From: Avi Kivity <[email protected]>
To: Jens Axboe <[email protected]>, [email protected]
Subject: Re: IORING_OP_POLL_ADD slower than linux-aio IOCB_CMD_POLL
Date: Tue, 19 Apr 2022 22:41:48 +0300 [thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>
On 19/04/2022 20.14, Jens Axboe wrote:
> On 4/19/22 9:21 AM, Jens Axboe wrote:
>> On 4/19/22 6:31 AM, Jens Axboe wrote:
>>> On 4/19/22 6:21 AM, Avi Kivity wrote:
>>>> On 19/04/2022 15.04, Jens Axboe wrote:
>>>>> On 4/19/22 5:57 AM, Avi Kivity wrote:
>>>>>> On 19/04/2022 14.38, Jens Axboe wrote:
>>>>>>> On 4/19/22 5:07 AM, Avi Kivity wrote:
>>>>>>>> A simple webserver shows about 5% loss compared to linux-aio.
>>>>>>>>
>>>>>>>>
>>>>>>>> I expect the loss is due to an optimization that io_uring lacks -
>>>>>>>> inline completion vs workqueue completion:
>>>>>>> I don't think that's it, io_uring never punts to a workqueue for
>>>>>>> completions.
>>>>>> I measured this:
>>>>>>
>>>>>>
>>>>>>
>>>>>> Performance counter stats for 'system wide':
>>>>>>
>>>>>> 1,273,756 io_uring:io_uring_task_add
>>>>>>
>>>>>> 12.288597765 seconds time elapsed
>>>>>>
>>>>>> Which exactly matches with the number of requests sent. If that's the
>>>>>> wrong counter to measure, I'm happy to try again with the correct
>>>>>> counter.
>>>>> io_uring_task_add() isn't a workqueue, it's task_work. So that is
>>>>> expected.
>> Might actually be implicated. Not because it's a async worker, but
>> because I think we might be losing some affinity in this case. Looking
>> at traces, we're definitely bouncing between the poll completion side
>> and then execution the completion.
>>
>> Can you try this hack? It's against -git + for-5.19/io_uring. If you let
>> me know what base you prefer, I can do a version against that. I see
>> about a 3% win with io_uring with this, and was slower before against
>> linux-aio as you saw as well.
> Another thing to try - get rid of the IPI for TWA_SIGNAL, which I
> believe may be the underlying cause of it.
>
Won't it delay notification until the next io_uring_enter? Or does
io_uring only guarantee completions when you call it (and earlier
completions are best-effort?)
For Seastar it's not a problem, asking about the general io_uring
completion philosophy.
I'll try it tomorrow (also the other patch).
next prev parent reply other threads:[~2022-04-19 19:41 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-04-19 11:07 IORING_OP_POLL_ADD slower than linux-aio IOCB_CMD_POLL Avi Kivity
2022-04-19 11:38 ` Jens Axboe
2022-04-19 11:57 ` Avi Kivity
2022-04-19 12:04 ` Jens Axboe
2022-04-19 12:21 ` Avi Kivity
2022-04-19 12:31 ` Jens Axboe
2022-04-19 15:21 ` Jens Axboe
2022-04-19 15:51 ` Avi Kivity
2022-04-19 17:14 ` Jens Axboe
2022-04-19 19:41 ` Avi Kivity [this message]
2022-04-19 19:58 ` Jens Axboe
2022-04-20 11:55 ` Avi Kivity
2022-04-20 12:09 ` Jens Axboe
2022-04-21 9:05 ` Avi Kivity
2022-06-15 10:12 ` Avi Kivity
2022-06-15 10:48 ` Pavel Begunkov
2022-06-15 11:04 ` Avi Kivity
2022-06-15 11:07 ` Avi Kivity
2022-06-15 11:38 ` Pavel Begunkov
2022-06-15 12:21 ` Jens Axboe
2022-06-15 13:43 ` Avi Kivity
2022-06-15 11:30 ` Pavel Begunkov
2022-06-15 11:36 ` Avi Kivity
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox