IO-Uring Archive on lore.kernel.org
 help / color / Atom feed
From: Jens Axboe <axboe@kernel.dk>
To: Pavel Begunkov <asml.silence@gmail.com>, io-uring@vger.kernel.org
Subject: Re: [PATCH 2/8] io_uring: fix commit_cqring() locking in iopoll
Date: Tue, 30 Jun 2020 08:46:52 -0600
Message-ID: <312bbba1-878c-5681-5e6c-b6de3f7feb55@kernel.dk> (raw)
In-Reply-To: <6c499cf3-418d-2edf-d308-2bb5a8d1d007@gmail.com>

On 6/30/20 8:36 AM, Pavel Begunkov wrote:
> On 30/06/2020 17:04, Jens Axboe wrote:
>> On 6/30/20 6:20 AM, Pavel Begunkov wrote:
>>> Don't call io_commit_cqring() without holding the completion spinlock
>>> in io_iopoll_complete(), it can race, e.g. with async request failing.
>>
>> Can you be more specific?
> 
> io_iopoll_complete()
> 	-> io_req_free_batch()
> 		-> io_queue_next()
> 			-> io_req_task_queue()
> 				-> task_work_add()
> 
> if this task_work_add() fails, it will be redirected to io-wq manager task
> to do io_req_task_cancel() -> commit_cqring().
> 
> 
> And probably something similar will happen if a request currently in io-wq
> is retried with
> io_rw_should_retry() -> io_async_buf_func() -> task_work_add()

For the IOPOLL setup, it should be protected by the ring mutex. Adding
the irq disable + lock for polling would be a nasty performance hit.

-- 
Jens Axboe


  reply index

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-06-30 12:20 [PATCH 0/8] iopoll and task_work fixes Pavel Begunkov
2020-06-30 12:20 ` [PATCH 1/8] io_uring: fix io_fail_links() locking Pavel Begunkov
2020-06-30 14:38   ` Jens Axboe
2020-06-30 14:45     ` Pavel Begunkov
2020-06-30 12:20 ` [PATCH 2/8] io_uring: fix commit_cqring() locking in iopoll Pavel Begunkov
2020-06-30 14:04   ` Jens Axboe
2020-06-30 14:36     ` Pavel Begunkov
2020-06-30 14:46       ` Jens Axboe [this message]
2020-06-30 15:00         ` Pavel Begunkov
2020-06-30 15:33           ` Jens Axboe
2020-06-30 12:20 ` [PATCH 3/8] io_uring: fix ignoring eventfd " Pavel Begunkov
2020-06-30 12:20 ` [PATCH 4/8] io_uring: fix missing ->mm on exit Pavel Begunkov
2020-06-30 12:20 ` [PATCH 5/8] io_uring: don't fail iopoll requeue without ->mm Pavel Begunkov
2020-06-30 12:20 ` [PATCH 6/8] io_uring: fix NULL mm in io_poll_task_func() Pavel Begunkov
2020-06-30 12:20 ` [PATCH 7/8] io_uring: simplify io_async_task_func() Pavel Begunkov
2020-06-30 12:20 ` [PATCH 8/8] io_uring: optimise io_req_find_next() fast check Pavel Begunkov
2020-06-30 12:39 ` [PATCH 0/8] iopoll and task_work fixes Pavel Begunkov

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=312bbba1-878c-5681-5e6c-b6de3f7feb55@kernel.dk \
    --to=axboe@kernel.dk \
    --cc=asml.silence@gmail.com \
    --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

IO-Uring Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/io-uring/0 io-uring/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 io-uring io-uring/ https://lore.kernel.org/io-uring \
		io-uring@vger.kernel.org
	public-inbox-index io-uring

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.io-uring


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git