Merge tag 'vfs-6.9.super' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull block handle updates from Christian Brauner:
 "Last cycle we changed opening of block devices, and opening a block
  device would return a bdev_handle. This allowed us to implement
  support for restricting and forbidding writes to mounted block
  devices. It was accompanied by converting and adding helpers to
  operate on bdev_handles instead of plain block devices.

  That was already a good step forward but ultimately it isn't necessary
  to have special purpose helpers for opening block devices internally
  that return a bdev_handle.

  Fundamentally, opening a block device internally should just be
  equivalent to opening files. So now all internal opens of block
  devices return files just as a userspace open would. Instead of
  introducing a separate indirection into bdev_open_by_*() via struct
  bdev_handle bdev_file_open_by_*() is made to just return a struct
  file. Opening and closing a block device just becomes equivalent to
  opening and closing a file.

  This all works well because internally we already have a pseudo fs for
  block devices and so opening block devices is simple. There's a few
  places where we needed to be careful such as during boot when the
  kernel is supposed to mount the rootfs directly without init doing it.
  Here we need to take care to ensure that we flush out any asynchronous
  file close. That's what we already do for opening, unpacking, and
  closing the initramfs. So nothing new here.

  The equivalence of opening and closing block devices to regular files
  is a win in and of itself. But it also has various other advantages.
  We can remove struct bdev_handle completely. Various low-level helpers
  are now private to the block layer. Other helpers were simply
  removable completely.

  A follow-up series that is already reviewed build on this and makes it
  possible to remove bdev->bd_inode and allows various clean ups of the
  buffer head code as well. All places where we stashed a bdev_handle
  now just stash a file and use simple accessors to get to the actual
  block device which was already the case for bdev_handle"

* tag 'vfs-6.9.super' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs: (35 commits)
  block: remove bdev_handle completely
  block: don't rely on BLK_OPEN_RESTRICT_WRITES when yielding write access
  bdev: remove bdev pointer from struct bdev_handle
  bdev: make struct bdev_handle private to the block layer
  bdev: make bdev_{release, open_by_dev}() private to block layer
  bdev: remove bdev_open_by_path()
  reiserfs: port block device access to file
  ocfs2: port block device access to file
  nfs: port block device access to files
  jfs: port block device access to file
  f2fs: port block device access to files
  ext4: port block device access to file
  erofs: port device access to file
  btrfs: port device access to file
  bcachefs: port block device access to file
  target: port block device access to file
  s390: port block device access to file
  nvme: port block device access to file
  block2mtd: port device access to files
  bcache: port block device access to files
  ...

5 files changed, 189 insertions, 134 deletions

diff --git a/block/bdev.c b/block/bdev.c
index e9f1b12bd75c..140093c99bdc 100644
--- a/block/bdev.c
+++ b/block/bdev.c
@@ -49,6 +49,12 @@ struct block_device *I_BDEV(struct inode *inode)
 }
 EXPORT_SYMBOL(I_BDEV);
 
+struct block_device *file_bdev(struct file *bdev_file)
+{
+	return I_BDEV(bdev_file->f_mapping->host);
+}
+EXPORT_SYMBOL(file_bdev);
+
 static void bdev_write_inode(struct block_device *bdev)
 {
 	struct inode *inode = bdev->bd_inode;
@@ -368,12 +374,12 @@ static struct file_system_type bd_type = {
 };
 
 struct super_block *blockdev_superblock __ro_after_init;
+struct vfsmount *blockdev_mnt __ro_after_init;
 EXPORT_SYMBOL_GPL(blockdev_superblock);
 
 void __init bdev_cache_init(void)
 {
 	int err;
-	static struct vfsmount *bd_mnt __ro_after_init;
 
 	bdev_cachep = kmem_cache_create("bdev_cache", sizeof(struct bdev_inode),
 			0, (SLAB_HWCACHE_ALIGN|SLAB_RECLAIM_ACCOUNT|
@@ -382,10 +388,10 @@ void __init bdev_cache_init(void)
 	err = register_filesystem(&bd_type);
 	if (err)
 		panic("Cannot register bdev pseudo-fs");
-	bd_mnt = kern_mount(&bd_type);
-	if (IS_ERR(bd_mnt))
+	blockdev_mnt = kern_mount(&bd_type);
+	if (IS_ERR(blockdev_mnt))
 		panic("Cannot create bdev pseudo-fs");
-	blockdev_superblock = bd_mnt->mnt_sb;   /* For writeback */
+	blockdev_superblock = blockdev_mnt->mnt_sb;   /* For writeback */
 }
 
 struct block_device *bdev_alloc(struct gendisk *disk, u8 partno)
@@ -696,6 +702,31 @@ out_blkdev_put:
 	return ret;
 }
 
+int bdev_permission(dev_t dev, blk_mode_t mode, void *holder)
+{
+	int ret;
+
+	ret = devcgroup_check_permission(DEVCG_DEV_BLOCK,
+			MAJOR(dev), MINOR(dev),
+			((mode & BLK_OPEN_READ) ? DEVCG_ACC_READ : 0) |
+			((mode & BLK_OPEN_WRITE) ? DEVCG_ACC_WRITE : 0));
+	if (ret)
+		return ret;
+
+	/* Blocking writes requires exclusive opener */
+	if (mode & BLK_OPEN_RESTRICT_WRITES && !holder)
+		return -EINVAL;
+
+	/*
+	 * We're using error pointers to indicate to ->release() when we
+	 * failed to open that block device. Also this doesn't make sense.
+	 */
+	if (WARN_ON_ONCE(IS_ERR(holder)))
+		return -EINVAL;
+
+	return 0;
+}
+
 static void blkdev_put_part(struct block_device *part)
 {
 	struct block_device *whole = bdev_whole(part);
@@ -775,83 +806,55 @@ static void bdev_claim_write_access(struct block_device *bdev, blk_mode_t mode)
 		bdev->bd_writers++;
 }
 
-static void bdev_yield_write_access(struct block_device *bdev, blk_mode_t mode)
+static void bdev_yield_write_access(struct file *bdev_file)
 {
+	struct block_device *bdev;
+
 	if (bdev_allow_write_mounted)
 		return;
 
+	bdev = file_bdev(bdev_file);
 	/* 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--;
+	if (bdev_file->f_mode & FMODE_WRITE) {
+		if (bdev_writes_blocked(bdev))
+			bdev_unblock_writes(bdev);
+		else
+			bdev->bd_writers--;
+	}
 }
 
 /**
- * bdev_open_by_dev - open a block device by device number
- * @dev: device number of block device to open
+ * bdev_open - open a block device
+ * @bdev: block device to open
  * @mode: open mode (BLK_OPEN_*)
  * @holder: exclusive holder identifier
  * @hops: holder operations
+ * @bdev_file: file for the block device
  *
- * Open the block device described by device number @dev. If @holder is not
- * %NULL, the block device is opened with exclusive access.  Exclusive opens may
- * nest for the same @holder.
- *
- * Use this interface ONLY if you really do not have anything better - i.e. when
- * you are behind a truly sucky interface and all you are given is a device
- * number.  Everything else should use bdev_open_by_path().
+ * Open the block device. If @holder is not %NULL, the block device is opened
+ * with exclusive access.  Exclusive opens may nest for the same @holder.
  *
  * CONTEXT:
  * Might sleep.
  *
  * RETURNS:
- * Handle with a reference to the block_device on success, ERR_PTR(-errno) on
- * failure.
+ * zero on success, -errno on failure.
  */
-struct bdev_handle *bdev_open_by_dev(dev_t dev, blk_mode_t mode, void *holder,
-				     const struct blk_holder_ops *hops)
+int bdev_open(struct block_device *bdev, blk_mode_t mode, void *holder,
+	      const struct blk_holder_ops *hops, struct file *bdev_file)
 {
-	struct bdev_handle *handle = kmalloc(sizeof(struct bdev_handle),
-					     GFP_KERNEL);
-	struct block_device *bdev;
 	bool unblock_events = true;
-	struct gendisk *disk;
+	struct gendisk *disk = bdev->bd_disk;
 	int ret;
 
-	if (!handle)
-		return ERR_PTR(-ENOMEM);
-
-	ret = devcgroup_check_permission(DEVCG_DEV_BLOCK,
-			MAJOR(dev), MINOR(dev),
-			((mode & BLK_OPEN_READ) ? DEVCG_ACC_READ : 0) |
-			((mode & BLK_OPEN_WRITE) ? DEVCG_ACC_WRITE : 0));
-	if (ret)
-		goto free_handle;
-
-	/* Blocking writes requires exclusive opener */
-	if (mode & BLK_OPEN_RESTRICT_WRITES && !holder) {
-		ret = -EINVAL;
-		goto free_handle;
-	}
-
-	bdev = blkdev_get_no_open(dev);
-	if (!bdev) {
-		ret = -ENXIO;
-		goto free_handle;
-	}
-	disk = bdev->bd_disk;
-
 	if (holder) {
 		mode |= BLK_OPEN_EXCL;
 		ret = bd_prepare_to_claim(bdev, holder, hops);
 		if (ret)
-			goto put_blkdev;
+			return ret;
 	} else {
-		if (WARN_ON_ONCE(mode & BLK_OPEN_EXCL)) {
-			ret = -EIO;
-			goto put_blkdev;
-		}
+		if (WARN_ON_ONCE(mode & BLK_OPEN_EXCL))
+			return -EIO;
 	}
 
 	disk_block_events(disk);
@@ -892,10 +895,16 @@ struct bdev_handle *bdev_open_by_dev(dev_t dev, blk_mode_t mode, void *holder,
 
 	if (unblock_events)
 		disk_unblock_events(disk);
-	handle->bdev = bdev;
-	handle->holder = holder;
-	handle->mode = mode;
-	return handle;
+
+	bdev_file->f_flags |= O_LARGEFILE;
+	bdev_file->f_mode |= FMODE_BUF_RASYNC | FMODE_CAN_ODIRECT;
+	if (bdev_nowait(bdev))
+		bdev_file->f_mode |= FMODE_NOWAIT;
+	bdev_file->f_mapping = bdev->bd_inode->i_mapping;
+	bdev_file->f_wb_err = filemap_sample_wb_err(bdev_file->f_mapping);
+	bdev_file->private_data = holder;
+
+	return 0;
 put_module:
 	module_put(disk->fops->owner);
 abort_claiming:
@@ -903,36 +912,80 @@ abort_claiming:
 		bd_abort_claiming(bdev, holder);
 	mutex_unlock(&disk->open_mutex);
 	disk_unblock_events(disk);
-put_blkdev:
-	blkdev_put_no_open(bdev);
-free_handle:
-	kfree(handle);
-	return ERR_PTR(ret);
+	return ret;
 }
-EXPORT_SYMBOL(bdev_open_by_dev);
 
-/**
- * bdev_open_by_path - open a block device by name
- * @path: path to the block device to open
- * @mode: open mode (BLK_OPEN_*)
- * @holder: exclusive holder identifier
- * @hops: holder operations
- *
- * Open the block device described by the device file at @path.  If @holder is
- * not %NULL, the block device is opened with exclusive access.  Exclusive opens
- * may nest for the same @holder.
- *
- * CONTEXT:
- * Might sleep.
+/*
+ * If BLK_OPEN_WRITE_IOCTL is set then this is a historical quirk
+ * associated with the floppy driver where it has allowed ioctls if the
+ * file was opened for writing, but does not allow reads or writes.
+ * Make sure that this quirk is reflected in @f_flags.
  *
- * RETURNS:
- * Handle with a reference to the block_device on success, ERR_PTR(-errno) on
- * failure.
+ * It can also happen if a block device is opened as O_RDWR | O_WRONLY.
  */
-struct bdev_handle *bdev_open_by_path(const char *path, blk_mode_t mode,
-		void *holder, const struct blk_holder_ops *hops)
+static unsigned blk_to_file_flags(blk_mode_t mode)
+{
+	unsigned int flags = 0;
+
+	if ((mode & (BLK_OPEN_READ | BLK_OPEN_WRITE)) ==
+	    (BLK_OPEN_READ | BLK_OPEN_WRITE))
+		flags |= O_RDWR;
+	else if (mode & BLK_OPEN_WRITE_IOCTL)
+		flags |= O_RDWR | O_WRONLY;
+	else if (mode & BLK_OPEN_WRITE)
+		flags |= O_WRONLY;
+	else if (mode & BLK_OPEN_READ)
+		flags |= O_RDONLY; /* homeopathic, because O_RDONLY is 0 */
+	else
+		WARN_ON_ONCE(true);
+
+	if (mode & BLK_OPEN_NDELAY)
+		flags |= O_NDELAY;
+
+	return flags;
+}
+
+struct file *bdev_file_open_by_dev(dev_t dev, blk_mode_t mode, void *holder,
+				   const struct blk_holder_ops *hops)
 {
-	struct bdev_handle *handle;
+	struct file *bdev_file;
+	struct block_device *bdev;
+	unsigned int flags;
+	int ret;
+
+	ret = bdev_permission(dev, mode, holder);
+	if (ret)
+		return ERR_PTR(ret);
+
+	bdev = blkdev_get_no_open(dev);
+	if (!bdev)
+		return ERR_PTR(-ENXIO);
+
+	flags = blk_to_file_flags(mode);
+	bdev_file = alloc_file_pseudo_noaccount(bdev->bd_inode,
+			blockdev_mnt, "", flags | O_LARGEFILE, &def_blk_fops);
+	if (IS_ERR(bdev_file)) {
+		blkdev_put_no_open(bdev);
+		return bdev_file;
+	}
+	ihold(bdev->bd_inode);
+
+	ret = bdev_open(bdev, mode, holder, hops, bdev_file);
+	if (ret) {
+		/* We failed to open the block device. Let ->release() know. */
+		bdev_file->private_data = ERR_PTR(ret);
+		fput(bdev_file);
+		return ERR_PTR(ret);
+	}
+	return bdev_file;
+}
+EXPORT_SYMBOL(bdev_file_open_by_dev);
+
+struct file *bdev_file_open_by_path(const char *path, blk_mode_t mode,
+				    void *holder,
+				    const struct blk_holder_ops *hops)
+{
+	struct file *file;
 	dev_t dev;
 	int error;
 
@@ -940,22 +993,28 @@ struct bdev_handle *bdev_open_by_path(const char *path, blk_mode_t mode,
 	if (error)
 		return ERR_PTR(error);
 
-	handle = bdev_open_by_dev(dev, mode, holder, hops);
-	if (!IS_ERR(handle) && (mode & BLK_OPEN_WRITE) &&
-	    bdev_read_only(handle->bdev)) {
-		bdev_release(handle);
-		return ERR_PTR(-EACCES);
+	file = bdev_file_open_by_dev(dev, mode, holder, hops);
+	if (!IS_ERR(file) && (mode & BLK_OPEN_WRITE)) {
+		if (bdev_read_only(file_bdev(file))) {
+			fput(file);
+			file = ERR_PTR(-EACCES);
+		}
 	}
 
-	return handle;
+	return file;
 }
-EXPORT_SYMBOL(bdev_open_by_path);
+EXPORT_SYMBOL(bdev_file_open_by_path);
 
-void bdev_release(struct bdev_handle *handle)
+void bdev_release(struct file *bdev_file)
 {
-	struct block_device *bdev = handle->bdev;
+	struct block_device *bdev = file_bdev(bdev_file);
+	void *holder = bdev_file->private_data;
 	struct gendisk *disk = bdev->bd_disk;
 
+	/* We failed to open that block device. */
+	if (IS_ERR(holder))
+		goto put_no_open;
+
 	/*
 	 * Sync early if it looks like we're the last one.  If someone else
 	 * opens the block device between now and the decrement of bd_openers
@@ -967,10 +1026,10 @@ void bdev_release(struct bdev_handle *handle)
 		sync_blockdev(bdev);
 
 	mutex_lock(&disk->open_mutex);
-	bdev_yield_write_access(bdev, handle->mode);
+	bdev_yield_write_access(bdev_file);
 
-	if (handle->holder)
-		bd_end_claim(bdev, handle->holder);
+	if (holder)
+		bd_end_claim(bdev, holder);
 
 	/*
 	 * Trigger event checking and tell drivers to flush MEDIA_CHANGE
@@ -986,10 +1045,9 @@ void bdev_release(struct bdev_handle *handle)
 	mutex_unlock(&disk->open_mutex);
 
 	module_put(disk->fops->owner);
+put_no_open:
 	blkdev_put_no_open(bdev);
-	kfree(handle);
 }
-EXPORT_SYMBOL(bdev_release);
 
 /**
  * lookup_bdev() - Look up a struct block_device by name.
diff --git a/block/blk.h b/block/blk.h
index 1ef920f72e0f..f02b25f22e8b 100644
--- a/block/blk.h
+++ b/block/blk.h
@@ -516,4 +516,8 @@ static inline int req_ref_read(struct request *req)
 	return atomic_read(&req->ref);
 }
 
+void bdev_release(struct file *bdev_file);
+int bdev_open(struct block_device *bdev, blk_mode_t mode, void *holder,
+	      const struct blk_holder_ops *hops, struct file *bdev_file);
+int bdev_permission(dev_t dev, blk_mode_t mode, void *holder);
 #endif /* BLK_INTERNAL_H */
diff --git a/block/fops.c b/block/fops.c
index 63566a3db186..679d9b752fe8 100644
--- a/block/fops.c
+++ b/block/fops.c
@@ -572,18 +572,17 @@ static int blkdev_fsync(struct file *filp, loff_t start, loff_t end,
 blk_mode_t file_to_blk_mode(struct file *file)
 {
 	blk_mode_t mode = 0;
-	struct bdev_handle *handle = file->private_data;
 
 	if (file->f_mode & FMODE_READ)
 		mode |= BLK_OPEN_READ;
 	if (file->f_mode & FMODE_WRITE)
 		mode |= BLK_OPEN_WRITE;
 	/*
-	 * do_dentry_open() clears O_EXCL from f_flags, use handle->mode to
-	 * determine whether the open was exclusive for already open files.
+	 * do_dentry_open() clears O_EXCL from f_flags, use file->private_data
+	 * to determine whether the open was exclusive for already open files.
 	 */
-	if (handle)
-		mode |= handle->mode & BLK_OPEN_EXCL;
+	if (file->private_data)
+		mode |= BLK_OPEN_EXCL;
 	else if (file->f_flags & O_EXCL)
 		mode |= BLK_OPEN_EXCL;
 	if (file->f_flags & O_NDELAY)
@@ -602,36 +601,31 @@ blk_mode_t file_to_blk_mode(struct file *file)
 
 static int blkdev_open(struct inode *inode, struct file *filp)
 {
-	struct bdev_handle *handle;
+	struct block_device *bdev;
 	blk_mode_t mode;
-
-	/*
-	 * Preserve backwards compatibility and allow large file access
-	 * even if userspace doesn't ask for it explicitly. Some mkfs
-	 * binary needs it. We might want to drop this workaround
-	 * during an unstable branch.
-	 */
-	filp->f_flags |= O_LARGEFILE;
-	filp->f_mode |= FMODE_BUF_RASYNC | FMODE_CAN_ODIRECT;
+	int ret;
 
 	mode = file_to_blk_mode(filp);
-	handle = bdev_open_by_dev(inode->i_rdev, mode,
-			mode & BLK_OPEN_EXCL ? filp : NULL, NULL);
-	if (IS_ERR(handle))
-		return PTR_ERR(handle);
+	/* Use the file as the holder. */
+	if (mode & BLK_OPEN_EXCL)
+		filp->private_data = filp;
+	ret = bdev_permission(inode->i_rdev, mode, filp->private_data);
+	if (ret)
+		return ret;
 
-	if (bdev_nowait(handle->bdev))
-		filp->f_mode |= FMODE_NOWAIT;
+	bdev = blkdev_get_no_open(inode->i_rdev);
+	if (!bdev)
+		return -ENXIO;
 
-	filp->f_mapping = handle->bdev->bd_inode->i_mapping;
-	filp->f_wb_err = filemap_sample_wb_err(filp->f_mapping);
-	filp->private_data = handle;
-	return 0;
+	ret = bdev_open(bdev, mode, filp->private_data, NULL, filp);
+	if (ret)
+		blkdev_put_no_open(bdev);
+	return ret;
 }
 
 static int blkdev_release(struct inode *inode, struct file *filp)
 {
-	bdev_release(filp->private_data);
+	bdev_release(filp);
 	return 0;
 }
 
diff --git a/block/genhd.c b/block/genhd.c
index d74fb5b4ae68..a911d2969c07 100644
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -342,7 +342,7 @@ EXPORT_SYMBOL_GPL(disk_uevent);
 
 int disk_scan_partitions(struct gendisk *disk, blk_mode_t mode)
 {
-	struct bdev_handle *handle;
+	struct file *file;
 	int ret = 0;
 
 	if (disk->flags & (GENHD_FL_NO_PART | GENHD_FL_HIDDEN))
@@ -366,12 +366,12 @@ int disk_scan_partitions(struct gendisk *disk, blk_mode_t mode)
 	}
 
 	set_bit(GD_NEED_PART_SCAN, &disk->state);
-	handle = bdev_open_by_dev(disk_devt(disk), mode & ~BLK_OPEN_EXCL, NULL,
-				  NULL);
-	if (IS_ERR(handle))
-		ret = PTR_ERR(handle);
+	file = bdev_file_open_by_dev(disk_devt(disk), mode & ~BLK_OPEN_EXCL,
+				     NULL, NULL);
+	if (IS_ERR(file))
+		ret = PTR_ERR(file);
 	else
-		bdev_release(handle);
+		fput(file);
 
 	/*
 	 * If blkdev_get_by_dev() failed early, GD_NEED_PART_SCAN is still set,
diff --git a/block/ioctl.c b/block/ioctl.c
index 438f79c564cf..4c8aebee595f 100644
--- a/block/ioctl.c
+++ b/block/ioctl.c
@@ -469,7 +469,7 @@ static int blkdev_bszset(struct block_device *bdev, blk_mode_t mode,
 		int __user *argp)
 {
 	int ret, n;
-	struct bdev_handle *handle;
+	struct file *file;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EACCES;
@@ -481,12 +481,11 @@ static int blkdev_bszset(struct block_device *bdev, blk_mode_t mode,
 	if (mode & BLK_OPEN_EXCL)
 		return set_blocksize(bdev, n);
 
-	handle = bdev_open_by_dev(bdev->bd_dev, mode, &bdev, NULL);
-	if (IS_ERR(handle))
+	file = bdev_file_open_by_dev(bdev->bd_dev, mode, &bdev, NULL);
+	if (IS_ERR(file))
 		return -EBUSY;
 	ret = set_blocksize(bdev, n);
-	bdev_release(handle);
-
+	fput(file);
 	return ret;
 }