block: Add config option to not allow writing to mounted devices

Writing to mounted devices is dangerous and can lead to filesystem
corruption as well as crashes. Furthermore syzbot comes with more and
more involved examples how to corrupt block device under a mounted
filesystem leading to kernel crashes and reports we can do nothing
about. Add tracking of writers to each block device and a kernel cmdline
argument which controls whether other writeable opens to block devices
open with BLK_OPEN_RESTRICT_WRITES flag are allowed. We will make
filesystems use this flag for used devices.

Note that this effectively only prevents modification of the particular
block device's page cache by other writers. The actual device content
can still be modified by other means - e.g. by issuing direct scsi
commands, by doing writes through devices lower in the storage stack
(e.g. in case loop devices, DM, or MD are involved) etc. But blocking
direct modifications of the block device page cache is enough to give
filesystems a chance to perform data validation when loading data from
the underlying storage and thus prevent kernel crashes.

Syzbot can use this cmdline argument option to avoid uninteresting
crashes. Also users whose userspace setup does not need writing to
mounted block devices can set this option for hardening.

Link: https://lore.kernel.org/all/60788e5d-5c7c-1142-e554-c21d709acfd9@linaro.org
Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20231101174325.10596-3-jack@suse.cz
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Christian Brauner <brauner@kernel.org>

4 files changed, 97 insertions, 1 deletions

diff --git a/block/Kconfig b/block/Kconfig
index 55ae2286a4de..1de4682d48cc 100644
--- a/block/Kconfig
+++ b/block/Kconfig
@@ -78,6 +78,26 @@ config BLK_DEV_INTEGRITY_T10
 	select CRC_T10DIF
 	select CRC64_ROCKSOFT
 
+config BLK_DEV_WRITE_MOUNTED
+	bool "Allow writing to mounted block devices"
+	default y
+	help
+	When a block device is mounted, writing to its buffer cache is very
+	likely going to cause filesystem corruption. It is also rather easy to
+	crash the kernel in this way since the filesystem has no practical way
+	of detecting these writes to buffer cache and verifying its metadata
+	integrity. However there are some setups that need this capability
+	like running fsck on read-only mounted root device, modifying some
+	features on mounted ext4 filesystem, and similar. If you say N, the
+	kernel will prevent processes from writing to block devices that are
+	mounted by filesystems which provides some more protection from runaway
+	privileged processes and generally makes it much harder to crash
+	filesystem drivers. Note however that this does not prevent
+	underlying device(s) from being modified by other means, e.g. by
+	directly submitting SCSI commands or through access to lower layers of
+	storage stack. If in doubt, say Y. The configuration can be overridden
+	with the bdev_allow_write_mounted boot option.
+
 config BLK_DEV_ZONED
 	bool "Zoned block device support"
 	select MQ_IOSCHED_DEADLINE
diff --git a/block/bdev.c b/block/bdev.c
index 0daf14e50dd0..fc8d28d77495 100644
--- a/block/bdev.c
+++ b/block/bdev.c
@@ -30,6 +30,9 @@
 #include "../fs/internal.h"
 #include "blk.h"
 
+/* Should we allow writing to mounted block devices? */
+static bool bdev_allow_write_mounted = IS_ENABLED(CONFIG_BLK_DEV_WRITE_MOUNTED);
+
 struct bdev_inode {
 	struct block_device bdev;
 	struct inode vfs_inode;
@@ -730,7 +733,58 @@ void blkdev_put_no_open(struct block_device *bdev)
 {
 	put_device(&bdev->bd_device);
 }
-	
+
+static bool bdev_writes_blocked(struct block_device *bdev)
+{
+	return bdev->bd_writers == -1;
+}
+
+static void bdev_block_writes(struct block_device *bdev)
+{
+	bdev->bd_writers = -1;
+}
+
+static void bdev_unblock_writes(struct block_device *bdev)
+{
+	bdev->bd_writers = 0;
+}
+
+static bool bdev_may_open(struct block_device *bdev, blk_mode_t mode)
+{
+	if (bdev_allow_write_mounted)
+		return true;
+	/* Writes blocked? */
+	if (mode & BLK_OPEN_WRITE && bdev_writes_blocked(bdev))
+		return false;
+	if (mode & BLK_OPEN_RESTRICT_WRITES && bdev->bd_writers > 0)
+		return false;
+	return true;
+}
+
+static void bdev_claim_write_access(struct block_device *bdev, blk_mode_t mode)
+{
+	if (bdev_allow_write_mounted)
+		return;
+
+	/* Claim exclusive or shared write access. */
+	if (mode & BLK_OPEN_RESTRICT_WRITES)
+		bdev_block_writes(bdev);
+	else if (mode & BLK_OPEN_WRITE)
+		bdev->bd_writers++;
+}
+
+static void bdev_yield_write_access(struct block_device *bdev, blk_mode_t mode)
+{
+	if (bdev_allow_write_mounted)
+		return;
+
+	/* Yield exclusive or shared write access. */
+	if (mode & BLK_OPEN_RESTRICT_WRITES)
+		bdev_unblock_writes(bdev);
+	else if (mode & BLK_OPEN_WRITE)
+		bdev->bd_writers--;
+}
+
 /**
  * bdev_open_by_dev - open a block device by device number
  * @dev: device number of block device to open
@@ -773,6 +827,10 @@ struct bdev_handle *bdev_open_by_dev(dev_t dev, blk_mode_t mode, void *holder,
 	if (ret)
 		goto free_handle;
 
+	/* Blocking writes requires exclusive opener */
+	if (mode & BLK_OPEN_RESTRICT_WRITES && !holder)
+		return ERR_PTR(-EINVAL);
+
 	bdev = blkdev_get_no_open(dev);
 	if (!bdev) {
 		ret = -ENXIO;
@@ -800,12 +858,16 @@ struct bdev_handle *bdev_open_by_dev(dev_t dev, blk_mode_t mode, void *holder,
 		goto abort_claiming;
 	if (!try_module_get(disk->fops->owner))
 		goto abort_claiming;
+	ret = -EBUSY;
+	if (!bdev_may_open(bdev, mode))
+		goto abort_claiming;
 	if (bdev_is_partition(bdev))
 		ret = blkdev_get_part(bdev, mode);
 	else
 		ret = blkdev_get_whole(bdev, mode);
 	if (ret)
 		goto put_module;
+	bdev_claim_write_access(bdev, mode);
 	if (holder) {
 		bd_finish_claiming(bdev, holder, hops);
 
@@ -901,6 +963,8 @@ void bdev_release(struct bdev_handle *handle)
 		sync_blockdev(bdev);
 
 	mutex_lock(&disk->open_mutex);
+	bdev_yield_write_access(bdev, handle->mode);
+
 	if (handle->holder)
 		bd_end_claim(bdev, handle->holder);
 
@@ -1069,3 +1133,12 @@ void bdev_statx_dioalign(struct inode *inode, struct kstat *stat)
 
 	blkdev_put_no_open(bdev);
 }
+
+static int __init setup_bdev_allow_write_mounted(char *str)
+{
+	if (kstrtobool(str, &bdev_allow_write_mounted))
+		pr_warn("Invalid option string for bdev_allow_write_mounted:"
+			" '%s'\n", str);
+	return 1;
+}
+__setup("bdev_allow_write_mounted=", setup_bdev_allow_write_mounted);
diff --git a/include/linux/blk_types.h b/include/linux/blk_types.h
index 749203277fee..52e264d5a830 100644
--- a/include/linux/blk_types.h
+++ b/include/linux/blk_types.h
@@ -66,6 +66,7 @@ struct block_device {
 #ifdef CONFIG_FAIL_MAKE_REQUEST
 	bool			bd_make_it_fail;
 #endif
+	int			bd_writers;
 	/*
 	 * keep this out-of-line as it's both big and not needed in the fast
 	 * path
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 7afc10315dd5..0e0c0186aa32 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -124,6 +124,8 @@ typedef unsigned int __bitwise blk_mode_t;
 #define BLK_OPEN_NDELAY		((__force blk_mode_t)(1 << 3))
 /* open for "writes" only for ioctls (specialy hack for floppy.c) */
 #define BLK_OPEN_WRITE_IOCTL	((__force blk_mode_t)(1 << 4))
+/* open is exclusive wrt all other BLK_OPEN_WRITE opens to the device */
+#define BLK_OPEN_RESTRICT_WRITES	((__force blk_mode_t)(1 << 5))
 
 struct gendisk {
 	/*