From: Pavel Begunkov <[email protected]>
To: [email protected]
Cc: Jens Axboe <[email protected]>
Subject: Re: [RFC v2 00/13] CQ waiting and wake up optimisations
Date: Wed, 4 Jan 2023 20:25:20 +0000 [thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>
On 1/3/23 03:03, Pavel Begunkov wrote:
> The series replaces waitqueues for CQ waiting with a custom waiting
> loop and adds a couple more perf tweak around it. Benchmarking is done
> for QD1 with simulated tw arrival right after we start waiting, it
> gets us from 7.5 MIOPS to 9.2, which is +22%, or double the number for
> the in-kernel io_uring overhead (i.e. without syscall and userspace).
> That matches profiles, wake_up() _without_ wake_up_state() was taking
> 12-14% and prepare_to_wait_exclusive() was around 4-6%.
The numbers are gathered with an in-kernel trick. Tried to quickly
measure without it:
modprobe null_blk no_sched=1 irqmode=2 completion_nsec=0
taskset -c 0 fio/t/io_uring -d1 -s1 -c1 -p0 -B1 -F1 -X -b512 -n4 /dev/nullb0
The important part here is using timers-backed nullblk and pinning
multiple workers to a single CPU. -n4 was enough for me to keep
the CPU 100% busy.
old:
IOPS=539.51K, BW=2.11GiB/s, IOS/call=1/1
IOPS=542.26K, BW=2.12GiB/s, IOS/call=1/1
IOPS=540.73K, BW=2.11GiB/s, IOS/call=1/1
IOPS=541.28K, BW=2.11GiB/s, IOS/call=0/0
new:
IOPS=561.85K, BW=2.19GiB/s, IOS/call=1/1
IOPS=561.58K, BW=2.19GiB/s, IOS/call=1/1
IOPS=561.56K, BW=2.19GiB/s, IOS/call=1/1
IOPS=559.94K, BW=2.19GiB/s, IOS/call=1/1
The different is only ~3.5% because of huge additional overhead
for nullb timers, block qos and other unnecessary bits.
P.S. tested with an out-of-tree patch adding a flag enabling/disabling
the feature to remove variance b/w reboots.
--
Pavel Begunkov
prev parent reply other threads:[~2023-01-04 20:26 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-01-03 3:03 [RFC v2 00/13] CQ waiting and wake up optimisations Pavel Begunkov
2023-01-03 3:03 ` [RFC v2 01/13] io_uring: rearrange defer list checks Pavel Begunkov
2023-01-03 3:03 ` [RFC v2 02/13] io_uring: don't iterate cq wait fast path Pavel Begunkov
2023-01-03 3:03 ` [RFC v2 03/13] io_uring: kill io_run_task_work_ctx Pavel Begunkov
2023-01-03 3:03 ` [RFC v2 04/13] io_uring: move defer tw task checks Pavel Begunkov
2023-01-03 3:03 ` [RFC v2 05/13] io_uring: parse check_cq out of wq waiting Pavel Begunkov
2023-01-03 3:03 ` [RFC v2 06/13] io_uring: mimimise io_cqring_wait_schedule Pavel Begunkov
2023-01-03 3:03 ` [RFC v2 07/13] io_uring: simplify io_has_work Pavel Begunkov
2023-01-03 3:03 ` [RFC v2 08/13] io_uring: set TASK_RUNNING right after schedule Pavel Begunkov
2023-01-03 3:04 ` [RFC v2 09/13] io_uring: separate wq for ring polling Pavel Begunkov
2023-01-04 18:08 ` Jens Axboe
2023-01-04 20:28 ` Pavel Begunkov
2023-01-04 20:34 ` Jens Axboe
2023-01-04 20:45 ` Pavel Begunkov
2023-01-04 20:53 ` Jens Axboe
2023-01-04 20:52 ` Jens Axboe
2023-01-03 3:04 ` [RFC v2 10/13] io_uring: add lazy poll_wq activation Pavel Begunkov
2023-01-03 3:04 ` [RFC v2 11/13] io_uring: wake up optimisations Pavel Begunkov
2023-01-03 3:04 ` [RFC v2 12/13] io_uring: waitqueue-less cq waiting Pavel Begunkov
2023-01-03 3:04 ` [RFC v2 13/13] io_uring: add io_req_local_work_add wake fast path Pavel Begunkov
2023-01-04 18:05 ` (subset) [RFC v2 00/13] CQ waiting and wake up optimisations Jens Axboe
2023-01-04 20:25 ` Pavel Begunkov [this message]
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