Re: [PATCH 7.3 14/38] io_uring/poll: avoid retry wakeup loss when dropping ownership

[Jiri Slaby <jirislaby@kernel.org>]

Hi,

On 28. 10. 26, 18:00, Greg Kroah-Hartman wrote:
> 7.3-stable review patch.  If anyone has any objections, please let me know.
> 
> ------------------
> 
> From: Juanlu Herrero <juanlu@fastmail.com>
> 
> commit 937d38dcacd675dc0572d6c9b6743960df3128ed upstream.

Our machinery flags this is needed as a fix:
67bb1ad9a517 ("io_uring/poll: don't loop on HUP with multiple references")

> If a wakeup sets IO_POLL_RETRY_FLAG between the owning task consuming
> the existing instance of IO_POLL_RETRY_FLAG and the loop check
> of io_poll_check_events(), then we stall on processing the request
> until an unrelated wakeup comes along.
> 
> Using atomic_cmpxchg instead of atomic_sub_return to guarantee both the
> count v and poll_refs are 0 and no wake-ups are missed.
> 
> Cc: stable@vger.kernel.org
> Fixes: a26a35e9019f ("io_uring: make poll refs more robust")
> Signed-off-by: Juanlu Herrero <juanlu@fastmail.com>
> Assisted-by: LLM
> Suggested-by: Anand Natarajan <anand@meta.com>
> Link: https://patch.msgid.link/20261013-pu-cmpxchg-v1-1-4ddc10fd8a03@fastmail.com
> Signed-off-by: Jens Axboe <axboe@kernel.dk>
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> ---
>   io_uring/poll.c |    3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)
> 
> --- a/io_uring/poll.c
> +++ b/io_uring/poll.c
> @@ -330,7 +330,8 @@ static int io_poll_check_events(struct i
>   		 * Release all references, retry if someone tried to restart
>   		 * task_work while we were executing it.
>   		 */
> -	} while (atomic_sub_return(v, &req->poll_refs) & IO_POLL_REF_MASK);
> +		v &= IO_POLL_REF_MASK;
> +	} while (atomic_cmpxchg(&req->poll_refs, v, 0) != v);
>   
>   	io_napi_add(req);
>   	return IOU_POLL_NO_ACTION;

thanks,
-- 
js
suse labs