linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Michal Suchanek <msuchanek@suse.de>
To: linux-scsi@vger.kernel.org
Cc: Michal Suchanek <msuchanek@suse.de>,
	Jonathan Corbet <corbet@lwn.net>, Jens Axboe <axboe@kernel.dk>,
	"James E.J. Bottomley" <jejb@linux.ibm.com>,
	"Martin K. Petersen" <martin.petersen@oracle.com>,
	Alexander Viro <viro@zeniv.linux.org.uk>,
	Mauro Carvalho Chehab <mchehab+samsung@kernel.org>,
	Eric Biggers <ebiggers@google.com>,
	"J. Bruce Fields" <bfields@redhat.com>,
	Benjamin Coddington <bcodding@redhat.com>,
	Hannes Reinecke <hare@suse.com>, Omar Sandoval <osandov@fb.com>,
	Ming Lei <ming.lei@redhat.com>,
	Damien Le Moal <damien.lemoal@wdc.com>,
	Bart Van Assche <bvanassche@acm.org>, Tejun Heo <tj@kernel.org>,
	linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org,
	linux-fsdevel@vger.kernel.org
Subject: [PATCH v2 6/8] bdev: add open_finish.
Date: Wed, 23 Oct 2019 14:52:45 +0200	[thread overview]
Message-ID: <ea2652294651cbc8549736728c650d16d2fe1808.1571834862.git.msuchanek@suse.de> (raw)
In-Reply-To: <cover.1571834862.git.msuchanek@suse.de>

Opening a block device may require a long operation such as waiting for
the cdrom tray to close. Performing this operation with locks held locks
out other attempts to open the device. These processes waiting to open
the device are not killable.

To avoid this issue and still be able to perform time-consuming checks
at open() time the block device driver can provide open_finish(). If it
does opening the device proceeds even when an error is returned from
open(), bd_mutex is released and open_finish() is called. If
open_finish() succeeds the device is now open, if it fails release() is
called.

When -ERESTARTSYS is returned from open() blkdev_get may loop without
calling open_finish(). On -ERESTARTSYS open_finish() is not called.

Move a ret = 0 assignment up in the if/else branching to avoid returning
-ENXIO. Previously the return value was ignored on the unhandled branch.

Signed-off-by: Michal Suchanek <msuchanek@suse.de>
---
 Documentation/filesystems/locking.rst |  2 ++
 fs/block_dev.c                        | 21 +++++++++++++++++----
 include/linux/blkdev.h                |  1 +
 3 files changed, 20 insertions(+), 4 deletions(-)

diff --git a/Documentation/filesystems/locking.rst b/Documentation/filesystems/locking.rst
index fc3a0704553c..2471ced5a8cf 100644
--- a/Documentation/filesystems/locking.rst
+++ b/Documentation/filesystems/locking.rst
@@ -456,6 +456,7 @@ block_device_operations
 prototypes::
 
 	int (*open) (struct block_device *, fmode_t);
+	int (*open_finish) (struct block_device *, fmode_t, int);
 	int (*release) (struct gendisk *, fmode_t);
 	int (*ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
 	int (*compat_ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
@@ -473,6 +474,7 @@ locking rules:
 ops			bd_mutex
 ======================= ===================
 open:			yes
+open_finish:		no
 release:		yes
 ioctl:			no
 compat_ioctl:		no
diff --git a/fs/block_dev.c b/fs/block_dev.c
index 9c073dbdc1b0..009b5dedb1f7 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -1526,6 +1526,7 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part)
 	int partno;
 	int perm = 0;
 	bool first_open = false;
+	bool need_finish = false;
 
 	if (mode & FMODE_READ)
 		perm |= MAY_READ;
@@ -1581,6 +1582,8 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part)
 					put_disk_and_module(disk);
 					goto restart;
 				}
+				if (bdev->bd_disk->fops->open_finish)
+					need_finish = true;
 			}
 
 			if (!ret) {
@@ -1601,7 +1604,7 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part)
 					invalidate_partitions(disk, bdev);
 			}
 
-			if (ret)
+			if (ret && !need_finish)
 				goto out_clear;
 		} else {
 			struct block_device *whole;
@@ -1627,10 +1630,14 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part)
 		if (bdev->bd_bdi == &noop_backing_dev_info)
 			bdev->bd_bdi = bdi_get(disk->queue->backing_dev_info);
 	} else {
+		ret = 0;
 		if (bdev->bd_contains == bdev) {
-			ret = 0;
-			if (bdev->bd_disk->fops->open)
+			if (bdev->bd_disk->fops->open) {
 				ret = bdev->bd_disk->fops->open(bdev, mode);
+				if ((ret != -ERESTARTSYS) &&
+				    bdev->bd_disk->fops->open_finish)
+					need_finish = true;
+			}
 			/* the same as first opener case, read comment there */
 			if (bdev->bd_invalidated) {
 				if (!ret)
@@ -1638,7 +1645,7 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part)
 				else if (ret == -ENOMEDIUM)
 					invalidate_partitions(bdev->bd_disk, bdev);
 			}
-			if (ret)
+			if (ret && !need_finish)
 				goto out_unlock_bdev;
 		}
 	}
@@ -1650,6 +1657,12 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part)
 	/* only one opener holds refs to the module and disk */
 	if (!first_open)
 		put_disk_and_module(disk);
+	if (ret && need_finish)
+		ret = bdev->bd_disk->fops->open_finish(bdev, mode, ret);
+	if (ret) {
+		__blkdev_put(bdev, mode, for_part);
+		return ret;
+	}
 	return 0;
 
  out_clear:
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index f3ea78b0c91c..b67e93c6afb7 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -1695,6 +1695,7 @@ static inline struct bio_vec *rq_integrity_vec(struct request *rq)
 
 struct block_device_operations {
 	int (*open) (struct block_device *, fmode_t);
+	int (*open_finish)(struct block_device *bdev, fmode_t mode, int ret);
 	void (*release) (struct gendisk *, fmode_t);
 	int (*rw_page)(struct block_device *, sector_t, struct page *, unsigned int);
 	int (*ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
-- 
2.23.0


  parent reply	other threads:[~2019-10-23 12:53 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-23 12:52 [PATCH v2 0/8] Fix cdrom autoclose Michal Suchanek
2019-10-23 12:52 ` [PATCH v2 1/8] cdrom: add poll_event_interruptible Michal Suchanek
2019-10-23 12:52 ` [PATCH v2 2/8] cdrom: factor out common open_for_* code Michal Suchanek
2019-10-24  2:19   ` Christoph Hellwig
2019-10-24  8:50     ` Michal Suchánek
2019-10-25  2:39       ` Christoph Hellwig
2019-10-25 10:42         ` Michal Suchánek
2019-10-26  6:46           ` Finn Thain
2019-10-24 13:23     ` Matthew Wilcox
2019-10-25  2:38       ` Christoph Hellwig
2019-10-23 12:52 ` [PATCH v2 3/8] cdrom: wait for the tray to close Michal Suchanek
2019-10-23 12:52 ` [PATCH v2 4/8] cdrom: separate autoclose into an IOCTL Michal Suchanek
2019-10-23 12:52 ` [PATCH v2 5/8] docs: cdrom: Add autoclose IOCTL Michal Suchanek
2019-10-23 12:52 ` Michal Suchanek [this message]
2019-10-24  2:22   ` [PATCH v2 6/8] bdev: add open_finish Christoph Hellwig
2019-10-24  8:55     ` Michal Suchánek
2019-10-24 13:12       ` Matthew Wilcox
2019-10-24 13:19         ` Michal Suchánek
2019-11-21 10:15         ` Michal Suchánek
2019-10-23 12:52 ` [PATCH v2 7/8] scsi: sr: workaround VMware ESXi cdrom emulation bug Michal Suchanek
2019-10-23 14:13   ` Hannes Reinecke
2019-10-23 16:23     ` Michal Suchánek
2019-10-23 21:44       ` Ewan D. Milne
2019-10-24  5:46       ` Hannes Reinecke
2019-10-24  8:56         ` Michal Suchánek
2019-10-24  9:41           ` Hannes Reinecke
2019-10-24 10:11             ` Michal Suchánek
2019-10-24 11:45             ` [PATCH RFC] scsi: blacklist: add VMware ESXi cdrom - broken tray emulation Michal Suchanek
2019-10-24  2:23   ` [PATCH v2 7/8] scsi: sr: workaround VMware ESXi cdrom emulation bug Christoph Hellwig
2019-10-24  8:53     ` Michal Suchánek
2019-11-21 15:21     ` Michal Suchánek
2019-10-23 12:52 ` [PATCH v2 8/8] scsi: sr: wait for the medium to become ready Michal Suchanek
2019-10-24  2:24   ` Christoph Hellwig
2019-10-24  8:51     ` Michal Suchánek
2019-10-24 13:14       ` Matthew Wilcox
2019-10-26 14:57   ` [scsi] 9ed2563662: BUG:kernel_NULL_pointer_dereference,address kernel test robot

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=ea2652294651cbc8549736728c650d16d2fe1808.1571834862.git.msuchanek@suse.de \
    --to=msuchanek@suse.de \
    --cc=axboe@kernel.dk \
    --cc=bcodding@redhat.com \
    --cc=bfields@redhat.com \
    --cc=bvanassche@acm.org \
    --cc=corbet@lwn.net \
    --cc=damien.lemoal@wdc.com \
    --cc=ebiggers@google.com \
    --cc=hare@suse.com \
    --cc=jejb@linux.ibm.com \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-scsi@vger.kernel.org \
    --cc=martin.petersen@oracle.com \
    --cc=mchehab+samsung@kernel.org \
    --cc=ming.lei@redhat.com \
    --cc=osandov@fb.com \
    --cc=tj@kernel.org \
    --cc=viro@zeniv.linux.org.uk \
    /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).