consistently use __skb_queue_head_init() instead of
>> skb_queue_head_init()
> 
> TIPC is using the list lock at message reception within the scope of tipc_sk_rcv()/tipc_skb_peek_port(), so it is fundamental that the lock always is correctly initialized.

Where is the lock acquired, why was it only acquired by queue purge and not normal dequeues ???

> 
>>
> [...]
>>
>> tipc_link_xmit() for example never acquires the spinlock, yet uses skb_peek()
>> and __skb_dequeue()
> 
> 
> You should look at tipc_node_xmit instead. Node local messages are sent directly to tipc_sk_rcv(), and never go through tipc_link_xmit()

tipc_node_xmit() calls tipc_link_xmit() eventually, right ?

Please show me where the head->lock is acquired, and why it needed.

If this is mandatory, then more fixes are needed than just initializing the lock for lockdep purposes.