All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ming Lei <ming.lei@redhat.com>
To: Christoph Hellwig <hch@lst.de>
Cc: "Jens Axboe" <axboe@kernel.dk>,
	"Stefan Haberland" <sth@linux.ibm.com>,
	"Jan Hoeppner" <hoeppner@linux.ibm.com>,
	"Martin K. Petersen" <martin.petersen@oracle.com>,
	"Doug Gilbert" <dgilbert@interlog.com>,
	"Kai Mäkisara" <Kai.Makisara@kolumbus.fi>,
	"Luis Chamberlain" <mcgrof@kernel.org>,
	linux-block@vger.kernel.org, linux-nvme@lists.infradead.org,
	linux-s390@vger.kernel.org, linux-scsi@vger.kernel.org,
	blankburian@uni-muenster.de
Subject: Re: [PATCH 8/9] block: hold a request_queue reference for the lifetime of struct gendisk
Date: Fri, 20 Aug 2021 21:42:58 +0800	[thread overview]
Message-ID: <YR+xYi3VX8MFilud@T590> (raw)
In-Reply-To: <20210816131910.615153-9-hch@lst.de>

On Mon, Aug 16, 2021 at 03:19:09PM +0200, Christoph Hellwig wrote:
> Acquire the queue ref dropped in disk_release in __blk_alloc_disk so any
> allocate gendisk always has a queue reference.

BTW, today Markus reported that request queue is released when the
disk is still live.

And looks it is triggered when running virtio-scsi hotplug from qemu
side, and the reason could be that we grab the request queue refcount
after disk is added to driver core, so there is small race window in
which the request queue is released before we grab it in __device_add_disk().

I guess this patch could fix the issue, but it is hard to verify
since it takes days to reproduce.


Thanks,
Ming


WARNING: multiple messages have this Message-ID (diff)
From: Ming Lei <ming.lei@redhat.com>
To: Christoph Hellwig <hch@lst.de>
Cc: "Jens Axboe" <axboe@kernel.dk>,
	"Stefan Haberland" <sth@linux.ibm.com>,
	"Jan Hoeppner" <hoeppner@linux.ibm.com>,
	"Martin K. Petersen" <martin.petersen@oracle.com>,
	"Doug Gilbert" <dgilbert@interlog.com>,
	"Kai Mäkisara" <Kai.Makisara@kolumbus.fi>,
	"Luis Chamberlain" <mcgrof@kernel.org>,
	linux-block@vger.kernel.org, linux-nvme@lists.infradead.org,
	linux-s390@vger.kernel.org, linux-scsi@vger.kernel.org,
	blankburian@uni-muenster.de
Subject: Re: [PATCH 8/9] block: hold a request_queue reference for the lifetime of struct gendisk
Date: Fri, 20 Aug 2021 21:42:58 +0800	[thread overview]
Message-ID: <YR+xYi3VX8MFilud@T590> (raw)
In-Reply-To: <20210816131910.615153-9-hch@lst.de>

On Mon, Aug 16, 2021 at 03:19:09PM +0200, Christoph Hellwig wrote:
> Acquire the queue ref dropped in disk_release in __blk_alloc_disk so any
> allocate gendisk always has a queue reference.

BTW, today Markus reported that request queue is released when the
disk is still live.

And looks it is triggered when running virtio-scsi hotplug from qemu
side, and the reason could be that we grab the request queue refcount
after disk is added to driver core, so there is small race window in
which the request queue is released before we grab it in __device_add_disk().

I guess this patch could fix the issue, but it is hard to verify
since it takes days to reproduce.


Thanks,
Ming


_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

  reply	other threads:[~2021-08-20 13:44 UTC|newest]

Thread overview: 42+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-08-16 13:19 ensure each gendisk always has a request_queue reference v2 Christoph Hellwig
2021-08-16 13:19 ` Christoph Hellwig
2021-08-16 13:19 ` [PATCH 1/9] nvme: use blk_mq_alloc_disk Christoph Hellwig
2021-08-16 13:19   ` Christoph Hellwig
2021-08-19 14:54   ` Keith Busch
2021-08-19 14:54     ` Keith Busch
2021-08-19 14:58     ` Keith Busch
2021-08-19 14:58       ` Keith Busch
2021-08-19 22:57   ` Luis Chamberlain
2021-08-19 22:57     ` Luis Chamberlain
2021-08-20  4:06     ` Christoph Hellwig
2021-08-20  4:06       ` Christoph Hellwig
2021-08-23 17:22   ` Sagi Grimberg
2021-08-23 17:22     ` Sagi Grimberg
2021-08-16 13:19 ` [PATCH 2/9] st: do not allocate a gendisk Christoph Hellwig
2021-08-16 13:19   ` Christoph Hellwig
2021-08-16 13:19 ` [PATCH 3/9] sg: " Christoph Hellwig
2021-08-16 13:19   ` Christoph Hellwig
2021-08-19 21:33   ` Luis Chamberlain
2021-08-19 21:33     ` Luis Chamberlain
2021-08-19 21:36     ` Luis Chamberlain
2021-08-19 21:36       ` Luis Chamberlain
2021-08-20  4:05     ` Christoph Hellwig
2021-08-20  4:05       ` Christoph Hellwig
2021-08-16 13:19 ` [PATCH 4/9] block: cleanup the lockdep handling in *alloc_disk Christoph Hellwig
2021-08-16 13:19   ` Christoph Hellwig
2021-08-16 13:19 ` [PATCH 5/9] block: remove alloc_disk and alloc_disk_node Christoph Hellwig
2021-08-16 13:19   ` Christoph Hellwig
2021-08-16 13:19 ` [PATCH 6/9] block: remove the minors argument to __alloc_disk_node Christoph Hellwig
2021-08-16 13:19   ` Christoph Hellwig
2021-08-16 13:19 ` [PATCH 7/9] block: pass a request_queue to __blk_alloc_disk Christoph Hellwig
2021-08-16 13:19   ` Christoph Hellwig
2021-08-16 13:19 ` [PATCH 8/9] block: hold a request_queue reference for the lifetime of struct gendisk Christoph Hellwig
2021-08-16 13:19   ` Christoph Hellwig
2021-08-20 13:42   ` Ming Lei [this message]
2021-08-20 13:42     ` Ming Lei
2021-08-16 13:46 ` [PATCH 9/9] block: add an explicit ->disk backpointer to the request_queue Christoph Hellwig
2021-08-16 13:46   ` Christoph Hellwig
2021-08-19 14:39   ` Sven Schnelle
2021-08-19 14:39     ` Sven Schnelle
2021-08-23 18:55 ` ensure each gendisk always has a request_queue reference v2 Jens Axboe
2021-08-23 18:55   ` 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 \
    --in-reply-to=YR+xYi3VX8MFilud@T590 \
    --to=ming.lei@redhat.com \
    --cc=Kai.Makisara@kolumbus.fi \
    --cc=axboe@kernel.dk \
    --cc=blankburian@uni-muenster.de \
    --cc=dgilbert@interlog.com \
    --cc=hch@lst.de \
    --cc=hoeppner@linux.ibm.com \
    --cc=linux-block@vger.kernel.org \
    --cc=linux-nvme@lists.infradead.org \
    --cc=linux-s390@vger.kernel.org \
    --cc=linux-scsi@vger.kernel.org \
    --cc=martin.petersen@oracle.com \
    --cc=mcgrof@kernel.org \
    --cc=sth@linux.ibm.com \
    /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.