2 * This file is part of UBIFS.
4 * Copyright (C) 2006-2008 Nokia Corporation.
6 * This program is free software; you can redistribute it and/or modify it
7 * under the terms of the GNU General Public License version 2 as published by
8 * the Free Software Foundation.
10 * This program is distributed in the hope that it will be useful, but WITHOUT
11 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
12 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
15 * You should have received a copy of the GNU General Public License along with
16 * this program; if not, write to the Free Software Foundation, Inc., 51
17 * Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
19 * Author: Adrian Hunter
25 * An orphan is an inode number whose inode node has been committed to the index
26 * with a link count of zero. That happens when an open file is deleted
27 * (unlinked) and then a commit is run. In the normal course of events the inode
28 * would be deleted when the file is closed. However in the case of an unclean
29 * unmount, orphans need to be accounted for. After an unclean unmount, the
30 * orphans' inodes must be deleted which means either scanning the entire index
31 * looking for them, or keeping a list on flash somewhere. This unit implements
32 * the latter approach.
34 * The orphan area is a fixed number of LEBs situated between the LPT area and
35 * the main area. The number of orphan area LEBs is specified when the file
36 * system is created. The minimum number is 1. The size of the orphan area
37 * should be so that it can hold the maximum number of orphans that are expected
38 * to ever exist at one time.
40 * The number of orphans that can fit in a LEB is:
42 * (c->leb_size - UBIFS_ORPH_NODE_SZ) / sizeof(__le64)
44 * For example: a 15872 byte LEB can fit 1980 orphans so 1 LEB may be enough.
46 * Orphans are accumulated in a rb-tree. When an inode's link count drops to
47 * zero, the inode number is added to the rb-tree. It is removed from the tree
48 * when the inode is deleted. Any new orphans that are in the orphan tree when
49 * the commit is run, are written to the orphan area in 1 or more orph nodes.
50 * If the orphan area is full, it is consolidated to make space. There is
51 * always enough space because validation prevents the user from creating more
52 * than the maximum number of orphans allowed.
55 #ifdef CONFIG_UBIFS_FS_DEBUG
56 static int dbg_check_orphans(struct ubifs_info *c);
58 #define dbg_check_orphans(c) 0
62 * ubifs_add_orphan - add an orphan.
63 * @c: UBIFS file-system description object
64 * @inum: orphan inode number
66 * Add an orphan. This function is called when an inodes link count drops to
69 int ubifs_add_orphan(struct ubifs_info *c, ino_t inum)
71 struct ubifs_orphan *orphan, *o;
72 struct rb_node **p, *parent = NULL;
74 orphan = kzalloc(sizeof(struct ubifs_orphan), GFP_NOFS);
80 spin_lock(&c->orphan_lock);
81 if (c->tot_orphans >= c->max_orphans) {
82 spin_unlock(&c->orphan_lock);
86 p = &c->orph_tree.rb_node;
89 o = rb_entry(parent, struct ubifs_orphan, rb);
92 else if (inum > o->inum)
95 dbg_err("orphaned twice");
96 spin_unlock(&c->orphan_lock);
103 rb_link_node(&orphan->rb, parent, p);
104 rb_insert_color(&orphan->rb, &c->orph_tree);
105 list_add_tail(&orphan->list, &c->orph_list);
106 list_add_tail(&orphan->new_list, &c->orph_new);
107 spin_unlock(&c->orphan_lock);
108 dbg_gen("ino %lu", inum);
113 * ubifs_delete_orphan - delete an orphan.
114 * @c: UBIFS file-system description object
115 * @inum: orphan inode number
117 * Delete an orphan. This function is called when an inode is deleted.
119 void ubifs_delete_orphan(struct ubifs_info *c, ino_t inum)
121 struct ubifs_orphan *o;
124 spin_lock(&c->orphan_lock);
125 p = c->orph_tree.rb_node;
127 o = rb_entry(p, struct ubifs_orphan, rb);
130 else if (inum > o->inum)
134 spin_unlock(&c->orphan_lock);
135 dbg_gen("deleted twice ino %lu", inum);
139 o->dnext = c->orph_dnext;
141 spin_unlock(&c->orphan_lock);
142 dbg_gen("delete later ino %lu", inum);
145 rb_erase(p, &c->orph_tree);
149 list_del(&o->new_list);
152 spin_unlock(&c->orphan_lock);
154 dbg_gen("inum %lu", inum);
158 spin_unlock(&c->orphan_lock);
159 dbg_err("missing orphan ino %lu", inum);
164 * ubifs_orphan_start_commit - start commit of orphans.
165 * @c: UBIFS file-system description object
167 * Start commit of orphans.
169 int ubifs_orphan_start_commit(struct ubifs_info *c)
171 struct ubifs_orphan *orphan, **last;
173 spin_lock(&c->orphan_lock);
174 last = &c->orph_cnext;
175 list_for_each_entry(orphan, &c->orph_new, new_list) {
176 ubifs_assert(orphan->new);
179 last = &orphan->cnext;
181 *last = orphan->cnext;
182 c->cmt_orphans = c->new_orphans;
184 dbg_cmt("%d orphans to commit", c->cmt_orphans);
185 INIT_LIST_HEAD(&c->orph_new);
186 if (c->tot_orphans == 0)
190 spin_unlock(&c->orphan_lock);
195 * avail_orphs - calculate available space.
196 * @c: UBIFS file-system description object
198 * This function returns the number of orphans that can be written in the
201 static int avail_orphs(struct ubifs_info *c)
203 int avail_lebs, avail, gap;
205 avail_lebs = c->orph_lebs - (c->ohead_lnum - c->orph_first) - 1;
207 ((c->leb_size - UBIFS_ORPH_NODE_SZ) / sizeof(__le64));
208 gap = c->leb_size - c->ohead_offs;
209 if (gap >= UBIFS_ORPH_NODE_SZ + sizeof(__le64))
210 avail += (gap - UBIFS_ORPH_NODE_SZ) / sizeof(__le64);
215 * tot_avail_orphs - calculate total space.
216 * @c: UBIFS file-system description object
218 * This function returns the number of orphans that can be written in half
219 * the total space. That leaves half the space for adding new orphans.
221 static int tot_avail_orphs(struct ubifs_info *c)
223 int avail_lebs, avail;
225 avail_lebs = c->orph_lebs;
227 ((c->leb_size - UBIFS_ORPH_NODE_SZ) / sizeof(__le64));
232 * do_write_orph_node - write a node
233 * @c: UBIFS file-system description object
234 * @len: length of node
235 * @atomic: write atomically
237 * This function writes a node to the orphan head from the orphan buffer. If
238 * %atomic is not zero, then the write is done atomically. On success, %0 is
239 * returned, otherwise a negative error code is returned.
241 static int do_write_orph_node(struct ubifs_info *c, int len, int atomic)
246 ubifs_assert(c->ohead_offs == 0);
247 ubifs_prepare_node(c, c->orph_buf, len, 1);
248 len = ALIGN(len, c->min_io_size);
249 err = ubifs_leb_change(c, c->ohead_lnum, c->orph_buf, len,
252 if (c->ohead_offs == 0) {
253 /* Ensure LEB has been unmapped */
254 err = ubifs_leb_unmap(c, c->ohead_lnum);
258 err = ubifs_write_node(c, c->orph_buf, len, c->ohead_lnum,
259 c->ohead_offs, UBI_SHORTTERM);
265 * write_orph_node - write an orph node
266 * @c: UBIFS file-system description object
267 * @atomic: write atomically
269 * This function builds an orph node from the cnext list and writes it to the
270 * orphan head. On success, %0 is returned, otherwise a negative error code
273 static int write_orph_node(struct ubifs_info *c, int atomic)
275 struct ubifs_orphan *orphan, *cnext;
276 struct ubifs_orph_node *orph;
277 int gap, err, len, cnt, i;
279 ubifs_assert(c->cmt_orphans > 0);
280 gap = c->leb_size - c->ohead_offs;
281 if (gap < UBIFS_ORPH_NODE_SZ + sizeof(__le64)) {
285 if (c->ohead_lnum > c->orph_last) {
287 * We limit the number of orphans so that this should
290 ubifs_err("out of space in orphan area");
294 cnt = (gap - UBIFS_ORPH_NODE_SZ) / sizeof(__le64);
295 if (cnt > c->cmt_orphans)
296 cnt = c->cmt_orphans;
297 len = UBIFS_ORPH_NODE_SZ + cnt * sizeof(__le64);
298 ubifs_assert(c->orph_buf);
300 orph->ch.node_type = UBIFS_ORPH_NODE;
301 spin_lock(&c->orphan_lock);
302 cnext = c->orph_cnext;
303 for (i = 0; i < cnt; i++) {
305 orph->inos[i] = cpu_to_le64(orphan->inum);
306 cnext = orphan->cnext;
307 orphan->cnext = NULL;
309 c->orph_cnext = cnext;
310 c->cmt_orphans -= cnt;
311 spin_unlock(&c->orphan_lock);
313 orph->cmt_no = cpu_to_le64(c->cmt_no);
315 /* Mark the last node of the commit */
316 orph->cmt_no = cpu_to_le64((c->cmt_no) | (1ULL << 63));
317 ubifs_assert(c->ohead_offs + len <= c->leb_size);
318 ubifs_assert(c->ohead_lnum >= c->orph_first);
319 ubifs_assert(c->ohead_lnum <= c->orph_last);
320 err = do_write_orph_node(c, len, atomic);
321 c->ohead_offs += ALIGN(len, c->min_io_size);
322 c->ohead_offs = ALIGN(c->ohead_offs, 8);
327 * write_orph_nodes - write orph nodes until there are no more to commit
328 * @c: UBIFS file-system description object
329 * @atomic: write atomically
331 * This function writes orph nodes for all the orphans to commit. On success,
332 * %0 is returned, otherwise a negative error code is returned.
334 static int write_orph_nodes(struct ubifs_info *c, int atomic)
338 while (c->cmt_orphans > 0) {
339 err = write_orph_node(c, atomic);
346 /* Unmap any unused LEBs after consolidation */
347 lnum = c->ohead_lnum + 1;
348 for (lnum = c->ohead_lnum + 1; lnum <= c->orph_last; lnum++) {
349 err = ubifs_leb_unmap(c, lnum);
358 * consolidate - consolidate the orphan area.
359 * @c: UBIFS file-system description object
361 * This function enables consolidation by putting all the orphans into the list
362 * to commit. The list is in the order that the orphans were added, and the
363 * LEBs are written atomically in order, so at no time can orphans be lost by
364 * an unclean unmount.
366 * This function returns %0 on success and a negative error code on failure.
368 static int consolidate(struct ubifs_info *c)
370 int tot_avail = tot_avail_orphs(c), err = 0;
372 spin_lock(&c->orphan_lock);
373 dbg_cmt("there is space for %d orphans and there are %d",
374 tot_avail, c->tot_orphans);
375 if (c->tot_orphans - c->new_orphans <= tot_avail) {
376 struct ubifs_orphan *orphan, **last;
379 /* Change the cnext list to include all non-new orphans */
380 last = &c->orph_cnext;
381 list_for_each_entry(orphan, &c->orph_list, list) {
385 last = &orphan->cnext;
388 *last = orphan->cnext;
389 ubifs_assert(cnt == c->tot_orphans - c->new_orphans);
390 c->cmt_orphans = cnt;
391 c->ohead_lnum = c->orph_first;
395 * We limit the number of orphans so that this should
398 ubifs_err("out of space in orphan area");
401 spin_unlock(&c->orphan_lock);
406 * commit_orphans - commit orphans.
407 * @c: UBIFS file-system description object
409 * This function commits orphans to flash. On success, %0 is returned,
410 * otherwise a negative error code is returned.
412 static int commit_orphans(struct ubifs_info *c)
414 int avail, atomic = 0, err;
416 ubifs_assert(c->cmt_orphans > 0);
417 avail = avail_orphs(c);
418 if (avail < c->cmt_orphans) {
419 /* Not enough space to write new orphans, so consolidate */
420 err = consolidate(c);
425 err = write_orph_nodes(c, atomic);
430 * erase_deleted - erase the orphans marked for deletion.
431 * @c: UBIFS file-system description object
433 * During commit, the orphans being committed cannot be deleted, so they are
434 * marked for deletion and deleted by this function. Also, the recovery
435 * adds killed orphans to the deletion list, and therefore they are deleted
438 static void erase_deleted(struct ubifs_info *c)
440 struct ubifs_orphan *orphan, *dnext;
442 spin_lock(&c->orphan_lock);
443 dnext = c->orph_dnext;
446 dnext = orphan->dnext;
447 ubifs_assert(!orphan->new);
448 rb_erase(&orphan->rb, &c->orph_tree);
449 list_del(&orphan->list);
451 dbg_gen("deleting orphan ino %lu", orphan->inum);
454 c->orph_dnext = NULL;
455 spin_unlock(&c->orphan_lock);
459 * ubifs_orphan_end_commit - end commit of orphans.
460 * @c: UBIFS file-system description object
462 * End commit of orphans.
464 int ubifs_orphan_end_commit(struct ubifs_info *c)
468 if (c->cmt_orphans != 0) {
469 err = commit_orphans(c);
474 err = dbg_check_orphans(c);
479 * clear_orphans - erase all LEBs used for orphans.
480 * @c: UBIFS file-system description object
482 * If recovery is not required, then the orphans from the previous session
483 * are not needed. This function locates the LEBs used to record
484 * orphans, and un-maps them.
486 static int clear_orphans(struct ubifs_info *c)
490 for (lnum = c->orph_first; lnum <= c->orph_last; lnum++) {
491 err = ubifs_leb_unmap(c, lnum);
495 c->ohead_lnum = c->orph_first;
501 * insert_dead_orphan - insert an orphan.
502 * @c: UBIFS file-system description object
503 * @inum: orphan inode number
505 * This function is a helper to the 'do_kill_orphans()' function. The orphan
506 * must be kept until the next commit, so it is added to the rb-tree and the
509 static int insert_dead_orphan(struct ubifs_info *c, ino_t inum)
511 struct ubifs_orphan *orphan, *o;
512 struct rb_node **p, *parent = NULL;
514 orphan = kzalloc(sizeof(struct ubifs_orphan), GFP_KERNEL);
519 p = &c->orph_tree.rb_node;
522 o = rb_entry(parent, struct ubifs_orphan, rb);
525 else if (inum > o->inum)
528 /* Already added - no problem */
534 rb_link_node(&orphan->rb, parent, p);
535 rb_insert_color(&orphan->rb, &c->orph_tree);
536 list_add_tail(&orphan->list, &c->orph_list);
537 orphan->dnext = c->orph_dnext;
538 c->orph_dnext = orphan;
539 dbg_mnt("ino %lu, new %d, tot %d",
540 inum, c->new_orphans, c->tot_orphans);
545 * do_kill_orphans - remove orphan inodes from the index.
546 * @c: UBIFS file-system description object
548 * @last_cmt_no: cmt_no of last orph node read is passed and returned here
549 * @outofdate: whether the LEB is out of date is returned here
550 * @last_flagged: whether the end orph node is encountered
552 * This function is a helper to the 'kill_orphans()' function. It goes through
553 * every orphan node in a LEB and for every inode number recorded, removes
554 * all keys for that inode from the TNC.
556 static int do_kill_orphans(struct ubifs_info *c, struct ubifs_scan_leb *sleb,
557 unsigned long long *last_cmt_no, int *outofdate,
560 struct ubifs_scan_node *snod;
561 struct ubifs_orph_node *orph;
562 unsigned long long cmt_no;
564 int i, n, err, first = 1;
566 list_for_each_entry(snod, &sleb->nodes, list) {
567 if (snod->type != UBIFS_ORPH_NODE) {
568 ubifs_err("invalid node type %d in orphan area at "
569 "%d:%d", snod->type, sleb->lnum, snod->offs);
570 dbg_dump_node(c, snod->node);
576 /* Check commit number */
577 cmt_no = le64_to_cpu(orph->cmt_no) & LLONG_MAX;
579 * The commit number on the master node may be less, because
580 * of a failed commit. If there are several failed commits in a
581 * row, the commit number written on orph nodes will continue to
582 * increase (because the commit number is adjusted here) even
583 * though the commit number on the master node stays the same
584 * because the master node has not been re-written.
586 if (cmt_no > c->cmt_no)
588 if (cmt_no < *last_cmt_no && *last_flagged) {
590 * The last orph node had a higher commit number and was
591 * flagged as the last written for that commit number.
592 * That makes this orph node, out of date.
595 ubifs_err("out of order commit number %llu in "
596 "orphan node at %d:%d",
597 cmt_no, sleb->lnum, snod->offs);
598 dbg_dump_node(c, snod->node);
601 dbg_rcvry("out of date LEB %d", sleb->lnum);
609 n = (le32_to_cpu(orph->ch.len) - UBIFS_ORPH_NODE_SZ) >> 3;
610 for (i = 0; i < n; i++) {
611 inum = le64_to_cpu(orph->inos[i]);
612 dbg_rcvry("deleting orphaned inode %lu", inum);
613 err = ubifs_tnc_remove_ino(c, inum);
616 err = insert_dead_orphan(c, inum);
621 *last_cmt_no = cmt_no;
622 if (le64_to_cpu(orph->cmt_no) & (1ULL << 63)) {
623 dbg_rcvry("last orph node for commit %llu at %d:%d",
624 cmt_no, sleb->lnum, snod->offs);
634 * kill_orphans - remove all orphan inodes from the index.
635 * @c: UBIFS file-system description object
637 * If recovery is required, then orphan inodes recorded during the previous
638 * session (which ended with an unclean unmount) must be deleted from the index.
639 * This is done by updating the TNC, but since the index is not updated until
640 * the next commit, the LEBs where the orphan information is recorded are not
641 * erased until the next commit.
643 static int kill_orphans(struct ubifs_info *c)
645 unsigned long long last_cmt_no = 0;
646 int lnum, err = 0, outofdate = 0, last_flagged = 0;
648 c->ohead_lnum = c->orph_first;
650 /* Check no-orphans flag and skip this if no orphans */
652 dbg_rcvry("no orphans");
656 * Orph nodes always start at c->orph_first and are written to each
657 * successive LEB in turn. Generally unused LEBs will have been unmapped
658 * but may contain out of date orph nodes if the unmap didn't go
659 * through. In addition, the last orph node written for each commit is
660 * marked (top bit of orph->cmt_no is set to 1). It is possible that
661 * there are orph nodes from the next commit (i.e. the commit did not
662 * complete successfully). In that case, no orphans will have been lost
663 * due to the way that orphans are written, and any orphans added will
664 * be valid orphans anyway and so can be deleted.
666 for (lnum = c->orph_first; lnum <= c->orph_last; lnum++) {
667 struct ubifs_scan_leb *sleb;
669 dbg_rcvry("LEB %d", lnum);
670 sleb = ubifs_scan(c, lnum, 0, c->sbuf);
672 sleb = ubifs_recover_leb(c, lnum, 0, c->sbuf, 0);
678 err = do_kill_orphans(c, sleb, &last_cmt_no, &outofdate,
680 if (err || outofdate) {
681 ubifs_scan_destroy(sleb);
685 c->ohead_lnum = lnum;
686 c->ohead_offs = sleb->endpt;
688 ubifs_scan_destroy(sleb);
694 * ubifs_mount_orphans - delete orphan inodes and erase LEBs that recorded them.
695 * @c: UBIFS file-system description object
696 * @unclean: indicates recovery from unclean unmount
697 * @read_only: indicates read only mount
699 * This function is called when mounting to erase orphans from the previous
700 * session. If UBIFS was not unmounted cleanly, then the inodes recorded as
701 * orphans are deleted.
703 int ubifs_mount_orphans(struct ubifs_info *c, int unclean, int read_only)
707 c->max_orphans = tot_avail_orphs(c);
710 c->orph_buf = vmalloc(c->leb_size);
716 err = kill_orphans(c);
718 err = clear_orphans(c);
723 #ifdef CONFIG_UBIFS_FS_DEBUG
725 struct check_orphan {
731 unsigned long last_ino;
732 unsigned long tot_inos;
733 unsigned long missing;
734 unsigned long long leaf_cnt;
735 struct ubifs_ino_node *node;
739 static int dbg_find_orphan(struct ubifs_info *c, ino_t inum)
741 struct ubifs_orphan *o;
744 spin_lock(&c->orphan_lock);
745 p = c->orph_tree.rb_node;
747 o = rb_entry(p, struct ubifs_orphan, rb);
750 else if (inum > o->inum)
753 spin_unlock(&c->orphan_lock);
757 spin_unlock(&c->orphan_lock);
761 static int dbg_ins_check_orphan(struct rb_root *root, ino_t inum)
763 struct check_orphan *orphan, *o;
764 struct rb_node **p, *parent = NULL;
766 orphan = kzalloc(sizeof(struct check_orphan), GFP_NOFS);
774 o = rb_entry(parent, struct check_orphan, rb);
777 else if (inum > o->inum)
784 rb_link_node(&orphan->rb, parent, p);
785 rb_insert_color(&orphan->rb, root);
789 static int dbg_find_check_orphan(struct rb_root *root, ino_t inum)
791 struct check_orphan *o;
796 o = rb_entry(p, struct check_orphan, rb);
799 else if (inum > o->inum)
807 static void dbg_free_check_tree(struct rb_root *root)
809 struct rb_node *this = root->rb_node;
810 struct check_orphan *o;
814 this = this->rb_left;
816 } else if (this->rb_right) {
817 this = this->rb_right;
820 o = rb_entry(this, struct check_orphan, rb);
821 this = rb_parent(this);
823 if (this->rb_left == &o->rb)
824 this->rb_left = NULL;
826 this->rb_right = NULL;
832 static int dbg_orphan_check(struct ubifs_info *c, struct ubifs_zbranch *zbr,
835 struct check_info *ci = priv;
839 inum = key_inum(c, &zbr->key);
840 if (inum != ci->last_ino) {
841 /* Lowest node type is the inode node, so it comes first */
842 if (key_type(c, &zbr->key) != UBIFS_INO_KEY)
843 ubifs_err("found orphan node ino %lu, type %d", inum,
844 key_type(c, &zbr->key));
847 err = ubifs_tnc_read_node(c, zbr, ci->node);
849 ubifs_err("node read failed, error %d", err);
852 if (ci->node->nlink == 0)
853 /* Must be recorded as an orphan */
854 if (!dbg_find_check_orphan(&ci->root, inum) &&
855 !dbg_find_orphan(c, inum)) {
856 ubifs_err("missing orphan, ino %lu", inum);
864 static int dbg_read_orphans(struct check_info *ci, struct ubifs_scan_leb *sleb)
866 struct ubifs_scan_node *snod;
867 struct ubifs_orph_node *orph;
871 list_for_each_entry(snod, &sleb->nodes, list) {
873 if (snod->type != UBIFS_ORPH_NODE)
876 n = (le32_to_cpu(orph->ch.len) - UBIFS_ORPH_NODE_SZ) >> 3;
877 for (i = 0; i < n; i++) {
878 inum = le64_to_cpu(orph->inos[i]);
879 err = dbg_ins_check_orphan(&ci->root, inum);
887 static int dbg_scan_orphans(struct ubifs_info *c, struct check_info *ci)
891 /* Check no-orphans flag and skip this if no orphans */
895 for (lnum = c->orph_first; lnum <= c->orph_last; lnum++) {
896 struct ubifs_scan_leb *sleb;
898 sleb = ubifs_scan(c, lnum, 0, c->dbg_buf);
904 err = dbg_read_orphans(ci, sleb);
905 ubifs_scan_destroy(sleb);
913 static int dbg_check_orphans(struct ubifs_info *c)
915 struct check_info ci;
918 if (!(ubifs_chk_flags & UBIFS_CHK_ORPH))
926 ci.node = kmalloc(UBIFS_MAX_INO_NODE_SZ, GFP_NOFS);
928 ubifs_err("out of memory");
932 err = dbg_scan_orphans(c, &ci);
936 err = dbg_walk_index(c, &dbg_orphan_check, NULL, &ci);
938 ubifs_err("cannot scan TNC, error %d", err);
943 ubifs_err("%lu missing orphan(s)", ci.missing);
948 dbg_cmt("last inode number is %lu", ci.last_ino);
949 dbg_cmt("total number of inodes is %lu", ci.tot_inos);
950 dbg_cmt("total number of leaf nodes is %llu", ci.leaf_cnt);
953 dbg_free_check_tree(&ci.root);
958 #endif /* CONFIG_UBIFS_FS_DEBUG */