Re: [PATCH net] net/iucv: fix use-after-free of a severed iucv_path

[Alexandra Winter <wintera@linux.ibm.com>]

On 21.07.26 09:30, patchwork-bot+netdevbpf@kernel.org wrote:
> Hello:
> 
> This patch was applied to netdev/net.git (main)
> by Paolo Abeni <pabeni@redhat.com>:
> 
-- Original Patch for reference:  --
> From: Bryam Vargas <hexlabsecurity@proton.me>
>
> af_iucv queues not-yet-received message notifications on iucv->message_q,
> each holding a raw pointer to the connection's iucv_path.  When the peer
> severs the connection, iucv_sever_path() frees that path with
> iucv_path_free() but leaves the notifications queued.  A later recvmsg()
> drains message_q via iucv_process_message_q() and hands the stale path to
> message_receive() -- a use-after-free of the freed iucv_path.
>
> Drop the queued notifications when the path is severed; once the path is
> gone they can no longer be received.  This also frees the notifications
> leaked when a socket is closed with messages still queued.
>
> Fixes: f0703c80e515 ("[AF_IUCV]: postpone receival of iucv-packets")
> Closes: https://sashiko.dev/#/patchset/20260705-b4-disp-fc79c0dc-v1-1-d2cdcb57afa9@proton.me?part=1
> Cc: stable@vger.kernel.org
> Signed-off-by: Bryam Vargas <hexlabsecurity@proton.me>
> ---
>  net/iucv/af_iucv.c | 14 ++++++++++++++
>  1 file changed, 14 insertions(+)
>
> diff --git a/net/iucv/af_iucv.c b/net/iucv/af_iucv.c
> index fed240b453bd..2869a103f7fa 100644
> --- a/net/iucv/af_iucv.c
> +++ b/net/iucv/af_iucv.c
> @@ -337,6 +337,7 @@ static void iucv_sever_path(struct sock *sk, int with_user_data)
>  	unsigned char user_data[16];
>  	struct iucv_sock *iucv = iucv_sk(sk);
>  	struct iucv_path *path = iucv->path;
> +	struct sock_msg_q *p, *n;
>
>  	/* Whoever resets the path pointer, must sever and free it. */
>  	if (xchg(&iucv->path, NULL)) {
> @@ -348,6 +349,19 @@ static void iucv_sever_path(struct sock *sk, int with_user_data)
>  		} else
>  			pr_iucv->path_sever(path, NULL);
>  		iucv_path_free(path);
> +
> +		/*
> +		 * Message notifications queued on message_q still reference
> +		 * the now freed path; drop them, otherwise a later recvmsg()
> +		 * would pass the freed iucv_path to message_receive() via
> +		 * iucv_process_message_q().
> +		 */
> +		spin_lock_bh(&iucv->message_q.lock);
> +		list_for_each_entry_safe(p, n, &iucv->message_q.list, list) {
> +			list_del(&p->list);
> +			kfree(p);
> +		}
> +		spin_unlock_bh(&iucv->message_q.lock);
>  	}
>  }
>
>
--- end of patch --

> 
> Here is the summary with links:
>   - [net] net/iucv: fix use-after-free of a severed iucv_path
>     https://git.kernel.org/netdev/net/c/be7cc4656eb1
> 
> You are awesome, thank you!


Ah, Paolo was faster than me.

@Bryam and for the records,
this should not be a use-after-free. After iucv_server_path it should not be possible to call
iucv_process_message_q() anymore. I agree that it is a message leak, the pending messages indicators
in iucv->message_q are not freed anywhere. I would have preferred to do that in iucv_sock_close()
instead of iucv_sever_path() for symmetry with iucv_sock_alloc(), but this should work as well.

Having said that, as we discussed in [1] the socket locking in af_iucv wrt receive path has
deficiencies, and we will follow up anyhow.

[1] https://lore.kernel.org/all/20260711041119.12764-1-hexlabsecurity@proton.me/