io-uring.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Pavel Begunkov <asml.silence@gmail.com>
To: Jens Axboe <axboe@kernel.dk>, io-uring@vger.kernel.org
Subject: [PATCH v2 00/24] rework and optimise submission+completion paths
Date: Fri, 24 Sep 2021 21:59:40 +0100	[thread overview]
Message-ID: <cover.1632516769.git.asml.silence@gmail.com> (raw)

24 MIOPS vs 31.5, or ~30% win for fio/t/io_uring nops batching=32
Jens mentioned that with his standard test against Optane it gave
yet another +3% to throughput.

1-14 are about optimising the completion path:
- replaces lists with single linked lists
- kills 64 * 8B of caches in ctx
- adds some shuffling of iopoll bits
- list splice instead of per-req list_add in one place
- inlines io_req_free_batch() and other helpers

15-22: inlines __io_queue_sqe() so all the submission path
up to io_issue_sqe() is inlined + little tweaks

v2: rebase for-5.16/io_uring

    multicqe_drain was hanging because it's a bit buggy, i.e. doesn't
    consider that requests may get punted, but still add 24th patch to
    avoid it.

Pavel Begunkov (24):
  io_uring: mark having different creds unlikely
  io_uring: force_nonspin
  io_uring: make io_do_iopoll return number of reqs
  io_uring: use slist for completion batching
  io_uring: remove allocation cache array
  io-wq: add io_wq_work_node based stack
  io_uring: replace list with stack for req caches
  io_uring: split iopoll loop
  io_uring: use single linked list for iopoll
  io_uring: add a helper for batch free
  io_uring: convert iopoll_completed to store_release
  io_uring: optimise batch completion
  io_uring: inline completion batching helpers
  io_uring: don't pass tail into io_free_batch_list
  io_uring: don't pass state to io_submit_state_end
  io_uring: deduplicate io_queue_sqe() call sites
  io_uring: remove drain_active check from hot path
  io_uring: split slow path from io_queue_sqe
  io_uring: inline hot path of __io_queue_sqe()
  io_uring: reshuffle queue_sqe completion handling
  io_uring: restructure submit sqes to_submit checks
  io_uring: kill off ->inflight_entry field
  io_uring: comment why inline complete calls io_clean_op()
  io_uring: disable draining earlier

 fs/io-wq.h    |  60 +++++-
 fs/io_uring.c | 508 +++++++++++++++++++++++---------------------------
 2 files changed, 287 insertions(+), 281 deletions(-)

-- 
2.33.0


             reply	other threads:[~2021-09-24 21:00 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-24 20:59 Pavel Begunkov [this message]
2021-09-24 20:59 ` [PATCH v2 01/24] io_uring: mark having different creds unlikely Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 02/24] io_uring: force_nonspin Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 03/24] io_uring: make io_do_iopoll return number of reqs Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 04/24] io_uring: use slist for completion batching Pavel Begunkov
2021-09-26  6:57   ` Hao Xu
2021-09-28  9:41     ` Pavel Begunkov
2021-09-28 15:32       ` Jens Axboe
2021-09-24 20:59 ` [PATCH v2 05/24] io_uring: remove allocation cache array Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 06/24] io-wq: add io_wq_work_node based stack Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 07/24] io_uring: replace list with stack for req caches Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 08/24] io_uring: split iopoll loop Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 09/24] io_uring: use single linked list for iopoll Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 10/24] io_uring: add a helper for batch free Pavel Begunkov
2021-09-26  3:36   ` Hao Xu
2021-09-28  9:33     ` Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 11/24] io_uring: convert iopoll_completed to store_release Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 12/24] io_uring: optimise batch completion Pavel Begunkov
     [not found]   ` <CAFUsyfLSXMvd_MBAp83qriW7LD=bg2=25TC4e_X4oMO1atoPYg@mail.gmail.com>
2021-09-28  9:35     ` Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 13/24] io_uring: inline completion batching helpers Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 14/24] io_uring: don't pass tail into io_free_batch_list Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 15/24] io_uring: don't pass state to io_submit_state_end Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 16/24] io_uring: deduplicate io_queue_sqe() call sites Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 17/24] io_uring: remove drain_active check from hot path Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 18/24] io_uring: split slow path from io_queue_sqe Pavel Begunkov
2021-09-24 20:59 ` [PATCH v2 19/24] io_uring: inline hot path of __io_queue_sqe() Pavel Begunkov
2021-09-24 21:00 ` [PATCH v2 20/24] io_uring: reshuffle queue_sqe completion handling Pavel Begunkov
2021-09-24 21:00 ` [PATCH v2 21/24] io_uring: restructure submit sqes to_submit checks Pavel Begunkov
2021-09-24 21:00 ` [PATCH v2 22/24] io_uring: kill off ->inflight_entry field Pavel Begunkov
2021-09-24 21:00 ` [PATCH v2 23/24] io_uring: comment why inline complete calls io_clean_op() Pavel Begunkov
2021-09-24 21:00 ` [PATCH v2 24/24] io_uring: disable draining earlier Pavel Begunkov
2021-09-30 16:04 ` [PATCH v2 00/24] rework and optimise submission+completion paths 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=cover.1632516769.git.asml.silence@gmail.com \
    --to=asml.silence@gmail.com \
    --cc=axboe@kernel.dk \
    --cc=io-uring@vger.kernel.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).