Re: [PATCH] splice: report related fsnotify events

[Jens Axboe <axboe@kernel.dk>]

On 4/3/23 11:15?AM, Amir Goldstein wrote:
>> On 4/3/23 11:00?AM, Amir Goldstein wrote:
>>> On Wed, Mar 22, 2023 at 9:08?AM Amir Goldstein <amir73il@gmail.com> wrote:
>>>>
>>>> On Wed, Mar 22, 2023 at 8:51?AM Chung-Chiang Cheng <cccheng@synology.com> wrote:
>>>>>
>>>>> The fsnotify ACCESS and MODIFY event are missing when manipulating a file
>>>>> with splice(2).
>>>>>
>>>
>>> Jens, Jan,
>>>
>>> FYI, I've audited aio routines and found that
>>> fsnotify_access()/modify() are also missing in aio_complete_rw()
>>> and in io_complete_rw_iopoll() (io_req_io_end() should be called?).
>>>
>>> I am not using/testing aio/io_uring usually, so I wasn't planning on sending
>>> a patch any time soon. I'll get to it someday.
>>> Just wanted to bring this to public attention in case someone is
>>> interested in testing/fixing.
>>
>> aio has never done fsnotify, but I think that's probably an oversight.
> 
> I know. and I am not keen either on fixing something that nobody
> complained about.

Nobody does buffered IO with aio (as it doesn't work), which is probably
why nobody complained.

>> io_uring does do it for non-polled IO, I don't think there's much point
>> in adding it to IOPOLL however. Not really seeing any use cases where
>> that would make sense.
>>
> 
> Users subscribe to fsnotify because they want to be notified of changes/
> access to a file.
> Why do you think that polled IO should be exempt?

Because it's a drastically different use case. If you're doing high
performance polled IO, then you'd never rely on something as slow as
fsnotify to tell you of any changes that happened to a device or file.
That would be counter productive.

-- 
Jens Axboe