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 <asm/div64.h>
  27 #include "compat.h"
  28 #include "ctree.h"
  29 #include "extent_map.h"
  30 #include "disk-io.h"
  31 #include "transaction.h"
  32 #include "print-tree.h"
  33 #include "volumes.h"
  34 #include "async-thread.h"
  35
  36 static int init_first_rw_device(struct btrfs_trans_handle *trans,
  37                                 struct btrfs_root *root,
  38                                 struct btrfs_device *device);
  39 static int btrfs_relocate_sys_chunks(struct btrfs_root *root);
  40
  41 static DEFINE_MUTEX(uuid_mutex);
  42 static LIST_HEAD(fs_uuids);
  43
  44 static void lock_chunks(struct btrfs_root *root)
  45 {
  46         mutex_lock(&root->fs_info->chunk_mutex);
  47 }
  48
  49 static void unlock_chunks(struct btrfs_root *root)
  50 {
  51         mutex_unlock(&root->fs_info->chunk_mutex);
  52 }
  53
  54 static void free_fs_devices(struct btrfs_fs_devices *fs_devices)
  55 {
  56         struct btrfs_device *device;
  57         WARN_ON(fs_devices->opened);
  58         while (!list_empty(&fs_devices->devices)) {
  59                 device = list_entry(fs_devices->devices.next,
  60                                     struct btrfs_device, dev_list);
  61                 list_del(&device->dev_list);
  62                 kfree(device->name);
  63                 kfree(device);
  64         }
  65         kfree(fs_devices);
  66 }
  67
  68 int btrfs_cleanup_fs_uuids(void)
  69 {
  70         struct btrfs_fs_devices *fs_devices;
  71
  72         while (!list_empty(&fs_uuids)) {
  73                 fs_devices = list_entry(fs_uuids.next,
  74                                         struct btrfs_fs_devices, list);
  75                 list_del(&fs_devices->list);
  76                 free_fs_devices(fs_devices);
  77         }
  78         return 0;
  79 }
  80
  81 static noinline struct btrfs_device *__find_device(struct list_head *head,
  82                                                    u64 devid, u8 *uuid)
  83 {
  84         struct btrfs_device *dev;
  85
  86         list_for_each_entry(dev, head, dev_list) {
  87                 if (dev->devid == devid &&
  88                     (!uuid || !memcmp(dev->uuid, uuid, BTRFS_UUID_SIZE))) {
  89                         return dev;
  90                 }
  91         }
  92         return NULL;
  93 }
  94
  95 static noinline struct btrfs_fs_devices *find_fsid(u8 *fsid)
  96 {
  97         struct btrfs_fs_devices *fs_devices;
  98
  99         list_for_each_entry(fs_devices, &fs_uuids, list) {
 100                 if (memcmp(fsid, fs_devices->fsid, BTRFS_FSID_SIZE) == 0)
 101                         return fs_devices;
 102         }
 103         return NULL;
 104 }
 105
 106 static void requeue_list(struct btrfs_pending_bios *pending_bios,
 107                         struct bio *head, struct bio *tail)
 108 {
 109
 110         struct bio *old_head;
 111
 112         old_head = pending_bios->head;
 113         pending_bios->head = head;
 114         if (pending_bios->tail)
 115                 tail->bi_next = old_head;
 116         else
 117                 pending_bios->tail = tail;
 118 }
 119
 120 /*
 121  * we try to collect pending bios for a device so we don't get a large
 122  * number of procs sending bios down to the same device.  This greatly
 123  * improves the schedulers ability to collect and merge the bios.
 124  *
 125  * But, it also turns into a long list of bios to process and that is sure
 126  * to eventually make the worker thread block.  The solution here is to
 127  * make some progress and then put this work struct back at the end of
 128  * the list if the block device is congested.  This way, multiple devices
 129  * can make progress from a single worker thread.
 130  */
 131 static noinline int run_scheduled_bios(struct btrfs_device *device)
 132 {
 133         struct bio *pending;
 134         struct backing_dev_info *bdi;
 135         struct btrfs_fs_info *fs_info;
 136         struct btrfs_pending_bios *pending_bios;
 137         struct bio *tail;
 138         struct bio *cur;
 139         int again = 0;
 140         unsigned long num_run;
 141         unsigned long batch_run = 0;
 142         unsigned long limit;
 143         unsigned long last_waited = 0;
 144         int force_reg = 0;
 145         int sync_pending = 0;
 146         struct blk_plug plug;
 147
 148         /*
 149          * this function runs all the bios we've collected for
 150          * a particular device.  We don't want to wander off to
 151          * another device without first sending all of these down.
 152          * So, setup a plug here and finish it off before we return
 153          */
 154         blk_start_plug(&plug);
 155
 156         bdi = blk_get_backing_dev_info(device->bdev);
 157         fs_info = device->dev_root->fs_info;
 158         limit = btrfs_async_submit_limit(fs_info);
 159         limit = limit * 2 / 3;
 160
 161 loop:
 162         spin_lock(&device->io_lock);
 163
 164 loop_lock:
 165         num_run = 0;
 166
 167         /* take all the bios off the list at once and process them
 168          * later on (without the lock held).  But, remember the
 169          * tail and other pointers so the bios can be properly reinserted
 170          * into the list if we hit congestion
 171          */
 172         if (!force_reg && device->pending_sync_bios.head) {
 173                 pending_bios = &device->pending_sync_bios;
 174                 force_reg = 1;
 175         } else {
 176                 pending_bios = &device->pending_bios;
 177                 force_reg = 0;
 178         }
 179
 180         pending = pending_bios->head;
 181         tail = pending_bios->tail;
 182         WARN_ON(pending && !tail);
 183
 184         /*
 185          * if pending was null this time around, no bios need processing
 186          * at all and we can stop.  Otherwise it'll loop back up again
 187          * and do an additional check so no bios are missed.
 188          *
 189          * device->running_pending is used to synchronize with the
 190          * schedule_bio code.
 191          */
 192         if (device->pending_sync_bios.head == NULL &&
 193             device->pending_bios.head == NULL) {
 194                 again = 0;
 195                 device->running_pending = 0;
 196         } else {
 197                 again = 1;
 198                 device->running_pending = 1;
 199         }
 200
 201         pending_bios->head = NULL;
 202         pending_bios->tail = NULL;
 203
 204         spin_unlock(&device->io_lock);
 205
 206         while (pending) {
 207
 208                 rmb();
 209                 /* we want to work on both lists, but do more bios on the
 210                  * sync list than the regular list
 211                  */
 212                 if ((num_run > 32 &&
 213                     pending_bios != &device->pending_sync_bios &&
 214                     device->pending_sync_bios.head) ||
 215                    (num_run > 64 && pending_bios == &device->pending_sync_bios &&
 216                     device->pending_bios.head)) {
 217                         spin_lock(&device->io_lock);
 218                         requeue_list(pending_bios, pending, tail);
 219                         goto loop_lock;
 220                 }
 221
 222                 cur = pending;
 223                 pending = pending->bi_next;
 224                 cur->bi_next = NULL;
 225                 atomic_dec(&fs_info->nr_async_bios);
 226
 227                 if (atomic_read(&fs_info->nr_async_bios) < limit &&
 228                     waitqueue_active(&fs_info->async_submit_wait))
 229                         wake_up(&fs_info->async_submit_wait);
 230
 231                 BUG_ON(atomic_read(&cur->bi_cnt) == 0);
 232
 233                 /*
 234                  * if we're doing the sync list, record that our
 235                  * plug has some sync requests on it
 236                  *
 237                  * If we're doing the regular list and there are
 238                  * sync requests sitting around, unplug before
 239                  * we add more
 240                  */
 241                 if (pending_bios == &device->pending_sync_bios) {
 242                         sync_pending = 1;
 243                 } else if (sync_pending) {
 244                         blk_finish_plug(&plug);
 245                         blk_start_plug(&plug);
 246                         sync_pending = 0;
 247                 }
 248
 249                 submit_bio(cur->bi_rw, cur);
 250                 num_run++;
 251                 batch_run++;
 252                 if (need_resched())
 253                         cond_resched();
 254
 255                 /*
 256                  * we made progress, there is more work to do and the bdi
 257                  * is now congested.  Back off and let other work structs
 258                  * run instead
 259                  */
 260                 if (pending && bdi_write_congested(bdi) && batch_run > 8 &&
 261                     fs_info->fs_devices->open_devices > 1) {
 262                         struct io_context *ioc;
 263
 264                         ioc = current->io_context;
 265
 266                         /*
 267                          * the main goal here is that we don't want to
 268                          * block if we're going to be able to submit
 269                          * more requests without blocking.
 270                          *
 271                          * This code does two great things, it pokes into
 272                          * the elevator code from a filesystem _and_
 273                          * it makes assumptions about how batching works.
 274                          */
 275                         if (ioc && ioc->nr_batch_requests > 0 &&
 276                             time_before(jiffies, ioc->last_waited + HZ/50UL) &&
 277                             (last_waited == 0 ||
 278                              ioc->last_waited == last_waited)) {
 279                                 /*
 280                                  * we want to go through our batch of
 281                                  * requests and stop.  So, we copy out
 282                                  * the ioc->last_waited time and test
 283                                  * against it before looping
 284                                  */
 285                                 last_waited = ioc->last_waited;
 286                                 if (need_resched())
 287                                         cond_resched();
 288                                 continue;
 289                         }
 290                         spin_lock(&device->io_lock);
 291                         requeue_list(pending_bios, pending, tail);
 292                         device->running_pending = 1;
 293
 294                         spin_unlock(&device->io_lock);
 295                         btrfs_requeue_work(&device->work);
 296                         goto done;
 297                 }
 298         }
 299
 300         cond_resched();
 301         if (again)
 302                 goto loop;
 303
 304         spin_lock(&device->io_lock);
 305         if (device->pending_bios.head || device->pending_sync_bios.head)
 306                 goto loop_lock;
 307         spin_unlock(&device->io_lock);
 308
 309 done:
 310         blk_finish_plug(&plug);
 311         return 0;
 312 }
 313
 314 static void pending_bios_fn(struct btrfs_work *work)
 315 {
 316         struct btrfs_device *device;
 317
 318         device = container_of(work, struct btrfs_device, work);
 319         run_scheduled_bios(device);
 320 }
 321
 322 static noinline int device_list_add(const char *path,
 323                            struct btrfs_super_block *disk_super,
 324                            u64 devid, struct btrfs_fs_devices **fs_devices_ret)
 325 {
 326         struct btrfs_device *device;
 327         struct btrfs_fs_devices *fs_devices;
 328         u64 found_transid = btrfs_super_generation(disk_super);
 329         char *name;
 330
 331         fs_devices = find_fsid(disk_super->fsid);
 332         if (!fs_devices) {
 333                 fs_devices = kzalloc(sizeof(*fs_devices), GFP_NOFS);
 334                 if (!fs_devices)
 335                         return -ENOMEM;
 336                 INIT_LIST_HEAD(&fs_devices->devices);
 337                 INIT_LIST_HEAD(&fs_devices->alloc_list);
 338                 list_add(&fs_devices->list, &fs_uuids);
 339                 memcpy(fs_devices->fsid, disk_super->fsid, BTRFS_FSID_SIZE);
 340                 fs_devices->latest_devid = devid;
 341                 fs_devices->latest_trans = found_transid;
 342                 mutex_init(&fs_devices->device_list_mutex);
 343                 device = NULL;
 344         } else {
 345                 device = __find_device(&fs_devices->devices, devid,
 346                                        disk_super->dev_item.uuid);
 347         }
 348         if (!device) {
 349                 if (fs_devices->opened)
 350                         return -EBUSY;
 351
 352                 device = kzalloc(sizeof(*device), GFP_NOFS);
 353                 if (!device) {
 354                         /* we can safely leave the fs_devices entry around */
 355                         return -ENOMEM;
 356                 }
 357                 device->devid = devid;
 358                 device->work.func = pending_bios_fn;
 359                 memcpy(device->uuid, disk_super->dev_item.uuid,
 360                        BTRFS_UUID_SIZE);
 361                 spin_lock_init(&device->io_lock);
 362                 device->name = kstrdup(path, GFP_NOFS);
 363                 if (!device->name) {
 364                         kfree(device);
 365                         return -ENOMEM;
 366                 }
 367                 INIT_LIST_HEAD(&device->dev_alloc_list);
 368
 369                 /* init readahead state */
 370                 spin_lock_init(&device->reada_lock);
 371                 device->reada_curr_zone = NULL;
 372                 atomic_set(&device->reada_in_flight, 0);
 373                 device->reada_next = 0;
 374                 INIT_RADIX_TREE(&device->reada_zones, GFP_NOFS & ~__GFP_WAIT);
 375                 INIT_RADIX_TREE(&device->reada_extents, GFP_NOFS & ~__GFP_WAIT);
 376
 377                 mutex_lock(&fs_devices->device_list_mutex);
 378                 list_add_rcu(&device->dev_list, &fs_devices->devices);
 379                 mutex_unlock(&fs_devices->device_list_mutex);
 380
 381                 device->fs_devices = fs_devices;
 382                 fs_devices->num_devices++;
 383         } else if (!device->name || strcmp(device->name, path)) {
 384                 name = kstrdup(path, GFP_NOFS);
 385                 if (!name)
 386                         return -ENOMEM;
 387                 kfree(device->name);
 388                 device->name = name;
 389                 if (device->missing) {
 390                         fs_devices->missing_devices--;
 391                         device->missing = 0;
 392                 }
 393         }
 394
 395         if (found_transid > fs_devices->latest_trans) {
 396                 fs_devices->latest_devid = devid;
 397                 fs_devices->latest_trans = found_transid;
 398         }
 399         *fs_devices_ret = fs_devices;
 400         return 0;
 401 }
 402
 403 static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
 404 {
 405         struct btrfs_fs_devices *fs_devices;
 406         struct btrfs_device *device;
 407         struct btrfs_device *orig_dev;
 408
 409         fs_devices = kzalloc(sizeof(*fs_devices), GFP_NOFS);
 410         if (!fs_devices)
 411                 return ERR_PTR(-ENOMEM);
 412
 413         INIT_LIST_HEAD(&fs_devices->devices);
 414         INIT_LIST_HEAD(&fs_devices->alloc_list);
 415         INIT_LIST_HEAD(&fs_devices->list);
 416         mutex_init(&fs_devices->device_list_mutex);
 417         fs_devices->latest_devid = orig->latest_devid;
 418         fs_devices->latest_trans = orig->latest_trans;
 419         memcpy(fs_devices->fsid, orig->fsid, sizeof(fs_devices->fsid));
 420
 421         /* We have held the volume lock, it is safe to get the devices. */
 422         list_for_each_entry(orig_dev, &orig->devices, dev_list) {
 423                 device = kzalloc(sizeof(*device), GFP_NOFS);
 424                 if (!device)
 425                         goto error;
 426
 427                 device->name = kstrdup(orig_dev->name, GFP_NOFS);
 428                 if (!device->name) {
 429                         kfree(device);
 430                         goto error;
 431                 }
 432
 433                 device->devid = orig_dev->devid;
 434                 device->work.func = pending_bios_fn;
 435                 memcpy(device->uuid, orig_dev->uuid, sizeof(device->uuid));
 436                 spin_lock_init(&device->io_lock);
 437                 INIT_LIST_HEAD(&device->dev_list);
 438                 INIT_LIST_HEAD(&device->dev_alloc_list);
 439
 440                 list_add(&device->dev_list, &fs_devices->devices);
 441                 device->fs_devices = fs_devices;
 442                 fs_devices->num_devices++;
 443         }
 444         return fs_devices;
 445 error:
 446         free_fs_devices(fs_devices);
 447         return ERR_PTR(-ENOMEM);
 448 }
 449
 450 int btrfs_close_extra_devices(struct btrfs_fs_devices *fs_devices)
 451 {
 452         struct btrfs_device *device, *next;
 453
 454         mutex_lock(&uuid_mutex);
 455 again:
 456         /* This is the initialized path, it is safe to release the devices. */
 457         list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
 458                 if (device->in_fs_metadata)
 459                         continue;
 460
 461                 if (device->bdev) {
 462                         blkdev_put(device->bdev, device->mode);
 463                         device->bdev = NULL;
 464                         fs_devices->open_devices--;
 465                 }
 466                 if (device->writeable) {
 467                         list_del_init(&device->dev_alloc_list);
 468                         device->writeable = 0;
 469                         fs_devices->rw_devices--;
 470                 }
 471                 list_del_init(&device->dev_list);
 472                 fs_devices->num_devices--;
 473                 kfree(device->name);
 474                 kfree(device);
 475         }
 476
 477         if (fs_devices->seed) {
 478                 fs_devices = fs_devices->seed;
 479                 goto again;
 480         }
 481
 482         mutex_unlock(&uuid_mutex);
 483         return 0;
 484 }
 485
 486 static void __free_device(struct work_struct *work)
 487 {
 488         struct btrfs_device *device;
 489
 490         device = container_of(work, struct btrfs_device, rcu_work);
 491
 492         if (device->bdev)
 493                 blkdev_put(device->bdev, device->mode);
 494
 495         kfree(device->name);
 496         kfree(device);
 497 }
 498
 499 static void free_device(struct rcu_head *head)
 500 {
 501         struct btrfs_device *device;
 502
 503         device = container_of(head, struct btrfs_device, rcu);
 504
 505         INIT_WORK(&device->rcu_work, __free_device);
 506         schedule_work(&device->rcu_work);
 507 }
 508
 509 static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 510 {
 511         struct btrfs_device *device;
 512
 513         if (--fs_devices->opened > 0)
 514                 return 0;
 515
 516         mutex_lock(&fs_devices->device_list_mutex);
 517         list_for_each_entry(device, &fs_devices->devices, dev_list) {
 518                 struct btrfs_device *new_device;
 519
 520                 if (device->bdev)
 521                         fs_devices->open_devices--;
 522
 523                 if (device->writeable) {
 524                         list_del_init(&device->dev_alloc_list);
 525                         fs_devices->rw_devices--;
 526                 }
 527
 528                 if (device->can_discard)
 529                         fs_devices->num_can_discard--;
 530
 531                 new_device = kmalloc(sizeof(*new_device), GFP_NOFS);
 532                 BUG_ON(!new_device);
 533                 memcpy(new_device, device, sizeof(*new_device));
 534                 new_device->name = kstrdup(device->name, GFP_NOFS);
 535                 BUG_ON(device->name && !new_device->name);
 536                 new_device->bdev = NULL;
 537                 new_device->writeable = 0;
 538                 new_device->in_fs_metadata = 0;
 539                 new_device->can_discard = 0;
 540                 list_replace_rcu(&device->dev_list, &new_device->dev_list);
 541
 542                 call_rcu(&device->rcu, free_device);
 543         }
 544         mutex_unlock(&fs_devices->device_list_mutex);
 545
 546         WARN_ON(fs_devices->open_devices);
 547         WARN_ON(fs_devices->rw_devices);
 548         fs_devices->opened = 0;
 549         fs_devices->seeding = 0;
 550
 551         return 0;
 552 }
 553
 554 int btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 555 {
 556         struct btrfs_fs_devices *seed_devices = NULL;
 557         int ret;
 558
 559         mutex_lock(&uuid_mutex);
 560         ret = __btrfs_close_devices(fs_devices);
 561         if (!fs_devices->opened) {
 562                 seed_devices = fs_devices->seed;
 563                 fs_devices->seed = NULL;
 564         }
 565         mutex_unlock(&uuid_mutex);
 566
 567         while (seed_devices) {
 568                 fs_devices = seed_devices;
 569                 seed_devices = fs_devices->seed;
 570                 __btrfs_close_devices(fs_devices);
 571                 free_fs_devices(fs_devices);
 572         }
 573         return ret;
 574 }
 575
 576 static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
 577                                 fmode_t flags, void *holder)
 578 {
 579         struct request_queue *q;
 580         struct block_device *bdev;
 581         struct list_head *head = &fs_devices->devices;
 582         struct btrfs_device *device;
 583         struct block_device *latest_bdev = NULL;
 584         struct buffer_head *bh;
 585         struct btrfs_super_block *disk_super;
 586         u64 latest_devid = 0;
 587         u64 latest_transid = 0;
 588         u64 devid;
 589         int seeding = 1;
 590         int ret = 0;
 591
 592         flags |= FMODE_EXCL;
 593
 594         list_for_each_entry(device, head, dev_list) {
 595                 if (device->bdev)
 596                         continue;
 597                 if (!device->name)
 598                         continue;
 599
 600                 bdev = blkdev_get_by_path(device->name, flags, holder);
 601                 if (IS_ERR(bdev)) {
 602                         printk(KERN_INFO "open %s failed\n", device->name);
 603                         goto error;
 604                 }
 605                 set_blocksize(bdev, 4096);
 606
 607                 bh = btrfs_read_dev_super(bdev);
 608                 if (!bh)
 609                         goto error_close;
 610
 611                 disk_super = (struct btrfs_super_block *)bh->b_data;
 612                 devid = btrfs_stack_device_id(&disk_super->dev_item);
 613                 if (devid != device->devid)
 614                         goto error_brelse;
 615
 616                 if (memcmp(device->uuid, disk_super->dev_item.uuid,
 617                            BTRFS_UUID_SIZE))
 618                         goto error_brelse;
 619
 620                 device->generation = btrfs_super_generation(disk_super);
 621                 if (!latest_transid || device->generation > latest_transid) {
 622                         latest_devid = devid;
 623                         latest_transid = device->generation;
 624                         latest_bdev = bdev;
 625                 }
 626
 627                 if (btrfs_super_flags(disk_super) & BTRFS_SUPER_FLAG_SEEDING) {
 628                         device->writeable = 0;
 629                 } else {
 630                         device->writeable = !bdev_read_only(bdev);
 631                         seeding = 0;
 632                 }
 633
 634                 q = bdev_get_queue(bdev);
 635                 if (blk_queue_discard(q)) {
 636                         device->can_discard = 1;
 637                         fs_devices->num_can_discard++;
 638                 }
 639
 640                 device->bdev = bdev;
 641                 device->in_fs_metadata = 0;
 642                 device->mode = flags;
 643
 644                 if (!blk_queue_nonrot(bdev_get_queue(bdev)))
 645                         fs_devices->rotating = 1;
 646
 647                 fs_devices->open_devices++;
 648                 if (device->writeable) {
 649                         fs_devices->rw_devices++;
 650                         list_add(&device->dev_alloc_list,
 651                                  &fs_devices->alloc_list);
 652                 }
 653                 brelse(bh);
 654                 continue;
 655
 656 error_brelse:
 657                 brelse(bh);
 658 error_close:
 659                 blkdev_put(bdev, flags);
 660 error:
 661                 continue;
 662         }
 663         if (fs_devices->open_devices == 0) {
 664                 ret = -EINVAL;
 665                 goto out;
 666         }
 667         fs_devices->seeding = seeding;
 668         fs_devices->opened = 1;
 669         fs_devices->latest_bdev = latest_bdev;
 670         fs_devices->latest_devid = latest_devid;
 671         fs_devices->latest_trans = latest_transid;
 672         fs_devices->total_rw_bytes = 0;
 673 out:
 674         return ret;
 675 }
 676
 677 int btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
 678                        fmode_t flags, void *holder)
 679 {
 680         int ret;
 681
 682         mutex_lock(&uuid_mutex);
 683         if (fs_devices->opened) {
 684                 fs_devices->opened++;
 685                 ret = 0;
 686         } else {
 687                 ret = __btrfs_open_devices(fs_devices, flags, holder);
 688         }
 689         mutex_unlock(&uuid_mutex);
 690         return ret;
 691 }
 692
 693 int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
 694                           struct btrfs_fs_devices **fs_devices_ret)
 695 {
 696         struct btrfs_super_block *disk_super;
 697         struct block_device *bdev;
 698         struct buffer_head *bh;
 699         int ret;
 700         u64 devid;
 701         u64 transid;
 702
 703         mutex_lock(&uuid_mutex);
 704
 705         flags |= FMODE_EXCL;
 706         bdev = blkdev_get_by_path(path, flags, holder);
 707
 708         if (IS_ERR(bdev)) {
 709                 ret = PTR_ERR(bdev);
 710                 goto error;
 711         }
 712
 713         ret = set_blocksize(bdev, 4096);
 714         if (ret)
 715                 goto error_close;
 716         bh = btrfs_read_dev_super(bdev);
 717         if (!bh) {
 718                 ret = -EINVAL;
 719                 goto error_close;
 720         }
 721         disk_super = (struct btrfs_super_block *)bh->b_data;
 722         devid = btrfs_stack_device_id(&disk_super->dev_item);
 723         transid = btrfs_super_generation(disk_super);
 724         if (disk_super->label[0])
 725                 printk(KERN_INFO "device label %s ", disk_super->label);
 726         else
 727                 printk(KERN_INFO "device fsid %pU ", disk_super->fsid);
 728         printk(KERN_CONT "devid %llu transid %llu %s\n",
 729                (unsigned long long)devid, (unsigned long long)transid, path);
 730         ret = device_list_add(path, disk_super, devid, fs_devices_ret);
 731
 732         brelse(bh);
 733 error_close:
 734         blkdev_put(bdev, flags);
 735 error:
 736         mutex_unlock(&uuid_mutex);
 737         return ret;
 738 }
 739
 740 /* helper to account the used device space in the range */
 741 int btrfs_account_dev_extents_size(struct btrfs_device *device, u64 start,
 742                                    u64 end, u64 *length)
 743 {
 744         struct btrfs_key key;
 745         struct btrfs_root *root = device->dev_root;
 746         struct btrfs_dev_extent *dev_extent;
 747         struct btrfs_path *path;
 748         u64 extent_end;
 749         int ret;
 750         int slot;
 751         struct extent_buffer *l;
 752
 753         *length = 0;
 754
 755         if (start >= device->total_bytes)
 756                 return 0;
 757
 758         path = btrfs_alloc_path();
 759         if (!path)
 760                 return -ENOMEM;
 761         path->reada = 2;
 762
 763         key.objectid = device->devid;
 764         key.offset = start;
 765         key.type = BTRFS_DEV_EXTENT_KEY;
 766
 767         ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
 768         if (ret < 0)
 769                 goto out;
 770         if (ret > 0) {
 771                 ret = btrfs_previous_item(root, path, key.objectid, key.type);
 772                 if (ret < 0)
 773                         goto out;
 774         }
 775
 776         while (1) {
 777                 l = path->nodes[0];
 778                 slot = path->slots[0];
 779                 if (slot >= btrfs_header_nritems(l)) {
 780                         ret = btrfs_next_leaf(root, path);
 781                         if (ret == 0)
 782                                 continue;
 783                         if (ret < 0)
 784                                 goto out;
 785
 786                         break;
 787                 }
 788                 btrfs_item_key_to_cpu(l, &key, slot);
 789
 790                 if (key.objectid < device->devid)
 791                         goto next;
 792
 793                 if (key.objectid > device->devid)
 794                         break;
 795
 796                 if (btrfs_key_type(&key) != BTRFS_DEV_EXTENT_KEY)
 797                         goto next;
 798
 799                 dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
 800                 extent_end = key.offset + btrfs_dev_extent_length(l,
 801                                                                   dev_extent);
 802                 if (key.offset <= start && extent_end > end) {
 803                         *length = end - start + 1;
 804                         break;
 805                 } else if (key.offset <= start && extent_end > start)
 806                         *length += extent_end - start;
 807                 else if (key.offset > start && extent_end <= end)
 808                         *length += extent_end - key.offset;
 809                 else if (key.offset > start && key.offset <= end) {
 810                         *length += end - key.offset + 1;
 811                         break;
 812                 } else if (key.offset > end)
 813                         break;
 814
 815 next:
 816                 path->slots[0]++;
 817         }
 818         ret = 0;
 819 out:
 820         btrfs_free_path(path);
 821         return ret;
 822 }
 823
 824 /*
 825  * find_free_dev_extent - find free space in the specified device
 826  * @trans:      transaction handler
 827  * @device:     the device which we search the free space in
 828  * @num_bytes:  the size of the free space that we need
 829  * @start:      store the start of the free space.
 830  * @len:        the size of the free space. that we find, or the size of the max
 831  *              free space if we don't find suitable free space
 832  *
 833  * this uses a pretty simple search, the expectation is that it is
 834  * called very infrequently and that a given device has a small number
 835  * of extents
 836  *
 837  * @start is used to store the start of the free space if we find. But if we
 838  * don't find suitable free space, it will be used to store the start position
 839  * of the max free space.
 840  *
 841  * @len is used to store the size of the free space that we find.
 842  * But if we don't find suitable free space, it is used to store the size of
 843  * the max free space.
 844  */
 845 int find_free_dev_extent(struct btrfs_trans_handle *trans,
 846                          struct btrfs_device *device, u64 num_bytes,
 847                          u64 *start, u64 *len)
 848 {
 849         struct btrfs_key key;
 850         struct btrfs_root *root = device->dev_root;
 851         struct btrfs_dev_extent *dev_extent;
 852         struct btrfs_path *path;
 853         u64 hole_size;
 854         u64 max_hole_start;
 855         u64 max_hole_size;
 856         u64 extent_end;
 857         u64 search_start;
 858         u64 search_end = device->total_bytes;
 859         int ret;
 860         int slot;
 861         struct extent_buffer *l;
 862
 863         /* FIXME use last free of some kind */
 864
 865         /* we don't want to overwrite the superblock on the drive,
 866          * so we make sure to start at an offset of at least 1MB
 867          */
 868         search_start = max(root->fs_info->alloc_start, 1024ull * 1024);
 869
 870         max_hole_start = search_start;
 871         max_hole_size = 0;
 872         hole_size = 0;
 873
 874         if (search_start >= search_end) {
 875                 ret = -ENOSPC;
 876                 goto error;
 877         }
 878
 879         path = btrfs_alloc_path();
 880         if (!path) {
 881                 ret = -ENOMEM;
 882                 goto error;
 883         }
 884         path->reada = 2;
 885
 886         key.objectid = device->devid;
 887         key.offset = search_start;
 888         key.type = BTRFS_DEV_EXTENT_KEY;
 889
 890         ret = btrfs_search_slot(trans, root, &key, path, 0, 0);
 891         if (ret < 0)
 892                 goto out;
 893         if (ret > 0) {
 894                 ret = btrfs_previous_item(root, path, key.objectid, key.type);
 895                 if (ret < 0)
 896                         goto out;
 897         }
 898
 899         while (1) {
 900                 l = path->nodes[0];
 901                 slot = path->slots[0];
 902                 if (slot >= btrfs_header_nritems(l)) {
 903                         ret = btrfs_next_leaf(root, path);
 904                         if (ret == 0)
 905                                 continue;
 906                         if (ret < 0)
 907                                 goto out;
 908
 909                         break;
 910                 }
 911                 btrfs_item_key_to_cpu(l, &key, slot);
 912
 913                 if (key.objectid < device->devid)
 914                         goto next;
 915
 916                 if (key.objectid > device->devid)
 917                         break;
 918
 919                 if (btrfs_key_type(&key) != BTRFS_DEV_EXTENT_KEY)
 920                         goto next;
 921
 922                 if (key.offset > search_start) {
 923                         hole_size = key.offset - search_start;
 924
 925                         if (hole_size > max_hole_size) {
 926                                 max_hole_start = search_start;
 927                                 max_hole_size = hole_size;
 928                         }
 929
 930                         /*
 931                          * If this free space is greater than which we need,
 932                          * it must be the max free space that we have found
 933                          * until now, so max_hole_start must point to the start
 934                          * of this free space and the length of this free space
 935                          * is stored in max_hole_size. Thus, we return
 936                          * max_hole_start and max_hole_size and go back to the
 937                          * caller.
 938                          */
 939                         if (hole_size >= num_bytes) {
 940                                 ret = 0;
 941                                 goto out;
 942                         }
 943                 }
 944
 945                 dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
 946                 extent_end = key.offset + btrfs_dev_extent_length(l,
 947                                                                   dev_extent);
 948                 if (extent_end > search_start)
 949                         search_start = extent_end;
 950 next:
 951                 path->slots[0]++;
 952                 cond_resched();
 953         }
 954
 955         /*
 956          * At this point, search_start should be the end of
 957          * allocated dev extents, and when shrinking the device,
 958          * search_end may be smaller than search_start.
 959          */
 960         if (search_end > search_start)
 961                 hole_size = search_end - search_start;
 962
 963         if (hole_size > max_hole_size) {
 964                 max_hole_start = search_start;
 965                 max_hole_size = hole_size;
 966         }
 967
 968         /* See above. */
 969         if (hole_size < num_bytes)
 970                 ret = -ENOSPC;
 971         else
 972                 ret = 0;
 973
 974 out:
 975         btrfs_free_path(path);
 976 error:
 977         *start = max_hole_start;
 978         if (len)
 979                 *len = max_hole_size;
 980         return ret;
 981 }
 982
 983 static int btrfs_free_dev_extent(struct btrfs_trans_handle *trans,
 984                           struct btrfs_device *device,
 985                           u64 start)
 986 {
 987         int ret;
 988         struct btrfs_path *path;
 989         struct btrfs_root *root = device->dev_root;
 990         struct btrfs_key key;
 991         struct btrfs_key found_key;
 992         struct extent_buffer *leaf = NULL;
 993         struct btrfs_dev_extent *extent = NULL;
 994
 995         path = btrfs_alloc_path();
 996         if (!path)
 997                 return -ENOMEM;
 998
 999         key.objectid = device->devid;
1000         key.offset = start;
1001         key.type = BTRFS_DEV_EXTENT_KEY;
1002 again:
1003         ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
1004         if (ret > 0) {
1005                 ret = btrfs_previous_item(root, path, key.objectid,
1006                                           BTRFS_DEV_EXTENT_KEY);
1007                 if (ret)
1008                         goto out;
1009                 leaf = path->nodes[0];
1010                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
1011                 extent = btrfs_item_ptr(leaf, path->slots[0],
1012                                         struct btrfs_dev_extent);
1013                 BUG_ON(found_key.offset > start || found_key.offset +
1014                        btrfs_dev_extent_length(leaf, extent) < start);
1015                 key = found_key;
1016                 btrfs_release_path(path);
1017                 goto again;
1018         } else if (ret == 0) {
1019                 leaf = path->nodes[0];
1020                 extent = btrfs_item_ptr(leaf, path->slots[0],
1021                                         struct btrfs_dev_extent);
1022         }
1023         BUG_ON(ret);
1024
1025         if (device->bytes_used > 0) {
1026                 u64 len = btrfs_dev_extent_length(leaf, extent);
1027                 device->bytes_used -= len;
1028                 spin_lock(&root->fs_info->free_chunk_lock);
1029                 root->fs_info->free_chunk_space += len;
1030                 spin_unlock(&root->fs_info->free_chunk_lock);
1031         }
1032         ret = btrfs_del_item(trans, root, path);
1033
1034 out:
1035         btrfs_free_path(path);
1036         return ret;
1037 }
1038
1039 int btrfs_alloc_dev_extent(struct btrfs_trans_handle *trans,
1040                            struct btrfs_device *device,
1041                            u64 chunk_tree, u64 chunk_objectid,
1042                            u64 chunk_offset, u64 start, u64 num_bytes)
1043 {
1044         int ret;
1045         struct btrfs_path *path;
1046         struct btrfs_root *root = device->dev_root;
1047         struct btrfs_dev_extent *extent;
1048         struct extent_buffer *leaf;
1049         struct btrfs_key key;
1050
1051         WARN_ON(!device->in_fs_metadata);
1052         path = btrfs_alloc_path();
1053         if (!path)
1054                 return -ENOMEM;
1055
1056         key.objectid = device->devid;
1057         key.offset = start;
1058         key.type = BTRFS_DEV_EXTENT_KEY;
1059         ret = btrfs_insert_empty_item(trans, root, path, &key,
1060                                       sizeof(*extent));
1061         BUG_ON(ret);
1062
1063         leaf = path->nodes[0];
1064         extent = btrfs_item_ptr(leaf, path->slots[0],
1065                                 struct btrfs_dev_extent);
1066         btrfs_set_dev_extent_chunk_tree(leaf, extent, chunk_tree);
1067         btrfs_set_dev_extent_chunk_objectid(leaf, extent, chunk_objectid);
1068         btrfs_set_dev_extent_chunk_offset(leaf, extent, chunk_offset);
1069
1070         write_extent_buffer(leaf, root->fs_info->chunk_tree_uuid,
1071                     (unsigned long)btrfs_dev_extent_chunk_tree_uuid(extent),
1072                     BTRFS_UUID_SIZE);
1073
1074         btrfs_set_dev_extent_length(leaf, extent, num_bytes);
1075         btrfs_mark_buffer_dirty(leaf);
1076         btrfs_free_path(path);
1077         return ret;
1078 }
1079
1080 static noinline int find_next_chunk(struct btrfs_root *root,
1081                                     u64 objectid, u64 *offset)
1082 {
1083         struct btrfs_path *path;
1084         int ret;
1085         struct btrfs_key key;
1086         struct btrfs_chunk *chunk;
1087         struct btrfs_key found_key;
1088
1089         path = btrfs_alloc_path();
1090         if (!path)
1091                 return -ENOMEM;
1092
1093         key.objectid = objectid;
1094         key.offset = (u64)-1;
1095         key.type = BTRFS_CHUNK_ITEM_KEY;
1096
1097         ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
1098         if (ret < 0)
1099                 goto error;
1100
1101         BUG_ON(ret == 0);
1102
1103         ret = btrfs_previous_item(root, path, 0, BTRFS_CHUNK_ITEM_KEY);
1104         if (ret) {
1105                 *offset = 0;
1106         } else {
1107                 btrfs_item_key_to_cpu(path->nodes[0], &found_key,
1108                                       path->slots[0]);
1109                 if (found_key.objectid != objectid)
1110                         *offset = 0;
1111                 else {
1112                         chunk = btrfs_item_ptr(path->nodes[0], path->slots[0],
1113                                                struct btrfs_chunk);
1114                         *offset = found_key.offset +
1115                                 btrfs_chunk_length(path->nodes[0], chunk);
1116                 }
1117         }
1118         ret = 0;
1119 error:
1120         btrfs_free_path(path);
1121         return ret;
1122 }
1123
1124 static noinline int find_next_devid(struct btrfs_root *root, u64 *objectid)
1125 {
1126         int ret;
1127         struct btrfs_key key;
1128         struct btrfs_key found_key;
1129         struct btrfs_path *path;
1130
1131         root = root->fs_info->chunk_root;
1132
1133         path = btrfs_alloc_path();
1134         if (!path)
1135                 return -ENOMEM;
1136
1137         key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
1138         key.type = BTRFS_DEV_ITEM_KEY;
1139         key.offset = (u64)-1;
1140
1141         ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
1142         if (ret < 0)
1143                 goto error;
1144
1145         BUG_ON(ret == 0);
1146
1147         ret = btrfs_previous_item(root, path, BTRFS_DEV_ITEMS_OBJECTID,
1148                                   BTRFS_DEV_ITEM_KEY);
1149         if (ret) {
1150                 *objectid = 1;
1151         } else {
1152                 btrfs_item_key_to_cpu(path->nodes[0], &found_key,
1153                                       path->slots[0]);
1154                 *objectid = found_key.offset + 1;
1155         }
1156         ret = 0;
1157 error:
1158         btrfs_free_path(path);
1159         return ret;
1160 }
1161
1162 /*
1163  * the device information is stored in the chunk root
1164  * the btrfs_device struct should be fully filled in
1165  */
1166 int btrfs_add_device(struct btrfs_trans_handle *trans,
1167                      struct btrfs_root *root,
1168                      struct btrfs_device *device)
1169 {
1170         int ret;
1171         struct btrfs_path *path;
1172         struct btrfs_dev_item *dev_item;
1173         struct extent_buffer *leaf;
1174         struct btrfs_key key;
1175         unsigned long ptr;
1176
1177         root = root->fs_info->chunk_root;
1178
1179         path = btrfs_alloc_path();
1180         if (!path)
1181                 return -ENOMEM;
1182
1183         key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
1184         key.type = BTRFS_DEV_ITEM_KEY;
1185         key.offset = device->devid;
1186
1187         ret = btrfs_insert_empty_item(trans, root, path, &key,
1188                                       sizeof(*dev_item));
1189         if (ret)
1190                 goto out;
1191
1192         leaf = path->nodes[0];
1193         dev_item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_dev_item);
1194
1195         btrfs_set_device_id(leaf, dev_item, device->devid);
1196         btrfs_set_device_generation(leaf, dev_item, 0);
1197         btrfs_set_device_type(leaf, dev_item, device->type);
1198         btrfs_set_device_io_align(leaf, dev_item, device->io_align);
1199         btrfs_set_device_io_width(leaf, dev_item, device->io_width);
1200         btrfs_set_device_sector_size(leaf, dev_item, device->sector_size);
1201         btrfs_set_device_total_bytes(leaf, dev_item, device->total_bytes);
1202         btrfs_set_device_bytes_used(leaf, dev_item, device->bytes_used);
1203         btrfs_set_device_group(leaf, dev_item, 0);
1204         btrfs_set_device_seek_speed(leaf, dev_item, 0);
1205         btrfs_set_device_bandwidth(leaf, dev_item, 0);
1206         btrfs_set_device_start_offset(leaf, dev_item, 0);
1207
1208         ptr = (unsigned long)btrfs_device_uuid(dev_item);
1209         write_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
1210         ptr = (unsigned long)btrfs_device_fsid(dev_item);
1211         write_extent_buffer(leaf, root->fs_info->fsid, ptr, BTRFS_UUID_SIZE);
1212         btrfs_mark_buffer_dirty(leaf);
1213
1214         ret = 0;
1215 out:
1216         btrfs_free_path(path);
1217         return ret;
1218 }
1219
1220 static int btrfs_rm_dev_item(struct btrfs_root *root,
1221                              struct btrfs_device *device)
1222 {
1223         int ret;
1224         struct btrfs_path *path;
1225         struct btrfs_key key;
1226         struct btrfs_trans_handle *trans;
1227
1228         root = root->fs_info->chunk_root;
1229
1230         path = btrfs_alloc_path();
1231         if (!path)
1232                 return -ENOMEM;
1233
1234         trans = btrfs_start_transaction(root, 0);
1235         if (IS_ERR(trans)) {
1236                 btrfs_free_path(path);
1237                 return PTR_ERR(trans);
1238         }
1239         key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
1240         key.type = BTRFS_DEV_ITEM_KEY;
1241         key.offset = device->devid;
1242         lock_chunks(root);
1243
1244         ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
1245         if (ret < 0)
1246                 goto out;
1247
1248         if (ret > 0) {
1249                 ret = -ENOENT;
1250                 goto out;
1251         }
1252
1253         ret = btrfs_del_item(trans, root, path);
1254         if (ret)
1255                 goto out;
1256 out:
1257         btrfs_free_path(path);
1258         unlock_chunks(root);
1259         btrfs_commit_transaction(trans, root);
1260         return ret;
1261 }
1262
1263 int btrfs_rm_device(struct btrfs_root *root, char *device_path)
1264 {
1265         struct btrfs_device *device;
1266         struct btrfs_device *next_device;
1267         struct block_device *bdev;
1268         struct buffer_head *bh = NULL;
1269         struct btrfs_super_block *disk_super;
1270         struct btrfs_fs_devices *cur_devices;
1271         u64 all_avail;
1272         u64 devid;
1273         u64 num_devices;
1274         u8 *dev_uuid;
1275         int ret = 0;
1276         bool clear_super = false;
1277
1278         mutex_lock(&uuid_mutex);
1279         mutex_lock(&root->fs_info->volume_mutex);
1280
1281         all_avail = root->fs_info->avail_data_alloc_bits |
1282                 root->fs_info->avail_system_alloc_bits |
1283                 root->fs_info->avail_metadata_alloc_bits;
1284
1285         if ((all_avail & BTRFS_BLOCK_GROUP_RAID10) &&
1286             root->fs_info->fs_devices->num_devices <= 4) {
1287                 printk(KERN_ERR "btrfs: unable to go below four devices "
1288                        "on raid10\n");
1289                 ret = -EINVAL;
1290                 goto out;
1291         }
1292
1293         if ((all_avail & BTRFS_BLOCK_GROUP_RAID1) &&
1294             root->fs_info->fs_devices->num_devices <= 2) {
1295                 printk(KERN_ERR "btrfs: unable to go below two "
1296                        "devices on raid1\n");
1297                 ret = -EINVAL;
1298                 goto out;
1299         }
1300
1301         if (strcmp(device_path, "missing") == 0) {
1302                 struct list_head *devices;
1303                 struct btrfs_device *tmp;
1304
1305                 device = NULL;
1306                 devices = &root->fs_info->fs_devices->devices;
1307                 /*
1308                  * It is safe to read the devices since the volume_mutex
1309                  * is held.
1310                  */
1311                 list_for_each_entry(tmp, devices, dev_list) {
1312                         if (tmp->in_fs_metadata && !tmp->bdev) {
1313                                 device = tmp;
1314                                 break;
1315                         }
1316                 }
1317                 bdev = NULL;
1318                 bh = NULL;
1319                 disk_super = NULL;
1320                 if (!device) {
1321                         printk(KERN_ERR "btrfs: no missing devices found to "
1322                                "remove\n");
1323                         goto out;
1324                 }
1325         } else {
1326                 bdev = blkdev_get_by_path(device_path, FMODE_READ | FMODE_EXCL,
1327                                           root->fs_info->bdev_holder);
1328                 if (IS_ERR(bdev)) {
1329                         ret = PTR_ERR(bdev);
1330                         goto out;
1331                 }
1332
1333                 set_blocksize(bdev, 4096);
1334                 bh = btrfs_read_dev_super(bdev);
1335                 if (!bh) {
1336                         ret = -EINVAL;
1337                         goto error_close;
1338                 }
1339                 disk_super = (struct btrfs_super_block *)bh->b_data;
1340                 devid = btrfs_stack_device_id(&disk_super->dev_item);
1341                 dev_uuid = disk_super->dev_item.uuid;
1342                 device = btrfs_find_device(root, devid, dev_uuid,
1343                                            disk_super->fsid);
1344                 if (!device) {
1345                         ret = -ENOENT;
1346                         goto error_brelse;
1347                 }
1348         }
1349
1350         if (device->writeable && root->fs_info->fs_devices->rw_devices == 1) {
1351                 printk(KERN_ERR "btrfs: unable to remove the only writeable "
1352                        "device\n");
1353                 ret = -EINVAL;
1354                 goto error_brelse;
1355         }
1356
1357         if (device->writeable) {
1358                 lock_chunks(root);
1359                 list_del_init(&device->dev_alloc_list);
1360                 unlock_chunks(root);
1361                 root->fs_info->fs_devices->rw_devices--;
1362                 clear_super = true;
1363         }
1364
1365         ret = btrfs_shrink_device(device, 0);
1366         if (ret)
1367                 goto error_undo;
1368
1369         ret = btrfs_rm_dev_item(root->fs_info->chunk_root, device);
1370         if (ret)
1371                 goto error_undo;
1372
1373         spin_lock(&root->fs_info->free_chunk_lock);
1374         root->fs_info->free_chunk_space = device->total_bytes -
1375                 device->bytes_used;
1376         spin_unlock(&root->fs_info->free_chunk_lock);
1377
1378         device->in_fs_metadata = 0;
1379         btrfs_scrub_cancel_dev(root, device);
1380
1381         /*
1382          * the device list mutex makes sure that we don't change
1383          * the device list while someone else is writing out all
1384          * the device supers.
1385          */
1386
1387         cur_devices = device->fs_devices;
1388         mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
1389         list_del_rcu(&device->dev_list);
1390
1391         device->fs_devices->num_devices--;
1392
1393         if (device->missing)
1394                 root->fs_info->fs_devices->missing_devices--;
1395
1396         next_device = list_entry(root->fs_info->fs_devices->devices.next,
1397                                  struct btrfs_device, dev_list);
1398         if (device->bdev == root->fs_info->sb->s_bdev)
1399                 root->fs_info->sb->s_bdev = next_device->bdev;
1400         if (device->bdev == root->fs_info->fs_devices->latest_bdev)
1401                 root->fs_info->fs_devices->latest_bdev = next_device->bdev;
1402
1403         if (device->bdev)
1404                 device->fs_devices->open_devices--;
1405
1406         call_rcu(&device->rcu, free_device);
1407         mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
1408
1409         num_devices = btrfs_super_num_devices(root->fs_info->super_copy) - 1;
1410         btrfs_set_super_num_devices(root->fs_info->super_copy, num_devices);
1411
1412         if (cur_devices->open_devices == 0) {
1413                 struct btrfs_fs_devices *fs_devices;
1414                 fs_devices = root->fs_info->fs_devices;
1415                 while (fs_devices) {
1416                         if (fs_devices->seed == cur_devices)
1417                                 break;
1418                         fs_devices = fs_devices->seed;
1419                 }
1420                 fs_devices->seed = cur_devices->seed;
1421                 cur_devices->seed = NULL;
1422                 lock_chunks(root);
1423                 __btrfs_close_devices(cur_devices);
1424                 unlock_chunks(root);
1425                 free_fs_devices(cur_devices);
1426         }
1427
1428         /*
1429          * at this point, the device is zero sized.  We want to
1430          * remove it from the devices list and zero out the old super
1431          */
1432         if (clear_super) {
1433                 /* make sure this device isn't detected as part of
1434                  * the FS anymore
1435                  */
1436                 memset(&disk_super->magic, 0, sizeof(disk_super->magic));
1437                 set_buffer_dirty(bh);
1438                 sync_dirty_buffer(bh);
1439         }
1440
1441         ret = 0;
1442
1443 error_brelse:
1444         brelse(bh);
1445 error_close:
1446         if (bdev)
1447                 blkdev_put(bdev, FMODE_READ | FMODE_EXCL);
1448 out:
1449         mutex_unlock(&root->fs_info->volume_mutex);
1450         mutex_unlock(&uuid_mutex);
1451         return ret;
1452 error_undo:
1453         if (device->writeable) {
1454                 lock_chunks(root);
1455                 list_add(&device->dev_alloc_list,
1456                          &root->fs_info->fs_devices->alloc_list);
1457                 unlock_chunks(root);
1458                 root->fs_info->fs_devices->rw_devices++;
1459         }
1460         goto error_brelse;
1461 }
1462
1463 /*
1464  * does all the dirty work required for changing file system's UUID.
1465  */
1466 static int btrfs_prepare_sprout(struct btrfs_trans_handle *trans,
1467                                 struct btrfs_root *root)
1468 {
1469         struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
1470         struct btrfs_fs_devices *old_devices;
1471         struct btrfs_fs_devices *seed_devices;
1472         struct btrfs_super_block *disk_super = root->fs_info->super_copy;
1473         struct btrfs_device *device;
1474         u64 super_flags;
1475
1476         BUG_ON(!mutex_is_locked(&uuid_mutex));
1477         if (!fs_devices->seeding)
1478                 return -EINVAL;
1479
1480         seed_devices = kzalloc(sizeof(*fs_devices), GFP_NOFS);
1481         if (!seed_devices)
1482                 return -ENOMEM;
1483
1484         old_devices = clone_fs_devices(fs_devices);
1485         if (IS_ERR(old_devices)) {
1486                 kfree(seed_devices);
1487                 return PTR_ERR(old_devices);
1488         }
1489
1490         list_add(&old_devices->list, &fs_uuids);
1491
1492         memcpy(seed_devices, fs_devices, sizeof(*seed_devices));
1493         seed_devices->opened = 1;
1494         INIT_LIST_HEAD(&seed_devices->devices);
1495         INIT_LIST_HEAD(&seed_devices->alloc_list);
1496         mutex_init(&seed_devices->device_list_mutex);
1497
1498         mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
1499         list_splice_init_rcu(&fs_devices->devices, &seed_devices->devices,
1500                               synchronize_rcu);
1501         mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
1502
1503         list_splice_init(&fs_devices->alloc_list, &seed_devices->alloc_list);
1504         list_for_each_entry(device, &seed_devices->devices, dev_list) {
1505                 device->fs_devices = seed_devices;
1506         }
1507
1508         fs_devices->seeding = 0;
1509         fs_devices->num_devices = 0;
1510         fs_devices->open_devices = 0;
1511         fs_devices->seed = seed_devices;
1512
1513         generate_random_uuid(fs_devices->fsid);
1514         memcpy(root->fs_info->fsid, fs_devices->fsid, BTRFS_FSID_SIZE);
1515         memcpy(disk_super->fsid, fs_devices->fsid, BTRFS_FSID_SIZE);
1516         super_flags = btrfs_super_flags(disk_super) &
1517                       ~BTRFS_SUPER_FLAG_SEEDING;
1518         btrfs_set_super_flags(disk_super, super_flags);
1519
1520         return 0;
1521 }
1522
1523 /*
1524  * strore the expected generation for seed devices in device items.
1525  */
1526 static int btrfs_finish_sprout(struct btrfs_trans_handle *trans,
1527                                struct btrfs_root *root)
1528 {
1529         struct btrfs_path *path;
1530         struct extent_buffer *leaf;
1531         struct btrfs_dev_item *dev_item;
1532         struct btrfs_device *device;
1533         struct btrfs_key key;
1534         u8 fs_uuid[BTRFS_UUID_SIZE];
1535         u8 dev_uuid[BTRFS_UUID_SIZE];
1536         u64 devid;
1537         int ret;
1538
1539         path = btrfs_alloc_path();
1540         if (!path)
1541                 return -ENOMEM;
1542
1543         root = root->fs_info->chunk_root;
1544         key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
1545         key.offset = 0;
1546         key.type = BTRFS_DEV_ITEM_KEY;
1547
1548         while (1) {
1549                 ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
1550                 if (ret < 0)
1551                         goto error;
1552
1553                 leaf = path->nodes[0];
1554 next_slot:
1555                 if (path->slots[0] >= btrfs_header_nritems(leaf)) {
1556                         ret = btrfs_next_leaf(root, path);
1557                         if (ret > 0)
1558                                 break;
1559                         if (ret < 0)
1560                                 goto error;
1561                         leaf = path->nodes[0];
1562                         btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
1563                         btrfs_release_path(path);
1564                         continue;
1565                 }
1566
1567                 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
1568                 if (key.objectid != BTRFS_DEV_ITEMS_OBJECTID ||
1569                     key.type != BTRFS_DEV_ITEM_KEY)
1570                         break;
1571
1572                 dev_item = btrfs_item_ptr(leaf, path->slots[0],
1573                                           struct btrfs_dev_item);
1574                 devid = btrfs_device_id(leaf, dev_item);
1575                 read_extent_buffer(leaf, dev_uuid,
1576                                    (unsigned long)btrfs_device_uuid(dev_item),
1577                                    BTRFS_UUID_SIZE);
1578                 read_extent_buffer(leaf, fs_uuid,
1579                                    (unsigned long)btrfs_device_fsid(dev_item),
1580                                    BTRFS_UUID_SIZE);
1581                 device = btrfs_find_device(root, devid, dev_uuid, fs_uuid);
1582                 BUG_ON(!device);
1583
1584                 if (device->fs_devices->seeding) {
1585                         btrfs_set_device_generation(leaf, dev_item,
1586                                                     device->generation);
1587                         btrfs_mark_buffer_dirty(leaf);
1588                 }
1589
1590                 path->slots[0]++;
1591                 goto next_slot;
1592         }
1593         ret = 0;
1594 error:
1595         btrfs_free_path(path);
1596         return ret;
1597 }
1598
1599 int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
1600 {
1601         struct request_queue *q;
1602         struct btrfs_trans_handle *trans;
1603         struct btrfs_device *device;
1604         struct block_device *bdev;
1605         struct list_head *devices;
1606         struct super_block *sb = root->fs_info->sb;
1607         u64 total_bytes;
1608         int seeding_dev = 0;
1609         int ret = 0;
1610
1611         if ((sb->s_flags & MS_RDONLY) && !root->fs_info->fs_devices->seeding)
1612                 return -EINVAL;
1613
1614         bdev = blkdev_get_by_path(device_path, FMODE_EXCL,
1615                                   root->fs_info->bdev_holder);
1616         if (IS_ERR(bdev))
1617                 return PTR_ERR(bdev);
1618
1619         if (root->fs_info->fs_devices->seeding) {
1620                 seeding_dev = 1;
1621                 down_write(&sb->s_umount);
1622                 mutex_lock(&uuid_mutex);
1623         }
1624
1625         filemap_write_and_wait(bdev->bd_inode->i_mapping);
1626         mutex_lock(&root->fs_info->volume_mutex);
1627
1628         devices = &root->fs_info->fs_devices->devices;
1629         /*
1630          * we have the volume lock, so we don't need the extra
1631          * device list mutex while reading the list here.
1632          */
1633         list_for_each_entry(device, devices, dev_list) {
1634                 if (device->bdev == bdev) {
1635                         ret = -EEXIST;
1636                         goto error;
1637                 }
1638         }
1639
1640         device = kzalloc(sizeof(*device), GFP_NOFS);
1641         if (!device) {
1642                 /* we can safely leave the fs_devices entry around */
1643                 ret = -ENOMEM;
1644                 goto error;
1645         }
1646
1647         device->name = kstrdup(device_path, GFP_NOFS);
1648         if (!device->name) {
1649                 kfree(device);
1650                 ret = -ENOMEM;
1651                 goto error;
1652         }
1653
1654         ret = find_next_devid(root, &device->devid);
1655         if (ret) {
1656                 kfree(device->name);
1657                 kfree(device);
1658                 goto error;
1659         }
1660
1661         trans = btrfs_start_transaction(root, 0);
1662         if (IS_ERR(trans)) {
1663                 kfree(device->name);
1664                 kfree(device);
1665                 ret = PTR_ERR(trans);
1666                 goto error;
1667         }
1668
1669         lock_chunks(root);
1670
1671         q = bdev_get_queue(bdev);
1672         if (blk_queue_discard(q))
1673                 device->can_discard = 1;
1674         device->writeable = 1;
1675         device->work.func = pending_bios_fn;
1676         generate_random_uuid(device->uuid);
1677         spin_lock_init(&device->io_lock);
1678         device->generation = trans->transid;
1679         device->io_width = root->sectorsize;
1680         device->io_align = root->sectorsize;
1681         device->sector_size = root->sectorsize;
1682         device->total_bytes = i_size_read(bdev->bd_inode);
1683         device->disk_total_bytes = device->total_bytes;
1684         device->dev_root = root->fs_info->dev_root;
1685         device->bdev = bdev;
1686         device->in_fs_metadata = 1;
1687         device->mode = FMODE_EXCL;
1688         set_blocksize(device->bdev, 4096);
1689
1690         if (seeding_dev) {
1691                 sb->s_flags &= ~MS_RDONLY;
1692                 ret = btrfs_prepare_sprout(trans, root);
1693                 BUG_ON(ret);
1694         }
1695
1696         device->fs_devices = root->fs_info->fs_devices;
1697
1698         /*
1699          * we don't want write_supers to jump in here with our device
1700          * half setup
1701          */
1702         mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
1703         list_add_rcu(&device->dev_list, &root->fs_info->fs_devices->devices);
1704         list_add(&device->dev_alloc_list,
1705                  &root->fs_info->fs_devices->alloc_list);
1706         root->fs_info->fs_devices->num_devices++;
1707         root->fs_info->fs_devices->open_devices++;
1708         root->fs_info->fs_devices->rw_devices++;
1709         if (device->can_discard)
1710                 root->fs_info->fs_devices->num_can_discard++;
1711         root->fs_info->fs_devices->total_rw_bytes += device->total_bytes;
1712
1713         spin_lock(&root->fs_info->free_chunk_lock);
1714         root->fs_info->free_chunk_space += device->total_bytes;
1715         spin_unlock(&root->fs_info->free_chunk_lock);
1716
1717         if (!blk_queue_nonrot(bdev_get_queue(bdev)))
1718                 root->fs_info->fs_devices->rotating = 1;
1719
1720         total_bytes = btrfs_super_total_bytes(root->fs_info->super_copy);
1721         btrfs_set_super_total_bytes(root->fs_info->super_copy,
1722                                     total_bytes + device->total_bytes);
1723
1724         total_bytes = btrfs_super_num_devices(root->fs_info->super_copy);
1725         btrfs_set_super_num_devices(root->fs_info->super_copy,
1726                                     total_bytes + 1);
1727         mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
1728
1729         if (seeding_dev) {
1730                 ret = init_first_rw_device(trans, root, device);
1731                 BUG_ON(ret);
1732                 ret = btrfs_finish_sprout(trans, root);
1733                 BUG_ON(ret);
1734         } else {
1735                 ret = btrfs_add_device(trans, root, device);
1736         }
1737
1738         /*
1739          * we've got more storage, clear any full flags on the space
1740          * infos
1741          */
1742         btrfs_clear_space_info_full(root->fs_info);
1743
1744         unlock_chunks(root);
1745         btrfs_commit_transaction(trans, root);
1746
1747         if (seeding_dev) {
1748                 mutex_unlock(&uuid_mutex);
1749                 up_write(&sb->s_umount);
1750
1751                 ret = btrfs_relocate_sys_chunks(root);
1752                 BUG_ON(ret);
1753         }
1754 out:
1755         mutex_unlock(&root->fs_info->volume_mutex);
1756         return ret;
1757 error:
1758         blkdev_put(bdev, FMODE_EXCL);
1759         if (seeding_dev) {
1760                 mutex_unlock(&uuid_mutex);
1761                 up_write(&sb->s_umount);
1762         }
1763         goto out;
1764 }
1765
1766 static noinline int btrfs_update_device(struct btrfs_trans_handle *trans,
1767                                         struct btrfs_device *device)
1768 {
1769         int ret;
1770         struct btrfs_path *path;
1771         struct btrfs_root *root;
1772         struct btrfs_dev_item *dev_item;
1773         struct extent_buffer *leaf;
1774         struct btrfs_key key;
1775
1776         root = device->dev_root->fs_info->chunk_root;
1777
1778         path = btrfs_alloc_path();
1779         if (!path)
1780                 return -ENOMEM;
1781
1782         key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
1783         key.type = BTRFS_DEV_ITEM_KEY;
1784         key.offset = device->devid;
1785
1786         ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
1787         if (ret < 0)
1788                 goto out;
1789
1790         if (ret > 0) {
1791                 ret = -ENOENT;
1792                 goto out;
1793         }
1794
1795         leaf = path->nodes[0];
1796         dev_item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_dev_item);
1797
1798         btrfs_set_device_id(leaf, dev_item, device->devid);
1799         btrfs_set_device_type(leaf, dev_item, device->type);
1800         btrfs_set_device_io_align(leaf, dev_item, device->io_align);
1801         btrfs_set_device_io_width(leaf, dev_item, device->io_width);
1802         btrfs_set_device_sector_size(leaf, dev_item, device->sector_size);
1803         btrfs_set_device_total_bytes(leaf, dev_item, device->disk_total_bytes);
1804         btrfs_set_device_bytes_used(leaf, dev_item, device->bytes_used);
1805         btrfs_mark_buffer_dirty(leaf);
1806
1807 out:
1808         btrfs_free_path(path);
1809         return ret;
1810 }
1811
1812 static int __btrfs_grow_device(struct btrfs_trans_handle *trans,
1813                       struct btrfs_device *device, u64 new_size)
1814 {
1815         struct btrfs_super_block *super_copy =
1816                 device->dev_root->fs_info->super_copy;
1817         u64 old_total = btrfs_super_total_bytes(super_copy);
1818         u64 diff = new_size - device->total_bytes;
1819
1820         if (!device->writeable)
1821                 return -EACCES;
1822         if (new_size <= device->total_bytes)
1823                 return -EINVAL;
1824
1825         btrfs_set_super_total_bytes(super_copy, old_total + diff);
1826         device->fs_devices->total_rw_bytes += diff;
1827
1828         device->total_bytes = new_size;
1829         device->disk_total_bytes = new_size;
1830         btrfs_clear_space_info_full(device->dev_root->fs_info);
1831
1832         return btrfs_update_device(trans, device);
1833 }
1834
1835 int btrfs_grow_device(struct btrfs_trans_handle *trans,
1836                       struct btrfs_device *device, u64 new_size)
1837 {
1838         int ret;
1839         lock_chunks(device->dev_root);
1840         ret = __btrfs_grow_device(trans, device, new_size);
1841         unlock_chunks(device->dev_root);
1842         return ret;
1843 }
1844
1845 static int btrfs_free_chunk(struct btrfs_trans_handle *trans,
1846                             struct btrfs_root *root,
1847                             u64 chunk_tree, u64 chunk_objectid,
1848                             u64 chunk_offset)
1849 {
1850         int ret;
1851         struct btrfs_path *path;
1852         struct btrfs_key key;
1853
1854         root = root->fs_info->chunk_root;
1855         path = btrfs_alloc_path();
1856         if (!path)
1857                 return -ENOMEM;
1858
1859         key.objectid = chunk_objectid;
1860         key.offset = chunk_offset;
1861         key.type = BTRFS_CHUNK_ITEM_KEY;
1862
1863         ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
1864         BUG_ON(ret);
1865
1866         ret = btrfs_del_item(trans, root, path);
1867
1868         btrfs_free_path(path);
1869         return ret;
1870 }
1871
1872 static int btrfs_del_sys_chunk(struct btrfs_root *root, u64 chunk_objectid, u64
1873                         chunk_offset)
1874 {
1875         struct btrfs_super_block *super_copy = root->fs_info->super_copy;
1876         struct btrfs_disk_key *disk_key;
1877         struct btrfs_chunk *chunk;
1878         u8 *ptr;
1879         int ret = 0;
1880         u32 num_stripes;
1881         u32 array_size;
1882         u32 len = 0;
1883         u32 cur;
1884         struct btrfs_key key;
1885
1886         array_size = btrfs_super_sys_array_size(super_copy);
1887
1888         ptr = super_copy->sys_chunk_array;
1889         cur = 0;
1890
1891         while (cur < array_size) {
1892                 disk_key = (struct btrfs_disk_key *)ptr;
1893                 btrfs_disk_key_to_cpu(&key, disk_key);
1894
1895                 len = sizeof(*disk_key);
1896
1897                 if (key.type == BTRFS_CHUNK_ITEM_KEY) {
1898                         chunk = (struct btrfs_chunk *)(ptr + len);
1899                         num_stripes = btrfs_stack_chunk_num_stripes(chunk);
1900                         len += btrfs_chunk_item_size(num_stripes);
1901                 } else {
1902                         ret = -EIO;
1903                         break;
1904                 }
1905                 if (key.objectid == chunk_objectid &&
1906                     key.offset == chunk_offset) {
1907                         memmove(ptr, ptr + len, array_size - (cur + len));
1908                         array_size -= len;
1909                         btrfs_set_super_sys_array_size(super_copy, array_size);
1910                 } else {
1911                         ptr += len;
1912                         cur += len;
1913                 }
1914         }
1915         return ret;
1916 }
1917
1918 static int btrfs_relocate_chunk(struct btrfs_root *root,
1919                          u64 chunk_tree, u64 chunk_objectid,
1920                          u64 chunk_offset)
1921 {
1922         struct extent_map_tree *em_tree;
1923         struct btrfs_root *extent_root;
1924         struct btrfs_trans_handle *trans;
1925         struct extent_map *em;
1926         struct map_lookup *map;
1927         int ret;
1928         int i;
1929
1930         root = root->fs_info->chunk_root;
1931         extent_root = root->fs_info->extent_root;
1932         em_tree = &root->fs_info->mapping_tree.map_tree;
1933
1934         ret = btrfs_can_relocate(extent_root, chunk_offset);
1935         if (ret)
1936                 return -ENOSPC;
1937
1938         /* step one, relocate all the extents inside this chunk */
1939         ret = btrfs_relocate_block_group(extent_root, chunk_offset);
1940         if (ret)
1941                 return ret;
1942
1943         trans = btrfs_start_transaction(root, 0);
1944         BUG_ON(IS_ERR(trans));
1945
1946         lock_chunks(root);
1947
1948         /*
1949          * step two, delete the device extents and the
1950          * chunk tree entries
1951          */
1952         read_lock(&em_tree->lock);
1953         em = lookup_extent_mapping(em_tree, chunk_offset, 1);
1954         read_unlock(&em_tree->lock);
1955
1956         BUG_ON(em->start > chunk_offset ||
1957                em->start + em->len < chunk_offset);
1958         map = (struct map_lookup *)em->bdev;
1959
1960         for (i = 0; i < map->num_stripes; i++) {
1961                 ret = btrfs_free_dev_extent(trans, map->stripes[i].dev,
1962                                             map->stripes[i].physical);
1963                 BUG_ON(ret);
1964
1965                 if (map->stripes[i].dev) {
1966                         ret = btrfs_update_device(trans, map->stripes[i].dev);
1967                         BUG_ON(ret);
1968                 }
1969         }
1970         ret = btrfs_free_chunk(trans, root, chunk_tree, chunk_objectid,
1971                                chunk_offset);
1972
1973         BUG_ON(ret);
1974
1975         trace_btrfs_chunk_free(root, map, chunk_offset, em->len);
1976
1977         if (map->type & BTRFS_BLOCK_GROUP_SYSTEM) {
1978                 ret = btrfs_del_sys_chunk(root, chunk_objectid, chunk_offset);
1979                 BUG_ON(ret);
1980         }
1981
1982         ret = btrfs_remove_block_group(trans, extent_root, chunk_offset);
1983         BUG_ON(ret);
1984
1985         write_lock(&em_tree->lock);
1986         remove_extent_mapping(em_tree, em);
1987         write_unlock(&em_tree->lock);
1988
1989         kfree(map);
1990         em->bdev = NULL;
1991
1992         /* once for the tree */
1993         free_extent_map(em);
1994         /* once for us */
1995         free_extent_map(em);
1996
1997         unlock_chunks(root);
1998         btrfs_end_transaction(trans, root);
1999         return 0;
2000 }
2001
2002 static int btrfs_relocate_sys_chunks(struct btrfs_root *root)
2003 {
2004         struct btrfs_root *chunk_root = root->fs_info->chunk_root;
2005         struct btrfs_path *path;
2006         struct extent_buffer *leaf;
2007         struct btrfs_chunk *chunk;
2008         struct btrfs_key key;
2009         struct btrfs_key found_key;
2010         u64 chunk_tree = chunk_root->root_key.objectid;
2011         u64 chunk_type;
2012         bool retried = false;
2013         int failed = 0;
2014         int ret;
2015
2016         path = btrfs_alloc_path();
2017         if (!path)
2018                 return -ENOMEM;
2019
2020 again:
2021         key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
2022         key.offset = (u64)-1;
2023         key.type = BTRFS_CHUNK_ITEM_KEY;
2024
2025         while (1) {
2026                 ret = btrfs_search_slot(NULL, chunk_root, &key, path, 0, 0);
2027                 if (ret < 0)
2028                         goto error;
2029                 BUG_ON(ret == 0);
2030
2031                 ret = btrfs_previous_item(chunk_root, path, key.objectid,
2032                                           key.type);
2033                 if (ret < 0)
2034                         goto error;
2035                 if (ret > 0)
2036                         break;
2037
2038                 leaf = path->nodes[0];
2039                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
2040
2041                 chunk = btrfs_item_ptr(leaf, path->slots[0],
2042                                        struct btrfs_chunk);
2043                 chunk_type = btrfs_chunk_type(leaf, chunk);
2044                 btrfs_release_path(path);
2045
2046                 if (chunk_type & BTRFS_BLOCK_GROUP_SYSTEM) {
2047                         ret = btrfs_relocate_chunk(chunk_root, chunk_tree,
2048                                                    found_key.objectid,
2049                                                    found_key.offset);
2050                         if (ret == -ENOSPC)
2051                                 failed++;
2052                         else if (ret)
2053                                 BUG();
2054                 }
2055
2056                 if (found_key.offset == 0)
2057                         break;
2058                 key.offset = found_key.offset - 1;
2059         }
2060         ret = 0;
2061         if (failed && !retried) {
2062                 failed = 0;
2063                 retried = true;
2064                 goto again;
2065         } else if (failed && retried) {
2066                 WARN_ON(1);
2067                 ret = -ENOSPC;
2068         }
2069 error:
2070         btrfs_free_path(path);
2071         return ret;
2072 }
2073
2074 static u64 div_factor(u64 num, int factor)
2075 {
2076         if (factor == 10)
2077                 return num;
2078         num *= factor;
2079         do_div(num, 10);
2080         return num;
2081 }
2082
2083 int btrfs_balance(struct btrfs_root *dev_root)
2084 {
2085         int ret;
2086         struct list_head *devices = &dev_root->fs_info->fs_devices->devices;
2087         struct btrfs_device *device;
2088         u64 old_size;
2089         u64 size_to_free;
2090         struct btrfs_path *path;
2091         struct btrfs_key key;
2092         struct btrfs_root *chunk_root = dev_root->fs_info->chunk_root;
2093         struct btrfs_trans_handle *trans;
2094         struct btrfs_key found_key;
2095
2096         if (dev_root->fs_info->sb->s_flags & MS_RDONLY)
2097                 return -EROFS;
2098
2099         if (!capable(CAP_SYS_ADMIN))
2100                 return -EPERM;
2101
2102         mutex_lock(&dev_root->fs_info->volume_mutex);
2103         dev_root = dev_root->fs_info->dev_root;
2104
2105         /* step one make some room on all the devices */
2106         list_for_each_entry(device, devices, dev_list) {
2107                 old_size = device->total_bytes;
2108                 size_to_free = div_factor(old_size, 1);
2109                 size_to_free = min(size_to_free, (u64)1 * 1024 * 1024);
2110                 if (!device->writeable ||
2111                     device->total_bytes - device->bytes_used > size_to_free)
2112                         continue;
2113
2114                 ret = btrfs_shrink_device(device, old_size - size_to_free);
2115                 if (ret == -ENOSPC)
2116                         break;
2117                 BUG_ON(ret);
2118
2119                 trans = btrfs_start_transaction(dev_root, 0);
2120                 BUG_ON(IS_ERR(trans));
2121
2122                 ret = btrfs_grow_device(trans, device, old_size);
2123                 BUG_ON(ret);
2124
2125                 btrfs_end_transaction(trans, dev_root);
2126         }
2127
2128         /* step two, relocate all the chunks */
2129         path = btrfs_alloc_path();
2130         if (!path) {
2131                 ret = -ENOMEM;
2132                 goto error;
2133         }
2134         key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
2135         key.offset = (u64)-1;
2136         key.type = BTRFS_CHUNK_ITEM_KEY;
2137
2138         while (1) {
2139                 ret = btrfs_search_slot(NULL, chunk_root, &key, path, 0, 0);
2140                 if (ret < 0)
2141                         goto error;
2142
2143                 /*
2144                  * this shouldn't happen, it means the last relocate
2145                  * failed
2146                  */
2147                 if (ret == 0)
2148                         break;
2149
2150                 ret = btrfs_previous_item(chunk_root, path, 0,
2151                                           BTRFS_CHUNK_ITEM_KEY);
2152                 if (ret)
2153                         break;
2154
2155                 btrfs_item_key_to_cpu(path->nodes[0], &found_key,
2156                                       path->slots[0]);
2157                 if (found_key.objectid != key.objectid)
2158                         break;
2159
2160                 /* chunk zero is special */
2161                 if (found_key.offset == 0)
2162                         break;
2163
2164                 btrfs_release_path(path);
2165                 ret = btrfs_relocate_chunk(chunk_root,
2166                                            chunk_root->root_key.objectid,
2167                                            found_key.objectid,
2168                                            found_key.offset);
2169                 if (ret && ret != -ENOSPC)
2170                         goto error;
2171                 key.offset = found_key.offset - 1;
2172         }
2173         ret = 0;
2174 error:
2175         btrfs_free_path(path);
2176         mutex_unlock(&dev_root->fs_info->volume_mutex);
2177         return ret;
2178 }
2179
2180 /*
2181  * shrinking a device means finding all of the device extents past
2182  * the new size, and then following the back refs to the chunks.
2183  * The chunk relocation code actually frees the device extent
2184  */
2185 int btrfs_shrink_device(struct btrfs_device *device, u64 new_size)
2186 {
2187         struct btrfs_trans_handle *trans;
2188         struct btrfs_root *root = device->dev_root;
2189         struct btrfs_dev_extent *dev_extent = NULL;
2190         struct btrfs_path *path;
2191         u64 length;
2192         u64 chunk_tree;
2193         u64 chunk_objectid;
2194         u64 chunk_offset;
2195         int ret;
2196         int slot;
2197         int failed = 0;
2198         bool retried = false;
2199         struct extent_buffer *l;
2200         struct btrfs_key key;
2201         struct btrfs_super_block *super_copy = root->fs_info->super_copy;
2202         u64 old_total = btrfs_super_total_bytes(super_copy);
2203         u64 old_size = device->total_bytes;
2204         u64 diff = device->total_bytes - new_size;
2205
2206         if (new_size >= device->total_bytes)
2207                 return -EINVAL;
2208
2209         path = btrfs_alloc_path();
2210         if (!path)
2211                 return -ENOMEM;
2212
2213         path->reada = 2;
2214
2215         lock_chunks(root);
2216
2217         device->total_bytes = new_size;
2218         if (device->writeable) {
2219                 device->fs_devices->total_rw_bytes -= diff;
2220                 spin_lock(&root->fs_info->free_chunk_lock);
2221                 root->fs_info->free_chunk_space -= diff;
2222                 spin_unlock(&root->fs_info->free_chunk_lock);
2223         }
2224         unlock_chunks(root);
2225
2226 again:
2227         key.objectid = device->devid;
2228         key.offset = (u64)-1;
2229         key.type = BTRFS_DEV_EXTENT_KEY;
2230
2231         while (1) {
2232                 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
2233                 if (ret < 0)
2234                         goto done;
2235
2236                 ret = btrfs_previous_item(root, path, 0, key.type);
2237                 if (ret < 0)
2238                         goto done;
2239                 if (ret) {
2240                         ret = 0;
2241                         btrfs_release_path(path);
2242                         break;
2243                 }
2244
2245                 l = path->nodes[0];
2246                 slot = path->slots[0];
2247                 btrfs_item_key_to_cpu(l, &key, path->slots[0]);
2248
2249                 if (key.objectid != device->devid) {
2250                         btrfs_release_path(path);
2251                         break;
2252                 }
2253
2254                 dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
2255                 length = btrfs_dev_extent_length(l, dev_extent);
2256
2257                 if (key.offset + length <= new_size) {
2258                         btrfs_release_path(path);
2259                         break;
2260                 }
2261
2262                 chunk_tree = btrfs_dev_extent_chunk_tree(l, dev_extent);
2263                 chunk_objectid = btrfs_dev_extent_chunk_objectid(l, dev_extent);
2264                 chunk_offset = btrfs_dev_extent_chunk_offset(l, dev_extent);
2265                 btrfs_release_path(path);
2266
2267                 ret = btrfs_relocate_chunk(root, chunk_tree, chunk_objectid,
2268                                            chunk_offset);
2269                 if (ret && ret != -ENOSPC)
2270                         goto done;
2271                 if (ret == -ENOSPC)
2272                         failed++;
2273                 key.offset -= 1;
2274         }
2275
2276         if (failed && !retried) {
2277                 failed = 0;
2278                 retried = true;
2279                 goto again;
2280         } else if (failed && retried) {
2281                 ret = -ENOSPC;
2282                 lock_chunks(root);
2283
2284                 device->total_bytes = old_size;
2285                 if (device->writeable)
2286                         device->fs_devices->total_rw_bytes += diff;
2287                 spin_lock(&root->fs_info->free_chunk_lock);
2288                 root->fs_info->free_chunk_space += diff;
2289                 spin_unlock(&root->fs_info->free_chunk_lock);
2290                 unlock_chunks(root);
2291                 goto done;
2292         }
2293
2294         /* Shrinking succeeded, else we would be at "done". */
2295         trans = btrfs_start_transaction(root, 0);
2296         if (IS_ERR(trans)) {
2297                 ret = PTR_ERR(trans);
2298                 goto done;
2299         }
2300
2301         lock_chunks(root);
2302
2303         device->disk_total_bytes = new_size;
2304         /* Now btrfs_update_device() will change the on-disk size. */
2305         ret = btrfs_update_device(trans, device);
2306         if (ret) {
2307                 unlock_chunks(root);
2308                 btrfs_end_transaction(trans, root);
2309                 goto done;
2310         }
2311         WARN_ON(diff > old_total);
2312         btrfs_set_super_total_bytes(super_copy, old_total - diff);
2313         unlock_chunks(root);
2314         btrfs_end_transaction(trans, root);
2315 done:
2316         btrfs_free_path(path);
2317         return ret;
2318 }
2319
2320 static int btrfs_add_system_chunk(struct btrfs_trans_handle *trans,
2321                            struct btrfs_root *root,
2322                            struct btrfs_key *key,
2323                            struct btrfs_chunk *chunk, int item_size)
2324 {
2325         struct btrfs_super_block *super_copy = root->fs_info->super_copy;
2326         struct btrfs_disk_key disk_key;
2327         u32 array_size;
2328         u8 *ptr;
2329
2330         array_size = btrfs_super_sys_array_size(super_copy);
2331         if (array_size + item_size > BTRFS_SYSTEM_CHUNK_ARRAY_SIZE)
2332                 return -EFBIG;
2333
2334         ptr = super_copy->sys_chunk_array + array_size;
2335         btrfs_cpu_key_to_disk(&disk_key, key);
2336         memcpy(ptr, &disk_key, sizeof(disk_key));
2337         ptr += sizeof(disk_key);
2338         memcpy(ptr, chunk, item_size);
2339         item_size += sizeof(disk_key);
2340         btrfs_set_super_sys_array_size(super_copy, array_size + item_size);
2341         return 0;
2342 }
2343
2344 /*
2345  * sort the devices in descending order by max_avail, total_avail
2346  */
2347 static int btrfs_cmp_device_info(const void *a, const void *b)
2348 {
2349         const struct btrfs_device_info *di_a = a;
2350         const struct btrfs_device_info *di_b = b;
2351
2352         if (di_a->max_avail > di_b->max_avail)
2353                 return -1;
2354         if (di_a->max_avail < di_b->max_avail)
2355                 return 1;
2356         if (di_a->total_avail > di_b->total_avail)
2357                 return -1;
2358         if (di_a->total_avail < di_b->total_avail)
2359                 return 1;
2360         return 0;
2361 }
2362
2363 static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
2364                                struct btrfs_root *extent_root,
2365                                struct map_lookup **map_ret,
2366                                u64 *num_bytes_out, u64 *stripe_size_out,
2367                                u64 start, u64 type)
2368 {
2369         struct btrfs_fs_info *info = extent_root->fs_info;
2370         struct btrfs_fs_devices *fs_devices = info->fs_devices;
2371         struct list_head *cur;
2372         struct map_lookup *map = NULL;
2373         struct extent_map_tree *em_tree;
2374         struct extent_map *em;
2375         struct btrfs_device_info *devices_info = NULL;
2376         u64 total_avail;
2377         int num_stripes;        /* total number of stripes to allocate */
2378         int sub_stripes;        /* sub_stripes info for map */
2379         int dev_stripes;        /* stripes per dev */
2380         int devs_max;           /* max devs to use */
2381         int devs_min;           /* min devs needed */
2382         int devs_increment;     /* ndevs has to be a multiple of this */
2383         int ncopies;            /* how many copies to data has */
2384         int ret;
2385         u64 max_stripe_size;
2386         u64 max_chunk_size;
2387         u64 stripe_size;
2388         u64 num_bytes;
2389         int ndevs;
2390         int i;
2391         int j;
2392
2393         if ((type & BTRFS_BLOCK_GROUP_RAID1) &&
2394             (type & BTRFS_BLOCK_GROUP_DUP)) {
2395                 WARN_ON(1);
2396                 type &= ~BTRFS_BLOCK_GROUP_DUP;
2397         }
2398
2399         if (list_empty(&fs_devices->alloc_list))
2400                 return -ENOSPC;
2401
2402         sub_stripes = 1;
2403         dev_stripes = 1;
2404         devs_increment = 1;
2405         ncopies = 1;
2406         devs_max = 0;   /* 0 == as many as possible */
2407         devs_min = 1;
2408
2409         /*
2410          * define the properties of each RAID type.
2411          * FIXME: move this to a global table and use it in all RAID
2412          * calculation code
2413          */
2414         if (type & (BTRFS_BLOCK_GROUP_DUP)) {
2415                 dev_stripes = 2;
2416                 ncopies = 2;
2417                 devs_max = 1;
2418         } else if (type & (BTRFS_BLOCK_GROUP_RAID0)) {
2419                 devs_min = 2;
2420         } else if (type & (BTRFS_BLOCK_GROUP_RAID1)) {
2421                 devs_increment = 2;
2422                 ncopies = 2;
2423                 devs_max = 2;
2424                 devs_min = 2;
2425         } else if (type & (BTRFS_BLOCK_GROUP_RAID10)) {
2426                 sub_stripes = 2;
2427                 devs_increment = 2;
2428                 ncopies = 2;
2429                 devs_min = 4;
2430         } else {
2431                 devs_max = 1;
2432         }
2433
2434         if (type & BTRFS_BLOCK_GROUP_DATA) {
2435                 max_stripe_size = 1024 * 1024 * 1024;
2436                 max_chunk_size = 10 * max_stripe_size;
2437         } else if (type & BTRFS_BLOCK_GROUP_METADATA) {
2438                 max_stripe_size = 256 * 1024 * 1024;
2439                 max_chunk_size = max_stripe_size;
2440         } else if (type & BTRFS_BLOCK_GROUP_SYSTEM) {
2441                 max_stripe_size = 8 * 1024 * 1024;
2442                 max_chunk_size = 2 * max_stripe_size;
2443         } else {
2444                 printk(KERN_ERR "btrfs: invalid chunk type 0x%llx requested\n",
2445                        type);
2446                 BUG_ON(1);
2447         }
2448
2449         /* we don't want a chunk larger than 10% of writeable space */
2450         max_chunk_size = min(div_factor(fs_devices->total_rw_bytes, 1),
2451                              max_chunk_size);
2452
2453         devices_info = kzalloc(sizeof(*devices_info) * fs_devices->rw_devices,
2454                                GFP_NOFS);
2455         if (!devices_info)
2456                 return -ENOMEM;
2457
2458         cur = fs_devices->alloc_list.next;
2459
2460         /*
2461          * in the first pass through the devices list, we gather information
2462          * about the available holes on each device.
2463          */
2464         ndevs = 0;
2465         while (cur != &fs_devices->alloc_list) {
2466                 struct btrfs_device *device;
2467                 u64 max_avail;
2468                 u64 dev_offset;
2469
2470                 device = list_entry(cur, struct btrfs_device, dev_alloc_list);
2471
2472                 cur = cur->next;
2473
2474                 if (!device->writeable) {
2475                         printk(KERN_ERR
2476                                "btrfs: read-only device in alloc_list\n");
2477                         WARN_ON(1);
2478                         continue;
2479                 }
2480
2481                 if (!device->in_fs_metadata)
2482                         continue;
2483
2484                 if (device->total_bytes > device->bytes_used)
2485                         total_avail = device->total_bytes - device->bytes_used;
2486                 else
2487                         total_avail = 0;
2488
2489                 /* If there is no space on this device, skip it. */
2490                 if (total_avail == 0)
2491                         continue;
2492
2493                 ret = find_free_dev_extent(trans, device,
2494                                            max_stripe_size * dev_stripes,
2495                                            &dev_offset, &max_avail);
2496                 if (ret && ret != -ENOSPC)
2497                         goto error;
2498
2499                 if (ret == 0)
2500                         max_avail = max_stripe_size * dev_stripes;
2501
2502                 if (max_avail < BTRFS_STRIPE_LEN * dev_stripes)
2503                         continue;
2504
2505                 devices_info[ndevs].dev_offset = dev_offset;
2506                 devices_info[ndevs].max_avail = max_avail;
2507                 devices_info[ndevs].total_avail = total_avail;
2508                 devices_info[ndevs].dev = device;
2509                 ++ndevs;
2510         }
2511
2512         /*
2513          * now sort the devices by hole size / available space
2514          */
2515         sort(devices_info, ndevs, sizeof(struct btrfs_device_info),
2516              btrfs_cmp_device_info, NULL);
2517
2518         /* round down to number of usable stripes */
2519         ndevs -= ndevs % devs_increment;
2520
2521         if (ndevs < devs_increment * sub_stripes || ndevs < devs_min) {
2522                 ret = -ENOSPC;
2523                 goto error;
2524         }
2525
2526         if (devs_max && ndevs > devs_max)
2527                 ndevs = devs_max;
2528         /*
2529          * the primary goal is to maximize the number of stripes, so use as many
2530          * devices as possible, even if the stripes are not maximum sized.
2531          */
2532         stripe_size = devices_info[ndevs-1].max_avail;
2533         num_stripes = ndevs * dev_stripes;
2534
2535         if (stripe_size * num_stripes > max_chunk_size * ncopies) {
2536                 stripe_size = max_chunk_size * ncopies;
2537                 do_div(stripe_size, num_stripes);
2538         }
2539
2540         do_div(stripe_size, dev_stripes);
2541         do_div(stripe_size, BTRFS_STRIPE_LEN);
2542         stripe_size *= BTRFS_STRIPE_LEN;
2543
2544         map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
2545         if (!map) {
2546                 ret = -ENOMEM;
2547                 goto error;
2548         }
2549         map->num_stripes = num_stripes;
2550
2551         for (i = 0; i < ndevs; ++i) {
2552                 for (j = 0; j < dev_stripes; ++j) {
2553                         int s = i * dev_stripes + j;
2554                         map->stripes[s].dev = devices_info[i].dev;
2555                         map->stripes[s].physical = devices_info[i].dev_offset +
2556                                                    j * stripe_size;
2557                 }
2558         }
2559         map->sector_size = extent_root->sectorsize;
2560         map->stripe_len = BTRFS_STRIPE_LEN;
2561         map->io_align = BTRFS_STRIPE_LEN;
2562         map->io_width = BTRFS_STRIPE_LEN;
2563         map->type = type;
2564         map->sub_stripes = sub_stripes;
2565
2566         *map_ret = map;
2567         num_bytes = stripe_size * (num_stripes / ncopies);
2568
2569         *stripe_size_out = stripe_size;
2570         *num_bytes_out = num_bytes;
2571
2572         trace_btrfs_chunk_alloc(info->chunk_root, map, start, num_bytes);
2573
2574         em = alloc_extent_map();
2575         if (!em) {
2576                 ret = -ENOMEM;
2577                 goto error;
2578         }
2579         em->bdev = (struct block_device *)map;
2580         em->start = start;
2581         em->len = num_bytes;
2582         em->block_start = 0;
2583         em->block_len = em->len;
2584
2585         em_tree = &extent_root->fs_info->mapping_tree.map_tree;
2586         write_lock(&em_tree->lock);
2587         ret = add_extent_mapping(em_tree, em);
2588         write_unlock(&em_tree->lock);
2589         BUG_ON(ret);
2590         free_extent_map(em);
2591
2592         ret = btrfs_make_block_group(trans, extent_root, 0, type,
2593                                      BTRFS_FIRST_CHUNK_TREE_OBJECTID,
2594                                      start, num_bytes);
2595         BUG_ON(ret);
2596
2597         for (i = 0; i < map->num_stripes; ++i) {
2598                 struct btrfs_device *device;
2599                 u64 dev_offset;
2600
2601                 device = map->stripes[i].dev;
2602                 dev_offset = map->stripes[i].physical;
2603
2604                 ret = btrfs_alloc_dev_extent(trans, device,
2605                                 info->chunk_root->root_key.objectid,
2606                                 BTRFS_FIRST_CHUNK_TREE_OBJECTID,
2607                                 start, dev_offset, stripe_size);
2608                 BUG_ON(ret);
2609         }
2610
2611         kfree(devices_info);
2612         return 0;
2613
2614 error:
2615         kfree(map);
2616         kfree(devices_info);
2617         return ret;
2618 }
2619
2620 static int __finish_chunk_alloc(struct btrfs_trans_handle *trans,
2621                                 struct btrfs_root *extent_root,
2622                                 struct map_lookup *map, u64 chunk_offset,
2623                                 u64 chunk_size, u64 stripe_size)
2624 {
2625         u64 dev_offset;
2626         struct btrfs_key key;
2627         struct btrfs_root *chunk_root = extent_root->fs_info->chunk_root;
2628         struct btrfs_device *device;
2629         struct btrfs_chunk *chunk;
2630         struct btrfs_stripe *stripe;
2631         size_t item_size = btrfs_chunk_item_size(map->num_stripes);
2632         int index = 0;
2633         int ret;
2634
2635         chunk = kzalloc(item_size, GFP_NOFS);
2636         if (!chunk)
2637                 return -ENOMEM;
2638
2639         index = 0;
2640         while (index < map->num_stripes) {
2641                 device = map->stripes[index].dev;
2642                 device->bytes_used += stripe_size;
2643                 ret = btrfs_update_device(trans, device);
2644                 BUG_ON(ret);
2645                 index++;
2646         }
2647
2648         spin_lock(&extent_root->fs_info->free_chunk_lock);
2649         extent_root->fs_info->free_chunk_space -= (stripe_size *
2650                                                    map->num_stripes);
2651         spin_unlock(&extent_root->fs_info->free_chunk_lock);
2652
2653         index = 0;
2654         stripe = &chunk->stripe;
2655         while (index < map->num_stripes) {
2656                 device = map->stripes[index].dev;
2657                 dev_offset = map->stripes[index].physical;
2658
2659                 btrfs_set_stack_stripe_devid(stripe, device->devid);
2660                 btrfs_set_stack_stripe_offset(stripe, dev_offset);
2661                 memcpy(stripe->dev_uuid, device->uuid, BTRFS_UUID_SIZE);
2662                 stripe++;
2663                 index++;
2664         }
2665
2666         btrfs_set_stack_chunk_length(chunk, chunk_size);
2667         btrfs_set_stack_chunk_owner(chunk, extent_root->root_key.objectid);
2668         btrfs_set_stack_chunk_stripe_len(chunk, map->stripe_len);
2669         btrfs_set_stack_chunk_type(chunk, map->type);
2670         btrfs_set_stack_chunk_num_stripes(chunk, map->num_stripes);
2671         btrfs_set_stack_chunk_io_align(chunk, map->stripe_len);
2672         btrfs_set_stack_chunk_io_width(chunk, map->stripe_len);
2673         btrfs_set_stack_chunk_sector_size(chunk, extent_root->sectorsize);
2674         btrfs_set_stack_chunk_sub_stripes(chunk, map->sub_stripes);
2675
2676         key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
2677         key.type = BTRFS_CHUNK_ITEM_KEY;
2678         key.offset = chunk_offset;
2679
2680         ret = btrfs_insert_item(trans, chunk_root, &key, chunk, item_size);
2681         BUG_ON(ret);
2682
2683         if (map->type & BTRFS_BLOCK_GROUP_SYSTEM) {
2684                 ret = btrfs_add_system_chunk(trans, chunk_root, &key, chunk,
2685                                              item_size);
2686                 BUG_ON(ret);
2687         }
2688
2689         kfree(chunk);
2690         return 0;
2691 }
2692
2693 /*
2694  * Chunk allocation falls into two parts. The first part does works
2695  * that make the new allocated chunk useable, but not do any operation
2696  * that modifies the chunk tree. The second part does the works that
2697  * require modifying the chunk tree. This division is important for the
2698  * bootstrap process of adding storage to a seed btrfs.
2699  */
2700 int btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
2701                       struct btrfs_root *extent_root, u64 type)
2702 {
2703         u64 chunk_offset;
2704         u64 chunk_size;
2705         u64 stripe_size;
2706         struct map_lookup *map;
2707         struct btrfs_root *chunk_root = extent_root->fs_info->chunk_root;
2708         int ret;
2709
2710         ret = find_next_chunk(chunk_root, BTRFS_FIRST_CHUNK_TREE_OBJECTID,
2711                               &chunk_offset);
2712         if (ret)
2713                 return ret;
2714
2715         ret = __btrfs_alloc_chunk(trans, extent_root, &map, &chunk_size,
2716                                   &stripe_size, chunk_offset, type);
2717         if (ret)
2718                 return ret;
2719
2720         ret = __finish_chunk_alloc(trans, extent_root, map, chunk_offset,
2721                                    chunk_size, stripe_size);
2722         BUG_ON(ret);
2723         return 0;
2724 }
2725
2726 static noinline int init_first_rw_device(struct btrfs_trans_handle *trans,
2727                                          struct btrfs_root *root,
2728                                          struct btrfs_device *device)
2729 {
2730         u64 chunk_offset;
2731         u64 sys_chunk_offset;
2732         u64 chunk_size;
2733         u64 sys_chunk_size;
2734         u64 stripe_size;
2735         u64 sys_stripe_size;
2736         u64 alloc_profile;
2737         struct map_lookup *map;
2738         struct map_lookup *sys_map;
2739         struct btrfs_fs_info *fs_info = root->fs_info;
2740         struct btrfs_root *extent_root = fs_info->extent_root;
2741         int ret;
2742
2743         ret = find_next_chunk(fs_info->chunk_root,
2744                               BTRFS_FIRST_CHUNK_TREE_OBJECTID, &chunk_offset);
2745         if (ret)
2746                 return ret;
2747
2748         alloc_profile = BTRFS_BLOCK_GROUP_METADATA |
2749                         (fs_info->metadata_alloc_profile &
2750                          fs_info->avail_metadata_alloc_bits);
2751         alloc_profile = btrfs_reduce_alloc_profile(root, alloc_profile);
2752
2753         ret = __btrfs_alloc_chunk(trans, extent_root, &map, &chunk_size,
2754                                   &stripe_size, chunk_offset, alloc_profile);
2755         BUG_ON(ret);
2756
2757         sys_chunk_offset = chunk_offset + chunk_size;
2758
2759         alloc_profile = BTRFS_BLOCK_GROUP_SYSTEM |
2760                         (fs_info->system_alloc_profile &
2761                          fs_info->avail_system_alloc_bits);
2762         alloc_profile = btrfs_reduce_alloc_profile(root, alloc_profile);
2763
2764         ret = __btrfs_alloc_chunk(trans, extent_root, &sys_map,
2765                                   &sys_chunk_size, &sys_stripe_size,
2766                                   sys_chunk_offset, alloc_profile);
2767         BUG_ON(ret);
2768
2769         ret = btrfs_add_device(trans, fs_info->chunk_root, device);
2770         BUG_ON(ret);
2771
2772         /*
2773          * Modifying chunk tree needs allocating new blocks from both
2774          * system block group and metadata block group. So we only can
2775          * do operations require modifying the chunk tree after both
2776          * block groups were created.
2777          */
2778         ret = __finish_chunk_alloc(trans, extent_root, map, chunk_offset,
2779                                    chunk_size, stripe_size);
2780         BUG_ON(ret);
2781
2782         ret = __finish_chunk_alloc(trans, extent_root, sys_map,
2783                                    sys_chunk_offset, sys_chunk_size,
2784                                    sys_stripe_size);
2785         BUG_ON(ret);
2786         return 0;
2787 }
2788
2789 int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset)
2790 {
2791         struct extent_map *em;
2792         struct map_lookup *map;
2793         struct btrfs_mapping_tree *map_tree = &root->fs_info->mapping_tree;
2794         int readonly = 0;
2795         int i;
2796
2797         read_lock(&map_tree->map_tree.lock);
2798         em = lookup_extent_mapping(&map_tree->map_tree, chunk_offset, 1);
2799         read_unlock(&map_tree->map_tree.lock);
2800         if (!em)
2801                 return 1;
2802
2803         if (btrfs_test_opt(root, DEGRADED)) {
2804                 free_extent_map(em);
2805                 return 0;
2806         }
2807
2808         map = (struct map_lookup *)em->bdev;
2809         for (i = 0; i < map->num_stripes; i++) {
2810                 if (!map->stripes[i].dev->writeable) {
2811                         readonly = 1;
2812                         break;
2813                 }
2814         }
2815         free_extent_map(em);
2816         return readonly;
2817 }
2818
2819 void btrfs_mapping_init(struct btrfs_mapping_tree *tree)
2820 {
2821         extent_map_tree_init(&tree->map_tree);
2822 }
2823
2824 void btrfs_mapping_tree_free(struct btrfs_mapping_tree *tree)
2825 {
2826         struct extent_map *em;
2827
2828         while (1) {
2829                 write_lock(&tree->map_tree.lock);
2830                 em = lookup_extent_mapping(&tree->map_tree, 0, (u64)-1);
2831                 if (em)
2832                         remove_extent_mapping(&tree->map_tree, em);
2833                 write_unlock(&tree->map_tree.lock);
2834                 if (!em)
2835                         break;
2836                 kfree(em->bdev);
2837                 /* once for us */
2838                 free_extent_map(em);
2839                 /* once for the tree */
2840                 free_extent_map(em);
2841         }
2842 }
2843
2844 int btrfs_num_copies(struct btrfs_mapping_tree *map_tree, u64 logical, u64 len)
2845 {
2846         struct extent_map *em;
2847         struct map_lookup *map;
2848         struct extent_map_tree *em_tree = &map_tree->map_tree;
2849         int ret;
2850
2851         read_lock(&em_tree->lock);
2852         em = lookup_extent_mapping(em_tree, logical, len);
2853         read_unlock(&em_tree->lock);
2854         BUG_ON(!em);
2855
2856         BUG_ON(em->start > logical || em->start + em->len < logical);
2857         map = (struct map_lookup *)em->bdev;
2858         if (map->type & (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1))
2859                 ret = map->num_stripes;
2860         else if (map->type & BTRFS_BLOCK_GROUP_RAID10)
2861                 ret = map->sub_stripes;
2862         else
2863                 ret = 1;
2864         free_extent_map(em);
2865         return ret;
2866 }
2867
2868 static int find_live_mirror(struct map_lookup *map, int first, int num,
2869                             int optimal)
2870 {
2871         int i;
2872         if (map->stripes[optimal].dev->bdev)
2873                 return optimal;
2874         for (i = first; i < first + num; i++) {
2875                 if (map->stripes[i].dev->bdev)
2876                         return i;
2877         }
2878         /* we couldn't find one that doesn't fail.  Just return something
2879          * and the io error handling code will clean up eventually
2880          */
2881         return optimal;
2882 }
2883
2884 static int __btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw,
2885                              u64 logical, u64 *length,
2886                              struct btrfs_bio **bbio_ret,
2887                              int mirror_num)
2888 {
2889         struct extent_map *em;
2890         struct map_lookup *map;
2891         struct extent_map_tree *em_tree = &map_tree->map_tree;
2892         u64 offset;
2893         u64 stripe_offset;
2894         u64 stripe_end_offset;
2895         u64 stripe_nr;
2896         u64 stripe_nr_orig;
2897         u64 stripe_nr_end;
2898         int stripes_allocated = 8;
2899         int stripes_required = 1;
2900         int stripe_index;
2901         int i;
2902         int num_stripes;
2903         int max_errors = 0;
2904         struct btrfs_bio *bbio = NULL;
2905
2906         if (bbio_ret && !(rw & (REQ_WRITE | REQ_DISCARD)))
2907                 stripes_allocated = 1;
2908 again:
2909         if (bbio_ret) {
2910                 bbio = kzalloc(btrfs_bio_size(stripes_allocated),
2911                                 GFP_NOFS);
2912                 if (!bbio)
2913                         return -ENOMEM;
2914
2915                 atomic_set(&bbio->error, 0);
2916         }
2917
2918         read_lock(&em_tree->lock);
2919         em = lookup_extent_mapping(em_tree, logical, *length);
2920         read_unlock(&em_tree->lock);
2921
2922         if (!em) {
2923                 printk(KERN_CRIT "unable to find logical %llu len %llu\n",
2924                        (unsigned long long)logical,
2925                        (unsigned long long)*length);
2926                 BUG();
2927         }
2928
2929         BUG_ON(em->start > logical || em->start + em->len < logical);
2930         map = (struct map_lookup *)em->bdev;
2931         offset = logical - em->start;
2932
2933         if (mirror_num > map->num_stripes)
2934                 mirror_num = 0;
2935
2936         /* if our btrfs_bio struct is too small, back off and try again */
2937         if (rw & REQ_WRITE) {
2938                 if (map->type & (BTRFS_BLOCK_GROUP_RAID1 |
2939                                  BTRFS_BLOCK_GROUP_DUP)) {
2940                         stripes_required = map->num_stripes;
2941                         max_errors = 1;
2942                 } else if (map->type & BTRFS_BLOCK_GROUP_RAID10) {
2943                         stripes_required = map->sub_stripes;
2944                         max_errors = 1;
2945                 }
2946         }
2947         if (rw & REQ_DISCARD) {
2948                 if (map->type & (BTRFS_BLOCK_GROUP_RAID0 |
2949                                  BTRFS_BLOCK_GROUP_RAID1 |
2950                                  BTRFS_BLOCK_GROUP_DUP |
2951                                  BTRFS_BLOCK_GROUP_RAID10)) {
2952                         stripes_required = map->num_stripes;
2953                 }
2954         }
2955         if (bbio_ret && (rw & (REQ_WRITE | REQ_DISCARD)) &&
2956             stripes_allocated < stripes_required) {
2957                 stripes_allocated = map->num_stripes;
2958                 free_extent_map(em);
2959                 kfree(bbio);
2960                 goto again;
2961         }
2962         stripe_nr = offset;
2963         /*
2964          * stripe_nr counts the total number of stripes we have to stride
2965          * to get to this block
2966          */
2967         do_div(stripe_nr, map->stripe_len);
2968
2969         stripe_offset = stripe_nr * map->stripe_len;
2970         BUG_ON(offset < stripe_offset);
2971
2972         /* stripe_offset is the offset of this block in its stripe*/
2973         stripe_offset = offset - stripe_offset;
2974
2975         if (rw & REQ_DISCARD)
2976                 *length = min_t(u64, em->len - offset, *length);
2977         else if (map->type & (BTRFS_BLOCK_GROUP_RAID0 |
2978                               BTRFS_BLOCK_GROUP_RAID1 |
2979                               BTRFS_BLOCK_GROUP_RAID10 |
2980                               BTRFS_BLOCK_GROUP_DUP)) {
2981                 /* we limit the length of each bio to what fits in a stripe */
2982                 *length = min_t(u64, em->len - offset,
2983                                 map->stripe_len - stripe_offset);
2984         } else {
2985                 *length = em->len - offset;
2986         }
2987
2988         if (!bbio_ret)
2989                 goto out;
2990
2991         num_stripes = 1;
2992         stripe_index = 0;
2993         stripe_nr_orig = stripe_nr;
2994         stripe_nr_end = (offset + *length + map->stripe_len - 1) &
2995                         (~(map->stripe_len - 1));
2996         do_div(stripe_nr_end, map->stripe_len);
2997         stripe_end_offset = stripe_nr_end * map->stripe_len -
2998                             (offset + *length);
2999         if (map->type & BTRFS_BLOCK_GROUP_RAID0) {
3000                 if (rw & REQ_DISCARD)
3001                         num_stripes = min_t(u64, map->num_stripes,
3002                                             stripe_nr_end - stripe_nr_orig);
3003                 stripe_index = do_div(stripe_nr, map->num_stripes);
3004         } else if (map->type & BTRFS_BLOCK_GROUP_RAID1) {
3005                 if (rw & (REQ_WRITE | REQ_DISCARD))
3006                         num_stripes = map->num_stripes;
3007                 else if (mirror_num)
3008                         stripe_index = mirror_num - 1;
3009                 else {
3010                         stripe_index = find_live_mirror(map, 0,
3011                                             map->num_stripes,
3012                                             current->pid % map->num_stripes);
3013                         mirror_num = stripe_index + 1;
3014                 }
3015
3016         } else if (map->type & BTRFS_BLOCK_GROUP_DUP) {
3017                 if (rw & (REQ_WRITE | REQ_DISCARD)) {
3018                         num_stripes = map->num_stripes;
3019                 } else if (mirror_num) {
3020                         stripe_index = mirror_num - 1;
3021                 } else {
3022                         mirror_num = 1;
3023                 }
3024
3025         } else if (map->type & BTRFS_BLOCK_GROUP_RAID10) {
3026                 int factor = map->num_stripes / map->sub_stripes;
3027
3028                 stripe_index = do_div(stripe_nr, factor);
3029                 stripe_index *= map->sub_stripes;
3030
3031                 if (rw & REQ_WRITE)
3032                         num_stripes = map->sub_stripes;
3033                 else if (rw & REQ_DISCARD)
3034                         num_stripes = min_t(u64, map->sub_stripes *
3035                                             (stripe_nr_end - stripe_nr_orig),
3036                                             map->num_stripes);
3037                 else if (mirror_num)
3038                         stripe_index += mirror_num - 1;
3039                 else {
3040                         stripe_index = find_live_mirror(map, stripe_index,
3041                                               map->sub_stripes, stripe_index +
3042                                               current->pid % map->sub_stripes);
3043                         mirror_num = stripe_index + 1;
3044                 }
3045         } else {
3046                 /*
3047                  * after this do_div call, stripe_nr is the number of stripes
3048                  * on this device we have to walk to find the data, and
3049                  * stripe_index is the number of our device in the stripe array
3050                  */
3051                 stripe_index = do_div(stripe_nr, map->num_stripes);
3052                 mirror_num = stripe_index + 1;
3053         }
3054         BUG_ON(stripe_index >= map->num_stripes);
3055
3056         if (rw & REQ_DISCARD) {
3057                 for (i = 0; i < num_stripes; i++) {
3058                         bbio->stripes[i].physical =
3059                                 map->stripes[stripe_index].physical +
3060                                 stripe_offset + stripe_nr * map->stripe_len;
3061                         bbio->stripes[i].dev = map->stripes[stripe_index].dev;
3062
3063                         if (map->type & BTRFS_BLOCK_GROUP_RAID0) {
3064                                 u64 stripes;
3065                                 u32 last_stripe = 0;
3066                                 int j;
3067
3068                                 div_u64_rem(stripe_nr_end - 1,
3069                                             map->num_stripes,
3070                                             &last_stripe);
3071
3072                                 for (j = 0; j < map->num_stripes; j++) {
3073                                         u32 test;
3074
3075                                         div_u64_rem(stripe_nr_end - 1 - j,
3076                                                     map->num_stripes, &test);
3077                                         if (test == stripe_index)
3078                                                 break;
3079                                 }
3080                                 stripes = stripe_nr_end - 1 - j;
3081                                 do_div(stripes, map->num_stripes);
3082                                 bbio->stripes[i].length = map->stripe_len *
3083                                         (stripes - stripe_nr + 1);
3084
3085                                 if (i == 0) {
3086                                         bbio->stripes[i].length -=
3087                                                 stripe_offset;
3088                                         stripe_offset = 0;
3089                                 }
3090                                 if (stripe_index == last_stripe)
3091                                         bbio->stripes[i].length -=
3092                                                 stripe_end_offset;
3093                         } else if (map->type & BTRFS_BLOCK_GROUP_RAID10) {
3094                                 u64 stripes;
3095                                 int j;
3096                                 int factor = map->num_stripes /
3097                                              map->sub_stripes;
3098                                 u32 last_stripe = 0;
3099
3100                                 div_u64_rem(stripe_nr_end - 1,
3101                                             factor, &last_stripe);
3102                                 last_stripe *= map->sub_stripes;
3103
3104                                 for (j = 0; j < factor; j++) {
3105                                         u32 test;
3106
3107                                         div_u64_rem(stripe_nr_end - 1 - j,
3108                                                     factor, &test);
3109
3110                                         if (test ==
3111                                             stripe_index / map->sub_stripes)
3112                                                 break;
3113                                 }
3114                                 stripes = stripe_nr_end - 1 - j;
3115                                 do_div(stripes, factor);
3116                                 bbio->stripes[i].length = map->stripe_len *
3117                                         (stripes - stripe_nr + 1);
3118
3119                                 if (i < map->sub_stripes) {
3120                                         bbio->stripes[i].length -=
3121                                                 stripe_offset;
3122                                         if (i == map->sub_stripes - 1)
3123                                                 stripe_offset = 0;
3124                                 }
3125                                 if (stripe_index >= last_stripe &&
3126                                     stripe_index <= (last_stripe +
3127                                                      map->sub_stripes - 1)) {
3128                                         bbio->stripes[i].length -=
3129                                                 stripe_end_offset;
3130                                 }
3131                         } else
3132                                 bbio->stripes[i].length = *length;
3133
3134                         stripe_index++;
3135                         if (stripe_index == map->num_stripes) {
3136                                 /* This could only happen for RAID0/10 */
3137                                 stripe_index = 0;
3138                                 stripe_nr++;
3139                         }
3140                 }
3141         } else {
3142                 for (i = 0; i < num_stripes; i++) {
3143                         bbio->stripes[i].physical =
3144                                 map->stripes[stripe_index].physical +
3145                                 stripe_offset +
3146                                 stripe_nr * map->stripe_len;
3147                         bbio->stripes[i].dev =
3148                                 map->stripes[stripe_index].dev;
3149                         stripe_index++;
3150                 }
3151         }
3152         if (bbio_ret) {
3153                 *bbio_ret = bbio;
3154                 bbio->num_stripes = num_stripes;
3155                 bbio->max_errors = max_errors;
3156                 bbio->mirror_num = mirror_num;
3157         }
3158 out:
3159         free_extent_map(em);
3160         return 0;
3161 }
3162
3163 int btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw,
3164                       u64 logical, u64 *length,
3165                       struct btrfs_bio **bbio_ret, int mirror_num)
3166 {
3167         return __btrfs_map_block(map_tree, rw, logical, length, bbio_ret,
3168                                  mirror_num);
3169 }
3170
3171 int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree,
3172                      u64 chunk_start, u64 physical, u64 devid,
3173                      u64 **logical, int *naddrs, int *stripe_len)
3174 {
3175         struct extent_map_tree *em_tree = &map_tree->map_tree;
3176         struct extent_map *em;
3177         struct map_lookup *map;
3178         u64 *buf;
3179         u64 bytenr;
3180         u64 length;
3181         u64 stripe_nr;
3182         int i, j, nr = 0;
3183
3184         read_lock(&em_tree->lock);
3185         em = lookup_extent_mapping(em_tree, chunk_start, 1);
3186         read_unlock(&em_tree->lock);
3187
3188         BUG_ON(!em || em->start != chunk_start);
3189         map = (struct map_lookup *)em->bdev;
3190
3191         length = em->len;
3192         if (map->type & BTRFS_BLOCK_GROUP_RAID10)
3193                 do_div(length, map->num_stripes / map->sub_stripes);
3194         else if (map->type & BTRFS_BLOCK_GROUP_RAID0)
3195                 do_div(length, map->num_stripes);
3196
3197         buf = kzalloc(sizeof(u64) * map->num_stripes, GFP_NOFS);
3198         BUG_ON(!buf);
3199
3200         for (i = 0; i < map->num_stripes; i++) {
3201                 if (devid && map->stripes[i].dev->devid != devid)
3202                         continue;
3203                 if (map->stripes[i].physical > physical ||
3204                     map->stripes[i].physical + length <= physical)
3205                         continue;
3206
3207                 stripe_nr = physical - map->stripes[i].physical;
3208                 do_div(stripe_nr, map->stripe_len);
3209
3210                 if (map->type & BTRFS_BLOCK_GROUP_RAID10) {
3211                         stripe_nr = stripe_nr * map->num_stripes + i;
3212                         do_div(stripe_nr, map->sub_stripes);
3213                 } else if (map->type & BTRFS_BLOCK_GROUP_RAID0) {
3214                         stripe_nr = stripe_nr * map->num_stripes + i;
3215                 }
3216                 bytenr = chunk_start + stripe_nr * map->stripe_len;
3217                 WARN_ON(nr >= map->num_stripes);
3218                 for (j = 0; j < nr; j++) {
3219                         if (buf[j] == bytenr)
3220                                 break;
3221                 }
3222                 if (j == nr) {
3223                         WARN_ON(nr >= map->num_stripes);
3224                         buf[nr++] = bytenr;
3225                 }
3226         }
3227
3228         *logical = buf;
3229         *naddrs = nr;
3230         *stripe_len = map->stripe_len;
3231
3232         free_extent_map(em);
3233         return 0;
3234 }
3235
3236 static void btrfs_end_bio(struct bio *bio, int err)
3237 {
3238         struct btrfs_bio *bbio = bio->bi_private;
3239         int is_orig_bio = 0;
3240
3241         if (err)
3242                 atomic_inc(&bbio->error);
3243
3244         if (bio == bbio->orig_bio)
3245                 is_orig_bio = 1;
3246
3247         if (atomic_dec_and_test(&bbio->stripes_pending)) {
3248                 if (!is_orig_bio) {
3249                         bio_put(bio);
3250                         bio = bbio->orig_bio;
3251                 }
3252                 bio->bi_private = bbio->private;
3253                 bio->bi_end_io = bbio->end_io;
3254                 bio->bi_bdev = (struct block_device *)
3255                                         (unsigned long)bbio->mirror_num;
3256                 /* only send an error to the higher layers if it is
3257                  * beyond the tolerance of the multi-bio
3258                  */
3259                 if (atomic_read(&bbio->error) > bbio->max_errors) {
3260                         err = -EIO;
3261                 } else if (err) {
3262                         /*
3263                          * this bio is actually up to date, we didn't
3264                          * go over the max number of errors
3265                          */
3266                         set_bit(BIO_UPTODATE, &bio->bi_flags);
3267                         err = 0;
3268                 }
3269                 kfree(bbio);
3270
3271                 bio_endio(bio, err);
3272         } else if (!is_orig_bio) {
3273                 bio_put(bio);
3274         }
3275 }
3276
3277 struct async_sched {
3278         struct bio *bio;
3279         int rw;
3280         struct btrfs_fs_info *info;
3281         struct btrfs_work work;
3282 };
3283
3284 /*
3285  * see run_scheduled_bios for a description of why bios are collected for
3286  * async submit.
3287  *
3288  * This will add one bio to the pending list for a device and make sure
3289  * the work struct is scheduled.
3290  */
3291 static noinline int schedule_bio(struct btrfs_root *root,
3292                                  struct btrfs_device *device,
3293                                  int rw, struct bio *bio)
3294 {
3295         int should_queue = 1;
3296         struct btrfs_pending_bios *pending_bios;
3297
3298         /* don't bother with additional async steps for reads, right now */
3299         if (!(rw & REQ_WRITE)) {
3300                 bio_get(bio);
3301                 submit_bio(rw, bio);
3302                 bio_put(bio);
3303                 return 0;
3304         }
3305
3306         /*
3307          * nr_async_bios allows us to reliably return congestion to the
3308          * higher layers.  Otherwise, the async bio makes it appear we have
3309          * made progress against dirty pages when we've really just put it
3310          * on a queue for later
3311          */
3312         atomic_inc(&root->fs_info->nr_async_bios);
3313         WARN_ON(bio->bi_next);
3314         bio->bi_next = NULL;
3315         bio->bi_rw |= rw;
3316
3317         spin_lock(&device->io_lock);
3318         if (bio->bi_rw & REQ_SYNC)
3319                 pending_bios = &device->pending_sync_bios;
3320         else
3321                 pending_bios = &device->pending_bios;
3322
3323         if (pending_bios->tail)
3324                 pending_bios->tail->bi_next = bio;
3325
3326         pending_bios->tail = bio;
3327         if (!pending_bios->head)
3328                 pending_bios->head = bio;
3329         if (device->running_pending)
3330                 should_queue = 0;
3331
3332         spin_unlock(&device->io_lock);
3333
3334         if (should_queue)
3335                 btrfs_queue_worker(&root->fs_info->submit_workers,
3336                                    &device->work);
3337         return 0;
3338 }
3339
3340 int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio,
3341                   int mirror_num, int async_submit)
3342 {
3343         struct btrfs_mapping_tree *map_tree;
3344         struct btrfs_device *dev;
3345         struct bio *first_bio = bio;
3346         u64 logical = (u64)bio->bi_sector << 9;
3347         u64 length = 0;
3348         u64 map_length;
3349         int ret;
3350         int dev_nr = 0;
3351         int total_devs = 1;
3352         struct btrfs_bio *bbio = NULL;
3353
3354         length = bio->bi_size;
3355         map_tree = &root->fs_info->mapping_tree;
3356         map_length = length;
3357
3358         ret = btrfs_map_block(map_tree, rw, logical, &map_length, &bbio,
3359                               mirror_num);
3360         BUG_ON(ret);
3361
3362         total_devs = bbio->num_stripes;
3363         if (map_length < length) {
3364                 printk(KERN_CRIT "mapping failed logical %llu bio len %llu "
3365                        "len %llu\n", (unsigned long long)logical,
3366                        (unsigned long long)length,
3367                        (unsigned long long)map_length);
3368                 BUG();
3369         }
3370
3371         bbio->orig_bio = first_bio;
3372         bbio->private = first_bio->bi_private;
3373         bbio->end_io = first_bio->bi_end_io;
3374         atomic_set(&bbio->stripes_pending, bbio->num_stripes);
3375
3376         while (dev_nr < total_devs) {
3377                 if (dev_nr < total_devs - 1) {
3378                         bio = bio_clone(first_bio, GFP_NOFS);
3379                         BUG_ON(!bio);
3380                 } else {
3381                         bio = first_bio;
3382                 }
3383                 bio->bi_private = bbio;
3384                 bio->bi_end_io = btrfs_end_bio;
3385                 bio->bi_sector = bbio->stripes[dev_nr].physical >> 9;
3386                 dev = bbio->stripes[dev_nr].dev;
3387                 if (dev && dev->bdev && (rw != WRITE || dev->writeable)) {
3388                         pr_debug("btrfs_map_bio: rw %d, secor=%llu, dev=%lu "
3389                                  "(%s id %llu), size=%u\n", rw,
3390                                  (u64)bio->bi_sector, (u_long)dev->bdev->bd_dev,
3391                                  dev->name, dev->devid, bio->bi_size);
3392                         bio->bi_bdev = dev->bdev;
3393                         if (async_submit)
3394                                 schedule_bio(root, dev, rw, bio);
3395                         else
3396                                 submit_bio(rw, bio);
3397                 } else {
3398                         bio->bi_bdev = root->fs_info->fs_devices->latest_bdev;
3399                         bio->bi_sector = logical >> 9;
3400                         bio_endio(bio, -EIO);
3401                 }
3402                 dev_nr++;
3403         }
3404         return 0;
3405 }
3406
3407 struct btrfs_device *btrfs_find_device(struct btrfs_root *root, u64 devid,
3408                                        u8 *uuid, u8 *fsid)
3409 {
3410         struct btrfs_device *device;
3411         struct btrfs_fs_devices *cur_devices;
3412
3413         cur_devices = root->fs_info->fs_devices;
3414         while (cur_devices) {
3415                 if (!fsid ||
3416                     !memcmp(cur_devices->fsid, fsid, BTRFS_UUID_SIZE)) {
3417                         device = __find_device(&cur_devices->devices,
3418                                                devid, uuid);
3419                         if (device)
3420                                 return device;
3421                 }
3422                 cur_devices = cur_devices->seed;
3423         }
3424         return NULL;
3425 }
3426
3427 static struct btrfs_device *add_missing_dev(struct btrfs_root *root,
3428                                             u64 devid, u8 *dev_uuid)
3429 {
3430         struct btrfs_device *device;
3431         struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
3432
3433         device = kzalloc(sizeof(*device), GFP_NOFS);
3434         if (!device)
3435                 return NULL;
3436         list_add(&device->dev_list,
3437                  &fs_devices->devices);
3438         device->dev_root = root->fs_info->dev_root;
3439         device->devid = devid;
3440         device->work.func = pending_bios_fn;
3441         device->fs_devices = fs_devices;
3442         device->missing = 1;
3443         fs_devices->num_devices++;
3444         fs_devices->missing_devices++;
3445         spin_lock_init(&device->io_lock);
3446         INIT_LIST_HEAD(&device->dev_alloc_list);
3447         memcpy(device->uuid, dev_uuid, BTRFS_UUID_SIZE);
3448         return device;
3449 }
3450
3451 static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key,
3452                           struct extent_buffer *leaf,
3453                           struct btrfs_chunk *chunk)
3454 {
3455         struct btrfs_mapping_tree *map_tree = &root->fs_info->mapping_tree;
3456         struct map_lookup *map;
3457         struct extent_map *em;
3458         u64 logical;
3459         u64 length;
3460         u64 devid;
3461         u8 uuid[BTRFS_UUID_SIZE];
3462         int num_stripes;
3463         int ret;
3464         int i;
3465
3466         logical = key->offset;
3467         length = btrfs_chunk_length(leaf, chunk);
3468
3469         read_lock(&map_tree->map_tree.lock);
3470         em = lookup_extent_mapping(&map_tree->map_tree, logical, 1);
3471         read_unlock(&map_tree->map_tree.lock);
3472
3473         /* already mapped? */
3474         if (em && em->start <= logical && em->start + em->len > logical) {
3475                 free_extent_map(em);
3476                 return 0;
3477         } else if (em) {
3478                 free_extent_map(em);
3479         }
3480
3481         em = alloc_extent_map();
3482         if (!em)
3483                 return -ENOMEM;
3484         num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
3485         map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
3486         if (!map) {
3487                 free_extent_map(em);
3488                 return -ENOMEM;
3489         }
3490
3491         em->bdev = (struct block_device *)map;
3492         em->start = logical;
3493         em->len = length;
3494         em->block_start = 0;
3495         em->block_len = em->len;
3496
3497         map->num_stripes = num_stripes;
3498         map->io_width = btrfs_chunk_io_width(leaf, chunk);
3499         map->io_align = btrfs_chunk_io_align(leaf, chunk);
3500         map->sector_size = btrfs_chunk_sector_size(leaf, chunk);
3501         map->stripe_len = btrfs_chunk_stripe_len(leaf, chunk);
3502         map->type = btrfs_chunk_type(leaf, chunk);
3503         map->sub_stripes = btrfs_chunk_sub_stripes(leaf, chunk);
3504         for (i = 0; i < num_stripes; i++) {
3505                 map->stripes[i].physical =
3506                         btrfs_stripe_offset_nr(leaf, chunk, i);
3507                 devid = btrfs_stripe_devid_nr(leaf, chunk, i);
3508                 read_extent_buffer(leaf, uuid, (unsigned long)
3509                                    btrfs_stripe_dev_uuid_nr(chunk, i),
3510                                    BTRFS_UUID_SIZE);
3511                 map->stripes[i].dev = btrfs_find_device(root, devid, uuid,
3512                                                         NULL);
3513                 if (!map->stripes[i].dev && !btrfs_test_opt(root, DEGRADED)) {
3514                         kfree(map);
3515                         free_extent_map(em);
3516                         return -EIO;
3517                 }
3518                 if (!map->stripes[i].dev) {
3519                         map->stripes[i].dev =
3520                                 add_missing_dev(root, devid, uuid);
3521                         if (!map->stripes[i].dev) {
3522                                 kfree(map);
3523                                 free_extent_map(em);
3524                                 return -EIO;
3525                         }
3526                 }
3527                 map->stripes[i].dev->in_fs_metadata = 1;
3528         }
3529
3530         write_lock(&map_tree->map_tree.lock);
3531         ret = add_extent_mapping(&map_tree->map_tree, em);
3532         write_unlock(&map_tree->map_tree.lock);
3533         BUG_ON(ret);
3534         free_extent_map(em);
3535
3536         return 0;
3537 }
3538
3539 static int fill_device_from_item(struct extent_buffer *leaf,
3540                                  struct btrfs_dev_item *dev_item,
3541                                  struct btrfs_device *device)
3542 {
3543         unsigned long ptr;
3544
3545         device->devid = btrfs_device_id(leaf, dev_item);
3546         device->disk_total_bytes = btrfs_device_total_bytes(leaf, dev_item);
3547         device->total_bytes = device->disk_total_bytes;
3548         device->bytes_used = btrfs_device_bytes_used(leaf, dev_item);
3549         device->type = btrfs_device_type(leaf, dev_item);
3550         device->io_align = btrfs_device_io_align(leaf, dev_item);
3551         device->io_width = btrfs_device_io_width(leaf, dev_item);
3552         device->sector_size = btrfs_device_sector_size(leaf, dev_item);
3553
3554         ptr = (unsigned long)btrfs_device_uuid(dev_item);
3555         read_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
3556
3557         return 0;
3558 }
3559
3560 static int open_seed_devices(struct btrfs_root *root, u8 *fsid)
3561 {
3562         struct btrfs_fs_devices *fs_devices;
3563         int ret;
3564
3565         mutex_lock(&uuid_mutex);
3566
3567         fs_devices = root->fs_info->fs_devices->seed;
3568         while (fs_devices) {
3569                 if (!memcmp(fs_devices->fsid, fsid, BTRFS_UUID_SIZE)) {
3570                         ret = 0;
3571                         goto out;
3572                 }
3573                 fs_devices = fs_devices->seed;
3574         }
3575
3576         fs_devices = find_fsid(fsid);
3577         if (!fs_devices) {
3578                 ret = -ENOENT;
3579                 goto out;
3580         }
3581
3582         fs_devices = clone_fs_devices(fs_devices);
3583         if (IS_ERR(fs_devices)) {
3584                 ret = PTR_ERR(fs_devices);
3585                 goto out;
3586         }
3587
3588         ret = __btrfs_open_devices(fs_devices, FMODE_READ,
3589                                    root->fs_info->bdev_holder);
3590         if (ret)
3591                 goto out;
3592
3593         if (!fs_devices->seeding) {
3594                 __btrfs_close_devices(fs_devices);
3595                 free_fs_devices(fs_devices);
3596                 ret = -EINVAL;
3597                 goto out;
3598         }
3599
3600         fs_devices->seed = root->fs_info->fs_devices->seed;
3601         root->fs_info->fs_devices->seed = fs_devices;
3602 out:
3603         mutex_unlock(&uuid_mutex);
3604         return ret;
3605 }
3606
3607 static int read_one_dev(struct btrfs_root *root,
3608                         struct extent_buffer *leaf,
3609                         struct btrfs_dev_item *dev_item)
3610 {
3611         struct btrfs_device *device;
3612         u64 devid;
3613         int ret;
3614         u8 fs_uuid[BTRFS_UUID_SIZE];
3615         u8 dev_uuid[BTRFS_UUID_SIZE];
3616
3617         devid = btrfs_device_id(leaf, dev_item);
3618         read_extent_buffer(leaf, dev_uuid,
3619                            (unsigned long)btrfs_device_uuid(dev_item),
3620                            BTRFS_UUID_SIZE);
3621         read_extent_buffer(leaf, fs_uuid,
3622                            (unsigned long)btrfs_device_fsid(dev_item),
3623                            BTRFS_UUID_SIZE);
3624
3625         if (memcmp(fs_uuid, root->fs_info->fsid, BTRFS_UUID_SIZE)) {
3626                 ret = open_seed_devices(root, fs_uuid);
3627                 if (ret && !btrfs_test_opt(root, DEGRADED))
3628                         return ret;
3629         }
3630
3631         device = btrfs_find_device(root, devid, dev_uuid, fs_uuid);
3632         if (!device || !device->bdev) {
3633                 if (!btrfs_test_opt(root, DEGRADED))
3634                         return -EIO;
3635
3636                 if (!device) {
3637                         printk(KERN_WARNING "warning devid %llu missing\n",
3638                                (unsigned long long)devid);
3639                         device = add_missing_dev(root, devid, dev_uuid);
3640                         if (!device)
3641                                 return -ENOMEM;
3642                 } else if (!device->missing) {
3643                         /*
3644                          * this happens when a device that was properly setup
3645                          * in the device info lists suddenly goes bad.
3646                          * device->bdev is NULL, and so we have to set
3647                          * device->missing to one here
3648                          */
3649                         root->fs_info->fs_devices->missing_devices++;
3650                         device->missing = 1;
3651                 }
3652         }
3653
3654         if (device->fs_devices != root->fs_info->fs_devices) {
3655                 BUG_ON(device->writeable);
3656                 if (device->generation !=
3657                     btrfs_device_generation(leaf, dev_item))
3658                         return -EINVAL;
3659         }
3660
3661         fill_device_from_item(leaf, dev_item, device);
3662         device->dev_root = root->fs_info->dev_root;
3663         device->in_fs_metadata = 1;
3664         if (device->writeable) {
3665                 device->fs_devices->total_rw_bytes += device->total_bytes;
3666                 spin_lock(&root->fs_info->free_chunk_lock);
3667                 root->fs_info->free_chunk_space += device->total_bytes -
3668                         device->bytes_used;
3669                 spin_unlock(&root->fs_info->free_chunk_lock);
3670         }
3671         ret = 0;
3672         return ret;
3673 }
3674
3675 int btrfs_read_sys_array(struct btrfs_root *root)
3676 {
3677         struct btrfs_super_block *super_copy = root->fs_info->super_copy;
3678         struct extent_buffer *sb;
3679         struct btrfs_disk_key *disk_key;
3680         struct btrfs_chunk *chunk;
3681         u8 *ptr;
3682         unsigned long sb_ptr;
3683         int ret = 0;
3684         u32 num_stripes;
3685         u32 array_size;
3686         u32 len = 0;
3687         u32 cur;
3688         struct btrfs_key key;
3689
3690         sb = btrfs_find_create_tree_block(root, BTRFS_SUPER_INFO_OFFSET,
3691                                           BTRFS_SUPER_INFO_SIZE);
3692         if (!sb)
3693                 return -ENOMEM;
3694         btrfs_set_buffer_uptodate(sb);
3695         btrfs_set_buffer_lockdep_class(root->root_key.objectid, sb, 0);
3696
3697         write_extent_buffer(sb, super_copy, 0, BTRFS_SUPER_INFO_SIZE);
3698         array_size = btrfs_super_sys_array_size(super_copy);
3699
3700         ptr = super_copy->sys_chunk_array;
3701         sb_ptr = offsetof(struct btrfs_super_block, sys_chunk_array);
3702         cur = 0;
3703
3704         while (cur < array_size) {
3705                 disk_key = (struct btrfs_disk_key *)ptr;
3706                 btrfs_disk_key_to_cpu(&key, disk_key);
3707
3708                 len = sizeof(*disk_key); ptr += len;
3709                 sb_ptr += len;
3710                 cur += len;
3711
3712                 if (key.type == BTRFS_CHUNK_ITEM_KEY) {
3713                         chunk = (struct btrfs_chunk *)sb_ptr;
3714                         ret = read_one_chunk(root, &key, sb, chunk);
3715                         if (ret)
3716                                 break;
3717                         num_stripes = btrfs_chunk_num_stripes(sb, chunk);
3718                         len = btrfs_chunk_item_size(num_stripes);
3719                 } else {
3720                         ret = -EIO;
3721                         break;
3722                 }
3723                 ptr += len;
3724                 sb_ptr += len;
3725                 cur += len;
3726         }
3727         free_extent_buffer(sb);
3728         return ret;
3729 }
3730
3731 int btrfs_read_chunk_tree(struct btrfs_root *root)
3732 {
3733         struct btrfs_path *path;
3734         struct extent_buffer *leaf;
3735         struct btrfs_key key;
3736         struct btrfs_key found_key;
3737         int ret;
3738         int slot;
3739
3740         root = root->fs_info->chunk_root;
3741
3742         path = btrfs_alloc_path();
3743         if (!path)
3744                 return -ENOMEM;
3745
3746         /* first we search for all of the device items, and then we
3747          * read in all of the chunk items.  This way we can create chunk
3748          * mappings that reference all of the devices that are afound
3749          */
3750         key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
3751         key.offset = 0;
3752         key.type = 0;
3753 again:
3754         ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
3755         if (ret < 0)
3756                 goto error;
3757         while (1) {
3758                 leaf = path->nodes[0];
3759                 slot = path->slots[0];
3760                 if (slot >= btrfs_header_nritems(leaf)) {
3761                         ret = btrfs_next_leaf(root, path);
3762                         if (ret == 0)
3763                                 continue;
3764                         if (ret < 0)
3765                                 goto error;
3766                         break;
3767                 }
3768                 btrfs_item_key_to_cpu(leaf, &found_key, slot);
3769                 if (key.objectid == BTRFS_DEV_ITEMS_OBJECTID) {
3770                         if (found_key.objectid != BTRFS_DEV_ITEMS_OBJECTID)
3771                                 break;
3772                         if (found_key.type == BTRFS_DEV_ITEM_KEY) {
3773                                 struct btrfs_dev_item *dev_item;
3774                                 dev_item = btrfs_item_ptr(leaf, slot,
3775                                                   struct btrfs_dev_item);
3776                                 ret = read_one_dev(root, leaf, dev_item);
3777                                 if (ret)
3778                                         goto error;
3779                         }
3780                 } else if (found_key.type == BTRFS_CHUNK_ITEM_KEY) {
3781                         struct btrfs_chunk *chunk;
3782                         chunk = btrfs_item_ptr(leaf, slot, struct btrfs_chunk);
3783                         ret = read_one_chunk(root, &found_key, leaf, chunk);
3784                         if (ret)
3785                                 goto error;
3786                 }
3787                 path->slots[0]++;
3788         }
3789         if (key.objectid == BTRFS_DEV_ITEMS_OBJECTID) {
3790                 key.objectid = 0;
3791                 btrfs_release_path(path);
3792                 goto again;
3793         }
3794         ret = 0;
3795 error:
3796         btrfs_free_path(path);
3797         return ret;
3798 }