From: Mike Snitzer <snitzer@redhat.com>
To: Bart Van Assche <bart.vanassche@gmail.com>
Cc: Bart Van Assche <bart.vanassche@sandisk.com>,
axboe@kernel.dk, linux-block@vger.kernel.org,
device-mapper development <dm-devel@redhat.com>,
hch@lst.de
Subject: Re: should blk-mq halt requeue processing while queue is frozen?
Date: Tue, 13 Sep 2016 10:36:47 -0400 [thread overview]
Message-ID: <20160913143647.GA23094@redhat.com> (raw)
In-Reply-To: <7c034c5b-db1b-e0c7-2824-81672c6683f9@acm.org>
On Tue, Sep 13 2016 at 4:01am -0400,
Bart Van Assche <bart.vanassche@gmail.com> wrote:
> On 09/07/2016 06:41 PM, Mike Snitzer wrote:
> >On Fri, Sep 02 2016 at 6:42pm -0400,
> >Bart Van Assche <bart.vanassche@sandisk.com> wrote:
> >>+/**
> >>+ * blk_mq_quiesce_queue - wait until all pending queue_rq calls have finished
> >>+ *
> >>+ * Prevent that new I/O requests are queued and wait until all pending
> >>+ * queue_rq() calls have finished.
> >>+ */
> >>+void blk_mq_quiesce_queue(struct request_queue *q)
> >>+{
> >>+ spin_lock_irq(q->queue_lock);
> >>+ WARN_ON_ONCE(blk_queue_quiescing(q));
> >>+ queue_flag_set(QUEUE_FLAG_QUIESCING, q);
> >>+ spin_unlock_irq(q->queue_lock);
> >>+
> >>+ atomic_inc_return(&q->mq_freeze_depth);
> >>+ blk_mq_run_hw_queues(q, false);
> >>+ synchronize_rcu();
> >
> >Why the synchronize_rcu()?
>
> Hello Mike,
>
> Adding read_lock() + read_unlock() in __blk_mq_run_hw_queue() and
> synchronize_rcu() in blk_mq_quiesce_queue() is the lowest overhead
> mechanism I know of to make the latter function wait until the
> former has finished.
OK.
> >Also, you're effectively open-coding blk_mq_freeze_queue_start() minus
> >the q->q_usage_counter mgmt. Why not add a flag to conditionally manage
> >q->q_usage_counter to blk_mq_freeze_queue_start()?
>
> I will consider this.
>
> >But I'm concerned about blk_mq_{quiesce,resume}_queue vs
> >blk_mq_{freeze,unfreeze}_queue -- e.g. if "freeze" is nested after
> >"queue" (but before "resume") it would still need the q->q_usage_counter
> >management. Your patch as-is would break the blk-mq freeze interface.
>
> Agreed. blk_mq_{quiesce,resume}_queue() has to manipulate
> q_usage_counter in the same way as blk_mq_{freeze,unfreeze}_queue().
> Once I am back in the office I will rework this patch and send it to
> Jens.
Please base any further work in this area ontop of
http://git.kernel.org/cgit/linux/kernel/git/snitzer/linux.git/log/?h=devel
And please verify all the mptest tests pass with your changes in place.
next prev parent reply other threads:[~2016-09-13 14:36 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-08-31 22:14 [PATCH 0/9] dm patches for kernel v4.9 Bart Van Assche
2016-08-31 22:15 ` [PATCH 1/9] blk-mq: Introduce blk_mq_queue_stopped() Bart Van Assche
2016-08-31 22:16 ` [PATCH 2/9] dm: Rename a function argument Bart Van Assche
2016-09-01 3:29 ` Mike Snitzer
2016-09-01 14:17 ` Bart Van Assche
2016-08-31 22:16 ` [PATCH 3/9] dm: Introduce signal_pending_state() Bart Van Assche
2016-08-31 22:16 ` [PATCH 4/9] dm: Convert wait loops Bart Van Assche
2016-08-31 22:17 ` [PATCH 5/9] dm: Add two lockdep_assert_held() statements Bart Van Assche
2016-08-31 22:17 ` [PATCH 6/9] dm: Simplify dm_old_stop_queue() Bart Van Assche
2016-08-31 22:17 ` [PATCH 7/9] dm: Mark block layer queue dead before destroying the dm device Bart Van Assche
2016-08-31 22:18 ` [PATCH 8/9] dm: Fix two race conditions related to stopping and starting queues Bart Van Assche
2016-09-01 3:13 ` Mike Snitzer
2016-09-01 14:23 ` Bart Van Assche
2016-09-01 15:05 ` Mike Snitzer
2016-09-01 15:31 ` Bart Van Assche
2016-09-01 15:50 ` Mike Snitzer
2016-09-01 16:12 ` Mike Snitzer
2016-09-01 17:59 ` Bart Van Assche
2016-09-01 19:05 ` Mike Snitzer
2016-09-01 19:35 ` Mike Snitzer
2016-09-01 20:15 ` Bart Van Assche
2016-09-01 20:33 ` Mike Snitzer
2016-09-01 20:39 ` Bart Van Assche
2016-09-01 20:48 ` Mike Snitzer
2016-09-01 20:52 ` Bart Van Assche
2016-09-01 21:17 ` Mike Snitzer
2016-09-01 22:18 ` Mike Snitzer
2016-09-01 22:22 ` Bart Van Assche
2016-09-01 22:26 ` Mike Snitzer
2016-09-01 23:17 ` Bart Van Assche
2016-09-01 23:47 ` Mike Snitzer
2016-09-02 0:03 ` Bart Van Assche
2016-09-02 15:12 ` Mike Snitzer
2016-09-02 16:10 ` should blk-mq halt requeue processing while queue is frozen? [was: Re: [PATCH 8/9] dm: Fix two race conditions related to stopping and starting queues] Mike Snitzer
2016-09-02 22:42 ` [dm-devel] should blk-mq halt requeue processing while queue is frozen? Bart Van Assche
2016-09-03 0:34 ` Mike Snitzer
2016-09-07 16:41 ` Mike Snitzer
2016-09-13 8:01 ` [dm-devel] " Bart Van Assche
2016-09-13 14:36 ` Mike Snitzer [this message]
2016-08-31 22:18 ` [PATCH 9/9] dm path selector: Avoid that device removal triggers an infinite loop Bart Van Assche
2016-09-01 2:49 ` Mike Snitzer
2016-09-01 14:14 ` Bart Van Assche
2016-09-01 15:06 ` Mike Snitzer
2016-09-01 15:22 ` Bart Van Assche
2016-09-01 15:26 ` Mike Snitzer
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=20160913143647.GA23094@redhat.com \
--to=snitzer@redhat.com \
--cc=axboe@kernel.dk \
--cc=bart.vanassche@gmail.com \
--cc=bart.vanassche@sandisk.com \
--cc=dm-devel@redhat.com \
--cc=hch@lst.de \
--cc=linux-block@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).