linux-block.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Pavel Begunkov <asml.silence@gmail.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Jens Axboe <axboe@kernel.dk>, Ingo Molnar <mingo@redhat.com>,
	linux-block@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 1/2] sched/wait: Add wait_threshold
Date: Mon, 23 Sep 2019 19:37:46 +0300	[thread overview]
Message-ID: <3e359937-5b19-8a4c-4243-ba2edff68504@gmail.com> (raw)
In-Reply-To: <20190923071932.GD2349@hirez.programming.kicks-ass.net>


[-- Attachment #1.1: Type: text/plain, Size: 1909 bytes --]

Just in case duplicating a mail from the cover-letter thread:


It could be done with @cond indeed, that's how it works for now.
However, this addresses performance issues only.

The problem with wait_event_*() is that, if we have a counter and are
trying to wake up tasks after each increment, it would schedule each
waiting task O(threshold) times just for it to spuriously check @cond
and go back to sleep. All that overhead (memory barriers, registers
save/load, accounting, etc) turned out to be enough for some workloads
to slow down the system.

With this specialisation it still traverses a wait list and makes
indirect calls to the checker callback, but the list supposedly is
fairly  small, so performance there shouldn't be a problem, at least for
now.

Regarding semantics; It should wake a task when a value passed to
wake_up_threshold() is greater or equal then a task's threshold, that is
specified individually for each task in wait_threshold_*().

In pseudo code:
```
def wake_up_threshold(n, wait_queue):
	for waiter in wait_queue:
		waiter.wake_up_if(n >= waiter.threshold);
```

Any thoughts how to do it better? Ideas are very welcome.

BTW, this monster is mostly a copy-paste from wait_event_*(),
wait_bit_*(). We could try to extract some common parts from these
three, but that's another topic.


On 23/09/2019 10:19, Peter Zijlstra wrote:
> On Sun, Sep 22, 2019 at 11:08:50AM +0300, Pavel Begunkov (Silence) wrote:
>> From: Pavel Begunkov <asml.silence@gmail.com>
>>
>> Add wait_threshold -- a custom wait_event derivative, that waits until
>> a value is equal to or greater than the specified threshold.
> 
> This is quite insufficient justification for this monster... what exact
> semantics do you want?
> 
> Why can't you do this exact same with a slightly more complicated @cond
> ?
> 

-- 
Yours sincerely,
Pavel Begunkov


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

  reply	other threads:[~2019-09-23 16:37 UTC|newest]

Thread overview: 38+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-09-22  8:08 [PATCH v2 0/2] Optimise io_uring completion waiting Pavel Begunkov (Silence)
2019-09-22  8:08 ` [PATCH v2 1/2] sched/wait: Add wait_threshold Pavel Begunkov (Silence)
2019-09-23  7:19   ` Peter Zijlstra
2019-09-23 16:37     ` Pavel Begunkov [this message]
2019-09-23 19:27       ` Peter Zijlstra
2019-09-23 20:23         ` Peter Zijlstra
2019-09-24  6:44         ` Pavel Begunkov
2019-09-22  8:08 ` [PATCH v2 2/2] io_uring: Optimise cq waiting with wait_threshold Pavel Begunkov (Silence)
2019-09-22 15:51 ` [PATCH v2 0/2] Optimise io_uring completion waiting Jens Axboe
2019-09-23  8:35   ` Ingo Molnar
2019-09-23 16:21     ` Pavel Begunkov
2019-09-23 16:32       ` Pavel Begunkov
2019-09-23 20:48         ` Jens Axboe
2019-09-23 23:00           ` Jens Axboe
2019-09-24  7:06             ` Pavel Begunkov
2019-09-24  8:02               ` Jens Axboe
2019-09-24  8:27                 ` Jens Axboe
2019-09-24  8:36                   ` Jens Axboe
2019-09-24  9:33                     ` Pavel Begunkov
2019-09-24 10:11                       ` Jens Axboe
2019-09-24  9:49                     ` Peter Zijlstra
2019-09-24 10:13                       ` Jens Axboe
2019-09-24 10:34                         ` Jens Axboe
2019-09-24 11:11                           ` Pavel Begunkov
2019-09-24 11:15                             ` Jens Axboe
2019-09-24 11:23                               ` Pavel Begunkov
2019-09-24 13:13                                 ` Jens Axboe
2019-09-24 17:33                                   ` Pavel Begunkov
2019-09-24 17:46                                     ` Jens Axboe
2019-09-24 18:28                                       ` Pavel Begunkov
2019-09-24 19:32                                         ` Jens Axboe
2019-09-24 11:43                             ` Peter Zijlstra
2019-09-24 12:57                               ` Jens Axboe
2019-09-24 11:33                           ` Peter Zijlstra
2019-09-24  9:20                   ` Pavel Begunkov
2019-09-24 10:09                     ` Jens Axboe
2019-09-24  9:21                 ` Pavel Begunkov
2019-09-24 10:09                   ` Jens Axboe

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 \
    --in-reply-to=3e359937-5b19-8a4c-4243-ba2edff68504@gmail.com \
    --to=asml.silence@gmail.com \
    --cc=axboe@kernel.dk \
    --cc=linux-block@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=peterz@infradead.org \
    /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;
as well as URLs for NNTP newsgroup(s).