fs/btrfs/volumes.c

// SPDX-License-Identifier: GPL-2.0
/*
 * Copyright (C) 2007 Oracle.  All rights reserved.
 */

#include <linux/sched.h>
#include <linux/sched/mm.h>
#include <linux/bio.h>
#include <linux/slab.h>
#include <linux/blkdev.h>
#include <linux/ratelimit.h>
#include <linux/kthread.h>
#include <linux/raid/pq.h>
#include <linux/semaphore.h>
#include <linux/uuid.h>
#include <linux/list_sort.h>
#include <linux/namei.h>
#include "misc.h"
#include "ctree.h"
#include "extent_map.h"
#include "disk-io.h"
#include "transaction.h"
#include "print-tree.h"
#include "volumes.h"
#include "raid56.h"
#include "async-thread.h"
#include "check-integrity.h"
#include "rcu-string.h"
#include "dev-replace.h"
#include "sysfs.h"
#include "tree-checker.h"
#include "space-info.h"
#include "block-group.h"
#include "discard.h"
#include "zoned.h"

const struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = {
	[BTRFS_RAID_RAID10] = {
		.sub_stripes	= 2,
		.dev_stripes	= 1,
		.devs_max	= 0,	/* 0 == as many as possible */
		.devs_min	= 2,
		.tolerated_failures = 1,
		.devs_increment	= 2,
		.ncopies	= 2,
		.nparity        = 0,
		.raid_name	= "raid10",
		.bg_flag	= BTRFS_BLOCK_GROUP_RAID10,
		.mindev_error	= BTRFS_ERROR_DEV_RAID10_MIN_NOT_MET,
	},
	[BTRFS_RAID_RAID1] = {
		.sub_stripes	= 1,
		.dev_stripes	= 1,
		.devs_max	= 2,
		.devs_min	= 2,
		.tolerated_failures = 1,
		.devs_increment	= 2,
		.ncopies	= 2,
		.nparity        = 0,
		.raid_name	= "raid1",
		.bg_flag	= BTRFS_BLOCK_GROUP_RAID1,
		.mindev_error	= BTRFS_ERROR_DEV_RAID1_MIN_NOT_MET,
	},
	[BTRFS_RAID_RAID1C3] = {
		.sub_stripes	= 1,
		.dev_stripes	= 1,
		.devs_max	= 3,
		.devs_min	= 3,
		.tolerated_failures = 2,
		.devs_increment	= 3,
		.ncopies	= 3,
		.nparity        = 0,
		.raid_name	= "raid1c3",
		.bg_flag	= BTRFS_BLOCK_GROUP_RAID1C3,
		.mindev_error	= BTRFS_ERROR_DEV_RAID1C3_MIN_NOT_MET,
	},
	[BTRFS_RAID_RAID1C4] = {
		.sub_stripes	= 1,
		.dev_stripes	= 1,
		.devs_max	= 4,
		.devs_min	= 4,
		.tolerated_failures = 3,
		.devs_increment	= 4,
		.ncopies	= 4,
		.nparity        = 0,
		.raid_name	= "raid1c4",
		.bg_flag	= BTRFS_BLOCK_GROUP_RAID1C4,
		.mindev_error	= BTRFS_ERROR_DEV_RAID1C4_MIN_NOT_MET,
	},
	[BTRFS_RAID_DUP] = {
		.sub_stripes	= 1,
		.dev_stripes	= 2,
		.devs_max	= 1,
		.devs_min	= 1,
		.tolerated_failures = 0,
		.devs_increment	= 1,
		.ncopies	= 2,
		.nparity        = 0,
		.raid_name	= "dup",
		.bg_flag	= BTRFS_BLOCK_GROUP_DUP,
		.mindev_error	= 0,
	},
	[BTRFS_RAID_RAID0] = {
		.sub_stripes	= 1,
		.dev_stripes	= 1,
		.devs_max	= 0,
		.devs_min	= 1,
		.tolerated_failures = 0,
		.devs_increment	= 1,
		.ncopies	= 1,
		.nparity        = 0,
		.raid_name	= "raid0",
		.bg_flag	= BTRFS_BLOCK_GROUP_RAID0,
		.mindev_error	= 0,
	},
	[BTRFS_RAID_SINGLE] = {
		.sub_stripes	= 1,
		.dev_stripes	= 1,
		.devs_max	= 1,
		.devs_min	= 1,
		.tolerated_failures = 0,
		.devs_increment	= 1,
		.ncopies	= 1,
		.nparity        = 0,
		.raid_name	= "single",
		.bg_flag	= 0,
		.mindev_error	= 0,
	},
	[BTRFS_RAID_RAID5] = {
		.sub_stripes	= 1,
		.dev_stripes	= 1,
		.devs_max	= 0,
		.devs_min	= 2,
		.tolerated_failures = 1,
		.devs_increment	= 1,
		.ncopies	= 1,
		.nparity        = 1,
		.raid_name	= "raid5",
		.bg_flag	= BTRFS_BLOCK_GROUP_RAID5,
		.mindev_error	= BTRFS_ERROR_DEV_RAID5_MIN_NOT_MET,
	},
	[BTRFS_RAID_RAID6] = {
		.sub_stripes	= 1,
		.dev_stripes	= 1,
		.devs_max	= 0,
		.devs_min	= 3,
		.tolerated_failures = 2,
		.devs_increment	= 1,
		.ncopies	= 1,
		.nparity        = 2,
		.raid_name	= "raid6",
		.bg_flag	= BTRFS_BLOCK_GROUP_RAID6,
		.mindev_error	= BTRFS_ERROR_DEV_RAID6_MIN_NOT_MET,
	},
};

/*
 * Convert block group flags (BTRFS_BLOCK_GROUP_*) to btrfs_raid_types, which
 * can be used as index to access btrfs_raid_array[].
 */
enum btrfs_raid_types __attribute_const__ btrfs_bg_flags_to_raid_index(u64 flags)
{
	if (flags & BTRFS_BLOCK_GROUP_RAID10)
		return BTRFS_RAID_RAID10;
	else if (flags & BTRFS_BLOCK_GROUP_RAID1)
		return BTRFS_RAID_RAID1;
	else if (flags & BTRFS_BLOCK_GROUP_RAID1C3)
		return BTRFS_RAID_RAID1C3;
	else if (flags & BTRFS_BLOCK_GROUP_RAID1C4)
		return BTRFS_RAID_RAID1C4;
	else if (flags & BTRFS_BLOCK_GROUP_DUP)
		return BTRFS_RAID_DUP;
	else if (flags & BTRFS_BLOCK_GROUP_RAID0)
		return BTRFS_RAID_RAID0;
	else if (flags & BTRFS_BLOCK_GROUP_RAID5)
		return BTRFS_RAID_RAID5;
	else if (flags & BTRFS_BLOCK_GROUP_RAID6)
		return BTRFS_RAID_RAID6;

	return BTRFS_RAID_SINGLE; /* BTRFS_BLOCK_GROUP_SINGLE */
}

const char *btrfs_bg_type_to_raid_name(u64 flags)
{
	const int index = btrfs_bg_flags_to_raid_index(flags);

	if (index >= BTRFS_NR_RAID_TYPES)
		return NULL;

	return btrfs_raid_array[index].raid_name;
}

/*
 * Fill @buf with textual description of @bg_flags, no more than @size_buf
 * bytes including terminating null byte.
 */
void btrfs_describe_block_groups(u64 bg_flags, char *buf, u32 size_buf)
{
	int i;
	int ret;
	char *bp = buf;
	u64 flags = bg_flags;
	u32 size_bp = size_buf;

	if (!flags) {
		strcpy(bp, "NONE");
		return;
	}

#define DESCRIBE_FLAG(flag, desc)						\
	do {								\
		if (flags & (flag)) {					\
			ret = snprintf(bp, size_bp, "%s|", (desc));	\
			if (ret < 0 || ret >= size_bp)			\
				goto out_overflow;			\
			size_bp -= ret;					\
			bp += ret;					\
			flags &= ~(flag);				\
		}							\
	} while (0)

	DESCRIBE_FLAG(BTRFS_BLOCK_GROUP_DATA, "data");
	DESCRIBE_FLAG(BTRFS_BLOCK_GROUP_SYSTEM, "system");
	DESCRIBE_FLAG(BTRFS_BLOCK_GROUP_METADATA, "metadata");

	DESCRIBE_FLAG(BTRFS_AVAIL_ALLOC_BIT_SINGLE, "single");
	for (i = 0; i < BTRFS_NR_RAID_TYPES; i++)
		DESCRIBE_FLAG(btrfs_raid_array[i].bg_flag,
			      btrfs_raid_array[i].raid_name);
#undef DESCRIBE_FLAG

	if (flags) {
		ret = snprintf(bp, size_bp, "0x%llx|", flags);
		size_bp -= ret;
	}

	if (size_bp < size_buf)
		buf[size_buf - size_bp - 1] = '\0'; /* remove last | */

	/*
	 * The text is trimmed, it's up to the caller to provide sufficiently
	 * large buffer
	 */
out_overflow:;
}

static int init_first_rw_device(struct btrfs_trans_handle *trans);
static int btrfs_relocate_sys_chunks(struct btrfs_fs_info *fs_info);
static void btrfs_dev_stat_print_on_error(struct btrfs_device *dev);
static void btrfs_dev_stat_print_on_load(struct btrfs_device *device);
static int __btrfs_map_block(struct btrfs_fs_info *fs_info,
			     enum btrfs_map_op op,
			     u64 logical, u64 *length,
			     struct btrfs_io_context **bioc_ret,
			     int mirror_num, int need_raid_map);

/*
 * Device locking
 * ==============
 *
 * There are several mutexes that protect manipulation of devices and low-level
 * structures like chunks but not block groups, extents or files
 *
 * uuid_mutex (global lock)
 * ------------------------
 * protects the fs_uuids list that tracks all per-fs fs_devices, resulting from
 * the SCAN_DEV ioctl registration or from mount either implicitly (the first
 * device) or requested by the device= mount option
 *
 * the mutex can be very coarse and can cover long-running operations
 *
 * protects: updates to fs_devices counters like missing devices, rw devices,
 * seeding, structure cloning, opening/closing devices at mount/umount time
 *
 * global::fs_devs - add, remove, updates to the global list
 *
 * does not protect: manipulation of the fs_devices::devices list in general
 * but in mount context it could be used to exclude list modifications by eg.
 * scan ioctl
 *
 * btrfs_device::name - renames (write side), read is RCU
 *
 * fs_devices::device_list_mutex (per-fs, with RCU)
 * ------------------------------------------------
 * protects updates to fs_devices::devices, ie. adding and deleting
 *
 * simple list traversal with read-only actions can be done with RCU protection
 *
 * may be used to exclude some operations from running concurrently without any
 * modifications to the list (see write_all_supers)
 *
 * Is not required at mount and close times, because our device list is
 * protected by the uuid_mutex at that point.
 *
 * balance_mutex
 * -------------
 * protects balance structures (status, state) and context accessed from
 * several places (internally, ioctl)
 *
 * chunk_mutex
 * -----------
 * protects chunks, adding or removing during allocation, trim or when a new
 * device is added/removed. Additionally it also protects post_commit_list of
 * individual devices, since they can be added to the transaction's
 * post_commit_list only with chunk_mutex held.
 *
 * cleaner_mutex
 * -------------
 * a big lock that is held by the cleaner thread and prevents running subvolume
 * cleaning together with relocation or delayed iputs
 *
 *
 * Lock nesting
 * ============
 *
 * uuid_mutex
 *   device_list_mutex
 *     chunk_mutex
 *   balance_mutex
 *
 *
 * Exclusive operations
 * ====================
 *
 * Maintains the exclusivity of the following operations that apply to the
 * whole filesystem and cannot run in parallel.
 *
 * - Balance (*)
 * - Device add
 * - Device remove
 * - Device replace (*)
 * - Resize
 *
 * The device operations (as above) can be in one of the following states:
 *
 * - Running state
 * - Paused state
 * - Completed state
 *
 * Only device operations marked with (*) can go into the Paused state for the
 * following reasons:
 *
 * - ioctl (only Balance can be Paused through ioctl)
 * - filesystem remounted as read-only
 * - filesystem unmounted and mounted as read-only
 * - system power-cycle and filesystem mounted as read-only
 * - filesystem or device errors leading to forced read-only
 *
 * The status of exclusive operation is set and cleared atomically.
 * During the course of Paused state, fs_info::exclusive_operation remains set.
 * A device operation in Paused or Running state can be canceled or resumed
 * either by ioctl (Balance only) or when remounted as read-write.
 * The exclusive status is cleared when the device operation is canceled or
 * completed.
 */

DEFINE_MUTEX(uuid_mutex);
static LIST_HEAD(fs_uuids);
struct list_head * __attribute_const__ btrfs_get_fs_uuids(void)
{
	return &fs_uuids;
}

/*
 * alloc_fs_devices - allocate struct btrfs_fs_devices
 * @fsid:		if not NULL, copy the UUID to fs_devices::fsid
 * @metadata_fsid:	if not NULL, copy the UUID to fs_devices::metadata_fsid
 *
 * Return a pointer to a new struct btrfs_fs_devices on success, or ERR_PTR().
 * The returned struct is not linked onto any lists and can be destroyed with
 * kfree() right away.
 */
static struct btrfs_fs_devices *alloc_fs_devices(const u8 *fsid,
						 const u8 *metadata_fsid)
{
	struct btrfs_fs_devices *fs_devs;

	fs_devs = kzalloc(sizeof(*fs_devs), GFP_KERNEL);
	if (!fs_devs)
		return ERR_PTR(-ENOMEM);

	mutex_init(&fs_devs->device_list_mutex);

	INIT_LIST_HEAD(&fs_devs->devices);
	INIT_LIST_HEAD(&fs_devs->alloc_list);
	INIT_LIST_HEAD(&fs_devs->fs_list);
	INIT_LIST_HEAD(&fs_devs->seed_list);
	if (fsid)
		memcpy(fs_devs->fsid, fsid, BTRFS_FSID_SIZE);

	if (metadata_fsid)
		memcpy(fs_devs->metadata_uuid, metadata_fsid, BTRFS_FSID_SIZE);
	else if (fsid)
		memcpy(fs_devs->metadata_uuid, fsid, BTRFS_FSID_SIZE);

	return fs_devs;
}

void btrfs_free_device(struct btrfs_device *device)
{
	WARN_ON(!list_empty(&device->post_commit_list));
	rcu_string_free(device->name);
	extent_io_tree_release(&device->alloc_state);
	bio_put(device->flush_bio);
	btrfs_destroy_dev_zone_info(device);
	kfree(device);
}

static void free_fs_devices(struct btrfs_fs_devices *fs_devices)
{
	struct btrfs_device *device;
	WARN_ON(fs_devices->opened);
	while (!list_empty(&fs_devices->devices)) {
		device = list_entry(fs_devices->devices.next,
				    struct btrfs_device, dev_list);
		list_del(&device->dev_list);
		btrfs_free_device(device);
	}
	kfree(fs_devices);
}

void __exit btrfs_cleanup_fs_uuids(void)
{
	struct btrfs_fs_devices *fs_devices;

	while (!list_empty(&fs_uuids)) {
		fs_devices = list_entry(fs_uuids.next,
					struct btrfs_fs_devices, fs_list);
		list_del(&fs_devices->fs_list);
		free_fs_devices(fs_devices);
	}
}

static noinline struct btrfs_fs_devices *find_fsid(
		const u8 *fsid, const u8 *metadata_fsid)
{
	struct btrfs_fs_devices *fs_devices;

	ASSERT(fsid);

	/* Handle non-split brain cases */
	list_for_each_entry(fs_devices, &fs_uuids, fs_list) {
		if (metadata_fsid) {
			if (memcmp(fsid, fs_devices->fsid, BTRFS_FSID_SIZE) == 0
			    && memcmp(metadata_fsid, fs_devices->metadata_uuid,
				      BTRFS_FSID_SIZE) == 0)
				return fs_devices;
		} else {
			if (memcmp(fsid, fs_devices->fsid, BTRFS_FSID_SIZE) == 0)
				return fs_devices;
		}
	}
	return NULL;
}

static struct btrfs_fs_devices *find_fsid_with_metadata_uuid(
				struct btrfs_super_block *disk_super)
{

	struct btrfs_fs_devices *fs_devices;

	/*
	 * Handle scanned device having completed its fsid change but
	 * belonging to a fs_devices that was created by first scanning
	 * a device which didn't have its fsid/metadata_uuid changed
	 * at all and the CHANGING_FSID_V2 flag set.
	 */
	list_for_each_entry(fs_devices, &fs_uuids, fs_list) {
		if (fs_devices->fsid_change &&
		    memcmp(disk_super->metadata_uuid, fs_devices->fsid,
			   BTRFS_FSID_SIZE) == 0 &&
		    memcmp(fs_devices->fsid, fs_devices->metadata_uuid,
			   BTRFS_FSID_SIZE) == 0) {
			return fs_devices;
		}
	}
	/*
	 * Handle scanned device having completed its fsid change but
	 * belonging to a fs_devices that was created by a device that
	 * has an outdated pair of fsid/metadata_uuid and
	 * CHANGING_FSID_V2 flag set.
	 */
	list_for_each_entry(fs_devices, &fs_uuids, fs_list) {
		if (fs_devices->fsid_change &&
		    memcmp(fs_devices->metadata_uuid,
			   fs_devices->fsid, BTRFS_FSID_SIZE) != 0 &&
		    memcmp(disk_super->metadata_uuid, fs_devices->metadata_uuid,
			   BTRFS_FSID_SIZE) == 0) {
			return fs_devices;
		}
	}

	return find_fsid(disk_super->fsid, disk_super->metadata_uuid);
}


static int
btrfs_get_bdev_and_sb(const char *device_path, fmode_t flags, void *holder,
		      int flush, struct block_device **bdev,
		      struct btrfs_super_block **disk_super)
{
	int ret;

	*bdev = blkdev_get_by_path(device_path, flags, holder);

	if (IS_ERR(*bdev)) {
		ret = PTR_ERR(*bdev);
		goto error;
	}

	if (flush)
		sync_blockdev(*bdev);
	ret = set_blocksize(*bdev, BTRFS_BDEV_BLOCKSIZE);
	if (ret) {
		blkdev_put(*bdev, flags);
		goto error;
	}
	invalidate_bdev(*bdev);
	*disk_super = btrfs_read_dev_super(*bdev);
	if (IS_ERR(*disk_super)) {
		ret = PTR_ERR(*disk_super);
		blkdev_put(*bdev, flags);
		goto error;
	}

	return 0;

error:
	*bdev = NULL;
	return ret;
}

static bool device_path_matched(const char *path, struct btrfs_device *device)
{
	int found;

	rcu_read_lock();
	found = strcmp(rcu_str_deref(device->name), path);
	rcu_read_unlock();

	return found == 0;
}

/*
 *  Search and remove all stale (devices which are not mounted) devices.
 *  When both inputs are NULL, it will search and release all stale devices.
 *  path:	Optional. When provided will it release all unmounted devices
 *		matching this path only.
 *  skip_dev:	Optional. Will skip this device when searching for the stale
 *		devices.
 *  Return:	0 for success or if @path is NULL.
 * 		-EBUSY if @path is a mounted device.
 * 		-ENOENT if @path does not match any device in the list.
 */
static int btrfs_free_stale_devices(const char *path,
				     struct btrfs_device *skip_device)
{
	struct btrfs_fs_devices *fs_devices, *tmp_fs_devices;
	struct btrfs_device *device, *tmp_device;
	int ret = 0;

	lockdep_assert_held(&uuid_mutex);

	if (path)
		ret = -ENOENT;

	list_for_each_entry_safe(fs_devices, tmp_fs_devices, &fs_uuids, fs_list) {

		mutex_lock(&fs_devices->device_list_mutex);
		list_for_each_entry_safe(device, tmp_device,
					 &fs_devices->devices, dev_list) {
			if (skip_device && skip_device == device)
				continue;
			if (path && !device->name)
				continue;
			if (path && !device_path_matched(path, device))
				continue;
			if (fs_devices->opened) {
				/* for an already deleted device return 0 */
				if (path && ret != 0)
					ret = -EBUSY;
				break;
			}

			/* delete the stale device */
			fs_devices->num_devices--;
			list_del(&device->dev_list);
			btrfs_free_device(device);

			ret = 0;
		}
		mutex_unlock(&fs_devices->device_list_mutex);

		if (fs_devices->num_devices == 0) {
			btrfs_sysfs_remove_fsid(fs_devices);
			list_del(&fs_devices->fs_list);
			free_fs_devices(fs_devices);
		}
	}

	return ret;
}

/*
 * This is only used on mount, and we are protected from competing things
 * messing with our fs_devices by the uuid_mutex, thus we do not need the
 * fs_devices->device_list_mutex here.
 */
static int btrfs_open_one_device(struct btrfs_fs_devices *fs_devices,
			struct btrfs_device *device, fmode_t flags,
			void *holder)
{
	struct request_queue *q;
	struct block_device *bdev;
	struct btrfs_super_block *disk_super;
	u64 devid;
	int ret;

	if (device->bdev)
		return -EINVAL;
	if (!device->name)
		return -EINVAL;

	ret = btrfs_get_bdev_and_sb(device->name->str, flags, holder, 1,
				    &bdev, &disk_super);
	if (ret)
		return ret;

	devid = btrfs_stack_device_id(&disk_super->dev_item);
	if (devid != device->devid)
		goto error_free_page;

	if (memcmp(device->uuid, disk_super->dev_item.uuid, BTRFS_UUID_SIZE))
		goto error_free_page;

	device->generation = btrfs_super_generation(disk_super);

	if (btrfs_super_flags(disk_super) & BTRFS_SUPER_FLAG_SEEDING) {
		if (btrfs_super_incompat_flags(disk_super) &
		    BTRFS_FEATURE_INCOMPAT_METADATA_UUID) {
			pr_err(
		"BTRFS: Invalid seeding and uuid-changed device detected\n");
			goto error_free_page;
		}

		clear_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state);
		fs_devices->seeding = true;
	} else {
		if (bdev_read_only(bdev))
			clear_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state);
		else
			set_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state);
	}

	q = bdev_get_queue(bdev);
	if (!blk_queue_nonrot(q))
		fs_devices->rotating = true;

	device->bdev = bdev;
	clear_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state);
	device->mode = flags;

	fs_devices->open_devices++;
	if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state) &&
	    device->devid != BTRFS_DEV_REPLACE_DEVID) {
		fs_devices->rw_devices++;
		list_add_tail(&device->dev_alloc_list, &fs_devices->alloc_list);
	}
	btrfs_release_disk_super(disk_super);

	return 0;

error_free_page:
	btrfs_release_disk_super(disk_super);
	blkdev_put(bdev, flags);

	return -EINVAL;
}

/*
 * Handle scanned device having its CHANGING_FSID_V2 flag set and the fs_devices
 * being created with a disk that has already completed its fsid change. Such
 * disk can belong to an fs which has its FSID changed or to one which doesn't.
 * Handle both cases here.
 */
static struct btrfs_fs_devices *find_fsid_inprogress(
					struct btrfs_super_block *disk_super)
{
	struct btrfs_fs_devices *fs_devices;

	list_for_each_entry(fs_devices, &fs_uuids, fs_list) {
		if (memcmp(fs_devices->metadata_uuid, fs_devices->fsid,
			   BTRFS_FSID_SIZE) != 0 &&
		    memcmp(fs_devices->metadata_uuid, disk_super->fsid,
			   BTRFS_FSID_SIZE) == 0 && !fs_devices->fsid_change) {
			return fs_devices;
		}
	}

	return find_fsid(disk_super->fsid, NULL);
}


static struct btrfs_fs_devices *find_fsid_changed(
					struct btrfs_super_block *disk_super)
{
	struct btrfs_fs_devices *fs_devices;

	/*
	 * Handles the case where scanned device is part of an fs that had
	 * multiple successful changes of FSID but currently device didn't
	 * observe it. Meaning our fsid will be different than theirs. We need
	 * to handle two subcases :
	 *  1 - The fs still continues to have different METADATA/FSID uuids.
	 *  2 - The fs is switched back to its original FSID (METADATA/FSID
	 *  are equal).
	 */
	list_for_each_entry(fs_devices, &fs_uuids, fs_list) {
		/* Changed UUIDs */
		if (memcmp(fs_devices->metadata_uuid, fs_devices->fsid,
			   BTRFS_FSID_SIZE) != 0 &&
		    memcmp(fs_devices->metadata_uuid, disk_super->metadata_uuid,
			   BTRFS_FSID_SIZE) == 0 &&
		    memcmp(fs_devices->fsid, disk_super->fsid,
			   BTRFS_FSID_SIZE) != 0)
			return fs_devices;

		/* Unchanged UUIDs */
		if (memcmp(fs_devices->metadata_uuid, fs_devices->fsid,
			   BTRFS_FSID_SIZE) == 0 &&
		    memcmp(fs_devices->fsid, disk_super->metadata_uuid,
			   BTRFS_FSID_SIZE) == 0)
			return fs_devices;
	}

	return NULL;
}

static struct btrfs_fs_devices *find_fsid_reverted_metadata(
				struct btrfs_super_block *disk_super)
{
	struct btrfs_fs_devices *fs_devices;

	/*
	 * Handle the case where the scanned device is part of an fs whose last
	 * metadata UUID change reverted it to the original FSID. At the same
	 * time * fs_devices was first created by another constitutent device
	 * which didn't fully observe the operation. This results in an
	 * btrfs_fs_devices created with metadata/fsid different AND
	 * btrfs_fs_devices::fsid_change set AND the metadata_uuid of the
	 * fs_devices equal to the FSID of the disk.
	 */
	list_for_each_entry(fs_devices, &fs_uuids, fs_list) {
		if (memcmp(fs_devices->fsid, fs_devices->metadata_uuid,
			   BTRFS_FSID_SIZE) != 0 &&
		    memcmp(fs_devices->metadata_uuid, disk_super->fsid,
			   BTRFS_FSID_SIZE) == 0 &&
		    fs_devices->fsid_change)
			return fs_devices;
	}

	return NULL;
}
/*
 * Add new device to list of registered devices
 *
 * Returns:
 * device pointer which was just added or updated when successful
 * error pointer when failed
 */
static noinline struct btrfs_device *device_list_add(const char *path,
			   struct btrfs_super_block *disk_super,
			   bool *new_device_added)
{
	struct btrfs_device *device;
	struct btrfs_fs_devices *fs_devices = NULL;
	struct rcu_string *name;
	u64 found_transid = btrfs_super_generation(disk_super);
	u64 devid = btrfs_stack_device_id(&disk_super->dev_item);
	bool has_metadata_uuid = (btrfs_super_incompat_flags(disk_super) &
		BTRFS_FEATURE_INCOMPAT_METADATA_UUID);
	bool fsid_change_in_progress = (btrfs_super_flags(disk_super) &
					BTRFS_SUPER_FLAG_CHANGING_FSID_V2);

	if (fsid_change_in_progress) {
		if (!has_metadata_uuid)
			fs_devices = find_fsid_inprogress(disk_super);
		else
			fs_devices = find_fsid_changed(disk_super);
	} else if (has_metadata_uuid) {
		fs_devices = find_fsid_with_metadata_uuid(disk_super);
	} else {
		fs_devices = find_fsid_reverted_metadata(disk_super);
		if (!fs_devices)
			fs_devices = find_fsid(disk_super->fsid, NULL);
	}


	if (!fs_devices) {
		if (has_metadata_uuid)
			fs_devices = alloc_fs_devices(disk_super->fsid,
						      disk_super->metadata_uuid);
		else
			fs_devices = alloc_fs_devices(disk_super->fsid, NULL);

		if (IS_ERR(fs_devices))
			return ERR_CAST(fs_devices);

		fs_devices->fsid_change = fsid_change_in_progress;

		mutex_lock(&fs_devices->device_list_mutex);
		list_add(&fs_devices->fs_list, &fs_uuids);

		device = NULL;
	} else {
		struct btrfs_dev_lookup_args args = {
			.devid = devid,
			.uuid = disk_super->dev_item.uuid,
		};

		mutex_lock(&fs_devices->device_list_mutex);
		device = btrfs_find_device(fs_devices, &args);

		/*
		 * If this disk has been pulled into an fs devices created by
		 * a device which had the CHANGING_FSID_V2 flag then replace the
		 * metadata_uuid/fsid values of the fs_devices.
		 */
		if (fs_devices->fsid_change &&
		    found_transid > fs_devices->latest_generation) {
			memcpy(fs_devices->fsid, disk_super->fsid,
					BTRFS_FSID_SIZE);

			if (has_metadata_uuid)
				memcpy(fs_devices->metadata_uuid,
				       disk_super->metadata_uuid,
				       BTRFS_FSID_SIZE);
			else
				memcpy(fs_devices->metadata_uuid,
				       disk_super->fsid, BTRFS_FSID_SIZE);

			fs_devices->fsid_change = false;
		}
	}

	if (!device) {
		if (fs_devices->opened) {
			mutex_unlock(&fs_devices->device_list_mutex);
			return ERR_PTR(-EBUSY);
		}

		device = btrfs_alloc_device(NULL, &devid,
					    disk_super->dev_item.uuid);
		if (IS_ERR(device)) {
			mutex_unlock(&fs_devices->device_list_mutex);
			/* we can safely leave the fs_devices entry around */
			return device;
		}

		name = rcu_string_strdup(path, GFP_NOFS);
		if (!name) {
			btrfs_free_device(device);
			mutex_unlock(&fs_devices->device_list_mutex);
			return ERR_PTR(-ENOMEM);
		}
		rcu_assign_pointer(device->name, name);

		list_add_rcu(&device->dev_list, &fs_devices->devices);
		fs_devices->num_devices++;

		device->fs_devices = fs_devices;
		*new_device_added = true;

		if (disk_super->label[0])
			pr_info(
	"BTRFS: device label %s devid %llu transid %llu %s scanned by %s (%d)\n",
				disk_super->label, devid, found_transid, path,
				current->comm, task_pid_nr(current));
		else
			pr_info(
	"BTRFS: device fsid %pU devid %llu transid %llu %s scanned by %s (%d)\n",
				disk_super->fsid, devid, found_transid, path,
				current->comm, task_pid_nr(current));

	} else if (!device->name || strcmp(device->name->str, path)) {
		/*
		 * When FS is already mounted.
		 * 1. If you are here and if the device->name is NULL that
		 *    means this device was missing at time of FS mount.
		 * 2. If you are here and if the device->name is different
		 *    from 'path' that means either
		 *      a. The same device disappeared and reappeared with
		 *         different name. or
		 *      b. The missing-disk-which-was-replaced, has
		 *         reappeared now.
		 *
		 * We must allow 1 and 2a above. But 2b would be a spurious
		 * and unintentional.
		 *
		 * Further in case of 1 and 2a above, the disk at 'path'
		 * would have missed some transaction when it was away and
		 * in case of 2a the stale bdev has to be updated as well.
		 * 2b must not be allowed at all time.
		 */

		/*
		 * For now, we do allow update to btrfs_fs_device through the
		 * btrfs dev scan cli after FS has been mounted.  We're still
		 * tracking a problem where systems fail mount by subvolume id
		 * when we reject replacement on a mounted FS.
		 */
		if (!fs_devices->opened && found_transid < device->generation) {
			/*
			 * That is if the FS is _not_ mounted and if you
			 * are here, that means there is more than one
			 * disk with same uuid and devid.We keep the one
			 * with larger generation number or the last-in if
			 * generation are equal.
			 */
			mutex_unlock(&fs_devices->device_list_mutex);
			return ERR_PTR(-EEXIST);
		}

		/*
		 * We are going to replace the device path for a given devid,
		 * make sure it's the same device if the device is mounted
		 */
		if (device->bdev) {
			int error;
			dev_t path_dev;

			error = lookup_bdev(path, &path_dev);
			if (error) {
				mutex_unlock(&fs_devices->device_list_mutex);
				return ERR_PTR(error);
			}

			if (device->bdev->bd_dev != path_dev) {
				mutex_unlock(&fs_devices->device_list_mutex);
				/*
				 * device->fs_info may not be reliable here, so
				 * pass in a NULL instead. This avoids a
				 * possible use-after-free when the fs_info and
				 * fs_info->sb are already torn down.
				 */
				btrfs_warn_in_rcu(NULL,
	"duplicate device %s devid %llu generation %llu scanned by %s (%d)",
						  path, devid, found_transid,
						  current->comm,
						  task_pid_nr(current));
				return ERR_PTR(-EEXIST);
			}
			btrfs_info_in_rcu(device->fs_info,
	"devid %llu device path %s changed to %s scanned by %s (%d)",
					  devid, rcu_str_deref(device->name),
					  path, current->comm,
					  task_pid_nr(current));
		}

		name = rcu_string_strdup(path, GFP_NOFS);
		if (!name) {
			mutex_unlock(&fs_devices->device_list_mutex);
			return ERR_PTR(-ENOMEM);
		}
		rcu_string_free(device->name);
		rcu_assign_pointer(device->name, name);
		if (test_bit(BTRFS_DEV_STATE_MISSING, &device->dev_state)) {
			fs_devices->missing_devices--;
			clear_bit(BTRFS_DEV_STATE_MISSING, &device->dev_state);
		}
	}

	/*
	 * Unmount does not free the btrfs_device struct but would zero
	 * generation along with most of the other members. So just update
	 * it back. We need it to pick the disk with largest generation
	 * (as above).
	 */
	if (!fs_devices->opened) {
		device->generation = found_transid;
		fs_devices->latest_generation = max_t(u64, found_transid,
						fs_devices->latest_generation);
	}

	fs_devices->total_devices = btrfs_super_num_devices(disk_super);

	mutex_unlock(&fs_devices->device_list_mutex);
	return device;
}

static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
{
	struct btrfs_fs_devices *fs_devices;
	struct btrfs_device *device;
	struct btrfs_device *orig_dev;
	int ret = 0;

	lockdep_assert_held(&uuid_mutex);

	fs_devices = alloc_fs_devices(orig->fsid, NULL);
	if (IS_ERR(fs_devices))
		return fs_devices;

	fs_devices->total_devices = orig->total_devices;

	list_for_each_entry(orig_dev, &orig->devices, dev_list) {
		struct rcu_string *name;

		device = btrfs_alloc_device(NULL, &orig_dev->devid,
					    orig_dev->uuid);
		if (IS_ERR(device)) {
			ret = PTR_ERR(device);
			goto error;
		}

		/*
		 * This is ok to do without rcu read locked because we hold the
		 * uuid mutex so nothing we touch in here is going to disappear.
		 */
		if (orig_dev->name) {
			name = rcu_string_strdup(orig_dev->name->str,
					GFP_KERNEL);
			if (!name) {
				btrfs_free_device(device);
				ret = -ENOMEM;
				goto error;
			}
			rcu_assign_pointer(device->name, name);
		}

		list_add(&device->dev_list, &fs_devices->devices);
		device->fs_devices = fs_devices;
		fs_devices->num_devices++;
	}
	return fs_devices;
error:
	free_fs_devices(fs_devices);
	return ERR_PTR(ret);
}

static void __btrfs_free_extra_devids(struct btrfs_fs_devices *fs_devices,
				      struct btrfs_device **latest_dev)
{
	struct btrfs_device *device, *next;

	/* This is the initialized path, it is safe to release the devices. */
	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
		if (test_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state)) {
			if (!test_bit(BTRFS_DEV_STATE_REPLACE_TGT,
				      &device->dev_state) &&
			    !test_bit(BTRFS_DEV_STATE_MISSING,
				      &device->dev_state) &&
			    (!*latest_dev ||
			     device->generation > (*latest_dev)->generation)) {
				*latest_dev = device;
			}
			continue;
		}

		/*
		 * We have already validated the presence of BTRFS_DEV_REPLACE_DEVID,
		 * in btrfs_init_dev_replace() so just continue.
		 */
		if (device->devid == BTRFS_DEV_REPLACE_DEVID)
			continue;

		if (device->bdev) {
			blkdev_put(device->bdev, device->mode);
			device->bdev = NULL;
			fs_devices->open_devices--;
		}
		if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state)) {
			list_del_init(&device->dev_alloc_list);
			clear_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state);
			fs_devices->rw_devices--;
		}
		list_del_init(&device->dev_list);
		fs_devices->num_devices--;
		btrfs_free_device(device);
	}

}

/*
 * After we have read the system tree and know devids belonging to this
 * filesystem, remove the device which does not belong there.
 */
void btrfs_free_extra_devids(struct btrfs_fs_devices *fs_devices)
{
	struct btrfs_device *latest_dev = NULL;
	struct btrfs_fs_devices *seed_dev;

	mutex_lock(&uuid_mutex);
	__btrfs_free_extra_devids(fs_devices, &latest_dev);

	list_for_each_entry(seed_dev, &fs_devices->seed_list, seed_list)
		__btrfs_free_extra_devids(seed_dev, &latest_dev);

	fs_devices->latest_dev = latest_dev;

	mutex_unlock(&uuid_mutex);
}

static void btrfs_close_bdev(struct btrfs_device *device)
{
	if (!device->bdev)
		return;

	if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state)) {
		sync_blockdev(device->bdev);
		invalidate_bdev(device->bdev);
	}

	blkdev_put(device->bdev, device->mode);
}

static void btrfs_close_one_device(struct btrfs_device *device)
{
	struct btrfs_fs_devices *fs_devices = device->fs_devices;

	if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state) &&
	    device->devid != BTRFS_DEV_REPLACE_DEVID) {
		list_del_init(&device->dev_alloc_list);
		fs_devices->rw_devices--;
	}

	if (device->devid == BTRFS_DEV_REPLACE_DEVID)
		clear_bit(BTRFS_DEV_STATE_REPLACE_TGT, &device->dev_state);

	if (test_bit(BTRFS_DEV_STATE_MISSING, &device->dev_state)) {
		clear_bit(BTRFS_DEV_STATE_MISSING, &device->dev_state);
		fs_devices->missing_devices--;
	}

	btrfs_close_bdev(device);
	if (device->bdev) {
		fs_devices->open_devices--;
		device->bdev = NULL;
	}
	clear_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state);
	btrfs_destroy_dev_zone_info(device);

	device->fs_info = NULL;
	atomic_set(&device->dev_stats_ccnt, 0);
	extent_io_tree_release(&device->alloc_state);

	/*
	 * Reset the flush error record. We might have a transient flush error
	 * in this mount, and if so we aborted the current transaction and set
	 * the fs to an error state, guaranteeing no super blocks can be further
	 * committed. However that error might be transient and if we unmount the
	 * filesystem and mount it again, we should allow the mount to succeed
	 * (btrfs_check_rw_degradable() should not fail) - if after mounting the
	 * filesystem again we still get flush errors, then we will again abort
	 * any transaction and set the error state, guaranteeing no commits of
	 * unsafe super blocks.
	 */
	device->last_flush_error = 0;

	/* Verify the device is back in a pristine state  */
	ASSERT(!test_bit(BTRFS_DEV_STATE_FLUSH_SENT, &device->dev_state));
	ASSERT(!test_bit(BTRFS_DEV_STATE_REPLACE_TGT, &device->dev_state));
	ASSERT(list_empty(&device->dev_alloc_list));
	ASSERT(list_empty(&device->post_commit_list));
	ASSERT(atomic_read(&device->reada_in_flight) == 0);
}

static void close_fs_devices(struct btrfs_fs_devices *fs_devices)
{
	struct btrfs_device *device, *tmp;

	lockdep_assert_held(&uuid_mutex);

	if (--fs_devices->opened > 0)
		return;

	list_for_each_entry_safe(device, tmp, &fs_devices->devices, dev_list)
		btrfs_close_one_device(device);

	WARN_ON(fs_devices->open_devices);
	WARN_ON(fs_devices->rw_devices);
	fs_devices->opened = 0;
	fs_devices->seeding = false;
	fs_devices->fs_info = NULL;
}

void btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
{
	LIST_HEAD(list);
	struct btrfs_fs_devices *tmp;

	mutex_lock(&uuid_mutex);
	close_fs_devices(fs_devices);
	if (!fs_devices->opened)
		list_splice_init(&fs_devices->seed_list, &list);

	list_for_each_entry_safe(fs_devices, tmp, &list, seed_list) {
		close_fs_devices(fs_devices);
		list_del(&fs_devices->seed_list);
		free_fs_devices(fs_devices);
	}
	mutex_unlock(&uuid_mutex);
}

static int open_fs_devices(struct btrfs_fs_devices *fs_devices,
				fmode_t flags, void *holder)
{
	struct btrfs_device *device;
	struct btrfs_device *latest_dev = NULL;
	struct btrfs_device *tmp_device;

	flags |= FMODE_EXCL;

	list_for_each_entry_safe(device, tmp_device, &fs_devices->devices,
				 dev_list) {
		int ret;

		ret = btrfs_open_one_device(fs_devices, device, flags, holder);
		if (ret == 0 &&
		    (!latest_dev || device->generation > latest_dev->generation)) {
			latest_dev = device;
		} else if (ret == -ENODATA) {
			fs_devices->num_devices--;
			list_del(&device->dev_list);
			btrfs_free_device(device);
		}
	}
	if (fs_devices->open_devices == 0)
		return -EINVAL;

	fs_devices->opened = 1;
	fs_devices->latest_dev = latest_dev;
	fs_devices->total_rw_bytes = 0;
	fs_devices->chunk_alloc_policy = BTRFS_CHUNK_ALLOC_REGULAR;
	fs_devices->read_policy = BTRFS_READ_POLICY_PID;

	return 0;
}

static int devid_cmp(void *priv, const struct list_head *a,
		     const struct list_head *b)
{
	const struct btrfs_device *dev1, *dev2;

	dev1 = list_entry(a, struct btrfs_device, dev_list);
	dev2 = list_entry(b, struct btrfs_device, dev_list);

	if (dev1->devid < dev2->devid)
		return -1;
	else if (dev1->devid > dev2->devid)
		return 1;
	return 0;
}

int btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
		       fmode_t flags, void *holder)
{
	int ret;

	lockdep_assert_held(&uuid_mutex);
	/*
	 * The device_list_mutex cannot be taken here in case opening the
	 * underlying device takes further locks like open_mutex.
	 *
	 * We also don't need the lock here as this is called during mount and
	 * exclusion is provided by uuid_mutex
	 */

	if (fs_devices->opened) {
		fs_devices->opened++;
		ret = 0;
	} else {
		list_sort(NULL, &fs_devices->devices, devid_cmp);
		ret = open_fs_devices(fs_devices, flags, holder);
	}

	return ret;
}

void btrfs_release_disk_super(struct btrfs_super_block *super)
{
	struct page *page = virt_to_page(super);

	put_page(page);
}

static struct btrfs_super_block *btrfs_read_disk_super(struct block_device *bdev,
						       u64 bytenr, u64 bytenr_orig)
{
	struct btrfs_super_block *disk_super;
	struct page *page;
	void *p;
	pgoff_t index;

	/* make sure our super fits in the device */
	if (bytenr + PAGE_SIZE >= bdev_nr_bytes(bdev))
		return ERR_PTR(-EINVAL);

	/* make sure our super fits in the page */
	if (sizeof(*disk_super) > PAGE_SIZE)
		return ERR_PTR(-EINVAL);

	/* make sure our super doesn't straddle pages on disk */
	index = bytenr >> PAGE_SHIFT;
	if ((bytenr + sizeof(*disk_super) - 1) >> PAGE_SHIFT != index)
		return ERR_PTR(-EINVAL);

	/* pull in the page with our super */
	page = read_cache_page_gfp(bdev->bd_inode->i_mapping, index, GFP_KERNEL);

	if (IS_ERR(page))
		return ERR_CAST(page);

	p = page_address(page);

	/* align our pointer to the offset of the super block */
	disk_super = p + offset_in_page(bytenr);

	if (btrfs_super_bytenr(disk_super) != bytenr_orig ||
	    btrfs_super_magic(disk_super) != BTRFS_MAGIC) {
		btrfs_release_disk_super(p);
		return ERR_PTR(-EINVAL);
	}

	if (disk_super->label[0] && disk_super->label[BTRFS_LABEL_SIZE - 1])
		disk_super->label[BTRFS_LABEL_SIZE - 1] = 0;

	return disk_super;
}

int btrfs_forget_devices(const char *path)
{
	int ret;

	mutex_lock(&uuid_mutex);
	ret = btrfs_free_stale_devices(strlen(path) ? path : NULL, NULL);
	mutex_unlock(&uuid_mutex);

	return ret;
}

/*
 * Look for a btrfs signature on a device. This may be called out of the mount path
 * and we are not allowed to call set_blocksize during the scan. The superblock
 * is read via pagecache
 */
struct btrfs_device *btrfs_scan_one_device(const char *path, fmode_t flags,
					   void *holder)
{
	struct btrfs_super_block *disk_super;
	bool new_device_added = false;
	struct btrfs_device *device = NULL;
	struct block_device *bdev;
	u64 bytenr, bytenr_orig;
	int ret;

	lockdep_assert_held(&uuid_mutex);

	/*
	 * we would like to check all the supers, but that would make
	 * a btrfs mount succeed after a mkfs from a different FS.
	 * So, we need to add a special mount option to scan for
	 * later supers, using BTRFS_SUPER_MIRROR_MAX instead
	 */
	flags |= FMODE_EXCL;

	bdev = blkdev_get_by_path(path, flags, holder);
	if (IS_ERR(bdev))
		return ERR_CAST(bdev);

	bytenr_orig = btrfs_sb_offset(0);
	ret = btrfs_sb_log_location_bdev(bdev, 0, READ, &bytenr);
	if (ret)
		return ERR_PTR(ret);

	disk_super = btrfs_read_disk_super(bdev, bytenr, bytenr_orig);
	if (IS_ERR(disk_super)) {
		device = ERR_CAST(disk_super);
		goto error_bdev_put;
	}

	device = device_list_add(path, disk_super, &new_device_added);
	if (!IS_ERR(device)) {
		if (new_device_added)
			btrfs_free_stale_devices(path, device);
	}

	btrfs_release_disk_super(disk_super);

error_bdev_put:
	blkdev_put(bdev, flags);

	return device;
}

/*
 * Try to find a chunk that intersects [start, start + len] range and when one
 * such is found, record the end of it in *start
 */
static bool contains_pending_extent(struct btrfs_device *device, u64 *start,
				    u64 len)
{
	u64 physical_start, physical_end;

	lockdep_assert_held(&device->fs_info->chunk_mutex);

	if (!find_first_extent_bit(&device->alloc_state, *start,
				   &physical_start, &physical_end,
				   CHUNK_ALLOCATED, NULL)) {

		if (in_range(physical_start, *start, len) ||
		    in_range(*start, physical_start,
			     physical_end - physical_start)) {
			*start = physical_end + 1;
			return true;
		}
	}
	return false;
}

static u64 dev_extent_search_start(struct btrfs_device *device, u64 start)
{
	switch (device->fs_devices->chunk_alloc_policy) {
	case BTRFS_CHUNK_ALLOC_REGULAR:
		/*
		 * We don't want to overwrite the superblock on the drive nor
		 * any area used by the boot loader (grub for example), so we
		 * make sure to start at an offset of at least 1MB.
		 */
		return max_t(u64, start, SZ_1M);
	case BTRFS_CHUNK_ALLOC_ZONED:
		/*
		 * We don't care about the starting region like regular
		 * allocator, because we anyway use/reserve the first two zones
		 * for superblock logging.
		 */
		return ALIGN(start, device->zone_info->zone_size);
	default:
		BUG();
	}
}

static bool dev_extent_hole_check_zoned(struct btrfs_device *device,
					u64 *hole_start, u64 *hole_size,
					u64 num_bytes)
{
	u64 zone_size = device->zone_info->zone_size;
	u64 pos;
	int ret;
	bool changed = false;

	ASSERT(IS_ALIGNED(*hole_start, zone_size));

	while (*hole_size > 0) {
		pos = btrfs_find_allocatable_zones(device, *hole_start,
						   *hole_start + *hole_size,
						   num_bytes);
		if (pos != *hole_start) {
			*hole_size = *hole_start + *hole_size - pos;
			*hole_start = pos;
			changed = true;
			if (*hole_size < num_bytes)
				break;
		}

		ret = btrfs_ensure_empty_zones(device, pos, num_bytes);

		/* Range is ensured to be empty */
		if (!ret)
			return changed;

		/* Given hole range was invalid (outside of device) */
		if (ret == -ERANGE) {
			*hole_start += *hole_size;
			*hole_size = 0;
			return true;
		}

		*hole_start += zone_size;
		*hole_size -= zone_size;
		changed = true;
	}

	return changed;
}

/**
 * dev_extent_hole_check - check if specified hole is suitable for allocation
 * @device:	the device which we have the hole
 * @hole_start: starting position of the hole
 * @hole_size:	the size of the hole
 * @num_bytes:	the size of the free space that we need
 *
 * This function may modify @hole_start and @hole_size to reflect the suitable
 * position for allocation. Returns 1 if hole position is updated, 0 otherwise.
 */
static bool dev_extent_hole_check(struct btrfs_device *device, u64 *hole_start,
				  u64 *hole_size, u64 num_bytes)
{
	bool changed = false;
	u64 hole_end = *hole_start + *hole_size;

	for (;;) {
		/*
		 * Check before we set max_hole_start, otherwise we could end up
		 * sending back this offset anyway.
		 */
		if (contains_pending_extent(device, hole_start, *hole_size)) {
			if (hole_end >= *hole_start)
				*hole_size = hole_end - *hole_start;
			else
				*hole_size = 0;
			changed = true;
		}

		switch (device->fs_devices->chunk_alloc_policy) {
		case BTRFS_CHUNK_ALLOC_REGULAR:
			/* No extra check */
			break;
		case BTRFS_CHUNK_ALLOC_ZONED:
			if (dev_extent_hole_check_zoned(device, hole_start,
							hole_size, num_bytes)) {
				changed = true;
				/*
				 * The changed hole can contain pending extent.
				 * Loop again to check that.
				 */
				continue;
			}
			break;
		default:
			BUG();
		}

		break;
	}

	return changed;
}

/*
 * find_free_dev_extent_start - find free space in the specified device
 * @device:	  the device which we search the free space in
 * @num_bytes:	  the size of the free space that we need
 * @search_start: the position from which to begin the search
 * @start:	  store the start of the free space.
 * @len:	  the size of the free space. that we find, or the size
 *		  of the max free space if we don't find suitable free space
 *
 * this uses a pretty simple search, the expectation is that it is
 * called very infrequently and that a given device has a small number
 * of extents
 *
 * @start is used to store the start of the free space if we find. But if we
 * don't find suitable free space, it will be used to store the start position
 * of the max free space.
 *
 * @len is used to store the size of the free space that we find.
 * But if we don't find suitable free space, it is used to store the size of
 * the max free space.
 *
 * NOTE: This function will search *commit* root of device tree, and does extra
 * check to ensure dev extents are not double allocated.
 * This makes the function safe to allocate dev extents but may not report
 * correct usable device space, as device extent freed in current transaction
 * is not reported as available.
 */
static int find_free_dev_extent_start(struct btrfs_device *device,
				u64 num_bytes, u64 search_start, u64 *start,
				u64 *len)
{
	struct btrfs_fs_info *fs_info = device->fs_info;
	struct btrfs_root *root = fs_info->dev_root;
	struct btrfs_key key;
	struct btrfs_dev_extent *dev_extent;
	struct btrfs_path *path;
	u64 hole_size;
	u64 max_hole_start;
	u64 max_hole_size;
	u64 extent_end;
	u64 search_end = device->total_bytes;
	int ret;
	int slot;
	struct extent_buffer *l;

	search_start = dev_extent_search_start(device, search_start);

	WARN_ON(device->zone_info &&
		!IS_ALIGNED(num_bytes, device->zone_info->zone_size));

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	max_hole_start = search_start;
	max_hole_size = 0;

again:
	if (search_start >= search_end ||
		test_bit(BTRFS_DEV_STATE_REPLACE_TGT, &device->dev_state)) {
		ret = -ENOSPC;
		goto out;
	}

	path->reada = READA_FORWARD;
	path->search_commit_root = 1;
	path->skip_locking = 1;

	key.objectid = device->devid;
	key.offset = search_start;
	key.type = BTRFS_DEV_EXTENT_KEY;

	ret = btrfs_search_backwards(root, &key, path);
	if (ret < 0)
		goto out;

	while (1) {
		l = path->nodes[0];
		slot = path->slots[0];
		if (slot >= btrfs_header_nritems(l)) {
			ret = btrfs_next_leaf(root, path);
			if (ret == 0)
				continue;
			if (ret < 0)
				goto out;

			break;
		}
		btrfs_item_key_to_cpu(l, &key, slot);

		if (key.objectid < device->devid)
			goto next;

		if (key.objectid > device->devid)
			break;

		if (key.type != BTRFS_DEV_EXTENT_KEY)
			goto next;

		if (key.offset > search_start) {
			hole_size = key.offset - search_start;
			dev_extent_hole_check(device, &search_start, &hole_size,
					      num_bytes);

			if (hole_size > max_hole_size) {
				max_hole_start = search_start;
				max_hole_size = hole_size;
			}

			/*
			 * If this free space is greater than which we need,
			 * it must be the max free space that we have found
			 * until now, so max_hole_start must point to the start
			 * of this free space and the length of this free space
			 * is stored in max_hole_size. Thus, we return
			 * max_hole_start and max_hole_size and go back to the
			 * caller.
			 */
			if (hole_size >= num_bytes) {
				ret = 0;
				goto out;
			}
		}

		dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
		extent_end = key.offset + btrfs_dev_extent_length(l,
								  dev_extent);
		if (extent_end > search_start)
			search_start = extent_end;
next:
		path->slots[0]++;
		cond_resched();
	}

	/*
	 * At this point, search_start should be the end of
	 * allocated dev extents, and when shrinking the device,
	 * search_end may be smaller than search_start.
	 */
	if (search_end > search_start) {
		hole_size = search_end - search_start;
		if (dev_extent_hole_check(device, &search_start, &hole_size,
					  num_bytes)) {
			btrfs_release_path(path);
			goto again;
		}

		if (hole_size > max_hole_size) {
			max_hole_start = search_start;
			max_hole_size = hole_size;
		}
	}

	/* See above. */
	if (max_hole_size < num_bytes)
		ret = -ENOSPC;
	else
		ret = 0;

out:
	btrfs_free_path(path);
	*start = max_hole_start;
	if (len)
		*len = max_hole_size;
	return ret;
}

int find_free_dev_extent(struct btrfs_device *device, u64 num_bytes,
			 u64 *start, u64 *len)
{
	/* FIXME use last free of some kind */
	return find_free_dev_extent_start(device, num_bytes, 0, start, len);
}

static int btrfs_free_dev_extent(struct btrfs_trans_handle *trans,
			  struct btrfs_device *device,
			  u64 start, u64 *dev_extent_len)
{
	struct btrfs_fs_info *fs_info = device->fs_info;
	struct btrfs_root *root = fs_info->dev_root;
	int ret;
	struct btrfs_path *path;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct extent_buffer *leaf = NULL;
	struct btrfs_dev_extent *extent = NULL;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	key.objectid = device->devid;
	key.offset = start;
	key.type = BTRFS_DEV_EXTENT_KEY;
again:
	ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
	if (ret > 0) {
		ret = btrfs_previous_item(root, path, key.objectid,
					  BTRFS_DEV_EXTENT_KEY);
		if (ret)
			goto out;
		leaf = path->nodes[0];
		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
		extent = btrfs_item_ptr(leaf, path->slots[0],
					struct btrfs_dev_extent);
		BUG_ON(found_key.offset > start || found_key.offset +
		       btrfs_dev_extent_length(leaf, extent) < start);
		key = found_key;
		btrfs_release_path(path);
		goto again;
	} else if (ret == 0) {
		leaf = path->nodes[0];
		extent = btrfs_item_ptr(leaf, path->slots[0],
					struct btrfs_dev_extent);
	} else {
		goto out;
	}

	*dev_extent_len = btrfs_dev_extent_length(leaf, extent);

	ret = btrfs_del_item(trans, root, path);
	if (ret == 0)
		set_bit(BTRFS_TRANS_HAVE_FREE_BGS, &trans->transaction->flags);
out:
	btrfs_free_path(path);
	return ret;
}

static u64 find_next_chunk(struct btrfs_fs_info *fs_info)
{
	struct extent_map_tree *em_tree;
	struct extent_map *em;
	struct rb_node *n;
	u64 ret = 0;

	em_tree = &fs_info->mapping_tree;
	read_lock(&em_tree->lock);
	n = rb_last(&em_tree->map.rb_root);
	if (n) {
		em = rb_entry(n, struct extent_map, rb_node);
		ret = em->start + em->len;
	}
	read_unlock(&em_tree->lock);

	return ret;
}

static noinline int find_next_devid(struct btrfs_fs_info *fs_info,
				    u64 *devid_ret)
{
	int ret;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct btrfs_path *path;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
	key.type = BTRFS_DEV_ITEM_KEY;
	key.offset = (u64)-1;

	ret = btrfs_search_slot(NULL, fs_info->chunk_root, &key, path, 0, 0);
	if (ret < 0)
		goto error;

	if (ret == 0) {
		/* Corruption */
		btrfs_err(fs_info, "corrupted chunk tree devid -1 matched");
		ret = -EUCLEAN;
		goto error;
	}

	ret = btrfs_previous_item(fs_info->chunk_root, path,
				  BTRFS_DEV_ITEMS_OBJECTID,
				  BTRFS_DEV_ITEM_KEY);
	if (ret) {
		*devid_ret = 1;
	} else {
		btrfs_item_key_to_cpu(path->nodes[0], &found_key,
				      path->slots[0]);
		*devid_ret = found_key.offset + 1;
	}
	ret = 0;
error:
	btrfs_free_path(path);
	return ret;
}

/*
 * the device information is stored in the chunk root
 * the btrfs_device struct should be fully filled in
 */
static int btrfs_add_dev_item(struct btrfs_trans_handle *trans,
			    struct btrfs_device *device)
{
	int ret;
	struct btrfs_path *path;
	struct btrfs_dev_item *dev_item;
	struct extent_buffer *leaf;
	struct btrfs_key key;
	unsigned long ptr;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
	key.type = BTRFS_DEV_ITEM_KEY;
	key.offset = device->devid;

	btrfs_reserve_chunk_metadata(trans, true);
	ret = btrfs_insert_empty_item(trans, trans->fs_info->chunk_root, path,
				      &key, sizeof(*dev_item));
	btrfs_trans_release_chunk_metadata(trans);
	if (ret)
		goto out;

	leaf = path->nodes[0];
	dev_item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_dev_item);

	btrfs_set_device_id(leaf, dev_item, device->devid);
	btrfs_set_device_generation(leaf, dev_item, 0);
	btrfs_set_device_type(leaf, dev_item, device->type);
	btrfs_set_device_io_align(leaf, dev_item, device->io_align);
	btrfs_set_device_io_width(leaf, dev_item, device->io_width);
	btrfs_set_device_sector_size(leaf, dev_item, device->sector_size);
	btrfs_set_device_total_bytes(leaf, dev_item,
				     btrfs_device_get_disk_total_bytes(device));
	btrfs_set_device_bytes_used(leaf, dev_item,
				    btrfs_device_get_bytes_used(device));
	btrfs_set_device_group(leaf, dev_item, 0);
	btrfs_set_device_seek_speed(leaf, dev_item, 0);
	btrfs_set_device_bandwidth(leaf, dev_item, 0);
	btrfs_set_device_start_offset(leaf, dev_item, 0);

	ptr = btrfs_device_uuid(dev_item);
	write_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
	ptr = btrfs_device_fsid(dev_item);
	write_extent_buffer(leaf, trans->fs_info->fs_devices->metadata_uuid,
			    ptr, BTRFS_FSID_SIZE);
	btrfs_mark_buffer_dirty(leaf);

	ret = 0;
out:
	btrfs_free_path(path);
	return ret;
}

/*
 * Function to update ctime/mtime for a given device path.
 * Mainly used for ctime/mtime based probe like libblkid.
 *
 * We don't care about errors here, this is just to be kind to userspace.
 */
static void update_dev_time(const char *device_path)
{
	struct path path;
	struct timespec64 now;
	int ret;

	ret = kern_path(device_path, LOOKUP_FOLLOW, &path);
	if (ret)
		return;

	now = current_time(d_inode(path.dentry));
	inode_update_time(d_inode(path.dentry), &now, S_MTIME | S_CTIME);
	path_put(&path);
}

static int btrfs_rm_dev_item(struct btrfs_device *device)
{
	struct btrfs_root *root = device->fs_info->chunk_root;
	int ret;
	struct btrfs_path *path;
	struct btrfs_key key;
	struct btrfs_trans_handle *trans;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		btrfs_free_path(path);
		return PTR_ERR(trans);
	}
	key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
	key.type = BTRFS_DEV_ITEM_KEY;
	key.offset = device->devid;

	btrfs_reserve_chunk_metadata(trans, false);
	ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
	btrfs_trans_release_chunk_metadata(trans);
	if (ret) {
		if (ret > 0)
			ret = -ENOENT;
		btrfs_abort_transaction(trans, ret);
		btrfs_end_transaction(trans);
		goto out;
	}

	ret = btrfs_del_item(trans, root, path);
	if (ret) {
		btrfs_abort_transaction(trans, ret);
		btrfs_end_transaction(trans);
	}

out:
	btrfs_free_path(path);
	if (!ret)
		ret = btrfs_commit_transaction(trans);
	return ret;
}

/*
 * Verify that @num_devices satisfies the RAID profile constraints in the whole
 * filesystem. It's up to the caller to adjust that number regarding eg. device
 * replace.
 */
static int btrfs_check_raid_min_devices(struct btrfs_fs_info *fs_info,
		u64 num_devices)
{
	u64 all_avail;
	unsigned seq;
	int i;

	do {
		seq = read_seqbegin(&fs_info->profiles_lock);

		all_avail = fs_info->avail_data_alloc_bits |
			    fs_info->avail_system_alloc_bits |
			    fs_info->avail_metadata_alloc_bits;
	} while (read_seqretry(&fs_info->profiles_lock, seq));

	for (i = 0; i < BTRFS_NR_RAID_TYPES; i++) {
		if (!(all_avail & btrfs_raid_array[i].bg_flag))
			continue;

		if (num_devices < btrfs_raid_array[i].devs_min)
			return btrfs_raid_array[i].mindev_error;
	}

	return 0;
}

static struct btrfs_device * btrfs_find_next_active_device(
		struct btrfs_fs_devices *fs_devs, struct btrfs_device *device)
{
	struct btrfs_device *next_device;

	list_for_each_entry(next_device, &fs_devs->devices, dev_list) {
		if (next_device != device &&
		    !test_bit(BTRFS_DEV_STATE_MISSING, &next_device->dev_state)
		    && next_device->bdev)
			return next_device;
	}

	return NULL;
}

/*
 * Helper function to check if the given device is part of s_bdev / latest_dev
 * and replace it with the provided or the next active device, in the context
 * where this function called, there should be always be another device (or
 * this_dev) which is active.
 */
void __cold btrfs_assign_next_active_device(struct btrfs_device *device,
					    struct btrfs_device *next_device)
{
	struct btrfs_fs_info *fs_info = device->fs_info;

	if (!next_device)
		next_device = btrfs_find_next_active_device(fs_info->fs_devices,
							    device);
	ASSERT(next_device);

	if (fs_info->sb->s_bdev &&
			(fs_info->sb->s_bdev == device->bdev))
		fs_info->sb->s_bdev = next_device->bdev;

	if (fs_info->fs_devices->latest_dev->bdev == device->bdev)
		fs_info->fs_devices->latest_dev = next_device;
}

/*
 * Return btrfs_fs_devices::num_devices excluding the device that's being
 * currently replaced.
 */
static u64 btrfs_num_devices(struct btrfs_fs_info *fs_info)
{
	u64 num_devices = fs_info->fs_devices->num_devices;

	down_read(&fs_info->dev_replace.rwsem);
	if (btrfs_dev_replace_is_ongoing(&fs_info->dev_replace)) {
		ASSERT(num_devices > 1);
		num_devices--;
	}
	up_read(&fs_info->dev_replace.rwsem);

	return num_devices;
}

void btrfs_scratch_superblocks(struct btrfs_fs_info *fs_info,
			       struct block_device *bdev,
			       const char *device_path)
{
	struct btrfs_super_block *disk_super;
	int copy_num;

	if (!bdev)
		return;

	for (copy_num = 0; copy_num < BTRFS_SUPER_MIRROR_MAX; copy_num++) {
		struct page *page;
		int ret;

		disk_super = btrfs_read_dev_one_super(bdev, copy_num);
		if (IS_ERR(disk_super))
			continue;

		if (bdev_is_zoned(bdev)) {
			btrfs_reset_sb_log_zones(bdev, copy_num);
			continue;
		}

		memset(&disk_super->magic, 0, sizeof(disk_super->magic));

		page = virt_to_page(disk_super);
		set_page_dirty(page);
		lock_page(page);
		/* write_on_page() unlocks the page */
		ret = write_one_page(page);
		if (ret)
			btrfs_warn(fs_info,
				"error clearing superblock number %d (%d)",
				copy_num, ret);
		btrfs_release_disk_super(disk_super);

	}

	/* Notify udev that device has changed */
	btrfs_kobject_uevent(bdev, KOBJ_CHANGE);

	/* Update ctime/mtime for device path for libblkid */
	update_dev_time(device_path);
}

int btrfs_rm_device(struct btrfs_fs_info *fs_info,
		    struct btrfs_dev_lookup_args *args,
		    struct block_device **bdev, fmode_t *mode)
{
	struct btrfs_device *device;
	struct btrfs_fs_devices *cur_devices;
	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
	u64 num_devices;
	int ret = 0;

	/*
	 * The device list in fs_devices is accessed without locks (neither
	 * uuid_mutex nor device_list_mutex) as it won't change on a mounted
	 * filesystem and another device rm cannot run.
	 */
	num_devices = btrfs_num_devices(fs_info);

	ret = btrfs_check_raid_min_devices(fs_info, num_devices - 1);
	if (ret)
		goto out;

	device = btrfs_find_device(fs_info->fs_devices, args);
	if (!device) {
		if (args->missing)
			ret = BTRFS_ERROR_DEV_MISSING_NOT_FOUND;
		else
			ret = -ENOENT;
		goto out;
	}

	if (btrfs_pinned_by_swapfile(fs_info, device)) {
		btrfs_warn_in_rcu(fs_info,
		  "cannot remove device %s (devid %llu) due to active swapfile",
				  rcu_str_deref(device->name), device->devid);
		ret = -ETXTBSY;
		goto out;
	}

	if (test_bit(BTRFS_DEV_STATE_REPLACE_TGT, &device->dev_state)) {
		ret = BTRFS_ERROR_DEV_TGT_REPLACE;
		goto out;
	}

	if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state) &&
	    fs_info->fs_devices->rw_devices == 1) {
		ret = BTRFS_ERROR_DEV_ONLY_WRITABLE;
		goto out;
	}

	if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state)) {
		mutex_lock(&fs_info->chunk_mutex);
		list_del_init(&device->dev_alloc_list);
		device->fs_devices->rw_devices--;
		mutex_unlock(&fs_info->chunk_mutex);
	}

	ret = btrfs_shrink_device(device, 0);
	if (!ret)
		btrfs_reada_remove_dev(device);
	if (ret)
		goto error_undo;

	/*
	 * TODO: the superblock still includes this device in its num_devices
	 * counter although write_all_supers() is not locked out. This
	 * could give a filesystem state which requires a degraded mount.
	 */
	ret = btrfs_rm_dev_item(device);
	if (ret)
		goto error_undo;

	clear_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state);
	btrfs_scrub_cancel_dev(device);

	/*
	 * the device list mutex makes sure that we don't change
	 * the device list while someone else is writing out all
	 * the device supers. Whoever is writing all supers, should
	 * lock the device list mutex before getting the number of
	 * devices in the super block (super_copy). Conversely,
	 * whoever updates the number of devices in the super block
	 * (super_copy) should hold the device list mutex.
	 */

	/*
	 * In normal cases the cur_devices == fs_devices. But in case
	 * of deleting a seed device, the cur_devices should point to
	 * its own fs_devices listed under the fs_devices->seed_list.
	 */
	cur_devices = device->fs_devices;
	mutex_lock(&fs_devices->device_list_mutex);
	list_del_rcu(&device->dev_list);

	cur_devices->num_devices--;
	cur_devices->total_devices--;
	/* Update total_devices of the parent fs_devices if it's seed */
	if (cur_devices != fs_devices)
		fs_devices->total_devices--;

	if (test_bit(BTRFS_DEV_STATE_MISSING, &device->dev_state))
		cur_devices->missing_devices--;

	btrfs_assign_next_active_device(device, NULL);

	if (device->bdev) {
		cur_devices->open_devices--;
		/* remove sysfs entry */
		btrfs_sysfs_remove_device(device);
	}

	num_devices = btrfs_super_num_devices(fs_info->super_copy) - 1;
	btrfs_set_super_num_devices(fs_info->super_copy, num_devices);
	mutex_unlock(&fs_devices->device_list_mutex);

	/*
	 * At this point, the device is zero sized and detached from the
	 * devices list.  All that's left is to zero out the old supers and
	 * free the device.
	 *
	 * We cannot call btrfs_close_bdev() here because we're holding the sb
	 * write lock, and blkdev_put() will pull in the ->open_mutex on the
	 * block device and it's dependencies.  Instead just flush the device
	 * and let the caller do the final blkdev_put.
	 */
	if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state)) {
		btrfs_scratch_superblocks(fs_info, device->bdev,
					  device->name->str);
		if (device->bdev) {
			sync_blockdev(device->bdev);
			invalidate_bdev(device->bdev);
		}
	}

	*bdev = device->bdev;
	*mode = device->mode;
	synchronize_rcu();
	btrfs_free_device(device);

	/*
	 * This can happen if cur_devices is the private seed devices list.  We
	 * cannot call close_fs_devices() here because it expects the uuid_mutex
	 * to be held, but in fact we don't need that for the private
	 * seed_devices, we can simply decrement cur_devices->opened and then
	 * remove it from our list and free the fs_devices.
	 */
	if (cur_devices->num_devices == 0) {
		list_del_init(&cur_devices->seed_list);
		ASSERT(cur_devices->opened == 1);
		cur_devices->opened--;
		free_fs_devices(cur_devices);
	}

out:
	return ret;

error_undo:
	btrfs_reada_undo_remove_dev(device);
	if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state)) {
		mutex_lock(&fs_info->chunk_mutex);
		list_add(&device->dev_alloc_list,
			 &fs_devices->alloc_list);
		device->fs_devices->rw_devices++;
		mutex_unlock(&fs_info->chunk_mutex);
	}
	goto out;
}

void btrfs_rm_dev_replace_remove_srcdev(struct btrfs_device *srcdev)
{
	struct btrfs_fs_devices *fs_devices;

	lockdep_assert_held(&srcdev->fs_info->fs_devices->device_list_mutex);

	/*
	 * in case of fs with no seed, srcdev->fs_devices will point
	 * to fs_devices of fs_info. However when the dev being replaced is
	 * a seed dev it will point to the seed's local fs_devices. In short
	 * srcdev will have its correct fs_devices in both the cases.
	 */
	fs_devices = srcdev->fs_devices;

	list_del_rcu(&srcdev->dev_list);
	list_del(&srcdev->dev_alloc_list);
	fs_devices->num_devices--;
	if (test_bit(BTRFS_DEV_STATE_MISSING, &srcdev->dev_state))
		fs_devices->missing_devices--;

	if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &srcdev->dev_state))
		fs_devices->rw_devices--;

	if (srcdev->bdev)
		fs_devices->open_devices--;
}

void btrfs_rm_dev_replace_free_srcdev(struct btrfs_device *srcdev)
{
	struct btrfs_fs_devices *fs_devices = srcdev->fs_devices;

	mutex_lock(&uuid_mutex);

	btrfs_close_bdev(srcdev);
	synchronize_rcu();
	btrfs_free_device(srcdev);

	/* if this is no devs we rather delete the fs_devices */
	if (!fs_devices->num_devices) {
		/*
		 * On a mounted FS, num_devices can't be zero unless it's a
		 * seed. In case of a seed device being replaced, the replace
		 * target added to the sprout FS, so there will be no more
		 * device left under the seed FS.
		 */
		ASSERT(fs_devices->seeding);

		list_del_init(&fs_devices->seed_list);
		close_fs_devices(fs_devices);
		free_fs_devices(fs_devices);
	}
	mutex_unlock(&uuid_mutex);
}

void btrfs_destroy_dev_replace_tgtdev(struct btrfs_device *tgtdev)
{
	struct btrfs_fs_devices *fs_devices = tgtdev->fs_info->fs_devices;

	mutex_lock(&fs_devices->device_list_mutex);

	btrfs_sysfs_remove_device(tgtdev);

	if (tgtdev->bdev)
		fs_devices->open_devices--;

	fs_devices->num_devices--;

	btrfs_assign_next_active_device(tgtdev, NULL);

	list_del_rcu(&tgtdev->dev_list);

	mutex_unlock(&fs_devices->device_list_mutex);

	btrfs_scratch_superblocks(tgtdev->fs_info, tgtdev->bdev,
				  tgtdev->name->str);

	btrfs_close_bdev(tgtdev);
	synchronize_rcu();
	btrfs_free_device(tgtdev);
}

/**
 * Populate args from device at path
 *
 * @fs_info:	the filesystem
 * @args:	the args to populate
 * @path:	the path to the device
 *
 * This will read the super block of the device at @path and populate @args with
 * the devid, fsid, and uuid.  This is meant to be used for ioctls that need to
 * lookup a device to operate on, but need to do it before we take any locks.
 * This properly handles the special case of "missing" that a user may pass in,
 * and does some basic sanity checks.  The caller must make sure that @path is
 * properly NUL terminated before calling in, and must call
 * btrfs_put_dev_args_from_path() in order to free up the temporary fsid and
 * uuid buffers.
 *
 * Return: 0 for success, -errno for failure
 */
int btrfs_get_dev_args_from_path(struct btrfs_fs_info *fs_info,
				 struct btrfs_dev_lookup_args *args,
				 const char *path)
{
	struct btrfs_super_block *disk_super;
	struct block_device *bdev;
	int ret;

	if (!path || !path[0])
		return -EINVAL;
	if (!strcmp(path, "missing")) {
		args->missing = true;
		return 0;
	}

	args->uuid = kzalloc(BTRFS_UUID_SIZE, GFP_KERNEL);
	args->fsid = kzalloc(BTRFS_FSID_SIZE, GFP_KERNEL);
	if (!args->uuid || !args->fsid) {
		btrfs_put_dev_args_from_path(args);
		return -ENOMEM;
	}

	ret = btrfs_get_bdev_and_sb(path, FMODE_READ, fs_info->bdev_holder, 0,
				    &bdev, &disk_super);
	if (ret)
		return ret;
	args->devid = btrfs_stack_device_id(&disk_super->dev_item);
	memcpy(args->uuid, disk_super->dev_item.uuid, BTRFS_UUID_SIZE);
	if (btrfs_fs_incompat(fs_info, METADATA_UUID))
		memcpy(args->fsid, disk_super->metadata_uuid, BTRFS_FSID_SIZE);
	else
		memcpy(args->fsid, disk_super->fsid, BTRFS_FSID_SIZE);
	btrfs_release_disk_super(disk_super);
	blkdev_put(bdev, FMODE_READ);
	return 0;
}

/*
 * Only use this jointly with btrfs_get_dev_args_from_path() because we will
 * allocate our ->uuid and ->fsid pointers, everybody else uses local variables
 * that don't need to be freed.
 */
void btrfs_put_dev_args_from_path(struct btrfs_dev_lookup_args *args)
{
	kfree(args->uuid);
	kfree(args->fsid);
	args->uuid = NULL;
	args->fsid = NULL;
}

struct btrfs_device *btrfs_find_device_by_devspec(
		struct btrfs_fs_info *fs_info, u64 devid,
		const char *device_path)
{
	BTRFS_DEV_LOOKUP_ARGS(args);
	struct btrfs_device *device;
	int ret;

	if (devid) {
		args.devid = devid;
		device = btrfs_find_device(fs_info->fs_devices, &args);
		if (!device)
			return ERR_PTR(-ENOENT);
		return device;
	}

	ret = btrfs_get_dev_args_from_path(fs_info, &args, device_path);
	if (ret)
		return ERR_PTR(ret);
	device = btrfs_find_device(fs_info->fs_devices, &args);
	btrfs_put_dev_args_from_path(&args);
	if (!device)
		return ERR_PTR(-ENOENT);
	return device;
}

/*
 * does all the dirty work required for changing file system's UUID.
 */
static int btrfs_prepare_sprout(struct btrfs_fs_info *fs_info)
{
	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
	struct btrfs_fs_devices *old_devices;
	struct btrfs_fs_devices *seed_devices;
	struct btrfs_super_block *disk_super = fs_info->super_copy;
	struct btrfs_device *device;
	u64 super_flags;

	lockdep_assert_held(&uuid_mutex);
	if (!fs_devices->seeding)
		return -EINVAL;

	/*
	 * Private copy of the seed devices, anchored at
	 * fs_info->fs_devices->seed_list
	 */
	seed_devices = alloc_fs_devices(NULL, NULL);
	if (IS_ERR(seed_devices))
		return PTR_ERR(seed_devices);

	/*
	 * It's necessary to retain a copy of the original seed fs_devices in
	 * fs_uuids so that filesystems which have been seeded can successfully
	 * reference the seed device from open_seed_devices. This also supports
	 * multiple fs seed.
	 */
	old_devices = clone_fs_devices(fs_devices);
	if (IS_ERR(old_devices)) {
		kfree(seed_devices);
		return PTR_ERR(old_devices);
	}

	list_add(&old_devices->fs_list, &fs_uuids);

	memcpy(seed_devices, fs_devices, sizeof(*seed_devices));
	seed_devices->opened = 1;
	INIT_LIST_HEAD(&seed_devices->devices);
	INIT_LIST_HEAD(&seed_devices->alloc_list);
	mutex_init(&seed_devices->device_list_mutex);

	mutex_lock(&fs_devices->device_list_mutex);
	list_splice_init_rcu(&fs_devices->devices, &seed_devices->devices,
			      synchronize_rcu);
	list_for_each_entry(device, &seed_devices->devices, dev_list)
		device->fs_devices = seed_devices;

	fs_devices->seeding = false;
	fs_devices->num_devices = 0;
	fs_devices->open_devices = 0;
	fs_devices->missing_devices = 0;
	fs_devices->rotating = false;
	list_add(&seed_devices->seed_list, &fs_devices->seed_list);

	generate_random_uuid(fs_devices->fsid);
	memcpy(fs_devices->metadata_uuid, fs_devices->fsid, BTRFS_FSID_SIZE);
	memcpy(disk_super->fsid, fs_devices->fsid, BTRFS_FSID_SIZE);
	mutex_unlock(&fs_devices->device_list_mutex);

	super_flags = btrfs_super_flags(disk_super) &
		      ~BTRFS_SUPER_FLAG_SEEDING;
	btrfs_set_super_flags(disk_super, super_flags);

	return 0;
}

/*
 * Store the expected generation for seed devices in device items.
 */
static int btrfs_finish_sprout(struct btrfs_trans_handle *trans)
{
	BTRFS_DEV_LOOKUP_ARGS(args);
	struct btrfs_fs_info *fs_info = trans->fs_info;
	struct btrfs_root *root = fs_info->chunk_root;
	struct btrfs_path *path;
	struct extent_buffer *leaf;
	struct btrfs_dev_item *dev_item;
	struct btrfs_device *device;
	struct btrfs_key key;
	u8 fs_uuid[BTRFS_FSID_SIZE];
	u8 dev_uuid[BTRFS_UUID_SIZE];
	int ret;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
	key.offset = 0;
	key.type = BTRFS_DEV_ITEM_KEY;

	while (1) {
		btrfs_reserve_chunk_metadata(trans, false);
		ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
		btrfs_trans_release_chunk_metadata(trans);
		if (ret < 0)
			goto error;

		leaf = path->nodes[0];
next_slot:
		if (path->slots[0] >= btrfs_header_nritems(leaf)) {
			ret = btrfs_next_leaf(root, path);
			if (ret > 0)
				break;
			if (ret < 0)
				goto error;
			leaf = path->nodes[0];
			btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
			btrfs_release_path(path);
			continue;
		}

		btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
		if (key.objectid != BTRFS_DEV_ITEMS_OBJECTID ||
		    key.type != BTRFS_DEV_ITEM_KEY)
			break;

		dev_item = btrfs_item_ptr(leaf, path->slots[0],
					  struct btrfs_dev_item);
		args.devid = btrfs_device_id(leaf, dev_item);
		read_extent_buffer(leaf, dev_uuid, btrfs_device_uuid(dev_item),
				   BTRFS_UUID_SIZE);
		read_extent_buffer(leaf, fs_uuid, btrfs_device_fsid(dev_item),
				   BTRFS_FSID_SIZE);
		args.uuid = dev_uuid;
		args.fsid = fs_uuid;
		device = btrfs_find_device(fs_info->fs_devices, &args);
		BUG_ON(!device); /* Logic error */

		if (device->fs_devices->seeding) {
			btrfs_set_device_generation(leaf, dev_item,
						    device->generation);
			btrfs_mark_buffer_dirty(leaf);
		}

		path->slots[0]++;
		goto next_slot;
	}
	ret = 0;
error:
	btrfs_free_path(path);
	return ret;
}

int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path)
{
	struct btrfs_root *root = fs_info->dev_root;
	struct request_queue *q;
	struct btrfs_trans_handle *trans;
	struct btrfs_device *device;
	struct block_device *bdev;
	struct super_block *sb = fs_info->sb;
	struct rcu_string *name;
	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
	u64 orig_super_total_bytes;
	u64 orig_super_num_devices;
	int seeding_dev = 0;
	int ret = 0;
	bool locked = false;

	if (sb_rdonly(sb) && !fs_devices->seeding)
		return -EROFS;

	bdev = blkdev_get_by_path(device_path, FMODE_WRITE | FMODE_EXCL,
				  fs_info->bdev_holder);
	if (IS_ERR(bdev))
		return PTR_ERR(bdev);

	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
		ret = -EINVAL;
		goto error;
	}

	if (fs_devices->seeding) {
		seeding_dev = 1;
		down_write(&sb->s_umount);
		mutex_lock(&uuid_mutex);
		locked = true;
	}

	sync_blockdev(bdev);

	rcu_read_lock();
	list_for_each_entry_rcu(device, &fs_devices->devices, dev_list) {
		if (device->bdev == bdev) {
			ret = -EEXIST;
			rcu_read_unlock();
			goto error;
		}
	}
	rcu_read_unlock();

	device = btrfs_alloc_device(fs_info, NULL, NULL);
	if (IS_ERR(device)) {
		/* we can safely leave the fs_devices entry around */
		ret = PTR_ERR(device);
		goto error;
	}

	name = rcu_string_strdup(device_path, GFP_KERNEL);
	if (!name) {
		ret = -ENOMEM;
		goto error_free_device;
	}
	rcu_assign_pointer(device->name, name);

	device->fs_info = fs_info;
	device->bdev = bdev;

	ret = btrfs_get_dev_zone_info(device);
	if (ret)
		goto error_free_device;

	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto error_free_zone;
	}

	q = bdev_get_queue(bdev);
	set_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state);
	device->generation = trans->transid;
	device->io_width = fs_info->sectorsize;
	device->io_align = fs_info->sectorsize;
	device->sector_size = fs_info->sectorsize;
	device->total_bytes =
		round_down(bdev_nr_bytes(bdev), fs_info->sectorsize);
	device->disk_total_bytes = device->total_bytes;
	device->commit_total_bytes = device->total_bytes;
	set_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state);
	clear_bit(BTRFS_DEV_STATE_REPLACE_TGT, &device->dev_state);
	device->mode = FMODE_EXCL;
	device->dev_stats_valid = 1;
	set_blocksize(device->bdev, BTRFS_BDEV_BLOCKSIZE);

	if (seeding_dev) {
		btrfs_clear_sb_rdonly(sb);
		ret = btrfs_prepare_sprout(fs_info);
		if (ret) {
			btrfs_abort_transaction(trans, ret);
			goto error_trans;
		}
		btrfs_assign_next_active_device(fs_info->fs_devices->latest_dev,
						device);
	}

	device->fs_devices = fs_devices;

	mutex_lock(&fs_devices->device_list_mutex);
	mutex_lock(&fs_info->chunk_mutex);
	list_add_rcu(&device->dev_list, &fs_devices->devices);
	list_add(&device->dev_alloc_list, &fs_devices->alloc_list);
	fs_devices->num_devices++;
	fs_devices->open_devices++;
	fs_devices->rw_devices++;
	fs_devices->total_devices++;
	fs_devices->total_rw_bytes += device->total_bytes;

	atomic64_add(device->total_bytes, &fs_info->free_chunk_space);

	if (!blk_queue_nonrot(q))
		fs_devices->rotating = true;

	orig_super_total_bytes = btrfs_super_total_bytes(fs_info->super_copy);
	btrfs_set_super_total_bytes(fs_info->super_copy,
		round_down(orig_super_total_bytes + device->total_bytes,
			   fs_info->sectorsize));

	orig_super_num_devices = btrfs_super_num_devices(fs_info->super_copy);
	btrfs_set_super_num_devices(fs_info->super_copy,
				    orig_super_num_devices + 1);

	/*
	 * we've got more storage, clear any full flags on the space
	 * infos
	 */
	btrfs_clear_space_info_full(fs_info);

	mutex_unlock(&fs_info->chunk_mutex);

	/* Add sysfs device entry */
	btrfs_sysfs_add_device(device);

	mutex_unlock(&fs_devices->device_list_mutex);

	if (seeding_dev) {
		mutex_lock(&fs_info->chunk_mutex);
		ret = init_first_rw_device(trans);
		mutex_unlock(&fs_info->chunk_mutex);
		if (ret) {
			btrfs_abort_transaction(trans, ret);
			goto error_sysfs;
		}
	}

	ret = btrfs_add_dev_item(trans, device);
	if (ret) {
		btrfs_abort_transaction(trans, ret);
		goto error_sysfs;
	}

	if (seeding_dev) {
		ret = btrfs_finish_sprout(trans);
		if (ret) {
			btrfs_abort_transaction(trans, ret);
			goto error_sysfs;
		}

		/*
		 * fs_devices now represents the newly sprouted filesystem and
		 * its fsid has been changed by btrfs_prepare_sprout
		 */
		btrfs_sysfs_update_sprout_fsid(fs_devices);
	}

	ret = btrfs_commit_transaction(trans);

	if (seeding_dev) {
		mutex_unlock(&uuid_mutex);
		up_write(&sb->s_umount);
		locked = false;

		if (ret) /* transaction commit */
			return ret;

		ret = btrfs_relocate_sys_chunks(fs_info);
		if (ret < 0)
			btrfs_handle_fs_error(fs_info, ret,
				    "Failed to relocate sys chunks after device initialization. This can be fixed using the \"btrfs balance\" command.");
		trans = btrfs_attach_transaction(root);
		if (IS_ERR(trans)) {
			if (PTR_ERR(trans) == -ENOENT)
				return 0;
			ret = PTR_ERR(trans);
			trans = NULL;
			goto error_sysfs;
		}
		ret = btrfs_commit_transaction(trans);
	}

	/*
	 * Now that we have written a new super block to this device, check all
	 * other fs_devices list if device_path alienates any other scanned
	 * device.
	 * We can ignore the return value as it typically returns -EINVAL and
	 * only succeeds if the device was an alien.
	 */
	btrfs_forget_devices(device_path);

	/* Update ctime/mtime for blkid or udev */
	update_dev_time(device_path);

	return ret;

error_sysfs:
	btrfs_sysfs_remove_device(device);
	mutex_lock(&fs_info->fs_devices->device_list_mutex);
	mutex_lock(&fs_info->chunk_mutex);
	list_del_rcu(&device->dev_list);
	list_del(&device->dev_alloc_list);
	fs_info->fs_devices->num_devices--;
	fs_info->fs_devices->open_devices--;
	fs_info->fs_devices->rw_devices--;
	fs_info->fs_devices->total_devices--;
	fs_info->fs_devices->total_rw_bytes -= device->total_bytes;
	atomic64_sub(device->total_bytes, &fs_info->free_chunk_space);
	btrfs_set_super_total_bytes(fs_info->super_copy,
				    orig_super_total_bytes);
	btrfs_set_super_num_devices(fs_info->super_copy,
				    orig_super_num_devices);
	mutex_unlock(&fs_info->chunk_mutex);
	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
error_trans:
	if (seeding_dev)
		btrfs_set_sb_rdonly(sb);
	if (trans)
		btrfs_end_transaction(trans);
error_free_zone:
	btrfs_destroy_dev_zone_info(device);
error_free_device:
	btrfs_free_device(device);
error:
	blkdev_put(bdev, FMODE_EXCL);
	if (locked) {
		mutex_unlock(&uuid_mutex);
		up_write(&sb->s_umount);
	}
	return ret;
}

static noinline int btrfs_update_device(struct btrfs_trans_handle *trans,
					struct btrfs_device *device)
{
	int ret;
	struct btrfs_path *path;
	struct btrfs_root *root = device->fs_info->chunk_root;
	struct btrfs_dev_item *dev_item;
	struct extent_buffer *leaf;
	struct btrfs_key key;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
	key.type = BTRFS_DEV_ITEM_KEY;
	key.offset = device->devid;

	ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
	if (ret < 0)
		goto out;

	if (ret > 0) {
		ret = -ENOENT;
		goto out;
	}

	leaf = path->nodes[0];
	dev_item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_dev_item);

	btrfs_set_device_id(leaf, dev_item, device->devid);
	btrfs_set_device_type(leaf, dev_item, device->type);
	btrfs_set_device_io_align(leaf, dev_item, device->io_align);
	btrfs_set_device_io_width(leaf, dev_item, device->io_width);
	btrfs_set_device_sector_size(leaf, dev_item, device->sector_size);
	btrfs_set_device_total_bytes(leaf, dev_item,
				     btrfs_device_get_disk_total_bytes(device));
	btrfs_set_device_bytes_used(leaf, dev_item,
				    btrfs_device_get_bytes_used(device));
	btrfs_mark_buffer_dirty(leaf);

out:
	btrfs_free_path(path);
	return ret;
}

int btrfs_grow_device(struct btrfs_trans_handle *trans,
		      struct btrfs_device *device, u64 new_size)
{
	struct btrfs_fs_info *fs_info = device->fs_info;
	struct btrfs_super_block *super_copy = fs_info->super_copy;
	u64 old_total;
	u64 diff;
	int ret;

	if (!test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state))
		return -EACCES;

	new_size = round_down(new_size, fs_info->sectorsize);

	mutex_lock(&fs_info->chunk_mutex);
	old_total = btrfs_super_total_bytes(super_copy);
	diff = round_down(new_size - device->total_bytes, fs_info->sectorsize);

	if (new_size <= device->total_bytes ||
	    test_bit(BTRFS_DEV_STATE_REPLACE_TGT, &device->dev_state)) {
		mutex_unlock(&fs_info->chunk_mutex);
		return -EINVAL;
	}

	btrfs_set_super_total_bytes(super_copy,
			round_down(old_total + diff, fs_info->sectorsize));
	device->fs_devices->total_rw_bytes += diff;

	btrfs_device_set_total_bytes(device, new_size);
	btrfs_device_set_disk_total_bytes(device, new_size);
	btrfs_clear_space_info_full(device->fs_info);
	if (list_empty(&device->post_commit_list))
		list_add_tail(&device->post_commit_list,
			      &trans->transaction->dev_update_list);
	mutex_unlock(&fs_info->chunk_mutex);

	btrfs_reserve_chunk_metadata(trans, false);
	ret = btrfs_update_device(trans, device);
	btrfs_trans_release_chunk_metadata(trans);

	return ret;
}

static int btrfs_free_chunk(struct btrfs_trans_handle *trans, u64 chunk_offset)
{
	struct btrfs_fs_info *fs_info = trans->fs_info;
	struct btrfs_root *root = fs_info->chunk_root;
	int ret;
	struct btrfs_path *path;
	struct btrfs_key key;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
	key.offset = chunk_offset;
	key.type = BTRFS_CHUNK_ITEM_KEY;

	ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
	if (ret < 0)
		goto out;
	else if (ret > 0) { /* Logic error or corruption */
		btrfs_handle_fs_error(fs_info, -ENOENT,
				      "Failed lookup while freeing chunk.");
		ret = -ENOENT;
		goto out;
	}

	ret = btrfs_del_item(trans, root, path);
	if (ret < 0)
		btrfs_handle_fs_error(fs_info, ret,
				      "Failed to delete chunk item.");
out:
	btrfs_free_path(path);
	return ret;
}

static int btrfs_del_sys_chunk(struct btrfs_fs_info *fs_info, u64 chunk_offset)
{
	struct btrfs_super_block *super_copy = fs_info->super_copy;
	struct btrfs_disk_key *disk_key;
	struct btrfs_chunk *chunk;
	u8 *ptr;
	int ret = 0;
	u32 num_stripes;
	u32 array_size;
	u32 len = 0;
	u32 cur;
	struct btrfs_key key;

	lockdep_assert_held(&fs_info->chunk_mutex);
	array_size = btrfs_super_sys_array_size(super_copy);

	ptr = super_copy->sys_chunk_array;
	cur = 0;

	while (cur < array_size) {
		disk_key = (struct btrfs_disk_key *)ptr;
		btrfs_disk_key_to_cpu(&key, disk_key);

		len = sizeof(*disk_key);

		if (key.type == BTRFS_CHUNK_ITEM_KEY) {
			chunk = (struct btrfs_chunk *)(ptr + len);
			num_stripes = btrfs_stack_chunk_num_stripes(chunk);
			len += btrfs_chunk_item_size(num_stripes);
		} else {
			ret = -EIO;
			break;
		}
		if (key.objectid == BTRFS_FIRST_CHUNK_TREE_OBJECTID &&
		    key.offset == chunk_offset) {
			memmove(ptr, ptr + len, array_size - (cur + len));
			array_size -= len;
			btrfs_set_super_sys_array_size(super_copy, array_size);
		} else {
			ptr += len;
			cur += len;
		}
	}
	return ret;
}

/*
 * btrfs_get_chunk_map() - Find the mapping containing the given logical extent.
 * @logical: Logical block offset in bytes.
 * @length: Length of extent in bytes.
 *
 * Return: Chunk mapping or ERR_PTR.
 */
struct extent_map *btrfs_get_chunk_map(struct btrfs_fs_info *fs_info,
				       u64 logical, u64 length)
{
	struct extent_map_tree *em_tree;
	struct extent_map *em;

	em_tree = &fs_info->mapping_tree;
	read_lock(&em_tree->lock);
	em = lookup_extent_mapping(em_tree, logical, length);
	read_unlock(&em_tree->lock);

	if (!em) {
		btrfs_crit(fs_info, "unable to find logical %llu length %llu",
			   logical, length);
		return ERR_PTR(-EINVAL);
	}

	if (em->start > logical || em->start + em->len < logical) {
		btrfs_crit(fs_info,
			   "found a bad mapping, wanted %llu-%llu, found %llu-%llu",
			   logical, length, em->start, em->start + em->len);
		free_extent_map(em);
		return ERR_PTR(-EINVAL);
	}

	/* callers are responsible for dropping em's ref. */
	return em;
}

static int remove_chunk_item(struct btrfs_trans_handle *trans,
			     struct map_lookup *map, u64 chunk_offset)
{
	int i;

	/*
	 * Removing chunk items and updating the device items in the chunks btree
	 * requires holding the chunk_mutex.
	 * See the comment at btrfs_chunk_alloc() for the details.
	 */
	lockdep_assert_held(&trans->fs_info->chunk_mutex);

	for (i = 0; i < map->num_stripes; i++) {
		int ret;

		ret = btrfs_update_device(trans, map->stripes[i].dev);
		if (ret)
			return ret;
	}

	return btrfs_free_chunk(trans, chunk_offset);
}

int btrfs_remove_chunk(struct btrfs_trans_handle *trans, u64 chunk_offset)
{
	struct btrfs_fs_info *fs_info = trans->fs_info;
	struct extent_map *em;
	struct map_lookup *map;
	u64 dev_extent_len = 0;
	int i, ret = 0;
	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;

	em = btrfs_get_chunk_map(fs_info, chunk_offset, 1);
	if (IS_ERR(em)) {
		/*
		 * This is a logic error, but we don't want to just rely on the
		 * user having built with ASSERT enabled, so if ASSERT doesn't
		 * do anything we still error out.
		 */
		ASSERT(0);
		return PTR_ERR(em);
	}
	map = em->map_lookup;

	/*
	 * First delete the device extent items from the devices btree.
	 * We take the device_list_mutex to avoid racing with the finishing phase
	 * of a device replace operation. See the comment below before acquiring
	 * fs_info->chunk_mutex. Note that here we do not acquire the chunk_mutex
	 * because that can result in a deadlock when deleting the device extent
	 * items from the devices btree - COWing an extent buffer from the btree
	 * may result in allocating a new metadata chunk, which would attempt to
	 * lock again fs_info->chunk_mutex.
	 */
	mutex_lock(&fs_devices->device_list_mutex);
	for (i = 0; i < map->num_stripes; i++) {
		struct btrfs_device *device = map->stripes[i].dev;
		ret = btrfs_free_dev_extent(trans, device,
					    map->stripes[i].physical,
					    &dev_extent_len);
		if (ret) {
			mutex_unlock(&fs_devices->device_list_mutex);
			btrfs_abort_transaction(trans, ret);
			goto out;
		}

		if (device->bytes_used > 0) {
			mutex_lock(&fs_info->chunk_mutex);
			btrfs_device_set_bytes_used(device,
					device->bytes_used - dev_extent_len);
			atomic64_add(dev_extent_len, &fs_info->free_chunk_space);
			btrfs_clear_space_info_full(fs_info);
			mutex_unlock(&fs_info->chunk_mutex);
		}
	}
	mutex_unlock(&fs_devices->device_list_mutex);

	/*
	 * We acquire fs_info->chunk_mutex for 2 reasons:
	 *
	 * 1) Just like with the first phase of the chunk allocation, we must
	 *    reserve system space, do all chunk btree updates and deletions, and
	 *    update the system chunk array in the superblock while holding this
	 *    mutex. This is for similar reasons as explained on the comment at
	 *    the top of btrfs_chunk_alloc();
	 *
	 * 2) Prevent races with the final phase of a device replace operation
	 *    that replaces the device object associated with the map's stripes,
	 *    because the device object's id can change at any time during that
	 *    final phase of the device replace operation
	 *    (dev-replace.c:btrfs_dev_replace_finishing()), so we could grab the
	 *    replaced device and then see it with an ID of
	 *    BTRFS_DEV_REPLACE_DEVID, which would cause a failure when updating
	 *    the device item, which does not exists on the chunk btree.
	 *    The finishing phase of device replace acquires both the
	 *    device_list_mutex and the chunk_mutex, in that order, so we are
	 *    safe by just acquiring the chunk_mutex.
	 */
	trans->removing_chunk = true;
	mutex_lock(&fs_info->chunk_mutex);

	check_system_chunk(trans, map->type);

	ret = remove_chunk_item(trans, map, chunk_offset);
	/*
	 * Normally we should not get -ENOSPC since we reserved space before
	 * through the call to check_system_chunk().
	 *
	 * Despite our system space_info having enough free space, we may not
	 * be able to allocate extents from its block groups, because all have
	 * an incompatible profile, which will force us to allocate a new system
	 * block group with the right profile, or right after we called
	 * check_system_space() above, a scrub turned the only system block group
	 * with enough free space into RO mode.
	 * This is explained with more detail at do_chunk_alloc().
	 *
	 * So if we get -ENOSPC, allocate a new system chunk and retry once.
	 */
	if (ret == -ENOSPC) {
		const u64 sys_flags = btrfs_system_alloc_profile(fs_info);
		struct btrfs_block_group *sys_bg;

		sys_bg = btrfs_create_chunk(trans, sys_flags);
		if (IS_ERR(sys_bg)) {
			ret = PTR_ERR(sys_bg);
			btrfs_abort_transaction(trans, ret);
			goto out;
		}

		ret = btrfs_chunk_alloc_add_chunk_item(trans, sys_bg);
		if (ret) {
			btrfs_abort_transaction(trans, ret);
			goto out;
		}

		ret = remove_chunk_item(trans, map, chunk_offset);
		if (ret) {
			btrfs_abort_transaction(trans, ret);
			goto out;
		}
	} else if (ret) {
		btrfs_abort_transaction(trans, ret);
		goto out;
	}

	trace_btrfs_chunk_free(fs_info, map, chunk_offset, em->len);

	if (map->type & BTRFS_BLOCK_GROUP_SYSTEM) {
		ret = btrfs_del_sys_chunk(fs_info, chunk_offset);
		if (ret) {
			btrfs_abort_transaction(trans, ret);