]> git.proxmox.com Git - mirror_ubuntu-bionic-kernel.git/blob - fs/btrfs/volumes.c
projects / mirror_ubuntu-bionic-kernel.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
blame | history | raw | HEAD
Merge branch 'cleanup/blocksize-diet-part2' of git://git.kernel.org/pub/scm/linux...
[mirror_ubuntu-bionic-kernel.git] / fs / btrfs / volumes.c
1 /*
2 * Copyright (C) 2007 Oracle. All rights reserved.
3 *
4 * This program is free software; you can redistribute it and/or
5 * modify it under the terms of the GNU General Public
6 * License v2 as published by the Free Software Foundation.
7 *
8 * This program is distributed in the hope that it will be useful,
9 * but WITHOUT ANY WARRANTY; without even the implied warranty of
10 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
11 * General Public License for more details.
12 *
13 * You should have received a copy of the GNU General Public
14 * License along with this program; if not, write to the
15 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
16 * Boston, MA 021110-1307, USA.
17 */
18 #include <linux/sched.h>
19 #include <linux/bio.h>
20 #include <linux/slab.h>
21 #include <linux/buffer_head.h>
22 #include <linux/blkdev.h>
23 #include <linux/random.h>
24 #include <linux/iocontext.h>
25 #include <linux/capability.h>
26 #include <linux/ratelimit.h>
27 #include <linux/kthread.h>
28 #include <linux/raid/pq.h>
29 #include <linux/semaphore.h>
30 #include <asm/div64.h>
31 #include "ctree.h"
32 #include "extent_map.h"
33 #include "disk-io.h"
34 #include "transaction.h"
35 #include "print-tree.h"
36 #include "volumes.h"
37 #include "raid56.h"
38 #include "async-thread.h"
39 #include "check-integrity.h"
40 #include "rcu-string.h"
41 #include "math.h"
42 #include "dev-replace.h"
43 #include "sysfs.h"
44
45 static int init_first_rw_device(struct btrfs_trans_handle *trans,
46 struct btrfs_root *root,
47 struct btrfs_device *device);
48 static int btrfs_relocate_sys_chunks(struct btrfs_root *root);
49 static void __btrfs_reset_dev_stats(struct btrfs_device *dev);
50 static void btrfs_dev_stat_print_on_error(struct btrfs_device *dev);
51 static void btrfs_dev_stat_print_on_load(struct btrfs_device *device);
52
53 DEFINE_MUTEX(uuid_mutex);
54 static LIST_HEAD(fs_uuids);
55
56 static struct btrfs_fs_devices *__alloc_fs_devices(void)
57 {
58 struct btrfs_fs_devices *fs_devs;
59
60 fs_devs = kzalloc(sizeof(*fs_devs), GFP_NOFS);
61 if (!fs_devs)
62 return ERR_PTR(-ENOMEM);
63
64 mutex_init(&fs_devs->device_list_mutex);
65
66 INIT_LIST_HEAD(&fs_devs->devices);
67 INIT_LIST_HEAD(&fs_devs->resized_devices);
68 INIT_LIST_HEAD(&fs_devs->alloc_list);
69 INIT_LIST_HEAD(&fs_devs->list);
70
71 return fs_devs;
72 }
73
74 /**
75 * alloc_fs_devices - allocate struct btrfs_fs_devices
76 * @fsid: a pointer to UUID for this FS. If NULL a new UUID is
77 * generated.
78 *
79 * Return: a pointer to a new &struct btrfs_fs_devices on success;
80 * ERR_PTR() on error. Returned struct is not linked onto any lists and
81 * can be destroyed with kfree() right away.
82 */
83 static struct btrfs_fs_devices *alloc_fs_devices(const u8 *fsid)
84 {
85 struct btrfs_fs_devices *fs_devs;
86
87 fs_devs = __alloc_fs_devices();
88 if (IS_ERR(fs_devs))
89 return fs_devs;
90
91 if (fsid)
92 memcpy(fs_devs->fsid, fsid, BTRFS_FSID_SIZE);
93 else
94 generate_random_uuid(fs_devs->fsid);
95
96 return fs_devs;
97 }
98
99 static void free_fs_devices(struct btrfs_fs_devices *fs_devices)
100 {
101 struct btrfs_device *device;
102 WARN_ON(fs_devices->opened);
103 while (!list_empty(&fs_devices->devices)) {
104 device = list_entry(fs_devices->devices.next,
105 struct btrfs_device, dev_list);
106 list_del(&device->dev_list);
107 rcu_string_free(device->name);
108 kfree(device);
109 }
110 kfree(fs_devices);
111 }
112
113 static void btrfs_kobject_uevent(struct block_device *bdev,
114 enum kobject_action action)
115 {
116 int ret;
117
118 ret = kobject_uevent(&disk_to_dev(bdev->bd_disk)->kobj, action);
119 if (ret)
120 pr_warn("BTRFS: Sending event '%d' to kobject: '%s' (%p): failed\n",
121 action,
122 kobject_name(&disk_to_dev(bdev->bd_disk)->kobj),
123 &disk_to_dev(bdev->bd_disk)->kobj);
124 }
125
126 void btrfs_cleanup_fs_uuids(void)
127 {
128 struct btrfs_fs_devices *fs_devices;
129
130 while (!list_empty(&fs_uuids)) {
131 fs_devices = list_entry(fs_uuids.next,
132 struct btrfs_fs_devices, list);
133 list_del(&fs_devices->list);
134 free_fs_devices(fs_devices);
135 }
136 }
137
138 static struct btrfs_device *__alloc_device(void)
139 {
140 struct btrfs_device *dev;
141
142 dev = kzalloc(sizeof(*dev), GFP_NOFS);
143 if (!dev)
144 return ERR_PTR(-ENOMEM);
145
146 INIT_LIST_HEAD(&dev->dev_list);
147 INIT_LIST_HEAD(&dev->dev_alloc_list);
148 INIT_LIST_HEAD(&dev->resized_list);
149
150 spin_lock_init(&dev->io_lock);
151
152 spin_lock_init(&dev->reada_lock);
153 atomic_set(&dev->reada_in_flight, 0);
154 atomic_set(&dev->dev_stats_ccnt, 0);
155 INIT_RADIX_TREE(&dev->reada_zones, GFP_NOFS & ~__GFP_WAIT);
156 INIT_RADIX_TREE(&dev->reada_extents, GFP_NOFS & ~__GFP_WAIT);
157
158 return dev;
159 }
160
161 static noinline struct btrfs_device *__find_device(struct list_head *head,
162 u64 devid, u8 *uuid)
163 {
164 struct btrfs_device *dev;
165
166 list_for_each_entry(dev, head, dev_list) {
167 if (dev->devid == devid &&
168 (!uuid || !memcmp(dev->uuid, uuid, BTRFS_UUID_SIZE))) {
169 return dev;
170 }
171 }
172 return NULL;
173 }
174
175 static noinline struct btrfs_fs_devices *find_fsid(u8 *fsid)
176 {
177 struct btrfs_fs_devices *fs_devices;
178
179 list_for_each_entry(fs_devices, &fs_uuids, list) {
180 if (memcmp(fsid, fs_devices->fsid, BTRFS_FSID_SIZE) == 0)
181 return fs_devices;
182 }
183 return NULL;
184 }
185
186 static int
187 btrfs_get_bdev_and_sb(const char *device_path, fmode_t flags, void *holder,
188 int flush, struct block_device **bdev,
189 struct buffer_head **bh)
190 {
191 int ret;
192
193 *bdev = blkdev_get_by_path(device_path, flags, holder);
194
195 if (IS_ERR(*bdev)) {
196 ret = PTR_ERR(*bdev);
197 printk(KERN_INFO "BTRFS: open %s failed\n", device_path);
198 goto error;
199 }
200
201 if (flush)
202 filemap_write_and_wait((*bdev)->bd_inode->i_mapping);
203 ret = set_blocksize(*bdev, 4096);
204 if (ret) {
205 blkdev_put(*bdev, flags);
206 goto error;
207 }
208 invalidate_bdev(*bdev);
209 *bh = btrfs_read_dev_super(*bdev);
210 if (!*bh) {
211 ret = -EINVAL;
212 blkdev_put(*bdev, flags);
213 goto error;
214 }
215
216 return 0;
217
218 error:
219 *bdev = NULL;
220 *bh = NULL;
221 return ret;
222 }
223
224 static void requeue_list(struct btrfs_pending_bios *pending_bios,
225 struct bio *head, struct bio *tail)
226 {
227
228 struct bio *old_head;
229
230 old_head = pending_bios->head;
231 pending_bios->head = head;
232 if (pending_bios->tail)
233 tail->bi_next = old_head;
234 else
235 pending_bios->tail = tail;
236 }
237
238 /*
239 * we try to collect pending bios for a device so we don't get a large
240 * number of procs sending bios down to the same device. This greatly
241 * improves the schedulers ability to collect and merge the bios.
242 *
243 * But, it also turns into a long list of bios to process and that is sure
244 * to eventually make the worker thread block. The solution here is to
245 * make some progress and then put this work struct back at the end of
246 * the list if the block device is congested. This way, multiple devices
247 * can make progress from a single worker thread.
248 */
249 static noinline void run_scheduled_bios(struct btrfs_device *device)
250 {
251 struct bio *pending;
252 struct backing_dev_info *bdi;
253 struct btrfs_fs_info *fs_info;
254 struct btrfs_pending_bios *pending_bios;
255 struct bio *tail;
256 struct bio *cur;
257 int again = 0;
258 unsigned long num_run;
259 unsigned long batch_run = 0;
260 unsigned long limit;
261 unsigned long last_waited = 0;
262 int force_reg = 0;
263 int sync_pending = 0;
264 struct blk_plug plug;
265
266 /*
267 * this function runs all the bios we've collected for
268 * a particular device. We don't want to wander off to
269 * another device without first sending all of these down.
270 * So, setup a plug here and finish it off before we return
271 */
272 blk_start_plug(&plug);
273
274 bdi = blk_get_backing_dev_info(device->bdev);
275 fs_info = device->dev_root->fs_info;
276 limit = btrfs_async_submit_limit(fs_info);
277 limit = limit * 2 / 3;
278
279 loop:
280 spin_lock(&device->io_lock);
281
282 loop_lock:
283 num_run = 0;
284
285 /* take all the bios off the list at once and process them
286 * later on (without the lock held). But, remember the
287 * tail and other pointers so the bios can be properly reinserted
288 * into the list if we hit congestion
289 */
290 if (!force_reg && device->pending_sync_bios.head) {
291 pending_bios = &device->pending_sync_bios;
292 force_reg = 1;
293 } else {
294 pending_bios = &device->pending_bios;
295 force_reg = 0;
296 }
297
298 pending = pending_bios->head;
299 tail = pending_bios->tail;
300 WARN_ON(pending && !tail);
301
302 /*
303 * if pending was null this time around, no bios need processing
304 * at all and we can stop. Otherwise it'll loop back up again
305 * and do an additional check so no bios are missed.
306 *
307 * device->running_pending is used to synchronize with the
308 * schedule_bio code.
309 */
310 if (device->pending_sync_bios.head == NULL &&
311 device->pending_bios.head == NULL) {
312 again = 0;
313 device->running_pending = 0;
314 } else {
315 again = 1;
316 device->running_pending = 1;
317 }
318
319 pending_bios->head = NULL;
320 pending_bios->tail = NULL;
321
322 spin_unlock(&device->io_lock);
323
324 while (pending) {
325
326 rmb();
327 /* we want to work on both lists, but do more bios on the
328 * sync list than the regular list
329 */
330 if ((num_run > 32 &&
331 pending_bios != &device->pending_sync_bios &&
332 device->pending_sync_bios.head) ||
333 (num_run > 64 && pending_bios == &device->pending_sync_bios &&
334 device->pending_bios.head)) {
335 spin_lock(&device->io_lock);
336 requeue_list(pending_bios, pending, tail);
337 goto loop_lock;
338 }
339
340 cur = pending;
341 pending = pending->bi_next;
342 cur->bi_next = NULL;
343
344 if (atomic_dec_return(&fs_info->nr_async_bios) < limit &&
345 waitqueue_active(&fs_info->async_submit_wait))
346 wake_up(&fs_info->async_submit_wait);
347
348 BUG_ON(atomic_read(&cur->bi_cnt) == 0);
349
350 /*
351 * if we're doing the sync list, record that our
352 * plug has some sync requests on it
353 *
354 * If we're doing the regular list and there are
355 * sync requests sitting around, unplug before
356 * we add more
357 */
358 if (pending_bios == &device->pending_sync_bios) {
359 sync_pending = 1;
360 } else if (sync_pending) {
361 blk_finish_plug(&plug);
362 blk_start_plug(&plug);
363 sync_pending = 0;
364 }
365
366 btrfsic_submit_bio(cur->bi_rw, cur);
367 num_run++;
368 batch_run++;
369 if (need_resched())
370 cond_resched();
371
372 /*
373 * we made progress, there is more work to do and the bdi
374 * is now congested. Back off and let other work structs
375 * run instead
376 */
377 if (pending && bdi_write_congested(bdi) && batch_run > 8 &&
378 fs_info->fs_devices->open_devices > 1) {
379 struct io_context *ioc;
380
381 ioc = current->io_context;
382
383 /*
384 * the main goal here is that we don't want to
385 * block if we're going to be able to submit
386 * more requests without blocking.
387 *
388 * This code does two great things, it pokes into
389 * the elevator code from a filesystem _and_
390 * it makes assumptions about how batching works.
391 */
392 if (ioc && ioc->nr_batch_requests > 0 &&
393 time_before(jiffies, ioc->last_waited + HZ/50UL) &&
394 (last_waited == 0 ||
395 ioc->last_waited == last_waited)) {
396 /*
397 * we want to go through our batch of
398 * requests and stop. So, we copy out
399 * the ioc->last_waited time and test
400 * against it before looping
401 */
402 last_waited = ioc->last_waited;
403 if (need_resched())
404 cond_resched();
405 continue;
406 }
407 spin_lock(&device->io_lock);
408 requeue_list(pending_bios, pending, tail);
409 device->running_pending = 1;
410
411 spin_unlock(&device->io_lock);
412 btrfs_queue_work(fs_info->submit_workers,
413 &device->work);
414 goto done;
415 }
416 /* unplug every 64 requests just for good measure */
417 if (batch_run % 64 == 0) {
418 blk_finish_plug(&plug);
419 blk_start_plug(&plug);
420 sync_pending = 0;
421 }
422 }
423
424 cond_resched();
425 if (again)
426 goto loop;
427
428 spin_lock(&device->io_lock);
429 if (device->pending_bios.head || device->pending_sync_bios.head)
430 goto loop_lock;
431 spin_unlock(&device->io_lock);
432
433 done:
434 blk_finish_plug(&plug);
435 }
436
437 static void pending_bios_fn(struct btrfs_work *work)
438 {
439 struct btrfs_device *device;
440
441 device = container_of(work, struct btrfs_device, work);
442 run_scheduled_bios(device);
443 }
444
445 /*
446 * Add new device to list of registered devices
447 *
448 * Returns:
449 * 1 - first time device is seen
450 * 0 - device already known
451 * < 0 - error
452 */
453 static noinline int device_list_add(const char *path,
454 struct btrfs_super_block *disk_super,
455 u64 devid, struct btrfs_fs_devices **fs_devices_ret)
456 {
457 struct btrfs_device *device;
458 struct btrfs_fs_devices *fs_devices;
459 struct rcu_string *name;
460 int ret = 0;
461 u64 found_transid = btrfs_super_generation(disk_super);
462
463 fs_devices = find_fsid(disk_super->fsid);
464 if (!fs_devices) {
465 fs_devices = alloc_fs_devices(disk_super->fsid);
466 if (IS_ERR(fs_devices))
467 return PTR_ERR(fs_devices);
468
469 list_add(&fs_devices->list, &fs_uuids);
470
471 device = NULL;
472 } else {
473 device = __find_device(&fs_devices->devices, devid,
474 disk_super->dev_item.uuid);
475 }
476
477 if (!device) {
478 if (fs_devices->opened)
479 return -EBUSY;
480
481 device = btrfs_alloc_device(NULL, &devid,
482 disk_super->dev_item.uuid);
483 if (IS_ERR(device)) {
484 /* we can safely leave the fs_devices entry around */
485 return PTR_ERR(device);
486 }
487
488 name = rcu_string_strdup(path, GFP_NOFS);
489 if (!name) {
490 kfree(device);
491 return -ENOMEM;
492 }
493 rcu_assign_pointer(device->name, name);
494
495 mutex_lock(&fs_devices->device_list_mutex);
496 list_add_rcu(&device->dev_list, &fs_devices->devices);
497 fs_devices->num_devices++;
498 mutex_unlock(&fs_devices->device_list_mutex);
499
500 ret = 1;
501 device->fs_devices = fs_devices;
502 } else if (!device->name || strcmp(device->name->str, path)) {
503 /*
504 * When FS is already mounted.
505 * 1. If you are here and if the device->name is NULL that
506 * means this device was missing at time of FS mount.
507 * 2. If you are here and if the device->name is different
508 * from 'path' that means either
509 * a. The same device disappeared and reappeared with
510 * different name. or
511 * b. The missing-disk-which-was-replaced, has
512 * reappeared now.
513 *
514 * We must allow 1 and 2a above. But 2b would be a spurious
515 * and unintentional.
516 *
517 * Further in case of 1 and 2a above, the disk at 'path'
518 * would have missed some transaction when it was away and
519 * in case of 2a the stale bdev has to be updated as well.
520 * 2b must not be allowed at all time.
521 */
522
523 /*
524 * For now, we do allow update to btrfs_fs_device through the
525 * btrfs dev scan cli after FS has been mounted. We're still
526 * tracking a problem where systems fail mount by subvolume id
527 * when we reject replacement on a mounted FS.
528 */
529 if (!fs_devices->opened && found_transid < device->generation) {
530 /*
531 * That is if the FS is _not_ mounted and if you
532 * are here, that means there is more than one
533 * disk with same uuid and devid.We keep the one
534 * with larger generation number or the last-in if
535 * generation are equal.
536 */
537 return -EEXIST;
538 }
539
540 name = rcu_string_strdup(path, GFP_NOFS);
541 if (!name)
542 return -ENOMEM;
543 rcu_string_free(device->name);
544 rcu_assign_pointer(device->name, name);
545 if (device->missing) {
546 fs_devices->missing_devices--;
547 device->missing = 0;
548 }
549 }
550
551 /*
552 * Unmount does not free the btrfs_device struct but would zero
553 * generation along with most of the other members. So just update
554 * it back. We need it to pick the disk with largest generation
555 * (as above).
556 */
557 if (!fs_devices->opened)
558 device->generation = found_transid;
559
560 *fs_devices_ret = fs_devices;
561
562 return ret;
563 }
564
565 static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
566 {
567 struct btrfs_fs_devices *fs_devices;
568 struct btrfs_device *device;
569 struct btrfs_device *orig_dev;
570
571 fs_devices = alloc_fs_devices(orig->fsid);
572 if (IS_ERR(fs_devices))
573 return fs_devices;
574
575 mutex_lock(&orig->device_list_mutex);
576 fs_devices->total_devices = orig->total_devices;
577
578 /* We have held the volume lock, it is safe to get the devices. */
579 list_for_each_entry(orig_dev, &orig->devices, dev_list) {
580 struct rcu_string *name;
581
582 device = btrfs_alloc_device(NULL, &orig_dev->devid,
583 orig_dev->uuid);
584 if (IS_ERR(device))
585 goto error;
586
587 /*
588 * This is ok to do without rcu read locked because we hold the
589 * uuid mutex so nothing we touch in here is going to disappear.
590 */
591 if (orig_dev->name) {
592 name = rcu_string_strdup(orig_dev->name->str, GFP_NOFS);
593 if (!name) {
594 kfree(device);
595 goto error;
596 }
597 rcu_assign_pointer(device->name, name);
598 }
599
600 list_add(&device->dev_list, &fs_devices->devices);
601 device->fs_devices = fs_devices;
602 fs_devices->num_devices++;
603 }
604 mutex_unlock(&orig->device_list_mutex);
605 return fs_devices;
606 error:
607 mutex_unlock(&orig->device_list_mutex);
608 free_fs_devices(fs_devices);
609 return ERR_PTR(-ENOMEM);
610 }
611
612 void btrfs_close_extra_devices(struct btrfs_fs_info *fs_info,
613 struct btrfs_fs_devices *fs_devices, int step)
614 {
615 struct btrfs_device *device, *next;
616 struct btrfs_device *latest_dev = NULL;
617
618 mutex_lock(&uuid_mutex);
619 again:
620 /* This is the initialized path, it is safe to release the devices. */
621 list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
622 if (device->in_fs_metadata) {
623 if (!device->is_tgtdev_for_dev_replace &&
624 (!latest_dev ||
625 device->generation > latest_dev->generation)) {
626 latest_dev = device;
627 }
628 continue;
629 }
630
631 if (device->devid == BTRFS_DEV_REPLACE_DEVID) {
632 /*
633 * In the first step, keep the device which has
634 * the correct fsid and the devid that is used
635 * for the dev_replace procedure.
636 * In the second step, the dev_replace state is
637 * read from the device tree and it is known
638 * whether the procedure is really active or
639 * not, which means whether this device is
640 * used or whether it should be removed.
641 */
642 if (step == 0 || device->is_tgtdev_for_dev_replace) {
643 continue;
644 }
645 }
646 if (device->bdev) {
647 blkdev_put(device->bdev, device->mode);
648 device->bdev = NULL;
649 fs_devices->open_devices--;
650 }
651 if (device->writeable) {
652 list_del_init(&device->dev_alloc_list);
653 device->writeable = 0;
654 if (!device->is_tgtdev_for_dev_replace)
655 fs_devices->rw_devices--;
656 }
657 list_del_init(&device->dev_list);
658 fs_devices->num_devices--;
659 rcu_string_free(device->name);
660 kfree(device);
661 }
662
663 if (fs_devices->seed) {
664 fs_devices = fs_devices->seed;
665 goto again;
666 }
667
668 fs_devices->latest_bdev = latest_dev->bdev;
669
670 mutex_unlock(&uuid_mutex);
671 }
672
673 static void __free_device(struct work_struct *work)
674 {
675 struct btrfs_device *device;
676
677 device = container_of(work, struct btrfs_device, rcu_work);
678
679 if (device->bdev)
680 blkdev_put(device->bdev, device->mode);
681
682 rcu_string_free(device->name);
683 kfree(device);
684 }
685
686 static void free_device(struct rcu_head *head)
687 {
688 struct btrfs_device *device;
689
690 device = container_of(head, struct btrfs_device, rcu);
691
692 INIT_WORK(&device->rcu_work, __free_device);
693 schedule_work(&device->rcu_work);
694 }
695
696 static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
697 {
698 struct btrfs_device *device;
699
700 if (--fs_devices->opened > 0)
701 return 0;
702
703 mutex_lock(&fs_devices->device_list_mutex);
704 list_for_each_entry(device, &fs_devices->devices, dev_list) {
705 struct btrfs_device *new_device;
706 struct rcu_string *name;
707
708 if (device->bdev)
709 fs_devices->open_devices--;
710
711 if (device->writeable &&
712 device->devid != BTRFS_DEV_REPLACE_DEVID) {
713 list_del_init(&device->dev_alloc_list);
714 fs_devices->rw_devices--;
715 }
716
717 if (device->missing)
718 fs_devices->missing_devices--;
719
720 new_device = btrfs_alloc_device(NULL, &device->devid,
721 device->uuid);
722 BUG_ON(IS_ERR(new_device)); /* -ENOMEM */
723
724 /* Safe because we are under uuid_mutex */
725 if (device->name) {
726 name = rcu_string_strdup(device->name->str, GFP_NOFS);
727 BUG_ON(!name); /* -ENOMEM */
728 rcu_assign_pointer(new_device->name, name);
729 }
730
731 list_replace_rcu(&device->dev_list, &new_device->dev_list);
732 new_device->fs_devices = device->fs_devices;
733
734 call_rcu(&device->rcu, free_device);
735 }
736 mutex_unlock(&fs_devices->device_list_mutex);
737
738 WARN_ON(fs_devices->open_devices);
739 WARN_ON(fs_devices->rw_devices);
740 fs_devices->opened = 0;
741 fs_devices->seeding = 0;
742
743 return 0;
744 }
745
746 int btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
747 {
748 struct btrfs_fs_devices *seed_devices = NULL;
749 int ret;
750
751 mutex_lock(&uuid_mutex);
752 ret = __btrfs_close_devices(fs_devices);
753 if (!fs_devices->opened) {
754 seed_devices = fs_devices->seed;
755 fs_devices->seed = NULL;
756 }
757 mutex_unlock(&uuid_mutex);
758
759 while (seed_devices) {
760 fs_devices = seed_devices;
761 seed_devices = fs_devices->seed;
762 __btrfs_close_devices(fs_devices);
763 free_fs_devices(fs_devices);
764 }
765 /*
766 * Wait for rcu kworkers under __btrfs_close_devices
767 * to finish all blkdev_puts so device is really
768 * free when umount is done.
769 */
770 rcu_barrier();
771 return ret;
772 }
773
774 static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
775 fmode_t flags, void *holder)
776 {
777 struct request_queue *q;
778 struct block_device *bdev;
779 struct list_head *head = &fs_devices->devices;
780 struct btrfs_device *device;
781 struct btrfs_device *latest_dev = NULL;
782 struct buffer_head *bh;
783 struct btrfs_super_block *disk_super;
784 u64 devid;
785 int seeding = 1;
786 int ret = 0;
787
788 flags |= FMODE_EXCL;
789
790 list_for_each_entry(device, head, dev_list) {
791 if (device->bdev)
792 continue;
793 if (!device->name)
794 continue;
795
796 /* Just open everything we can; ignore failures here */
797 if (btrfs_get_bdev_and_sb(device->name->str, flags, holder, 1,
798 &bdev, &bh))
799 continue;
800
801 disk_super = (struct btrfs_super_block *)bh->b_data;
802 devid = btrfs_stack_device_id(&disk_super->dev_item);
803 if (devid != device->devid)
804 goto error_brelse;
805
806 if (memcmp(device->uuid, disk_super->dev_item.uuid,
807 BTRFS_UUID_SIZE))
808 goto error_brelse;
809
810 device->generation = btrfs_super_generation(disk_super);
811 if (!latest_dev ||
812 device->generation > latest_dev->generation)
813 latest_dev = device;
814
815 if (btrfs_super_flags(disk_super) & BTRFS_SUPER_FLAG_SEEDING) {
816 device->writeable = 0;
817 } else {
818 device->writeable = !bdev_read_only(bdev);
819 seeding = 0;
820 }
821
822 q = bdev_get_queue(bdev);
823 if (blk_queue_discard(q))
824 device->can_discard = 1;
825
826 device->bdev = bdev;
827 device->in_fs_metadata = 0;
828 device->mode = flags;
829
830 if (!blk_queue_nonrot(bdev_get_queue(bdev)))
831 fs_devices->rotating = 1;
832
833 fs_devices->open_devices++;
834 if (device->writeable &&
835 device->devid != BTRFS_DEV_REPLACE_DEVID) {
836 fs_devices->rw_devices++;
837 list_add(&device->dev_alloc_list,
838 &fs_devices->alloc_list);
839 }
840 brelse(bh);
841 continue;
842
843 error_brelse:
844 brelse(bh);
845 blkdev_put(bdev, flags);
846 continue;
847 }
848 if (fs_devices->open_devices == 0) {
849 ret = -EINVAL;
850 goto out;
851 }
852 fs_devices->seeding = seeding;
853 fs_devices->opened = 1;
854 fs_devices->latest_bdev = latest_dev->bdev;
855 fs_devices->total_rw_bytes = 0;
856 out:
857 return ret;
858 }
859
860 int btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
861 fmode_t flags, void *holder)
862 {
863 int ret;
864
865 mutex_lock(&uuid_mutex);
866 if (fs_devices->opened) {
867 fs_devices->opened++;
868 ret = 0;
869 } else {
870 ret = __btrfs_open_devices(fs_devices, flags, holder);
871 }
872 mutex_unlock(&uuid_mutex);
873 return ret;
874 }
875
876 /*
877 * Look for a btrfs signature on a device. This may be called out of the mount path
878 * and we are not allowed to call set_blocksize during the scan. The superblock
879 * is read via pagecache
880 */
881 int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
882 struct btrfs_fs_devices **fs_devices_ret)
883 {
884 struct btrfs_super_block *disk_super;
885 struct block_device *bdev;
886 struct page *page;
887 void *p;
888 int ret = -EINVAL;
889 u64 devid;
890 u64 transid;
891 u64 total_devices;
892 u64 bytenr;
893 pgoff_t index;
894
895 /*
896 * we would like to check all the supers, but that would make
897 * a btrfs mount succeed after a mkfs from a different FS.
898 * So, we need to add a special mount option to scan for
899 * later supers, using BTRFS_SUPER_MIRROR_MAX instead
900 */
901 bytenr = btrfs_sb_offset(0);
902 flags |= FMODE_EXCL;
903 mutex_lock(&uuid_mutex);
904
905 bdev = blkdev_get_by_path(path, flags, holder);
906
907 if (IS_ERR(bdev)) {
908 ret = PTR_ERR(bdev);
909 goto error;
910 }
911
912 /* make sure our super fits in the device */
913 if (bytenr + PAGE_CACHE_SIZE >= i_size_read(bdev->bd_inode))
914 goto error_bdev_put;
915
916 /* make sure our super fits in the page */
917 if (sizeof(*disk_super) > PAGE_CACHE_SIZE)
918 goto error_bdev_put;
919
920 /* make sure our super doesn't straddle pages on disk */
921 index = bytenr >> PAGE_CACHE_SHIFT;
922 if ((bytenr + sizeof(*disk_super) - 1) >> PAGE_CACHE_SHIFT != index)
923 goto error_bdev_put;
924
925 /* pull in the page with our super */
926 page = read_cache_page_gfp(bdev->bd_inode->i_mapping,
927 index, GFP_NOFS);
928
929 if (IS_ERR_OR_NULL(page))
930 goto error_bdev_put;
931
932 p = kmap(page);
933
934 /* align our pointer to the offset of the super block */
935 disk_super = p + (bytenr & ~PAGE_CACHE_MASK);
936
937 if (btrfs_super_bytenr(disk_super) != bytenr ||
938 btrfs_super_magic(disk_super) != BTRFS_MAGIC)
939 goto error_unmap;
940
941 devid = btrfs_stack_device_id(&disk_super->dev_item);
942 transid = btrfs_super_generation(disk_super);
943 total_devices = btrfs_super_num_devices(disk_super);
944
945 ret = device_list_add(path, disk_super, devid, fs_devices_ret);
946 if (ret > 0) {
947 if (disk_super->label[0]) {
948 if (disk_super->label[BTRFS_LABEL_SIZE - 1])
949 disk_super->label[BTRFS_LABEL_SIZE - 1] = '\0';
950 printk(KERN_INFO "BTRFS: device label %s ", disk_super->label);
951 } else {
952 printk(KERN_INFO "BTRFS: device fsid %pU ", disk_super->fsid);
953 }
954
955 printk(KERN_CONT "devid %llu transid %llu %s\n", devid, transid, path);
956 ret = 0;
957 }
958 if (!ret && fs_devices_ret)
959 (*fs_devices_ret)->total_devices = total_devices;
960
961 error_unmap:
962 kunmap(page);
963 page_cache_release(page);
964
965 error_bdev_put:
966 blkdev_put(bdev, flags);
967 error:
968 mutex_unlock(&uuid_mutex);
969 return ret;
970 }
971
972 /* helper to account the used device space in the range */
973 int btrfs_account_dev_extents_size(struct btrfs_device *device, u64 start,
974 u64 end, u64 *length)
975 {
976 struct btrfs_key key;
977 struct btrfs_root *root = device->dev_root;
978 struct btrfs_dev_extent *dev_extent;
979 struct btrfs_path *path;
980 u64 extent_end;
981 int ret;
982 int slot;
983 struct extent_buffer *l;
984
985 *length = 0;
986
987 if (start >= device->total_bytes || device->is_tgtdev_for_dev_replace)
988 return 0;
989
990 path = btrfs_alloc_path();
991 if (!path)
992 return -ENOMEM;
993 path->reada = 2;
994
995 key.objectid = device->devid;
996 key.offset = start;
997 key.type = BTRFS_DEV_EXTENT_KEY;
998
999 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
1000 if (ret < 0)
1001 goto out;
1002 if (ret > 0) {
1003 ret = btrfs_previous_item(root, path, key.objectid, key.type);
1004 if (ret < 0)
1005 goto out;
1006 }
1007
1008 while (1) {
1009 l = path->nodes[0];
1010 slot = path->slots[0];
1011 if (slot >= btrfs_header_nritems(l)) {
1012 ret = btrfs_next_leaf(root, path);
1013 if (ret == 0)
1014 continue;
1015 if (ret < 0)
1016 goto out;
1017
1018 break;
1019 }
1020 btrfs_item_key_to_cpu(l, &key, slot);
1021
1022 if (key.objectid < device->devid)
1023 goto next;
1024
1025 if (key.objectid > device->devid)
1026 break;
1027
1028 if (key.type != BTRFS_DEV_EXTENT_KEY)
1029 goto next;
1030
1031 dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
1032 extent_end = key.offset + btrfs_dev_extent_length(l,
1033 dev_extent);
1034 if (key.offset <= start && extent_end > end) {
1035 *length = end - start + 1;
1036 break;
1037 } else if (key.offset <= start && extent_end > start)
1038 *length += extent_end - start;
1039 else if (key.offset > start && extent_end <= end)
1040 *length += extent_end - key.offset;
1041 else if (key.offset > start && key.offset <= end) {
1042 *length += end - key.offset + 1;
1043 break;
1044 } else if (key.offset > end)
1045 break;
1046
1047 next:
1048 path->slots[0]++;
1049 }
1050 ret = 0;
1051 out:
1052 btrfs_free_path(path);
1053 return ret;
1054 }
1055
1056 static int contains_pending_extent(struct btrfs_trans_handle *trans,
1057 struct btrfs_device *device,
1058 u64 *start, u64 len)
1059 {
1060 struct extent_map *em;
1061 struct list_head *search_list = &trans->transaction->pending_chunks;
1062 int ret = 0;
1063
1064 again:
1065 list_for_each_entry(em, search_list, list) {
1066 struct map_lookup *map;
1067 int i;
1068
1069 map = (struct map_lookup *)em->bdev;
1070 for (i = 0; i < map->num_stripes; i++) {
1071 if (map->stripes[i].dev != device)
1072 continue;
1073 if (map->stripes[i].physical >= *start + len ||
1074 map->stripes[i].physical + em->orig_block_len <=
1075 *start)
1076 continue;
1077 *start = map->stripes[i].physical +
1078 em->orig_block_len;
1079 ret = 1;
1080 }
1081 }
1082 if (search_list == &trans->transaction->pending_chunks) {
1083 search_list = &trans->root->fs_info->pinned_chunks;
1084 goto again;
1085 }
1086
1087 return ret;
1088 }
1089
1090
1091 /*
1092 * find_free_dev_extent - find free space in the specified device
1093 * @device: the device which we search the free space in
1094 * @num_bytes: the size of the free space that we need
1095 * @start: store the start of the free space.
1096 * @len: the size of the free space. that we find, or the size of the max
1097 * free space if we don't find suitable free space
1098 *
1099 * this uses a pretty simple search, the expectation is that it is
1100 * called very infrequently and that a given device has a small number
1101 * of extents
1102 *
1103 * @start is used to store the start of the free space if we find. But if we
1104 * don't find suitable free space, it will be used to store the start position
1105 * of the max free space.
1106 *
1107 * @len is used to store the size of the free space that we find.
1108 * But if we don't find suitable free space, it is used to store the size of
1109 * the max free space.
1110 */
1111 int find_free_dev_extent(struct btrfs_trans_handle *trans,
1112 struct btrfs_device *device, u64 num_bytes,
1113 u64 *start, u64 *len)
1114 {
1115 struct btrfs_key key;
1116 struct btrfs_root *root = device->dev_root;
1117 struct btrfs_dev_extent *dev_extent;
1118 struct btrfs_path *path;
1119 u64 hole_size;
1120 u64 max_hole_start;
1121 u64 max_hole_size;
1122 u64 extent_end;
1123 u64 search_start;
1124 u64 search_end = device->total_bytes;
1125 int ret;
1126 int slot;
1127 struct extent_buffer *l;
1128
1129 /* FIXME use last free of some kind */
1130
1131 /* we don't want to overwrite the superblock on the drive,
1132 * so we make sure to start at an offset of at least 1MB
1133 */
1134 search_start = max(root->fs_info->alloc_start, 1024ull * 1024);
1135
1136 path = btrfs_alloc_path();
1137 if (!path)
1138 return -ENOMEM;
1139 again:
1140 max_hole_start = search_start;
1141 max_hole_size = 0;
1142 hole_size = 0;
1143
1144 if (search_start >= search_end || device->is_tgtdev_for_dev_replace) {
1145 ret = -ENOSPC;
1146 goto out;
1147 }
1148
1149 path->reada = 2;
1150 path->search_commit_root = 1;
1151 path->skip_locking = 1;
1152
1153 key.objectid = device->devid;
1154 key.offset = search_start;
1155 key.type = BTRFS_DEV_EXTENT_KEY;
1156
1157 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
1158 if (ret < 0)
1159 goto out;
1160 if (ret > 0) {
1161 ret = btrfs_previous_item(root, path, key.objectid, key.type);
1162 if (ret < 0)
1163 goto out;
1164 }
1165
1166 while (1) {
1167 l = path->nodes[0];
1168 slot = path->slots[0];
1169 if (slot >= btrfs_header_nritems(l)) {
1170 ret = btrfs_next_leaf(root, path);
1171 if (ret == 0)
1172 continue;
1173 if (ret < 0)
1174 goto out;
1175
1176 break;
1177 }
1178 btrfs_item_key_to_cpu(l, &key, slot);
1179
1180 if (key.objectid < device->devid)
1181 goto next;
1182
1183 if (key.objectid > device->devid)
1184 break;
1185
1186 if (key.type != BTRFS_DEV_EXTENT_KEY)
1187 goto next;
1188
1189 if (key.offset > search_start) {
1190 hole_size = key.offset - search_start;
1191
1192 /*
1193 * Have to check before we set max_hole_start, otherwise
1194 * we could end up sending back this offset anyway.
1195 */
1196 if (contains_pending_extent(trans, device,
1197 &search_start,
1198 hole_size))
1199 hole_size = 0;
1200
1201 if (hole_size > max_hole_size) {
1202 max_hole_start = search_start;
1203 max_hole_size = hole_size;
1204 }
1205
1206 /*
1207 * If this free space is greater than which we need,
1208 * it must be the max free space that we have found
1209 * until now, so max_hole_start must point to the start
1210 * of this free space and the length of this free space
1211 * is stored in max_hole_size. Thus, we return
1212 * max_hole_start and max_hole_size and go back to the
1213 * caller.
1214 */
1215 if (hole_size >= num_bytes) {
1216 ret = 0;
1217 goto out;
1218 }
1219 }
1220
1221 dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
1222 extent_end = key.offset + btrfs_dev_extent_length(l,
1223 dev_extent);
1224 if (extent_end > search_start)
1225 search_start = extent_end;
1226 next:
1227 path->slots[0]++;
1228 cond_resched();
1229 }
1230
1231 /*
1232 * At this point, search_start should be the end of
1233 * allocated dev extents, and when shrinking the device,
1234 * search_end may be smaller than search_start.
1235 */
1236 if (search_end > search_start)
1237 hole_size = search_end - search_start;
1238
1239 if (hole_size > max_hole_size) {
1240 max_hole_start = search_start;
1241 max_hole_size = hole_size;
1242 }
1243
1244 if (contains_pending_extent(trans, device, &search_start, hole_size)) {
1245 btrfs_release_path(path);
1246 goto again;
1247 }
1248
1249 /* See above. */
1250 if (hole_size < num_bytes)
1251 ret = -ENOSPC;
1252 else
1253 ret = 0;
1254
1255 out:
1256 btrfs_free_path(path);
1257 *start = max_hole_start;
1258 if (len)
1259 *len = max_hole_size;
1260 return ret;
1261 }
1262
1263 static int btrfs_free_dev_extent(struct btrfs_trans_handle *trans,
1264 struct btrfs_device *device,
1265 u64 start, u64 *dev_extent_len)
1266 {
1267 int ret;
1268 struct btrfs_path *path;
1269 struct btrfs_root *root = device->dev_root;
1270 struct btrfs_key key;
1271 struct btrfs_key found_key;
1272 struct extent_buffer *leaf = NULL;
1273 struct btrfs_dev_extent *extent = NULL;
1274
1275 path = btrfs_alloc_path();
1276 if (!path)
1277 return -ENOMEM;
1278
1279 key.objectid = device->devid;
1280 key.offset = start;
1281 key.type = BTRFS_DEV_EXTENT_KEY;
1282 again:
1283 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
1284 if (ret > 0) {
1285 ret = btrfs_previous_item(root, path, key.objectid,
1286 BTRFS_DEV_EXTENT_KEY);
1287 if (ret)
1288 goto out;
1289 leaf = path->nodes[0];
1290 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
1291 extent = btrfs_item_ptr(leaf, path->slots[0],
1292 struct btrfs_dev_extent);
1293 BUG_ON(found_key.offset > start || found_key.offset +
1294 btrfs_dev_extent_length(leaf, extent) < start);
1295 key = found_key;
1296 btrfs_release_path(path);
1297 goto again;
1298 } else if (ret == 0) {
1299 leaf = path->nodes[0];
1300 extent = btrfs_item_ptr(leaf, path->slots[0],
1301 struct btrfs_dev_extent);
1302 } else {
1303 btrfs_error(root->fs_info, ret, "Slot search failed");
1304 goto out;
1305 }
1306
1307 *dev_extent_len = btrfs_dev_extent_length(leaf, extent);
1308
1309 ret = btrfs_del_item(trans, root, path);
1310 if (ret) {
1311 btrfs_error(root->fs_info, ret,
1312 "Failed to remove dev extent item");
1313 }
1314 out:
1315 btrfs_free_path(path);
1316 return ret;
1317 }
1318
1319 static int btrfs_alloc_dev_extent(struct btrfs_trans_handle *trans,
1320 struct btrfs_device *device,
1321 u64 chunk_tree, u64 chunk_objectid,
1322 u64 chunk_offset, u64 start, u64 num_bytes)
1323 {
1324 int ret;
1325 struct btrfs_path *path;
1326 struct btrfs_root *root = device->dev_root;
1327 struct btrfs_dev_extent *extent;
1328 struct extent_buffer *leaf;
1329 struct btrfs_key key;
1330
1331 WARN_ON(!device->in_fs_metadata);
1332 WARN_ON(device->is_tgtdev_for_dev_replace);
1333 path = btrfs_alloc_path();
1334 if (!path)
1335 return -ENOMEM;
1336
1337 key.objectid = device->devid;
1338 key.offset = start;
1339 key.type = BTRFS_DEV_EXTENT_KEY;
1340 ret = btrfs_insert_empty_item(trans, root, path, &key,
1341 sizeof(*extent));
1342 if (ret)
1343 goto out;
1344
1345 leaf = path->nodes[0];
1346 extent = btrfs_item_ptr(leaf, path->slots[0],
1347 struct btrfs_dev_extent);
1348 btrfs_set_dev_extent_chunk_tree(leaf, extent, chunk_tree);
1349 btrfs_set_dev_extent_chunk_objectid(leaf, extent, chunk_objectid);
1350 btrfs_set_dev_extent_chunk_offset(leaf, extent, chunk_offset);
1351
1352 write_extent_buffer(leaf, root->fs_info->chunk_tree_uuid,
1353 btrfs_dev_extent_chunk_tree_uuid(extent), BTRFS_UUID_SIZE);
1354
1355 btrfs_set_dev_extent_length(leaf, extent, num_bytes);
1356 btrfs_mark_buffer_dirty(leaf);
1357 out:
1358 btrfs_free_path(path);
1359 return ret;
1360 }
1361
1362 static u64 find_next_chunk(struct btrfs_fs_info *fs_info)
1363 {
1364 struct extent_map_tree *em_tree;
1365 struct extent_map *em;
1366 struct rb_node *n;
1367 u64 ret = 0;
1368
1369 em_tree = &fs_info->mapping_tree.map_tree;
1370 read_lock(&em_tree->lock);
1371 n = rb_last(&em_tree->map);
1372 if (n) {
1373 em = rb_entry(n, struct extent_map, rb_node);
1374 ret = em->start + em->len;
1375 }
1376 read_unlock(&em_tree->lock);
1377
1378 return ret;
1379 }
1380
1381 static noinline int find_next_devid(struct btrfs_fs_info *fs_info,
1382 u64 *devid_ret)
1383 {
1384 int ret;
1385 struct btrfs_key key;
1386 struct btrfs_key found_key;
1387 struct btrfs_path *path;
1388
1389 path = btrfs_alloc_path();
1390 if (!path)
1391 return -ENOMEM;
1392
1393 key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
1394 key.type = BTRFS_DEV_ITEM_KEY;
1395 key.offset = (u64)-1;
1396
1397 ret = btrfs_search_slot(NULL, fs_info->chunk_root, &key, path, 0, 0);
1398 if (ret < 0)
1399 goto error;
1400
1401 BUG_ON(ret == 0); /* Corruption */
1402
1403 ret = btrfs_previous_item(fs_info->chunk_root, path,
1404 BTRFS_DEV_ITEMS_OBJECTID,
1405 BTRFS_DEV_ITEM_KEY);
1406 if (ret) {
1407 *devid_ret = 1;
1408 } else {
1409 btrfs_item_key_to_cpu(path->nodes[0], &found_key,
1410 path->slots[0]);
1411 *devid_ret = found_key.offset + 1;
1412 }
1413 ret = 0;
1414 error:
1415 btrfs_free_path(path);
1416 return ret;
1417 }
1418
1419 /*
1420 * the device information is stored in the chunk root
1421 * the btrfs_device struct should be fully filled in
1422 */
1423 static int btrfs_add_device(struct btrfs_trans_handle *trans,
1424 struct btrfs_root *root,
1425 struct btrfs_device *device)
1426 {
1427 int ret;
1428 struct btrfs_path *path;
1429 struct btrfs_dev_item *dev_item;
1430 struct extent_buffer *leaf;
1431 struct btrfs_key key;
1432 unsigned long ptr;
1433
1434 root = root->fs_info->chunk_root;
1435
1436 path = btrfs_alloc_path();
1437 if (!path)
1438 return -ENOMEM;
1439
1440 key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
1441 key.type = BTRFS_DEV_ITEM_KEY;
1442 key.offset = device->devid;
1443
1444 ret = btrfs_insert_empty_item(trans, root, path, &key,
1445 sizeof(*dev_item));
1446 if (ret)
1447 goto out;
1448
1449 leaf = path->nodes[0];
1450 dev_item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_dev_item);
1451
1452 btrfs_set_device_id(leaf, dev_item, device->devid);
1453 btrfs_set_device_generation(leaf, dev_item, 0);
1454 btrfs_set_device_type(leaf, dev_item, device->type);
1455 btrfs_set_device_io_align(leaf, dev_item, device->io_align);
1456 btrfs_set_device_io_width(leaf, dev_item, device->io_width);
1457 btrfs_set_device_sector_size(leaf, dev_item, device->sector_size);
1458 btrfs_set_device_total_bytes(leaf, dev_item,
1459 btrfs_device_get_disk_total_bytes(device));
1460 btrfs_set_device_bytes_used(leaf, dev_item,
1461 btrfs_device_get_bytes_used(device));
1462 btrfs_set_device_group(leaf, dev_item, 0);
1463 btrfs_set_device_seek_speed(leaf, dev_item, 0);
1464 btrfs_set_device_bandwidth(leaf, dev_item, 0);
1465 btrfs_set_device_start_offset(leaf, dev_item, 0);
1466
1467 ptr = btrfs_device_uuid(dev_item);
1468 write_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
1469 ptr = btrfs_device_fsid(dev_item);
1470 write_extent_buffer(leaf, root->fs_info->fsid, ptr, BTRFS_UUID_SIZE);
1471 btrfs_mark_buffer_dirty(leaf);
1472
1473 ret = 0;
1474 out:
1475 btrfs_free_path(path);
1476 return ret;
1477 }
1478
1479 /*
1480 * Function to update ctime/mtime for a given device path.
1481 * Mainly used for ctime/mtime based probe like libblkid.
1482 */
1483 static void update_dev_time(char *path_name)
1484 {
1485 struct file *filp;
1486
1487 filp = filp_open(path_name, O_RDWR, 0);
1488 if (IS_ERR(filp))
1489 return;
1490 file_update_time(filp);
1491 filp_close(filp, NULL);
1492 return;
1493 }
1494
1495 static int btrfs_rm_dev_item(struct btrfs_root *root,
1496 struct btrfs_device *device)
1497 {
1498 int ret;
1499 struct btrfs_path *path;
1500 struct btrfs_key key;
1501 struct btrfs_trans_handle *trans;
1502
1503 root = root->fs_info->chunk_root;
1504
1505 path = btrfs_alloc_path();
1506 if (!path)
1507 return -ENOMEM;
1508
1509 trans = btrfs_start_transaction(root, 0);
1510 if (IS_ERR(trans)) {
1511 btrfs_free_path(path);
1512 return PTR_ERR(trans);
1513 }
1514 key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
1515 key.type = BTRFS_DEV_ITEM_KEY;
1516 key.offset = device->devid;
1517
1518 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
1519 if (ret < 0)
1520 goto out;
1521
1522 if (ret > 0) {
1523 ret = -ENOENT;
1524 goto out;
1525 }
1526
1527 ret = btrfs_del_item(trans, root, path);
1528 if (ret)
1529 goto out;
1530 out:
1531 btrfs_free_path(path);
1532 btrfs_commit_transaction(trans, root);
1533 return ret;
1534 }
1535
1536 int btrfs_rm_device(struct btrfs_root *root, char *device_path)
1537 {
1538 struct btrfs_device *device;
1539 struct btrfs_device *next_device;
1540 struct block_device *bdev;
1541 struct buffer_head *bh = NULL;
1542 struct btrfs_super_block *disk_super;
1543 struct btrfs_fs_devices *cur_devices;
1544 u64 all_avail;
1545 u64 devid;
1546 u64 num_devices;
1547 u8 *dev_uuid;
1548 unsigned seq;
1549 int ret = 0;
1550 bool clear_super = false;
1551
1552 mutex_lock(&uuid_mutex);
1553
1554 do {
1555 seq = read_seqbegin(&root->fs_info->profiles_lock);
1556
1557 all_avail = root->fs_info->avail_data_alloc_bits |
1558 root->fs_info->avail_system_alloc_bits |
1559 root->fs_info->avail_metadata_alloc_bits;
1560 } while (read_seqretry(&root->fs_info->profiles_lock, seq));
1561
1562 num_devices = root->fs_info->fs_devices->num_devices;
1563 btrfs_dev_replace_lock(&root->fs_info->dev_replace);
1564 if (btrfs_dev_replace_is_ongoing(&root->fs_info->dev_replace)) {
1565 WARN_ON(num_devices < 1);
1566 num_devices--;
1567 }
1568 btrfs_dev_replace_unlock(&root->fs_info->dev_replace);
1569
1570 if ((all_avail & BTRFS_BLOCK_GROUP_RAID10) && num_devices <= 4) {
1571 ret = BTRFS_ERROR_DEV_RAID10_MIN_NOT_MET;
1572 goto out;
1573 }
1574
1575 if ((all_avail & BTRFS_BLOCK_GROUP_RAID1) && num_devices <= 2) {
1576 ret = BTRFS_ERROR_DEV_RAID1_MIN_NOT_MET;
1577 goto out;
1578 }
1579
1580 if ((all_avail & BTRFS_BLOCK_GROUP_RAID5) &&
1581 root->fs_info->fs_devices->rw_devices <= 2) {
1582 ret = BTRFS_ERROR_DEV_RAID5_MIN_NOT_MET;
1583 goto out;
1584 }
1585 if ((all_avail & BTRFS_BLOCK_GROUP_RAID6) &&
1586 root->fs_info->fs_devices->rw_devices <= 3) {
1587 ret = BTRFS_ERROR_DEV_RAID6_MIN_NOT_MET;
1588 goto out;
1589 }
1590
1591 if (strcmp(device_path, "missing") == 0) {
1592 struct list_head *devices;
1593 struct btrfs_device *tmp;
1594
1595 device = NULL;
1596 devices = &root->fs_info->fs_devices->devices;
1597 /*
1598 * It is safe to read the devices since the volume_mutex
1599 * is held.
1600 */
1601 list_for_each_entry(tmp, devices, dev_list) {
1602 if (tmp->in_fs_metadata &&
1603 !tmp->is_tgtdev_for_dev_replace &&
1604 !tmp->bdev) {
1605 device = tmp;
1606 break;
1607 }
1608 }
1609 bdev = NULL;
1610 bh = NULL;
1611 disk_super = NULL;
1612 if (!device) {
1613 ret = BTRFS_ERROR_DEV_MISSING_NOT_FOUND;
1614 goto out;
1615 }
1616 } else {
1617 ret = btrfs_get_bdev_and_sb(device_path,
1618 FMODE_WRITE | FMODE_EXCL,
1619 root->fs_info->bdev_holder, 0,
1620 &bdev, &bh);
1621 if (ret)
1622 goto out;
1623 disk_super = (struct btrfs_super_block *)bh->b_data;
1624 devid = btrfs_stack_device_id(&disk_super->dev_item);
1625 dev_uuid = disk_super->dev_item.uuid;
1626 device = btrfs_find_device(root->fs_info, devid, dev_uuid,
1627 disk_super->fsid);
1628 if (!device) {
1629 ret = -ENOENT;
1630 goto error_brelse;
1631 }
1632 }
1633
1634 if (device->is_tgtdev_for_dev_replace) {
1635 ret = BTRFS_ERROR_DEV_TGT_REPLACE;
1636 goto error_brelse;
1637 }
1638
1639 if (device->writeable && root->fs_info->fs_devices->rw_devices == 1) {
1640 ret = BTRFS_ERROR_DEV_ONLY_WRITABLE;
1641 goto error_brelse;
1642 }
1643
1644 if (device->writeable) {
1645 lock_chunks(root);
1646 list_del_init(&device->dev_alloc_list);
1647 device->fs_devices->rw_devices--;
1648 unlock_chunks(root);
1649 clear_super = true;
1650 }
1651
1652 mutex_unlock(&uuid_mutex);
1653 ret = btrfs_shrink_device(device, 0);
1654 mutex_lock(&uuid_mutex);
1655 if (ret)
1656 goto error_undo;
1657
1658 /*
1659 * TODO: the superblock still includes this device in its num_devices
1660 * counter although write_all_supers() is not locked out. This
1661 * could give a filesystem state which requires a degraded mount.
1662 */
1663 ret = btrfs_rm_dev_item(root->fs_info->chunk_root, device);
1664 if (ret)
1665 goto error_undo;
1666
1667 device->in_fs_metadata = 0;
1668 btrfs_scrub_cancel_dev(root->fs_info, device);
1669
1670 /*
1671 * the device list mutex makes sure that we don't change
1672 * the device list while someone else is writing out all
1673 * the device supers. Whoever is writing all supers, should
1674 * lock the device list mutex before getting the number of
1675 * devices in the super block (super_copy). Conversely,
1676 * whoever updates the number of devices in the super block
1677 * (super_copy) should hold the device list mutex.
1678 */
1679
1680 cur_devices = device->fs_devices;
1681 mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
1682 list_del_rcu(&device->dev_list);
1683
1684 device->fs_devices->num_devices--;
1685 device->fs_devices->total_devices--;
1686
1687 if (device->missing)
1688 device->fs_devices->missing_devices--;
1689
1690 next_device = list_entry(root->fs_info->fs_devices->devices.next,
1691 struct btrfs_device, dev_list);
1692 if (device->bdev == root->fs_info->sb->s_bdev)
1693 root->fs_info->sb->s_bdev = next_device->bdev;
1694 if (device->bdev == root->fs_info->fs_devices->latest_bdev)
1695 root->fs_info->fs_devices->latest_bdev = next_device->bdev;
1696
1697 if (device->bdev) {
1698 device->fs_devices->open_devices--;
1699 /* remove sysfs entry */
1700 btrfs_kobj_rm_device(root->fs_info, device);
1701 }
1702
1703 call_rcu(&device->rcu, free_device);
1704
1705 num_devices = btrfs_super_num_devices(root->fs_info->super_copy) - 1;
1706 btrfs_set_super_num_devices(root->fs_info->super_copy, num_devices);
1707 mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
1708
1709 if (cur_devices->open_devices == 0) {
1710 struct btrfs_fs_devices *fs_devices;
1711 fs_devices = root->fs_info->fs_devices;
1712 while (fs_devices) {
1713 if (fs_devices->seed == cur_devices) {
1714 fs_devices->seed = cur_devices->seed;
1715 break;
1716 }
1717 fs_devices = fs_devices->seed;
1718 }
1719 cur_devices->seed = NULL;
1720 __btrfs_close_devices(cur_devices);
1721 free_fs_devices(cur_devices);
1722 }
1723
1724 root->fs_info->num_tolerated_disk_barrier_failures =
1725 btrfs_calc_num_tolerated_disk_barrier_failures(root->fs_info);
1726
1727 /*
1728 * at this point, the device is zero sized. We want to
1729 * remove it from the devices list and zero out the old super
1730 */
1731 if (clear_super && disk_super) {
1732 u64 bytenr;
1733 int i;
1734
1735 /* make sure this device isn't detected as part of
1736 * the FS anymore
1737 */
1738 memset(&disk_super->magic, 0, sizeof(disk_super->magic));
1739 set_buffer_dirty(bh);
1740 sync_dirty_buffer(bh);
1741
1742 /* clear the mirror copies of super block on the disk
1743 * being removed, 0th copy is been taken care above and
1744 * the below would take of the rest
1745 */
1746 for (i = 1; i < BTRFS_SUPER_MIRROR_MAX; i++) {
1747 bytenr = btrfs_sb_offset(i);
1748 if (bytenr + BTRFS_SUPER_INFO_SIZE >=
1749 i_size_read(bdev->bd_inode))
1750 break;
1751
1752 brelse(bh);
1753 bh = __bread(bdev, bytenr / 4096,
1754 BTRFS_SUPER_INFO_SIZE);
1755 if (!bh)
1756 continue;
1757
1758 disk_super = (struct btrfs_super_block *)bh->b_data;
1759
1760 if (btrfs_super_bytenr(disk_super) != bytenr ||
1761 btrfs_super_magic(disk_super) != BTRFS_MAGIC) {
1762 continue;
1763 }
1764 memset(&disk_super->magic, 0,
1765 sizeof(disk_super->magic));
1766 set_buffer_dirty(bh);
1767 sync_dirty_buffer(bh);
1768 }
1769 }
1770
1771 ret = 0;
1772
1773 if (bdev) {
1774 /* Notify udev that device has changed */
1775 btrfs_kobject_uevent(bdev, KOBJ_CHANGE);
1776
1777 /* Update ctime/mtime for device path for libblkid */
1778 update_dev_time(device_path);
1779 }
1780
1781 error_brelse:
1782 brelse(bh);
1783 if (bdev)
1784 blkdev_put(bdev, FMODE_READ | FMODE_EXCL);
1785 out:
1786 mutex_unlock(&uuid_mutex);
1787 return ret;
1788 error_undo:
1789 if (device->writeable) {
1790 lock_chunks(root);
1791 list_add(&device->dev_alloc_list,
1792 &root->fs_info->fs_devices->alloc_list);
1793 device->fs_devices->rw_devices++;
1794 unlock_chunks(root);
1795 }
1796 goto error_brelse;
1797 }
1798
1799 void btrfs_rm_dev_replace_remove_srcdev(struct btrfs_fs_info *fs_info,
1800 struct btrfs_device *srcdev)
1801 {
1802 struct btrfs_fs_devices *fs_devices;
1803
1804 WARN_ON(!mutex_is_locked(&fs_info->fs_devices->device_list_mutex));
1805
1806 /*
1807 * in case of fs with no seed, srcdev->fs_devices will point
1808 * to fs_devices of fs_info. However when the dev being replaced is
1809 * a seed dev it will point to the seed's local fs_devices. In short
1810 * srcdev will have its correct fs_devices in both the cases.
1811 */
1812 fs_devices = srcdev->fs_devices;
1813
1814 list_del_rcu(&srcdev->dev_list);
1815 list_del_rcu(&srcdev->dev_alloc_list);
1816 fs_devices->num_devices--;
1817 if (srcdev->missing)
1818 fs_devices->missing_devices--;
1819
1820 if (srcdev->writeable) {
1821 fs_devices->rw_devices--;
1822 /* zero out the old super if it is writable */
1823 btrfs_scratch_superblock(srcdev);
1824 }
1825
1826 if (srcdev->bdev)
1827 fs_devices->open_devices--;
1828 }
1829
1830 void btrfs_rm_dev_replace_free_srcdev(struct btrfs_fs_info *fs_info,
1831 struct btrfs_device *srcdev)
1832 {
1833 struct btrfs_fs_devices *fs_devices = srcdev->fs_devices;
1834
1835 call_rcu(&srcdev->rcu, free_device);
1836
1837 /*
1838 * unless fs_devices is seed fs, num_devices shouldn't go
1839 * zero
1840 */
1841 BUG_ON(!fs_devices->num_devices && !fs_devices->seeding);
1842
1843 /* if this is no devs we rather delete the fs_devices */
1844 if (!fs_devices->num_devices) {
1845 struct btrfs_fs_devices *tmp_fs_devices;
1846
1847 tmp_fs_devices = fs_info->fs_devices;
1848 while (tmp_fs_devices) {
1849 if (tmp_fs_devices->seed == fs_devices) {
1850 tmp_fs_devices->seed = fs_devices->seed;
1851 break;
1852 }
1853 tmp_fs_devices = tmp_fs_devices->seed;
1854 }
1855 fs_devices->seed = NULL;
1856 __btrfs_close_devices(fs_devices);
1857 free_fs_devices(fs_devices);
1858 }
1859 }
1860
1861 void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
1862 struct btrfs_device *tgtdev)
1863 {
1864 struct btrfs_device *next_device;
1865
1866 mutex_lock(&uuid_mutex);
1867 WARN_ON(!tgtdev);
1868 mutex_lock(&fs_info->fs_devices->device_list_mutex);
1869 if (tgtdev->bdev) {
1870 btrfs_scratch_superblock(tgtdev);
1871 fs_info->fs_devices->open_devices--;
1872 }
1873 fs_info->fs_devices->num_devices--;
1874
1875 next_device = list_entry(fs_info->fs_devices->devices.next,
1876 struct btrfs_device, dev_list);
1877 if (tgtdev->bdev == fs_info->sb->s_bdev)
1878 fs_info->sb->s_bdev = next_device->bdev;
1879 if (tgtdev->bdev == fs_info->fs_devices->latest_bdev)
1880 fs_info->fs_devices->latest_bdev = next_device->bdev;
1881 list_del_rcu(&tgtdev->dev_list);
1882
1883 call_rcu(&tgtdev->rcu, free_device);
1884
1885 mutex_unlock(&fs_info->fs_devices->device_list_mutex);
1886 mutex_unlock(&uuid_mutex);
1887 }
1888
1889 static int btrfs_find_device_by_path(struct btrfs_root *root, char *device_path,
1890 struct btrfs_device **device)
1891 {
1892 int ret = 0;
1893 struct btrfs_super_block *disk_super;
1894 u64 devid;
1895 u8 *dev_uuid;
1896 struct block_device *bdev;
1897 struct buffer_head *bh;
1898
1899 *device = NULL;
1900 ret = btrfs_get_bdev_and_sb(device_path, FMODE_READ,
1901 root->fs_info->bdev_holder, 0, &bdev, &bh);
1902 if (ret)
1903 return ret;
1904 disk_super = (struct btrfs_super_block *)bh->b_data;
1905 devid = btrfs_stack_device_id(&disk_super->dev_item);
1906 dev_uuid = disk_super->dev_item.uuid;
1907 *device = btrfs_find_device(root->fs_info, devid, dev_uuid,
1908 disk_super->fsid);
1909 brelse(bh);
1910 if (!*device)
1911 ret = -ENOENT;
1912 blkdev_put(bdev, FMODE_READ);
1913 return ret;
1914 }
1915
1916 int btrfs_find_device_missing_or_by_path(struct btrfs_root *root,
1917 char *device_path,
1918 struct btrfs_device **device)
1919 {
1920 *device = NULL;
1921 if (strcmp(device_path, "missing") == 0) {
1922 struct list_head *devices;
1923 struct btrfs_device *tmp;
1924
1925 devices = &root->fs_info->fs_devices->devices;
1926 /*
1927 * It is safe to read the devices since the volume_mutex
1928 * is held by the caller.
1929 */
1930 list_for_each_entry(tmp, devices, dev_list) {
1931 if (tmp->in_fs_metadata && !tmp->bdev) {
1932 *device = tmp;
1933 break;
1934 }
1935 }
1936
1937 if (!*device) {
1938 btrfs_err(root->fs_info, "no missing device found");
1939 return -ENOENT;
1940 }
1941
1942 return 0;
1943 } else {
1944 return btrfs_find_device_by_path(root, device_path, device);
1945 }
1946 }
1947
1948 /*
1949 * does all the dirty work required for changing file system's UUID.
1950 */
1951 static int btrfs_prepare_sprout(struct btrfs_root *root)
1952 {
1953 struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
1954 struct btrfs_fs_devices *old_devices;
1955 struct btrfs_fs_devices *seed_devices;
1956 struct btrfs_super_block *disk_super = root->fs_info->super_copy;
1957 struct btrfs_device *device;
1958 u64 super_flags;
1959
1960 BUG_ON(!mutex_is_locked(&uuid_mutex));
1961 if (!fs_devices->seeding)
1962 return -EINVAL;
1963
1964 seed_devices = __alloc_fs_devices();
1965 if (IS_ERR(seed_devices))
1966 return PTR_ERR(seed_devices);
1967
1968 old_devices = clone_fs_devices(fs_devices);
1969 if (IS_ERR(old_devices)) {
1970 kfree(seed_devices);
1971 return PTR_ERR(old_devices);
1972 }
1973
1974 list_add(&old_devices->list, &fs_uuids);
1975
1976 memcpy(seed_devices, fs_devices, sizeof(*seed_devices));
1977 seed_devices->opened = 1;
1978 INIT_LIST_HEAD(&seed_devices->devices);
1979 INIT_LIST_HEAD(&seed_devices->alloc_list);
1980 mutex_init(&seed_devices->device_list_mutex);
1981
1982 mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
1983 list_splice_init_rcu(&fs_devices->devices, &seed_devices->devices,
1984 synchronize_rcu);
1985 list_for_each_entry(device, &seed_devices->devices, dev_list)
1986 device->fs_devices = seed_devices;
1987
1988 lock_chunks(root);
1989 list_splice_init(&fs_devices->alloc_list, &seed_devices->alloc_list);
1990 unlock_chunks(root);
1991
1992 fs_devices->seeding = 0;
1993 fs_devices->num_devices = 0;
1994 fs_devices->open_devices = 0;
1995 fs_devices->missing_devices = 0;
1996 fs_devices->rotating = 0;
1997 fs_devices->seed = seed_devices;
1998
1999 generate_random_uuid(fs_devices->fsid);
2000 memcpy(root->fs_info->fsid, fs_devices->fsid, BTRFS_FSID_SIZE);
2001 memcpy(disk_super->fsid, fs_devices->fsid, BTRFS_FSID_SIZE);
2002 mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
2003
2004 super_flags = btrfs_super_flags(disk_super) &
2005 ~BTRFS_SUPER_FLAG_SEEDING;
2006 btrfs_set_super_flags(disk_super, super_flags);
2007
2008 return 0;
2009 }
2010
2011 /*
2012 * strore the expected generation for seed devices in device items.
2013 */
2014 static int btrfs_finish_sprout(struct btrfs_trans_handle *trans,
2015 struct btrfs_root *root)
2016 {
2017 struct btrfs_path *path;
2018 struct extent_buffer *leaf;
2019 struct btrfs_dev_item *dev_item;
2020 struct btrfs_device *device;
2021 struct btrfs_key key;
2022 u8 fs_uuid[BTRFS_UUID_SIZE];
2023 u8 dev_uuid[BTRFS_UUID_SIZE];
2024 u64 devid;
2025 int ret;
2026
2027 path = btrfs_alloc_path();
2028 if (!path)
2029 return -ENOMEM;
2030
2031 root = root->fs_info->chunk_root;
2032 key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
2033 key.offset = 0;
2034 key.type = BTRFS_DEV_ITEM_KEY;
2035
2036 while (1) {
2037 ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
2038 if (ret < 0)
2039 goto error;
2040
2041 leaf = path->nodes[0];
2042 next_slot:
2043 if (path->slots[0] >= btrfs_header_nritems(leaf)) {
2044 ret = btrfs_next_leaf(root, path);
2045 if (ret > 0)
2046 break;
2047 if (ret < 0)
2048 goto error;
2049 leaf = path->nodes[0];
2050 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
2051 btrfs_release_path(path);
2052 continue;
2053 }
2054
2055 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
2056 if (key.objectid != BTRFS_DEV_ITEMS_OBJECTID ||
2057 key.type != BTRFS_DEV_ITEM_KEY)
2058 break;
2059
2060 dev_item = btrfs_item_ptr(leaf, path->slots[0],
2061 struct btrfs_dev_item);
2062 devid = btrfs_device_id(leaf, dev_item);
2063 read_extent_buffer(leaf, dev_uuid, btrfs_device_uuid(dev_item),
2064 BTRFS_UUID_SIZE);
2065 read_extent_buffer(leaf, fs_uuid, btrfs_device_fsid(dev_item),
2066 BTRFS_UUID_SIZE);
2067 device = btrfs_find_device(root->fs_info, devid, dev_uuid,
2068 fs_uuid);
2069 BUG_ON(!device); /* Logic error */
2070
2071 if (device->fs_devices->seeding) {
2072 btrfs_set_device_generation(leaf, dev_item,
2073 device->generation);
2074 btrfs_mark_buffer_dirty(leaf);
2075 }
2076
2077 path->slots[0]++;
2078 goto next_slot;
2079 }
2080 ret = 0;
2081 error:
2082 btrfs_free_path(path);
2083 return ret;
2084 }
2085
2086 int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
2087 {
2088 struct request_queue *q;
2089 struct btrfs_trans_handle *trans;
2090 struct btrfs_device *device;
2091 struct block_device *bdev;
2092 struct list_head *devices;
2093 struct super_block *sb = root->fs_info->sb;
2094 struct rcu_string *name;
2095 u64 tmp;
2096 int seeding_dev = 0;
2097 int ret = 0;
2098
2099 if ((sb->s_flags & MS_RDONLY) && !root->fs_info->fs_devices->seeding)
2100 return -EROFS;
2101
2102 bdev = blkdev_get_by_path(device_path, FMODE_WRITE | FMODE_EXCL,
2103 root->fs_info->bdev_holder);
2104 if (IS_ERR(bdev))
2105 return PTR_ERR(bdev);
2106
2107 if (root->fs_info->fs_devices->seeding) {
2108 seeding_dev = 1;
2109 down_write(&sb->s_umount);
2110 mutex_lock(&uuid_mutex);
2111 }
2112
2113 filemap_write_and_wait(bdev->bd_inode->i_mapping);
2114
2115 devices = &root->fs_info->fs_devices->devices;
2116
2117 mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
2118 list_for_each_entry(device, devices, dev_list) {
2119 if (device->bdev == bdev) {
2120 ret = -EEXIST;
2121 mutex_unlock(
2122 &root->fs_info->fs_devices->device_list_mutex);
2123 goto error;
2124 }
2125 }
2126 mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
2127
2128 device = btrfs_alloc_device(root->fs_info, NULL, NULL);
2129 if (IS_ERR(device)) {
2130 /* we can safely leave the fs_devices entry around */
2131 ret = PTR_ERR(device);
2132 goto error;
2133 }
2134
2135 name = rcu_string_strdup(device_path, GFP_NOFS);
2136 if (!name) {
2137 kfree(device);
2138 ret = -ENOMEM;
2139 goto error;
2140 }
2141 rcu_assign_pointer(device->name, name);
2142
2143 trans = btrfs_start_transaction(root, 0);
2144 if (IS_ERR(trans)) {
2145 rcu_string_free(device->name);
2146 kfree(device);
2147 ret = PTR_ERR(trans);
2148 goto error;
2149 }
2150
2151 q = bdev_get_queue(bdev);
2152 if (blk_queue_discard(q))
2153 device->can_discard = 1;
2154 device->writeable = 1;
2155 device->generation = trans->transid;
2156 device->io_width = root->sectorsize;
2157 device->io_align = root->sectorsize;
2158 device->sector_size = root->sectorsize;
2159 device->total_bytes = i_size_read(bdev->bd_inode);
2160 device->disk_total_bytes = device->total_bytes;
2161 device->commit_total_bytes = device->total_bytes;
2162 device->dev_root = root->fs_info->dev_root;
2163 device->bdev = bdev;
2164 device->in_fs_metadata = 1;
2165 device->is_tgtdev_for_dev_replace = 0;
2166 device->mode = FMODE_EXCL;
2167 device->dev_stats_valid = 1;
2168 set_blocksize(device->bdev, 4096);
2169
2170 if (seeding_dev) {
2171 sb->s_flags &= ~MS_RDONLY;
2172 ret = btrfs_prepare_sprout(root);
2173 BUG_ON(ret); /* -ENOMEM */
2174 }
2175
2176 device->fs_devices = root->fs_info->fs_devices;
2177
2178 mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
2179 lock_chunks(root);
2180 list_add_rcu(&device->dev_list, &root->fs_info->fs_devices->devices);
2181 list_add(&device->dev_alloc_list,
2182 &root->fs_info->fs_devices->alloc_list);
2183 root->fs_info->fs_devices->num_devices++;
2184 root->fs_info->fs_devices->open_devices++;
2185 root->fs_info->fs_devices->rw_devices++;
2186 root->fs_info->fs_devices->total_devices++;
2187 root->fs_info->fs_devices->total_rw_bytes += device->total_bytes;
2188
2189 spin_lock(&root->fs_info->free_chunk_lock);
2190 root->fs_info->free_chunk_space += device->total_bytes;
2191 spin_unlock(&root->fs_info->free_chunk_lock);
2192
2193 if (!blk_queue_nonrot(bdev_get_queue(bdev)))
2194 root->fs_info->fs_devices->rotating = 1;
2195
2196 tmp = btrfs_super_total_bytes(root->fs_info->super_copy);
2197 btrfs_set_super_total_bytes(root->fs_info->super_copy,
2198 tmp + device->total_bytes);
2199
2200 tmp = btrfs_super_num_devices(root->fs_info->super_copy);
2201 btrfs_set_super_num_devices(root->fs_info->super_copy,
2202 tmp + 1);
2203
2204 /* add sysfs device entry */
2205 btrfs_kobj_add_device(root->fs_info, device);
2206
2207 /*
2208 * we've got more storage, clear any full flags on the space
2209 * infos
2210 */
2211 btrfs_clear_space_info_full(root->fs_info);
2212
2213 unlock_chunks(root);
2214 mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
2215
2216 if (seeding_dev) {
2217 lock_chunks(root);
2218 ret = init_first_rw_device(trans, root, device);
2219 unlock_chunks(root);
2220 if (ret) {
2221 btrfs_abort_transaction(trans, root, ret);
2222 goto error_trans;
2223 }
2224 }
2225
2226 ret = btrfs_add_device(trans, root, device);
2227 if (ret) {
2228 btrfs_abort_transaction(trans, root, ret);
2229 goto error_trans;
2230 }
2231
2232 if (seeding_dev) {
2233 char fsid_buf[BTRFS_UUID_UNPARSED_SIZE];
2234
2235 ret = btrfs_finish_sprout(trans, root);
2236 if (ret) {
2237 btrfs_abort_transaction(trans, root, ret);
2238 goto error_trans;
2239 }
2240
2241 /* Sprouting would change fsid of the mounted root,
2242 * so rename the fsid on the sysfs
2243 */
2244 snprintf(fsid_buf, BTRFS_UUID_UNPARSED_SIZE, "%pU",
2245 root->fs_info->fsid);
2246 if (kobject_rename(&root->fs_info->super_kobj, fsid_buf))
2247 goto error_trans;
2248 }
2249
2250 root->fs_info->num_tolerated_disk_barrier_failures =
2251 btrfs_calc_num_tolerated_disk_barrier_failures(root->fs_info);
2252 ret = btrfs_commit_transaction(trans, root);
2253
2254 if (seeding_dev) {
2255 mutex_unlock(&uuid_mutex);
2256 up_write(&sb->s_umount);
2257
2258 if (ret) /* transaction commit */
2259 return ret;
2260
2261 ret = btrfs_relocate_sys_chunks(root);
2262 if (ret < 0)
2263 btrfs_error(root->fs_info, ret,
2264 "Failed to relocate sys chunks after "
2265 "device initialization. This can be fixed "
2266 "using the \"btrfs balance\" command.");
2267 trans = btrfs_attach_transaction(root);
2268 if (IS_ERR(trans)) {
2269 if (PTR_ERR(trans) == -ENOENT)
2270 return 0;
2271 return PTR_ERR(trans);
2272 }
2273 ret = btrfs_commit_transaction(trans, root);
2274 }
2275
2276 /* Update ctime/mtime for libblkid */
2277 update_dev_time(device_path);
2278 return ret;
2279
2280 error_trans:
2281 btrfs_end_transaction(trans, root);
2282 rcu_string_free(device->name);
2283 btrfs_kobj_rm_device(root->fs_info, device);
2284 kfree(device);
2285 error:
2286 blkdev_put(bdev, FMODE_EXCL);
2287 if (seeding_dev) {
2288 mutex_unlock(&uuid_mutex);
2289 up_write(&sb->s_umount);
2290 }
2291 return ret;
2292 }
2293
2294 int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
2295 struct btrfs_device *srcdev,
2296 struct btrfs_device **device_out)
2297 {
2298 struct request_queue *q;
2299 struct btrfs_device *device;
2300 struct block_device *bdev;
2301 struct btrfs_fs_info *fs_info = root->fs_info;
2302 struct list_head *devices;
2303 struct rcu_string *name;
2304 u64 devid = BTRFS_DEV_REPLACE_DEVID;
2305 int ret = 0;
2306
2307 *device_out = NULL;
2308 if (fs_info->fs_devices->seeding) {
2309 btrfs_err(fs_info, "the filesystem is a seed filesystem!");
2310 return -EINVAL;
2311 }
2312
2313 bdev = blkdev_get_by_path(device_path, FMODE_WRITE | FMODE_EXCL,
2314 fs_info->bdev_holder);
2315 if (IS_ERR(bdev)) {
2316 btrfs_err(fs_info, "target device %s is invalid!", device_path);
2317 return PTR_ERR(bdev);
2318 }
2319
2320 filemap_write_and_wait(bdev->bd_inode->i_mapping);
2321
2322 devices = &fs_info->fs_devices->devices;
2323 list_for_each_entry(device, devices, dev_list) {
2324 if (device->bdev == bdev) {
2325 btrfs_err(fs_info, "target device is in the filesystem!");
2326 ret = -EEXIST;
2327 goto error;
2328 }
2329 }
2330
2331
2332 if (i_size_read(bdev->bd_inode) <
2333 btrfs_device_get_total_bytes(srcdev)) {
2334 btrfs_err(fs_info, "target device is smaller than source device!");
2335 ret = -EINVAL;
2336 goto error;
2337 }
2338
2339
2340 device = btrfs_alloc_device(NULL, &devid, NULL);
2341 if (IS_ERR(device)) {
2342 ret = PTR_ERR(device);
2343 goto error;
2344 }
2345
2346 name = rcu_string_strdup(device_path, GFP_NOFS);
2347 if (!name) {
2348 kfree(device);
2349 ret = -ENOMEM;
2350 goto error;
2351 }
2352 rcu_assign_pointer(device->name, name);
2353
2354 q = bdev_get_queue(bdev);
2355 if (blk_queue_discard(q))
2356 device->can_discard = 1;
2357 mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
2358 device->writeable = 1;
2359 device->generation = 0;
2360 device->io_width = root->sectorsize;
2361 device->io_align = root->sectorsize;
2362 device->sector_size = root->sectorsize;
2363 device->total_bytes = btrfs_device_get_total_bytes(srcdev);
2364 device->disk_total_bytes = btrfs_device_get_disk_total_bytes(srcdev);
2365 device->bytes_used = btrfs_device_get_bytes_used(srcdev);
2366 ASSERT(list_empty(&srcdev->resized_list));
2367 device->commit_total_bytes = srcdev->commit_total_bytes;
2368 device->commit_bytes_used = device->bytes_used;
2369 device->dev_root = fs_info->dev_root;
2370 device->bdev = bdev;
2371 device->in_fs_metadata = 1;
2372 device->is_tgtdev_for_dev_replace = 1;
2373 device->mode = FMODE_EXCL;
2374 device->dev_stats_valid = 1;
2375 set_blocksize(device->bdev, 4096);
2376 device->fs_devices = fs_info->fs_devices;
2377 list_add(&device->dev_list, &fs_info->fs_devices->devices);
2378 fs_info->fs_devices->num_devices++;
2379 fs_info->fs_devices->open_devices++;
2380 mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
2381
2382 *device_out = device;
2383 return ret;
2384
2385 error:
2386 blkdev_put(bdev, FMODE_EXCL);
2387 return ret;
2388 }
2389
2390 void btrfs_init_dev_replace_tgtdev_for_resume(struct btrfs_fs_info *fs_info,
2391 struct btrfs_device *tgtdev)
2392 {
2393 WARN_ON(fs_info->fs_devices->rw_devices == 0);
2394 tgtdev->io_width = fs_info->dev_root->sectorsize;
2395 tgtdev->io_align = fs_info->dev_root->sectorsize;
2396 tgtdev->sector_size = fs_info->dev_root->sectorsize;
2397 tgtdev->dev_root = fs_info->dev_root;
2398 tgtdev->in_fs_metadata = 1;
2399 }
2400
2401 static noinline int btrfs_update_device(struct btrfs_trans_handle *trans,
2402 struct btrfs_device *device)
2403 {
2404 int ret;
2405 struct btrfs_path *path;
2406 struct btrfs_root *root;
2407 struct btrfs_dev_item *dev_item;
2408 struct extent_buffer *leaf;
2409 struct btrfs_key key;
2410
2411 root = device->dev_root->fs_info->chunk_root;
2412
2413 path = btrfs_alloc_path();
2414 if (!path)
2415 return -ENOMEM;
2416
2417 key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
2418 key.type = BTRFS_DEV_ITEM_KEY;
2419 key.offset = device->devid;
2420
2421 ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
2422 if (ret < 0)
2423 goto out;
2424
2425 if (ret > 0) {
2426 ret = -ENOENT;
2427 goto out;
2428 }
2429
2430 leaf = path->nodes[0];
2431 dev_item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_dev_item);
2432
2433 btrfs_set_device_id(leaf, dev_item, device->devid);
2434 btrfs_set_device_type(leaf, dev_item, device->type);
2435 btrfs_set_device_io_align(leaf, dev_item, device->io_align);
2436 btrfs_set_device_io_width(leaf, dev_item, device->io_width);
2437 btrfs_set_device_sector_size(leaf, dev_item, device->sector_size);
2438 btrfs_set_device_total_bytes(leaf, dev_item,
2439 btrfs_device_get_disk_total_bytes(device));
2440 btrfs_set_device_bytes_used(leaf, dev_item,
2441 btrfs_device_get_bytes_used(device));
2442 btrfs_mark_buffer_dirty(leaf);
2443
2444 out:
2445 btrfs_free_path(path);
2446 return ret;
2447 }
2448
2449 int btrfs_grow_device(struct btrfs_trans_handle *trans,
2450 struct btrfs_device *device, u64 new_size)
2451 {
2452 struct btrfs_super_block *super_copy =
2453 device->dev_root->fs_info->super_copy;
2454 struct btrfs_fs_devices *fs_devices;
2455 u64 old_total;
2456 u64 diff;
2457
2458 if (!device->writeable)
2459 return -EACCES;
2460
2461 lock_chunks(device->dev_root);
2462 old_total = btrfs_super_total_bytes(super_copy);
2463 diff = new_size - device->total_bytes;
2464
2465 if (new_size <= device->total_bytes ||
2466 device->is_tgtdev_for_dev_replace) {
2467 unlock_chunks(device->dev_root);
2468 return -EINVAL;
2469 }
2470
2471 fs_devices = device->dev_root->fs_info->fs_devices;
2472
2473 btrfs_set_super_total_bytes(super_copy, old_total + diff);
2474 device->fs_devices->total_rw_bytes += diff;
2475
2476 btrfs_device_set_total_bytes(device, new_size);
2477 btrfs_device_set_disk_total_bytes(device, new_size);
2478 btrfs_clear_space_info_full(device->dev_root->fs_info);
2479 if (list_empty(&device->resized_list))
2480 list_add_tail(&device->resized_list,
2481 &fs_devices->resized_devices);
2482 unlock_chunks(device->dev_root);
2483
2484 return btrfs_update_device(trans, device);
2485 }
2486
2487 static int btrfs_free_chunk(struct btrfs_trans_handle *trans,
2488 struct btrfs_root *root,
2489 u64 chunk_tree, u64 chunk_objectid,
2490 u64 chunk_offset)
2491 {
2492 int ret;
2493 struct btrfs_path *path;
2494 struct btrfs_key key;
2495
2496 root = root->fs_info->chunk_root;
2497 path = btrfs_alloc_path();
2498 if (!path)
2499 return -ENOMEM;
2500
2501 key.objectid = chunk_objectid;
2502 key.offset = chunk_offset;
2503 key.type = BTRFS_CHUNK_ITEM_KEY;
2504
2505 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
2506 if (ret < 0)
2507 goto out;
2508 else if (ret > 0) { /* Logic error or corruption */
2509 btrfs_error(root->fs_info, -ENOENT,
2510 "Failed lookup while freeing chunk.");
2511 ret = -ENOENT;
2512 goto out;
2513 }
2514
2515 ret = btrfs_del_item(trans, root, path);
2516 if (ret < 0)
2517 btrfs_error(root->fs_info, ret,
2518 "Failed to delete chunk item.");
2519 out:
2520 btrfs_free_path(path);
2521 return ret;
2522 }
2523
2524 static int btrfs_del_sys_chunk(struct btrfs_root *root, u64 chunk_objectid, u64
2525 chunk_offset)
2526 {
2527 struct btrfs_super_block *super_copy = root->fs_info->super_copy;
2528 struct btrfs_disk_key *disk_key;
2529 struct btrfs_chunk *chunk;
2530 u8 *ptr;
2531 int ret = 0;
2532 u32 num_stripes;
2533 u32 array_size;
2534 u32 len = 0;
2535 u32 cur;
2536 struct btrfs_key key;
2537
2538 lock_chunks(root);
2539 array_size = btrfs_super_sys_array_size(super_copy);
2540
2541 ptr = super_copy->sys_chunk_array;
2542 cur = 0;
2543
2544 while (cur < array_size) {
2545 disk_key = (struct btrfs_disk_key *)ptr;
2546 btrfs_disk_key_to_cpu(&key, disk_key);
2547
2548 len = sizeof(*disk_key);
2549
2550 if (key.type == BTRFS_CHUNK_ITEM_KEY) {
2551 chunk = (struct btrfs_chunk *)(ptr + len);
2552 num_stripes = btrfs_stack_chunk_num_stripes(chunk);
2553 len += btrfs_chunk_item_size(num_stripes);
2554 } else {
2555 ret = -EIO;
2556 break;
2557 }
2558 if (key.objectid == chunk_objectid &&
2559 key.offset == chunk_offset) {
2560 memmove(ptr, ptr + len, array_size - (cur + len));
2561 array_size -= len;
2562 btrfs_set_super_sys_array_size(super_copy, array_size);
2563 } else {
2564 ptr += len;
2565 cur += len;
2566 }
2567 }
2568 unlock_chunks(root);
2569 return ret;
2570 }
2571
2572 int btrfs_remove_chunk(struct btrfs_trans_handle *trans,
2573 struct btrfs_root *root, u64 chunk_offset)
2574 {
2575 struct extent_map_tree *em_tree;
2576 struct extent_map *em;
2577 struct btrfs_root *extent_root = root->fs_info->extent_root;
2578 struct map_lookup *map;
2579 u64 dev_extent_len = 0;
2580 u64 chunk_objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
2581 u64 chunk_tree = root->fs_info->chunk_root->objectid;
2582 int i, ret = 0;
2583
2584 /* Just in case */
2585 root = root->fs_info->chunk_root;
2586 em_tree = &root->fs_info->mapping_tree.map_tree;
2587
2588 read_lock(&em_tree->lock);
2589 em = lookup_extent_mapping(em_tree, chunk_offset, 1);
2590 read_unlock(&em_tree->lock);
2591
2592 if (!em || em->start > chunk_offset ||
2593 em->start + em->len < chunk_offset) {
2594 /*
2595 * This is a logic error, but we don't want to just rely on the
2596 * user having built with ASSERT enabled, so if ASSERT doens't
2597 * do anything we still error out.
2598 */
2599 ASSERT(0);
2600 if (em)
2601 free_extent_map(em);
2602 return -EINVAL;
2603 }
2604 map = (struct map_lookup *)em->bdev;
2605
2606 for (i = 0; i < map->num_stripes; i++) {
2607 struct btrfs_device *device = map->stripes[i].dev;
2608 ret = btrfs_free_dev_extent(trans, device,
2609 map->stripes[i].physical,
2610 &dev_extent_len);
2611 if (ret) {
2612 btrfs_abort_transaction(trans, root, ret);
2613 goto out;
2614 }
2615
2616 if (device->bytes_used > 0) {
2617 lock_chunks(root);
2618 btrfs_device_set_bytes_used(device,
2619 device->bytes_used - dev_extent_len);
2620 spin_lock(&root->fs_info->free_chunk_lock);
2621 root->fs_info->free_chunk_space += dev_extent_len;
2622 spin_unlock(&root->fs_info->free_chunk_lock);
2623 btrfs_clear_space_info_full(root->fs_info);
2624 unlock_chunks(root);
2625 }
2626
2627 if (map->stripes[i].dev) {
2628 ret = btrfs_update_device(trans, map->stripes[i].dev);
2629 if (ret) {
2630 btrfs_abort_transaction(trans, root, ret);
2631 goto out;
2632 }
2633 }
2634 }
2635 ret = btrfs_free_chunk(trans, root, chunk_tree, chunk_objectid,
2636 chunk_offset);
2637 if (ret) {
2638 btrfs_abort_transaction(trans, root, ret);
2639 goto out;
2640 }
2641
2642 trace_btrfs_chunk_free(root, map, chunk_offset, em->len);
2643
2644 if (map->type & BTRFS_BLOCK_GROUP_SYSTEM) {
2645 ret = btrfs_del_sys_chunk(root, chunk_objectid, chunk_offset);
2646 if (ret) {
2647 btrfs_abort_transaction(trans, root, ret);
2648 goto out;
2649 }
2650 }
2651
2652 ret = btrfs_remove_block_group(trans, extent_root, chunk_offset, em);
2653 if (ret) {
2654 btrfs_abort_transaction(trans, extent_root, ret);
2655 goto out;
2656 }
2657
2658 out:
2659 /* once for us */
2660 free_extent_map(em);
2661 return ret;
2662 }
2663
2664 static int btrfs_relocate_chunk(struct btrfs_root *root,
2665 u64 chunk_tree, u64 chunk_objectid,
2666 u64 chunk_offset)
2667 {
2668 struct btrfs_root *extent_root;
2669 struct btrfs_trans_handle *trans;
2670 int ret;
2671
2672 root = root->fs_info->chunk_root;
2673 extent_root = root->fs_info->extent_root;
2674
2675 ret = btrfs_can_relocate(extent_root, chunk_offset);
2676 if (ret)
2677 return -ENOSPC;
2678
2679 /* step one, relocate all the extents inside this chunk */
2680 ret = btrfs_relocate_block_group(extent_root, chunk_offset);
2681 if (ret)
2682 return ret;
2683
2684 trans = btrfs_start_transaction(root, 0);
2685 if (IS_ERR(trans)) {
2686 ret = PTR_ERR(trans);
2687 btrfs_std_error(root->fs_info, ret);
2688 return ret;
2689 }
2690
2691 /*
2692 * step two, delete the device extents and the
2693 * chunk tree entries
2694 */
2695 ret = btrfs_remove_chunk(trans, root, chunk_offset);
2696 btrfs_end_transaction(trans, root);
2697 return ret;
2698 }
2699
2700 static int btrfs_relocate_sys_chunks(struct btrfs_root *root)
2701 {
2702 struct btrfs_root *chunk_root = root->fs_info->chunk_root;
2703 struct btrfs_path *path;
2704 struct extent_buffer *leaf;
2705 struct btrfs_chunk *chunk;
2706 struct btrfs_key key;
2707 struct btrfs_key found_key;
2708 u64 chunk_tree = chunk_root->root_key.objectid;
2709 u64 chunk_type;
2710 bool retried = false;
2711 int failed = 0;
2712 int ret;
2713
2714 path = btrfs_alloc_path();
2715 if (!path)
2716 return -ENOMEM;
2717
2718 again:
2719 key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
2720 key.offset = (u64)-1;
2721 key.type = BTRFS_CHUNK_ITEM_KEY;
2722
2723 while (1) {
2724 ret = btrfs_search_slot(NULL, chunk_root, &key, path, 0, 0);
2725 if (ret < 0)
2726 goto error;
2727 BUG_ON(ret == 0); /* Corruption */
2728
2729 ret = btrfs_previous_item(chunk_root, path, key.objectid,
2730 key.type);
2731 if (ret < 0)
2732 goto error;
2733 if (ret > 0)
2734 break;
2735
2736 leaf = path->nodes[0];
2737 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
2738
2739 chunk = btrfs_item_ptr(leaf, path->slots[0],
2740 struct btrfs_chunk);
2741 chunk_type = btrfs_chunk_type(leaf, chunk);
2742 btrfs_release_path(path);
2743
2744 if (chunk_type & BTRFS_BLOCK_GROUP_SYSTEM) {
2745 ret = btrfs_relocate_chunk(chunk_root, chunk_tree,
2746 found_key.objectid,
2747 found_key.offset);
2748 if (ret == -ENOSPC)
2749 failed++;
2750 else
2751 BUG_ON(ret);
2752 }
2753
2754 if (found_key.offset == 0)
2755 break;
2756 key.offset = found_key.offset - 1;
2757 }
2758 ret = 0;
2759 if (failed && !retried) {
2760 failed = 0;
2761 retried = true;
2762 goto again;
2763 } else if (WARN_ON(failed && retried)) {
2764 ret = -ENOSPC;
2765 }
2766 error:
2767 btrfs_free_path(path);
2768 return ret;
2769 }
2770
2771 static int insert_balance_item(struct btrfs_root *root,
2772 struct btrfs_balance_control *bctl)
2773 {
2774 struct btrfs_trans_handle *trans;
2775 struct btrfs_balance_item *item;
2776 struct btrfs_disk_balance_args disk_bargs;
2777 struct btrfs_path *path;
2778 struct extent_buffer *leaf;
2779 struct btrfs_key key;
2780 int ret, err;
2781
2782 path = btrfs_alloc_path();
2783 if (!path)
2784 return -ENOMEM;
2785
2786 trans = btrfs_start_transaction(root, 0);
2787 if (IS_ERR(trans)) {
2788 btrfs_free_path(path);
2789 return PTR_ERR(trans);
2790 }
2791
2792 key.objectid = BTRFS_BALANCE_OBJECTID;
2793 key.type = BTRFS_BALANCE_ITEM_KEY;
2794 key.offset = 0;
2795
2796 ret = btrfs_insert_empty_item(trans, root, path, &key,
2797 sizeof(*item));
2798 if (ret)
2799 goto out;
2800
2801 leaf = path->nodes[0];
2802 item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_balance_item);
2803
2804 memset_extent_buffer(leaf, 0, (unsigned long)item, sizeof(*item));
2805
2806 btrfs_cpu_balance_args_to_disk(&disk_bargs, &bctl->data);
2807 btrfs_set_balance_data(leaf, item, &disk_bargs);
2808 btrfs_cpu_balance_args_to_disk(&disk_bargs, &bctl->meta);
2809 btrfs_set_balance_meta(leaf, item, &disk_bargs);
2810 btrfs_cpu_balance_args_to_disk(&disk_bargs, &bctl->sys);
2811 btrfs_set_balance_sys(leaf, item, &disk_bargs);
2812
2813 btrfs_set_balance_flags(leaf, item, bctl->flags);
2814
2815 btrfs_mark_buffer_dirty(leaf);
2816 out:
2817 btrfs_free_path(path);
2818 err = btrfs_commit_transaction(trans, root);
2819 if (err && !ret)
2820 ret = err;
2821 return ret;
2822 }
2823
2824 static int del_balance_item(struct btrfs_root *root)
2825 {
2826 struct btrfs_trans_handle *trans;
2827 struct btrfs_path *path;
2828 struct btrfs_key key;
2829 int ret, err;
2830
2831 path = btrfs_alloc_path();
2832 if (!path)
2833 return -ENOMEM;
2834
2835 trans = btrfs_start_transaction(root, 0);
2836 if (IS_ERR(trans)) {
2837 btrfs_free_path(path);
2838 return PTR_ERR(trans);
2839 }
2840
2841 key.objectid = BTRFS_BALANCE_OBJECTID;
2842 key.type = BTRFS_BALANCE_ITEM_KEY;
2843 key.offset = 0;
2844
2845 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
2846 if (ret < 0)
2847 goto out;
2848 if (ret > 0) {
2849 ret = -ENOENT;
2850 goto out;
2851 }
2852
2853 ret = btrfs_del_item(trans, root, path);
2854 out:
2855 btrfs_free_path(path);
2856 err = btrfs_commit_transaction(trans, root);
2857 if (err && !ret)
2858 ret = err;
2859 return ret;
2860 }
2861
2862 /*
2863 * This is a heuristic used to reduce the number of chunks balanced on
2864 * resume after balance was interrupted.
2865 */
2866 static void update_balance_args(struct btrfs_balance_control *bctl)
2867 {
2868 /*
2869 * Turn on soft mode for chunk types that were being converted.
2870 */
2871 if (bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT)
2872 bctl->data.flags |= BTRFS_BALANCE_ARGS_SOFT;
2873 if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT)
2874 bctl->sys.flags |= BTRFS_BALANCE_ARGS_SOFT;
2875 if (bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT)
2876 bctl->meta.flags |= BTRFS_BALANCE_ARGS_SOFT;
2877
2878 /*
2879 * Turn on usage filter if is not already used. The idea is
2880 * that chunks that we have already balanced should be
2881 * reasonably full. Don't do it for chunks that are being
2882 * converted - that will keep us from relocating unconverted
2883 * (albeit full) chunks.
2884 */
2885 if (!(bctl->data.flags & BTRFS_BALANCE_ARGS_USAGE) &&
2886 !(bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT)) {
2887 bctl->data.flags |= BTRFS_BALANCE_ARGS_USAGE;
2888 bctl->data.usage = 90;
2889 }
2890 if (!(bctl->sys.flags & BTRFS_BALANCE_ARGS_USAGE) &&
2891 !(bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT)) {
2892 bctl->sys.flags |= BTRFS_BALANCE_ARGS_USAGE;
2893 bctl->sys.usage = 90;
2894 }
2895 if (!(bctl->meta.flags & BTRFS_BALANCE_ARGS_USAGE) &&
2896 !(bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT)) {
2897 bctl->meta.flags |= BTRFS_BALANCE_ARGS_USAGE;
2898 bctl->meta.usage = 90;
2899 }
2900 }
2901
2902 /*
2903 * Should be called with both balance and volume mutexes held to
2904 * serialize other volume operations (add_dev/rm_dev/resize) with
2905 * restriper. Same goes for unset_balance_control.
2906 */
2907 static void set_balance_control(struct btrfs_balance_control *bctl)
2908 {
2909 struct btrfs_fs_info *fs_info = bctl->fs_info;
2910
2911 BUG_ON(fs_info->balance_ctl);
2912
2913 spin_lock(&fs_info->balance_lock);
2914 fs_info->balance_ctl = bctl;
2915 spin_unlock(&fs_info->balance_lock);
2916 }
2917
2918 static void unset_balance_control(struct btrfs_fs_info *fs_info)
2919 {
2920 struct btrfs_balance_control *bctl = fs_info->balance_ctl;
2921
2922 BUG_ON(!fs_info->balance_ctl);
2923
2924 spin_lock(&fs_info->balance_lock);
2925 fs_info->balance_ctl = NULL;
2926 spin_unlock(&fs_info->balance_lock);
2927
2928 kfree(bctl);
2929 }
2930
2931 /*
2932 * Balance filters. Return 1 if chunk should be filtered out
2933 * (should not be balanced).
2934 */
2935 static int chunk_profiles_filter(u64 chunk_type,
2936 struct btrfs_balance_args *bargs)
2937 {
2938 chunk_type = chunk_to_extended(chunk_type) &
2939 BTRFS_EXTENDED_PROFILE_MASK;
2940
2941 if (bargs->profiles & chunk_type)
2942 return 0;
2943
2944 return 1;
2945 }
2946
2947 static int chunk_usage_filter(struct btrfs_fs_info *fs_info, u64 chunk_offset,
2948 struct btrfs_balance_args *bargs)
2949 {
2950 struct btrfs_block_group_cache *cache;
2951 u64 chunk_used, user_thresh;
2952 int ret = 1;
2953
2954 cache = btrfs_lookup_block_group(fs_info, chunk_offset);
2955 chunk_used = btrfs_block_group_used(&cache->item);
2956
2957 if (bargs->usage == 0)
2958 user_thresh = 1;
2959 else if (bargs->usage > 100)
2960 user_thresh = cache->key.offset;
2961 else
2962 user_thresh = div_factor_fine(cache->key.offset,
2963 bargs->usage);
2964
2965 if (chunk_used < user_thresh)
2966 ret = 0;
2967
2968 btrfs_put_block_group(cache);
2969 return ret;
2970 }
2971
2972 static int chunk_devid_filter(struct extent_buffer *leaf,
2973 struct btrfs_chunk *chunk,
2974 struct btrfs_balance_args *bargs)
2975 {
2976 struct btrfs_stripe *stripe;
2977 int num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
2978 int i;
2979
2980 for (i = 0; i < num_stripes; i++) {
2981 stripe = btrfs_stripe_nr(chunk, i);
2982 if (btrfs_stripe_devid(leaf, stripe) == bargs->devid)
2983 return 0;
2984 }
2985
2986 return 1;
2987 }
2988
2989 /* [pstart, pend) */
2990 static int chunk_drange_filter(struct extent_buffer *leaf,
2991 struct btrfs_chunk *chunk,
2992 u64 chunk_offset,
2993 struct btrfs_balance_args *bargs)
2994 {
2995 struct btrfs_stripe *stripe;
2996 int num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
2997 u64 stripe_offset;
2998 u64 stripe_length;
2999 int factor;
3000 int i;
3001
3002 if (!(bargs->flags & BTRFS_BALANCE_ARGS_DEVID))
3003 return 0;
3004
3005 if (btrfs_chunk_type(leaf, chunk) & (BTRFS_BLOCK_GROUP_DUP |
3006 BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID10)) {
3007 factor = num_stripes / 2;
3008 } else if (btrfs_chunk_type(leaf, chunk) & BTRFS_BLOCK_GROUP_RAID5) {
3009 factor = num_stripes - 1;
3010 } else if (btrfs_chunk_type(leaf, chunk) & BTRFS_BLOCK_GROUP_RAID6) {
3011 factor = num_stripes - 2;
3012 } else {
3013 factor = num_stripes;
3014 }
3015
3016 for (i = 0; i < num_stripes; i++) {
3017 stripe = btrfs_stripe_nr(chunk, i);
3018 if (btrfs_stripe_devid(leaf, stripe) != bargs->devid)
3019 continue;
3020
3021 stripe_offset = btrfs_stripe_offset(leaf, stripe);
3022 stripe_length = btrfs_chunk_length(leaf, chunk);
3023 do_div(stripe_length, factor);
3024
3025 if (stripe_offset < bargs->pend &&
3026 stripe_offset + stripe_length > bargs->pstart)
3027 return 0;
3028 }
3029
3030 return 1;
3031 }
3032
3033 /* [vstart, vend) */
3034 static int chunk_vrange_filter(struct extent_buffer *leaf,
3035 struct btrfs_chunk *chunk,
3036 u64 chunk_offset,
3037 struct btrfs_balance_args *bargs)
3038 {
3039 if (chunk_offset < bargs->vend &&
3040 chunk_offset + btrfs_chunk_length(leaf, chunk) > bargs->vstart)
3041 /* at least part of the chunk is inside this vrange */
3042 return 0;
3043
3044 return 1;
3045 }
3046
3047 static int chunk_soft_convert_filter(u64 chunk_type,
3048 struct btrfs_balance_args *bargs)
3049 {
3050 if (!(bargs->flags & BTRFS_BALANCE_ARGS_CONVERT))
3051 return 0;
3052
3053 chunk_type = chunk_to_extended(chunk_type) &
3054 BTRFS_EXTENDED_PROFILE_MASK;
3055
3056 if (bargs->target == chunk_type)
3057 return 1;
3058
3059 return 0;
3060 }
3061
3062 static int should_balance_chunk(struct btrfs_root *root,
3063 struct extent_buffer *leaf,
3064 struct btrfs_chunk *chunk, u64 chunk_offset)
3065 {
3066 struct btrfs_balance_control *bctl = root->fs_info->balance_ctl;
3067 struct btrfs_balance_args *bargs = NULL;
3068 u64 chunk_type = btrfs_chunk_type(leaf, chunk);
3069
3070 /* type filter */
3071 if (!((chunk_type & BTRFS_BLOCK_GROUP_TYPE_MASK) &
3072 (bctl->flags & BTRFS_BALANCE_TYPE_MASK))) {
3073 return 0;
3074 }
3075
3076 if (chunk_type & BTRFS_BLOCK_GROUP_DATA)
3077 bargs = &bctl->data;
3078 else if (chunk_type & BTRFS_BLOCK_GROUP_SYSTEM)
3079 bargs = &bctl->sys;
3080 else if (chunk_type & BTRFS_BLOCK_GROUP_METADATA)
3081 bargs = &bctl->meta;
3082
3083 /* profiles filter */
3084 if ((bargs->flags & BTRFS_BALANCE_ARGS_PROFILES) &&
3085 chunk_profiles_filter(chunk_type, bargs)) {
3086 return 0;
3087 }
3088
3089 /* usage filter */
3090 if ((bargs->flags & BTRFS_BALANCE_ARGS_USAGE) &&
3091 chunk_usage_filter(bctl->fs_info, chunk_offset, bargs)) {
3092 return 0;
3093 }
3094
3095 /* devid filter */
3096 if ((bargs->flags & BTRFS_BALANCE_ARGS_DEVID) &&
3097 chunk_devid_filter(leaf, chunk, bargs)) {
3098 return 0;
3099 }
3100
3101 /* drange filter, makes sense only with devid filter */
3102 if ((bargs->flags & BTRFS_BALANCE_ARGS_DRANGE) &&
3103 chunk_drange_filter(leaf, chunk, chunk_offset, bargs)) {
3104 return 0;
3105 }
3106
3107 /* vrange filter */
3108 if ((bargs->flags & BTRFS_BALANCE_ARGS_VRANGE) &&
3109 chunk_vrange_filter(leaf, chunk, chunk_offset, bargs)) {
3110 return 0;
3111 }
3112
3113 /* soft profile changing mode */
3114 if ((bargs->flags & BTRFS_BALANCE_ARGS_SOFT) &&
3115 chunk_soft_convert_filter(chunk_type, bargs)) {
3116 return 0;
3117 }
3118
3119 /*
3120 * limited by count, must be the last filter
3121 */
3122 if ((bargs->flags & BTRFS_BALANCE_ARGS_LIMIT)) {
3123 if (bargs->limit == 0)
3124 return 0;
3125 else
3126 bargs->limit--;
3127 }
3128
3129 return 1;
3130 }
3131
3132 static int __btrfs_balance(struct btrfs_fs_info *fs_info)
3133 {
3134 struct btrfs_balance_control *bctl = fs_info->balance_ctl;
3135 struct btrfs_root *chunk_root = fs_info->chunk_root;
3136 struct btrfs_root *dev_root = fs_info->dev_root;
3137 struct list_head *devices;
3138 struct btrfs_device *device;
3139 u64 old_size;
3140 u64 size_to_free;
3141 struct btrfs_chunk *chunk;
3142 struct btrfs_path *path;
3143 struct btrfs_key key;
3144 struct btrfs_key found_key;
3145 struct btrfs_trans_handle *trans;
3146 struct extent_buffer *leaf;
3147 int slot;
3148 int ret;
3149 int enospc_errors = 0;
3150 bool counting = true;
3151 u64 limit_data = bctl->data.limit;
3152 u64 limit_meta = bctl->meta.limit;
3153 u64 limit_sys = bctl->sys.limit;
3154
3155 /* step one make some room on all the devices */
3156 devices = &fs_info->fs_devices->devices;
3157 list_for_each_entry(device, devices, dev_list) {
3158 old_size = btrfs_device_get_total_bytes(device);
3159 size_to_free = div_factor(old_size, 1);
3160 size_to_free = min(size_to_free, (u64)1 * 1024 * 1024);
3161 if (!device->writeable ||
3162 btrfs_device_get_total_bytes(device) -
3163 btrfs_device_get_bytes_used(device) > size_to_free ||
3164 device->is_tgtdev_for_dev_replace)
3165 continue;
3166
3167 ret = btrfs_shrink_device(device, old_size - size_to_free);
3168 if (ret == -ENOSPC)
3169 break;
3170 BUG_ON(ret);
3171
3172 trans = btrfs_start_transaction(dev_root, 0);
3173 BUG_ON(IS_ERR(trans));
3174
3175 ret = btrfs_grow_device(trans, device, old_size);
3176 BUG_ON(ret);
3177
3178 btrfs_end_transaction(trans, dev_root);
3179 }
3180
3181 /* step two, relocate all the chunks */
3182 path = btrfs_alloc_path();
3183 if (!path) {
3184 ret = -ENOMEM;
3185 goto error;
3186 }
3187
3188 /* zero out stat counters */
3189 spin_lock(&fs_info->balance_lock);
3190 memset(&bctl->stat, 0, sizeof(bctl->stat));
3191 spin_unlock(&fs_info->balance_lock);
3192 again:
3193 if (!counting) {
3194 bctl->data.limit = limit_data;
3195 bctl->meta.limit = limit_meta;
3196 bctl->sys.limit = limit_sys;
3197 }
3198 key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
3199 key.offset = (u64)-1;
3200 key.type = BTRFS_CHUNK_ITEM_KEY;
3201
3202 while (1) {
3203 if ((!counting && atomic_read(&fs_info->balance_pause_req)) ||
3204 atomic_read(&fs_info->balance_cancel_req)) {
3205 ret = -ECANCELED;
3206 goto error;
3207 }
3208
3209 ret = btrfs_search_slot(NULL, chunk_root, &key, path, 0, 0);
3210 if (ret < 0)
3211 goto error;
3212
3213 /*
3214 * this shouldn't happen, it means the last relocate
3215 * failed
3216 */
3217 if (ret == 0)
3218 BUG(); /* FIXME break ? */
3219
3220 ret = btrfs_previous_item(chunk_root, path, 0,
3221 BTRFS_CHUNK_ITEM_KEY);
3222 if (ret) {
3223 ret = 0;
3224 break;
3225 }
3226
3227 leaf = path->nodes[0];
3228 slot = path->slots[0];
3229 btrfs_item_key_to_cpu(leaf, &found_key, slot);
3230
3231 if (found_key.objectid != key.objectid)
3232 break;
3233
3234 chunk = btrfs_item_ptr(leaf, slot, struct btrfs_chunk);
3235
3236 if (!counting) {
3237 spin_lock(&fs_info->balance_lock);
3238 bctl->stat.considered++;
3239 spin_unlock(&fs_info->balance_lock);
3240 }
3241
3242 ret = should_balance_chunk(chunk_root, leaf, chunk,
3243 found_key.offset);
3244 btrfs_release_path(path);
3245 if (!ret)
3246 goto loop;
3247
3248 if (counting) {
3249 spin_lock(&fs_info->balance_lock);
3250 bctl->stat.expected++;
3251 spin_unlock(&fs_info->balance_lock);
3252 goto loop;
3253 }
3254
3255 ret = btrfs_relocate_chunk(chunk_root,
3256 chunk_root->root_key.objectid,
3257 found_key.objectid,
3258 found_key.offset);
3259 if (ret && ret != -ENOSPC)
3260 goto error;
3261 if (ret == -ENOSPC) {
3262 enospc_errors++;
3263 } else {
3264 spin_lock(&fs_info->balance_lock);
3265 bctl->stat.completed++;
3266 spin_unlock(&fs_info->balance_lock);
3267 }
3268 loop:
3269 if (found_key.offset == 0)
3270 break;
3271 key.offset = found_key.offset - 1;
3272 }
3273
3274 if (counting) {
3275 btrfs_release_path(path);
3276 counting = false;
3277 goto again;
3278 }
3279 error:
3280 btrfs_free_path(path);
3281 if (enospc_errors) {
3282 btrfs_info(fs_info, "%d enospc errors during balance",
3283 enospc_errors);
3284 if (!ret)
3285 ret = -ENOSPC;
3286 }
3287
3288 return ret;
3289 }
3290
3291 /**
3292 * alloc_profile_is_valid - see if a given profile is valid and reduced
3293 * @flags: profile to validate
3294 * @extended: if true @flags is treated as an extended profile
3295 */
3296 static int alloc_profile_is_valid(u64 flags, int extended)
3297 {
3298 u64 mask = (extended ? BTRFS_EXTENDED_PROFILE_MASK :
3299 BTRFS_BLOCK_GROUP_PROFILE_MASK);
3300
3301 flags &= ~BTRFS_BLOCK_GROUP_TYPE_MASK;
3302
3303 /* 1) check that all other bits are zeroed */
3304 if (flags & ~mask)
3305 return 0;
3306
3307 /* 2) see if profile is reduced */
3308 if (flags == 0)
3309 return !extended; /* "0" is valid for usual profiles */
3310
3311 /* true if exactly one bit set */
3312 return (flags & (flags - 1)) == 0;
3313 }
3314
3315 static inline int balance_need_close(struct btrfs_fs_info *fs_info)
3316 {
3317 /* cancel requested || normal exit path */
3318 return atomic_read(&fs_info->balance_cancel_req) ||
3319 (atomic_read(&fs_info->balance_pause_req) == 0 &&
3320 atomic_read(&fs_info->balance_cancel_req) == 0);
3321 }
3322
3323 static void __cancel_balance(struct btrfs_fs_info *fs_info)
3324 {
3325 int ret;
3326
3327 unset_balance_control(fs_info);
3328 ret = del_balance_item(fs_info->tree_root);
3329 if (ret)
3330 btrfs_std_error(fs_info, ret);
3331
3332 atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
3333 }
3334
3335 /*
3336 * Should be called with both balance and volume mutexes held
3337 */
3338 int btrfs_balance(struct btrfs_balance_control *bctl,
3339 struct btrfs_ioctl_balance_args *bargs)
3340 {
3341 struct btrfs_fs_info *fs_info = bctl->fs_info;
3342 u64 allowed;
3343 int mixed = 0;
3344 int ret;
3345 u64 num_devices;
3346 unsigned seq;
3347
3348 if (btrfs_fs_closing(fs_info) ||
3349 atomic_read(&fs_info->balance_pause_req) ||
3350 atomic_read(&fs_info->balance_cancel_req)) {
3351 ret = -EINVAL;
3352 goto out;
3353 }
3354
3355 allowed = btrfs_super_incompat_flags(fs_info->super_copy);
3356 if (allowed & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS)
3357 mixed = 1;
3358
3359 /*
3360 * In case of mixed groups both data and meta should be picked,
3361 * and identical options should be given for both of them.
3362 */
3363 allowed = BTRFS_BALANCE_DATA | BTRFS_BALANCE_METADATA;
3364 if (mixed && (bctl->flags & allowed)) {
3365 if (!(bctl->flags & BTRFS_BALANCE_DATA) ||
3366 !(bctl->flags & BTRFS_BALANCE_METADATA) ||
3367 memcmp(&bctl->data, &bctl->meta, sizeof(bctl->data))) {
3368 btrfs_err(fs_info, "with mixed groups data and "
3369 "metadata balance options must be the same");
3370 ret = -EINVAL;
3371 goto out;
3372 }
3373 }
3374
3375 num_devices = fs_info->fs_devices->num_devices;
3376 btrfs_dev_replace_lock(&fs_info->dev_replace);
3377 if (btrfs_dev_replace_is_ongoing(&fs_info->dev_replace)) {
3378 BUG_ON(num_devices < 1);
3379 num_devices--;
3380 }
3381 btrfs_dev_replace_unlock(&fs_info->dev_replace);
3382 allowed = BTRFS_AVAIL_ALLOC_BIT_SINGLE;
3383 if (num_devices == 1)
3384 allowed |= BTRFS_BLOCK_GROUP_DUP;
3385 else if (num_devices > 1)
3386 allowed |= (BTRFS_BLOCK_GROUP_RAID0 | BTRFS_BLOCK_GROUP_RAID1);
3387 if (num_devices > 2)
3388 allowed |= BTRFS_BLOCK_GROUP_RAID5;
3389 if (num_devices > 3)
3390 allowed |= (BTRFS_BLOCK_GROUP_RAID10 |
3391 BTRFS_BLOCK_GROUP_RAID6);
3392 if ((bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
3393 (!alloc_profile_is_valid(bctl->data.target, 1) ||
3394 (bctl->data.target & ~allowed))) {
3395 btrfs_err(fs_info, "unable to start balance with target "
3396 "data profile %llu",
3397 bctl->data.target);
3398 ret = -EINVAL;
3399 goto out;
3400 }
3401 if ((bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
3402 (!alloc_profile_is_valid(bctl->meta.target, 1) ||
3403 (bctl->meta.target & ~allowed))) {
3404 btrfs_err(fs_info,
3405 "unable to start balance with target metadata profile %llu",
3406 bctl->meta.target);
3407 ret = -EINVAL;
3408 goto out;
3409 }
3410 if ((bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
3411 (!alloc_profile_is_valid(bctl->sys.target, 1) ||
3412 (bctl->sys.target & ~allowed))) {
3413 btrfs_err(fs_info,
3414 "unable to start balance with target system profile %llu",
3415 bctl->sys.target);
3416 ret = -EINVAL;
3417 goto out;
3418 }
3419
3420 /* allow dup'ed data chunks only in mixed mode */
3421 if (!mixed && (bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
3422 (bctl->data.target & BTRFS_BLOCK_GROUP_DUP)) {
3423 btrfs_err(fs_info, "dup for data is not allowed");
3424 ret = -EINVAL;
3425 goto out;
3426 }
3427
3428 /* allow to reduce meta or sys integrity only if force set */
3429 allowed = BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1 |
3430 BTRFS_BLOCK_GROUP_RAID10 |
3431 BTRFS_BLOCK_GROUP_RAID5 |
3432 BTRFS_BLOCK_GROUP_RAID6;
3433 do {
3434 seq = read_seqbegin(&fs_info->profiles_lock);
3435
3436 if (((bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
3437 (fs_info->avail_system_alloc_bits & allowed) &&
3438 !(bctl->sys.target & allowed)) ||
3439 ((bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
3440 (fs_info->avail_metadata_alloc_bits & allowed) &&
3441 !(bctl->meta.target & allowed))) {
3442 if (bctl->flags & BTRFS_BALANCE_FORCE) {
3443 btrfs_info(fs_info, "force reducing metadata integrity");
3444 } else {
3445 btrfs_err(fs_info, "balance will reduce metadata "
3446 "integrity, use force if you want this");
3447 ret = -EINVAL;
3448 goto out;
3449 }
3450 }
3451 } while (read_seqretry(&fs_info->profiles_lock, seq));
3452
3453 if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) {
3454 int num_tolerated_disk_barrier_failures;
3455 u64 target = bctl->sys.target;
3456
3457 num_tolerated_disk_barrier_failures =
3458 btrfs_calc_num_tolerated_disk_barrier_failures(fs_info);
3459 if (num_tolerated_disk_barrier_failures > 0 &&
3460 (target &
3461 (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID0 |
3462 BTRFS_AVAIL_ALLOC_BIT_SINGLE)))
3463 num_tolerated_disk_barrier_failures = 0;
3464 else if (num_tolerated_disk_barrier_failures > 1 &&
3465 (target &
3466 (BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID10)))
3467 num_tolerated_disk_barrier_failures = 1;
3468
3469 fs_info->num_tolerated_disk_barrier_failures =
3470 num_tolerated_disk_barrier_failures;
3471 }
3472
3473 ret = insert_balance_item(fs_info->tree_root, bctl);
3474 if (ret && ret != -EEXIST)
3475 goto out;
3476
3477 if (!(bctl->flags & BTRFS_BALANCE_RESUME)) {
3478 BUG_ON(ret == -EEXIST);
3479 set_balance_control(bctl);
3480 } else {
3481 BUG_ON(ret != -EEXIST);
3482 spin_lock(&fs_info->balance_lock);
3483 update_balance_args(bctl);
3484 spin_unlock(&fs_info->balance_lock);
3485 }
3486
3487 atomic_inc(&fs_info->balance_running);
3488 mutex_unlock(&fs_info->balance_mutex);
3489
3490 ret = __btrfs_balance(fs_info);
3491
3492 mutex_lock(&fs_info->balance_mutex);
3493 atomic_dec(&fs_info->balance_running);
3494
3495 if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) {
3496 fs_info->num_tolerated_disk_barrier_failures =
3497 btrfs_calc_num_tolerated_disk_barrier_failures(fs_info);
3498 }
3499
3500 if (bargs) {
3501 memset(bargs, 0, sizeof(*bargs));
3502 update_ioctl_balance_args(fs_info, 0, bargs);
3503 }
3504
3505 if ((ret && ret != -ECANCELED && ret != -ENOSPC) ||
3506 balance_need_close(fs_info)) {
3507 __cancel_balance(fs_info);
3508 }
3509
3510 wake_up(&fs_info->balance_wait_q);
3511
3512 return ret;
3513 out:
3514 if (bctl->flags & BTRFS_BALANCE_RESUME)
3515 __cancel_balance(fs_info);
3516 else {
3517 kfree(bctl);
3518 atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
3519 }
3520 return ret;
3521 }
3522
3523 static int balance_kthread(void *data)
3524 {
3525 struct btrfs_fs_info *fs_info = data;
3526 int ret = 0;
3527
3528 mutex_lock(&fs_info->volume_mutex);
3529 mutex_lock(&fs_info->balance_mutex);
3530
3531 if (fs_info->balance_ctl) {
3532 btrfs_info(fs_info, "continuing balance");
3533 ret = btrfs_balance(fs_info->balance_ctl, NULL);
3534 }
3535
3536 mutex_unlock(&fs_info->balance_mutex);
3537 mutex_unlock(&fs_info->volume_mutex);
3538
3539 return ret;
3540 }
3541
3542 int btrfs_resume_balance_async(struct btrfs_fs_info *fs_info)
3543 {
3544 struct task_struct *tsk;
3545
3546 spin_lock(&fs_info->balance_lock);
3547 if (!fs_info->balance_ctl) {
3548 spin_unlock(&fs_info->balance_lock);
3549 return 0;
3550 }
3551 spin_unlock(&fs_info->balance_lock);
3552
3553 if (btrfs_test_opt(fs_info->tree_root, SKIP_BALANCE)) {
3554 btrfs_info(fs_info, "force skipping balance");
3555 return 0;
3556 }
3557
3558 tsk = kthread_run(balance_kthread, fs_info, "btrfs-balance");
3559 return PTR_ERR_OR_ZERO(tsk);
3560 }
3561
3562 int btrfs_recover_balance(struct btrfs_fs_info *fs_info)
3563 {
3564 struct btrfs_balance_control *bctl;
3565 struct btrfs_balance_item *item;
3566 struct btrfs_disk_balance_args disk_bargs;
3567 struct btrfs_path *path;
3568 struct extent_buffer *leaf;
3569 struct btrfs_key key;
3570 int ret;
3571
3572 path = btrfs_alloc_path();
3573 if (!path)
3574 return -ENOMEM;
3575
3576 key.objectid = BTRFS_BALANCE_OBJECTID;
3577 key.type = BTRFS_BALANCE_ITEM_KEY;
3578 key.offset = 0;
3579
3580 ret = btrfs_search_slot(NULL, fs_info->tree_root, &key, path, 0, 0);
3581 if (ret < 0)
3582 goto out;
3583 if (ret > 0) { /* ret = -ENOENT; */
3584 ret = 0;
3585 goto out;
3586 }
3587
3588 bctl = kzalloc(sizeof(*bctl), GFP_NOFS);
3589 if (!bctl) {
3590 ret = -ENOMEM;
3591 goto out;
3592 }
3593
3594 leaf = path->nodes[0];
3595 item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_balance_item);
3596
3597 bctl->fs_info = fs_info;
3598 bctl->flags = btrfs_balance_flags(leaf, item);
3599 bctl->flags |= BTRFS_BALANCE_RESUME;
3600
3601 btrfs_balance_data(leaf, item, &disk_bargs);
3602 btrfs_disk_balance_args_to_cpu(&bctl->data, &disk_bargs);
3603 btrfs_balance_meta(leaf, item, &disk_bargs);
3604 btrfs_disk_balance_args_to_cpu(&bctl->meta, &disk_bargs);
3605 btrfs_balance_sys(leaf, item, &disk_bargs);
3606 btrfs_disk_balance_args_to_cpu(&bctl->sys, &disk_bargs);
3607
3608 WARN_ON(atomic_xchg(&fs_info->mutually_exclusive_operation_running, 1));
3609
3610 mutex_lock(&fs_info->volume_mutex);
3611 mutex_lock(&fs_info->balance_mutex);
3612
3613 set_balance_control(bctl);
3614
3615 mutex_unlock(&fs_info->balance_mutex);
3616 mutex_unlock(&fs_info->volume_mutex);
3617 out:
3618 btrfs_free_path(path);
3619 return ret;
3620 }
3621
3622 int btrfs_pause_balance(struct btrfs_fs_info *fs_info)
3623 {
3624 int ret = 0;
3625
3626 mutex_lock(&fs_info->balance_mutex);
3627 if (!fs_info->balance_ctl) {
3628 mutex_unlock(&fs_info->balance_mutex);
3629 return -ENOTCONN;
3630 }
3631
3632 if (atomic_read(&fs_info->balance_running)) {
3633 atomic_inc(&fs_info->balance_pause_req);
3634 mutex_unlock(&fs_info->balance_mutex);
3635
3636 wait_event(fs_info->balance_wait_q,
3637 atomic_read(&fs_info->balance_running) == 0);
3638
3639 mutex_lock(&fs_info->balance_mutex);
3640 /* we are good with balance_ctl ripped off from under us */
3641 BUG_ON(atomic_read(&fs_info->balance_running));
3642 atomic_dec(&fs_info->balance_pause_req);
3643 } else {
3644 ret = -ENOTCONN;
3645 }
3646
3647 mutex_unlock(&fs_info->balance_mutex);
3648 return ret;
3649 }
3650
3651 int btrfs_cancel_balance(struct btrfs_fs_info *fs_info)
3652 {
3653 if (fs_info->sb->s_flags & MS_RDONLY)
3654 return -EROFS;
3655
3656 mutex_lock(&fs_info->balance_mutex);
3657 if (!fs_info->balance_ctl) {
3658 mutex_unlock(&fs_info->balance_mutex);
3659 return -ENOTCONN;
3660 }
3661
3662 atomic_inc(&fs_info->balance_cancel_req);
3663 /*
3664 * if we are running just wait and return, balance item is
3665 * deleted in btrfs_balance in this case
3666 */
3667 if (atomic_read(&fs_info->balance_running)) {
3668 mutex_unlock(&fs_info->balance_mutex);
3669 wait_event(fs_info->balance_wait_q,
3670 atomic_read(&fs_info->balance_running) == 0);
3671 mutex_lock(&fs_info->balance_mutex);
3672 } else {
3673 /* __cancel_balance needs volume_mutex */
3674 mutex_unlock(&fs_info->balance_mutex);
3675 mutex_lock(&fs_info->volume_mutex);
3676 mutex_lock(&fs_info->balance_mutex);
3677
3678 if (fs_info->balance_ctl)
3679 __cancel_balance(fs_info);
3680
3681 mutex_unlock(&fs_info->volume_mutex);
3682 }
3683
3684 BUG_ON(fs_info->balance_ctl || atomic_read(&fs_info->balance_running));
3685 atomic_dec(&fs_info->balance_cancel_req);
3686 mutex_unlock(&fs_info->balance_mutex);
3687 return 0;
3688 }
3689
3690 static int btrfs_uuid_scan_kthread(void *data)
3691 {
3692 struct btrfs_fs_info *fs_info = data;
3693 struct btrfs_root *root = fs_info->tree_root;
3694 struct btrfs_key key;
3695 struct btrfs_key max_key;
3696 struct btrfs_path *path = NULL;
3697 int ret = 0;
3698 struct extent_buffer *eb;
3699 int slot;
3700 struct btrfs_root_item root_item;
3701 u32 item_size;
3702 struct btrfs_trans_handle *trans = NULL;
3703
3704 path = btrfs_alloc_path();
3705 if (!path) {
3706 ret = -ENOMEM;
3707 goto out;
3708 }
3709
3710 key.objectid = 0;
3711 key.type = BTRFS_ROOT_ITEM_KEY;
3712 key.offset = 0;
3713
3714 max_key.objectid = (u64)-1;
3715 max_key.type = BTRFS_ROOT_ITEM_KEY;
3716 max_key.offset = (u64)-1;
3717
3718 while (1) {
3719 ret = btrfs_search_forward(root, &key, path, 0);
3720 if (ret) {
3721 if (ret > 0)
3722 ret = 0;
3723 break;
3724 }
3725
3726 if (key.type != BTRFS_ROOT_ITEM_KEY ||
3727 (key.objectid < BTRFS_FIRST_FREE_OBJECTID &&
3728 key.objectid != BTRFS_FS_TREE_OBJECTID) ||
3729 key.objectid > BTRFS_LAST_FREE_OBJECTID)
3730 goto skip;
3731
3732 eb = path->nodes[0];
3733 slot = path->slots[0];
3734 item_size = btrfs_item_size_nr(eb, slot);
3735 if (item_size < sizeof(root_item))
3736 goto skip;
3737
3738 read_extent_buffer(eb, &root_item,
3739 btrfs_item_ptr_offset(eb, slot),
3740 (int)sizeof(root_item));
3741 if (btrfs_root_refs(&root_item) == 0)
3742 goto skip;
3743
3744 if (!btrfs_is_empty_uuid(root_item.uuid) ||
3745 !btrfs_is_empty_uuid(root_item.received_uuid)) {
3746 if (trans)
3747 goto update_tree;
3748
3749 btrfs_release_path(path);
3750 /*
3751 * 1 - subvol uuid item
3752 * 1 - received_subvol uuid item
3753 */
3754 trans = btrfs_start_transaction(fs_info->uuid_root, 2);
3755 if (IS_ERR(trans)) {
3756 ret = PTR_ERR(trans);
3757 break;
3758 }
3759 continue;
3760 } else {
3761 goto skip;
3762 }
3763 update_tree:
3764 if (!btrfs_is_empty_uuid(root_item.uuid)) {
3765 ret = btrfs_uuid_tree_add(trans, fs_info->uuid_root,
3766 root_item.uuid,
3767 BTRFS_UUID_KEY_SUBVOL,
3768 key.objectid);
3769 if (ret < 0) {
3770 btrfs_warn(fs_info, "uuid_tree_add failed %d",
3771 ret);
3772 break;
3773 }
3774 }
3775
3776 if (!btrfs_is_empty_uuid(root_item.received_uuid)) {
3777 ret = btrfs_uuid_tree_add(trans, fs_info->uuid_root,
3778 root_item.received_uuid,
3779 BTRFS_UUID_KEY_RECEIVED_SUBVOL,
3780 key.objectid);
3781 if (ret < 0) {
3782 btrfs_warn(fs_info, "uuid_tree_add failed %d",
3783 ret);
3784 break;
3785 }
3786 }
3787
3788 skip:
3789 if (trans) {
3790 ret = btrfs_end_transaction(trans, fs_info->uuid_root);
3791 trans = NULL;
3792 if (ret)
3793 break;
3794 }
3795
3796 btrfs_release_path(path);
3797 if (key.offset < (u64)-1) {
3798 key.offset++;
3799 } else if (key.type < BTRFS_ROOT_ITEM_KEY) {
3800 key.offset = 0;
3801 key.type = BTRFS_ROOT_ITEM_KEY;
3802 } else if (key.objectid < (u64)-1) {
3803 key.offset = 0;
3804 key.type = BTRFS_ROOT_ITEM_KEY;
3805 key.objectid++;
3806 } else {
3807 break;
3808 }
3809 cond_resched();
3810 }
3811
3812 out:
3813 btrfs_free_path(path);
3814 if (trans && !IS_ERR(trans))
3815 btrfs_end_transaction(trans, fs_info->uuid_root);
3816 if (ret)
3817 btrfs_warn(fs_info, "btrfs_uuid_scan_kthread failed %d", ret);
3818 else
3819 fs_info->update_uuid_tree_gen = 1;
3820 up(&fs_info->uuid_tree_rescan_sem);
3821 return 0;
3822 }
3823
3824 /*
3825 * Callback for btrfs_uuid_tree_iterate().
3826 * returns:
3827 * 0 check succeeded, the entry is not outdated.
3828 * < 0 if an error occured.
3829 * > 0 if the check failed, which means the caller shall remove the entry.
3830 */
3831 static int btrfs_check_uuid_tree_entry(struct btrfs_fs_info *fs_info,
3832 u8 *uuid, u8 type, u64 subid)
3833 {
3834 struct btrfs_key key;
3835 int ret = 0;
3836 struct btrfs_root *subvol_root;
3837
3838 if (type != BTRFS_UUID_KEY_SUBVOL &&
3839 type != BTRFS_UUID_KEY_RECEIVED_SUBVOL)
3840 goto out;
3841
3842 key.objectid = subid;
3843 key.type = BTRFS_ROOT_ITEM_KEY;
3844 key.offset = (u64)-1;
3845 subvol_root = btrfs_read_fs_root_no_name(fs_info, &key);
3846 if (IS_ERR(subvol_root)) {
3847 ret = PTR_ERR(subvol_root);
3848 if (ret == -ENOENT)
3849 ret = 1;
3850 goto out;
3851 }
3852
3853 switch (type) {
3854 case BTRFS_UUID_KEY_SUBVOL:
3855 if (memcmp(uuid, subvol_root->root_item.uuid, BTRFS_UUID_SIZE))
3856 ret = 1;
3857 break;
3858 case BTRFS_UUID_KEY_RECEIVED_SUBVOL:
3859 if (memcmp(uuid, subvol_root->root_item.received_uuid,
3860 BTRFS_UUID_SIZE))
3861 ret = 1;
3862 break;
3863 }
3864
3865 out:
3866 return ret;
3867 }
3868
3869 static int btrfs_uuid_rescan_kthread(void *data)
3870 {
3871 struct btrfs_fs_info *fs_info = (struct btrfs_fs_info *)data;
3872 int ret;
3873
3874 /*
3875 * 1st step is to iterate through the existing UUID tree and
3876 * to delete all entries that contain outdated data.
3877 * 2nd step is to add all missing entries to the UUID tree.
3878 */
3879 ret = btrfs_uuid_tree_iterate(fs_info, btrfs_check_uuid_tree_entry);
3880 if (ret < 0) {
3881 btrfs_warn(fs_info, "iterating uuid_tree failed %d", ret);
3882 up(&fs_info->uuid_tree_rescan_sem);
3883 return ret;
3884 }
3885 return btrfs_uuid_scan_kthread(data);
3886 }
3887
3888 int btrfs_create_uuid_tree(struct btrfs_fs_info *fs_info)
3889 {
3890 struct btrfs_trans_handle *trans;
3891 struct btrfs_root *tree_root = fs_info->tree_root;
3892 struct btrfs_root *uuid_root;
3893 struct task_struct *task;
3894 int ret;
3895
3896 /*
3897 * 1 - root node
3898 * 1 - root item
3899 */
3900 trans = btrfs_start_transaction(tree_root, 2);
3901 if (IS_ERR(trans))
3902 return PTR_ERR(trans);
3903
3904 uuid_root = btrfs_create_tree(trans, fs_info,
3905 BTRFS_UUID_TREE_OBJECTID);
3906 if (IS_ERR(uuid_root)) {
3907 btrfs_abort_transaction(trans, tree_root,
3908 PTR_ERR(uuid_root));
3909 return PTR_ERR(uuid_root);
3910 }
3911
3912 fs_info->uuid_root = uuid_root;
3913
3914 ret = btrfs_commit_transaction(trans, tree_root);
3915 if (ret)
3916 return ret;
3917
3918 down(&fs_info->uuid_tree_rescan_sem);
3919 task = kthread_run(btrfs_uuid_scan_kthread, fs_info, "btrfs-uuid");
3920 if (IS_ERR(task)) {
3921 /* fs_info->update_uuid_tree_gen remains 0 in all error case */
3922 btrfs_warn(fs_info, "failed to start uuid_scan task");
3923 up(&fs_info->uuid_tree_rescan_sem);
3924 return PTR_ERR(task);
3925 }
3926
3927 return 0;
3928 }
3929
3930 int btrfs_check_uuid_tree(struct btrfs_fs_info *fs_info)
3931 {
3932 struct task_struct *task;
3933
3934 down(&fs_info->uuid_tree_rescan_sem);
3935 task = kthread_run(btrfs_uuid_rescan_kthread, fs_info, "btrfs-uuid");
3936 if (IS_ERR(task)) {
3937 /* fs_info->update_uuid_tree_gen remains 0 in all error case */
3938 btrfs_warn(fs_info, "failed to start uuid_rescan task");
3939 up(&fs_info->uuid_tree_rescan_sem);
3940 return PTR_ERR(task);
3941 }
3942
3943 return 0;
3944 }
3945
3946 /*
3947 * shrinking a device means finding all of the device extents past
3948 * the new size, and then following the back refs to the chunks.
3949 * The chunk relocation code actually frees the device extent
3950 */
3951 int btrfs_shrink_device(struct btrfs_device *device, u64 new_size)
3952 {
3953 struct btrfs_trans_handle *trans;
3954 struct btrfs_root *root = device->dev_root;
3955 struct btrfs_dev_extent *dev_extent = NULL;
3956 struct btrfs_path *path;
3957 u64 length;
3958 u64 chunk_tree;
3959 u64 chunk_objectid;
3960 u64 chunk_offset;
3961 int ret;
3962 int slot;
3963 int failed = 0;
3964 bool retried = false;
3965 struct extent_buffer *l;
3966 struct btrfs_key key;
3967 struct btrfs_super_block *super_copy = root->fs_info->super_copy;
3968 u64 old_total = btrfs_super_total_bytes(super_copy);
3969 u64 old_size = btrfs_device_get_total_bytes(device);
3970 u64 diff = old_size - new_size;
3971
3972 if (device->is_tgtdev_for_dev_replace)
3973 return -EINVAL;
3974
3975 path = btrfs_alloc_path();
3976 if (!path)
3977 return -ENOMEM;
3978
3979 path->reada = 2;
3980
3981 lock_chunks(root);
3982
3983 btrfs_device_set_total_bytes(device, new_size);
3984 if (device->writeable) {
3985 device->fs_devices->total_rw_bytes -= diff;
3986 spin_lock(&root->fs_info->free_chunk_lock);
3987 root->fs_info->free_chunk_space -= diff;
3988 spin_unlock(&root->fs_info->free_chunk_lock);
3989 }
3990 unlock_chunks(root);
3991
3992 again:
3993 key.objectid = device->devid;
3994 key.offset = (u64)-1;
3995 key.type = BTRFS_DEV_EXTENT_KEY;
3996
3997 do {
3998 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
3999 if (ret < 0)
4000 goto done;
4001
4002 ret = btrfs_previous_item(root, path, 0, key.type);
4003 if (ret < 0)
4004 goto done;
4005 if (ret) {
4006 ret = 0;
4007 btrfs_release_path(path);
4008 break;
4009 }
4010
4011 l = path->nodes[0];
4012 slot = path->slots[0];
4013 btrfs_item_key_to_cpu(l, &key, path->slots[0]);
4014
4015 if (key.objectid != device->devid) {
4016 btrfs_release_path(path);
4017 break;
4018 }
4019
4020 dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
4021 length = btrfs_dev_extent_length(l, dev_extent);
4022
4023 if (key.offset + length <= new_size) {
4024 btrfs_release_path(path);
4025 break;
4026 }
4027
4028 chunk_tree = btrfs_dev_extent_chunk_tree(l, dev_extent);
4029 chunk_objectid = btrfs_dev_extent_chunk_objectid(l, dev_extent);
4030 chunk_offset = btrfs_dev_extent_chunk_offset(l, dev_extent);
4031 btrfs_release_path(path);
4032
4033 ret = btrfs_relocate_chunk(root, chunk_tree, chunk_objectid,
4034 chunk_offset);
4035 if (ret && ret != -ENOSPC)
4036 goto done;
4037 if (ret == -ENOSPC)
4038 failed++;
4039 } while (key.offset-- > 0);
4040
4041 if (failed && !retried) {
4042 failed = 0;
4043 retried = true;
4044 goto again;
4045 } else if (failed && retried) {
4046 ret = -ENOSPC;
4047 lock_chunks(root);
4048
4049 btrfs_device_set_total_bytes(device, old_size);
4050 if (device->writeable)
4051 device->fs_devices->total_rw_bytes += diff;
4052 spin_lock(&root->fs_info->free_chunk_lock);
4053 root->fs_info->free_chunk_space += diff;
4054 spin_unlock(&root->fs_info->free_chunk_lock);
4055 unlock_chunks(root);
4056 goto done;
4057 }
4058
4059 /* Shrinking succeeded, else we would be at "done". */
4060 trans = btrfs_start_transaction(root, 0);
4061 if (IS_ERR(trans)) {
4062 ret = PTR_ERR(trans);
4063 goto done;
4064 }
4065
4066 lock_chunks(root);
4067 btrfs_device_set_disk_total_bytes(device, new_size);
4068 if (list_empty(&device->resized_list))
4069 list_add_tail(&device->resized_list,
4070 &root->fs_info->fs_devices->resized_devices);
4071
4072 WARN_ON(diff > old_total);
4073 btrfs_set_super_total_bytes(super_copy, old_total - diff);
4074 unlock_chunks(root);
4075
4076 /* Now btrfs_update_device() will change the on-disk size. */
4077 ret = btrfs_update_device(trans, device);
4078 btrfs_end_transaction(trans, root);
4079 done:
4080 btrfs_free_path(path);
4081 return ret;
4082 }
4083
4084 static int btrfs_add_system_chunk(struct btrfs_root *root,
4085 struct btrfs_key *key,
4086 struct btrfs_chunk *chunk, int item_size)
4087 {
4088 struct btrfs_super_block *super_copy = root->fs_info->super_copy;
4089 struct btrfs_disk_key disk_key;
4090 u32 array_size;
4091 u8 *ptr;
4092
4093 lock_chunks(root);
4094 array_size = btrfs_super_sys_array_size(super_copy);
4095 if (array_size + item_size + sizeof(disk_key)
4096 > BTRFS_SYSTEM_CHUNK_ARRAY_SIZE) {
4097 unlock_chunks(root);
4098 return -EFBIG;
4099 }
4100
4101 ptr = super_copy->sys_chunk_array + array_size;
4102 btrfs_cpu_key_to_disk(&disk_key, key);
4103 memcpy(ptr, &disk_key, sizeof(disk_key));
4104 ptr += sizeof(disk_key);
4105 memcpy(ptr, chunk, item_size);
4106 item_size += sizeof(disk_key);
4107 btrfs_set_super_sys_array_size(super_copy, array_size + item_size);
4108 unlock_chunks(root);
4109
4110 return 0;
4111 }
4112
4113 /*
4114 * sort the devices in descending order by max_avail, total_avail
4115 */
4116 static int btrfs_cmp_device_info(const void *a, const void *b)
4117 {
4118 const struct btrfs_device_info *di_a = a;
4119 const struct btrfs_device_info *di_b = b;
4120
4121 if (di_a->max_avail > di_b->max_avail)
4122 return -1;
4123 if (di_a->max_avail < di_b->max_avail)
4124 return 1;
4125 if (di_a->total_avail > di_b->total_avail)
4126 return -1;
4127 if (di_a->total_avail < di_b->total_avail)
4128 return 1;
4129 return 0;
4130 }
4131
4132 static struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = {
4133 [BTRFS_RAID_RAID10] = {
4134 .sub_stripes = 2,
4135 .dev_stripes = 1,
4136 .devs_max = 0, /* 0 == as many as possible */
4137 .devs_min = 4,
4138 .devs_increment = 2,
4139 .ncopies = 2,
4140 },
4141 [BTRFS_RAID_RAID1] = {
4142 .sub_stripes = 1,
4143 .dev_stripes = 1,
4144 .devs_max = 2,
4145 .devs_min = 2,
4146 .devs_increment = 2,
4147 .ncopies = 2,
4148 },
4149 [BTRFS_RAID_DUP] = {
4150 .sub_stripes = 1,
4151 .dev_stripes = 2,
4152 .devs_max = 1,
4153 .devs_min = 1,
4154 .devs_increment = 1,
4155 .ncopies = 2,
4156 },
4157 [BTRFS_RAID_RAID0] = {
4158 .sub_stripes = 1,
4159 .dev_stripes = 1,
4160 .devs_max = 0,
4161 .devs_min = 2,
4162 .devs_increment = 1,
4163 .ncopies = 1,
4164 },
4165 [BTRFS_RAID_SINGLE] = {
4166 .sub_stripes = 1,
4167 .dev_stripes = 1,
4168 .devs_max = 1,
4169 .devs_min = 1,
4170 .devs_increment = 1,
4171 .ncopies = 1,
4172 },
4173 [BTRFS_RAID_RAID5] = {
4174 .sub_stripes = 1,
4175 .dev_stripes = 1,
4176 .devs_max = 0,
4177 .devs_min = 2,
4178 .devs_increment = 1,
4179 .ncopies = 2,
4180 },
4181 [BTRFS_RAID_RAID6] = {
4182 .sub_stripes = 1,
4183 .dev_stripes = 1,
4184 .devs_max = 0,
4185 .devs_min = 3,
4186 .devs_increment = 1,
4187 .ncopies = 3,
4188 },
4189 };
4190
4191 static u32 find_raid56_stripe_len(u32 data_devices, u32 dev_stripe_target)
4192 {
4193 /* TODO allow them to set a preferred stripe size */
4194 return 64 * 1024;
4195 }
4196
4197 static void check_raid56_incompat_flag(struct btrfs_fs_info *info, u64 type)
4198 {
4199 if (!(type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)))
4200 return;
4201
4202 btrfs_set_fs_incompat(info, RAID56);
4203 }
4204
4205 #define BTRFS_MAX_DEVS(r) ((BTRFS_LEAF_DATA_SIZE(r) \
4206 - sizeof(struct btrfs_item) \
4207 - sizeof(struct btrfs_chunk)) \
4208 / sizeof(struct btrfs_stripe) + 1)
4209
4210 #define BTRFS_MAX_DEVS_SYS_CHUNK ((BTRFS_SYSTEM_CHUNK_ARRAY_SIZE \
4211 - 2 * sizeof(struct btrfs_disk_key) \
4212 - 2 * sizeof(struct btrfs_chunk)) \
4213 / sizeof(struct btrfs_stripe) + 1)
4214
4215 static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
4216 struct btrfs_root *extent_root, u64 start,
4217 u64 type)
4218 {
4219 struct btrfs_fs_info *info = extent_root->fs_info;
4220 struct btrfs_fs_devices *fs_devices = info->fs_devices;
4221 struct list_head *cur;
4222 struct map_lookup *map = NULL;
4223 struct extent_map_tree *em_tree;
4224 struct extent_map *em;
4225 struct btrfs_device_info *devices_info = NULL;
4226 u64 total_avail;
4227 int num_stripes; /* total number of stripes to allocate */
4228 int data_stripes; /* number of stripes that count for
4229 block group size */
4230 int sub_stripes; /* sub_stripes info for map */
4231 int dev_stripes; /* stripes per dev */
4232 int devs_max; /* max devs to use */
4233 int devs_min; /* min devs needed */
4234 int devs_increment; /* ndevs has to be a multiple of this */
4235 int ncopies; /* how many copies to data has */
4236 int ret;
4237 u64 max_stripe_size;
4238 u64 max_chunk_size;
4239 u64 stripe_size;
4240 u64 num_bytes;
4241 u64 raid_stripe_len = BTRFS_STRIPE_LEN;
4242 int ndevs;
4243 int i;
4244 int j;
4245 int index;
4246
4247 BUG_ON(!alloc_profile_is_valid(type, 0));
4248
4249 if (list_empty(&fs_devices->alloc_list))
4250 return -ENOSPC;
4251
4252 index = __get_raid_index(type);
4253
4254 sub_stripes = btrfs_raid_array[index].sub_stripes;
4255 dev_stripes = btrfs_raid_array[index].dev_stripes;
4256 devs_max = btrfs_raid_array[index].devs_max;
4257 devs_min = btrfs_raid_array[index].devs_min;
4258 devs_increment = btrfs_raid_array[index].devs_increment;
4259 ncopies = btrfs_raid_array[index].ncopies;
4260
4261 if (type & BTRFS_BLOCK_GROUP_DATA) {
4262 max_stripe_size = 1024 * 1024 * 1024;
4263 max_chunk_size = 10 * max_stripe_size;
4264 if (!devs_max)
4265 devs_max = BTRFS_MAX_DEVS(info->chunk_root);
4266 } else if (type & BTRFS_BLOCK_GROUP_METADATA) {
4267 /* for larger filesystems, use larger metadata chunks */
4268 if (fs_devices->total_rw_bytes > 50ULL * 1024 * 1024 * 1024)
4269 max_stripe_size = 1024 * 1024 * 1024;
4270 else
4271 max_stripe_size = 256 * 1024 * 1024;
4272 max_chunk_size = max_stripe_size;
4273 if (!devs_max)
4274 devs_max = BTRFS_MAX_DEVS(info->chunk_root);
4275 } else if (type & BTRFS_BLOCK_GROUP_SYSTEM) {
4276 max_stripe_size = 32 * 1024 * 1024;
4277 max_chunk_size = 2 * max_stripe_size;
4278 if (!devs_max)
4279 devs_max = BTRFS_MAX_DEVS_SYS_CHUNK;
4280 } else {
4281 btrfs_err(info, "invalid chunk type 0x%llx requested",
4282 type);
4283 BUG_ON(1);
4284 }
4285
4286 /* we don't want a chunk larger than 10% of writeable space */
4287 max_chunk_size = min(div_factor(fs_devices->total_rw_bytes, 1),
4288 max_chunk_size);
4289
4290 devices_info = kzalloc(sizeof(*devices_info) * fs_devices->rw_devices,
4291 GFP_NOFS);
4292 if (!devices_info)
4293 return -ENOMEM;
4294
4295 cur = fs_devices->alloc_list.next;
4296
4297 /*
4298 * in the first pass through the devices list, we gather information
4299 * about the available holes on each device.
4300 */
4301 ndevs = 0;
4302 while (cur != &fs_devices->alloc_list) {
4303 struct btrfs_device *device;
4304 u64 max_avail;
4305 u64 dev_offset;
4306
4307 device = list_entry(cur, struct btrfs_device, dev_alloc_list);
4308
4309 cur = cur->next;
4310
4311 if (!device->writeable) {
4312 WARN(1, KERN_ERR
4313 "BTRFS: read-only device in alloc_list\n");
4314 continue;
4315 }
4316
4317 if (!device->in_fs_metadata ||
4318 device->is_tgtdev_for_dev_replace)
4319 continue;
4320
4321 if (device->total_bytes > device->bytes_used)
4322 total_avail = device->total_bytes - device->bytes_used;
4323 else
4324 total_avail = 0;
4325
4326 /* If there is no space on this device, skip it. */
4327 if (total_avail == 0)
4328 continue;
4329
4330 ret = find_free_dev_extent(trans, device,
4331 max_stripe_size * dev_stripes,
4332 &dev_offset, &max_avail);
4333 if (ret && ret != -ENOSPC)
4334 goto error;
4335
4336 if (ret == 0)
4337 max_avail = max_stripe_size * dev_stripes;
4338
4339 if (max_avail < BTRFS_STRIPE_LEN * dev_stripes)
4340 continue;
4341
4342 if (ndevs == fs_devices->rw_devices) {
4343 WARN(1, "%s: found more than %llu devices\n",
4344 __func__, fs_devices->rw_devices);
4345 break;
4346 }
4347 devices_info[ndevs].dev_offset = dev_offset;
4348 devices_info[ndevs].max_avail = max_avail;
4349 devices_info[ndevs].total_avail = total_avail;
4350 devices_info[ndevs].dev = device;
4351 ++ndevs;
4352 }
4353
4354 /*
4355 * now sort the devices by hole size / available space
4356 */
4357 sort(devices_info, ndevs, sizeof(struct btrfs_device_info),
4358 btrfs_cmp_device_info, NULL);
4359
4360 /* round down to number of usable stripes */
4361 ndevs -= ndevs % devs_increment;
4362
4363 if (ndevs < devs_increment * sub_stripes || ndevs < devs_min) {
4364 ret = -ENOSPC;
4365 goto error;
4366 }
4367
4368 if (devs_max && ndevs > devs_max)
4369 ndevs = devs_max;
4370 /*
4371 * the primary goal is to maximize the number of stripes, so use as many
4372 * devices as possible, even if the stripes are not maximum sized.
4373 */
4374 stripe_size = devices_info[ndevs-1].max_avail;
4375 num_stripes = ndevs * dev_stripes;
4376
4377 /*
4378 * this will have to be fixed for RAID1 and RAID10 over
4379 * more drives
4380 */
4381 data_stripes = num_stripes / ncopies;
4382
4383 if (type & BTRFS_BLOCK_GROUP_RAID5) {
4384 raid_stripe_len = find_raid56_stripe_len(ndevs - 1,
4385 btrfs_super_stripesize(info->super_copy));
4386 data_stripes = num_stripes - 1;
4387 }
4388 if (type & BTRFS_BLOCK_GROUP_RAID6) {
4389 raid_stripe_len = find_raid56_stripe_len(ndevs - 2,
4390 btrfs_super_stripesize(info->super_copy));
4391 data_stripes = num_stripes - 2;
4392 }
4393
4394 /*
4395 * Use the number of data stripes to figure out how big this chunk
4396 * is really going to be in terms of logical address space,
4397 * and compare that answer with the max chunk size
4398 */
4399 if (stripe_size * data_stripes > max_chunk_size) {
4400 u64 mask = (1ULL << 24) - 1;
4401 stripe_size = max_chunk_size;
4402 do_div(stripe_size, data_stripes);
4403
4404 /* bump the answer up to a 16MB boundary */
4405 stripe_size = (stripe_size + mask) & ~mask;
4406
4407 /* but don't go higher than the limits we found
4408 * while searching for free extents
4409 */
4410 if (stripe_size > devices_info[ndevs-1].max_avail)
4411 stripe_size = devices_info[ndevs-1].max_avail;
4412 }
4413
4414 do_div(stripe_size, dev_stripes);
4415
4416 /* align to BTRFS_STRIPE_LEN */
4417 do_div(stripe_size, raid_stripe_len);
4418 stripe_size *= raid_stripe_len;
4419
4420 map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
4421 if (!map) {
4422 ret = -ENOMEM;
4423 goto error;
4424 }
4425 map->num_stripes = num_stripes;
4426
4427 for (i = 0; i < ndevs; ++i) {
4428 for (j = 0; j < dev_stripes; ++j) {
4429 int s = i * dev_stripes + j;
4430 map->stripes[s].dev = devices_info[i].dev;
4431 map->stripes[s].physical = devices_info[i].dev_offset +
4432 j * stripe_size;
4433 }
4434 }
4435 map->sector_size = extent_root->sectorsize;
4436 map->stripe_len = raid_stripe_len;
4437 map->io_align = raid_stripe_len;
4438 map->io_width = raid_stripe_len;
4439 map->type = type;
4440 map->sub_stripes = sub_stripes;
4441
4442 num_bytes = stripe_size * data_stripes;
4443
4444 trace_btrfs_chunk_alloc(info->chunk_root, map, start, num_bytes);
4445
4446 em = alloc_extent_map();
4447 if (!em) {
4448 kfree(map);
4449 ret = -ENOMEM;
4450 goto error;
4451 }
4452 set_bit(EXTENT_FLAG_FS_MAPPING, &em->flags);
4453 em->bdev = (struct block_device *)map;
4454 em->start = start;
4455 em->len = num_bytes;
4456 em->block_start = 0;
4457 em->block_len = em->len;
4458 em->orig_block_len = stripe_size;
4459
4460 em_tree = &extent_root->fs_info->mapping_tree.map_tree;
4461 write_lock(&em_tree->lock);
4462 ret = add_extent_mapping(em_tree, em, 0);
4463 if (!ret) {
4464 list_add_tail(&em->list, &trans->transaction->pending_chunks);
4465 atomic_inc(&em->refs);
4466 }
4467 write_unlock(&em_tree->lock);
4468 if (ret) {
4469 free_extent_map(em);
4470 goto error;
4471 }
4472
4473 ret = btrfs_make_block_group(trans, extent_root, 0, type,
4474 BTRFS_FIRST_CHUNK_TREE_OBJECTID,
4475 start, num_bytes);
4476 if (ret)
4477 goto error_del_extent;
4478
4479 for (i = 0; i < map->num_stripes; i++) {
4480 num_bytes = map->stripes[i].dev->bytes_used + stripe_size;
4481 btrfs_device_set_bytes_used(map->stripes[i].dev, num_bytes);
4482 }
4483
4484 spin_lock(&extent_root->fs_info->free_chunk_lock);
4485 extent_root->fs_info->free_chunk_space -= (stripe_size *
4486 map->num_stripes);
4487 spin_unlock(&extent_root->fs_info->free_chunk_lock);
4488
4489 free_extent_map(em);
4490 check_raid56_incompat_flag(extent_root->fs_info, type);
4491
4492 kfree(devices_info);
4493 return 0;
4494
4495 error_del_extent:
4496 write_lock(&em_tree->lock);
4497 remove_extent_mapping(em_tree, em);
4498 write_unlock(&em_tree->lock);
4499
4500 /* One for our allocation */
4501 free_extent_map(em);
4502 /* One for the tree reference */
4503 free_extent_map(em);
4504 /* One for the pending_chunks list reference */
4505 free_extent_map(em);
4506 error:
4507 kfree(devices_info);
4508 return ret;
4509 }
4510
4511 int btrfs_finish_chunk_alloc(struct btrfs_trans_handle *trans,
4512 struct btrfs_root *extent_root,
4513 u64 chunk_offset, u64 chunk_size)
4514 {
4515 struct btrfs_key key;
4516 struct btrfs_root *chunk_root = extent_root->fs_info->chunk_root;
4517 struct btrfs_device *device;
4518 struct btrfs_chunk *chunk;
4519 struct btrfs_stripe *stripe;
4520 struct extent_map_tree *em_tree;
4521 struct extent_map *em;
4522 struct map_lookup *map;
4523 size_t item_size;
4524 u64 dev_offset;
4525 u64 stripe_size;
4526 int i = 0;
4527 int ret;
4528
4529 em_tree = &extent_root->fs_info->mapping_tree.map_tree;
4530 read_lock(&em_tree->lock);
4531 em = lookup_extent_mapping(em_tree, chunk_offset, chunk_size);
4532 read_unlock(&em_tree->lock);
4533
4534 if (!em) {
4535 btrfs_crit(extent_root->fs_info, "unable to find logical "
4536 "%Lu len %Lu", chunk_offset, chunk_size);
4537 return -EINVAL;
4538 }
4539
4540 if (em->start != chunk_offset || em->len != chunk_size) {
4541 btrfs_crit(extent_root->fs_info, "found a bad mapping, wanted"
4542 " %Lu-%Lu, found %Lu-%Lu", chunk_offset,
4543 chunk_size, em->start, em->len);
4544 free_extent_map(em);
4545 return -EINVAL;
4546 }
4547
4548 map = (struct map_lookup *)em->bdev;
4549 item_size = btrfs_chunk_item_size(map->num_stripes);
4550 stripe_size = em->orig_block_len;
4551
4552 chunk = kzalloc(item_size, GFP_NOFS);
4553 if (!chunk) {
4554 ret = -ENOMEM;
4555 goto out;
4556 }
4557
4558 for (i = 0; i < map->num_stripes; i++) {
4559 device = map->stripes[i].dev;
4560 dev_offset = map->stripes[i].physical;
4561
4562 ret = btrfs_update_device(trans, device);
4563 if (ret)
4564 goto out;
4565 ret = btrfs_alloc_dev_extent(trans, device,
4566 chunk_root->root_key.objectid,
4567 BTRFS_FIRST_CHUNK_TREE_OBJECTID,
4568 chunk_offset, dev_offset,
4569 stripe_size);
4570 if (ret)
4571 goto out;
4572 }
4573
4574 stripe = &chunk->stripe;
4575 for (i = 0; i < map->num_stripes; i++) {
4576 device = map->stripes[i].dev;
4577 dev_offset = map->stripes[i].physical;
4578
4579 btrfs_set_stack_stripe_devid(stripe, device->devid);
4580 btrfs_set_stack_stripe_offset(stripe, dev_offset);
4581 memcpy(stripe->dev_uuid, device->uuid, BTRFS_UUID_SIZE);
4582 stripe++;
4583 }
4584
4585 btrfs_set_stack_chunk_length(chunk, chunk_size);
4586 btrfs_set_stack_chunk_owner(chunk, extent_root->root_key.objectid);
4587 btrfs_set_stack_chunk_stripe_len(chunk, map->stripe_len);
4588 btrfs_set_stack_chunk_type(chunk, map->type);
4589 btrfs_set_stack_chunk_num_stripes(chunk, map->num_stripes);
4590 btrfs_set_stack_chunk_io_align(chunk, map->stripe_len);
4591 btrfs_set_stack_chunk_io_width(chunk, map->stripe_len);
4592 btrfs_set_stack_chunk_sector_size(chunk, extent_root->sectorsize);
4593 btrfs_set_stack_chunk_sub_stripes(chunk, map->sub_stripes);
4594
4595 key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
4596 key.type = BTRFS_CHUNK_ITEM_KEY;
4597 key.offset = chunk_offset;
4598
4599 ret = btrfs_insert_item(trans, chunk_root, &key, chunk, item_size);
4600 if (ret == 0 && map->type & BTRFS_BLOCK_GROUP_SYSTEM) {
4601 /*
4602 * TODO: Cleanup of inserted chunk root in case of
4603 * failure.
4604 */
4605 ret = btrfs_add_system_chunk(chunk_root, &key, chunk,
4606 item_size);
4607 }
4608
4609 out:
4610 kfree(chunk);
4611 free_extent_map(em);
4612 return ret;
4613 }
4614
4615 /*
4616 * Chunk allocation falls into two parts. The first part does works
4617 * that make the new allocated chunk useable, but not do any operation
4618 * that modifies the chunk tree. The second part does the works that
4619 * require modifying the chunk tree. This division is important for the
4620 * bootstrap process of adding storage to a seed btrfs.
4621 */
4622 int btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
4623 struct btrfs_root *extent_root, u64 type)
4624 {
4625 u64 chunk_offset;
4626
4627 chunk_offset = find_next_chunk(extent_root->fs_info);
4628 return __btrfs_alloc_chunk(trans, extent_root, chunk_offset, type);
4629 }
4630
4631 static noinline int init_first_rw_device(struct btrfs_trans_handle *trans,
4632 struct btrfs_root *root,
4633 struct btrfs_device *device)
4634 {
4635 u64 chunk_offset;
4636 u64 sys_chunk_offset;
4637 u64 alloc_profile;
4638 struct btrfs_fs_info *fs_info = root->fs_info;
4639 struct btrfs_root *extent_root = fs_info->extent_root;
4640 int ret;
4641
4642 chunk_offset = find_next_chunk(fs_info);
4643 alloc_profile = btrfs_get_alloc_profile(extent_root, 0);
4644 ret = __btrfs_alloc_chunk(trans, extent_root, chunk_offset,
4645 alloc_profile);
4646 if (ret)
4647 return ret;
4648
4649 sys_chunk_offset = find_next_chunk(root->fs_info);
4650 alloc_profile = btrfs_get_alloc_profile(fs_info->chunk_root, 0);
4651 ret = __btrfs_alloc_chunk(trans, extent_root, sys_chunk_offset,
4652 alloc_profile);
4653 return ret;
4654 }
4655
4656 static inline int btrfs_chunk_max_errors(struct map_lookup *map)
4657 {
4658 int max_errors;
4659
4660 if (map->type & (BTRFS_BLOCK_GROUP_RAID1 |
4661 BTRFS_BLOCK_GROUP_RAID10 |
4662 BTRFS_BLOCK_GROUP_RAID5 |
4663 BTRFS_BLOCK_GROUP_DUP)) {
4664 max_errors = 1;
4665 } else if (map->type & BTRFS_BLOCK_GROUP_RAID6) {
4666 max_errors = 2;
4667 } else {
4668 max_errors = 0;
4669 }
4670
4671 return max_errors;
4672 }
4673
4674 int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset)
4675 {
4676 struct extent_map *em;
4677 struct map_lookup *map;
4678 struct btrfs_mapping_tree *map_tree = &root->fs_info->mapping_tree;
4679 int readonly = 0;
4680 int miss_ndevs = 0;
4681 int i;
4682
4683 read_lock(&map_tree->map_tree.lock);
4684 em = lookup_extent_mapping(&map_tree->map_tree, chunk_offset, 1);
4685 read_unlock(&map_tree->map_tree.lock);
4686 if (!em)
4687 return 1;
4688
4689 map = (struct map_lookup *)em->bdev;
4690 for (i = 0; i < map->num_stripes; i++) {
4691 if (map->stripes[i].dev->missing) {
4692 miss_ndevs++;
4693 continue;
4694 }
4695
4696 if (!map->stripes[i].dev->writeable) {
4697 readonly = 1;
4698 goto end;
4699 }
4700 }
4701
4702 /*
4703 * If the number of missing devices is larger than max errors,
4704 * we can not write the data into that chunk successfully, so
4705 * set it readonly.
4706 */
4707 if (miss_ndevs > btrfs_chunk_max_errors(map))
4708 readonly = 1;
4709 end:
4710 free_extent_map(em);
4711 return readonly;
4712 }
4713
4714 void btrfs_mapping_init(struct btrfs_mapping_tree *tree)
4715 {
4716 extent_map_tree_init(&tree->map_tree);
4717 }
4718
4719 void btrfs_mapping_tree_free(struct btrfs_mapping_tree *tree)
4720 {
4721 struct extent_map *em;
4722
4723 while (1) {
4724 write_lock(&tree->map_tree.lock);
4725 em = lookup_extent_mapping(&tree->map_tree, 0, (u64)-1);
4726 if (em)
4727 remove_extent_mapping(&tree->map_tree, em);
4728 write_unlock(&tree->map_tree.lock);
4729 if (!em)
4730 break;
4731 /* once for us */
4732 free_extent_map(em);
4733 /* once for the tree */
4734 free_extent_map(em);
4735 }
4736 }
4737
4738 int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len)
4739 {
4740 struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree;
4741 struct extent_map *em;
4742 struct map_lookup *map;
4743 struct extent_map_tree *em_tree = &map_tree->map_tree;
4744 int ret;
4745
4746 read_lock(&em_tree->lock);
4747 em = lookup_extent_mapping(em_tree, logical, len);
4748 read_unlock(&em_tree->lock);
4749
4750 /*
4751 * We could return errors for these cases, but that could get ugly and
4752 * we'd probably do the same thing which is just not do anything else
4753 * and exit, so return 1 so the callers don't try to use other copies.
4754 */
4755 if (!em) {
4756 btrfs_crit(fs_info, "No mapping for %Lu-%Lu", logical,
4757 logical+len);
4758 return 1;
4759 }
4760
4761 if (em->start > logical || em->start + em->len < logical) {
4762 btrfs_crit(fs_info, "Invalid mapping for %Lu-%Lu, got "
4763 "%Lu-%Lu", logical, logical+len, em->start,
4764 em->start + em->len);
4765 free_extent_map(em);
4766 return 1;
4767 }
4768
4769 map = (struct map_lookup *)em->bdev;
4770 if (map->type & (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1))
4771 ret = map->num_stripes;
4772 else if (map->type & BTRFS_BLOCK_GROUP_RAID10)
4773 ret = map->sub_stripes;
4774 else if (map->type & BTRFS_BLOCK_GROUP_RAID5)
4775 ret = 2;
4776 else if (map->type & BTRFS_BLOCK_GROUP_RAID6)
4777 ret = 3;
4778 else
4779 ret = 1;
4780 free_extent_map(em);
4781
4782 btrfs_dev_replace_lock(&fs_info->dev_replace);
4783 if (btrfs_dev_replace_is_ongoing(&fs_info->dev_replace))
4784 ret++;
4785 btrfs_dev_replace_unlock(&fs_info->dev_replace);
4786
4787 return ret;
4788 }
4789
4790 unsigned long btrfs_full_stripe_len(struct btrfs_root *root,
4791 struct btrfs_mapping_tree *map_tree,
4792 u64 logical)
4793 {
4794 struct extent_map *em;
4795 struct map_lookup *map;
4796 struct extent_map_tree *em_tree = &map_tree->map_tree;
4797 unsigned long len = root->sectorsize;
4798
4799 read_lock(&em_tree->lock);
4800 em = lookup_extent_mapping(em_tree, logical, len);
4801 read_unlock(&em_tree->lock);
4802 BUG_ON(!em);
4803
4804 BUG_ON(em->start > logical || em->start + em->len < logical);
4805 map = (struct map_lookup *)em->bdev;
4806 if (map->type & (BTRFS_BLOCK_GROUP_RAID5 |
4807 BTRFS_BLOCK_GROUP_RAID6)) {
4808 len = map->stripe_len * nr_data_stripes(map);
4809 }
4810 free_extent_map(em);
4811 return len;
4812 }
4813
4814 int btrfs_is_parity_mirror(struct btrfs_mapping_tree *map_tree,
4815 u64 logical, u64 len, int mirror_num)
4816 {
4817 struct extent_map *em;
4818 struct map_lookup *map;
4819 struct extent_map_tree *em_tree = &map_tree->map_tree;
4820 int ret = 0;
4821
4822 read_lock(&em_tree->lock);
4823 em = lookup_extent_mapping(em_tree, logical, len);
4824 read_unlock(&em_tree->lock);
4825 BUG_ON(!em);
4826
4827 BUG_ON(em->start > logical || em->start + em->len < logical);
4828 map = (struct map_lookup *)em->bdev;
4829 if (map->type & (BTRFS_BLOCK_GROUP_RAID5 |
4830 BTRFS_BLOCK_GROUP_RAID6))
4831 ret = 1;
4832 free_extent_map(em);
4833 return ret;
4834 }
4835
4836 static int find_live_mirror(struct btrfs_fs_info *fs_info,
4837 struct map_lookup *map, int first, int num,
4838 int optimal, int dev_replace_is_ongoing)
4839 {
4840 int i;
4841 int tolerance;
4842 struct btrfs_device *srcdev;
4843
4844 if (dev_replace_is_ongoing &&
4845 fs_info->dev_replace.cont_reading_from_srcdev_mode ==
4846 BTRFS_DEV_REPLACE_ITEM_CONT_READING_FROM_SRCDEV_MODE_AVOID)
4847 srcdev = fs_info->dev_replace.srcdev;
4848 else
4849 srcdev = NULL;
4850
4851 /*
4852 * try to avoid the drive that is the source drive for a
4853 * dev-replace procedure, only choose it if no other non-missing
4854 * mirror is available
4855 */
4856 for (tolerance = 0; tolerance < 2; tolerance++) {
4857 if (map->stripes[optimal].dev->bdev &&
4858 (tolerance || map->stripes[optimal].dev != srcdev))
4859 return optimal;
4860 for (i = first; i < first + num; i++) {
4861 if (map->stripes[i].dev->bdev &&
4862 (tolerance || map->stripes[i].dev != srcdev))
4863 return i;
4864 }
4865 }
4866
4867 /* we couldn't find one that doesn't fail. Just return something
4868 * and the io error handling code will clean up eventually
4869 */
4870 return optimal;
4871 }
4872
4873 static inline int parity_smaller(u64 a, u64 b)
4874 {
4875 return a > b;
4876 }
4877
4878 /* Bubble-sort the stripe set to put the parity/syndrome stripes last */
4879 static void sort_parity_stripes(struct btrfs_bio *bbio, u64 *raid_map)
4880 {
4881 struct btrfs_bio_stripe s;
4882 int real_stripes = bbio->num_stripes - bbio->num_tgtdevs;
4883 int i;
4884 u64 l;
4885 int again = 1;
4886 int m;
4887
4888 while (again) {
4889 again = 0;
4890 for (i = 0; i < real_stripes - 1; i++) {
4891 if (parity_smaller(raid_map[i], raid_map[i+1])) {
4892 s = bbio->stripes[i];
4893 l = raid_map[i];
4894 bbio->stripes[i] = bbio->stripes[i+1];
4895 raid_map[i] = raid_map[i+1];
4896 bbio->stripes[i+1] = s;
4897 raid_map[i+1] = l;
4898
4899 if (bbio->tgtdev_map) {
4900 m = bbio->tgtdev_map[i];
4901 bbio->tgtdev_map[i] =
4902 bbio->tgtdev_map[i + 1];
4903 bbio->tgtdev_map[i + 1] = m;
4904 }
4905
4906 again = 1;
4907 }
4908 }
4909 }
4910 }
4911
4912 static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
4913 u64 logical, u64 *length,
4914 struct btrfs_bio **bbio_ret,
4915 int mirror_num, u64 **raid_map_ret)
4916 {
4917 struct extent_map *em;
4918 struct map_lookup *map;
4919 struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree;
4920 struct extent_map_tree *em_tree = &map_tree->map_tree;
4921 u64 offset;
4922 u64 stripe_offset;
4923 u64 stripe_end_offset;
4924 u64 stripe_nr;
4925 u64 stripe_nr_orig;
4926 u64 stripe_nr_end;
4927 u64 stripe_len;
4928 u64 *raid_map = NULL;
4929 int stripe_index;
4930 int i;
4931 int ret = 0;
4932 int num_stripes;
4933 int max_errors = 0;
4934 int tgtdev_indexes = 0;
4935 struct btrfs_bio *bbio = NULL;
4936 struct btrfs_dev_replace *dev_replace = &fs_info->dev_replace;
4937 int dev_replace_is_ongoing = 0;
4938 int num_alloc_stripes;
4939 int patch_the_first_stripe_for_dev_replace = 0;
4940 u64 physical_to_patch_in_first_stripe = 0;
4941 u64 raid56_full_stripe_start = (u64)-1;
4942
4943 read_lock(&em_tree->lock);
4944 em = lookup_extent_mapping(em_tree, logical, *length);
4945 read_unlock(&em_tree->lock);
4946
4947 if (!em) {
4948 btrfs_crit(fs_info, "unable to find logical %llu len %llu",
4949 logical, *length);
4950 return -EINVAL;
4951 }
4952
4953 if (em->start > logical || em->start + em->len < logical) {
4954 btrfs_crit(fs_info, "found a bad mapping, wanted %Lu, "
4955 "found %Lu-%Lu", logical, em->start,
4956 em->start + em->len);
4957 free_extent_map(em);
4958 return -EINVAL;
4959 }
4960
4961 map = (struct map_lookup *)em->bdev;
4962 offset = logical - em->start;
4963
4964 stripe_len = map->stripe_len;
4965 stripe_nr = offset;
4966 /*
4967 * stripe_nr counts the total number of stripes we have to stride
4968 * to get to this block
4969 */
4970 do_div(stripe_nr, stripe_len);
4971
4972 stripe_offset = stripe_nr * stripe_len;
4973 BUG_ON(offset < stripe_offset);
4974
4975 /* stripe_offset is the offset of this block in its stripe*/
4976 stripe_offset = offset - stripe_offset;
4977
4978 /* if we're here for raid56, we need to know the stripe aligned start */
4979 if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) {
4980 unsigned long full_stripe_len = stripe_len * nr_data_stripes(map);
4981 raid56_full_stripe_start = offset;
4982
4983 /* allow a write of a full stripe, but make sure we don't
4984 * allow straddling of stripes
4985 */
4986 do_div(raid56_full_stripe_start, full_stripe_len);
4987 raid56_full_stripe_start *= full_stripe_len;
4988 }
4989
4990 if (rw & REQ_DISCARD) {
4991 /* we don't discard raid56 yet */
4992 if (map->type &
4993 (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) {
4994 ret = -EOPNOTSUPP;
4995 goto out;
4996 }
4997 *length = min_t(u64, em->len - offset, *length);
4998 } else if (map->type & BTRFS_BLOCK_GROUP_PROFILE_MASK) {
4999 u64 max_len;
5000 /* For writes to RAID[56], allow a full stripeset across all disks.
5001 For other RAID types and for RAID[56] reads, just allow a single
5002 stripe (on a single disk). */
5003 if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6) &&
5004 (rw & REQ_WRITE)) {
5005 max_len = stripe_len * nr_data_stripes(map) -
5006 (offset - raid56_full_stripe_start);
5007 } else {
5008 /* we limit the length of each bio to what fits in a stripe */
5009 max_len = stripe_len - stripe_offset;
5010 }
5011 *length = min_t(u64, em->len - offset, max_len);
5012 } else {
5013 *length = em->len - offset;
5014 }
5015
5016 /* This is for when we're called from btrfs_merge_bio_hook() and all
5017 it cares about is the length */
5018 if (!bbio_ret)
5019 goto out;
5020
5021 btrfs_dev_replace_lock(dev_replace);
5022 dev_replace_is_ongoing = btrfs_dev_replace_is_ongoing(dev_replace);
5023 if (!dev_replace_is_ongoing)
5024 btrfs_dev_replace_unlock(dev_replace);
5025
5026 if (dev_replace_is_ongoing && mirror_num == map->num_stripes + 1 &&
5027 !(rw & (REQ_WRITE | REQ_DISCARD | REQ_GET_READ_MIRRORS)) &&
5028 dev_replace->tgtdev != NULL) {
5029 /*
5030 * in dev-replace case, for repair case (that's the only
5031 * case where the mirror is selected explicitly when
5032 * calling btrfs_map_block), blocks left of the left cursor
5033 * can also be read from the target drive.
5034 * For REQ_GET_READ_MIRRORS, the target drive is added as
5035 * the last one to the array of stripes. For READ, it also
5036 * needs to be supported using the same mirror number.
5037 * If the requested block is not left of the left cursor,
5038 * EIO is returned. This can happen because btrfs_num_copies()
5039 * returns one more in the dev-replace case.
5040 */
5041 u64 tmp_length = *length;
5042 struct btrfs_bio *tmp_bbio = NULL;
5043 int tmp_num_stripes;
5044 u64 srcdev_devid = dev_replace->srcdev->devid;
5045 int index_srcdev = 0;
5046 int found = 0;
5047 u64 physical_of_found = 0;
5048
5049 ret = __btrfs_map_block(fs_info, REQ_GET_READ_MIRRORS,
5050 logical, &tmp_length, &tmp_bbio, 0, NULL);
5051 if (ret) {
5052 WARN_ON(tmp_bbio != NULL);
5053 goto out;
5054 }
5055
5056 tmp_num_stripes = tmp_bbio->num_stripes;
5057 if (mirror_num > tmp_num_stripes) {
5058 /*
5059 * REQ_GET_READ_MIRRORS does not contain this
5060 * mirror, that means that the requested area
5061 * is not left of the left cursor
5062 */
5063 ret = -EIO;
5064 kfree(tmp_bbio);
5065 goto out;
5066 }
5067
5068 /*
5069 * process the rest of the function using the mirror_num
5070 * of the source drive. Therefore look it up first.
5071 * At the end, patch the device pointer to the one of the
5072 * target drive.
5073 */
5074 for (i = 0; i < tmp_num_stripes; i++) {
5075 if (tmp_bbio->stripes[i].dev->devid == srcdev_devid) {
5076 /*
5077 * In case of DUP, in order to keep it
5078 * simple, only add the mirror with the
5079 * lowest physical address
5080 */
5081 if (found &&
5082 physical_of_found <=
5083 tmp_bbio->stripes[i].physical)
5084 continue;
5085 index_srcdev = i;
5086 found = 1;
5087 physical_of_found =
5088 tmp_bbio->stripes[i].physical;
5089 }
5090 }
5091
5092 if (found) {
5093 mirror_num = index_srcdev + 1;
5094 patch_the_first_stripe_for_dev_replace = 1;
5095 physical_to_patch_in_first_stripe = physical_of_found;
5096 } else {
5097 WARN_ON(1);
5098 ret = -EIO;
5099 kfree(tmp_bbio);
5100 goto out;
5101 }
5102
5103 kfree(tmp_bbio);
5104 } else if (mirror_num > map->num_stripes) {
5105 mirror_num = 0;
5106 }
5107
5108 num_stripes = 1;
5109 stripe_index = 0;
5110 stripe_nr_orig = stripe_nr;
5111 stripe_nr_end = ALIGN(offset + *length, map->stripe_len);
5112 do_div(stripe_nr_end, map->stripe_len);
5113 stripe_end_offset = stripe_nr_end * map->stripe_len -
5114 (offset + *length);
5115
5116 if (map->type & BTRFS_BLOCK_GROUP_RAID0) {
5117 if (rw & REQ_DISCARD)
5118 num_stripes = min_t(u64, map->num_stripes,
5119 stripe_nr_end - stripe_nr_orig);
5120 stripe_index = do_div(stripe_nr, map->num_stripes);
5121 if (!(rw & (REQ_WRITE | REQ_DISCARD | REQ_GET_READ_MIRRORS)))
5122 mirror_num = 1;
5123 } else if (map->type & BTRFS_BLOCK_GROUP_RAID1) {
5124 if (rw & (REQ_WRITE | REQ_DISCARD | REQ_GET_READ_MIRRORS))
5125 num_stripes = map->num_stripes;
5126 else if (mirror_num)
5127 stripe_index = mirror_num - 1;
5128 else {
5129 stripe_index = find_live_mirror(fs_info, map, 0,
5130 map->num_stripes,
5131 current->pid % map->num_stripes,
5132 dev_replace_is_ongoing);
5133 mirror_num = stripe_index + 1;
5134 }
5135
5136 } else if (map->type & BTRFS_BLOCK_GROUP_DUP) {
5137 if (rw & (REQ_WRITE | REQ_DISCARD | REQ_GET_READ_MIRRORS)) {
5138 num_stripes = map->num_stripes;
5139 } else if (mirror_num) {
5140 stripe_index = mirror_num - 1;
5141 } else {
5142 mirror_num = 1;
5143 }
5144
5145 } else if (map->type & BTRFS_BLOCK_GROUP_RAID10) {
5146 int factor = map->num_stripes / map->sub_stripes;
5147
5148 stripe_index = do_div(stripe_nr, factor);
5149 stripe_index *= map->sub_stripes;
5150
5151 if (rw & (REQ_WRITE | REQ_GET_READ_MIRRORS))
5152 num_stripes = map->sub_stripes;
5153 else if (rw & REQ_DISCARD)
5154 num_stripes = min_t(u64, map->sub_stripes *
5155 (stripe_nr_end - stripe_nr_orig),
5156 map->num_stripes);
5157 else if (mirror_num)
5158 stripe_index += mirror_num - 1;
5159 else {
5160 int old_stripe_index = stripe_index;
5161 stripe_index = find_live_mirror(fs_info, map,
5162 stripe_index,
5163 map->sub_stripes, stripe_index +
5164 current->pid % map->sub_stripes,
5165 dev_replace_is_ongoing);
5166 mirror_num = stripe_index - old_stripe_index + 1;
5167 }
5168
5169 } else if (map->type & (BTRFS_BLOCK_GROUP_RAID5 |
5170 BTRFS_BLOCK_GROUP_RAID6)) {
5171 u64 tmp;
5172
5173 if (raid_map_ret &&
5174 ((rw & (REQ_WRITE | REQ_GET_READ_MIRRORS)) ||
5175 mirror_num > 1)) {
5176 int i, rot;
5177
5178 /* push stripe_nr back to the start of the full stripe */
5179 stripe_nr = raid56_full_stripe_start;
5180 do_div(stripe_nr, stripe_len * nr_data_stripes(map));
5181
5182 /* RAID[56] write or recovery. Return all stripes */
5183 num_stripes = map->num_stripes;
5184 max_errors = nr_parity_stripes(map);
5185
5186 raid_map = kmalloc_array(num_stripes, sizeof(u64),
5187 GFP_NOFS);
5188 if (!raid_map) {
5189 ret = -ENOMEM;
5190 goto out;
5191 }
5192
5193 /* Work out the disk rotation on this stripe-set */
5194 tmp = stripe_nr;
5195 rot = do_div(tmp, num_stripes);
5196
5197 /* Fill in the logical address of each stripe */
5198 tmp = stripe_nr * nr_data_stripes(map);
5199 for (i = 0; i < nr_data_stripes(map); i++)
5200 raid_map[(i+rot) % num_stripes] =
5201 em->start + (tmp + i) * map->stripe_len;
5202
5203 raid_map[(i+rot) % map->num_stripes] = RAID5_P_STRIPE;
5204 if (map->type & BTRFS_BLOCK_GROUP_RAID6)
5205 raid_map[(i+rot+1) % num_stripes] =
5206 RAID6_Q_STRIPE;
5207
5208 *length = map->stripe_len;
5209 stripe_index = 0;
5210 stripe_offset = 0;
5211 } else {
5212 /*
5213 * Mirror #0 or #1 means the original data block.
5214 * Mirror #2 is RAID5 parity block.
5215 * Mirror #3 is RAID6 Q block.
5216 */
5217 stripe_index = do_div(stripe_nr, nr_data_stripes(map));
5218 if (mirror_num > 1)
5219 stripe_index = nr_data_stripes(map) +
5220 mirror_num - 2;
5221
5222 /* We distribute the parity blocks across stripes */
5223 tmp = stripe_nr + stripe_index;
5224 stripe_index = do_div(tmp, map->num_stripes);
5225 if (!(rw & (REQ_WRITE | REQ_DISCARD |
5226 REQ_GET_READ_MIRRORS)) && mirror_num <= 1)
5227 mirror_num = 1;
5228 }
5229 } else {
5230 /*
5231 * after this do_div call, stripe_nr is the number of stripes
5232 * on this device we have to walk to find the data, and
5233 * stripe_index is the number of our device in the stripe array
5234 */
5235 stripe_index = do_div(stripe_nr, map->num_stripes);
5236 mirror_num = stripe_index + 1;
5237 }
5238 BUG_ON(stripe_index >= map->num_stripes);
5239
5240 num_alloc_stripes = num_stripes;
5241 if (dev_replace_is_ongoing) {
5242 if (rw & (REQ_WRITE | REQ_DISCARD))
5243 num_alloc_stripes <<= 1;
5244 if (rw & REQ_GET_READ_MIRRORS)
5245 num_alloc_stripes++;
5246 tgtdev_indexes = num_stripes;
5247 }
5248
5249 bbio = kzalloc(btrfs_bio_size(num_alloc_stripes, tgtdev_indexes),
5250 GFP_NOFS);
5251 if (!bbio) {
5252 kfree(raid_map);
5253 ret = -ENOMEM;
5254 goto out;
5255 }
5256 atomic_set(&bbio->error, 0);
5257 if (dev_replace_is_ongoing)
5258 bbio->tgtdev_map = (int *)(bbio->stripes + num_alloc_stripes);
5259
5260 if (rw & REQ_DISCARD) {
5261 int factor = 0;
5262 int sub_stripes = 0;
5263 u64 stripes_per_dev = 0;
5264 u32 remaining_stripes = 0;
5265 u32 last_stripe = 0;
5266
5267 if (map->type &
5268 (BTRFS_BLOCK_GROUP_RAID0 | BTRFS_BLOCK_GROUP_RAID10)) {
5269 if (map->type & BTRFS_BLOCK_GROUP_RAID0)
5270 sub_stripes = 1;
5271 else
5272 sub_stripes = map->sub_stripes;
5273
5274 factor = map->num_stripes / sub_stripes;
5275 stripes_per_dev = div_u64_rem(stripe_nr_end -
5276 stripe_nr_orig,
5277 factor,
5278 &remaining_stripes);
5279 div_u64_rem(stripe_nr_end - 1, factor, &last_stripe);
5280 last_stripe *= sub_stripes;
5281 }
5282
5283 for (i = 0; i < num_stripes; i++) {
5284 bbio->stripes[i].physical =
5285 map->stripes[stripe_index].physical +
5286 stripe_offset + stripe_nr * map->stripe_len;
5287 bbio->stripes[i].dev = map->stripes[stripe_index].dev;
5288
5289 if (map->type & (BTRFS_BLOCK_GROUP_RAID0 |
5290 BTRFS_BLOCK_GROUP_RAID10)) {
5291 bbio->stripes[i].length = stripes_per_dev *
5292 map->stripe_len;
5293
5294 if (i / sub_stripes < remaining_stripes)
5295 bbio->stripes[i].length +=
5296 map->stripe_len;
5297
5298 /*
5299 * Special for the first stripe and
5300 * the last stripe:
5301 *
5302 * |-------|...|-------|
5303 * |----------|
5304 * off end_off
5305 */
5306 if (i < sub_stripes)
5307 bbio->stripes[i].length -=
5308 stripe_offset;
5309
5310 if (stripe_index >= last_stripe &&
5311 stripe_index <= (last_stripe +
5312 sub_stripes - 1))
5313 bbio->stripes[i].length -=
5314 stripe_end_offset;
5315
5316 if (i == sub_stripes - 1)
5317 stripe_offset = 0;
5318 } else
5319 bbio->stripes[i].length = *length;
5320
5321 stripe_index++;
5322 if (stripe_index == map->num_stripes) {
5323 /* This could only happen for RAID0/10 */
5324 stripe_index = 0;
5325 stripe_nr++;
5326 }
5327 }
5328 } else {
5329 for (i = 0; i < num_stripes; i++) {
5330 bbio->stripes[i].physical =
5331 map->stripes[stripe_index].physical +
5332 stripe_offset +
5333 stripe_nr * map->stripe_len;
5334 bbio->stripes[i].dev =
5335 map->stripes[stripe_index].dev;
5336 stripe_index++;
5337 }
5338 }
5339
5340 if (rw & (REQ_WRITE | REQ_GET_READ_MIRRORS))
5341 max_errors = btrfs_chunk_max_errors(map);
5342
5343 tgtdev_indexes = 0;
5344 if (dev_replace_is_ongoing && (rw & (REQ_WRITE | REQ_DISCARD)) &&
5345 dev_replace->tgtdev != NULL) {
5346 int index_where_to_add;
5347 u64 srcdev_devid = dev_replace->srcdev->devid;
5348
5349 /*
5350 * duplicate the write operations while the dev replace
5351 * procedure is running. Since the copying of the old disk
5352 * to the new disk takes place at run time while the
5353 * filesystem is mounted writable, the regular write
5354 * operations to the old disk have to be duplicated to go
5355 * to the new disk as well.
5356 * Note that device->missing is handled by the caller, and
5357 * that the write to the old disk is already set up in the
5358 * stripes array.
5359 */
5360 index_where_to_add = num_stripes;
5361 for (i = 0; i < num_stripes; i++) {
5362 if (bbio->stripes[i].dev->devid == srcdev_devid) {
5363 /* write to new disk, too */
5364 struct btrfs_bio_stripe *new =
5365 bbio->stripes + index_where_to_add;
5366 struct btrfs_bio_stripe *old =
5367 bbio->stripes + i;
5368
5369 new->physical = old->physical;
5370 new->length = old->length;
5371 new->dev = dev_replace->tgtdev;
5372 bbio->tgtdev_map[i] = index_where_to_add;
5373 index_where_to_add++;
5374 max_errors++;
5375 tgtdev_indexes++;
5376 }
5377 }
5378 num_stripes = index_where_to_add;
5379 } else if (dev_replace_is_ongoing && (rw & REQ_GET_READ_MIRRORS) &&
5380 dev_replace->tgtdev != NULL) {
5381 u64 srcdev_devid = dev_replace->srcdev->devid;
5382 int index_srcdev = 0;
5383 int found = 0;
5384 u64 physical_of_found = 0;
5385
5386 /*
5387 * During the dev-replace procedure, the target drive can
5388 * also be used to read data in case it is needed to repair
5389 * a corrupt block elsewhere. This is possible if the
5390 * requested area is left of the left cursor. In this area,
5391 * the target drive is a full copy of the source drive.
5392 */
5393 for (i = 0; i < num_stripes; i++) {
5394 if (bbio->stripes[i].dev->devid == srcdev_devid) {
5395 /*
5396 * In case of DUP, in order to keep it
5397 * simple, only add the mirror with the
5398 * lowest physical address
5399 */
5400 if (found &&
5401 physical_of_found <=
5402 bbio->stripes[i].physical)
5403 continue;
5404 index_srcdev = i;
5405 found = 1;
5406 physical_of_found = bbio->stripes[i].physical;
5407 }
5408 }
5409 if (found) {
5410 u64 length = map->stripe_len;
5411
5412 if (physical_of_found + length <=
5413 dev_replace->cursor_left) {
5414 struct btrfs_bio_stripe *tgtdev_stripe =
5415 bbio->stripes + num_stripes;
5416
5417 tgtdev_stripe->physical = physical_of_found;
5418 tgtdev_stripe->length =
5419 bbio->stripes[index_srcdev].length;
5420 tgtdev_stripe->dev = dev_replace->tgtdev;
5421 bbio->tgtdev_map[index_srcdev] = num_stripes;
5422
5423 tgtdev_indexes++;
5424 num_stripes++;
5425 }
5426 }
5427 }
5428
5429 *bbio_ret = bbio;
5430 bbio->num_stripes = num_stripes;
5431 bbio->max_errors = max_errors;
5432 bbio->mirror_num = mirror_num;
5433 bbio->num_tgtdevs = tgtdev_indexes;
5434
5435 /*
5436 * this is the case that REQ_READ && dev_replace_is_ongoing &&
5437 * mirror_num == num_stripes + 1 && dev_replace target drive is
5438 * available as a mirror
5439 */
5440 if (patch_the_first_stripe_for_dev_replace && num_stripes > 0) {
5441 WARN_ON(num_stripes > 1);
5442 bbio->stripes[0].dev = dev_replace->tgtdev;
5443 bbio->stripes[0].physical = physical_to_patch_in_first_stripe;
5444 bbio->mirror_num = map->num_stripes + 1;
5445 }
5446 if (raid_map) {
5447 sort_parity_stripes(bbio, raid_map);
5448 *raid_map_ret = raid_map;
5449 }
5450 out:
5451 if (dev_replace_is_ongoing)
5452 btrfs_dev_replace_unlock(dev_replace);
5453 free_extent_map(em);
5454 return ret;
5455 }
5456
5457 int btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
5458 u64 logical, u64 *length,
5459 struct btrfs_bio **bbio_ret, int mirror_num)
5460 {
5461 return __btrfs_map_block(fs_info, rw, logical, length, bbio_ret,
5462 mirror_num, NULL);
5463 }
5464
5465 /* For Scrub/replace */
5466 int btrfs_map_sblock(struct btrfs_fs_info *fs_info, int rw,
5467 u64 logical, u64 *length,
5468 struct btrfs_bio **bbio_ret, int mirror_num,
5469 u64 **raid_map_ret)
5470 {
5471 return __btrfs_map_block(fs_info, rw, logical, length, bbio_ret,
5472 mirror_num, raid_map_ret);
5473 }
5474
5475 int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree,
5476 u64 chunk_start, u64 physical, u64 devid,
5477 u64 **logical, int *naddrs, int *stripe_len)
5478 {
5479 struct extent_map_tree *em_tree = &map_tree->map_tree;
5480 struct extent_map *em;
5481 struct map_lookup *map;
5482 u64 *buf;
5483 u64 bytenr;
5484 u64 length;
5485 u64 stripe_nr;
5486 u64 rmap_len;
5487 int i, j, nr = 0;
5488
5489 read_lock(&em_tree->lock);
5490 em = lookup_extent_mapping(em_tree, chunk_start, 1);
5491 read_unlock(&em_tree->lock);
5492
5493 if (!em) {
5494 printk(KERN_ERR "BTRFS: couldn't find em for chunk %Lu\n",
5495 chunk_start);
5496 return -EIO;
5497 }
5498
5499 if (em->start != chunk_start) {
5500 printk(KERN_ERR "BTRFS: bad chunk start, em=%Lu, wanted=%Lu\n",
5501 em->start, chunk_start);
5502 free_extent_map(em);
5503 return -EIO;
5504 }
5505 map = (struct map_lookup *)em->bdev;
5506
5507 length = em->len;
5508 rmap_len = map->stripe_len;
5509
5510 if (map->type & BTRFS_BLOCK_GROUP_RAID10)
5511 do_div(length, map->num_stripes / map->sub_stripes);
5512 else if (map->type & BTRFS_BLOCK_GROUP_RAID0)
5513 do_div(length, map->num_stripes);
5514 else if (map->type & (BTRFS_BLOCK_GROUP_RAID5 |
5515 BTRFS_BLOCK_GROUP_RAID6)) {
5516 do_div(length, nr_data_stripes(map));
5517 rmap_len = map->stripe_len * nr_data_stripes(map);
5518 }
5519
5520 buf = kzalloc(sizeof(u64) * map->num_stripes, GFP_NOFS);
5521 BUG_ON(!buf); /* -ENOMEM */
5522
5523 for (i = 0; i < map->num_stripes; i++) {
5524 if (devid && map->stripes[i].dev->devid != devid)
5525 continue;
5526 if (map->stripes[i].physical > physical ||
5527 map->stripes[i].physical + length <= physical)
5528 continue;
5529
5530 stripe_nr = physical - map->stripes[i].physical;
5531 do_div(stripe_nr, map->stripe_len);
5532
5533 if (map->type & BTRFS_BLOCK_GROUP_RAID10) {
5534 stripe_nr = stripe_nr * map->num_stripes + i;
5535 do_div(stripe_nr, map->sub_stripes);
5536 } else if (map->type & BTRFS_BLOCK_GROUP_RAID0) {
5537 stripe_nr = stripe_nr * map->num_stripes + i;
5538 } /* else if RAID[56], multiply by nr_data_stripes().
5539 * Alternatively, just use rmap_len below instead of
5540 * map->stripe_len */
5541
5542 bytenr = chunk_start + stripe_nr * rmap_len;
5543 WARN_ON(nr >= map->num_stripes);
5544 for (j = 0; j < nr; j++) {
5545 if (buf[j] == bytenr)
5546 break;
5547 }
5548 if (j == nr) {
5549 WARN_ON(nr >= map->num_stripes);
5550 buf[nr++] = bytenr;
5551 }
5552 }
5553
5554 *logical = buf;
5555 *naddrs = nr;
5556 *stripe_len = rmap_len;
5557
5558 free_extent_map(em);
5559 return 0;
5560 }
5561
5562 static inline void btrfs_end_bbio(struct btrfs_bio *bbio, struct bio *bio, int err)
5563 {
5564 if (likely(bbio->flags & BTRFS_BIO_ORIG_BIO_SUBMITTED))
5565 bio_endio_nodec(bio, err);
5566 else
5567 bio_endio(bio, err);
5568 kfree(bbio);
5569 }
5570
5571 static void btrfs_end_bio(struct bio *bio, int err)
5572 {
5573 struct btrfs_bio *bbio = bio->bi_private;
5574 struct btrfs_device *dev = bbio->stripes[0].dev;
5575 int is_orig_bio = 0;
5576
5577 if (err) {
5578 atomic_inc(&bbio->error);
5579 if (err == -EIO || err == -EREMOTEIO) {
5580 unsigned int stripe_index =
5581 btrfs_io_bio(bio)->stripe_index;
5582
5583 BUG_ON(stripe_index >= bbio->num_stripes);
5584 dev = bbio->stripes[stripe_index].dev;
5585 if (dev->bdev) {
5586 if (bio->bi_rw & WRITE)
5587 btrfs_dev_stat_inc(dev,
5588 BTRFS_DEV_STAT_WRITE_ERRS);
5589 else
5590 btrfs_dev_stat_inc(dev,
5591 BTRFS_DEV_STAT_READ_ERRS);
5592 if ((bio->bi_rw & WRITE_FLUSH) == WRITE_FLUSH)
5593 btrfs_dev_stat_inc(dev,
5594 BTRFS_DEV_STAT_FLUSH_ERRS);
5595 btrfs_dev_stat_print_on_error(dev);
5596 }
5597 }
5598 }
5599
5600 if (bio == bbio->orig_bio)
5601 is_orig_bio = 1;
5602
5603 btrfs_bio_counter_dec(bbio->fs_info);
5604
5605 if (atomic_dec_and_test(&bbio->stripes_pending)) {
5606 if (!is_orig_bio) {
5607 bio_put(bio);
5608 bio = bbio->orig_bio;
5609 }
5610
5611 bio->bi_private = bbio->private;
5612 bio->bi_end_io = bbio->end_io;
5613 btrfs_io_bio(bio)->mirror_num = bbio->mirror_num;
5614 /* only send an error to the higher layers if it is
5615 * beyond the tolerance of the btrfs bio
5616 */
5617 if (atomic_read(&bbio->error) > bbio->max_errors) {
5618 err = -EIO;
5619 } else {
5620 /*
5621 * this bio is actually up to date, we didn't
5622 * go over the max number of errors
5623 */
5624 set_bit(BIO_UPTODATE, &bio->bi_flags);
5625 err = 0;
5626 }
5627
5628 btrfs_end_bbio(bbio, bio, err);
5629 } else if (!is_orig_bio) {
5630 bio_put(bio);
5631 }
5632 }
5633
5634 /*
5635 * see run_scheduled_bios for a description of why bios are collected for
5636 * async submit.
5637 *
5638 * This will add one bio to the pending list for a device and make sure
5639 * the work struct is scheduled.
5640 */
5641 static noinline void btrfs_schedule_bio(struct btrfs_root *root,
5642 struct btrfs_device *device,
5643 int rw, struct bio *bio)
5644 {
5645 int should_queue = 1;
5646 struct btrfs_pending_bios *pending_bios;
5647
5648 if (device->missing || !device->bdev) {
5649 bio_endio(bio, -EIO);
5650 return;
5651 }
5652
5653 /* don't bother with additional async steps for reads, right now */
5654 if (!(rw & REQ_WRITE)) {
5655 bio_get(bio);
5656 btrfsic_submit_bio(rw, bio);
5657 bio_put(bio);
5658 return;
5659 }
5660
5661 /*
5662 * nr_async_bios allows us to reliably return congestion to the
5663 * higher layers. Otherwise, the async bio makes it appear we have
5664 * made progress against dirty pages when we've really just put it
5665 * on a queue for later
5666 */
5667 atomic_inc(&root->fs_info->nr_async_bios);
5668 WARN_ON(bio->bi_next);
5669 bio->bi_next = NULL;
5670 bio->bi_rw |= rw;
5671
5672 spin_lock(&device->io_lock);
5673 if (bio->bi_rw & REQ_SYNC)
5674 pending_bios = &device->pending_sync_bios;
5675 else
5676 pending_bios = &device->pending_bios;
5677
5678 if (pending_bios->tail)
5679 pending_bios->tail->bi_next = bio;
5680
5681 pending_bios->tail = bio;
5682 if (!pending_bios->head)
5683 pending_bios->head = bio;
5684 if (device->running_pending)
5685 should_queue = 0;
5686
5687 spin_unlock(&device->io_lock);
5688
5689 if (should_queue)
5690 btrfs_queue_work(root->fs_info->submit_workers,
5691 &device->work);
5692 }
5693
5694 static int bio_size_ok(struct block_device *bdev, struct bio *bio,
5695 sector_t sector)
5696 {
5697 struct bio_vec *prev;
5698 struct request_queue *q = bdev_get_queue(bdev);
5699 unsigned int max_sectors = queue_max_sectors(q);
5700 struct bvec_merge_data bvm = {
5701 .bi_bdev = bdev,
5702 .bi_sector = sector,
5703 .bi_rw = bio->bi_rw,
5704 };
5705
5706 if (WARN_ON(bio->bi_vcnt == 0))
5707 return 1;
5708
5709 prev = &bio->bi_io_vec[bio->bi_vcnt - 1];
5710 if (bio_sectors(bio) > max_sectors)
5711 return 0;
5712
5713 if (!q->merge_bvec_fn)
5714 return 1;
5715
5716 bvm.bi_size = bio->bi_iter.bi_size - prev->bv_len;
5717 if (q->merge_bvec_fn(q, &bvm, prev) < prev->bv_len)
5718 return 0;
5719 return 1;
5720 }
5721
5722 static void submit_stripe_bio(struct btrfs_root *root, struct btrfs_bio *bbio,
5723 struct bio *bio, u64 physical, int dev_nr,
5724 int rw, int async)
5725 {
5726 struct btrfs_device *dev = bbio->stripes[dev_nr].dev;
5727
5728 bio->bi_private = bbio;
5729 btrfs_io_bio(bio)->stripe_index = dev_nr;
5730 bio->bi_end_io = btrfs_end_bio;
5731 bio->bi_iter.bi_sector = physical >> 9;
5732 #ifdef DEBUG
5733 {
5734 struct rcu_string *name;
5735
5736 rcu_read_lock();
5737 name = rcu_dereference(dev->name);
5738 pr_debug("btrfs_map_bio: rw %d, sector=%llu, dev=%lu "
5739 "(%s id %llu), size=%u\n", rw,
5740 (u64)bio->bi_iter.bi_sector, (u_long)dev->bdev->bd_dev,
5741 name->str, dev->devid, bio->bi_iter.bi_size);
5742 rcu_read_unlock();
5743 }
5744 #endif
5745 bio->bi_bdev = dev->bdev;
5746
5747 btrfs_bio_counter_inc_noblocked(root->fs_info);
5748
5749 if (async)
5750 btrfs_schedule_bio(root, dev, rw, bio);
5751 else
5752 btrfsic_submit_bio(rw, bio);
5753 }
5754
5755 static int breakup_stripe_bio(struct btrfs_root *root, struct btrfs_bio *bbio,
5756 struct bio *first_bio, struct btrfs_device *dev,
5757 int dev_nr, int rw, int async)
5758 {
5759 struct bio_vec *bvec = first_bio->bi_io_vec;
5760 struct bio *bio;
5761 int nr_vecs = bio_get_nr_vecs(dev->bdev);
5762 u64 physical = bbio->stripes[dev_nr].physical;
5763
5764 again:
5765 bio = btrfs_bio_alloc(dev->bdev, physical >> 9, nr_vecs, GFP_NOFS);
5766 if (!bio)
5767 return -ENOMEM;
5768
5769 while (bvec <= (first_bio->bi_io_vec + first_bio->bi_vcnt - 1)) {
5770 if (bio_add_page(bio, bvec->bv_page, bvec->bv_len,
5771 bvec->bv_offset) < bvec->bv_len) {
5772 u64 len = bio->bi_iter.bi_size;
5773
5774 atomic_inc(&bbio->stripes_pending);
5775 submit_stripe_bio(root, bbio, bio, physical, dev_nr,
5776 rw, async);
5777 physical += len;
5778 goto again;
5779 }
5780 bvec++;
5781 }
5782
5783 submit_stripe_bio(root, bbio, bio, physical, dev_nr, rw, async);
5784 return 0;
5785 }
5786
5787 static void bbio_error(struct btrfs_bio *bbio, struct bio *bio, u64 logical)
5788 {
5789 atomic_inc(&bbio->error);
5790 if (atomic_dec_and_test(&bbio->stripes_pending)) {
5791 /* Shoud be the original bio. */
5792 WARN_ON(bio != bbio->orig_bio);
5793
5794 bio->bi_private = bbio->private;
5795 bio->bi_end_io = bbio->end_io;
5796 btrfs_io_bio(bio)->mirror_num = bbio->mirror_num;
5797 bio->bi_iter.bi_sector = logical >> 9;
5798
5799 btrfs_end_bbio(bbio, bio, -EIO);
5800 }
5801 }
5802
5803 int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio,
5804 int mirror_num, int async_submit)
5805 {
5806 struct btrfs_device *dev;
5807 struct bio *first_bio = bio;
5808 u64 logical = (u64)bio->bi_iter.bi_sector << 9;
5809 u64 length = 0;
5810 u64 map_length;
5811 u64 *raid_map = NULL;
5812 int ret;
5813 int dev_nr = 0;
5814 int total_devs = 1;
5815 struct btrfs_bio *bbio = NULL;
5816
5817 length = bio->bi_iter.bi_size;
5818 map_length = length;
5819
5820 btrfs_bio_counter_inc_blocked(root->fs_info);
5821 ret = __btrfs_map_block(root->fs_info, rw, logical, &map_length, &bbio,
5822 mirror_num, &raid_map);
5823 if (ret) {
5824 btrfs_bio_counter_dec(root->fs_info);
5825 return ret;
5826 }
5827
5828 total_devs = bbio->num_stripes;
5829 bbio->orig_bio = first_bio;
5830 bbio->private = first_bio->bi_private;
5831 bbio->end_io = first_bio->bi_end_io;
5832 bbio->fs_info = root->fs_info;
5833 atomic_set(&bbio->stripes_pending, bbio->num_stripes);
5834
5835 if (raid_map) {
5836 /* In this case, map_length has been set to the length of
5837 a single stripe; not the whole write */
5838 if (rw & WRITE) {
5839 ret = raid56_parity_write(root, bio, bbio,
5840 raid_map, map_length);
5841 } else {
5842 ret = raid56_parity_recover(root, bio, bbio,
5843 raid_map, map_length,
5844 mirror_num, 1);
5845 }
5846
5847 btrfs_bio_counter_dec(root->fs_info);
5848 return ret;
5849 }
5850
5851 if (map_length < length) {
5852 btrfs_crit(root->fs_info, "mapping failed logical %llu bio len %llu len %llu",
5853 logical, length, map_length);
5854 BUG();
5855 }
5856
5857 while (dev_nr < total_devs) {
5858 dev = bbio->stripes[dev_nr].dev;
5859 if (!dev || !dev->bdev || (rw & WRITE && !dev->writeable)) {
5860 bbio_error(bbio, first_bio, logical);
5861 dev_nr++;
5862 continue;
5863 }
5864
5865 /*
5866 * Check and see if we're ok with this bio based on it's size
5867 * and offset with the given device.
5868 */
5869 if (!bio_size_ok(dev->bdev, first_bio,
5870 bbio->stripes[dev_nr].physical >> 9)) {
5871 ret = breakup_stripe_bio(root, bbio, first_bio, dev,
5872 dev_nr, rw, async_submit);
5873 BUG_ON(ret);
5874 dev_nr++;
5875 continue;
5876 }
5877
5878 if (dev_nr < total_devs - 1) {
5879 bio = btrfs_bio_clone(first_bio, GFP_NOFS);
5880 BUG_ON(!bio); /* -ENOMEM */
5881 } else {
5882 bio = first_bio;
5883 bbio->flags |= BTRFS_BIO_ORIG_BIO_SUBMITTED;
5884 }
5885
5886 submit_stripe_bio(root, bbio, bio,
5887 bbio->stripes[dev_nr].physical, dev_nr, rw,
5888 async_submit);
5889 dev_nr++;
5890 }
5891 btrfs_bio_counter_dec(root->fs_info);
5892 return 0;
5893 }
5894
5895 struct btrfs_device *btrfs_find_device(struct btrfs_fs_info *fs_info, u64 devid,
5896 u8 *uuid, u8 *fsid)
5897 {
5898 struct btrfs_device *device;
5899 struct btrfs_fs_devices *cur_devices;
5900
5901 cur_devices = fs_info->fs_devices;
5902 while (cur_devices) {
5903 if (!fsid ||
5904 !memcmp(cur_devices->fsid, fsid, BTRFS_UUID_SIZE)) {
5905 device = __find_device(&cur_devices->devices,
5906 devid, uuid);
5907 if (device)
5908 return device;
5909 }
5910 cur_devices = cur_devices->seed;
5911 }
5912 return NULL;
5913 }
5914
5915 static struct btrfs_device *add_missing_dev(struct btrfs_root *root,
5916 struct btrfs_fs_devices *fs_devices,
5917 u64 devid, u8 *dev_uuid)
5918 {
5919 struct btrfs_device *device;
5920
5921 device = btrfs_alloc_device(NULL, &devid, dev_uuid);
5922 if (IS_ERR(device))
5923 return NULL;
5924
5925 list_add(&device->dev_list, &fs_devices->devices);
5926 device->fs_devices = fs_devices;
5927 fs_devices->num_devices++;
5928
5929 device->missing = 1;
5930 fs_devices->missing_devices++;
5931
5932 return device;
5933 }
5934
5935 /**
5936 * btrfs_alloc_device - allocate struct btrfs_device
5937 * @fs_info: used only for generating a new devid, can be NULL if
5938 * devid is provided (i.e. @devid != NULL).
5939 * @devid: a pointer to devid for this device. If NULL a new devid
5940 * is generated.
5941 * @uuid: a pointer to UUID for this device. If NULL a new UUID
5942 * is generated.
5943 *
5944 * Return: a pointer to a new &struct btrfs_device on success; ERR_PTR()
5945 * on error. Returned struct is not linked onto any lists and can be
5946 * destroyed with kfree() right away.
5947 */
5948 struct btrfs_device *btrfs_alloc_device(struct btrfs_fs_info *fs_info,
5949 const u64 *devid,
5950 const u8 *uuid)
5951 {
5952 struct btrfs_device *dev;
5953 u64 tmp;
5954
5955 if (WARN_ON(!devid && !fs_info))
5956 return ERR_PTR(-EINVAL);
5957
5958 dev = __alloc_device();
5959 if (IS_ERR(dev))
5960 return dev;
5961
5962 if (devid)
5963 tmp = *devid;
5964 else {
5965 int ret;
5966
5967 ret = find_next_devid(fs_info, &tmp);
5968 if (ret) {
5969 kfree(dev);
5970 return ERR_PTR(ret);
5971 }
5972 }
5973 dev->devid = tmp;
5974
5975 if (uuid)
5976 memcpy(dev->uuid, uuid, BTRFS_UUID_SIZE);
5977 else
5978 generate_random_uuid(dev->uuid);
5979
5980 btrfs_init_work(&dev->work, btrfs_submit_helper,
5981 pending_bios_fn, NULL, NULL);
5982
5983 return dev;
5984 }
5985
5986 static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key,
5987 struct extent_buffer *leaf,
5988 struct btrfs_chunk *chunk)
5989 {
5990 struct btrfs_mapping_tree *map_tree = &root->fs_info->mapping_tree;
5991 struct map_lookup *map;
5992 struct extent_map *em;
5993 u64 logical;
5994 u64 length;
5995 u64 devid;
5996 u8 uuid[BTRFS_UUID_SIZE];
5997 int num_stripes;
5998 int ret;
5999 int i;
6000
6001 logical = key->offset;
6002 length = btrfs_chunk_length(leaf, chunk);
6003
6004 read_lock(&map_tree->map_tree.lock);
6005 em = lookup_extent_mapping(&map_tree->map_tree, logical, 1);
6006 read_unlock(&map_tree->map_tree.lock);
6007
6008 /* already mapped? */
6009 if (em && em->start <= logical && em->start + em->len > logical) {
6010 free_extent_map(em);
6011 return 0;
6012 } else if (em) {
6013 free_extent_map(em);
6014 }
6015
6016 em = alloc_extent_map();
6017 if (!em)
6018 return -ENOMEM;
6019 num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
6020 map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
6021 if (!map) {
6022 free_extent_map(em);
6023 return -ENOMEM;
6024 }
6025
6026 set_bit(EXTENT_FLAG_FS_MAPPING, &em->flags);
6027 em->bdev = (struct block_device *)map;
6028 em->start = logical;
6029 em->len = length;
6030 em->orig_start = 0;
6031 em->block_start = 0;
6032 em->block_len = em->len;
6033
6034 map->num_stripes = num_stripes;
6035 map->io_width = btrfs_chunk_io_width(leaf, chunk);
6036 map->io_align = btrfs_chunk_io_align(leaf, chunk);
6037 map->sector_size = btrfs_chunk_sector_size(leaf, chunk);
6038 map->stripe_len = btrfs_chunk_stripe_len(leaf, chunk);
6039 map->type = btrfs_chunk_type(leaf, chunk);
6040 map->sub_stripes = btrfs_chunk_sub_stripes(leaf, chunk);
6041 for (i = 0; i < num_stripes; i++) {
6042 map->stripes[i].physical =
6043 btrfs_stripe_offset_nr(leaf, chunk, i);
6044 devid = btrfs_stripe_devid_nr(leaf, chunk, i);
6045 read_extent_buffer(leaf, uuid, (unsigned long)
6046 btrfs_stripe_dev_uuid_nr(chunk, i),
6047 BTRFS_UUID_SIZE);
6048 map->stripes[i].dev = btrfs_find_device(root->fs_info, devid,
6049 uuid, NULL);
6050 if (!map->stripes[i].dev && !btrfs_test_opt(root, DEGRADED)) {
6051 free_extent_map(em);
6052 return -EIO;
6053 }
6054 if (!map->stripes[i].dev) {
6055 map->stripes[i].dev =
6056 add_missing_dev(root, root->fs_info->fs_devices,
6057 devid, uuid);
6058 if (!map->stripes[i].dev) {
6059 free_extent_map(em);
6060 return -EIO;
6061 }
6062 }
6063 map->stripes[i].dev->in_fs_metadata = 1;
6064 }
6065
6066 write_lock(&map_tree->map_tree.lock);
6067 ret = add_extent_mapping(&map_tree->map_tree, em, 0);
6068 write_unlock(&map_tree->map_tree.lock);
6069 BUG_ON(ret); /* Tree corruption */
6070 free_extent_map(em);
6071
6072 return 0;
6073 }
6074
6075 static void fill_device_from_item(struct extent_buffer *leaf,
6076 struct btrfs_dev_item *dev_item,
6077 struct btrfs_device *device)
6078 {
6079 unsigned long ptr;
6080
6081 device->devid = btrfs_device_id(leaf, dev_item);
6082 device->disk_total_bytes = btrfs_device_total_bytes(leaf, dev_item);
6083 device->total_bytes = device->disk_total_bytes;
6084 device->commit_total_bytes = device->disk_total_bytes;
6085 device->bytes_used = btrfs_device_bytes_used(leaf, dev_item);
6086 device->commit_bytes_used = device->bytes_used;
6087 device->type = btrfs_device_type(leaf, dev_item);
6088 device->io_align = btrfs_device_io_align(leaf, dev_item);
6089 device->io_width = btrfs_device_io_width(leaf, dev_item);
6090 device->sector_size = btrfs_device_sector_size(leaf, dev_item);
6091 WARN_ON(device->devid == BTRFS_DEV_REPLACE_DEVID);
6092 device->is_tgtdev_for_dev_replace = 0;
6093
6094 ptr = btrfs_device_uuid(dev_item);
6095 read_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
6096 }
6097
6098 static struct btrfs_fs_devices *open_seed_devices(struct btrfs_root *root,
6099 u8 *fsid)
6100 {
6101 struct btrfs_fs_devices *fs_devices;
6102 int ret;
6103
6104 BUG_ON(!mutex_is_locked(&uuid_mutex));
6105
6106 fs_devices = root->fs_info->fs_devices->seed;
6107 while (fs_devices) {
6108 if (!memcmp(fs_devices->fsid, fsid, BTRFS_UUID_SIZE))
6109 return fs_devices;
6110
6111 fs_devices = fs_devices->seed;
6112 }
6113
6114 fs_devices = find_fsid(fsid);
6115 if (!fs_devices) {
6116 if (!btrfs_test_opt(root, DEGRADED))
6117 return ERR_PTR(-ENOENT);
6118
6119 fs_devices = alloc_fs_devices(fsid);
6120 if (IS_ERR(fs_devices))
6121 return fs_devices;
6122
6123 fs_devices->seeding = 1;
6124 fs_devices->opened = 1;
6125 return fs_devices;
6126 }
6127
6128 fs_devices = clone_fs_devices(fs_devices);
6129 if (IS_ERR(fs_devices))
6130 return fs_devices;
6131
6132 ret = __btrfs_open_devices(fs_devices, FMODE_READ,
6133 root->fs_info->bdev_holder);
6134 if (ret) {
6135 free_fs_devices(fs_devices);
6136 fs_devices = ERR_PTR(ret);
6137 goto out;
6138 }
6139
6140 if (!fs_devices->seeding) {
6141 __btrfs_close_devices(fs_devices);
6142 free_fs_devices(fs_devices);
6143 fs_devices = ERR_PTR(-EINVAL);
6144 goto out;
6145 }
6146
6147 fs_devices->seed = root->fs_info->fs_devices->seed;
6148 root->fs_info->fs_devices->seed = fs_devices;
6149 out:
6150 return fs_devices;
6151 }
6152
6153 static int read_one_dev(struct btrfs_root *root,
6154 struct extent_buffer *leaf,
6155 struct btrfs_dev_item *dev_item)
6156 {
6157 struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
6158 struct btrfs_device *device;
6159 u64 devid;
6160 int ret;
6161 u8 fs_uuid[BTRFS_UUID_SIZE];
6162 u8 dev_uuid[BTRFS_UUID_SIZE];
6163
6164 devid = btrfs_device_id(leaf, dev_item);
6165 read_extent_buffer(leaf, dev_uuid, btrfs_device_uuid(dev_item),
6166 BTRFS_UUID_SIZE);
6167 read_extent_buffer(leaf, fs_uuid, btrfs_device_fsid(dev_item),
6168 BTRFS_UUID_SIZE);
6169
6170 if (memcmp(fs_uuid, root->fs_info->fsid, BTRFS_UUID_SIZE)) {
6171 fs_devices = open_seed_devices(root, fs_uuid);
6172 if (IS_ERR(fs_devices))
6173 return PTR_ERR(fs_devices);
6174 }
6175
6176 device = btrfs_find_device(root->fs_info, devid, dev_uuid, fs_uuid);
6177 if (!device) {
6178 if (!btrfs_test_opt(root, DEGRADED))
6179 return -EIO;
6180
6181 btrfs_warn(root->fs_info, "devid %llu missing", devid);
6182 device = add_missing_dev(root, fs_devices, devid, dev_uuid);
6183 if (!device)
6184 return -ENOMEM;
6185 } else {
6186 if (!device->bdev && !btrfs_test_opt(root, DEGRADED))
6187 return -EIO;
6188
6189 if(!device->bdev && !device->missing) {
6190 /*
6191 * this happens when a device that was properly setup
6192 * in the device info lists suddenly goes bad.
6193 * device->bdev is NULL, and so we have to set
6194 * device->missing to one here
6195 */
6196 device->fs_devices->missing_devices++;
6197 device->missing = 1;
6198 }
6199
6200 /* Move the device to its own fs_devices */
6201 if (device->fs_devices != fs_devices) {
6202 ASSERT(device->missing);
6203
6204 list_move(&device->dev_list, &fs_devices->devices);
6205 device->fs_devices->num_devices--;
6206 fs_devices->num_devices++;
6207
6208 device->fs_devices->missing_devices--;
6209 fs_devices->missing_devices++;
6210
6211 device->fs_devices = fs_devices;
6212 }
6213 }
6214
6215 if (device->fs_devices != root->fs_info->fs_devices) {
6216 BUG_ON(device->writeable);
6217 if (device->generation !=
6218 btrfs_device_generation(leaf, dev_item))
6219 return -EINVAL;
6220 }
6221
6222 fill_device_from_item(leaf, dev_item, device);
6223 device->in_fs_metadata = 1;
6224 if (device->writeable && !device->is_tgtdev_for_dev_replace) {
6225 device->fs_devices->total_rw_bytes += device->total_bytes;
6226 spin_lock(&root->fs_info->free_chunk_lock);
6227 root->fs_info->free_chunk_space += device->total_bytes -
6228 device->bytes_used;
6229 spin_unlock(&root->fs_info->free_chunk_lock);
6230 }
6231 ret = 0;
6232 return ret;
6233 }
6234
6235 int btrfs_read_sys_array(struct btrfs_root *root)
6236 {
6237 struct btrfs_super_block *super_copy = root->fs_info->super_copy;
6238 struct extent_buffer *sb;
6239 struct btrfs_disk_key *disk_key;
6240 struct btrfs_chunk *chunk;
6241 u8 *ptr;
6242 unsigned long sb_ptr;
6243 int ret = 0;
6244 u32 num_stripes;
6245 u32 array_size;
6246 u32 len = 0;
6247 u32 cur;
6248 struct btrfs_key key;
6249
6250 ASSERT(BTRFS_SUPER_INFO_SIZE <= root->nodesize);
6251 /*
6252 * This will create extent buffer of nodesize, superblock size is
6253 * fixed to BTRFS_SUPER_INFO_SIZE. If nodesize > sb size, this will
6254 * overallocate but we can keep it as-is, only the first page is used.
6255 */
6256 sb = btrfs_find_create_tree_block(root, BTRFS_SUPER_INFO_OFFSET);
6257 if (!sb)
6258 return -ENOMEM;
6259 btrfs_set_buffer_uptodate(sb);
6260 btrfs_set_buffer_lockdep_class(root->root_key.objectid, sb, 0);
6261 /*
6262 * The sb extent buffer is artifical and just used to read the system array.
6263 * btrfs_set_buffer_uptodate() call does not properly mark all it's
6264 * pages up-to-date when the page is larger: extent does not cover the
6265 * whole page and consequently check_page_uptodate does not find all
6266 * the page's extents up-to-date (the hole beyond sb),
6267 * write_extent_buffer then triggers a WARN_ON.
6268 *
6269 * Regular short extents go through mark_extent_buffer_dirty/writeback cycle,
6270 * but sb spans only this function. Add an explicit SetPageUptodate call
6271 * to silence the warning eg. on PowerPC 64.
6272 */
6273 if (PAGE_CACHE_SIZE > BTRFS_SUPER_INFO_SIZE)
6274 SetPageUptodate(sb->pages[0]);
6275
6276 write_extent_buffer(sb, super_copy, 0, BTRFS_SUPER_INFO_SIZE);
6277 array_size = btrfs_super_sys_array_size(super_copy);
6278
6279 ptr = super_copy->sys_chunk_array;
6280 sb_ptr = offsetof(struct btrfs_super_block, sys_chunk_array);
6281 cur = 0;
6282
6283 while (cur < array_size) {
6284 disk_key = (struct btrfs_disk_key *)ptr;
6285 btrfs_disk_key_to_cpu(&key, disk_key);
6286
6287 len = sizeof(*disk_key); ptr += len;
6288 sb_ptr += len;
6289 cur += len;
6290
6291 if (key.type == BTRFS_CHUNK_ITEM_KEY) {
6292 chunk = (struct btrfs_chunk *)sb_ptr;
6293 ret = read_one_chunk(root, &key, sb, chunk);
6294 if (ret)
6295 break;
6296 num_stripes = btrfs_chunk_num_stripes(sb, chunk);
6297 len = btrfs_chunk_item_size(num_stripes);
6298 } else {
6299 ret = -EIO;
6300 break;
6301 }
6302 ptr += len;
6303 sb_ptr += len;
6304 cur += len;
6305 }
6306 free_extent_buffer(sb);
6307 return ret;
6308 }
6309
6310 int btrfs_read_chunk_tree(struct btrfs_root *root)
6311 {
6312 struct btrfs_path *path;
6313 struct extent_buffer *leaf;
6314 struct btrfs_key key;
6315 struct btrfs_key found_key;
6316 int ret;
6317 int slot;
6318
6319 root = root->fs_info->chunk_root;
6320
6321 path = btrfs_alloc_path();
6322 if (!path)
6323 return -ENOMEM;
6324
6325 mutex_lock(&uuid_mutex);
6326 lock_chunks(root);
6327
6328 /*
6329 * Read all device items, and then all the chunk items. All
6330 * device items are found before any chunk item (their object id
6331 * is smaller than the lowest possible object id for a chunk
6332 * item - BTRFS_FIRST_CHUNK_TREE_OBJECTID).
6333 */
6334 key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
6335 key.offset = 0;
6336 key.type = 0;
6337 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
6338 if (ret < 0)
6339 goto error;
6340 while (1) {
6341 leaf = path->nodes[0];
6342 slot = path->slots[0];
6343 if (slot >= btrfs_header_nritems(leaf)) {
6344 ret = btrfs_next_leaf(root, path);
6345 if (ret == 0)
6346 continue;
6347 if (ret < 0)
6348 goto error;
6349 break;
6350 }
6351 btrfs_item_key_to_cpu(leaf, &found_key, slot);
6352 if (found_key.type == BTRFS_DEV_ITEM_KEY) {
6353 struct btrfs_dev_item *dev_item;
6354 dev_item = btrfs_item_ptr(leaf, slot,
6355 struct btrfs_dev_item);
6356 ret = read_one_dev(root, leaf, dev_item);
6357 if (ret)
6358 goto error;
6359 } else if (found_key.type == BTRFS_CHUNK_ITEM_KEY) {
6360 struct btrfs_chunk *chunk;
6361 chunk = btrfs_item_ptr(leaf, slot, struct btrfs_chunk);
6362 ret = read_one_chunk(root, &found_key, leaf, chunk);
6363 if (ret)
6364 goto error;
6365 }
6366 path->slots[0]++;
6367 }
6368 ret = 0;
6369 error:
6370 unlock_chunks(root);
6371 mutex_unlock(&uuid_mutex);
6372
6373 btrfs_free_path(path);
6374 return ret;
6375 }
6376
6377 void btrfs_init_devices_late(struct btrfs_fs_info *fs_info)
6378 {
6379 struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
6380 struct btrfs_device *device;
6381
6382 while (fs_devices) {
6383 mutex_lock(&fs_devices->device_list_mutex);
6384 list_for_each_entry(device, &fs_devices->devices, dev_list)
6385 device->dev_root = fs_info->dev_root;
6386 mutex_unlock(&fs_devices->device_list_mutex);
6387
6388 fs_devices = fs_devices->seed;
6389 }
6390 }
6391
6392 static void __btrfs_reset_dev_stats(struct btrfs_device *dev)
6393 {
6394 int i;
6395
6396 for (i = 0; i < BTRFS_DEV_STAT_VALUES_MAX; i++)
6397 btrfs_dev_stat_reset(dev, i);
6398 }
6399
6400 int btrfs_init_dev_stats(struct btrfs_fs_info *fs_info)
6401 {
6402 struct btrfs_key key;
6403 struct btrfs_key found_key;
6404 struct btrfs_root *dev_root = fs_info->dev_root;
6405 struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
6406 struct extent_buffer *eb;
6407 int slot;
6408 int ret = 0;
6409 struct btrfs_device *device;
6410 struct btrfs_path *path = NULL;
6411 int i;
6412
6413 path = btrfs_alloc_path();
6414 if (!path) {
6415 ret = -ENOMEM;
6416 goto out;
6417 }
6418
6419 mutex_lock(&fs_devices->device_list_mutex);
6420 list_for_each_entry(device, &fs_devices->devices, dev_list) {
6421 int item_size;
6422 struct btrfs_dev_stats_item *ptr;
6423
6424 key.objectid = 0;
6425 key.type = BTRFS_DEV_STATS_KEY;
6426 key.offset = device->devid;
6427 ret = btrfs_search_slot(NULL, dev_root, &key, path, 0, 0);
6428 if (ret) {
6429 __btrfs_reset_dev_stats(device);
6430 device->dev_stats_valid = 1;
6431 btrfs_release_path(path);
6432 continue;
6433 }
6434 slot = path->slots[0];
6435 eb = path->nodes[0];
6436 btrfs_item_key_to_cpu(eb, &found_key, slot);
6437 item_size = btrfs_item_size_nr(eb, slot);
6438
6439 ptr = btrfs_item_ptr(eb, slot,
6440 struct btrfs_dev_stats_item);
6441
6442 for (i = 0; i < BTRFS_DEV_STAT_VALUES_MAX; i++) {
6443 if (item_size >= (1 + i) * sizeof(__le64))
6444 btrfs_dev_stat_set(device, i,
6445 btrfs_dev_stats_value(eb, ptr, i));
6446 else
6447 btrfs_dev_stat_reset(device, i);
6448 }
6449
6450 device->dev_stats_valid = 1;
6451 btrfs_dev_stat_print_on_load(device);
6452 btrfs_release_path(path);
6453 }
6454 mutex_unlock(&fs_devices->device_list_mutex);
6455
6456 out:
6457 btrfs_free_path(path);
6458 return ret < 0 ? ret : 0;
6459 }
6460
6461 static int update_dev_stat_item(struct btrfs_trans_handle *trans,
6462 struct btrfs_root *dev_root,
6463 struct btrfs_device *device)
6464 {
6465 struct btrfs_path *path;
6466 struct btrfs_key key;
6467 struct extent_buffer *eb;
6468 struct btrfs_dev_stats_item *ptr;
6469 int ret;
6470 int i;
6471
6472 key.objectid = 0;
6473 key.type = BTRFS_DEV_STATS_KEY;
6474 key.offset = device->devid;
6475
6476 path = btrfs_alloc_path();
6477 BUG_ON(!path);
6478 ret = btrfs_search_slot(trans, dev_root, &key, path, -1, 1);
6479 if (ret < 0) {
6480 printk_in_rcu(KERN_WARNING "BTRFS: "
6481 "error %d while searching for dev_stats item for device %s!\n",
6482 ret, rcu_str_deref(device->name));
6483 goto out;
6484 }
6485
6486 if (ret == 0 &&
6487 btrfs_item_size_nr(path->nodes[0], path->slots[0]) < sizeof(*ptr)) {
6488 /* need to delete old one and insert a new one */
6489 ret = btrfs_del_item(trans, dev_root, path);
6490 if (ret != 0) {
6491 printk_in_rcu(KERN_WARNING "BTRFS: "
6492 "delete too small dev_stats item for device %s failed %d!\n",
6493 rcu_str_deref(device->name), ret);
6494 goto out;
6495 }
6496 ret = 1;
6497 }
6498
6499 if (ret == 1) {
6500 /* need to insert a new item */
6501 btrfs_release_path(path);
6502 ret = btrfs_insert_empty_item(trans, dev_root, path,
6503 &key, sizeof(*ptr));
6504 if (ret < 0) {
6505 printk_in_rcu(KERN_WARNING "BTRFS: "
6506 "insert dev_stats item for device %s failed %d!\n",
6507 rcu_str_deref(device->name), ret);
6508 goto out;
6509 }
6510 }
6511
6512 eb = path->nodes[0];
6513 ptr = btrfs_item_ptr(eb, path->slots[0], struct btrfs_dev_stats_item);
6514 for (i = 0; i < BTRFS_DEV_STAT_VALUES_MAX; i++)
6515 btrfs_set_dev_stats_value(eb, ptr, i,
6516 btrfs_dev_stat_read(device, i));
6517 btrfs_mark_buffer_dirty(eb);
6518
6519 out:
6520 btrfs_free_path(path);
6521 return ret;
6522 }
6523
6524 /*
6525 * called from commit_transaction. Writes all changed device stats to disk.
6526 */
6527 int btrfs_run_dev_stats(struct btrfs_trans_handle *trans,
6528 struct btrfs_fs_info *fs_info)
6529 {
6530 struct btrfs_root *dev_root = fs_info->dev_root;
6531 struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
6532 struct btrfs_device *device;
6533 int stats_cnt;
6534 int ret = 0;
6535
6536 mutex_lock(&fs_devices->device_list_mutex);
6537 list_for_each_entry(device, &fs_devices->devices, dev_list) {
6538 if (!device->dev_stats_valid || !btrfs_dev_stats_dirty(device))
6539 continue;
6540
6541 stats_cnt = atomic_read(&device->dev_stats_ccnt);
6542 ret = update_dev_stat_item(trans, dev_root, device);
6543 if (!ret)
6544 atomic_sub(stats_cnt, &device->dev_stats_ccnt);
6545 }
6546 mutex_unlock(&fs_devices->device_list_mutex);
6547
6548 return ret;
6549 }
6550
6551 void btrfs_dev_stat_inc_and_print(struct btrfs_device *dev, int index)
6552 {
6553 btrfs_dev_stat_inc(dev, index);
6554 btrfs_dev_stat_print_on_error(dev);
6555 }
6556
6557 static void btrfs_dev_stat_print_on_error(struct btrfs_device *dev)
6558 {
6559 if (!dev->dev_stats_valid)
6560 return;
6561 printk_ratelimited_in_rcu(KERN_ERR "BTRFS: "
6562 "bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u\n",
6563 rcu_str_deref(dev->name),
6564 btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_WRITE_ERRS),
6565 btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_READ_ERRS),
6566 btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_FLUSH_ERRS),
6567 btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_CORRUPTION_ERRS),
6568 btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_GENERATION_ERRS));
6569 }
6570
6571 static void btrfs_dev_stat_print_on_load(struct btrfs_device *dev)
6572 {
6573 int i;
6574
6575 for (i = 0; i < BTRFS_DEV_STAT_VALUES_MAX; i++)
6576 if (btrfs_dev_stat_read(dev, i) != 0)
6577 break;
6578 if (i == BTRFS_DEV_STAT_VALUES_MAX)
6579 return; /* all values == 0, suppress message */
6580
6581 printk_in_rcu(KERN_INFO "BTRFS: "
6582 "bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u\n",
6583 rcu_str_deref(dev->name),
6584 btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_WRITE_ERRS),
6585 btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_READ_ERRS),
6586 btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_FLUSH_ERRS),
6587 btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_CORRUPTION_ERRS),
6588 btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_GENERATION_ERRS));
6589 }
6590
6591 int btrfs_get_dev_stats(struct btrfs_root *root,
6592 struct btrfs_ioctl_get_dev_stats *stats)
6593 {
6594 struct btrfs_device *dev;
6595 struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
6596 int i;
6597
6598 mutex_lock(&fs_devices->device_list_mutex);
6599 dev = btrfs_find_device(root->fs_info, stats->devid, NULL, NULL);
6600 mutex_unlock(&fs_devices->device_list_mutex);
6601
6602 if (!dev) {
6603 btrfs_warn(root->fs_info, "get dev_stats failed, device not found");
6604 return -ENODEV;
6605 } else if (!dev->dev_stats_valid) {
6606 btrfs_warn(root->fs_info, "get dev_stats failed, not yet valid");
6607 return -ENODEV;
6608 } else if (stats->flags & BTRFS_DEV_STATS_RESET) {
6609 for (i = 0; i < BTRFS_DEV_STAT_VALUES_MAX; i++) {
6610 if (stats->nr_items > i)
6611 stats->values[i] =
6612 btrfs_dev_stat_read_and_reset(dev, i);
6613 else
6614 btrfs_dev_stat_reset(dev, i);
6615 }
6616 } else {
6617 for (i = 0; i < BTRFS_DEV_STAT_VALUES_MAX; i++)
6618 if (stats->nr_items > i)
6619 stats->values[i] = btrfs_dev_stat_read(dev, i);
6620 }
6621 if (stats->nr_items > BTRFS_DEV_STAT_VALUES_MAX)
6622 stats->nr_items = BTRFS_DEV_STAT_VALUES_MAX;
6623 return 0;
6624 }
6625
6626 int btrfs_scratch_superblock(struct btrfs_device *device)
6627 {
6628 struct buffer_head *bh;
6629 struct btrfs_super_block *disk_super;
6630
6631 bh = btrfs_read_dev_super(device->bdev);
6632 if (!bh)
6633 return -EINVAL;
6634 disk_super = (struct btrfs_super_block *)bh->b_data;
6635
6636 memset(&disk_super->magic, 0, sizeof(disk_super->magic));
6637 set_buffer_dirty(bh);
6638 sync_dirty_buffer(bh);
6639 brelse(bh);
6640
6641 return 0;
6642 }
6643
6644 /*
6645 * Update the size of all devices, which is used for writing out the
6646 * super blocks.
6647 */
6648 void btrfs_update_commit_device_size(struct btrfs_fs_info *fs_info)
6649 {
6650 struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
6651 struct btrfs_device *curr, *next;
6652
6653 if (list_empty(&fs_devices->resized_devices))
6654 return;
6655
6656 mutex_lock(&fs_devices->device_list_mutex);
6657 lock_chunks(fs_info->dev_root);
6658 list_for_each_entry_safe(curr, next, &fs_devices->resized_devices,
6659 resized_list) {
6660 list_del_init(&curr->resized_list);
6661 curr->commit_total_bytes = curr->disk_total_bytes;
6662 }
6663 unlock_chunks(fs_info->dev_root);
6664 mutex_unlock(&fs_devices->device_list_mutex);
6665 }
6666
6667 /* Must be invoked during the transaction commit */
6668 void btrfs_update_commit_device_bytes_used(struct btrfs_root *root,
6669 struct btrfs_transaction *transaction)
6670 {
6671 struct extent_map *em;
6672 struct map_lookup *map;
6673 struct btrfs_device *dev;
6674 int i;
6675
6676 if (list_empty(&transaction->pending_chunks))
6677 return;
6678
6679 /* In order to kick the device replace finish process */
6680 lock_chunks(root);
6681 list_for_each_entry(em, &transaction->pending_chunks, list) {
6682 map = (struct map_lookup *)em->bdev;
6683
6684 for (i = 0; i < map->num_stripes; i++) {
6685 dev = map->stripes[i].dev;
6686 dev->commit_bytes_used = dev->bytes_used;
6687 }
6688 }
6689 unlock_chunks(root);
6690 }
ubuntu-bionic-kernel mirror