From: Ming Lei <ming.lei@redhat.com>
To: Bart Van Assche <bvanassche@acm.org>
Cc: Jens Axboe <axboe@kernel.dk>, Keith Busch <keith.busch@intel.com>,
Sagi Grimberg <sagi@grimberg.me>,
linux-nvme@lists.infradead.org, linux-block@vger.kernel.org,
Max Gurtovoy <maxg@mellanox.com>, Christoph Hellwig <hch@lst.de>
Subject: Re: [PATCH 2/5] blk-mq: introduce blk_mq_tagset_wait_completed_request()
Date: Wed, 24 Jul 2019 09:34:34 +0800 [thread overview]
Message-ID: <20190724013432.GB22421@ming.t460p> (raw)
In-Reply-To: <d4d3ded9-0012-68c1-7511-f5ac3aa7b1fb@acm.org>
On Tue, Jul 23, 2019 at 01:54:52PM -0700, Bart Van Assche wrote:
> On 7/22/19 6:06 PM, Ming Lei wrote:
> > On Mon, Jul 22, 2019 at 08:25:07AM -0700, Bart Van Assche wrote:
> > > On 7/21/19 10:39 PM, Ming Lei wrote:
> > > > blk-mq may schedule to call queue's complete function on remote CPU via
> > > > IPI, but doesn't provide any way to synchronize the request's complete
> > > > fn.
> > > >
> > > > In some driver's EH(such as NVMe), hardware queue's resource may be freed &
> > > > re-allocated. If the completed request's complete fn is run finally after the
> > > > hardware queue's resource is released, kernel crash will be triggered.
> > > >
> > > > Prepare for fixing this kind of issue by introducing
> > > > blk_mq_tagset_wait_completed_request().
> > >
> > > An explanation is missing of why the block layer is modified to fix this
> > > instead of the NVMe driver.
> >
> > The above commit log has explained that there isn't sync mechanism in
> > blk-mq wrt. request completion, and there might be similar issue in other
> > future drivers.
>
> That is not sufficient as a motivation to modify the block layer because
> there is already a way to wait until request completions have finished,
> namely the request queue freeze mechanism. Have you considered to use that
> mechanism instead of introducing blk_mq_tagset_wait_completed_request()?
The introduced interface is used in EH, during which the aborted
requests will stay at blk-mq sw/scheduler queue, so queue freeze will
cause deadlock. We simply can't use it.
Thanks,
Ming
WARNING: multiple messages have this Message-ID (diff)
From: ming.lei@redhat.com (Ming Lei)
Subject: [PATCH 2/5] blk-mq: introduce blk_mq_tagset_wait_completed_request()
Date: Wed, 24 Jul 2019 09:34:34 +0800 [thread overview]
Message-ID: <20190724013432.GB22421@ming.t460p> (raw)
In-Reply-To: <d4d3ded9-0012-68c1-7511-f5ac3aa7b1fb@acm.org>
On Tue, Jul 23, 2019@01:54:52PM -0700, Bart Van Assche wrote:
> On 7/22/19 6:06 PM, Ming Lei wrote:
> > On Mon, Jul 22, 2019@08:25:07AM -0700, Bart Van Assche wrote:
> > > On 7/21/19 10:39 PM, Ming Lei wrote:
> > > > blk-mq may schedule to call queue's complete function on remote CPU via
> > > > IPI, but doesn't provide any way to synchronize the request's complete
> > > > fn.
> > > >
> > > > In some driver's EH(such as NVMe), hardware queue's resource may be freed &
> > > > re-allocated. If the completed request's complete fn is run finally after the
> > > > hardware queue's resource is released, kernel crash will be triggered.
> > > >
> > > > Prepare for fixing this kind of issue by introducing
> > > > blk_mq_tagset_wait_completed_request().
> > >
> > > An explanation is missing of why the block layer is modified to fix this
> > > instead of the NVMe driver.
> >
> > The above commit log has explained that there isn't sync mechanism in
> > blk-mq wrt. request completion, and there might be similar issue in other
> > future drivers.
>
> That is not sufficient as a motivation to modify the block layer because
> there is already a way to wait until request completions have finished,
> namely the request queue freeze mechanism. Have you considered to use that
> mechanism instead of introducing blk_mq_tagset_wait_completed_request()?
The introduced interface is used in EH, during which the aborted
requests will stay at blk-mq sw/scheduler queue, so queue freeze will
cause deadlock. We simply can't use it.
Thanks,
Ming
next prev parent reply other threads:[~2019-07-24 1:34 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-07-22 5:39 [PATCH 0/5] blk-mq: wait until completed req's complete fn is run Ming Lei
2019-07-22 5:39 ` Ming Lei
2019-07-22 5:39 ` [PATCH 1/5] blk-mq: introduce blk_mq_request_completed() Ming Lei
2019-07-22 5:39 ` Ming Lei
2019-07-23 20:26 ` Sagi Grimberg
2019-07-23 20:26 ` Sagi Grimberg
2019-07-22 5:39 ` [PATCH 2/5] blk-mq: introduce blk_mq_tagset_wait_completed_request() Ming Lei
2019-07-22 5:39 ` Ming Lei
2019-07-22 15:25 ` Bart Van Assche
2019-07-22 15:25 ` Bart Van Assche
2019-07-23 1:06 ` Ming Lei
2019-07-23 1:06 ` Ming Lei
2019-07-23 20:54 ` Bart Van Assche
2019-07-23 20:54 ` Bart Van Assche
2019-07-24 1:34 ` Ming Lei [this message]
2019-07-24 1:34 ` Ming Lei
2019-07-23 20:27 ` Sagi Grimberg
2019-07-23 20:27 ` Sagi Grimberg
2019-07-22 5:39 ` [PATCH 3/5] nvme: don't abort completed request in nvme_cancel_request Ming Lei
2019-07-22 5:39 ` Ming Lei
2019-07-22 15:27 ` Bart Van Assche
2019-07-22 15:27 ` Bart Van Assche
2019-07-22 23:22 ` Keith Busch
2019-07-22 23:22 ` Keith Busch
2019-07-23 0:07 ` Sagi Grimberg
2019-07-23 0:07 ` Sagi Grimberg
2019-07-23 1:08 ` Ming Lei
2019-07-23 1:08 ` Ming Lei
2019-07-23 19:22 ` Bart Van Assche
2019-07-23 19:22 ` Bart Van Assche
2019-07-23 20:27 ` Sagi Grimberg
2019-07-23 20:27 ` Sagi Grimberg
2019-07-22 5:39 ` [PATCH 4/5] nvme: wait until all completed request's complete fn is called Ming Lei
2019-07-22 5:39 ` Ming Lei
2019-07-23 16:14 ` Dongli Zhang
2019-07-23 16:14 ` Dongli Zhang
2019-07-24 2:05 ` Ming Lei
2019-07-24 2:05 ` Ming Lei
2019-07-23 20:29 ` Sagi Grimberg
2019-07-23 20:29 ` Sagi Grimberg
2019-07-24 1:43 ` Ming Lei
2019-07-24 1:43 ` Ming Lei
2019-07-22 5:39 ` [PATCH 5/5] blk-mq: remove blk_mq_complete_request_sync Ming Lei
2019-07-22 5:39 ` Ming Lei
2019-07-23 20:30 ` Sagi Grimberg
2019-07-23 20:30 ` Sagi Grimberg
2019-07-22 23:27 ` [PATCH 0/5] blk-mq: wait until completed req's complete fn is run Bob Liu
2019-07-22 23:27 ` Bob Liu
2019-07-23 1:10 ` Ming Lei
2019-07-23 1:10 ` 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=20190724013432.GB22421@ming.t460p \
--to=ming.lei@redhat.com \
--cc=axboe@kernel.dk \
--cc=bvanassche@acm.org \
--cc=hch@lst.de \
--cc=keith.busch@intel.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-nvme@lists.infradead.org \
--cc=maxg@mellanox.com \
--cc=sagi@grimberg.me \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.