Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <22175690a4e89a78abcb8244dfd0bdd0005267a5.camel@redhat.com>
Subject: Re: [PATCH v3] net: ax25: Fix deadlock caused by skb_recv_datagram
 in ax25_recvmsg
From:   Paolo Abeni <pabeni@redhat.com>
To:     Duoming Zhou <duoming@zju.edu.cn>, linux-kernel@vger.kernel.org
Cc:     jreuter@yaina.de, ralf@linux-mips.org, davem@davemloft.net,
        edumazet@google.com, kuba@kernel.org, netdev@vger.kernel.org,
        linux-hams@vger.kernel.org, thomas@osterried.de
Date:   Thu, 09 Jun 2022 10:41:02 +0200
In-Reply-To: <20220608012923.17505-1-duoming@zju.edu.cn>
References: <20220608012923.17505-1-duoming@zju.edu.cn>
Content-Type: text/plain; charset="UTF-8"
User-Agent: Evolution 3.42.4 (3.42.4-2.fc35) 
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Wed, 2022-06-08 at 09:29 +0800, Duoming Zhou wrote:
> The skb_recv_datagram() in ax25_recvmsg() will hold lock_sock
> and block until it receives a packet from the remote. If the client
> doesn`t connect to server and calls read() directly, it will not
> receive any packets forever. As a result, the deadlock will happen.
> 
> The fail log caused by deadlock is shown below:
> 
> [  369.606973] INFO: task ax25_deadlock:157 blocked for more than 245 seconds.
> [  369.608919] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
> [  369.613058] Call Trace:
> [  369.613315]  <TASK>
> [  369.614072]  __schedule+0x2f9/0xb20
> [  369.615029]  schedule+0x49/0xb0
> [  369.615734]  __lock_sock+0x92/0x100
> [  369.616763]  ? destroy_sched_domains_rcu+0x20/0x20
> [  369.617941]  lock_sock_nested+0x6e/0x70
> [  369.618809]  ax25_bind+0xaa/0x210
> [  369.619736]  __sys_bind+0xca/0xf0
> [  369.620039]  ? do_futex+0xae/0x1b0
> [  369.620387]  ? __x64_sys_futex+0x7c/0x1c0
> [  369.620601]  ? fpregs_assert_state_consistent+0x19/0x40
> [  369.620613]  __x64_sys_bind+0x11/0x20
> [  369.621791]  do_syscall_64+0x3b/0x90
> [  369.622423]  entry_SYSCALL_64_after_hwframe+0x46/0xb0
> [  369.623319] RIP: 0033:0x7f43c8aa8af7
> [  369.624301] RSP: 002b:00007f43c8197ef8 EFLAGS: 00000246 ORIG_RAX: 0000000000000031
> [  369.625756] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f43c8aa8af7
> [  369.626724] RDX: 0000000000000010 RSI: 000055768e2021d0 RDI: 0000000000000005
> [  369.628569] RBP: 00007f43c8197f00 R08: 0000000000000011 R09: 00007f43c8198700
> [  369.630208] R10: 0000000000000000 R11: 0000000000000246 R12: 00007fff845e6afe
> [  369.632240] R13: 00007fff845e6aff R14: 00007f43c8197fc0 R15: 00007f43c8198700
> 
> This patch moves the skb_recv_datagram() before lock_sock() in order that
> other functions that need lock_sock could be executed. What`s more, we
> add skb_free_datagram() before goto out in order to mitigate memory leak.
> 
> Suggested-by: Thomas Osterried <thomas@osterried.de>
> Signed-off-by: Duoming Zhou <duoming@zju.edu.cn>
> Reported-by: Thomas Habets <thomas@@habets.se>
> ---
> Changes in v3:
>   - Add skb_free_datagram() before goto out in order to mitigate memory leak.
> 
>  net/ax25/af_ax25.c | 12 +++++++-----
>  1 file changed, 7 insertions(+), 5 deletions(-)
> 
> diff --git a/net/ax25/af_ax25.c b/net/ax25/af_ax25.c
> index 95393bb2760..62aa5993093 100644
> --- a/net/ax25/af_ax25.c
> +++ b/net/ax25/af_ax25.c
> @@ -1665,6 +1665,11 @@ static int ax25_recvmsg(struct socket *sock, struct msghdr *msg, size_t size,
>  	int copied;
>  	int err = 0;
>  
> +	/* Now we can treat all alike */
> +	skb = skb_recv_datagram(sk, flags, &err);
> +	if (!skb)
> +		goto done;
> +

Note that this causes a behavior change: before this patch, calling
recvmsg() on unconnected seqpacket sockets returned immediatelly with
an error (due to the the check below), now it blocks. 

The change may confuse (== break) user-space applications. I think it
would be better replacing skb_recv_datagram with an open-coded variant
of it releasing the socket lock before the
__skb_wait_for_more_packets() call and re-acquiring it after such call.
Somewhat alike __unix_dgram_recvmsg().

In any case this lacks a 'Fixes' pointing to the commit introducing the
issue addressed here.

Thanks,

Paolo