Received-SPF: pass (google.com: domain of linux-kernel+bounces-104689-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Message-ID: <7b82679f-9b69-4568-a61d-03eb1e4afc18@gmail.com>
Date: Fri, 15 Mar 2024 17:02:05 +0000
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH] net: Do not break out of sk_stream_wait_memory() with
 TIF_NOTIFY_SIGNAL
Content-Language: en-US
To: Sascha Hauer <s.hauer@pengutronix.de>, netdev@vger.kernel.org
Cc: kernel@pengutronix.de, linux-kernel@vger.kernel.org,
 Paolo Abeni <pabeni@redhat.com>, Jens Axboe <axboe@kernel.dk>,
 io-uring@vger.kernel.org
References: <20240315100159.3898944-1-s.hauer@pengutronix.de>
From: Pavel Begunkov <asml.silence@gmail.com>
In-Reply-To: <20240315100159.3898944-1-s.hauer@pengutronix.de>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit

On 3/15/24 10:01, Sascha Hauer wrote:
> It can happen that a socket sends the remaining data at close() time.
> With io_uring and KTLS it can happen that sk_stream_wait_memory() bails
> out with -512 (-ERESTARTSYS) because TIF_NOTIFY_SIGNAL is set for the
> current task. This flag has been set in io_req_normal_work_add() by
> calling task_work_add().

The entire idea of task_work is to interrupt syscalls and let io_uring
do its job, otherwise it wouldn't free resources it might be holding,
and even potentially forever block the syscall.

I'm not that sure about connect / close (are they not restartable?),
but it doesn't seem to be a good idea for sk_stream_wait_memory(),
which is the normal TCP blocking send path. I'm thinking of some kinds
of cases with a local TCP socket pair, the tx queue is full as well
and the rx queue of the other end, and io_uring has to run to receive
the data.

If interruptions are not welcome you can use different io_uring flags,
see IORING_SETUP_COOP_TASKRUN and/or IORING_SETUP_DEFER_TASKRUN.

Maybe I'm missing something, why not restart your syscall?


> This patch replaces signal_pending() with task_sigpending(), thus ignoring
> the TIF_NOTIFY_SIGNAL flag.
> 
> A discussion of this issue can be found at
> https://lore.kernel.org/20231010141932.GD3114228@pengutronix.de
> 
> Suggested-by: Jens Axboe <axboe@kernel.dk>
> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
> ---
>   net/core/stream.c | 6 +++---
>   1 file changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/net/core/stream.c b/net/core/stream.c
> index 96fbcb9bbb30a..e9e17b48e0122 100644
> --- a/net/core/stream.c
> +++ b/net/core/stream.c
> @@ -67,7 +67,7 @@ int sk_stream_wait_connect(struct sock *sk, long *timeo_p)
>   			return -EPIPE;
>   		if (!*timeo_p)
>   			return -EAGAIN;
> -		if (signal_pending(tsk))
> +		if (task_sigpending(tsk))
>   			return sock_intr_errno(*timeo_p);
>   
>   		add_wait_queue(sk_sleep(sk), &wait);
> @@ -103,7 +103,7 @@ void sk_stream_wait_close(struct sock *sk, long timeout)
>   		do {
>   			if (sk_wait_event(sk, &timeout, !sk_stream_closing(sk), &wait))
>   				break;
> -		} while (!signal_pending(current) && timeout);
> +		} while (!task_sigpending(current) && timeout);
>   
>   		remove_wait_queue(sk_sleep(sk), &wait);
>   	}
> @@ -134,7 +134,7 @@ int sk_stream_wait_memory(struct sock *sk, long *timeo_p)
>   			goto do_error;
>   		if (!*timeo_p)
>   			goto do_eagain;
> -		if (signal_pending(current))
> +		if (task_sigpending(current))
>   			goto do_interrupted;
>   		sk_clear_bit(SOCKWQ_ASYNC_NOSPACE, sk);
>   		if (sk_stream_memory_free(sk) && !vm_wait)

-- 
Pavel Begunkov