public inbox for [email protected]
 help / color / mirror / Atom feed
From: Jens Axboe <[email protected]>
To: [email protected]
Cc: [email protected], Jens Axboe <[email protected]>,
	[email protected], Josef <[email protected]>
Subject: [PATCH 2/2] io_uring: use TWA_SIGNAL for task_work if the task isn't running
Date: Sat,  8 Aug 2020 12:34:39 -0600	[thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>

An earlier commit:

b7db41c9e03b ("io_uring: fix regression with always ignoring signals in io_cqring_wait()")

ensured that we didn't get stuck waiting for eventfd reads when it's
registered with the io_uring ring for event notification, but we still
have a gap where the task can be waiting on other events in the kernel
and need a bigger nudge to make forward progress.

Ensure that we use signaled notifications for a task that isn't currently
running, to be certain the work is seen and processed immediately.

Cc: [email protected] # v5.7+
Reported-by: Josef <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
---
 fs/io_uring.c | 22 ++++++++++++++--------
 1 file changed, 14 insertions(+), 8 deletions(-)

diff --git a/fs/io_uring.c b/fs/io_uring.c
index e9b27cdaa735..443eecdfeda9 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -1712,21 +1712,27 @@ static int io_req_task_work_add(struct io_kiocb *req, struct callback_head *cb)
 	struct io_ring_ctx *ctx = req->ctx;
 	int ret, notify = TWA_RESUME;
 
+	ret = __task_work_add(tsk, cb);
+	if (unlikely(ret))
+		return ret;
+
 	/*
 	 * SQPOLL kernel thread doesn't need notification, just a wakeup.
-	 * If we're not using an eventfd, then TWA_RESUME is always fine,
-	 * as we won't have dependencies between request completions for
-	 * other kernel wait conditions.
+	 * For any other work, use signaled wakeups if the task isn't
+	 * running to avoid dependencies between tasks or threads. If
+	 * the issuing task is currently waiting in the kernel on a thread,
+	 * and same thread is waiting for a completion event, then we need
+	 * to ensure that the issuing task processes task_work. TWA_SIGNAL
+	 * is needed for that.
 	 */
 	if (ctx->flags & IORING_SETUP_SQPOLL)
 		notify = 0;
-	else if (ctx->cq_ev_fd)
+	else if (READ_ONCE(tsk->state) != TASK_RUNNING)
 		notify = TWA_SIGNAL;
 
-	ret = task_work_add(tsk, cb, notify);
-	if (!ret)
-		wake_up_process(tsk);
-	return ret;
+	__task_work_notify(tsk, notify);
+	wake_up_process(tsk);
+	return 0;
 }
 
 static void __io_req_task_cancel(struct io_kiocb *req, int error)
-- 
2.28.0


  parent reply	other threads:[~2020-08-08 18:34 UTC|newest]

Thread overview: 41+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-08-08 18:34 [PATCHSET 0/2] io_uring: use TWA_SIGNAL more carefully Jens Axboe
2020-08-08 18:34 ` [PATCH 1/2] kernel: split task_work_add() into two separate helpers Jens Axboe
2020-08-10 11:37   ` peterz
2020-08-10 15:01     ` Jens Axboe
2020-08-10 15:28       ` peterz
2020-08-10 17:51       ` Jens Axboe
2020-08-10 19:53         ` Peter Zijlstra
2020-08-08 18:34 ` Jens Axboe [this message]
2020-08-10 11:42   ` [PATCH 2/2] io_uring: use TWA_SIGNAL for task_work if the task isn't running peterz
2020-08-10 15:02     ` Jens Axboe
2020-08-10 19:21       ` Jens Axboe
2020-08-10 20:12         ` Peter Zijlstra
2020-08-10 20:13           ` Jens Axboe
2020-08-10 20:25             ` Jens Axboe
2020-08-10 20:32               ` Peter Zijlstra
2020-08-10 20:35                 ` Jens Axboe
2020-08-10 20:35               ` Jann Horn
2020-08-10 21:06                 ` Jens Axboe
2020-08-10 21:10                   ` Peter Zijlstra
2020-08-10 21:12                     ` Jens Axboe
2020-08-10 21:26                       ` Jann Horn
2020-08-10 21:28                         ` Jens Axboe
2020-08-10 22:01                           ` Jens Axboe
2020-08-10 22:41                             ` Jann Horn
2020-08-11  1:25                               ` Jens Axboe
2020-08-11  6:45                               ` Oleg Nesterov
2020-08-11  6:56                                 ` Peter Zijlstra
2020-08-11  7:14                                   ` Oleg Nesterov
2020-08-11  7:26                                     ` Oleg Nesterov
2020-08-11  7:49                                       ` Peter Zijlstra
2020-08-11  7:45                                     ` Peter Zijlstra
2020-08-11  8:10                                       ` Oleg Nesterov
2020-08-11 13:06                                         ` Jens Axboe
2020-08-11 14:05                                           ` Oleg Nesterov
2020-08-11 14:12                                             ` Jens Axboe
2020-08-10 21:27                       ` Jens Axboe
2020-08-10 20:16           ` Jens Axboe
2020-08-13 16:25   ` Sasha Levin
2020-08-19 23:57   ` Sasha Levin
2020-08-19 23:59     ` Jens Axboe
2020-08-20  0:02       ` 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 \
    [email protected] \
    [email protected] \
    [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