Merge tag 'pull-set_blocksize' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs

Pull vfs blocksize updates from Al Viro:
 "This gets rid of bogus set_blocksize() uses, switches it over
  to be based on a 'struct file *' and verifies that the caller
  has the device opened exclusively"

* tag 'pull-set_blocksize' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  make set_blocksize() fail unless block device is opened exclusive
  set_blocksize(): switch to passing struct file *
  btrfs_get_bdev_and_sb(): call set_blocksize() only for exclusive opens
  swsusp: don't bother with setting block size
  zram: don't bother with reopening - just use O_EXCL for open
  swapon(2): open swap with O_EXCL
  swapon(2)/swapoff(2): don't bother with block size
  pktcdvd: sort set_blocksize() calls out
  bcache_register(): don't bother with set_blocksize()

Documentation/filesystems/porting.rst

@@ -1134,3 +1134,10 @@ superblock of the main block device, i.e., the one stored in sb->s_bdev. Block
 device freezing now works for any block device owned by a given superblock, not
 just the main block device. The get_active_super() helper and bd_fsfreeze_sb
 pointer are gone.
+
+---
+
+**mandatory**
+
+set_blocksize() takes opened struct file instead of struct block_device now
+and it *must* be opened exclusive.

block/bdev.c

@@ -144,8 +144,11 @@ static void set_init_blocksize(struct block_device *bdev)
 	bdev->bd_inode->i_blkbits = blksize_bits(bsize);
 }
 
-int set_blocksize(struct block_device *bdev, int size)
+int set_blocksize(struct file *file, int size)
 {
+	struct inode *inode = file->f_mapping->host;
+	struct block_device *bdev = I_BDEV(inode);
+
 	/* Size must be a power of two, and between 512 and PAGE_SIZE */
 	if (size > PAGE_SIZE || size < 512 || !is_power_of_2(size))
 		return -EINVAL;
@@ -154,10 +157,13 @@ int set_blocksize(struct block_device *bdev, int size)
 	if (size < bdev_logical_block_size(bdev))
 		return -EINVAL;
 
+	if (!file->private_data)
+		return -EINVAL;
+
 	/* Don't change the size if it is same as current */
-	if (bdev->bd_inode->i_blkbits != blksize_bits(size)) {
+	if (inode->i_blkbits != blksize_bits(size)) {
 		sync_blockdev(bdev);
-		bdev->bd_inode->i_blkbits = blksize_bits(size);
+		inode->i_blkbits = blksize_bits(size);
 		kill_bdev(bdev);
 	}
 	return 0;
@@ -167,7 +173,7 @@ EXPORT_SYMBOL(set_blocksize);
 
 int sb_set_blocksize(struct super_block *sb, int size)
 {
-	if (set_blocksize(sb->s_bdev, size))
+	if (set_blocksize(sb->s_bdev_file, size))
 		return 0;
 	/* If we get here, we know size is power of two
 	 * and it's value is between 512 and PAGE_SIZE */

block/ioctl.c

@@ -503,11 +503,14 @@ static int compat_hdio_getgeo(struct block_device *bdev,
 #endif
 
 /* set the logical block size */
-static int blkdev_bszset(struct block_device *bdev, blk_mode_t mode,
+static int blkdev_bszset(struct file *file, blk_mode_t mode,
 		int __user *argp)
 {
+	// this one might be file_inode(file)->i_rdev - a rare valid
+	// use of file_inode() for those.
+	dev_t dev = I_BDEV(file->f_mapping->host)->bd_dev;
+	struct file *excl_file;
 	int ret, n;
-	struct file *file;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EACCES;
@@ -517,13 +520,13 @@ static int blkdev_bszset(struct block_device *bdev, blk_mode_t mode,
 		return -EFAULT;
 
 	if (mode & BLK_OPEN_EXCL)
-		return set_blocksize(bdev, n);
+		return set_blocksize(file, n);
 
-	file = bdev_file_open_by_dev(bdev->bd_dev, mode, &bdev, NULL);
-	if (IS_ERR(file))
+	excl_file = bdev_file_open_by_dev(dev, mode, &dev, NULL);
+	if (IS_ERR(excl_file))
 		return -EBUSY;
-	ret = set_blocksize(bdev, n);
-	fput(file);
+	ret = set_blocksize(excl_file, n);
+	fput(excl_file);
 	return ret;
 }
 
@@ -652,7 +655,7 @@ long blkdev_ioctl(struct file *file, unsigned cmd, unsigned long arg)
 	case BLKBSZGET: /* get block device soft block size (cf. BLKSSZGET) */
 		return put_int(argp, block_size(bdev));
 	case BLKBSZSET:
-		return blkdev_bszset(bdev, mode, argp);
+		return blkdev_bszset(file, mode, argp);
 	case BLKGETSIZE64:
 		return put_u64(argp, bdev_nr_bytes(bdev));
 
@@ -712,7 +715,7 @@ long compat_blkdev_ioctl(struct file *file, unsigned cmd, unsigned long arg)
 	case BLKBSZGET_32: /* get the logical block size (cf. BLKSSZGET) */
 		return put_int(argp, bdev_logical_block_size(bdev));
 	case BLKBSZSET_32:
-		return blkdev_bszset(bdev, mode, argp);
+		return blkdev_bszset(file, mode, argp);
 	case BLKGETSIZE64_32:
 		return put_u64(argp, bdev_nr_bytes(bdev));
 

drivers/block/pktcdvd.c

@@ -2215,6 +2215,7 @@ static int pkt_open_dev(struct pktcdvd_device *pd, bool write)
 		}
 		dev_info(ddev, "%lukB available on disc\n", lba << 1);
 	}
+	set_blocksize(bdev_file, CD_FRAMESIZE);
 
 	return 0;
 
@@ -2278,11 +2279,6 @@ static int pkt_open(struct gendisk *disk, blk_mode_t mode)
 		ret = pkt_open_dev(pd, mode & BLK_OPEN_WRITE);
 		if (ret)
 			goto out_dec;
-		/*
-		 * needed here as well, since ext2 (among others) may change
-		 * the blocksize at mount time
-		 */
-		set_blocksize(disk->part0, CD_FRAMESIZE);
 	}
 	mutex_unlock(&ctl_mutex);
 	mutex_unlock(&pktcdvd_mutex);
@@ -2526,7 +2522,6 @@ static int pkt_new_dev(struct pktcdvd_device *pd, dev_t dev)
 	__module_get(THIS_MODULE);
 
 	pd->bdev_file = bdev_file;
-	set_blocksize(file_bdev(bdev_file), CD_FRAMESIZE);
 
 	atomic_set(&pd->cdrw.pending_bios, 0);
 	pd->cdrw.thread = kthread_run(kcdrwd, pd, "%s", pd->disk->disk_name);

drivers/block/zram/zram_drv.c

@@ -426,11 +426,10 @@ static void reset_bdev(struct zram *zram)
 	if (!zram->backing_dev)
 		return;
 
-	fput(zram->bdev_file);
 	/* hope filp_close flush all of IO */
 	filp_close(zram->backing_dev, NULL);
 	zram->backing_dev = NULL;
-	zram->bdev_file = NULL;
+	zram->bdev = NULL;
 	zram->disk->fops = &zram_devops;
 	kvfree(zram->bitmap);
 	zram->bitmap = NULL;
@@ -473,10 +472,8 @@ static ssize_t backing_dev_store(struct device *dev,
 	size_t sz;
 	struct file *backing_dev = NULL;
 	struct inode *inode;
-	struct address_space *mapping;
 	unsigned int bitmap_sz;
 	unsigned long nr_pages, *bitmap = NULL;
-	struct file *bdev_file = NULL;
 	int err;
 	struct zram *zram = dev_to_zram(dev);
 
@@ -497,15 +494,14 @@ static ssize_t backing_dev_store(struct device *dev,
 	if (sz > 0 && file_name[sz - 1] == '\n')
 		file_name[sz - 1] = 0x00;
 
-	backing_dev = filp_open(file_name, O_RDWR|O_LARGEFILE, 0);
+	backing_dev = filp_open(file_name, O_RDWR | O_LARGEFILE | O_EXCL, 0);
 	if (IS_ERR(backing_dev)) {
 		err = PTR_ERR(backing_dev);
 		backing_dev = NULL;
 		goto out;
 	}
 
-	mapping = backing_dev->f_mapping;
-	inode = mapping->host;
+	inode = backing_dev->f_mapping->host;
 
 	/* Support only block device in this moment */
 	if (!S_ISBLK(inode->i_mode)) {
@@ -513,14 +509,6 @@ static ssize_t backing_dev_store(struct device *dev,
 		goto out;
 	}
 
-	bdev_file = bdev_file_open_by_dev(inode->i_rdev,
-				BLK_OPEN_READ | BLK_OPEN_WRITE, zram, NULL);
-	if (IS_ERR(bdev_file)) {
-		err = PTR_ERR(bdev_file);
-		bdev_file = NULL;
-		goto out;
-	}
-
 	nr_pages = i_size_read(inode) >> PAGE_SHIFT;
 	bitmap_sz = BITS_TO_LONGS(nr_pages) * sizeof(long);
 	bitmap = kvzalloc(bitmap_sz, GFP_KERNEL);
@@ -531,7 +519,7 @@ static ssize_t backing_dev_store(struct device *dev,
 
 	reset_bdev(zram);
 
-	zram->bdev_file = bdev_file;
+	zram->bdev = I_BDEV(inode);
 	zram->backing_dev = backing_dev;
 	zram->bitmap = bitmap;
 	zram->nr_pages = nr_pages;
@@ -544,9 +532,6 @@ static ssize_t backing_dev_store(struct device *dev,
 out:
 	kvfree(bitmap);
 
-	if (bdev_file)
-		fput(bdev_file);
-
 	if (backing_dev)
 		filp_close(backing_dev, NULL);
 
@@ -587,7 +572,7 @@ static void read_from_bdev_async(struct zram *zram, struct page *page,
 {
 	struct bio *bio;
 
-	bio = bio_alloc(file_bdev(zram->bdev_file), 1, parent->bi_opf, GFP_NOIO);
+	bio = bio_alloc(zram->bdev, 1, parent->bi_opf, GFP_NOIO);
 	bio->bi_iter.bi_sector = entry * (PAGE_SIZE >> 9);
 	__bio_add_page(bio, page, PAGE_SIZE, 0);
 	bio_chain(bio, parent);
@@ -703,7 +688,7 @@ static ssize_t writeback_store(struct device *dev,
 			continue;
 		}
 
-		bio_init(&bio, file_bdev(zram->bdev_file), &bio_vec, 1,
+		bio_init(&bio, zram->bdev, &bio_vec, 1,
 			 REQ_OP_WRITE | REQ_SYNC);
 		bio.bi_iter.bi_sector = blk_idx * (PAGE_SIZE >> 9);
 		__bio_add_page(&bio, page, PAGE_SIZE, 0);
@@ -785,7 +770,7 @@ static void zram_sync_read(struct work_struct *work)
 	struct bio_vec bv;
 	struct bio bio;
 
-	bio_init(&bio, file_bdev(zw->zram->bdev_file), &bv, 1, REQ_OP_READ);
+	bio_init(&bio, zw->zram->bdev, &bv, 1, REQ_OP_READ);
 	bio.bi_iter.bi_sector = zw->entry * (PAGE_SIZE >> 9);
 	__bio_add_page(&bio, zw->page, PAGE_SIZE, 0);
 	zw->error = submit_bio_wait(&bio);

drivers/block/zram/zram_drv.h

@@ -132,7 +132,7 @@ struct zram {
 	spinlock_t wb_limit_lock;
 	bool wb_limit_enable;
 	u64 bd_wb_limit;
-	struct file *bdev_file;
+	struct block_device *bdev;
 	unsigned long *bitmap;
 	unsigned long nr_pages;
 #endif

drivers/md/bcache/super.c

@@ -2555,10 +2555,6 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 	if (IS_ERR(bdev_file))
 		goto out_free_sb;
 
-	err = "failed to set blocksize";
-	if (set_blocksize(file_bdev(bdev_file), 4096))
-		goto out_blkdev_put;
-
 	err = read_super(sb, file_bdev(bdev_file), &sb_disk);
 	if (err)
 		goto out_blkdev_put;

fs/btrfs/dev-replace.c

@@ -316,7 +316,7 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 	set_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state);
 	set_bit(BTRFS_DEV_STATE_REPLACE_TGT, &device->dev_state);
 	device->dev_stats_valid = 1;
-	set_blocksize(device->bdev, BTRFS_BDEV_BLOCKSIZE);
+	set_blocksize(bdev_file, BTRFS_BDEV_BLOCKSIZE);
 	device->fs_devices = fs_devices;
 
 	ret = btrfs_get_dev_zone_info(device, false);

fs/btrfs/volumes.c

@@ -482,10 +482,12 @@ btrfs_get_bdev_and_sb(const char *device_path, blk_mode_t flags, void *holder,
 
 	if (flush)
 		sync_blockdev(bdev);
-	ret = set_blocksize(bdev, BTRFS_BDEV_BLOCKSIZE);
-	if (ret) {
-		fput(*bdev_file);
-		goto error;
+	if (holder) {
+		ret = set_blocksize(*bdev_file, BTRFS_BDEV_BLOCKSIZE);
+		if (ret) {
+			fput(*bdev_file);
+			goto error;
+		}
 	}
 	invalidate_bdev(bdev);
 	*disk_super = btrfs_read_dev_super(bdev);
@@ -498,6 +500,7 @@ btrfs_get_bdev_and_sb(const char *device_path, blk_mode_t flags, void *holder,
 	return 0;
 
 error:
+	*disk_super = NULL;
 	*bdev_file = NULL;
 	return ret;
 }
@@ -2714,7 +2717,7 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
 	set_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state);
 	clear_bit(BTRFS_DEV_STATE_REPLACE_TGT, &device->dev_state);
 	device->dev_stats_valid = 1;
-	set_blocksize(device->bdev, BTRFS_BDEV_BLOCKSIZE);
+	set_blocksize(device->bdev_file, BTRFS_BDEV_BLOCKSIZE);
 
 	if (seeding_dev) {
 		btrfs_clear_sb_rdonly(sb);

fs/ext4/super.c

@@ -5866,7 +5866,7 @@ static struct file *ext4_get_journal_blkdev(struct super_block *sb,
 
 	sb_block = EXT4_MIN_BLOCK_SIZE / blocksize;
 	offset = EXT4_MIN_BLOCK_SIZE % blocksize;
-	set_blocksize(bdev, blocksize);
+	set_blocksize(bdev_file, blocksize);
 	bh = __bread(bdev, sb_block, blocksize);
 	if (!bh) {
 		ext4_msg(sb, KERN_ERR, "couldn't read superblock of "

fs/reiserfs/journal.c

@@ -2626,8 +2626,7 @@ static int journal_init_dev(struct super_block *super,
 					 MAJOR(jdev), MINOR(jdev), result);
 			return result;
 		} else if (jdev != super->s_dev)
-			set_blocksize(file_bdev(journal->j_bdev_file),
-				      super->s_blocksize);
+			set_blocksize(journal->j_bdev_file, super->s_blocksize);
 
 		return 0;
 	}
@@ -2643,7 +2642,7 @@ static int journal_init_dev(struct super_block *super,
 		return result;
 	}
 
-	set_blocksize(file_bdev(journal->j_bdev_file), super->s_blocksize);
+	set_blocksize(journal->j_bdev_file, super->s_blocksize);
 	reiserfs_info(super,
 		      "journal_init_dev: journal device: %pg\n",
 		      file_bdev(journal->j_bdev_file));

fs/xfs/xfs_buf.c

@@ -2046,7 +2046,7 @@ xfs_setsize_buftarg(
 	btp->bt_meta_sectorsize = sectorsize;
 	btp->bt_meta_sectormask = sectorsize - 1;
 
-	if (set_blocksize(btp->bt_bdev, sectorsize)) {
+	if (set_blocksize(btp->bt_bdev_file, sectorsize)) {
 		xfs_warn(btp->bt_mount,
 			"Cannot set_blocksize to %u on device %pg",
 			sectorsize, btp->bt_bdev);

include/linux/blkdev.h

@@ -1481,7 +1481,7 @@ static inline void bio_end_io_acct(struct bio *bio, unsigned long start_time)
 }
 
 int bdev_read_only(struct block_device *bdev);
-int set_blocksize(struct block_device *bdev, int size);
+int set_blocksize(struct file *file, int size);
 
 int lookup_bdev(const char *pathname, dev_t *dev);
 

include/linux/swap.h

@@ -310,10 +310,8 @@ struct swap_info_struct {
 	unsigned int __percpu *cluster_next_cpu; /*percpu index for next allocation */
 	struct percpu_cluster __percpu *percpu_cluster; /* per cpu's swap location */
 	struct rb_root swap_extent_root;/* root of the swap extent rbtree */
-	struct file *bdev_file;		/* open handle of the bdev */
 	struct block_device *bdev;	/* swap device or bdev of swap file */
 	struct file *swap_file;		/* seldom referenced */
-	unsigned int old_block_size;	/* seldom referenced */
 	struct completion comp;		/* seldom referenced */
 	spinlock_t lock;		/*
 					 * protect map scan related fields like

kernel/power/swap.c

@@ -368,11 +368,7 @@ static int swsusp_swap_check(void)
 	if (IS_ERR(hib_resume_bdev_file))
 		return PTR_ERR(hib_resume_bdev_file);
 
-	res = set_blocksize(file_bdev(hib_resume_bdev_file), PAGE_SIZE);
-	if (res < 0)
-		fput(hib_resume_bdev_file);
-
-	return res;
+	return 0;
 }
 
 /**
@@ -1574,7 +1570,6 @@ int swsusp_check(bool exclusive)
 	hib_resume_bdev_file = bdev_file_open_by_dev(swsusp_resume_device,
 				BLK_OPEN_READ, holder, NULL);
 	if (!IS_ERR(hib_resume_bdev_file)) {
-		set_blocksize(file_bdev(hib_resume_bdev_file), PAGE_SIZE);
 		clear_page(swsusp_header);
 		error = hib_submit_io(REQ_OP_READ, swsusp_resume_block,
 					swsusp_header, NULL);

mm/swapfile.c

@@ -2469,7 +2469,6 @@ SYSCALL_DEFINE1(swapoff, const char __user *, specialfile)
 	struct inode *inode;
 	struct filename *pathname;
 	int err, found = 0;
-	unsigned int old_block_size;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
@@ -2582,7 +2581,6 @@ SYSCALL_DEFINE1(swapoff, const char __user *, specialfile)
 	}
 
 	swap_file = p->swap_file;
-	old_block_size = p->old_block_size;
 	p->swap_file = NULL;
 	p->max = 0;
 	swap_map = p->swap_map;
@@ -2605,11 +2603,6 @@ SYSCALL_DEFINE1(swapoff, const char __user *, specialfile)
 	exit_swap_address_space(p->type);
 
 	inode = mapping->host;
-	if (p->bdev_file) {
-		set_blocksize(p->bdev, old_block_size);
-		fput(p->bdev_file);
-		p->bdev_file = NULL;
-	}
 
 	inode_lock(inode);
 	inode->i_flags &= ~S_SWAPFILE;
@@ -2835,21 +2828,8 @@ static struct swap_info_struct *alloc_swap_info(void)
 
 static int claim_swapfile(struct swap_info_struct *p, struct inode *inode)
 {
-	int error;
-
 	if (S_ISBLK(inode->i_mode)) {
-		p->bdev_file = bdev_file_open_by_dev(inode->i_rdev,
-				BLK_OPEN_READ | BLK_OPEN_WRITE, p, NULL);
-		if (IS_ERR(p->bdev_file)) {
-			error = PTR_ERR(p->bdev_file);
-			p->bdev_file = NULL;
-			return error;
-		}
-		p->bdev = file_bdev(p->bdev_file);
-		p->old_block_size = block_size(p->bdev);
-		error = set_blocksize(p->bdev, PAGE_SIZE);
-		if (error < 0)
-			return error;
+		p->bdev = I_BDEV(inode);
 		/*
 		 * Zoned block devices contain zones that have a sequential
 		 * write only restriction.  Hence zoned block devices are not
@@ -3090,7 +3070,7 @@ SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags)
 		name = NULL;
 		goto bad_swap;
 	}
-	swap_file = file_open_name(name, O_RDWR|O_LARGEFILE, 0);
+	swap_file = file_open_name(name, O_RDWR | O_LARGEFILE | O_EXCL, 0);
 	if (IS_ERR(swap_file)) {
 		error = PTR_ERR(swap_file);
 		swap_file = NULL;
@@ -3289,11 +3269,6 @@ SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags)
 	p->percpu_cluster = NULL;
 	free_percpu(p->cluster_next_cpu);
 	p->cluster_next_cpu = NULL;
-	if (p->bdev_file) {
-		set_blocksize(p->bdev, p->old_block_size);
-		fput(p->bdev_file);
-		p->bdev_file = NULL;
-	}
 	inode = NULL;
 	destroy_swap_extents(p);
 	swap_cgroup_swapoff(p->type);