Blame view
fs/btrfs/qgroup.c
107 KB
c1d7c514f
|
1 |
// SPDX-License-Identifier: GPL-2.0 |
bed92eae2
|
2 3 |
/* * Copyright (C) 2011 STRATO. All rights reserved. |
bed92eae2
|
4 5 6 7 8 9 10 11 12 |
*/ #include <linux/sched.h> #include <linux/pagemap.h> #include <linux/writeback.h> #include <linux/blkdev.h> #include <linux/rbtree.h> #include <linux/slab.h> #include <linux/workqueue.h> |
55e301fd5
|
13 |
#include <linux/btrfs.h> |
bed92eae2
|
14 15 16 17 18 19 |
#include "ctree.h" #include "transaction.h" #include "disk-io.h" #include "locking.h" #include "ulist.h" |
bed92eae2
|
20 |
#include "backref.h" |
2f2320360
|
21 |
#include "extent_io.h" |
fcebe4562
|
22 |
#include "qgroup.h" |
aac0023c2
|
23 |
#include "block-group.h" |
49e5fb462
|
24 |
#include "sysfs.h" |
e69bcee37
|
25 |
|
bed92eae2
|
26 27 28 29 30 |
/* TODO XXX FIXME * - subvol delete -> delete when ref goes to 0? delete limits also? * - reorganize keys * - compressed * - sync |
bed92eae2
|
31 32 |
* - copy also limits on subvol creation * - limit |
52042d8e8
|
33 |
* - caches for ulists |
bed92eae2
|
34 35 36 |
* - performance benchmarks * - check all ioctl parameters */ |
f59c0347d
|
37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 |
/* * Helpers to access qgroup reservation * * Callers should ensure the lock context and type are valid */ static u64 qgroup_rsv_total(const struct btrfs_qgroup *qgroup) { u64 ret = 0; int i; for (i = 0; i < BTRFS_QGROUP_RSV_LAST; i++) ret += qgroup->rsv.values[i]; return ret; } #ifdef CONFIG_BTRFS_DEBUG static const char *qgroup_rsv_type_str(enum btrfs_qgroup_rsv_type type) { if (type == BTRFS_QGROUP_RSV_DATA) return "data"; |
733e03a0b
|
59 60 61 62 |
if (type == BTRFS_QGROUP_RSV_META_PERTRANS) return "meta_pertrans"; if (type == BTRFS_QGROUP_RSV_META_PREALLOC) return "meta_prealloc"; |
f59c0347d
|
63 64 65 |
return NULL; } #endif |
64ee4e751
|
66 67 |
static void qgroup_rsv_add(struct btrfs_fs_info *fs_info, struct btrfs_qgroup *qgroup, u64 num_bytes, |
f59c0347d
|
68 69 |
enum btrfs_qgroup_rsv_type type) { |
64ee4e751
|
70 |
trace_qgroup_update_reserve(fs_info, qgroup, num_bytes, type); |
f59c0347d
|
71 72 |
qgroup->rsv.values[type] += num_bytes; } |
64ee4e751
|
73 74 |
static void qgroup_rsv_release(struct btrfs_fs_info *fs_info, struct btrfs_qgroup *qgroup, u64 num_bytes, |
f59c0347d
|
75 76 |
enum btrfs_qgroup_rsv_type type) { |
64ee4e751
|
77 |
trace_qgroup_update_reserve(fs_info, qgroup, -(s64)num_bytes, type); |
f59c0347d
|
78 79 80 81 82 83 84 85 86 87 88 89 |
if (qgroup->rsv.values[type] >= num_bytes) { qgroup->rsv.values[type] -= num_bytes; return; } #ifdef CONFIG_BTRFS_DEBUG WARN_RATELIMIT(1, "qgroup %llu %s reserved space underflow, have %llu to free %llu", qgroup->qgroupid, qgroup_rsv_type_str(type), qgroup->rsv.values[type], num_bytes); #endif qgroup->rsv.values[type] = 0; } |
64ee4e751
|
90 91 92 |
static void qgroup_rsv_add_by_qgroup(struct btrfs_fs_info *fs_info, struct btrfs_qgroup *dest, struct btrfs_qgroup *src) |
f59c0347d
|
93 94 95 96 |
{ int i; for (i = 0; i < BTRFS_QGROUP_RSV_LAST; i++) |
64ee4e751
|
97 |
qgroup_rsv_add(fs_info, dest, src->rsv.values[i], i); |
f59c0347d
|
98 |
} |
64ee4e751
|
99 100 |
static void qgroup_rsv_release_by_qgroup(struct btrfs_fs_info *fs_info, struct btrfs_qgroup *dest, |
f59c0347d
|
101 102 103 104 105 |
struct btrfs_qgroup *src) { int i; for (i = 0; i < BTRFS_QGROUP_RSV_LAST; i++) |
64ee4e751
|
106 |
qgroup_rsv_release(fs_info, dest, src->rsv.values[i], i); |
f59c0347d
|
107 |
} |
9c542136f
|
108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 |
static void btrfs_qgroup_update_old_refcnt(struct btrfs_qgroup *qg, u64 seq, int mod) { if (qg->old_refcnt < seq) qg->old_refcnt = seq; qg->old_refcnt += mod; } static void btrfs_qgroup_update_new_refcnt(struct btrfs_qgroup *qg, u64 seq, int mod) { if (qg->new_refcnt < seq) qg->new_refcnt = seq; qg->new_refcnt += mod; } static inline u64 btrfs_qgroup_get_old_refcnt(struct btrfs_qgroup *qg, u64 seq) { if (qg->old_refcnt < seq) return 0; return qg->old_refcnt - seq; } static inline u64 btrfs_qgroup_get_new_refcnt(struct btrfs_qgroup *qg, u64 seq) { if (qg->new_refcnt < seq) return 0; return qg->new_refcnt - seq; } |
bed92eae2
|
137 138 139 140 141 142 143 144 145 |
/* * glue structure to represent the relations between qgroups. */ struct btrfs_qgroup_list { struct list_head next_group; struct list_head next_member; struct btrfs_qgroup *group; struct btrfs_qgroup *member; }; |
ef2fff64f
|
146 147 148 149 150 151 152 153 154 |
static inline u64 qgroup_to_aux(struct btrfs_qgroup *qg) { return (u64)(uintptr_t)qg; } static inline struct btrfs_qgroup* unode_aux_to_qgroup(struct ulist_node *n) { return (struct btrfs_qgroup *)(uintptr_t)n->aux; } |
fcebe4562
|
155 |
|
b382a324b
|
156 157 158 159 |
static int qgroup_rescan_init(struct btrfs_fs_info *fs_info, u64 progress_objectid, int init_flags); static void qgroup_rescan_zero_tracking(struct btrfs_fs_info *fs_info); |
2f2320360
|
160 |
|
58400fce5
|
161 |
/* must be called with qgroup_ioctl_lock held */ |
bed92eae2
|
162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 |
static struct btrfs_qgroup *find_qgroup_rb(struct btrfs_fs_info *fs_info, u64 qgroupid) { struct rb_node *n = fs_info->qgroup_tree.rb_node; struct btrfs_qgroup *qgroup; while (n) { qgroup = rb_entry(n, struct btrfs_qgroup, node); if (qgroup->qgroupid < qgroupid) n = n->rb_left; else if (qgroup->qgroupid > qgroupid) n = n->rb_right; else return qgroup; } return NULL; } /* must be called with qgroup_lock held */ static struct btrfs_qgroup *add_qgroup_rb(struct btrfs_fs_info *fs_info, u64 qgroupid) { struct rb_node **p = &fs_info->qgroup_tree.rb_node; struct rb_node *parent = NULL; struct btrfs_qgroup *qgroup; while (*p) { parent = *p; qgroup = rb_entry(parent, struct btrfs_qgroup, node); if (qgroup->qgroupid < qgroupid) p = &(*p)->rb_left; else if (qgroup->qgroupid > qgroupid) p = &(*p)->rb_right; else return qgroup; } qgroup = kzalloc(sizeof(*qgroup), GFP_ATOMIC); if (!qgroup) return ERR_PTR(-ENOMEM); qgroup->qgroupid = qgroupid; INIT_LIST_HEAD(&qgroup->groups); INIT_LIST_HEAD(&qgroup->members); INIT_LIST_HEAD(&qgroup->dirty); rb_link_node(&qgroup->node, parent, p); rb_insert_color(&qgroup->node, &fs_info->qgroup_tree); return qgroup; } |
49e5fb462
|
214 215 |
static void __del_qgroup_rb(struct btrfs_fs_info *fs_info, struct btrfs_qgroup *qgroup) |
bed92eae2
|
216 |
{ |
bed92eae2
|
217 |
struct btrfs_qgroup_list *list; |
49e5fb462
|
218 |
btrfs_sysfs_del_one_qgroup(fs_info, qgroup); |
bed92eae2
|
219 |
list_del(&qgroup->dirty); |
bed92eae2
|
220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 |
while (!list_empty(&qgroup->groups)) { list = list_first_entry(&qgroup->groups, struct btrfs_qgroup_list, next_group); list_del(&list->next_group); list_del(&list->next_member); kfree(list); } while (!list_empty(&qgroup->members)) { list = list_first_entry(&qgroup->members, struct btrfs_qgroup_list, next_member); list_del(&list->next_group); list_del(&list->next_member); kfree(list); } kfree(qgroup); |
4082bd3d7
|
236 |
} |
bed92eae2
|
237 |
|
4082bd3d7
|
238 239 240 241 242 243 244 245 246 |
/* must be called with qgroup_lock held */ static int del_qgroup_rb(struct btrfs_fs_info *fs_info, u64 qgroupid) { struct btrfs_qgroup *qgroup = find_qgroup_rb(fs_info, qgroupid); if (!qgroup) return -ENOENT; rb_erase(&qgroup->node, &fs_info->qgroup_tree); |
49e5fb462
|
247 |
__del_qgroup_rb(fs_info, qgroup); |
bed92eae2
|
248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 |
return 0; } /* must be called with qgroup_lock held */ static int add_relation_rb(struct btrfs_fs_info *fs_info, u64 memberid, u64 parentid) { struct btrfs_qgroup *member; struct btrfs_qgroup *parent; struct btrfs_qgroup_list *list; member = find_qgroup_rb(fs_info, memberid); parent = find_qgroup_rb(fs_info, parentid); if (!member || !parent) return -ENOENT; list = kzalloc(sizeof(*list), GFP_ATOMIC); if (!list) return -ENOMEM; list->group = parent; list->member = member; list_add_tail(&list->next_group, &member->groups); list_add_tail(&list->next_member, &parent->members); return 0; } /* must be called with qgroup_lock held */ static int del_relation_rb(struct btrfs_fs_info *fs_info, u64 memberid, u64 parentid) { struct btrfs_qgroup *member; struct btrfs_qgroup *parent; struct btrfs_qgroup_list *list; member = find_qgroup_rb(fs_info, memberid); parent = find_qgroup_rb(fs_info, parentid); if (!member || !parent) return -ENOENT; list_for_each_entry(list, &member->groups, next_group) { if (list->group == parent) { list_del(&list->next_group); list_del(&list->next_member); kfree(list); return 0; } } return -ENOENT; } |
faa2dbf00
|
299 300 301 302 303 304 305 306 307 308 309 310 311 312 |
#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS int btrfs_verify_qgroup_counts(struct btrfs_fs_info *fs_info, u64 qgroupid, u64 rfer, u64 excl) { struct btrfs_qgroup *qgroup; qgroup = find_qgroup_rb(fs_info, qgroupid); if (!qgroup) return -EINVAL; if (qgroup->rfer != rfer || qgroup->excl != excl) return -EINVAL; return 0; } #endif |
bed92eae2
|
313 314 315 316 317 318 319 320 321 322 323 324 325 326 |
/* * The full config is read in one go, only called from open_ctree() * It doesn't use any locking, as at this point we're still single-threaded */ int btrfs_read_qgroup_config(struct btrfs_fs_info *fs_info) { struct btrfs_key key; struct btrfs_key found_key; struct btrfs_root *quota_root = fs_info->quota_root; struct btrfs_path *path = NULL; struct extent_buffer *l; int slot; int ret = 0; u64 flags = 0; |
b382a324b
|
327 |
u64 rescan_progress = 0; |
bed92eae2
|
328 |
|
afcdd129e
|
329 |
if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) |
bed92eae2
|
330 |
return 0; |
323b88f4a
|
331 |
fs_info->qgroup_ulist = ulist_alloc(GFP_KERNEL); |
1e8f91586
|
332 333 334 335 |
if (!fs_info->qgroup_ulist) { ret = -ENOMEM; goto out; } |
bed92eae2
|
336 337 338 339 340 |
path = btrfs_alloc_path(); if (!path) { ret = -ENOMEM; goto out; } |
49e5fb462
|
341 342 343 |
ret = btrfs_sysfs_add_qgroups(fs_info); if (ret < 0) goto out; |
bed92eae2
|
344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 |
/* default this to quota off, in case no status key is found */ fs_info->qgroup_flags = 0; /* * pass 1: read status, all qgroup infos and limits */ key.objectid = 0; key.type = 0; key.offset = 0; ret = btrfs_search_slot_for_read(quota_root, &key, path, 1, 1); if (ret) goto out; while (1) { struct btrfs_qgroup *qgroup; slot = path->slots[0]; l = path->nodes[0]; btrfs_item_key_to_cpu(l, &found_key, slot); if (found_key.type == BTRFS_QGROUP_STATUS_KEY) { struct btrfs_qgroup_status_item *ptr; ptr = btrfs_item_ptr(l, slot, struct btrfs_qgroup_status_item); if (btrfs_qgroup_status_version(l, ptr) != BTRFS_QGROUP_STATUS_VERSION) { |
efe120a06
|
372 373 |
btrfs_err(fs_info, "old qgroup version, quota disabled"); |
bed92eae2
|
374 375 376 377 378 |
goto out; } if (btrfs_qgroup_status_generation(l, ptr) != fs_info->generation) { flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; |
efe120a06
|
379 |
btrfs_err(fs_info, |
5d163e0e6
|
380 |
"qgroup generation mismatch, marked as inconsistent"); |
bed92eae2
|
381 382 383 |
} fs_info->qgroup_flags = btrfs_qgroup_status_flags(l, ptr); |
b382a324b
|
384 |
rescan_progress = btrfs_qgroup_status_rescan(l, ptr); |
bed92eae2
|
385 386 387 388 389 390 391 392 393 394 |
goto next1; } if (found_key.type != BTRFS_QGROUP_INFO_KEY && found_key.type != BTRFS_QGROUP_LIMIT_KEY) goto next1; qgroup = find_qgroup_rb(fs_info, found_key.offset); if ((qgroup && found_key.type == BTRFS_QGROUP_INFO_KEY) || (!qgroup && found_key.type == BTRFS_QGROUP_LIMIT_KEY)) { |
d41e36a0a
|
395 |
btrfs_err(fs_info, "inconsistent qgroup config"); |
bed92eae2
|
396 397 398 399 400 401 402 403 404 |
flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; } if (!qgroup) { qgroup = add_qgroup_rb(fs_info, found_key.offset); if (IS_ERR(qgroup)) { ret = PTR_ERR(qgroup); goto out; } } |
49e5fb462
|
405 406 407 |
ret = btrfs_sysfs_add_one_qgroup(fs_info, qgroup); if (ret < 0) goto out; |
bed92eae2
|
408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 |
switch (found_key.type) { case BTRFS_QGROUP_INFO_KEY: { struct btrfs_qgroup_info_item *ptr; ptr = btrfs_item_ptr(l, slot, struct btrfs_qgroup_info_item); qgroup->rfer = btrfs_qgroup_info_rfer(l, ptr); qgroup->rfer_cmpr = btrfs_qgroup_info_rfer_cmpr(l, ptr); qgroup->excl = btrfs_qgroup_info_excl(l, ptr); qgroup->excl_cmpr = btrfs_qgroup_info_excl_cmpr(l, ptr); /* generation currently unused */ break; } case BTRFS_QGROUP_LIMIT_KEY: { struct btrfs_qgroup_limit_item *ptr; ptr = btrfs_item_ptr(l, slot, struct btrfs_qgroup_limit_item); qgroup->lim_flags = btrfs_qgroup_limit_flags(l, ptr); qgroup->max_rfer = btrfs_qgroup_limit_max_rfer(l, ptr); qgroup->max_excl = btrfs_qgroup_limit_max_excl(l, ptr); qgroup->rsv_rfer = btrfs_qgroup_limit_rsv_rfer(l, ptr); qgroup->rsv_excl = btrfs_qgroup_limit_rsv_excl(l, ptr); break; } } next1: ret = btrfs_next_item(quota_root, path); if (ret < 0) goto out; if (ret) break; } btrfs_release_path(path); /* * pass 2: read all qgroup relations */ key.objectid = 0; key.type = BTRFS_QGROUP_RELATION_KEY; key.offset = 0; ret = btrfs_search_slot_for_read(quota_root, &key, path, 1, 0); if (ret) goto out; while (1) { slot = path->slots[0]; l = path->nodes[0]; btrfs_item_key_to_cpu(l, &found_key, slot); if (found_key.type != BTRFS_QGROUP_RELATION_KEY) goto next2; if (found_key.objectid > found_key.offset) { /* parent <- member, not needed to build config */ /* FIXME should we omit the key completely? */ goto next2; } ret = add_relation_rb(fs_info, found_key.objectid, found_key.offset); |
ff24858c6
|
468 |
if (ret == -ENOENT) { |
efe120a06
|
469 470 |
btrfs_warn(fs_info, "orphan qgroup relation 0x%llx->0x%llx", |
c1c9ff7c9
|
471 |
found_key.objectid, found_key.offset); |
ff24858c6
|
472 473 |
ret = 0; /* ignore the error */ } |
bed92eae2
|
474 475 476 477 478 479 480 481 482 483 484 |
if (ret) goto out; next2: ret = btrfs_next_item(quota_root, path); if (ret < 0) goto out; if (ret) break; } out: fs_info->qgroup_flags |= flags; |
afcdd129e
|
485 486 487 488 |
if (!(fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_ON)) clear_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags); else if (fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN && ret >= 0) |
b382a324b
|
489 |
ret = qgroup_rescan_init(fs_info, rescan_progress, 0); |
bed92eae2
|
490 |
btrfs_free_path(path); |
eb1716af8
|
491 |
if (ret < 0) { |
1e8f91586
|
492 |
ulist_free(fs_info->qgroup_ulist); |
eb1716af8
|
493 |
fs_info->qgroup_ulist = NULL; |
b382a324b
|
494 |
fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_RESCAN; |
49e5fb462
|
495 |
btrfs_sysfs_del_qgroups(fs_info); |
eb1716af8
|
496 |
} |
1e8f91586
|
497 |
|
bed92eae2
|
498 499 |
return ret < 0 ? ret : 0; } |
5958253cf
|
500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 |
/* * Called in close_ctree() when quota is still enabled. This verifies we don't * leak some reserved space. * * Return false if no reserved space is left. * Return true if some reserved space is leaked. */ bool btrfs_check_quota_leak(struct btrfs_fs_info *fs_info) { struct rb_node *node; bool ret = false; if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) return ret; /* * Since we're unmounting, there is no race and no need to grab qgroup * lock. And here we don't go post-order to provide a more user * friendly sorted result. */ for (node = rb_first(&fs_info->qgroup_tree); node; node = rb_next(node)) { struct btrfs_qgroup *qgroup; int i; qgroup = rb_entry(node, struct btrfs_qgroup, node); for (i = 0; i < BTRFS_QGROUP_RSV_LAST; i++) { if (qgroup->rsv.values[i]) { ret = true; btrfs_warn(fs_info, |
06f67c470
|
528 |
"qgroup %hu/%llu has unreleased space, type %d rsv %llu", |
5958253cf
|
529 530 531 532 533 534 535 536 |
btrfs_qgroup_level(qgroup->qgroupid), btrfs_qgroup_subvolid(qgroup->qgroupid), i, qgroup->rsv.values[i]); } } } return ret; } |
bed92eae2
|
537 |
/* |
e685da14a
|
538 539 540 541 |
* This is called from close_ctree() or open_ctree() or btrfs_quota_disable(), * first two are in single-threaded paths.And for the third one, we have set * quota_root to be null with qgroup_lock held before, so it is safe to clean * up the in-memory structures without qgroup_lock held. |
bed92eae2
|
542 543 544 545 546 |
*/ void btrfs_free_qgroup_config(struct btrfs_fs_info *fs_info) { struct rb_node *n; struct btrfs_qgroup *qgroup; |
bed92eae2
|
547 548 549 550 |
while ((n = rb_first(&fs_info->qgroup_tree))) { qgroup = rb_entry(n, struct btrfs_qgroup, node); rb_erase(n, &fs_info->qgroup_tree); |
49e5fb462
|
551 |
__del_qgroup_rb(fs_info, qgroup); |
bed92eae2
|
552 |
} |
1e7bac1ef
|
553 |
/* |
52042d8e8
|
554 |
* We call btrfs_free_qgroup_config() when unmounting |
013276101
|
555 |
* filesystem and disabling quota, so we set qgroup_ulist |
1e7bac1ef
|
556 557 |
* to be null here to avoid double free. */ |
1e8f91586
|
558 |
ulist_free(fs_info->qgroup_ulist); |
1e7bac1ef
|
559 |
fs_info->qgroup_ulist = NULL; |
49e5fb462
|
560 |
btrfs_sysfs_del_qgroups(fs_info); |
bed92eae2
|
561 |
} |
711169c40
|
562 563 |
static int add_qgroup_relation_item(struct btrfs_trans_handle *trans, u64 src, u64 dst) |
bed92eae2
|
564 565 |
{ int ret; |
711169c40
|
566 |
struct btrfs_root *quota_root = trans->fs_info->quota_root; |
bed92eae2
|
567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 |
struct btrfs_path *path; struct btrfs_key key; path = btrfs_alloc_path(); if (!path) return -ENOMEM; key.objectid = src; key.type = BTRFS_QGROUP_RELATION_KEY; key.offset = dst; ret = btrfs_insert_empty_item(trans, quota_root, path, &key, 0); btrfs_mark_buffer_dirty(path->nodes[0]); btrfs_free_path(path); return ret; } |
99d7f09ac
|
585 586 |
static int del_qgroup_relation_item(struct btrfs_trans_handle *trans, u64 src, u64 dst) |
bed92eae2
|
587 588 |
{ int ret; |
99d7f09ac
|
589 |
struct btrfs_root *quota_root = trans->fs_info->quota_root; |
bed92eae2
|
590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 |
struct btrfs_path *path; struct btrfs_key key; path = btrfs_alloc_path(); if (!path) return -ENOMEM; key.objectid = src; key.type = BTRFS_QGROUP_RELATION_KEY; key.offset = dst; ret = btrfs_search_slot(trans, quota_root, &key, path, -1, 1); if (ret < 0) goto out; if (ret > 0) { ret = -ENOENT; goto out; } ret = btrfs_del_item(trans, quota_root, path); out: btrfs_free_path(path); return ret; } static int add_qgroup_item(struct btrfs_trans_handle *trans, struct btrfs_root *quota_root, u64 qgroupid) { int ret; struct btrfs_path *path; struct btrfs_qgroup_info_item *qgroup_info; struct btrfs_qgroup_limit_item *qgroup_limit; struct extent_buffer *leaf; struct btrfs_key key; |
f5ee5c9ac
|
625 |
if (btrfs_is_testing(quota_root->fs_info)) |
faa2dbf00
|
626 |
return 0; |
fccb84c94
|
627 |
|
bed92eae2
|
628 629 630 631 632 633 634 |
path = btrfs_alloc_path(); if (!path) return -ENOMEM; key.objectid = 0; key.type = BTRFS_QGROUP_INFO_KEY; key.offset = qgroupid; |
0b4699dcb
|
635 636 637 638 639 |
/* * Avoid a transaction abort by catching -EEXIST here. In that * case, we proceed by re-initializing the existing structure * on disk. */ |
bed92eae2
|
640 641 |
ret = btrfs_insert_empty_item(trans, quota_root, path, &key, sizeof(*qgroup_info)); |
0b4699dcb
|
642 |
if (ret && ret != -EEXIST) |
bed92eae2
|
643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 |
goto out; leaf = path->nodes[0]; qgroup_info = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_qgroup_info_item); btrfs_set_qgroup_info_generation(leaf, qgroup_info, trans->transid); btrfs_set_qgroup_info_rfer(leaf, qgroup_info, 0); btrfs_set_qgroup_info_rfer_cmpr(leaf, qgroup_info, 0); btrfs_set_qgroup_info_excl(leaf, qgroup_info, 0); btrfs_set_qgroup_info_excl_cmpr(leaf, qgroup_info, 0); btrfs_mark_buffer_dirty(leaf); btrfs_release_path(path); key.type = BTRFS_QGROUP_LIMIT_KEY; ret = btrfs_insert_empty_item(trans, quota_root, path, &key, sizeof(*qgroup_limit)); |
0b4699dcb
|
661 |
if (ret && ret != -EEXIST) |
bed92eae2
|
662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 |
goto out; leaf = path->nodes[0]; qgroup_limit = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_qgroup_limit_item); btrfs_set_qgroup_limit_flags(leaf, qgroup_limit, 0); btrfs_set_qgroup_limit_max_rfer(leaf, qgroup_limit, 0); btrfs_set_qgroup_limit_max_excl(leaf, qgroup_limit, 0); btrfs_set_qgroup_limit_rsv_rfer(leaf, qgroup_limit, 0); btrfs_set_qgroup_limit_rsv_excl(leaf, qgroup_limit, 0); btrfs_mark_buffer_dirty(leaf); ret = 0; out: btrfs_free_path(path); return ret; } |
69104618f
|
680 |
static int del_qgroup_item(struct btrfs_trans_handle *trans, u64 qgroupid) |
bed92eae2
|
681 682 |
{ int ret; |
69104618f
|
683 |
struct btrfs_root *quota_root = trans->fs_info->quota_root; |
bed92eae2
|
684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 |
struct btrfs_path *path; struct btrfs_key key; path = btrfs_alloc_path(); if (!path) return -ENOMEM; key.objectid = 0; key.type = BTRFS_QGROUP_INFO_KEY; key.offset = qgroupid; ret = btrfs_search_slot(trans, quota_root, &key, path, -1, 1); if (ret < 0) goto out; if (ret > 0) { ret = -ENOENT; goto out; } ret = btrfs_del_item(trans, quota_root, path); if (ret) goto out; btrfs_release_path(path); key.type = BTRFS_QGROUP_LIMIT_KEY; ret = btrfs_search_slot(trans, quota_root, &key, path, -1, 1); if (ret < 0) goto out; if (ret > 0) { ret = -ENOENT; goto out; } ret = btrfs_del_item(trans, quota_root, path); out: btrfs_free_path(path); return ret; } static int update_qgroup_limit_item(struct btrfs_trans_handle *trans, |
1510e71c6
|
727 |
struct btrfs_qgroup *qgroup) |
bed92eae2
|
728 |
{ |
ac8a866af
|
729 |
struct btrfs_root *quota_root = trans->fs_info->quota_root; |
bed92eae2
|
730 731 732 733 734 735 736 737 738 |
struct btrfs_path *path; struct btrfs_key key; struct extent_buffer *l; struct btrfs_qgroup_limit_item *qgroup_limit; int ret; int slot; key.objectid = 0; key.type = BTRFS_QGROUP_LIMIT_KEY; |
1510e71c6
|
739 |
key.offset = qgroup->qgroupid; |
bed92eae2
|
740 741 |
path = btrfs_alloc_path(); |
84cbe2f72
|
742 743 |
if (!path) return -ENOMEM; |
ac8a866af
|
744 |
ret = btrfs_search_slot(trans, quota_root, &key, path, 0, 1); |
bed92eae2
|
745 746 747 748 749 750 751 752 |
if (ret > 0) ret = -ENOENT; if (ret) goto out; l = path->nodes[0]; slot = path->slots[0]; |
a3df41ee3
|
753 |
qgroup_limit = btrfs_item_ptr(l, slot, struct btrfs_qgroup_limit_item); |
1510e71c6
|
754 755 756 757 758 |
btrfs_set_qgroup_limit_flags(l, qgroup_limit, qgroup->lim_flags); btrfs_set_qgroup_limit_max_rfer(l, qgroup_limit, qgroup->max_rfer); btrfs_set_qgroup_limit_max_excl(l, qgroup_limit, qgroup->max_excl); btrfs_set_qgroup_limit_rsv_rfer(l, qgroup_limit, qgroup->rsv_rfer); btrfs_set_qgroup_limit_rsv_excl(l, qgroup_limit, qgroup->rsv_excl); |
bed92eae2
|
759 760 761 762 763 764 765 766 767 |
btrfs_mark_buffer_dirty(l); out: btrfs_free_path(path); return ret; } static int update_qgroup_info_item(struct btrfs_trans_handle *trans, |
bed92eae2
|
768 769 |
struct btrfs_qgroup *qgroup) { |
3e07e9a09
|
770 771 |
struct btrfs_fs_info *fs_info = trans->fs_info; struct btrfs_root *quota_root = fs_info->quota_root; |
bed92eae2
|
772 773 774 775 776 777 |
struct btrfs_path *path; struct btrfs_key key; struct extent_buffer *l; struct btrfs_qgroup_info_item *qgroup_info; int ret; int slot; |
3e07e9a09
|
778 |
if (btrfs_is_testing(fs_info)) |
faa2dbf00
|
779 |
return 0; |
fccb84c94
|
780 |
|
bed92eae2
|
781 782 783 784 785 |
key.objectid = 0; key.type = BTRFS_QGROUP_INFO_KEY; key.offset = qgroup->qgroupid; path = btrfs_alloc_path(); |
84cbe2f72
|
786 787 |
if (!path) return -ENOMEM; |
3e07e9a09
|
788 |
ret = btrfs_search_slot(trans, quota_root, &key, path, 0, 1); |
bed92eae2
|
789 790 791 792 793 794 795 796 |
if (ret > 0) ret = -ENOENT; if (ret) goto out; l = path->nodes[0]; slot = path->slots[0]; |
a3df41ee3
|
797 |
qgroup_info = btrfs_item_ptr(l, slot, struct btrfs_qgroup_info_item); |
bed92eae2
|
798 799 800 801 802 803 804 805 806 807 808 809 |
btrfs_set_qgroup_info_generation(l, qgroup_info, trans->transid); btrfs_set_qgroup_info_rfer(l, qgroup_info, qgroup->rfer); btrfs_set_qgroup_info_rfer_cmpr(l, qgroup_info, qgroup->rfer_cmpr); btrfs_set_qgroup_info_excl(l, qgroup_info, qgroup->excl); btrfs_set_qgroup_info_excl_cmpr(l, qgroup_info, qgroup->excl_cmpr); btrfs_mark_buffer_dirty(l); out: btrfs_free_path(path); return ret; } |
2e980acdd
|
810 |
static int update_qgroup_status_item(struct btrfs_trans_handle *trans) |
bed92eae2
|
811 |
{ |
2e980acdd
|
812 813 |
struct btrfs_fs_info *fs_info = trans->fs_info; struct btrfs_root *quota_root = fs_info->quota_root; |
bed92eae2
|
814 815 816 817 818 819 820 821 822 823 824 825 |
struct btrfs_path *path; struct btrfs_key key; struct extent_buffer *l; struct btrfs_qgroup_status_item *ptr; int ret; int slot; key.objectid = 0; key.type = BTRFS_QGROUP_STATUS_KEY; key.offset = 0; path = btrfs_alloc_path(); |
84cbe2f72
|
826 827 |
if (!path) return -ENOMEM; |
2e980acdd
|
828 |
ret = btrfs_search_slot(trans, quota_root, &key, path, 0, 1); |
bed92eae2
|
829 830 831 832 833 834 835 836 837 838 839 |
if (ret > 0) ret = -ENOENT; if (ret) goto out; l = path->nodes[0]; slot = path->slots[0]; ptr = btrfs_item_ptr(l, slot, struct btrfs_qgroup_status_item); btrfs_set_qgroup_status_flags(l, ptr, fs_info->qgroup_flags); btrfs_set_qgroup_status_generation(l, ptr, trans->transid); |
2f2320360
|
840 841 |
btrfs_set_qgroup_status_rescan(l, ptr, fs_info->qgroup_rescan_progress.objectid); |
bed92eae2
|
842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 |
btrfs_mark_buffer_dirty(l); out: btrfs_free_path(path); return ret; } /* * called with qgroup_lock held */ static int btrfs_clean_quota_tree(struct btrfs_trans_handle *trans, struct btrfs_root *root) { struct btrfs_path *path; struct btrfs_key key; |
06b3a860d
|
858 |
struct extent_buffer *leaf = NULL; |
bed92eae2
|
859 |
int ret; |
06b3a860d
|
860 |
int nr = 0; |
bed92eae2
|
861 |
|
bed92eae2
|
862 863 864 |
path = btrfs_alloc_path(); if (!path) return -ENOMEM; |
06b3a860d
|
865 866 867 868 869 |
path->leave_spinning = 1; key.objectid = 0; key.offset = 0; key.type = 0; |
bed92eae2
|
870 |
|
06b3a860d
|
871 |
while (1) { |
bed92eae2
|
872 |
ret = btrfs_search_slot(trans, root, &key, path, -1, 1); |
06b3a860d
|
873 874 875 876 877 |
if (ret < 0) goto out; leaf = path->nodes[0]; nr = btrfs_header_nritems(leaf); if (!nr) |
bed92eae2
|
878 |
break; |
06b3a860d
|
879 880 881 882 883 884 885 |
/* * delete the leaf one by one * since the whole tree is going * to be deleted. */ path->slots[0] = 0; ret = btrfs_del_items(trans, root, path, 0, nr); |
bed92eae2
|
886 887 |
if (ret) goto out; |
06b3a860d
|
888 |
|
bed92eae2
|
889 890 891 892 |
btrfs_release_path(path); } ret = 0; out: |
bed92eae2
|
893 894 895 |
btrfs_free_path(path); return ret; } |
340f1aa27
|
896 |
int btrfs_quota_enable(struct btrfs_fs_info *fs_info) |
bed92eae2
|
897 898 |
{ struct btrfs_root *quota_root; |
7708f029d
|
899 |
struct btrfs_root *tree_root = fs_info->tree_root; |
bed92eae2
|
900 901 902 903 |
struct btrfs_path *path = NULL; struct btrfs_qgroup_status_item *ptr; struct extent_buffer *leaf; struct btrfs_key key; |
7708f029d
|
904 905 |
struct btrfs_key found_key; struct btrfs_qgroup *qgroup = NULL; |
340f1aa27
|
906 |
struct btrfs_trans_handle *trans = NULL; |
bed92eae2
|
907 |
int ret = 0; |
7708f029d
|
908 |
int slot; |
bed92eae2
|
909 |
|
f2f6ed3d5
|
910 |
mutex_lock(&fs_info->qgroup_ioctl_lock); |
5d23515be
|
911 |
if (fs_info->quota_root) |
bed92eae2
|
912 |
goto out; |
bed92eae2
|
913 |
|
7503b83d8
|
914 915 916 917 918 |
fs_info->qgroup_ulist = ulist_alloc(GFP_KERNEL); if (!fs_info->qgroup_ulist) { ret = -ENOMEM; goto out; } |
49e5fb462
|
919 920 921 |
ret = btrfs_sysfs_add_qgroups(fs_info); if (ret < 0) goto out; |
340f1aa27
|
922 923 924 925 926 927 928 929 930 931 932 933 934 935 |
/* * 1 for quota root item * 1 for BTRFS_QGROUP_STATUS item * * Yet we also need 2*n items for a QGROUP_INFO/QGROUP_LIMIT items * per subvolume. However those are not currently reserved since it * would be a lot of overkill. */ trans = btrfs_start_transaction(tree_root, 2); if (IS_ERR(trans)) { ret = PTR_ERR(trans); trans = NULL; goto out; } |
bed92eae2
|
936 937 938 |
/* * initially create the quota tree */ |
9b7a2440a
|
939 |
quota_root = btrfs_create_tree(trans, BTRFS_QUOTA_TREE_OBJECTID); |
bed92eae2
|
940 941 |
if (IS_ERR(quota_root)) { ret = PTR_ERR(quota_root); |
340f1aa27
|
942 |
btrfs_abort_transaction(trans, ret); |
bed92eae2
|
943 944 945 946 |
goto out; } path = btrfs_alloc_path(); |
5b7ff5b3c
|
947 948 |
if (!path) { ret = -ENOMEM; |
340f1aa27
|
949 |
btrfs_abort_transaction(trans, ret); |
5b7ff5b3c
|
950 951 |
goto out_free_root; } |
bed92eae2
|
952 953 954 955 956 957 958 |
key.objectid = 0; key.type = BTRFS_QGROUP_STATUS_KEY; key.offset = 0; ret = btrfs_insert_empty_item(trans, quota_root, path, &key, sizeof(*ptr)); |
340f1aa27
|
959 960 |
if (ret) { btrfs_abort_transaction(trans, ret); |
5b7ff5b3c
|
961 |
goto out_free_path; |
340f1aa27
|
962 |
} |
bed92eae2
|
963 964 965 966 967 968 969 970 971 |
leaf = path->nodes[0]; ptr = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_qgroup_status_item); btrfs_set_qgroup_status_generation(leaf, ptr, trans->transid); btrfs_set_qgroup_status_version(leaf, ptr, BTRFS_QGROUP_STATUS_VERSION); fs_info->qgroup_flags = BTRFS_QGROUP_STATUS_FLAG_ON | BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; btrfs_set_qgroup_status_flags(leaf, ptr, fs_info->qgroup_flags); |
2f2320360
|
972 |
btrfs_set_qgroup_status_rescan(leaf, ptr, 0); |
bed92eae2
|
973 974 |
btrfs_mark_buffer_dirty(leaf); |
7708f029d
|
975 976 977 978 979 980 981 982 |
key.objectid = 0; key.type = BTRFS_ROOT_REF_KEY; key.offset = 0; btrfs_release_path(path); ret = btrfs_search_slot_for_read(tree_root, &key, path, 1, 0); if (ret > 0) goto out_add_root; |
340f1aa27
|
983 984 |
if (ret < 0) { btrfs_abort_transaction(trans, ret); |
7708f029d
|
985 |
goto out_free_path; |
340f1aa27
|
986 |
} |
7708f029d
|
987 988 989 990 991 992 993 994 995 |
while (1) { slot = path->slots[0]; leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &found_key, slot); if (found_key.type == BTRFS_ROOT_REF_KEY) { ret = add_qgroup_item(trans, quota_root, found_key.offset); |
340f1aa27
|
996 997 |
if (ret) { btrfs_abort_transaction(trans, ret); |
7708f029d
|
998 |
goto out_free_path; |
340f1aa27
|
999 |
} |
7708f029d
|
1000 |
|
7708f029d
|
1001 1002 |
qgroup = add_qgroup_rb(fs_info, found_key.offset); if (IS_ERR(qgroup)) { |
7708f029d
|
1003 |
ret = PTR_ERR(qgroup); |
340f1aa27
|
1004 |
btrfs_abort_transaction(trans, ret); |
7708f029d
|
1005 1006 |
goto out_free_path; } |
49e5fb462
|
1007 1008 1009 1010 1011 |
ret = btrfs_sysfs_add_one_qgroup(fs_info, qgroup); if (ret < 0) { btrfs_abort_transaction(trans, ret); goto out_free_path; } |
7708f029d
|
1012 1013 |
} ret = btrfs_next_item(tree_root, path); |
340f1aa27
|
1014 1015 |
if (ret < 0) { btrfs_abort_transaction(trans, ret); |
7708f029d
|
1016 |
goto out_free_path; |
340f1aa27
|
1017 |
} |
7708f029d
|
1018 1019 1020 1021 1022 1023 1024 |
if (ret) break; } out_add_root: btrfs_release_path(path); ret = add_qgroup_item(trans, quota_root, BTRFS_FS_TREE_OBJECTID); |
340f1aa27
|
1025 1026 |
if (ret) { btrfs_abort_transaction(trans, ret); |
7708f029d
|
1027 |
goto out_free_path; |
340f1aa27
|
1028 |
} |
7708f029d
|
1029 |
|
7708f029d
|
1030 1031 |
qgroup = add_qgroup_rb(fs_info, BTRFS_FS_TREE_OBJECTID); if (IS_ERR(qgroup)) { |
7708f029d
|
1032 |
ret = PTR_ERR(qgroup); |
340f1aa27
|
1033 |
btrfs_abort_transaction(trans, ret); |
7708f029d
|
1034 1035 |
goto out_free_path; } |
49e5fb462
|
1036 1037 1038 1039 1040 |
ret = btrfs_sysfs_add_one_qgroup(fs_info, qgroup); if (ret < 0) { btrfs_abort_transaction(trans, ret); goto out_free_path; } |
340f1aa27
|
1041 1042 |
ret = btrfs_commit_transaction(trans); |
b9b8a41ad
|
1043 1044 |
trans = NULL; if (ret) |
340f1aa27
|
1045 |
goto out_free_path; |
340f1aa27
|
1046 |
|
9a6f209e3
|
1047 1048 1049 1050 1051 1052 1053 1054 1055 |
/* * Set quota enabled flag after committing the transaction, to avoid * deadlocks on fs_info->qgroup_ioctl_lock with concurrent snapshot * creation. */ spin_lock(&fs_info->qgroup_lock); fs_info->quota_root = quota_root; set_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags); spin_unlock(&fs_info->qgroup_lock); |
5d23515be
|
1056 1057 1058 |
ret = qgroup_rescan_init(fs_info, 0, 1); if (!ret) { qgroup_rescan_zero_tracking(fs_info); |
d61acbbf5
|
1059 |
fs_info->qgroup_rescan_running = true; |
5d23515be
|
1060 1061 1062 |
btrfs_queue_work(fs_info->qgroup_rescan_workers, &fs_info->qgroup_rescan_work); } |
5b7ff5b3c
|
1063 |
out_free_path: |
bed92eae2
|
1064 |
btrfs_free_path(path); |
5b7ff5b3c
|
1065 |
out_free_root: |
8c38938c7
|
1066 |
if (ret) |
002465289
|
1067 |
btrfs_put_root(quota_root); |
5b7ff5b3c
|
1068 |
out: |
eb1716af8
|
1069 |
if (ret) { |
1e8f91586
|
1070 |
ulist_free(fs_info->qgroup_ulist); |
eb1716af8
|
1071 |
fs_info->qgroup_ulist = NULL; |
340f1aa27
|
1072 1073 |
if (trans) btrfs_end_transaction(trans); |
49e5fb462
|
1074 |
btrfs_sysfs_del_qgroups(fs_info); |
eb1716af8
|
1075 |
} |
f2f6ed3d5
|
1076 |
mutex_unlock(&fs_info->qgroup_ioctl_lock); |
bed92eae2
|
1077 1078 |
return ret; } |
340f1aa27
|
1079 |
int btrfs_quota_disable(struct btrfs_fs_info *fs_info) |
bed92eae2
|
1080 |
{ |
bed92eae2
|
1081 |
struct btrfs_root *quota_root; |
340f1aa27
|
1082 |
struct btrfs_trans_handle *trans = NULL; |
bed92eae2
|
1083 |
int ret = 0; |
f2f6ed3d5
|
1084 |
mutex_lock(&fs_info->qgroup_ioctl_lock); |
58400fce5
|
1085 |
if (!fs_info->quota_root) |
f2f6ed3d5
|
1086 |
goto out; |
340f1aa27
|
1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 |
/* * 1 For the root item * * We should also reserve enough items for the quota tree deletion in * btrfs_clean_quota_tree but this is not done. */ trans = btrfs_start_transaction(fs_info->tree_root, 1); if (IS_ERR(trans)) { ret = PTR_ERR(trans); goto out; } |
afcdd129e
|
1099 |
clear_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags); |
d06f23d6a
|
1100 |
btrfs_qgroup_wait_for_completion(fs_info, false); |
967ef5131
|
1101 |
spin_lock(&fs_info->qgroup_lock); |
bed92eae2
|
1102 1103 |
quota_root = fs_info->quota_root; fs_info->quota_root = NULL; |
8ea0ec9e0
|
1104 |
fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_ON; |
bed92eae2
|
1105 |
spin_unlock(&fs_info->qgroup_lock); |
e685da14a
|
1106 |
btrfs_free_qgroup_config(fs_info); |
bed92eae2
|
1107 |
ret = btrfs_clean_quota_tree(trans, quota_root); |
340f1aa27
|
1108 1109 1110 1111 |
if (ret) { btrfs_abort_transaction(trans, ret); goto end_trans; } |
bed92eae2
|
1112 |
|
ab9ce7d42
|
1113 |
ret = btrfs_del_root(trans, "a_root->root_key); |
340f1aa27
|
1114 1115 1116 1117 |
if (ret) { btrfs_abort_transaction(trans, ret); goto end_trans; } |
bed92eae2
|
1118 1119 1120 1121 |
list_del("a_root->dirty_list); btrfs_tree_lock(quota_root->node); |
6a884d7d5
|
1122 |
btrfs_clean_tree_block(quota_root->node); |
bed92eae2
|
1123 1124 |
btrfs_tree_unlock(quota_root->node); btrfs_free_tree_block(trans, quota_root, quota_root->node, 0, 1); |
002465289
|
1125 |
btrfs_put_root(quota_root); |
340f1aa27
|
1126 1127 1128 |
end_trans: ret = btrfs_end_transaction(trans); |
bed92eae2
|
1129 |
out: |
f2f6ed3d5
|
1130 |
mutex_unlock(&fs_info->qgroup_ioctl_lock); |
bed92eae2
|
1131 1132 |
return ret; } |
2f2320360
|
1133 1134 |
static void qgroup_dirty(struct btrfs_fs_info *fs_info, struct btrfs_qgroup *qgroup) |
bed92eae2
|
1135 |
{ |
2f2320360
|
1136 1137 |
if (list_empty(&qgroup->dirty)) list_add(&qgroup->dirty, &fs_info->dirty_qgroups); |
bed92eae2
|
1138 |
} |
9c8b35b1b
|
1139 |
/* |
429d6275d
|
1140 1141 1142 |
* The easy accounting, we're updating qgroup relationship whose child qgroup * only has exclusive extents. * |
52042d8e8
|
1143 |
* In this case, all exclusive extents will also be exclusive for parent, so |
429d6275d
|
1144 1145 1146 1147 1148 1149 |
* excl/rfer just get added/removed. * * So is qgroup reservation space, which should also be added/removed to * parent. * Or when child tries to release reservation space, parent will underflow its * reservation (for relationship adding case). |
9c8b35b1b
|
1150 1151 1152 1153 1154 |
* * Caller should hold fs_info->qgroup_lock. */ static int __qgroup_excl_accounting(struct btrfs_fs_info *fs_info, struct ulist *tmp, u64 ref_root, |
429d6275d
|
1155 |
struct btrfs_qgroup *src, int sign) |
9c8b35b1b
|
1156 1157 1158 1159 1160 |
{ struct btrfs_qgroup *qgroup; struct btrfs_qgroup_list *glist; struct ulist_node *unode; struct ulist_iterator uiter; |
429d6275d
|
1161 |
u64 num_bytes = src->excl; |
9c8b35b1b
|
1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 |
int ret = 0; qgroup = find_qgroup_rb(fs_info, ref_root); if (!qgroup) goto out; qgroup->rfer += sign * num_bytes; qgroup->rfer_cmpr += sign * num_bytes; WARN_ON(sign < 0 && qgroup->excl < num_bytes); qgroup->excl += sign * num_bytes; qgroup->excl_cmpr += sign * num_bytes; |
429d6275d
|
1174 1175 |
if (sign > 0) |
64ee4e751
|
1176 |
qgroup_rsv_add_by_qgroup(fs_info, qgroup, src); |
429d6275d
|
1177 |
else |
64ee4e751
|
1178 |
qgroup_rsv_release_by_qgroup(fs_info, qgroup, src); |
9c8b35b1b
|
1179 1180 1181 1182 1183 1184 |
qgroup_dirty(fs_info, qgroup); /* Get all of the parent groups that contain this qgroup */ list_for_each_entry(glist, &qgroup->groups, next_group) { ret = ulist_add(tmp, glist->group->qgroupid, |
ef2fff64f
|
1185 |
qgroup_to_aux(glist->group), GFP_ATOMIC); |
9c8b35b1b
|
1186 1187 1188 1189 1190 1191 1192 |
if (ret < 0) goto out; } /* Iterate all of the parents and adjust their reference counts */ ULIST_ITER_INIT(&uiter); while ((unode = ulist_next(tmp, &uiter))) { |
ef2fff64f
|
1193 |
qgroup = unode_aux_to_qgroup(unode); |
9c8b35b1b
|
1194 1195 1196 1197 |
qgroup->rfer += sign * num_bytes; qgroup->rfer_cmpr += sign * num_bytes; WARN_ON(sign < 0 && qgroup->excl < num_bytes); qgroup->excl += sign * num_bytes; |
429d6275d
|
1198 |
if (sign > 0) |
64ee4e751
|
1199 |
qgroup_rsv_add_by_qgroup(fs_info, qgroup, src); |
429d6275d
|
1200 |
else |
64ee4e751
|
1201 |
qgroup_rsv_release_by_qgroup(fs_info, qgroup, src); |
9c8b35b1b
|
1202 1203 1204 1205 1206 1207 |
qgroup->excl_cmpr += sign * num_bytes; qgroup_dirty(fs_info, qgroup); /* Add any parents of the parents */ list_for_each_entry(glist, &qgroup->groups, next_group) { ret = ulist_add(tmp, glist->group->qgroupid, |
ef2fff64f
|
1208 |
qgroup_to_aux(glist->group), GFP_ATOMIC); |
9c8b35b1b
|
1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 |
if (ret < 0) goto out; } } ret = 0; out: return ret; } /* * Quick path for updating qgroup with only excl refs. * * In that case, just update all parent will be enough. * Or we needs to do a full rescan. * Caller should also hold fs_info->qgroup_lock. * * Return 0 for quick update, return >0 for need to full rescan * and mark INCONSISTENT flag. * Return < 0 for other error. */ static int quick_update_accounting(struct btrfs_fs_info *fs_info, struct ulist *tmp, u64 src, u64 dst, int sign) { struct btrfs_qgroup *qgroup; int ret = 1; int err = 0; qgroup = find_qgroup_rb(fs_info, src); if (!qgroup) goto out; if (qgroup->excl == qgroup->rfer) { ret = 0; err = __qgroup_excl_accounting(fs_info, tmp, dst, |
429d6275d
|
1244 |
qgroup, sign); |
9c8b35b1b
|
1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 |
if (err < 0) { ret = err; goto out; } } out: if (ret) fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; return ret; } |
9f8a6ce6b
|
1255 1256 |
int btrfs_add_qgroup_relation(struct btrfs_trans_handle *trans, u64 src, u64 dst) |
bed92eae2
|
1257 |
{ |
9f8a6ce6b
|
1258 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
b7fef4f59
|
1259 1260 |
struct btrfs_qgroup *parent; struct btrfs_qgroup *member; |
534e6623b
|
1261 |
struct btrfs_qgroup_list *list; |
9c8b35b1b
|
1262 |
struct ulist *tmp; |
bed92eae2
|
1263 |
int ret = 0; |
8465ecec9
|
1264 1265 1266 |
/* Check the level of src and dst first */ if (btrfs_qgroup_level(src) >= btrfs_qgroup_level(dst)) return -EINVAL; |
6602caf14
|
1267 |
tmp = ulist_alloc(GFP_KERNEL); |
ab3680dd1
|
1268 1269 |
if (!tmp) return -ENOMEM; |
f2f6ed3d5
|
1270 |
mutex_lock(&fs_info->qgroup_ioctl_lock); |
e3b0edd29
|
1271 |
if (!fs_info->quota_root) { |
8a36e408d
|
1272 |
ret = -ENOTCONN; |
f2f6ed3d5
|
1273 1274 |
goto out; } |
b7fef4f59
|
1275 1276 1277 1278 1279 1280 |
member = find_qgroup_rb(fs_info, src); parent = find_qgroup_rb(fs_info, dst); if (!member || !parent) { ret = -EINVAL; goto out; } |
bed92eae2
|
1281 |
|
534e6623b
|
1282 1283 1284 1285 1286 1287 1288 |
/* check if such qgroup relation exist firstly */ list_for_each_entry(list, &member->groups, next_group) { if (list->group == parent) { ret = -EEXIST; goto out; } } |
711169c40
|
1289 |
ret = add_qgroup_relation_item(trans, src, dst); |
bed92eae2
|
1290 |
if (ret) |
f2f6ed3d5
|
1291 |
goto out; |
bed92eae2
|
1292 |
|
711169c40
|
1293 |
ret = add_qgroup_relation_item(trans, dst, src); |
bed92eae2
|
1294 |
if (ret) { |
99d7f09ac
|
1295 |
del_qgroup_relation_item(trans, src, dst); |
f2f6ed3d5
|
1296 |
goto out; |
bed92eae2
|
1297 1298 1299 |
} spin_lock(&fs_info->qgroup_lock); |
0b246afa6
|
1300 |
ret = add_relation_rb(fs_info, src, dst); |
9c8b35b1b
|
1301 1302 1303 1304 1305 |
if (ret < 0) { spin_unlock(&fs_info->qgroup_lock); goto out; } ret = quick_update_accounting(fs_info, tmp, src, dst, 1); |
bed92eae2
|
1306 |
spin_unlock(&fs_info->qgroup_lock); |
f2f6ed3d5
|
1307 1308 |
out: mutex_unlock(&fs_info->qgroup_ioctl_lock); |
9c8b35b1b
|
1309 |
ulist_free(tmp); |
bed92eae2
|
1310 1311 |
return ret; } |
6b36f1aa5
|
1312 1313 |
static int __del_qgroup_relation(struct btrfs_trans_handle *trans, u64 src, u64 dst) |
bed92eae2
|
1314 |
{ |
6b36f1aa5
|
1315 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
534e6623b
|
1316 1317 1318 |
struct btrfs_qgroup *parent; struct btrfs_qgroup *member; struct btrfs_qgroup_list *list; |
9c8b35b1b
|
1319 |
struct ulist *tmp; |
73798c465
|
1320 |
bool found = false; |
bed92eae2
|
1321 |
int ret = 0; |
73798c465
|
1322 |
int ret2; |
bed92eae2
|
1323 |
|
6602caf14
|
1324 |
tmp = ulist_alloc(GFP_KERNEL); |
9c8b35b1b
|
1325 1326 |
if (!tmp) return -ENOMEM; |
e3b0edd29
|
1327 |
if (!fs_info->quota_root) { |
8a36e408d
|
1328 |
ret = -ENOTCONN; |
f2f6ed3d5
|
1329 1330 |
goto out; } |
bed92eae2
|
1331 |
|
534e6623b
|
1332 1333 |
member = find_qgroup_rb(fs_info, src); parent = find_qgroup_rb(fs_info, dst); |
73798c465
|
1334 1335 1336 1337 1338 1339 |
/* * The parent/member pair doesn't exist, then try to delete the dead * relation items only. */ if (!member || !parent) goto delete_item; |
534e6623b
|
1340 1341 1342 |
/* check if such qgroup relation exist firstly */ list_for_each_entry(list, &member->groups, next_group) { |
73798c465
|
1343 1344 1345 1346 |
if (list->group == parent) { found = true; break; } |
534e6623b
|
1347 |
} |
73798c465
|
1348 1349 |
delete_item: |
99d7f09ac
|
1350 |
ret = del_qgroup_relation_item(trans, src, dst); |
73798c465
|
1351 1352 1353 1354 1355 |
if (ret < 0 && ret != -ENOENT) goto out; ret2 = del_qgroup_relation_item(trans, dst, src); if (ret2 < 0 && ret2 != -ENOENT) goto out; |
bed92eae2
|
1356 |
|
73798c465
|
1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 |
/* At least one deletion succeeded, return 0 */ if (!ret || !ret2) ret = 0; if (found) { spin_lock(&fs_info->qgroup_lock); del_relation_rb(fs_info, src, dst); ret = quick_update_accounting(fs_info, tmp, src, dst, -1); spin_unlock(&fs_info->qgroup_lock); } |
f2f6ed3d5
|
1367 |
out: |
9c8b35b1b
|
1368 |
ulist_free(tmp); |
f5a6b1c53
|
1369 1370 |
return ret; } |
39616c273
|
1371 1372 |
int btrfs_del_qgroup_relation(struct btrfs_trans_handle *trans, u64 src, u64 dst) |
f5a6b1c53
|
1373 |
{ |
39616c273
|
1374 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
f5a6b1c53
|
1375 1376 1377 |
int ret = 0; mutex_lock(&fs_info->qgroup_ioctl_lock); |
6b36f1aa5
|
1378 |
ret = __del_qgroup_relation(trans, src, dst); |
f2f6ed3d5
|
1379 |
mutex_unlock(&fs_info->qgroup_ioctl_lock); |
f5a6b1c53
|
1380 |
|
bed92eae2
|
1381 1382 |
return ret; } |
49a05ecde
|
1383 |
int btrfs_create_qgroup(struct btrfs_trans_handle *trans, u64 qgroupid) |
bed92eae2
|
1384 |
{ |
49a05ecde
|
1385 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
bed92eae2
|
1386 1387 1388 |
struct btrfs_root *quota_root; struct btrfs_qgroup *qgroup; int ret = 0; |
f2f6ed3d5
|
1389 |
mutex_lock(&fs_info->qgroup_ioctl_lock); |
e3b0edd29
|
1390 |
if (!fs_info->quota_root) { |
8a36e408d
|
1391 |
ret = -ENOTCONN; |
f2f6ed3d5
|
1392 1393 |
goto out; } |
e3b0edd29
|
1394 |
quota_root = fs_info->quota_root; |
534e6623b
|
1395 1396 1397 1398 1399 |
qgroup = find_qgroup_rb(fs_info, qgroupid); if (qgroup) { ret = -EEXIST; goto out; } |
bed92eae2
|
1400 1401 |
ret = add_qgroup_item(trans, quota_root, qgroupid); |
534e6623b
|
1402 1403 |
if (ret) goto out; |
bed92eae2
|
1404 1405 1406 1407 |
spin_lock(&fs_info->qgroup_lock); qgroup = add_qgroup_rb(fs_info, qgroupid); spin_unlock(&fs_info->qgroup_lock); |
49e5fb462
|
1408 |
if (IS_ERR(qgroup)) { |
bed92eae2
|
1409 |
ret = PTR_ERR(qgroup); |
49e5fb462
|
1410 1411 1412 |
goto out; } ret = btrfs_sysfs_add_one_qgroup(fs_info, qgroup); |
f2f6ed3d5
|
1413 1414 |
out: mutex_unlock(&fs_info->qgroup_ioctl_lock); |
bed92eae2
|
1415 1416 |
return ret; } |
3efbee1d0
|
1417 |
int btrfs_remove_qgroup(struct btrfs_trans_handle *trans, u64 qgroupid) |
bed92eae2
|
1418 |
{ |
3efbee1d0
|
1419 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
2cf687039
|
1420 |
struct btrfs_qgroup *qgroup; |
f5a6b1c53
|
1421 |
struct btrfs_qgroup_list *list; |
bed92eae2
|
1422 |
int ret = 0; |
f2f6ed3d5
|
1423 |
mutex_lock(&fs_info->qgroup_ioctl_lock); |
e3b0edd29
|
1424 |
if (!fs_info->quota_root) { |
8a36e408d
|
1425 |
ret = -ENOTCONN; |
f2f6ed3d5
|
1426 1427 |
goto out; } |
bed92eae2
|
1428 |
|
2cf687039
|
1429 |
qgroup = find_qgroup_rb(fs_info, qgroupid); |
534e6623b
|
1430 1431 1432 |
if (!qgroup) { ret = -ENOENT; goto out; |
2cf687039
|
1433 |
} |
b90e22ba4
|
1434 1435 1436 1437 1438 1439 |
/* Check if there are no children of this qgroup */ if (!list_empty(&qgroup->members)) { ret = -EBUSY; goto out; } |
69104618f
|
1440 |
ret = del_qgroup_item(trans, qgroupid); |
36b96fdc6
|
1441 1442 |
if (ret && ret != -ENOENT) goto out; |
bed92eae2
|
1443 |
|
f5a6b1c53
|
1444 1445 1446 |
while (!list_empty(&qgroup->groups)) { list = list_first_entry(&qgroup->groups, struct btrfs_qgroup_list, next_group); |
6b36f1aa5
|
1447 1448 |
ret = __del_qgroup_relation(trans, qgroupid, list->group->qgroupid); |
f5a6b1c53
|
1449 1450 1451 |
if (ret) goto out; } |
bed92eae2
|
1452 |
spin_lock(&fs_info->qgroup_lock); |
0b246afa6
|
1453 |
del_qgroup_rb(fs_info, qgroupid); |
bed92eae2
|
1454 |
spin_unlock(&fs_info->qgroup_lock); |
f2f6ed3d5
|
1455 1456 |
out: mutex_unlock(&fs_info->qgroup_ioctl_lock); |
bed92eae2
|
1457 1458 |
return ret; } |
f0042d5e9
|
1459 |
int btrfs_limit_qgroup(struct btrfs_trans_handle *trans, u64 qgroupid, |
bed92eae2
|
1460 1461 |
struct btrfs_qgroup_limit *limit) { |
f0042d5e9
|
1462 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
bed92eae2
|
1463 1464 |
struct btrfs_qgroup *qgroup; int ret = 0; |
fe7599079
|
1465 1466 1467 1468 1469 |
/* Sometimes we would want to clear the limit on this qgroup. * To meet this requirement, we treat the -1 as a special value * which tell kernel to clear the limit on this qgroup. */ const u64 CLEAR_VALUE = -1; |
bed92eae2
|
1470 |
|
f2f6ed3d5
|
1471 |
mutex_lock(&fs_info->qgroup_ioctl_lock); |
e3b0edd29
|
1472 |
if (!fs_info->quota_root) { |
8a36e408d
|
1473 |
ret = -ENOTCONN; |
f2f6ed3d5
|
1474 1475 |
goto out; } |
bed92eae2
|
1476 |
|
ddb47afa5
|
1477 1478 1479 1480 1481 |
qgroup = find_qgroup_rb(fs_info, qgroupid); if (!qgroup) { ret = -ENOENT; goto out; } |
bed92eae2
|
1482 |
|
58400fce5
|
1483 |
spin_lock(&fs_info->qgroup_lock); |
fe7599079
|
1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 |
if (limit->flags & BTRFS_QGROUP_LIMIT_MAX_RFER) { if (limit->max_rfer == CLEAR_VALUE) { qgroup->lim_flags &= ~BTRFS_QGROUP_LIMIT_MAX_RFER; limit->flags &= ~BTRFS_QGROUP_LIMIT_MAX_RFER; qgroup->max_rfer = 0; } else { qgroup->max_rfer = limit->max_rfer; } } if (limit->flags & BTRFS_QGROUP_LIMIT_MAX_EXCL) { if (limit->max_excl == CLEAR_VALUE) { qgroup->lim_flags &= ~BTRFS_QGROUP_LIMIT_MAX_EXCL; limit->flags &= ~BTRFS_QGROUP_LIMIT_MAX_EXCL; qgroup->max_excl = 0; } else { qgroup->max_excl = limit->max_excl; } } if (limit->flags & BTRFS_QGROUP_LIMIT_RSV_RFER) { if (limit->rsv_rfer == CLEAR_VALUE) { qgroup->lim_flags &= ~BTRFS_QGROUP_LIMIT_RSV_RFER; limit->flags &= ~BTRFS_QGROUP_LIMIT_RSV_RFER; qgroup->rsv_rfer = 0; } else { qgroup->rsv_rfer = limit->rsv_rfer; } } if (limit->flags & BTRFS_QGROUP_LIMIT_RSV_EXCL) { if (limit->rsv_excl == CLEAR_VALUE) { qgroup->lim_flags &= ~BTRFS_QGROUP_LIMIT_RSV_EXCL; limit->flags &= ~BTRFS_QGROUP_LIMIT_RSV_EXCL; qgroup->rsv_excl = 0; } else { qgroup->rsv_excl = limit->rsv_excl; } } |
03477d945
|
1520 |
qgroup->lim_flags |= limit->flags; |
bed92eae2
|
1521 |
spin_unlock(&fs_info->qgroup_lock); |
1510e71c6
|
1522 |
|
ac8a866af
|
1523 |
ret = update_qgroup_limit_item(trans, qgroup); |
1510e71c6
|
1524 1525 1526 1527 1528 |
if (ret) { fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; btrfs_info(fs_info, "unable to update quota limit for %llu", qgroupid); } |
f2f6ed3d5
|
1529 1530 |
out: mutex_unlock(&fs_info->qgroup_ioctl_lock); |
bed92eae2
|
1531 1532 |
return ret; } |
1152651a0
|
1533 |
|
50b3e040b
|
1534 |
int btrfs_qgroup_trace_extent_nolock(struct btrfs_fs_info *fs_info, |
cb93b52cc
|
1535 1536 |
struct btrfs_delayed_ref_root *delayed_refs, struct btrfs_qgroup_extent_record *record) |
3368d001b
|
1537 1538 1539 1540 1541 |
{ struct rb_node **p = &delayed_refs->dirty_extent_root.rb_node; struct rb_node *parent_node = NULL; struct btrfs_qgroup_extent_record *entry; u64 bytenr = record->bytenr; |
a4666e688
|
1542 |
lockdep_assert_held(&delayed_refs->lock); |
50b3e040b
|
1543 |
trace_btrfs_qgroup_trace_extent(fs_info, record); |
82bd101b5
|
1544 |
|
3368d001b
|
1545 1546 1547 1548 |
while (*p) { parent_node = *p; entry = rb_entry(parent_node, struct btrfs_qgroup_extent_record, node); |
1418bae1c
|
1549 |
if (bytenr < entry->bytenr) { |
3368d001b
|
1550 |
p = &(*p)->rb_left; |
1418bae1c
|
1551 |
} else if (bytenr > entry->bytenr) { |
3368d001b
|
1552 |
p = &(*p)->rb_right; |
1418bae1c
|
1553 1554 1555 1556 1557 1558 |
} else { if (record->data_rsv && !entry->data_rsv) { entry->data_rsv = record->data_rsv; entry->data_rsv_refroot = record->data_rsv_refroot; } |
cb93b52cc
|
1559 |
return 1; |
1418bae1c
|
1560 |
} |
3368d001b
|
1561 1562 1563 1564 |
} rb_link_node(&record->node, parent_node, p); rb_insert_color(&record->node, &delayed_refs->dirty_extent_root); |
cb93b52cc
|
1565 1566 |
return 0; } |
fb235dc06
|
1567 1568 1569 1570 1571 1572 |
int btrfs_qgroup_trace_extent_post(struct btrfs_fs_info *fs_info, struct btrfs_qgroup_extent_record *qrecord) { struct ulist *old_root; u64 bytenr = qrecord->bytenr; int ret; |
c995ab3cd
|
1573 |
ret = btrfs_find_all_roots(NULL, fs_info, bytenr, 0, &old_root, false); |
952bd3db0
|
1574 1575 1576 1577 1578 1579 1580 |
if (ret < 0) { fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; btrfs_warn(fs_info, "error accounting new delayed refs extent (err code: %d), quota inconsistent", ret); return 0; } |
fb235dc06
|
1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 |
/* * Here we don't need to get the lock of * trans->transaction->delayed_refs, since inserted qrecord won't * be deleted, only qrecord->node may be modified (new qrecord insert) * * So modifying qrecord->old_roots is safe here */ qrecord->old_roots = old_root; return 0; } |
a95f3aafd
|
1592 1593 |
int btrfs_qgroup_trace_extent(struct btrfs_trans_handle *trans, u64 bytenr, u64 num_bytes, gfp_t gfp_flag) |
cb93b52cc
|
1594 |
{ |
a95f3aafd
|
1595 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
cb93b52cc
|
1596 1597 1598 |
struct btrfs_qgroup_extent_record *record; struct btrfs_delayed_ref_root *delayed_refs; int ret; |
afcdd129e
|
1599 1600 |
if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags) || bytenr == 0 || num_bytes == 0) |
cb93b52cc
|
1601 |
return 0; |
1418bae1c
|
1602 |
record = kzalloc(sizeof(*record), gfp_flag); |
cb93b52cc
|
1603 1604 1605 1606 1607 1608 1609 1610 1611 |
if (!record) return -ENOMEM; delayed_refs = &trans->transaction->delayed_refs; record->bytenr = bytenr; record->num_bytes = num_bytes; record->old_roots = NULL; spin_lock(&delayed_refs->lock); |
2ff7e61e0
|
1612 |
ret = btrfs_qgroup_trace_extent_nolock(fs_info, delayed_refs, record); |
cb93b52cc
|
1613 |
spin_unlock(&delayed_refs->lock); |
fb235dc06
|
1614 |
if (ret > 0) { |
cb93b52cc
|
1615 |
kfree(record); |
fb235dc06
|
1616 1617 1618 |
return 0; } return btrfs_qgroup_trace_extent_post(fs_info, record); |
3368d001b
|
1619 |
} |
33d1f05cc
|
1620 |
int btrfs_qgroup_trace_leaf_items(struct btrfs_trans_handle *trans, |
33d1f05cc
|
1621 1622 |
struct extent_buffer *eb) { |
8d38d7eb7
|
1623 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
33d1f05cc
|
1624 1625 1626 1627 1628 1629 1630 |
int nr = btrfs_header_nritems(eb); int i, extent_type, ret; struct btrfs_key key; struct btrfs_file_extent_item *fi; u64 bytenr, num_bytes; /* We can be called directly from walk_up_proc() */ |
0b246afa6
|
1631 |
if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) |
33d1f05cc
|
1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 |
return 0; for (i = 0; i < nr; i++) { btrfs_item_key_to_cpu(eb, &key, i); if (key.type != BTRFS_EXTENT_DATA_KEY) continue; fi = btrfs_item_ptr(eb, i, struct btrfs_file_extent_item); /* filter out non qgroup-accountable extents */ extent_type = btrfs_file_extent_type(eb, fi); if (extent_type == BTRFS_FILE_EXTENT_INLINE) continue; bytenr = btrfs_file_extent_disk_bytenr(eb, fi); if (!bytenr) continue; num_bytes = btrfs_file_extent_disk_num_bytes(eb, fi); |
a95f3aafd
|
1652 1653 |
ret = btrfs_qgroup_trace_extent(trans, bytenr, num_bytes, GFP_NOFS); |
33d1f05cc
|
1654 1655 1656 |
if (ret) return ret; } |
cddf3b2cb
|
1657 |
cond_resched(); |
33d1f05cc
|
1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 |
return 0; } /* * Walk up the tree from the bottom, freeing leaves and any interior * nodes which have had all slots visited. If a node (leaf or * interior) is freed, the node above it will have it's slot * incremented. The root node will never be freed. * * At the end of this function, we should have a path which has all * slots incremented to the next position for a search. If we need to * read a new node it will be NULL and the node above it will have the * correct slot selected for a later read. * * If we increment the root nodes slot counter past the number of * elements, 1 is returned to signal completion of the search. */ |
15b34517a
|
1675 |
static int adjust_slots_upwards(struct btrfs_path *path, int root_level) |
33d1f05cc
|
1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 |
{ int level = 0; int nr, slot; struct extent_buffer *eb; if (root_level == 0) return 1; while (level <= root_level) { eb = path->nodes[level]; nr = btrfs_header_nritems(eb); path->slots[level]++; slot = path->slots[level]; if (slot >= nr || level == 0) { /* * Don't free the root - we will detect this * condition after our loop and return a * positive value for caller to stop walking the tree. */ if (level != root_level) { btrfs_tree_unlock_rw(eb, path->locks[level]); path->locks[level] = 0; free_extent_buffer(eb); path->nodes[level] = NULL; path->slots[level] = 0; } } else { /* * We have a valid slot to walk back down * from. Stop here so caller can process these * new nodes. */ break; } level++; } eb = path->nodes[root_level]; if (path->slots[root_level] >= btrfs_header_nritems(eb)) return 1; return 0; } |
25982561d
|
1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 |
/* * Helper function to trace a subtree tree block swap. * * The swap will happen in highest tree block, but there may be a lot of * tree blocks involved. * * For example: * OO = Old tree blocks * NN = New tree blocks allocated during balance * * File tree (257) Reloc tree for 257 * L2 OO NN * / \ / \ * L1 OO OO (a) OO NN (a) * / \ / \ / \ / \ * L0 OO OO OO OO OO OO NN NN * (b) (c) (b) (c) * * When calling qgroup_trace_extent_swap(), we will pass: * @src_eb = OO(a) * @dst_path = [ nodes[1] = NN(a), nodes[0] = NN(c) ] * @dst_level = 0 * @root_level = 1 * * In that case, qgroup_trace_extent_swap() will search from OO(a) to * reach OO(c), then mark both OO(c) and NN(c) as qgroup dirty. * * The main work of qgroup_trace_extent_swap() can be split into 3 parts: * * 1) Tree search from @src_eb * It should acts as a simplified btrfs_search_slot(). * The key for search can be extracted from @dst_path->nodes[dst_level] * (first key). * * 2) Mark the final tree blocks in @src_path and @dst_path qgroup dirty * NOTE: In above case, OO(a) and NN(a) won't be marked qgroup dirty. |
52042d8e8
|
1757 |
* They should be marked during previous (@dst_level = 1) iteration. |
25982561d
|
1758 1759 1760 1761 1762 1763 |
* * 3) Mark file extents in leaves dirty * We don't have good way to pick out new file extents only. * So we still follow the old method by scanning all file extents in * the leave. * |
52042d8e8
|
1764 |
* This function can free us from keeping two paths, thus later we only need |
25982561d
|
1765 1766 1767 1768 1769 |
* to care about how to iterate all new tree blocks in reloc tree. */ static int qgroup_trace_extent_swap(struct btrfs_trans_handle* trans, struct extent_buffer *src_eb, struct btrfs_path *dst_path, |
3d0174f78
|
1770 1771 |
int dst_level, int root_level, bool trace_leaf) |
25982561d
|
1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 |
{ struct btrfs_key key; struct btrfs_path *src_path; struct btrfs_fs_info *fs_info = trans->fs_info; u32 nodesize = fs_info->nodesize; int cur_level = root_level; int ret; BUG_ON(dst_level > root_level); /* Level mismatch */ if (btrfs_header_level(src_eb) != root_level) return -EINVAL; src_path = btrfs_alloc_path(); if (!src_path) { ret = -ENOMEM; goto out; } if (dst_level) btrfs_node_key_to_cpu(dst_path->nodes[dst_level], &key, 0); else btrfs_item_key_to_cpu(dst_path->nodes[dst_level], &key, 0); /* For src_path */ |
67439dadb
|
1797 |
atomic_inc(&src_eb->refs); |
25982561d
|
1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 |
src_path->nodes[root_level] = src_eb; src_path->slots[root_level] = dst_path->slots[root_level]; src_path->locks[root_level] = 0; /* A simplified version of btrfs_search_slot() */ while (cur_level >= dst_level) { struct btrfs_key src_key; struct btrfs_key dst_key; if (src_path->nodes[cur_level] == NULL) { struct btrfs_key first_key; struct extent_buffer *eb; int parent_slot; u64 child_gen; u64 child_bytenr; eb = src_path->nodes[cur_level + 1]; parent_slot = src_path->slots[cur_level + 1]; child_bytenr = btrfs_node_blockptr(eb, parent_slot); child_gen = btrfs_node_ptr_generation(eb, parent_slot); btrfs_node_key_to_cpu(eb, &first_key, parent_slot); eb = read_tree_block(fs_info, child_bytenr, child_gen, cur_level, &first_key); if (IS_ERR(eb)) { ret = PTR_ERR(eb); goto out; } else if (!extent_buffer_uptodate(eb)) { free_extent_buffer(eb); ret = -EIO; goto out; } src_path->nodes[cur_level] = eb; btrfs_tree_read_lock(eb); |
300aa896e
|
1834 |
btrfs_set_lock_blocking_read(eb); |
25982561d
|
1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 |
src_path->locks[cur_level] = BTRFS_READ_LOCK_BLOCKING; } src_path->slots[cur_level] = dst_path->slots[cur_level]; if (cur_level) { btrfs_node_key_to_cpu(dst_path->nodes[cur_level], &dst_key, dst_path->slots[cur_level]); btrfs_node_key_to_cpu(src_path->nodes[cur_level], &src_key, src_path->slots[cur_level]); } else { btrfs_item_key_to_cpu(dst_path->nodes[cur_level], &dst_key, dst_path->slots[cur_level]); btrfs_item_key_to_cpu(src_path->nodes[cur_level], &src_key, src_path->slots[cur_level]); } /* Content mismatch, something went wrong */ if (btrfs_comp_cpu_keys(&dst_key, &src_key)) { ret = -ENOENT; goto out; } cur_level--; } /* * Now both @dst_path and @src_path have been populated, record the tree * blocks for qgroup accounting. */ ret = btrfs_qgroup_trace_extent(trans, src_path->nodes[dst_level]->start, nodesize, GFP_NOFS); if (ret < 0) goto out; ret = btrfs_qgroup_trace_extent(trans, dst_path->nodes[dst_level]->start, nodesize, GFP_NOFS); if (ret < 0) goto out; /* Record leaf file extents */ |
3d0174f78
|
1873 |
if (dst_level == 0 && trace_leaf) { |
25982561d
|
1874 1875 1876 1877 1878 1879 1880 1881 1882 |
ret = btrfs_qgroup_trace_leaf_items(trans, src_path->nodes[0]); if (ret < 0) goto out; ret = btrfs_qgroup_trace_leaf_items(trans, dst_path->nodes[0]); } out: btrfs_free_path(src_path); return ret; } |
ea49f3e73
|
1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 |
/* * Helper function to do recursive generation-aware depth-first search, to * locate all new tree blocks in a subtree of reloc tree. * * E.g. (OO = Old tree blocks, NN = New tree blocks, whose gen == last_snapshot) * reloc tree * L2 NN (a) * / \ * L1 OO NN (b) * / \ / \ * L0 OO OO OO NN * (c) (d) * If we pass: * @dst_path = [ nodes[1] = NN(b), nodes[0] = NULL ], * @cur_level = 1 * @root_level = 1 * * We will iterate through tree blocks NN(b), NN(d) and info qgroup to trace * above tree blocks along with their counter parts in file tree. |
52042d8e8
|
1902 |
* While during search, old tree blocks OO(c) will be skipped as tree block swap |
ea49f3e73
|
1903 1904 1905 1906 1907 1908 |
* won't affect OO(c). */ static int qgroup_trace_new_subtree_blocks(struct btrfs_trans_handle* trans, struct extent_buffer *src_eb, struct btrfs_path *dst_path, int cur_level, int root_level, |
3d0174f78
|
1909 |
u64 last_snapshot, bool trace_leaf) |
ea49f3e73
|
1910 1911 1912 1913 1914 1915 1916 1917 |
{ struct btrfs_fs_info *fs_info = trans->fs_info; struct extent_buffer *eb; bool need_cleanup = false; int ret = 0; int i; /* Level sanity check */ |
7ff2c2a1a
|
1918 1919 |
if (cur_level < 0 || cur_level >= BTRFS_MAX_LEVEL - 1 || root_level < 0 || root_level >= BTRFS_MAX_LEVEL - 1 || |
ea49f3e73
|
1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 |
root_level < cur_level) { btrfs_err_rl(fs_info, "%s: bad levels, cur_level=%d root_level=%d", __func__, cur_level, root_level); return -EUCLEAN; } /* Read the tree block if needed */ if (dst_path->nodes[cur_level] == NULL) { struct btrfs_key first_key; int parent_slot; u64 child_gen; u64 child_bytenr; /* * dst_path->nodes[root_level] must be initialized before * calling this function. */ if (cur_level == root_level) { btrfs_err_rl(fs_info, "%s: dst_path->nodes[%d] not initialized, root_level=%d cur_level=%d", __func__, root_level, root_level, cur_level); return -EUCLEAN; } /* * We need to get child blockptr/gen from parent before we can * read it. */ eb = dst_path->nodes[cur_level + 1]; parent_slot = dst_path->slots[cur_level + 1]; child_bytenr = btrfs_node_blockptr(eb, parent_slot); child_gen = btrfs_node_ptr_generation(eb, parent_slot); btrfs_node_key_to_cpu(eb, &first_key, parent_slot); /* This node is old, no need to trace */ if (child_gen < last_snapshot) goto out; eb = read_tree_block(fs_info, child_bytenr, child_gen, cur_level, &first_key); if (IS_ERR(eb)) { ret = PTR_ERR(eb); goto out; } else if (!extent_buffer_uptodate(eb)) { free_extent_buffer(eb); ret = -EIO; goto out; } dst_path->nodes[cur_level] = eb; dst_path->slots[cur_level] = 0; btrfs_tree_read_lock(eb); |
300aa896e
|
1974 |
btrfs_set_lock_blocking_read(eb); |
ea49f3e73
|
1975 1976 1977 1978 1979 1980 |
dst_path->locks[cur_level] = BTRFS_READ_LOCK_BLOCKING; need_cleanup = true; } /* Now record this tree block and its counter part for qgroups */ ret = qgroup_trace_extent_swap(trans, src_eb, dst_path, cur_level, |
3d0174f78
|
1981 |
root_level, trace_leaf); |
ea49f3e73
|
1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 |
if (ret < 0) goto cleanup; eb = dst_path->nodes[cur_level]; if (cur_level > 0) { /* Iterate all child tree blocks */ for (i = 0; i < btrfs_header_nritems(eb); i++) { /* Skip old tree blocks as they won't be swapped */ if (btrfs_node_ptr_generation(eb, i) < last_snapshot) continue; dst_path->slots[cur_level] = i; /* Recursive call (at most 7 times) */ ret = qgroup_trace_new_subtree_blocks(trans, src_eb, dst_path, cur_level - 1, root_level, |
3d0174f78
|
1998 |
last_snapshot, trace_leaf); |
ea49f3e73
|
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 |
if (ret < 0) goto cleanup; } } cleanup: if (need_cleanup) { /* Clean up */ btrfs_tree_unlock_rw(dst_path->nodes[cur_level], dst_path->locks[cur_level]); free_extent_buffer(dst_path->nodes[cur_level]); dst_path->nodes[cur_level] = NULL; dst_path->slots[cur_level] = 0; dst_path->locks[cur_level] = 0; } out: return ret; } |
5aea1a4fc
|
2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 |
static int qgroup_trace_subtree_swap(struct btrfs_trans_handle *trans, struct extent_buffer *src_eb, struct extent_buffer *dst_eb, u64 last_snapshot, bool trace_leaf) { struct btrfs_fs_info *fs_info = trans->fs_info; struct btrfs_path *dst_path = NULL; int level; int ret; if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) return 0; /* Wrong parameter order */ if (btrfs_header_generation(src_eb) > btrfs_header_generation(dst_eb)) { btrfs_err_rl(fs_info, "%s: bad parameter order, src_gen=%llu dst_gen=%llu", __func__, btrfs_header_generation(src_eb), btrfs_header_generation(dst_eb)); return -EUCLEAN; } if (!extent_buffer_uptodate(src_eb) || !extent_buffer_uptodate(dst_eb)) { ret = -EIO; goto out; } level = btrfs_header_level(dst_eb); dst_path = btrfs_alloc_path(); if (!dst_path) { ret = -ENOMEM; goto out; } /* For dst_path */ |
67439dadb
|
2051 |
atomic_inc(&dst_eb->refs); |
5aea1a4fc
|
2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 |
dst_path->nodes[level] = dst_eb; dst_path->slots[level] = 0; dst_path->locks[level] = 0; /* Do the generation aware breadth-first search */ ret = qgroup_trace_new_subtree_blocks(trans, src_eb, dst_path, level, level, last_snapshot, trace_leaf); if (ret < 0) goto out; ret = 0; out: btrfs_free_path(dst_path); if (ret < 0) fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; return ret; } |
33d1f05cc
|
2069 |
int btrfs_qgroup_trace_subtree(struct btrfs_trans_handle *trans, |
33d1f05cc
|
2070 2071 2072 |
struct extent_buffer *root_eb, u64 root_gen, int root_level) { |
deb406274
|
2073 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
33d1f05cc
|
2074 2075 2076 2077 |
int ret = 0; int level; struct extent_buffer *eb = root_eb; struct btrfs_path *path = NULL; |
b6e6bca51
|
2078 |
BUG_ON(root_level < 0 || root_level >= BTRFS_MAX_LEVEL); |
33d1f05cc
|
2079 |
BUG_ON(root_eb == NULL); |
0b246afa6
|
2080 |
if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) |
33d1f05cc
|
2081 2082 2083 |
return 0; if (!extent_buffer_uptodate(root_eb)) { |
581c17604
|
2084 |
ret = btrfs_read_buffer(root_eb, root_gen, root_level, NULL); |
33d1f05cc
|
2085 2086 2087 2088 2089 |
if (ret) goto out; } if (root_level == 0) { |
8d38d7eb7
|
2090 |
ret = btrfs_qgroup_trace_leaf_items(trans, root_eb); |
33d1f05cc
|
2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 |
goto out; } path = btrfs_alloc_path(); if (!path) return -ENOMEM; /* * Walk down the tree. Missing extent blocks are filled in as * we go. Metadata is accounted every time we read a new * extent block. * * When we reach a leaf, we account for file extent items in it, * walk back up the tree (adjusting slot pointers as we go) * and restart the search process. */ |
67439dadb
|
2107 |
atomic_inc(&root_eb->refs); /* For path */ |
33d1f05cc
|
2108 2109 2110 2111 2112 2113 2114 |
path->nodes[root_level] = root_eb; path->slots[root_level] = 0; path->locks[root_level] = 0; /* so release_path doesn't try to unlock */ walk_down: level = root_level; while (level >= 0) { if (path->nodes[level] == NULL) { |
581c17604
|
2115 |
struct btrfs_key first_key; |
33d1f05cc
|
2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 |
int parent_slot; u64 child_gen; u64 child_bytenr; /* * We need to get child blockptr/gen from parent before * we can read it. */ eb = path->nodes[level + 1]; parent_slot = path->slots[level + 1]; child_bytenr = btrfs_node_blockptr(eb, parent_slot); child_gen = btrfs_node_ptr_generation(eb, parent_slot); |
581c17604
|
2128 |
btrfs_node_key_to_cpu(eb, &first_key, parent_slot); |
33d1f05cc
|
2129 |
|
581c17604
|
2130 2131 |
eb = read_tree_block(fs_info, child_bytenr, child_gen, level, &first_key); |
33d1f05cc
|
2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 |
if (IS_ERR(eb)) { ret = PTR_ERR(eb); goto out; } else if (!extent_buffer_uptodate(eb)) { free_extent_buffer(eb); ret = -EIO; goto out; } path->nodes[level] = eb; path->slots[level] = 0; btrfs_tree_read_lock(eb); |
300aa896e
|
2145 |
btrfs_set_lock_blocking_read(eb); |
33d1f05cc
|
2146 |
path->locks[level] = BTRFS_READ_LOCK_BLOCKING; |
a95f3aafd
|
2147 |
ret = btrfs_qgroup_trace_extent(trans, child_bytenr, |
0b246afa6
|
2148 2149 |
fs_info->nodesize, GFP_NOFS); |
33d1f05cc
|
2150 2151 2152 2153 2154 |
if (ret) goto out; } if (level == 0) { |
8d38d7eb7
|
2155 2156 |
ret = btrfs_qgroup_trace_leaf_items(trans, path->nodes[level]); |
33d1f05cc
|
2157 2158 2159 2160 |
if (ret) goto out; /* Nonzero return here means we completed our search */ |
15b34517a
|
2161 |
ret = adjust_slots_upwards(path, root_level); |
33d1f05cc
|
2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 |
if (ret) break; /* Restart search with new slots */ goto walk_down; } level--; } ret = 0; out: btrfs_free_path(path); return ret; } |
d810ef2be
|
2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 |
#define UPDATE_NEW 0 #define UPDATE_OLD 1 /* * Walk all of the roots that points to the bytenr and adjust their refcnts. */ static int qgroup_update_refcnt(struct btrfs_fs_info *fs_info, struct ulist *roots, struct ulist *tmp, struct ulist *qgroups, u64 seq, int update_old) { struct ulist_node *unode; struct ulist_iterator uiter; struct ulist_node *tmp_unode; struct ulist_iterator tmp_uiter; struct btrfs_qgroup *qg; int ret = 0; if (!roots) return 0; ULIST_ITER_INIT(&uiter); while ((unode = ulist_next(roots, &uiter))) { qg = find_qgroup_rb(fs_info, unode->val); if (!qg) continue; ulist_reinit(tmp); |
ef2fff64f
|
2203 |
ret = ulist_add(qgroups, qg->qgroupid, qgroup_to_aux(qg), |
d810ef2be
|
2204 2205 2206 |
GFP_ATOMIC); if (ret < 0) return ret; |
ef2fff64f
|
2207 |
ret = ulist_add(tmp, qg->qgroupid, qgroup_to_aux(qg), GFP_ATOMIC); |
d810ef2be
|
2208 2209 2210 2211 2212 |
if (ret < 0) return ret; ULIST_ITER_INIT(&tmp_uiter); while ((tmp_unode = ulist_next(tmp, &tmp_uiter))) { struct btrfs_qgroup_list *glist; |
ef2fff64f
|
2213 |
qg = unode_aux_to_qgroup(tmp_unode); |
d810ef2be
|
2214 2215 2216 2217 2218 2219 |
if (update_old) btrfs_qgroup_update_old_refcnt(qg, seq, 1); else btrfs_qgroup_update_new_refcnt(qg, seq, 1); list_for_each_entry(glist, &qg->groups, next_group) { ret = ulist_add(qgroups, glist->group->qgroupid, |
ef2fff64f
|
2220 |
qgroup_to_aux(glist->group), |
d810ef2be
|
2221 2222 2223 2224 |
GFP_ATOMIC); if (ret < 0) return ret; ret = ulist_add(tmp, glist->group->qgroupid, |
ef2fff64f
|
2225 |
qgroup_to_aux(glist->group), |
d810ef2be
|
2226 2227 2228 2229 2230 2231 2232 2233 |
GFP_ATOMIC); if (ret < 0) return ret; } } } return 0; } |
fcebe4562
|
2234 |
/* |
823ae5b8e
|
2235 2236 |
* Update qgroup rfer/excl counters. * Rfer update is easy, codes can explain themselves. |
e69bcee37
|
2237 |
* |
823ae5b8e
|
2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 |
* Excl update is tricky, the update is split into 2 part. * Part 1: Possible exclusive <-> sharing detect: * | A | !A | * ------------------------------------- * B | * | - | * ------------------------------------- * !B | + | ** | * ------------------------------------- * * Conditions: * A: cur_old_roots < nr_old_roots (not exclusive before) * !A: cur_old_roots == nr_old_roots (possible exclusive before) * B: cur_new_roots < nr_new_roots (not exclusive now) |
013276101
|
2251 |
* !B: cur_new_roots == nr_new_roots (possible exclusive now) |
823ae5b8e
|
2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 |
* * Results: * +: Possible sharing -> exclusive -: Possible exclusive -> sharing * *: Definitely not changed. **: Possible unchanged. * * For !A and !B condition, the exception is cur_old/new_roots == 0 case. * * To make the logic clear, we first use condition A and B to split * combination into 4 results. * * Then, for result "+" and "-", check old/new_roots == 0 case, as in them * only on variant maybe 0. * * Lastly, check result **, since there are 2 variants maybe 0, split them * again(2x2). * But this time we don't need to consider other things, the codes and logic * is easy to understand now. */ static int qgroup_update_counters(struct btrfs_fs_info *fs_info, struct ulist *qgroups, u64 nr_old_roots, u64 nr_new_roots, u64 num_bytes, u64 seq) { struct ulist_node *unode; struct ulist_iterator uiter; struct btrfs_qgroup *qg; u64 cur_new_count, cur_old_count; ULIST_ITER_INIT(&uiter); while ((unode = ulist_next(qgroups, &uiter))) { bool dirty = false; |
ef2fff64f
|
2284 |
qg = unode_aux_to_qgroup(unode); |
823ae5b8e
|
2285 2286 |
cur_old_count = btrfs_qgroup_get_old_refcnt(qg, seq); cur_new_count = btrfs_qgroup_get_new_refcnt(qg, seq); |
8b317901d
|
2287 2288 |
trace_qgroup_update_counters(fs_info, qg, cur_old_count, cur_new_count); |
0f5dcf8de
|
2289 |
|
823ae5b8e
|
2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 |
/* Rfer update part */ if (cur_old_count == 0 && cur_new_count > 0) { qg->rfer += num_bytes; qg->rfer_cmpr += num_bytes; dirty = true; } if (cur_old_count > 0 && cur_new_count == 0) { qg->rfer -= num_bytes; qg->rfer_cmpr -= num_bytes; dirty = true; } /* Excl update part */ /* Exclusive/none -> shared case */ if (cur_old_count == nr_old_roots && cur_new_count < nr_new_roots) { /* Exclusive -> shared */ if (cur_old_count != 0) { qg->excl -= num_bytes; qg->excl_cmpr -= num_bytes; dirty = true; } } /* Shared -> exclusive/none case */ if (cur_old_count < nr_old_roots && cur_new_count == nr_new_roots) { /* Shared->exclusive */ if (cur_new_count != 0) { qg->excl += num_bytes; qg->excl_cmpr += num_bytes; dirty = true; } } /* Exclusive/none -> exclusive/none case */ if (cur_old_count == nr_old_roots && cur_new_count == nr_new_roots) { if (cur_old_count == 0) { /* None -> exclusive/none */ if (cur_new_count != 0) { /* None -> exclusive */ qg->excl += num_bytes; qg->excl_cmpr += num_bytes; dirty = true; } /* None -> none, nothing changed */ } else { /* Exclusive -> exclusive/none */ if (cur_new_count == 0) { /* Exclusive -> none */ qg->excl -= num_bytes; qg->excl_cmpr -= num_bytes; dirty = true; } /* Exclusive -> exclusive, nothing changed */ } } |
c05f9429e
|
2350 |
|
823ae5b8e
|
2351 2352 2353 2354 2355 |
if (dirty) qgroup_dirty(fs_info, qg); } return 0; } |
5edfd9fdc
|
2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 |
/* * Check if the @roots potentially is a list of fs tree roots * * Return 0 for definitely not a fs/subvol tree roots ulist * Return 1 for possible fs/subvol tree roots in the list (considering an empty * one as well) */ static int maybe_fs_roots(struct ulist *roots) { struct ulist_node *unode; struct ulist_iterator uiter; /* Empty one, still possible for fs roots */ if (!roots || roots->nnodes == 0) return 1; ULIST_ITER_INIT(&uiter); unode = ulist_next(roots, &uiter); if (!unode) return 1; /* * If it contains fs tree roots, then it must belong to fs/subvol * trees. * If it contains a non-fs tree, it won't be shared with fs/subvol trees. */ return is_fstree(unode->val); } |
8696d7604
|
2384 2385 2386 |
int btrfs_qgroup_account_extent(struct btrfs_trans_handle *trans, u64 bytenr, u64 num_bytes, struct ulist *old_roots, struct ulist *new_roots) |
550d7a2ed
|
2387 |
{ |
8696d7604
|
2388 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
550d7a2ed
|
2389 2390 2391 2392 2393 2394 |
struct ulist *qgroups = NULL; struct ulist *tmp = NULL; u64 seq; u64 nr_new_roots = 0; u64 nr_old_roots = 0; int ret = 0; |
26ef8493e
|
2395 2396 2397 2398 |
/* * If quotas get disabled meanwhile, the resouces need to be freed and * we can't just exit here. */ |
81353d50f
|
2399 |
if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) |
26ef8493e
|
2400 |
goto out_free; |
81353d50f
|
2401 |
|
5edfd9fdc
|
2402 2403 2404 |
if (new_roots) { if (!maybe_fs_roots(new_roots)) goto out_free; |
550d7a2ed
|
2405 |
nr_new_roots = new_roots->nnodes; |
5edfd9fdc
|
2406 2407 2408 2409 |
} if (old_roots) { if (!maybe_fs_roots(old_roots)) goto out_free; |
550d7a2ed
|
2410 |
nr_old_roots = old_roots->nnodes; |
5edfd9fdc
|
2411 2412 2413 2414 2415 |
} /* Quick exit, either not fs tree roots, or won't affect any qgroup */ if (nr_old_roots == 0 && nr_new_roots == 0) goto out_free; |
550d7a2ed
|
2416 |
|
550d7a2ed
|
2417 |
BUG_ON(!fs_info->quota_root); |
c9f6f3cd1
|
2418 2419 |
trace_btrfs_qgroup_account_extent(fs_info, trans->transid, bytenr, num_bytes, nr_old_roots, nr_new_roots); |
0f5dcf8de
|
2420 |
|
550d7a2ed
|
2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 |
qgroups = ulist_alloc(GFP_NOFS); if (!qgroups) { ret = -ENOMEM; goto out_free; } tmp = ulist_alloc(GFP_NOFS); if (!tmp) { ret = -ENOMEM; goto out_free; } mutex_lock(&fs_info->qgroup_rescan_lock); if (fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN) { if (fs_info->qgroup_rescan_progress.objectid <= bytenr) { mutex_unlock(&fs_info->qgroup_rescan_lock); ret = 0; goto out_free; } } mutex_unlock(&fs_info->qgroup_rescan_lock); spin_lock(&fs_info->qgroup_lock); seq = fs_info->qgroup_seq; /* Update old refcnts using old_roots */ ret = qgroup_update_refcnt(fs_info, old_roots, tmp, qgroups, seq, UPDATE_OLD); if (ret < 0) goto out; /* Update new refcnts using new_roots */ ret = qgroup_update_refcnt(fs_info, new_roots, tmp, qgroups, seq, UPDATE_NEW); if (ret < 0) goto out; qgroup_update_counters(fs_info, qgroups, nr_old_roots, nr_new_roots, num_bytes, seq); /* * Bump qgroup_seq to avoid seq overlap */ fs_info->qgroup_seq += max(nr_old_roots, nr_new_roots) + 1; out: spin_unlock(&fs_info->qgroup_lock); out_free: ulist_free(tmp); ulist_free(qgroups); ulist_free(old_roots); ulist_free(new_roots); return ret; } |
460fb20a4
|
2473 |
int btrfs_qgroup_account_extents(struct btrfs_trans_handle *trans) |
550d7a2ed
|
2474 |
{ |
460fb20a4
|
2475 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
550d7a2ed
|
2476 2477 2478 2479 |
struct btrfs_qgroup_extent_record *record; struct btrfs_delayed_ref_root *delayed_refs; struct ulist *new_roots = NULL; struct rb_node *node; |
c337e7b02
|
2480 |
u64 num_dirty_extents = 0; |
9086db86e
|
2481 |
u64 qgroup_to_skip; |
550d7a2ed
|
2482 2483 2484 |
int ret = 0; delayed_refs = &trans->transaction->delayed_refs; |
9086db86e
|
2485 |
qgroup_to_skip = delayed_refs->qgroup_to_skip; |
550d7a2ed
|
2486 2487 2488 |
while ((node = rb_first(&delayed_refs->dirty_extent_root))) { record = rb_entry(node, struct btrfs_qgroup_extent_record, node); |
c337e7b02
|
2489 |
num_dirty_extents++; |
bc074524e
|
2490 |
trace_btrfs_qgroup_account_extents(fs_info, record); |
0f5dcf8de
|
2491 |
|
550d7a2ed
|
2492 2493 |
if (!ret) { /* |
d1b8b94a2
|
2494 2495 2496 2497 2498 2499 2500 |
* Old roots should be searched when inserting qgroup * extent record */ if (WARN_ON(!record->old_roots)) { /* Search commit root to find old_roots */ ret = btrfs_find_all_roots(NULL, fs_info, record->bytenr, 0, |
c995ab3cd
|
2501 |
&record->old_roots, false); |
d1b8b94a2
|
2502 2503 2504 |
if (ret < 0) goto cleanup; } |
1418bae1c
|
2505 2506 2507 2508 2509 |
/* Free the reserved data space */ btrfs_qgroup_free_refroot(fs_info, record->data_rsv_refroot, record->data_rsv, BTRFS_QGROUP_RSV_DATA); |
d1b8b94a2
|
2510 |
/* |
de47c9d3f
|
2511 |
* Use SEQ_LAST as time_seq to do special search, which |
550d7a2ed
|
2512 2513 2514 2515 |
* doesn't lock tree or delayed_refs and search current * root. It's safe inside commit_transaction(). */ ret = btrfs_find_all_roots(trans, fs_info, |
c995ab3cd
|
2516 |
record->bytenr, SEQ_LAST, &new_roots, false); |
550d7a2ed
|
2517 2518 |
if (ret < 0) goto cleanup; |
d1b8b94a2
|
2519 |
if (qgroup_to_skip) { |
9086db86e
|
2520 |
ulist_del(new_roots, qgroup_to_skip, 0); |
d1b8b94a2
|
2521 2522 2523 |
ulist_del(record->old_roots, qgroup_to_skip, 0); } |
8696d7604
|
2524 2525 2526 2527 |
ret = btrfs_qgroup_account_extent(trans, record->bytenr, record->num_bytes, record->old_roots, new_roots); |
550d7a2ed
|
2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 |
record->old_roots = NULL; new_roots = NULL; } cleanup: ulist_free(record->old_roots); ulist_free(new_roots); new_roots = NULL; rb_erase(node, &delayed_refs->dirty_extent_root); kfree(record); } |
c337e7b02
|
2539 2540 |
trace_qgroup_num_dirty_extents(fs_info, trans->transid, num_dirty_extents); |
550d7a2ed
|
2541 2542 |
return ret; } |
fcebe4562
|
2543 |
/* |
bed92eae2
|
2544 2545 |
* called from commit_transaction. Writes all changed qgroups to disk. */ |
280f8bd2c
|
2546 |
int btrfs_run_qgroups(struct btrfs_trans_handle *trans) |
bed92eae2
|
2547 |
{ |
280f8bd2c
|
2548 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
bed92eae2
|
2549 |
int ret = 0; |
e3b0edd29
|
2550 |
if (!fs_info->quota_root) |
5d23515be
|
2551 |
return ret; |
bed92eae2
|
2552 2553 2554 2555 2556 2557 2558 2559 |
spin_lock(&fs_info->qgroup_lock); while (!list_empty(&fs_info->dirty_qgroups)) { struct btrfs_qgroup *qgroup; qgroup = list_first_entry(&fs_info->dirty_qgroups, struct btrfs_qgroup, dirty); list_del_init(&qgroup->dirty); spin_unlock(&fs_info->qgroup_lock); |
3e07e9a09
|
2560 |
ret = update_qgroup_info_item(trans, qgroup); |
bed92eae2
|
2561 2562 |
if (ret) fs_info->qgroup_flags |= |
d3001ed3a
|
2563 |
BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; |
ac8a866af
|
2564 |
ret = update_qgroup_limit_item(trans, qgroup); |
d3001ed3a
|
2565 2566 |
if (ret) fs_info->qgroup_flags |= |
bed92eae2
|
2567 2568 2569 |
BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; spin_lock(&fs_info->qgroup_lock); } |
afcdd129e
|
2570 |
if (test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) |
bed92eae2
|
2571 2572 2573 2574 |
fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_ON; else fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_ON; spin_unlock(&fs_info->qgroup_lock); |
2e980acdd
|
2575 |
ret = update_qgroup_status_item(trans); |
bed92eae2
|
2576 2577 |
if (ret) fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; |
bed92eae2
|
2578 2579 2580 2581 |
return ret; } /* |
013276101
|
2582 |
* Copy the accounting information between qgroups. This is necessary |
918c2ee10
|
2583 2584 2585 |
* when a snapshot or a subvolume is created. Throwing an error will * cause a transaction abort so we take extra care here to only error * when a readonly fs is a reasonable outcome. |
bed92eae2
|
2586 |
*/ |
a93774225
|
2587 2588 |
int btrfs_qgroup_inherit(struct btrfs_trans_handle *trans, u64 srcid, u64 objectid, struct btrfs_qgroup_inherit *inherit) |
bed92eae2
|
2589 2590 2591 2592 |
{ int ret = 0; int i; u64 *i_qgroups; |
e88439deb
|
2593 |
bool committing = false; |
a93774225
|
2594 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
552f0329c
|
2595 |
struct btrfs_root *quota_root; |
bed92eae2
|
2596 2597 |
struct btrfs_qgroup *srcgroup; struct btrfs_qgroup *dstgroup; |
cbab8ade5
|
2598 |
bool need_rescan = false; |
bed92eae2
|
2599 |
u32 level_size = 0; |
3f5e2d3b3
|
2600 |
u64 nums; |
bed92eae2
|
2601 |
|
e88439deb
|
2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 |
/* * There are only two callers of this function. * * One in create_subvol() in the ioctl context, which needs to hold * the qgroup_ioctl_lock. * * The other one in create_pending_snapshot() where no other qgroup * code can modify the fs as they all need to either start a new trans * or hold a trans handler, thus we don't need to hold * qgroup_ioctl_lock. * This would avoid long and complex lock chain and make lockdep happy. */ spin_lock(&fs_info->trans_lock); if (trans->transaction->state == TRANS_STATE_COMMIT_DOING) committing = true; spin_unlock(&fs_info->trans_lock); if (!committing) mutex_lock(&fs_info->qgroup_ioctl_lock); |
afcdd129e
|
2621 |
if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) |
f2f6ed3d5
|
2622 |
goto out; |
bed92eae2
|
2623 |
|
552f0329c
|
2624 |
quota_root = fs_info->quota_root; |
f2f6ed3d5
|
2625 2626 2627 2628 |
if (!quota_root) { ret = -EINVAL; goto out; } |
bed92eae2
|
2629 |
|
3f5e2d3b3
|
2630 2631 2632 2633 2634 2635 |
if (inherit) { i_qgroups = (u64 *)(inherit + 1); nums = inherit->num_qgroups + 2 * inherit->num_ref_copies + 2 * inherit->num_excl_copies; for (i = 0; i < nums; ++i) { srcgroup = find_qgroup_rb(fs_info, *i_qgroups); |
09870d277
|
2636 |
|
918c2ee10
|
2637 2638 2639 2640 2641 2642 2643 |
/* * Zero out invalid groups so we can ignore * them later. */ if (!srcgroup || ((srcgroup->qgroupid >> 48) <= (objectid >> 48))) *i_qgroups = 0ULL; |
3f5e2d3b3
|
2644 2645 2646 |
++i_qgroups; } } |
bed92eae2
|
2647 2648 2649 2650 2651 2652 |
/* * create a tracking group for the subvol itself */ ret = add_qgroup_item(trans, quota_root, objectid); if (ret) goto out; |
bed92eae2
|
2653 2654 2655 2656 2657 |
/* * add qgroup to all inherited groups */ if (inherit) { i_qgroups = (u64 *)(inherit + 1); |
918c2ee10
|
2658 2659 2660 |
for (i = 0; i < inherit->num_qgroups; ++i, ++i_qgroups) { if (*i_qgroups == 0) continue; |
711169c40
|
2661 2662 |
ret = add_qgroup_relation_item(trans, objectid, *i_qgroups); |
918c2ee10
|
2663 |
if (ret && ret != -EEXIST) |
bed92eae2
|
2664 |
goto out; |
711169c40
|
2665 2666 |
ret = add_qgroup_relation_item(trans, *i_qgroups, objectid); |
918c2ee10
|
2667 |
if (ret && ret != -EEXIST) |
bed92eae2
|
2668 |
goto out; |
bed92eae2
|
2669 |
} |
918c2ee10
|
2670 |
ret = 0; |
bed92eae2
|
2671 2672 2673 2674 2675 2676 |
} spin_lock(&fs_info->qgroup_lock); dstgroup = add_qgroup_rb(fs_info, objectid); |
57a5a8820
|
2677 2678 |
if (IS_ERR(dstgroup)) { ret = PTR_ERR(dstgroup); |
bed92eae2
|
2679 |
goto unlock; |
57a5a8820
|
2680 |
} |
bed92eae2
|
2681 |
|
e8c8541ac
|
2682 |
if (inherit && inherit->flags & BTRFS_QGROUP_INHERIT_SET_LIMITS) { |
e8c8541ac
|
2683 2684 2685 2686 2687 |
dstgroup->lim_flags = inherit->lim.flags; dstgroup->max_rfer = inherit->lim.max_rfer; dstgroup->max_excl = inherit->lim.max_excl; dstgroup->rsv_rfer = inherit->lim.rsv_rfer; dstgroup->rsv_excl = inherit->lim.rsv_excl; |
1510e71c6
|
2688 |
|
ac8a866af
|
2689 |
ret = update_qgroup_limit_item(trans, dstgroup); |
1510e71c6
|
2690 2691 |
if (ret) { fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; |
5d163e0e6
|
2692 2693 2694 |
btrfs_info(fs_info, "unable to update quota limit for %llu", dstgroup->qgroupid); |
1510e71c6
|
2695 2696 |
goto unlock; } |
e8c8541ac
|
2697 |
} |
bed92eae2
|
2698 2699 |
if (srcid) { srcgroup = find_qgroup_rb(fs_info, srcid); |
f3a87f1b0
|
2700 |
if (!srcgroup) |
bed92eae2
|
2701 |
goto unlock; |
fcebe4562
|
2702 2703 2704 2705 2706 2707 |
/* * We call inherit after we clone the root in order to make sure * our counts don't go crazy, so at this point the only * difference between the two roots should be the root node. */ |
c8389d4c0
|
2708 |
level_size = fs_info->nodesize; |
fcebe4562
|
2709 2710 2711 2712 |
dstgroup->rfer = srcgroup->rfer; dstgroup->rfer_cmpr = srcgroup->rfer_cmpr; dstgroup->excl = level_size; dstgroup->excl_cmpr = level_size; |
bed92eae2
|
2713 2714 |
srcgroup->excl = level_size; srcgroup->excl_cmpr = level_size; |
3eeb4d597
|
2715 2716 2717 2718 2719 2720 2721 |
/* inherit the limit info */ dstgroup->lim_flags = srcgroup->lim_flags; dstgroup->max_rfer = srcgroup->max_rfer; dstgroup->max_excl = srcgroup->max_excl; dstgroup->rsv_rfer = srcgroup->rsv_rfer; dstgroup->rsv_excl = srcgroup->rsv_excl; |
bed92eae2
|
2722 2723 2724 |
qgroup_dirty(fs_info, dstgroup); qgroup_dirty(fs_info, srcgroup); } |
f3a87f1b0
|
2725 |
if (!inherit) |
bed92eae2
|
2726 2727 2728 2729 |
goto unlock; i_qgroups = (u64 *)(inherit + 1); for (i = 0; i < inherit->num_qgroups; ++i) { |
918c2ee10
|
2730 |
if (*i_qgroups) { |
0b246afa6
|
2731 |
ret = add_relation_rb(fs_info, objectid, *i_qgroups); |
918c2ee10
|
2732 2733 2734 |
if (ret) goto unlock; } |
bed92eae2
|
2735 |
++i_qgroups; |
cbab8ade5
|
2736 2737 2738 2739 2740 2741 2742 |
/* * If we're doing a snapshot, and adding the snapshot to a new * qgroup, the numbers are guaranteed to be incorrect. */ if (srcid) need_rescan = true; |
bed92eae2
|
2743 |
} |
918c2ee10
|
2744 |
for (i = 0; i < inherit->num_ref_copies; ++i, i_qgroups += 2) { |
bed92eae2
|
2745 2746 |
struct btrfs_qgroup *src; struct btrfs_qgroup *dst; |
918c2ee10
|
2747 2748 |
if (!i_qgroups[0] || !i_qgroups[1]) continue; |
bed92eae2
|
2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 |
src = find_qgroup_rb(fs_info, i_qgroups[0]); dst = find_qgroup_rb(fs_info, i_qgroups[1]); if (!src || !dst) { ret = -EINVAL; goto unlock; } dst->rfer = src->rfer - level_size; dst->rfer_cmpr = src->rfer_cmpr - level_size; |
cbab8ade5
|
2759 2760 2761 |
/* Manually tweaking numbers certainly needs a rescan */ need_rescan = true; |
bed92eae2
|
2762 |
} |
918c2ee10
|
2763 |
for (i = 0; i < inherit->num_excl_copies; ++i, i_qgroups += 2) { |
bed92eae2
|
2764 2765 |
struct btrfs_qgroup *src; struct btrfs_qgroup *dst; |
918c2ee10
|
2766 2767 |
if (!i_qgroups[0] || !i_qgroups[1]) continue; |
bed92eae2
|
2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 |
src = find_qgroup_rb(fs_info, i_qgroups[0]); dst = find_qgroup_rb(fs_info, i_qgroups[1]); if (!src || !dst) { ret = -EINVAL; goto unlock; } dst->excl = src->excl + level_size; dst->excl_cmpr = src->excl_cmpr + level_size; |
cbab8ade5
|
2778 |
need_rescan = true; |
bed92eae2
|
2779 2780 2781 2782 |
} unlock: spin_unlock(&fs_info->qgroup_lock); |
49e5fb462
|
2783 2784 |
if (!ret) ret = btrfs_sysfs_add_one_qgroup(fs_info, dstgroup); |
bed92eae2
|
2785 |
out: |
e88439deb
|
2786 2787 |
if (!committing) mutex_unlock(&fs_info->qgroup_ioctl_lock); |
cbab8ade5
|
2788 2789 |
if (need_rescan) fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; |
bed92eae2
|
2790 2791 |
return ret; } |
adca4d945
|
2792 |
static bool qgroup_check_limits(const struct btrfs_qgroup *qg, u64 num_bytes) |
003d7c59e
|
2793 2794 |
{ if ((qg->lim_flags & BTRFS_QGROUP_LIMIT_MAX_RFER) && |
dba213242
|
2795 |
qgroup_rsv_total(qg) + (s64)qg->rfer + num_bytes > qg->max_rfer) |
003d7c59e
|
2796 2797 2798 |
return false; if ((qg->lim_flags & BTRFS_QGROUP_LIMIT_MAX_EXCL) && |
dba213242
|
2799 |
qgroup_rsv_total(qg) + (s64)qg->excl + num_bytes > qg->max_excl) |
003d7c59e
|
2800 2801 2802 2803 |
return false; return true; } |
dba213242
|
2804 2805 |
static int qgroup_reserve(struct btrfs_root *root, u64 num_bytes, bool enforce, enum btrfs_qgroup_rsv_type type) |
bed92eae2
|
2806 |
{ |
bed92eae2
|
2807 2808 2809 2810 |
struct btrfs_qgroup *qgroup; struct btrfs_fs_info *fs_info = root->fs_info; u64 ref_root = root->root_key.objectid; int ret = 0; |
bed92eae2
|
2811 2812 2813 2814 2815 2816 2817 2818 |
struct ulist_node *unode; struct ulist_iterator uiter; if (!is_fstree(ref_root)) return 0; if (num_bytes == 0) return 0; |
f29efe292
|
2819 2820 2821 2822 |
if (test_bit(BTRFS_FS_QUOTA_OVERRIDE, &fs_info->flags) && capable(CAP_SYS_RESOURCE)) enforce = false; |
bed92eae2
|
2823 |
spin_lock(&fs_info->qgroup_lock); |
e3b0edd29
|
2824 |
if (!fs_info->quota_root) |
bed92eae2
|
2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 |
goto out; qgroup = find_qgroup_rb(fs_info, ref_root); if (!qgroup) goto out; /* * in a first step, we check all affected qgroups if any limits would * be exceeded */ |
1e8f91586
|
2835 2836 |
ulist_reinit(fs_info->qgroup_ulist); ret = ulist_add(fs_info->qgroup_ulist, qgroup->qgroupid, |
a1840b502
|
2837 |
qgroup_to_aux(qgroup), GFP_ATOMIC); |
3c97185c6
|
2838 2839 |
if (ret < 0) goto out; |
bed92eae2
|
2840 |
ULIST_ITER_INIT(&uiter); |
1e8f91586
|
2841 |
while ((unode = ulist_next(fs_info->qgroup_ulist, &uiter))) { |
bed92eae2
|
2842 2843 |
struct btrfs_qgroup *qg; struct btrfs_qgroup_list *glist; |
ef2fff64f
|
2844 |
qg = unode_aux_to_qgroup(unode); |
bed92eae2
|
2845 |
|
adca4d945
|
2846 |
if (enforce && !qgroup_check_limits(qg, num_bytes)) { |
bed92eae2
|
2847 |
ret = -EDQUOT; |
720f1e206
|
2848 2849 |
goto out; } |
bed92eae2
|
2850 2851 |
list_for_each_entry(glist, &qg->groups, next_group) { |
1e8f91586
|
2852 2853 |
ret = ulist_add(fs_info->qgroup_ulist, glist->group->qgroupid, |
a1840b502
|
2854 |
qgroup_to_aux(glist->group), GFP_ATOMIC); |
3c97185c6
|
2855 2856 |
if (ret < 0) goto out; |
bed92eae2
|
2857 2858 |
} } |
3c97185c6
|
2859 |
ret = 0; |
bed92eae2
|
2860 2861 2862 2863 |
/* * no limits exceeded, now record the reservation into all qgroups */ ULIST_ITER_INIT(&uiter); |
1e8f91586
|
2864 |
while ((unode = ulist_next(fs_info->qgroup_ulist, &uiter))) { |
bed92eae2
|
2865 |
struct btrfs_qgroup *qg; |
ef2fff64f
|
2866 |
qg = unode_aux_to_qgroup(unode); |
bed92eae2
|
2867 |
|
64ee4e751
|
2868 |
qgroup_rsv_add(fs_info, qg, num_bytes, type); |
bed92eae2
|
2869 2870 2871 2872 |
} out: spin_unlock(&fs_info->qgroup_lock); |
bed92eae2
|
2873 2874 |
return ret; } |
e1211d0e8
|
2875 2876 2877 2878 2879 2880 2881 2882 2883 |
/* * Free @num_bytes of reserved space with @type for qgroup. (Normally level 0 * qgroup). * * Will handle all higher level qgroup too. * * NOTE: If @num_bytes is (u64)-1, this means to free all bytes of this qgroup. * This special case is only used for META_PERTRANS type. */ |
297d750b9
|
2884 |
void btrfs_qgroup_free_refroot(struct btrfs_fs_info *fs_info, |
d4e5c9205
|
2885 2886 |
u64 ref_root, u64 num_bytes, enum btrfs_qgroup_rsv_type type) |
bed92eae2
|
2887 |
{ |
bed92eae2
|
2888 |
struct btrfs_qgroup *qgroup; |
bed92eae2
|
2889 2890 |
struct ulist_node *unode; struct ulist_iterator uiter; |
3c97185c6
|
2891 |
int ret = 0; |
bed92eae2
|
2892 2893 2894 2895 2896 2897 |
if (!is_fstree(ref_root)) return; if (num_bytes == 0) return; |
e1211d0e8
|
2898 2899 2900 2901 |
if (num_bytes == (u64)-1 && type != BTRFS_QGROUP_RSV_META_PERTRANS) { WARN(1, "%s: Invalid type to free", __func__); return; } |
bed92eae2
|
2902 |
spin_lock(&fs_info->qgroup_lock); |
e3b0edd29
|
2903 |
if (!fs_info->quota_root) |
bed92eae2
|
2904 2905 2906 2907 2908 |
goto out; qgroup = find_qgroup_rb(fs_info, ref_root); if (!qgroup) goto out; |
e1211d0e8
|
2909 |
if (num_bytes == (u64)-1) |
8287475a2
|
2910 2911 2912 2913 |
/* * We're freeing all pertrans rsv, get reserved value from * level 0 qgroup as real num_bytes to free. */ |
e1211d0e8
|
2914 |
num_bytes = qgroup->rsv.values[type]; |
1e8f91586
|
2915 2916 |
ulist_reinit(fs_info->qgroup_ulist); ret = ulist_add(fs_info->qgroup_ulist, qgroup->qgroupid, |
a1840b502
|
2917 |
qgroup_to_aux(qgroup), GFP_ATOMIC); |
3c97185c6
|
2918 2919 |
if (ret < 0) goto out; |
bed92eae2
|
2920 |
ULIST_ITER_INIT(&uiter); |
1e8f91586
|
2921 |
while ((unode = ulist_next(fs_info->qgroup_ulist, &uiter))) { |
bed92eae2
|
2922 2923 |
struct btrfs_qgroup *qg; struct btrfs_qgroup_list *glist; |
ef2fff64f
|
2924 |
qg = unode_aux_to_qgroup(unode); |
bed92eae2
|
2925 |
|
64ee4e751
|
2926 |
qgroup_rsv_release(fs_info, qg, num_bytes, type); |
bed92eae2
|
2927 2928 |
list_for_each_entry(glist, &qg->groups, next_group) { |
1e8f91586
|
2929 2930 |
ret = ulist_add(fs_info->qgroup_ulist, glist->group->qgroupid, |
a1840b502
|
2931 |
qgroup_to_aux(glist->group), GFP_ATOMIC); |
3c97185c6
|
2932 2933 |
if (ret < 0) goto out; |
bed92eae2
|
2934 2935 2936 2937 2938 |
} } out: spin_unlock(&fs_info->qgroup_lock); |
bed92eae2
|
2939 |
} |
2f2320360
|
2940 |
/* |
ff3d27a04
|
2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 |
* Check if the leaf is the last leaf. Which means all node pointers * are at their last position. */ static bool is_last_leaf(struct btrfs_path *path) { int i; for (i = 1; i < BTRFS_MAX_LEVEL && path->nodes[i]; i++) { if (path->slots[i] != btrfs_header_nritems(path->nodes[i]) - 1) return false; } return true; } /* |
2f2320360
|
2956 |
* returns < 0 on error, 0 when more leafs are to be scanned. |
3393168d2
|
2957 |
* returns 1 when done. |
2f2320360
|
2958 |
*/ |
62088ca74
|
2959 2960 |
static int qgroup_rescan_leaf(struct btrfs_trans_handle *trans, struct btrfs_path *path) |
2f2320360
|
2961 |
{ |
62088ca74
|
2962 |
struct btrfs_fs_info *fs_info = trans->fs_info; |
2f2320360
|
2963 |
struct btrfs_key found; |
0a0e8b893
|
2964 |
struct extent_buffer *scratch_leaf = NULL; |
2f2320360
|
2965 |
struct ulist *roots = NULL; |
fcebe4562
|
2966 |
u64 num_bytes; |
ff3d27a04
|
2967 |
bool done; |
2f2320360
|
2968 2969 |
int slot; int ret; |
2f2320360
|
2970 2971 2972 2973 |
mutex_lock(&fs_info->qgroup_rescan_lock); ret = btrfs_search_slot_for_read(fs_info->extent_root, &fs_info->qgroup_rescan_progress, path, 1, 0); |
ab8d0fc48
|
2974 2975 2976 2977 2978 |
btrfs_debug(fs_info, "current progress key (%llu %u %llu), search_slot ret %d", fs_info->qgroup_rescan_progress.objectid, fs_info->qgroup_rescan_progress.type, fs_info->qgroup_rescan_progress.offset, ret); |
2f2320360
|
2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 |
if (ret) { /* * The rescan is about to end, we will not be scanning any * further blocks. We cannot unset the RESCAN flag here, because * we want to commit the transaction if everything went well. * To make the live accounting work in this phase, we set our * scan progress pointer such that every real extent objectid * will be smaller. */ fs_info->qgroup_rescan_progress.objectid = (u64)-1; btrfs_release_path(path); mutex_unlock(&fs_info->qgroup_rescan_lock); return ret; } |
ff3d27a04
|
2994 |
done = is_last_leaf(path); |
2f2320360
|
2995 2996 2997 2998 |
btrfs_item_key_to_cpu(path->nodes[0], &found, btrfs_header_nritems(path->nodes[0]) - 1); fs_info->qgroup_rescan_progress.objectid = found.objectid + 1; |
0a0e8b893
|
2999 3000 3001 3002 3003 3004 |
scratch_leaf = btrfs_clone_extent_buffer(path->nodes[0]); if (!scratch_leaf) { ret = -ENOMEM; mutex_unlock(&fs_info->qgroup_rescan_lock); goto out; } |
2f2320360
|
3005 3006 3007 3008 3009 3010 |
slot = path->slots[0]; btrfs_release_path(path); mutex_unlock(&fs_info->qgroup_rescan_lock); for (; slot < btrfs_header_nritems(scratch_leaf); ++slot) { btrfs_item_key_to_cpu(scratch_leaf, &found, slot); |
3a6d75e84
|
3011 3012 |
if (found.type != BTRFS_EXTENT_ITEM_KEY && found.type != BTRFS_METADATA_ITEM_KEY) |
2f2320360
|
3013 |
continue; |
3a6d75e84
|
3014 |
if (found.type == BTRFS_METADATA_ITEM_KEY) |
da17066c4
|
3015 |
num_bytes = fs_info->nodesize; |
3a6d75e84
|
3016 3017 |
else num_bytes = found.offset; |
fcebe4562
|
3018 |
ret = btrfs_find_all_roots(NULL, fs_info, found.objectid, 0, |
c995ab3cd
|
3019 |
&roots, false); |
2f2320360
|
3020 3021 |
if (ret < 0) goto out; |
9d220c95f
|
3022 |
/* For rescan, just pass old_roots as NULL */ |
8696d7604
|
3023 3024 |
ret = btrfs_qgroup_account_extent(trans, found.objectid, num_bytes, NULL, roots); |
9d220c95f
|
3025 |
if (ret < 0) |
fcebe4562
|
3026 |
goto out; |
2f2320360
|
3027 |
} |
2f2320360
|
3028 |
out: |
df4497146
|
3029 |
if (scratch_leaf) |
0a0e8b893
|
3030 |
free_extent_buffer(scratch_leaf); |
2f2320360
|
3031 |
|
6f7de19ed
|
3032 |
if (done && !ret) { |
ff3d27a04
|
3033 |
ret = 1; |
6f7de19ed
|
3034 3035 |
fs_info->qgroup_rescan_progress.objectid = (u64)-1; } |
2f2320360
|
3036 3037 |
return ret; } |
d458b0540
|
3038 |
static void btrfs_qgroup_rescan_worker(struct btrfs_work *work) |
2f2320360
|
3039 |
{ |
b382a324b
|
3040 3041 |
struct btrfs_fs_info *fs_info = container_of(work, struct btrfs_fs_info, qgroup_rescan_work); |
2f2320360
|
3042 3043 |
struct btrfs_path *path; struct btrfs_trans_handle *trans = NULL; |
2f2320360
|
3044 |
int err = -ENOMEM; |
53b7cde9d
|
3045 |
int ret = 0; |
2f2320360
|
3046 3047 3048 3049 |
path = btrfs_alloc_path(); if (!path) goto out; |
b6debf15d
|
3050 3051 3052 3053 3054 3055 |
/* * Rescan should only search for commit root, and any later difference * should be recorded by qgroup */ path->search_commit_root = 1; path->skip_locking = 1; |
2f2320360
|
3056 3057 |
err = 0; |
7343dd61f
|
3058 |
while (!err && !btrfs_fs_closing(fs_info)) { |
2f2320360
|
3059 3060 3061 3062 3063 |
trans = btrfs_start_transaction(fs_info->fs_root, 0); if (IS_ERR(trans)) { err = PTR_ERR(trans); break; } |
afcdd129e
|
3064 |
if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) { |
2f2320360
|
3065 3066 |
err = -EINTR; } else { |
62088ca74
|
3067 |
err = qgroup_rescan_leaf(trans, path); |
2f2320360
|
3068 3069 |
} if (err > 0) |
3a45bb207
|
3070 |
btrfs_commit_transaction(trans); |
2f2320360
|
3071 |
else |
3a45bb207
|
3072 |
btrfs_end_transaction(trans); |
2f2320360
|
3073 3074 3075 |
} out: |
2f2320360
|
3076 |
btrfs_free_path(path); |
2f2320360
|
3077 3078 |
mutex_lock(&fs_info->qgroup_rescan_lock); |
3393168d2
|
3079 |
if (err > 0 && |
2f2320360
|
3080 3081 3082 3083 3084 3085 |
fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT) { fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; } else if (err < 0) { fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; } mutex_unlock(&fs_info->qgroup_rescan_lock); |
53b7cde9d
|
3086 |
/* |
013276101
|
3087 |
* only update status, since the previous part has already updated the |
53b7cde9d
|
3088 3089 3090 3091 3092 |
* qgroup info. */ trans = btrfs_start_transaction(fs_info->quota_root, 1); if (IS_ERR(trans)) { err = PTR_ERR(trans); |
13fc1d271
|
3093 |
trans = NULL; |
53b7cde9d
|
3094 |
btrfs_err(fs_info, |
913e15357
|
3095 |
"fail to start transaction for status update: %d", |
53b7cde9d
|
3096 |
err); |
53b7cde9d
|
3097 |
} |
13fc1d271
|
3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 |
mutex_lock(&fs_info->qgroup_rescan_lock); if (!btrfs_fs_closing(fs_info)) fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_RESCAN; if (trans) { ret = update_qgroup_status_item(trans); if (ret < 0) { err = ret; btrfs_err(fs_info, "fail to update qgroup status: %d", err); } |
53b7cde9d
|
3109 |
} |
13fc1d271
|
3110 3111 3112 3113 3114 3115 |
fs_info->qgroup_rescan_running = false; complete_all(&fs_info->qgroup_rescan_completion); mutex_unlock(&fs_info->qgroup_rescan_lock); if (!trans) return; |
3a45bb207
|
3116 |
btrfs_end_transaction(trans); |
53b7cde9d
|
3117 |
|
7343dd61f
|
3118 3119 3120 |
if (btrfs_fs_closing(fs_info)) { btrfs_info(fs_info, "qgroup scan paused"); } else if (err >= 0) { |
efe120a06
|
3121 |
btrfs_info(fs_info, "qgroup scan completed%s", |
3393168d2
|
3122 |
err > 0 ? " (inconsistency flag cleared)" : ""); |
2f2320360
|
3123 |
} else { |
efe120a06
|
3124 |
btrfs_err(fs_info, "qgroup scan failed with %d", err); |
2f2320360
|
3125 3126 |
} } |
b382a324b
|
3127 3128 3129 3130 3131 3132 3133 |
/* * Checks that (a) no rescan is running and (b) quota is enabled. Allocates all * memory required for the rescan context. */ static int qgroup_rescan_init(struct btrfs_fs_info *fs_info, u64 progress_objectid, int init_flags) |
2f2320360
|
3134 3135 |
{ int ret = 0; |
2f2320360
|
3136 |
|
9593bf496
|
3137 3138 |
if (!init_flags) { /* we're resuming qgroup rescan at mount time */ |
e4e7ede73
|
3139 3140 |
if (!(fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN)) { |
9593bf496
|
3141 |
btrfs_warn(fs_info, |
37d02592f
|
3142 |
"qgroup rescan init failed, qgroup rescan is not queued"); |
e4e7ede73
|
3143 3144 3145 |
ret = -EINVAL; } else if (!(fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_ON)) { |
9593bf496
|
3146 |
btrfs_warn(fs_info, |
37d02592f
|
3147 |
"qgroup rescan init failed, qgroup is not enabled"); |
e4e7ede73
|
3148 3149 3150 3151 3152 |
ret = -EINVAL; } if (ret) return ret; |
b382a324b
|
3153 |
} |
2f2320360
|
3154 3155 |
mutex_lock(&fs_info->qgroup_rescan_lock); |
b382a324b
|
3156 3157 |
if (init_flags) { |
9593bf496
|
3158 3159 3160 |
if (fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN) { btrfs_warn(fs_info, "qgroup rescan is already in progress"); |
b382a324b
|
3161 |
ret = -EINPROGRESS; |
9593bf496
|
3162 3163 3164 3165 |
} else if (!(fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_ON)) { btrfs_warn(fs_info, "qgroup rescan init failed, qgroup is not enabled"); |
b382a324b
|
3166 |
ret = -EINVAL; |
9593bf496
|
3167 |
} |
b382a324b
|
3168 3169 |
if (ret) { |
b382a324b
|
3170 |
mutex_unlock(&fs_info->qgroup_rescan_lock); |
9593bf496
|
3171 |
return ret; |
b382a324b
|
3172 |
} |
b382a324b
|
3173 |
fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_RESCAN; |
2f2320360
|
3174 |
} |
2f2320360
|
3175 3176 |
memset(&fs_info->qgroup_rescan_progress, 0, sizeof(fs_info->qgroup_rescan_progress)); |
b382a324b
|
3177 |
fs_info->qgroup_rescan_progress.objectid = progress_objectid; |
190631f1c
|
3178 |
init_completion(&fs_info->qgroup_rescan_completion); |
b382a324b
|
3179 |
mutex_unlock(&fs_info->qgroup_rescan_lock); |
fc97fab0e
|
3180 3181 |
btrfs_init_work(&fs_info->qgroup_rescan_work, btrfs_qgroup_rescan_worker, NULL, NULL); |
b382a324b
|
3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 |
return 0; } static void qgroup_rescan_zero_tracking(struct btrfs_fs_info *fs_info) { struct rb_node *n; struct btrfs_qgroup *qgroup; spin_lock(&fs_info->qgroup_lock); |
2f2320360
|
3192 3193 3194 3195 3196 3197 3198 |
/* clear all current qgroup tracking information */ for (n = rb_first(&fs_info->qgroup_tree); n; n = rb_next(n)) { qgroup = rb_entry(n, struct btrfs_qgroup, node); qgroup->rfer = 0; qgroup->rfer_cmpr = 0; qgroup->excl = 0; qgroup->excl_cmpr = 0; |
9c7b0c2e8
|
3199 |
qgroup_dirty(fs_info, qgroup); |
2f2320360
|
3200 3201 |
} spin_unlock(&fs_info->qgroup_lock); |
b382a324b
|
3202 |
} |
2f2320360
|
3203 |
|
b382a324b
|
3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 |
int btrfs_qgroup_rescan(struct btrfs_fs_info *fs_info) { int ret = 0; struct btrfs_trans_handle *trans; ret = qgroup_rescan_init(fs_info, 0, 1); if (ret) return ret; /* * We have set the rescan_progress to 0, which means no more * delayed refs will be accounted by btrfs_qgroup_account_ref. * However, btrfs_qgroup_account_ref may be right after its call * to btrfs_find_all_roots, in which case it would still do the * accounting. * To solve this, we're committing the transaction, which will * ensure we run all delayed refs and only after that, we are * going to clear all tracking information for a clean start. */ trans = btrfs_join_transaction(fs_info->fs_root); if (IS_ERR(trans)) { fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_RESCAN; return PTR_ERR(trans); } |
3a45bb207
|
3230 |
ret = btrfs_commit_transaction(trans); |
b382a324b
|
3231 3232 3233 3234 3235 3236 |
if (ret) { fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_RESCAN; return ret; } qgroup_rescan_zero_tracking(fs_info); |
d61acbbf5
|
3237 3238 |
mutex_lock(&fs_info->qgroup_rescan_lock); fs_info->qgroup_rescan_running = true; |
fc97fab0e
|
3239 3240 |
btrfs_queue_work(fs_info->qgroup_rescan_workers, &fs_info->qgroup_rescan_work); |
d61acbbf5
|
3241 |
mutex_unlock(&fs_info->qgroup_rescan_lock); |
2f2320360
|
3242 3243 3244 |
return 0; } |
57254b6eb
|
3245 |
|
d06f23d6a
|
3246 3247 |
int btrfs_qgroup_wait_for_completion(struct btrfs_fs_info *fs_info, bool interruptible) |
57254b6eb
|
3248 3249 3250 3251 3252 |
{ int running; int ret = 0; mutex_lock(&fs_info->qgroup_rescan_lock); |
d2c609b83
|
3253 |
running = fs_info->qgroup_rescan_running; |
57254b6eb
|
3254 |
mutex_unlock(&fs_info->qgroup_rescan_lock); |
d06f23d6a
|
3255 3256 3257 3258 |
if (!running) return 0; if (interruptible) |
57254b6eb
|
3259 3260 |
ret = wait_for_completion_interruptible( &fs_info->qgroup_rescan_completion); |
d06f23d6a
|
3261 3262 |
else wait_for_completion(&fs_info->qgroup_rescan_completion); |
57254b6eb
|
3263 3264 3265 |
return ret; } |
b382a324b
|
3266 3267 3268 3269 3270 3271 3272 3273 |
/* * this is only called from open_ctree where we're still single threaded, thus * locking is omitted here. */ void btrfs_qgroup_rescan_resume(struct btrfs_fs_info *fs_info) { |
d61acbbf5
|
3274 3275 3276 |
if (fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN) { mutex_lock(&fs_info->qgroup_rescan_lock); fs_info->qgroup_rescan_running = true; |
fc97fab0e
|
3277 3278 |
btrfs_queue_work(fs_info->qgroup_rescan_workers, &fs_info->qgroup_rescan_work); |
d61acbbf5
|
3279 3280 |
mutex_unlock(&fs_info->qgroup_rescan_lock); } |
b382a324b
|
3281 |
} |
524725537
|
3282 |
|
263da812e
|
3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 |
#define rbtree_iterate_from_safe(node, next, start) \ for (node = start; node && ({ next = rb_next(node); 1;}); node = next) static int qgroup_unreserve_range(struct btrfs_inode *inode, struct extent_changeset *reserved, u64 start, u64 len) { struct rb_node *node; struct rb_node *next; struct ulist_node *entry = NULL; int ret = 0; node = reserved->range_changed.root.rb_node; while (node) { entry = rb_entry(node, struct ulist_node, rb_node); if (entry->val < start) node = node->rb_right; else if (entry) node = node->rb_left; else break; } /* Empty changeset */ if (!entry) return 0; if (entry->val > start && rb_prev(&entry->rb_node)) entry = rb_entry(rb_prev(&entry->rb_node), struct ulist_node, rb_node); rbtree_iterate_from_safe(node, next, &entry->rb_node) { u64 entry_start; u64 entry_end; u64 entry_len; int clear_ret; entry = rb_entry(node, struct ulist_node, rb_node); entry_start = entry->val; entry_end = entry->aux; entry_len = entry_end - entry_start + 1; if (entry_start >= start + len) break; if (entry_start + entry_len <= start) continue; /* * Now the entry is in [start, start + len), revert the * EXTENT_QGROUP_RESERVED bit. */ clear_ret = clear_extent_bits(&inode->io_tree, entry_start, entry_end, EXTENT_QGROUP_RESERVED); if (!ret && clear_ret < 0) ret = clear_ret; ulist_del(&reserved->range_changed, entry->val, entry->aux); if (likely(reserved->bytes_changed >= entry_len)) { reserved->bytes_changed -= entry_len; } else { WARN_ON(1); reserved->bytes_changed = 0; } } return ret; } |
524725537
|
3349 |
/* |
c53e96536
|
3350 |
* Try to free some space for qgroup. |
524725537
|
3351 |
* |
c53e96536
|
3352 3353 3354 3355 3356 |
* For qgroup, there are only 3 ways to free qgroup space: * - Flush nodatacow write * Any nodatacow write will free its reserved data space at run_delalloc_range(). * In theory, we should only flush nodatacow inodes, but it's not yet * possible, so we need to flush the whole root. |
524725537
|
3357 |
* |
c53e96536
|
3358 3359 3360 3361 |
* - Wait for ordered extents * When ordered extents are finished, their reserved metadata is finally * converted to per_trans status, which can be freed by later commit * transaction. |
524725537
|
3362 |
* |
c53e96536
|
3363 3364 3365 3366 |
* - Commit transaction * This would free the meta_per_trans space. * In theory this shouldn't provide much space, but any more qgroup space * is needed. |
524725537
|
3367 |
*/ |
c53e96536
|
3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 |
static int try_flush_qgroup(struct btrfs_root *root) { struct btrfs_trans_handle *trans; int ret; /* * We don't want to run flush again and again, so if there is a running * one, we won't try to start a new flush, but exit directly. */ if (test_and_set_bit(BTRFS_ROOT_QGROUP_FLUSHING, &root->state)) { wait_event(root->qgroup_flush_wait, !test_bit(BTRFS_ROOT_QGROUP_FLUSHING, &root->state)); return 0; } ret = btrfs_start_delalloc_snapshot(root); if (ret < 0) goto out; btrfs_wait_ordered_extents(root, U64_MAX, 0, (u64)-1); trans = btrfs_join_transaction(root); if (IS_ERR(trans)) { ret = PTR_ERR(trans); goto out; } ret = btrfs_commit_transaction(trans); out: clear_bit(BTRFS_ROOT_QGROUP_FLUSHING, &root->state); wake_up(&root->qgroup_flush_wait); return ret; } static int qgroup_reserve_data(struct btrfs_inode *inode, |
364ecf365
|
3402 3403 |
struct extent_changeset **reserved_ret, u64 start, u64 len) |
524725537
|
3404 |
{ |
7661a3e03
|
3405 |
struct btrfs_root *root = inode->root; |
364ecf365
|
3406 |
struct extent_changeset *reserved; |
263da812e
|
3407 |
bool new_reserved = false; |
364ecf365
|
3408 3409 |
u64 orig_reserved; u64 to_reserve; |
524725537
|
3410 |
int ret; |
afcdd129e
|
3411 |
if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &root->fs_info->flags) || |
4fd786e6c
|
3412 |
!is_fstree(root->root_key.objectid) || len == 0) |
524725537
|
3413 |
return 0; |
364ecf365
|
3414 3415 3416 3417 |
/* @reserved parameter is mandatory for qgroup */ if (WARN_ON(!reserved_ret)) return -EINVAL; if (!*reserved_ret) { |
263da812e
|
3418 |
new_reserved = true; |
364ecf365
|
3419 3420 3421 3422 3423 3424 3425 |
*reserved_ret = extent_changeset_alloc(); if (!*reserved_ret) return -ENOMEM; } reserved = *reserved_ret; /* Record already reserved space */ orig_reserved = reserved->bytes_changed; |
7661a3e03
|
3426 |
ret = set_record_extent_bits(&inode->io_tree, start, |
364ecf365
|
3427 3428 3429 3430 |
start + len -1, EXTENT_QGROUP_RESERVED, reserved); /* Newly reserved space */ to_reserve = reserved->bytes_changed - orig_reserved; |
7661a3e03
|
3431 |
trace_btrfs_qgroup_reserve_data(&inode->vfs_inode, start, len, |
364ecf365
|
3432 |
to_reserve, QGROUP_RESERVE); |
524725537
|
3433 |
if (ret < 0) |
263da812e
|
3434 |
goto out; |
dba213242
|
3435 |
ret = qgroup_reserve(root, to_reserve, true, BTRFS_QGROUP_RSV_DATA); |
524725537
|
3436 3437 |
if (ret < 0) goto cleanup; |
524725537
|
3438 3439 3440 |
return ret; cleanup: |
263da812e
|
3441 3442 3443 3444 3445 3446 3447 |
qgroup_unreserve_range(inode, reserved, start, len); out: if (new_reserved) { extent_changeset_release(reserved); kfree(reserved); *reserved_ret = NULL; } |
524725537
|
3448 3449 |
return ret; } |
f695fdcef
|
3450 |
|
c53e96536
|
3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 |
/* * Reserve qgroup space for range [start, start + len). * * This function will either reserve space from related qgroups or do nothing * if the range is already reserved. * * Return 0 for successful reservation * Return <0 for error (including -EQUOT) * * NOTE: This function may sleep for memory allocation, dirty page flushing and * commit transaction. So caller should not hold any dirty page locked. */ int btrfs_qgroup_reserve_data(struct btrfs_inode *inode, struct extent_changeset **reserved_ret, u64 start, u64 len) { int ret; ret = qgroup_reserve_data(inode, reserved_ret, start, len); if (ret <= 0 && ret != -EDQUOT) return ret; ret = try_flush_qgroup(inode->root); if (ret < 0) return ret; return qgroup_reserve_data(inode, reserved_ret, start, len); } |
bc42bda22
|
3478 |
/* Free ranges specified by @reserved, normally in error path */ |
df2cfd131
|
3479 |
static int qgroup_free_reserved_data(struct btrfs_inode *inode, |
bc42bda22
|
3480 3481 |
struct extent_changeset *reserved, u64 start, u64 len) { |
df2cfd131
|
3482 |
struct btrfs_root *root = inode->root; |
bc42bda22
|
3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 |
struct ulist_node *unode; struct ulist_iterator uiter; struct extent_changeset changeset; int freed = 0; int ret; extent_changeset_init(&changeset); len = round_up(start + len, root->fs_info->sectorsize); start = round_down(start, root->fs_info->sectorsize); ULIST_ITER_INIT(&uiter); while ((unode = ulist_next(&reserved->range_changed, &uiter))) { u64 range_start = unode->val; /* unode->aux is the inclusive end */ u64 range_len = unode->aux - range_start + 1; u64 free_start; u64 free_len; extent_changeset_release(&changeset); /* Only free range in range [start, start + len) */ if (range_start >= start + len || range_start + range_len <= start) continue; free_start = max(range_start, start); free_len = min(start + len, range_start + range_len) - free_start; /* * TODO: To also modify reserved->ranges_reserved to reflect * the modification. * * However as long as we free qgroup reserved according to * EXTENT_QGROUP_RESERVED, we won't double free. * So not need to rush. */ |
df2cfd131
|
3518 3519 |
ret = clear_record_extent_bits(&inode->io_tree, free_start, free_start + free_len - 1, |
bc42bda22
|
3520 3521 3522 3523 3524 |
EXTENT_QGROUP_RESERVED, &changeset); if (ret < 0) goto out; freed += changeset.bytes_changed; } |
4fd786e6c
|
3525 |
btrfs_qgroup_free_refroot(root->fs_info, root->root_key.objectid, freed, |
d4e5c9205
|
3526 |
BTRFS_QGROUP_RSV_DATA); |
bc42bda22
|
3527 3528 3529 3530 3531 |
ret = freed; out: extent_changeset_release(&changeset); return ret; } |
8769af96c
|
3532 |
static int __btrfs_qgroup_release_data(struct btrfs_inode *inode, |
bc42bda22
|
3533 3534 |
struct extent_changeset *reserved, u64 start, u64 len, int free) |
f695fdcef
|
3535 3536 |
{ struct extent_changeset changeset; |
81fb6f77a
|
3537 |
int trace_op = QGROUP_RELEASE; |
f695fdcef
|
3538 |
int ret; |
8769af96c
|
3539 |
if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &inode->root->fs_info->flags)) |
3628b4ca6
|
3540 |
return 0; |
bc42bda22
|
3541 3542 3543 |
/* In release case, we shouldn't have @reserved */ WARN_ON(!free && reserved); if (free && reserved) |
8769af96c
|
3544 |
return qgroup_free_reserved_data(inode, reserved, start, len); |
364ecf365
|
3545 |
extent_changeset_init(&changeset); |
8769af96c
|
3546 3547 |
ret = clear_record_extent_bits(&inode->io_tree, start, start + len -1, EXTENT_QGROUP_RESERVED, &changeset); |
f695fdcef
|
3548 3549 |
if (ret < 0) goto out; |
d51ea5dd2
|
3550 |
if (free) |
81fb6f77a
|
3551 |
trace_op = QGROUP_FREE; |
8769af96c
|
3552 |
trace_btrfs_qgroup_release_data(&inode->vfs_inode, start, len, |
81fb6f77a
|
3553 |
changeset.bytes_changed, trace_op); |
d51ea5dd2
|
3554 |
if (free) |
8769af96c
|
3555 3556 |
btrfs_qgroup_free_refroot(inode->root->fs_info, inode->root->root_key.objectid, |
d4e5c9205
|
3557 |
changeset.bytes_changed, BTRFS_QGROUP_RSV_DATA); |
7bc329c18
|
3558 |
ret = changeset.bytes_changed; |
f695fdcef
|
3559 |
out: |
364ecf365
|
3560 |
extent_changeset_release(&changeset); |
f695fdcef
|
3561 3562 3563 3564 3565 3566 3567 3568 |
return ret; } /* * Free a reserved space range from io_tree and related qgroups * * Should be called when a range of pages get invalidated before reaching disk. * Or for error cleanup case. |
bc42bda22
|
3569 3570 |
* if @reserved is given, only reserved range in [@start, @start + @len) will * be freed. |
f695fdcef
|
3571 3572 3573 3574 3575 |
* * For data written to disk, use btrfs_qgroup_release_data(). * * NOTE: This function may sleep for memory allocation. */ |
8b8a979f1
|
3576 |
int btrfs_qgroup_free_data(struct btrfs_inode *inode, |
bc42bda22
|
3577 |
struct extent_changeset *reserved, u64 start, u64 len) |
f695fdcef
|
3578 |
{ |
8b8a979f1
|
3579 |
return __btrfs_qgroup_release_data(inode, reserved, start, len, 1); |
f695fdcef
|
3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 |
} /* * Release a reserved space range from io_tree only. * * Should be called when a range of pages get written to disk and corresponding * FILE_EXTENT is inserted into corresponding root. * * Since new qgroup accounting framework will only update qgroup numbers at * commit_transaction() time, its reserved space shouldn't be freed from * related qgroups. * * But we should release the range from io_tree, to allow further write to be * COWed. * * NOTE: This function may sleep for memory allocation. */ |
72b7d15bf
|
3597 |
int btrfs_qgroup_release_data(struct btrfs_inode *inode, u64 start, u64 len) |
f695fdcef
|
3598 |
{ |
72b7d15bf
|
3599 |
return __btrfs_qgroup_release_data(inode, NULL, start, len, 0); |
f695fdcef
|
3600 |
} |
55eeaf057
|
3601 |
|
8287475a2
|
3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 |
static void add_root_meta_rsv(struct btrfs_root *root, int num_bytes, enum btrfs_qgroup_rsv_type type) { if (type != BTRFS_QGROUP_RSV_META_PREALLOC && type != BTRFS_QGROUP_RSV_META_PERTRANS) return; if (num_bytes == 0) return; spin_lock(&root->qgroup_meta_rsv_lock); if (type == BTRFS_QGROUP_RSV_META_PREALLOC) root->qgroup_meta_rsv_prealloc += num_bytes; else root->qgroup_meta_rsv_pertrans += num_bytes; spin_unlock(&root->qgroup_meta_rsv_lock); } static int sub_root_meta_rsv(struct btrfs_root *root, int num_bytes, enum btrfs_qgroup_rsv_type type) { if (type != BTRFS_QGROUP_RSV_META_PREALLOC && type != BTRFS_QGROUP_RSV_META_PERTRANS) return 0; if (num_bytes == 0) return 0; spin_lock(&root->qgroup_meta_rsv_lock); if (type == BTRFS_QGROUP_RSV_META_PREALLOC) { num_bytes = min_t(u64, root->qgroup_meta_rsv_prealloc, num_bytes); root->qgroup_meta_rsv_prealloc -= num_bytes; } else { num_bytes = min_t(u64, root->qgroup_meta_rsv_pertrans, num_bytes); root->qgroup_meta_rsv_pertrans -= num_bytes; } spin_unlock(&root->qgroup_meta_rsv_lock); return num_bytes; } |
c53e96536
|
3641 |
static int qgroup_reserve_meta(struct btrfs_root *root, int num_bytes, |
733e03a0b
|
3642 |
enum btrfs_qgroup_rsv_type type, bool enforce) |
55eeaf057
|
3643 |
{ |
0b246afa6
|
3644 |
struct btrfs_fs_info *fs_info = root->fs_info; |
55eeaf057
|
3645 |
int ret; |
0b246afa6
|
3646 |
if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags) || |
4fd786e6c
|
3647 |
!is_fstree(root->root_key.objectid) || num_bytes == 0) |
55eeaf057
|
3648 |
return 0; |
0b246afa6
|
3649 |
BUG_ON(num_bytes != round_down(num_bytes, fs_info->nodesize)); |
fd2b007ea
|
3650 |
trace_qgroup_meta_reserve(root, (s64)num_bytes, type); |
733e03a0b
|
3651 |
ret = qgroup_reserve(root, num_bytes, enforce, type); |
55eeaf057
|
3652 3653 |
if (ret < 0) return ret; |
8287475a2
|
3654 3655 3656 3657 3658 3659 3660 3661 3662 |
/* * Record what we have reserved into root. * * To avoid quota disabled->enabled underflow. * In that case, we may try to free space we haven't reserved * (since quota was disabled), so record what we reserved into root. * And ensure later release won't underflow this number. */ add_root_meta_rsv(root, num_bytes, type); |
55eeaf057
|
3663 3664 |
return ret; } |
c53e96536
|
3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678 |
int __btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes, enum btrfs_qgroup_rsv_type type, bool enforce) { int ret; ret = qgroup_reserve_meta(root, num_bytes, type, enforce); if (ret <= 0 && ret != -EDQUOT) return ret; ret = try_flush_qgroup(root); if (ret < 0) return ret; return qgroup_reserve_meta(root, num_bytes, type, enforce); } |
733e03a0b
|
3679 |
void btrfs_qgroup_free_meta_all_pertrans(struct btrfs_root *root) |
55eeaf057
|
3680 |
{ |
0b246afa6
|
3681 |
struct btrfs_fs_info *fs_info = root->fs_info; |
55eeaf057
|
3682 |
|
0b246afa6
|
3683 |
if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags) || |
4fd786e6c
|
3684 |
!is_fstree(root->root_key.objectid)) |
55eeaf057
|
3685 |
return; |
e1211d0e8
|
3686 |
/* TODO: Update trace point to handle such free */ |
4ee0d8832
|
3687 |
trace_qgroup_meta_free_all_pertrans(root); |
e1211d0e8
|
3688 |
/* Special value -1 means to free all reserved space */ |
4fd786e6c
|
3689 |
btrfs_qgroup_free_refroot(fs_info, root->root_key.objectid, (u64)-1, |
733e03a0b
|
3690 |
BTRFS_QGROUP_RSV_META_PERTRANS); |
55eeaf057
|
3691 |
} |
733e03a0b
|
3692 3693 |
void __btrfs_qgroup_free_meta(struct btrfs_root *root, int num_bytes, enum btrfs_qgroup_rsv_type type) |
55eeaf057
|
3694 |
{ |
0b246afa6
|
3695 3696 3697 |
struct btrfs_fs_info *fs_info = root->fs_info; if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags) || |
4fd786e6c
|
3698 |
!is_fstree(root->root_key.objectid)) |
55eeaf057
|
3699 |
return; |
8287475a2
|
3700 3701 3702 3703 3704 3705 |
/* * reservation for META_PREALLOC can happen before quota is enabled, * which can lead to underflow. * Here ensure we will only free what we really have reserved. */ num_bytes = sub_root_meta_rsv(root, num_bytes, type); |
0b246afa6
|
3706 |
BUG_ON(num_bytes != round_down(num_bytes, fs_info->nodesize)); |
fd2b007ea
|
3707 |
trace_qgroup_meta_reserve(root, -(s64)num_bytes, type); |
4fd786e6c
|
3708 3709 |
btrfs_qgroup_free_refroot(fs_info, root->root_key.objectid, num_bytes, type); |
55eeaf057
|
3710 |
} |
56fa9d076
|
3711 |
|
64cfaef63
|
3712 3713 3714 |
static void qgroup_convert_meta(struct btrfs_fs_info *fs_info, u64 ref_root, int num_bytes) { |
64cfaef63
|
3715 3716 3717 3718 3719 3720 3721 |
struct btrfs_qgroup *qgroup; struct ulist_node *unode; struct ulist_iterator uiter; int ret = 0; if (num_bytes == 0) return; |
e3b0edd29
|
3722 |
if (!fs_info->quota_root) |
64cfaef63
|
3723 3724 3725 3726 3727 3728 3729 3730 |
return; spin_lock(&fs_info->qgroup_lock); qgroup = find_qgroup_rb(fs_info, ref_root); if (!qgroup) goto out; ulist_reinit(fs_info->qgroup_ulist); ret = ulist_add(fs_info->qgroup_ulist, qgroup->qgroupid, |
a1840b502
|
3731 |
qgroup_to_aux(qgroup), GFP_ATOMIC); |
64cfaef63
|
3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 |
if (ret < 0) goto out; ULIST_ITER_INIT(&uiter); while ((unode = ulist_next(fs_info->qgroup_ulist, &uiter))) { struct btrfs_qgroup *qg; struct btrfs_qgroup_list *glist; qg = unode_aux_to_qgroup(unode); qgroup_rsv_release(fs_info, qg, num_bytes, BTRFS_QGROUP_RSV_META_PREALLOC); qgroup_rsv_add(fs_info, qg, num_bytes, BTRFS_QGROUP_RSV_META_PERTRANS); list_for_each_entry(glist, &qg->groups, next_group) { ret = ulist_add(fs_info->qgroup_ulist, glist->group->qgroupid, |
a1840b502
|
3748 |
qgroup_to_aux(glist->group), GFP_ATOMIC); |
64cfaef63
|
3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 |
if (ret < 0) goto out; } } out: spin_unlock(&fs_info->qgroup_lock); } void btrfs_qgroup_convert_reserved_meta(struct btrfs_root *root, int num_bytes) { struct btrfs_fs_info *fs_info = root->fs_info; if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags) || |
4fd786e6c
|
3762 |
!is_fstree(root->root_key.objectid)) |
64cfaef63
|
3763 |
return; |
8287475a2
|
3764 3765 3766 |
/* Same as btrfs_qgroup_free_meta_prealloc() */ num_bytes = sub_root_meta_rsv(root, num_bytes, BTRFS_QGROUP_RSV_META_PREALLOC); |
4ee0d8832
|
3767 |
trace_qgroup_meta_convert(root, num_bytes); |
4fd786e6c
|
3768 |
qgroup_convert_meta(fs_info, root->root_key.objectid, num_bytes); |
64cfaef63
|
3769 |
} |
56fa9d076
|
3770 |
/* |
013276101
|
3771 |
* Check qgroup reserved space leaking, normally at destroy inode |
56fa9d076
|
3772 3773 |
* time */ |
cfdd45921
|
3774 |
void btrfs_qgroup_check_reserved_leak(struct btrfs_inode *inode) |
56fa9d076
|
3775 3776 3777 3778 3779 |
{ struct extent_changeset changeset; struct ulist_node *unode; struct ulist_iterator iter; int ret; |
364ecf365
|
3780 |
extent_changeset_init(&changeset); |
cfdd45921
|
3781 |
ret = clear_record_extent_bits(&inode->io_tree, 0, (u64)-1, |
f734c44a1
|
3782 |
EXTENT_QGROUP_RESERVED, &changeset); |
56fa9d076
|
3783 3784 3785 3786 |
WARN_ON(ret < 0); if (WARN_ON(changeset.bytes_changed)) { ULIST_ITER_INIT(&iter); |
53d323599
|
3787 |
while ((unode = ulist_next(&changeset.range_changed, &iter))) { |
cfdd45921
|
3788 3789 3790 |
btrfs_warn(inode->root->fs_info, "leaking qgroup reserved space, ino: %llu, start: %llu, end: %llu", btrfs_ino(inode), unode->val, unode->aux); |
56fa9d076
|
3791 |
} |
cfdd45921
|
3792 3793 |
btrfs_qgroup_free_refroot(inode->root->fs_info, inode->root->root_key.objectid, |
d4e5c9205
|
3794 |
changeset.bytes_changed, BTRFS_QGROUP_RSV_DATA); |
0b08e1f4f
|
3795 |
|
56fa9d076
|
3796 |
} |
364ecf365
|
3797 |
extent_changeset_release(&changeset); |
56fa9d076
|
3798 |
} |
370a11b81
|
3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 |
void btrfs_qgroup_init_swapped_blocks( struct btrfs_qgroup_swapped_blocks *swapped_blocks) { int i; spin_lock_init(&swapped_blocks->lock); for (i = 0; i < BTRFS_MAX_LEVEL; i++) swapped_blocks->blocks[i] = RB_ROOT; swapped_blocks->swapped = false; } /* * Delete all swapped blocks record of @root. * Every record here means we skipped a full subtree scan for qgroup. * * Gets called when committing one transaction. */ void btrfs_qgroup_clean_swapped_blocks(struct btrfs_root *root) { struct btrfs_qgroup_swapped_blocks *swapped_blocks; int i; swapped_blocks = &root->swapped_blocks; spin_lock(&swapped_blocks->lock); if (!swapped_blocks->swapped) goto out; for (i = 0; i < BTRFS_MAX_LEVEL; i++) { struct rb_root *cur_root = &swapped_blocks->blocks[i]; struct btrfs_qgroup_swapped_block *entry; struct btrfs_qgroup_swapped_block *next; rbtree_postorder_for_each_entry_safe(entry, next, cur_root, node) kfree(entry); swapped_blocks->blocks[i] = RB_ROOT; } swapped_blocks->swapped = false; out: spin_unlock(&swapped_blocks->lock); } /* * Add subtree roots record into @subvol_root. * * @subvol_root: tree root of the subvolume tree get swapped * @bg: block group under balance * @subvol_parent/slot: pointer to the subtree root in subvolume tree * @reloc_parent/slot: pointer to the subtree root in reloc tree * BOTH POINTERS ARE BEFORE TREE SWAP * @last_snapshot: last snapshot generation of the subvolume tree */ int btrfs_qgroup_add_swapped_blocks(struct btrfs_trans_handle *trans, struct btrfs_root *subvol_root, |
32da5386d
|
3854 |
struct btrfs_block_group *bg, |
370a11b81
|
3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 |
struct extent_buffer *subvol_parent, int subvol_slot, struct extent_buffer *reloc_parent, int reloc_slot, u64 last_snapshot) { struct btrfs_fs_info *fs_info = subvol_root->fs_info; struct btrfs_qgroup_swapped_blocks *blocks = &subvol_root->swapped_blocks; struct btrfs_qgroup_swapped_block *block; struct rb_node **cur; struct rb_node *parent = NULL; int level = btrfs_header_level(subvol_parent) - 1; int ret = 0; if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) return 0; if (btrfs_node_ptr_generation(subvol_parent, subvol_slot) > btrfs_node_ptr_generation(reloc_parent, reloc_slot)) { btrfs_err_rl(fs_info, "%s: bad parameter order, subvol_gen=%llu reloc_gen=%llu", __func__, btrfs_node_ptr_generation(subvol_parent, subvol_slot), btrfs_node_ptr_generation(reloc_parent, reloc_slot)); return -EUCLEAN; } block = kmalloc(sizeof(*block), GFP_NOFS); if (!block) { ret = -ENOMEM; goto out; } /* * @reloc_parent/slot is still before swap, while @block is going to * record the bytenr after swap, so we do the swap here. */ block->subvol_bytenr = btrfs_node_blockptr(reloc_parent, reloc_slot); block->subvol_generation = btrfs_node_ptr_generation(reloc_parent, reloc_slot); block->reloc_bytenr = btrfs_node_blockptr(subvol_parent, subvol_slot); block->reloc_generation = btrfs_node_ptr_generation(subvol_parent, subvol_slot); block->last_snapshot = last_snapshot; block->level = level; |
57949d033
|
3898 3899 3900 3901 3902 3903 3904 |
/* * If we have bg == NULL, we're called from btrfs_recover_relocation(), * no one else can modify tree blocks thus we qgroup will not change * no matter the value of trace_leaf. */ if (bg && bg->flags & BTRFS_BLOCK_GROUP_DATA) |
370a11b81
|
3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 |
block->trace_leaf = true; else block->trace_leaf = false; btrfs_node_key_to_cpu(reloc_parent, &block->first_key, reloc_slot); /* Insert @block into @blocks */ spin_lock(&blocks->lock); cur = &blocks->blocks[level].rb_node; while (*cur) { struct btrfs_qgroup_swapped_block *entry; parent = *cur; entry = rb_entry(parent, struct btrfs_qgroup_swapped_block, node); if (entry->subvol_bytenr < block->subvol_bytenr) { cur = &(*cur)->rb_left; } else if (entry->subvol_bytenr > block->subvol_bytenr) { cur = &(*cur)->rb_right; } else { if (entry->subvol_generation != block->subvol_generation || entry->reloc_bytenr != block->reloc_bytenr || entry->reloc_generation != block->reloc_generation) { /* * Duplicated but mismatch entry found. * Shouldn't happen. * * Marking qgroup inconsistent should be enough * for end users. */ WARN_ON(IS_ENABLED(CONFIG_BTRFS_DEBUG)); ret = -EEXIST; } kfree(block); goto out_unlock; } } rb_link_node(&block->node, parent, cur); rb_insert_color(&block->node, &blocks->blocks[level]); blocks->swapped = true; out_unlock: spin_unlock(&blocks->lock); out: if (ret < 0) fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; return ret; } |
f616f5cd9
|
3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 |
/* * Check if the tree block is a subtree root, and if so do the needed * delayed subtree trace for qgroup. * * This is called during btrfs_cow_block(). */ int btrfs_qgroup_trace_subtree_after_cow(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *subvol_eb) { struct btrfs_fs_info *fs_info = root->fs_info; struct btrfs_qgroup_swapped_blocks *blocks = &root->swapped_blocks; struct btrfs_qgroup_swapped_block *block; struct extent_buffer *reloc_eb = NULL; struct rb_node *node; bool found = false; bool swapped = false; int level = btrfs_header_level(subvol_eb); int ret = 0; int i; if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) return 0; if (!is_fstree(root->root_key.objectid) || !root->reloc_root) return 0; spin_lock(&blocks->lock); if (!blocks->swapped) { spin_unlock(&blocks->lock); return 0; } node = blocks->blocks[level].rb_node; while (node) { block = rb_entry(node, struct btrfs_qgroup_swapped_block, node); if (block->subvol_bytenr < subvol_eb->start) { node = node->rb_left; } else if (block->subvol_bytenr > subvol_eb->start) { node = node->rb_right; } else { found = true; break; } } if (!found) { spin_unlock(&blocks->lock); goto out; } /* Found one, remove it from @blocks first and update blocks->swapped */ rb_erase(&block->node, &blocks->blocks[level]); for (i = 0; i < BTRFS_MAX_LEVEL; i++) { if (RB_EMPTY_ROOT(&blocks->blocks[i])) { swapped = true; break; } } blocks->swapped = swapped; spin_unlock(&blocks->lock); /* Read out reloc subtree root */ reloc_eb = read_tree_block(fs_info, block->reloc_bytenr, block->reloc_generation, block->level, &block->first_key); if (IS_ERR(reloc_eb)) { ret = PTR_ERR(reloc_eb); reloc_eb = NULL; goto free_out; } if (!extent_buffer_uptodate(reloc_eb)) { ret = -EIO; goto free_out; } ret = qgroup_trace_subtree_swap(trans, reloc_eb, subvol_eb, block->last_snapshot, block->trace_leaf); free_out: kfree(block); free_extent_buffer(reloc_eb); out: if (ret < 0) { btrfs_err_rl(fs_info, "failed to account subtree at bytenr %llu: %d", subvol_eb->start, ret); fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT; } return ret; } |
81f7eb00f
|
4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 |
void btrfs_qgroup_destroy_extent_records(struct btrfs_transaction *trans) { struct btrfs_qgroup_extent_record *entry; struct btrfs_qgroup_extent_record *next; struct rb_root *root; root = &trans->delayed_refs.dirty_extent_root; rbtree_postorder_for_each_entry_safe(entry, next, root, node) { ulist_free(entry->old_roots); kfree(entry); } } |