public inbox for [email protected]
 help / color / mirror / Atom feed
From: Ming Lei <[email protected]>
To: Jens Axboe <[email protected]>
Cc: [email protected], [email protected],
	David Howells <[email protected]>,
	Chengming Zhou <[email protected]>,
	[email protected]
Subject: Re: [PATCH] io_uring: fix IO hang in io_wq_put_and_exit from do_exit()
Date: Fri, 1 Sep 2023 09:56:47 +0800	[thread overview]
Message-ID: <ZPFE3+vOI/Wm7/3e@fedora> (raw)
In-Reply-To: <[email protected]>

On Thu, Aug 31, 2023 at 11:59:42AM -0600, Jens Axboe wrote:
> On 8/31/23 1:42 AM, Ming Lei wrote:
> > @@ -3313,11 +3323,12 @@ __cold void io_uring_cancel_generic(bool cancel_all, struct io_sq_data *sqd)
> >  	atomic_inc(&tctx->in_cancel);
> >  	do {
> >  		bool loop = false;
> > +		bool wq_cancelled;
> 
> Minor nit, but io_uring generally uses US spelling, so this should be
> wq_canceled.

OK.

> 
> >  
> >  		io_uring_drop_tctx_refs(current);
> >  		/* read completions before cancelations */
> >  		inflight = tctx_inflight(tctx, !cancel_all);
> > -		if (!inflight)
> > +		if (!inflight && !tctx->io_wq)
> >  			break;
> 
> Not sure I follow this one, is it just for checking of io_wq was ever
> setup? How could it not be?

Here if we have io_wq, all requests in io_wq have to be canceled no
matter if inflight is zero or not because tctx_inflight(tctx, true)
doesn't track FIXED_FILE IOs, but there still can be such IOs in io_wq.

> 
> > -		if (loop) {
> > +		if (!wq_cancelled || (inflight && loop)) {
> >  			cond_resched();
> >  			continue;
> >  		}
> 
> And this one is a bit puzzling to me too - if we didn't cancel anything
> but don't have anything inflight, why are we looping? Should it be
> something ala:
> 
> if (inflight && (!wq_cancelled || loop)) {

There can be FIXED_FILE IOs in io_wq even though inflight is zero.


Thanks,
Ming


  reply	other threads:[~2023-09-01  1:57 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-31  7:42 [PATCH] io_uring: fix IO hang in io_wq_put_and_exit from do_exit() Ming Lei
2023-08-31 17:59 ` Jens Axboe
2023-09-01  1:56   ` Ming Lei [this message]
2023-09-01  1:50 ` Chengming Zhou
2023-09-01  2:09   ` Ming Lei
2023-09-01  2:17     ` Chengming Zhou
2023-09-01  8:56       ` Ming Lei

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=ZPFE3+vOI/Wm7/3e@fedora \
    [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