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 * Authors: Adrian Hunter
20 * Artem Bityutskiy (Битюцкий Артём)
23 /* This file implements TNC functions for committing */
25 #include <linux/random.h>
29 * make_idx_node - make an index node for fill-the-gaps method of TNC commit.
30 * @c: UBIFS file-system description object
31 * @idx: buffer in which to place new index node
32 * @znode: znode from which to make new index node
33 * @lnum: LEB number where new index node will be written
34 * @offs: offset where new index node will be written
35 * @len: length of new index node
37 static int make_idx_node(struct ubifs_info
*c
, struct ubifs_idx_node
*idx
,
38 struct ubifs_znode
*znode
, int lnum
, int offs
, int len
)
40 struct ubifs_znode
*zp
;
44 idx
->ch
.node_type
= UBIFS_IDX_NODE
;
45 idx
->child_cnt
= cpu_to_le16(znode
->child_cnt
);
46 idx
->level
= cpu_to_le16(znode
->level
);
47 for (i
= 0; i
< znode
->child_cnt
; i
++) {
48 struct ubifs_branch
*br
= ubifs_idx_branch(c
, idx
, i
);
49 struct ubifs_zbranch
*zbr
= &znode
->zbranch
[i
];
51 key_write_idx(c
, &zbr
->key
, &br
->key
);
52 br
->lnum
= cpu_to_le32(zbr
->lnum
);
53 br
->offs
= cpu_to_le32(zbr
->offs
);
54 br
->len
= cpu_to_le32(zbr
->len
);
55 if (!zbr
->lnum
|| !zbr
->len
) {
56 ubifs_err(c
, "bad ref in znode");
57 ubifs_dump_znode(c
, znode
);
59 ubifs_dump_znode(c
, zbr
->znode
);
62 ubifs_prepare_node(c
, idx
, len
, 0);
68 err
= insert_old_idx_znode(c
, znode
);
70 /* Update the parent */
73 struct ubifs_zbranch
*zbr
;
75 zbr
= &zp
->zbranch
[znode
->iip
];
84 c
->calc_idx_sz
+= ALIGN(len
, 8);
86 atomic_long_dec(&c
->dirty_zn_cnt
);
88 ubifs_assert(ubifs_zn_dirty(znode
));
89 ubifs_assert(ubifs_zn_cow(znode
));
92 * Note, unlike 'write_index()' we do not add memory barriers here
93 * because this function is called with @c->tnc_mutex locked.
95 __clear_bit(DIRTY_ZNODE
, &znode
->flags
);
96 __clear_bit(COW_ZNODE
, &znode
->flags
);
102 * fill_gap - make index nodes in gaps in dirty index LEBs.
103 * @c: UBIFS file-system description object
104 * @lnum: LEB number that gap appears in
105 * @gap_start: offset of start of gap
106 * @gap_end: offset of end of gap
107 * @dirt: adds dirty space to this
109 * This function returns the number of index nodes written into the gap.
111 static int fill_gap(struct ubifs_info
*c
, int lnum
, int gap_start
, int gap_end
,
114 int len
, gap_remains
, gap_pos
, written
, pad_len
;
116 ubifs_assert((gap_start
& 7) == 0);
117 ubifs_assert((gap_end
& 7) == 0);
118 ubifs_assert(gap_end
>= gap_start
);
120 gap_remains
= gap_end
- gap_start
;
126 len
= ubifs_idx_node_sz(c
, c
->enext
->child_cnt
);
127 if (len
< gap_remains
) {
128 struct ubifs_znode
*znode
= c
->enext
;
129 const int alen
= ALIGN(len
, 8);
132 ubifs_assert(alen
<= gap_remains
);
133 err
= make_idx_node(c
, c
->ileb_buf
+ gap_pos
, znode
,
139 c
->enext
= znode
->cnext
;
140 if (c
->enext
== c
->cnext
)
146 if (gap_end
== c
->leb_size
) {
147 c
->ileb_len
= ALIGN(gap_pos
, c
->min_io_size
);
148 /* Pad to end of min_io_size */
149 pad_len
= c
->ileb_len
- gap_pos
;
151 /* Pad to end of gap */
152 pad_len
= gap_remains
;
153 dbg_gc("LEB %d:%d to %d len %d nodes written %d wasted bytes %d",
154 lnum
, gap_start
, gap_end
, gap_end
- gap_start
, written
, pad_len
);
155 ubifs_pad(c
, c
->ileb_buf
+ gap_pos
, pad_len
);
161 * find_old_idx - find an index node obsoleted since the last commit start.
162 * @c: UBIFS file-system description object
163 * @lnum: LEB number of obsoleted index node
164 * @offs: offset of obsoleted index node
166 * Returns %1 if found and %0 otherwise.
168 static int find_old_idx(struct ubifs_info
*c
, int lnum
, int offs
)
170 struct ubifs_old_idx
*o
;
173 p
= c
->old_idx
.rb_node
;
175 o
= rb_entry(p
, struct ubifs_old_idx
, rb
);
178 else if (lnum
> o
->lnum
)
180 else if (offs
< o
->offs
)
182 else if (offs
> o
->offs
)
191 * is_idx_node_in_use - determine if an index node can be overwritten.
192 * @c: UBIFS file-system description object
193 * @key: key of index node
194 * @level: index node level
195 * @lnum: LEB number of index node
196 * @offs: offset of index node
198 * If @key / @lnum / @offs identify an index node that was not part of the old
199 * index, then this function returns %0 (obsolete). Else if the index node was
200 * part of the old index but is now dirty %1 is returned, else if it is clean %2
201 * is returned. A negative error code is returned on failure.
203 static int is_idx_node_in_use(struct ubifs_info
*c
, union ubifs_key
*key
,
204 int level
, int lnum
, int offs
)
208 ret
= is_idx_node_in_tnc(c
, key
, level
, lnum
, offs
);
210 return ret
; /* Error code */
212 if (find_old_idx(c
, lnum
, offs
))
218 * layout_leb_in_gaps - layout index nodes using in-the-gaps method.
219 * @c: UBIFS file-system description object
220 * @p: return LEB number here
222 * This function lays out new index nodes for dirty znodes using in-the-gaps
223 * method of TNC commit.
224 * This function merely puts the next znode into the next gap, making no attempt
225 * to try to maximise the number of znodes that fit.
226 * This function returns the number of index nodes written into the gaps, or a
227 * negative error code on failure.
229 static int layout_leb_in_gaps(struct ubifs_info
*c
, int *p
)
231 struct ubifs_scan_leb
*sleb
;
232 struct ubifs_scan_node
*snod
;
233 int lnum
, dirt
= 0, gap_start
, gap_end
, err
, written
, tot_written
;
236 /* Get an index LEB with lots of obsolete index nodes */
237 lnum
= ubifs_find_dirty_idx_leb(c
);
240 * There also may be dirt in the index head that could be
241 * filled, however we do not check there at present.
243 return lnum
; /* Error code */
245 dbg_gc("LEB %d", lnum
);
247 * Scan the index LEB. We use the generic scan for this even though
248 * it is more comprehensive and less efficient than is needed for this
251 sleb
= ubifs_scan(c
, lnum
, 0, c
->ileb_buf
, 0);
254 return PTR_ERR(sleb
);
256 list_for_each_entry(snod
, &sleb
->nodes
, list
) {
257 struct ubifs_idx_node
*idx
;
260 ubifs_assert(snod
->type
== UBIFS_IDX_NODE
);
262 key_read(c
, ubifs_idx_key(c
, idx
), &snod
->key
);
263 level
= le16_to_cpu(idx
->level
);
264 /* Determine if the index node is in use (not obsolete) */
265 in_use
= is_idx_node_in_use(c
, &snod
->key
, level
, lnum
,
268 ubifs_scan_destroy(sleb
);
269 return in_use
; /* Error code */
273 dirt
+= ALIGN(snod
->len
, 8);
275 * The obsolete index nodes form gaps that can be
276 * overwritten. This gap has ended because we have
277 * found an index node that is still in use
280 gap_end
= snod
->offs
;
281 /* Try to fill gap */
282 written
= fill_gap(c
, lnum
, gap_start
, gap_end
, &dirt
);
284 ubifs_scan_destroy(sleb
);
285 return written
; /* Error code */
287 tot_written
+= written
;
288 gap_start
= ALIGN(snod
->offs
+ snod
->len
, 8);
291 ubifs_scan_destroy(sleb
);
292 c
->ileb_len
= c
->leb_size
;
293 gap_end
= c
->leb_size
;
294 /* Try to fill gap */
295 written
= fill_gap(c
, lnum
, gap_start
, gap_end
, &dirt
);
297 return written
; /* Error code */
298 tot_written
+= written
;
299 if (tot_written
== 0) {
300 struct ubifs_lprops lp
;
302 dbg_gc("LEB %d wrote %d index nodes", lnum
, tot_written
);
303 err
= ubifs_read_one_lp(c
, lnum
, &lp
);
306 if (lp
.free
== c
->leb_size
) {
308 * We must have snatched this LEB from the idx_gc list
309 * so we need to correct the free and dirty space.
311 err
= ubifs_change_one_lp(c
, lnum
,
312 c
->leb_size
- c
->ileb_len
,
319 err
= ubifs_change_one_lp(c
, lnum
, c
->leb_size
- c
->ileb_len
, dirt
,
323 err
= ubifs_leb_change(c
, lnum
, c
->ileb_buf
, c
->ileb_len
);
326 dbg_gc("LEB %d wrote %d index nodes", lnum
, tot_written
);
331 * get_leb_cnt - calculate the number of empty LEBs needed to commit.
332 * @c: UBIFS file-system description object
333 * @cnt: number of znodes to commit
335 * This function returns the number of empty LEBs needed to commit @cnt znodes
336 * to the current index head. The number is not exact and may be more than
339 static int get_leb_cnt(struct ubifs_info
*c
, int cnt
)
343 /* Assume maximum index node size (i.e. overestimate space needed) */
344 cnt
-= (c
->leb_size
- c
->ihead_offs
) / c
->max_idx_node_sz
;
347 d
= c
->leb_size
/ c
->max_idx_node_sz
;
348 return DIV_ROUND_UP(cnt
, d
);
352 * layout_in_gaps - in-the-gaps method of committing TNC.
353 * @c: UBIFS file-system description object
354 * @cnt: number of dirty znodes to commit.
356 * This function lays out new index nodes for dirty znodes using in-the-gaps
357 * method of TNC commit.
359 * This function returns %0 on success and a negative error code on failure.
361 static int layout_in_gaps(struct ubifs_info
*c
, int cnt
)
363 int err
, leb_needed_cnt
, written
, *p
;
365 dbg_gc("%d znodes to write", cnt
);
367 c
->gap_lebs
= kmalloc(sizeof(int) * (c
->lst
.idx_lebs
+ 1), GFP_NOFS
);
373 ubifs_assert(p
< c
->gap_lebs
+ sizeof(int) * c
->lst
.idx_lebs
);
374 written
= layout_leb_in_gaps(c
, p
);
377 if (err
!= -ENOSPC
) {
382 if (!dbg_is_chk_index(c
)) {
384 * Do not print scary warnings if the debugging
385 * option which forces in-the-gaps is enabled.
387 ubifs_warn(c
, "out of space");
388 ubifs_dump_budg(c
, &c
->bi
);
389 ubifs_dump_lprops(c
);
391 /* Try to commit anyway */
396 leb_needed_cnt
= get_leb_cnt(c
, cnt
);
397 dbg_gc("%d znodes remaining, need %d LEBs, have %d", cnt
,
398 leb_needed_cnt
, c
->ileb_cnt
);
399 } while (leb_needed_cnt
> c
->ileb_cnt
);
406 * layout_in_empty_space - layout index nodes in empty space.
407 * @c: UBIFS file-system description object
409 * This function lays out new index nodes for dirty znodes using empty LEBs.
411 * This function returns %0 on success and a negative error code on failure.
413 static int layout_in_empty_space(struct ubifs_info
*c
)
415 struct ubifs_znode
*znode
, *cnext
, *zp
;
416 int lnum
, offs
, len
, next_len
, buf_len
, buf_offs
, used
, avail
;
423 lnum
= c
->ihead_lnum
;
424 buf_offs
= c
->ihead_offs
;
426 buf_len
= ubifs_idx_node_sz(c
, c
->fanout
);
427 buf_len
= ALIGN(buf_len
, c
->min_io_size
);
431 /* Ensure there is enough room for first write */
432 next_len
= ubifs_idx_node_sz(c
, cnext
->child_cnt
);
433 if (buf_offs
+ next_len
> c
->leb_size
)
439 len
= ubifs_idx_node_sz(c
, znode
->child_cnt
);
441 /* Determine the index node position */
443 if (c
->ileb_nxt
>= c
->ileb_cnt
) {
444 ubifs_err(c
, "out of space");
447 lnum
= c
->ilebs
[c
->ileb_nxt
++];
453 offs
= buf_offs
+ used
;
459 /* Update the parent */
462 struct ubifs_zbranch
*zbr
;
466 zbr
= &zp
->zbranch
[i
];
471 c
->zroot
.lnum
= lnum
;
472 c
->zroot
.offs
= offs
;
475 c
->calc_idx_sz
+= ALIGN(len
, 8);
478 * Once lprops is updated, we can decrease the dirty znode count
479 * but it is easier to just do it here.
481 atomic_long_dec(&c
->dirty_zn_cnt
);
484 * Calculate the next index node length to see if there is
487 cnext
= znode
->cnext
;
488 if (cnext
== c
->cnext
)
491 next_len
= ubifs_idx_node_sz(c
, cnext
->child_cnt
);
493 /* Update buffer positions */
495 used
+= ALIGN(len
, 8);
496 avail
-= ALIGN(len
, 8);
499 buf_offs
+ used
+ next_len
<= c
->leb_size
&&
503 if (avail
<= 0 && next_len
&&
504 buf_offs
+ used
+ next_len
<= c
->leb_size
)
507 blen
= ALIGN(wlen
, c
->min_io_size
);
509 /* The buffer is full or there are no more znodes to do */
512 if (buf_offs
+ next_len
> c
->leb_size
) {
513 err
= ubifs_update_one_lp(c
, lnum
,
514 c
->leb_size
- buf_offs
, blen
- used
,
523 avail
= buf_len
- used
;
526 err
= ubifs_update_one_lp(c
, lnum
, c
->leb_size
- buf_offs
,
533 c
->dbg
->new_ihead_lnum
= lnum
;
534 c
->dbg
->new_ihead_offs
= buf_offs
;
540 * layout_commit - determine positions of index nodes to commit.
541 * @c: UBIFS file-system description object
542 * @no_space: indicates that insufficient empty LEBs were allocated
543 * @cnt: number of znodes to commit
545 * Calculate and update the positions of index nodes to commit. If there were
546 * an insufficient number of empty LEBs allocated, then index nodes are placed
547 * into the gaps created by obsolete index nodes in non-empty index LEBs. For
548 * this purpose, an obsolete index node is one that was not in the index as at
549 * the end of the last commit. To write "in-the-gaps" requires that those index
550 * LEBs are updated atomically in-place.
552 static int layout_commit(struct ubifs_info
*c
, int no_space
, int cnt
)
557 err
= layout_in_gaps(c
, cnt
);
561 err
= layout_in_empty_space(c
);
566 * find_first_dirty - find first dirty znode.
567 * @znode: znode to begin searching from
569 static struct ubifs_znode
*find_first_dirty(struct ubifs_znode
*znode
)
577 if (znode
->level
== 0) {
578 if (ubifs_zn_dirty(znode
))
583 for (i
= 0; i
< znode
->child_cnt
; i
++) {
584 struct ubifs_zbranch
*zbr
= &znode
->zbranch
[i
];
586 if (zbr
->znode
&& ubifs_zn_dirty(zbr
->znode
)) {
593 if (ubifs_zn_dirty(znode
))
601 * find_next_dirty - find next dirty znode.
602 * @znode: znode to begin searching from
604 static struct ubifs_znode
*find_next_dirty(struct ubifs_znode
*znode
)
606 int n
= znode
->iip
+ 1;
608 znode
= znode
->parent
;
611 for (; n
< znode
->child_cnt
; n
++) {
612 struct ubifs_zbranch
*zbr
= &znode
->zbranch
[n
];
614 if (zbr
->znode
&& ubifs_zn_dirty(zbr
->znode
))
615 return find_first_dirty(zbr
->znode
);
621 * get_znodes_to_commit - create list of dirty znodes to commit.
622 * @c: UBIFS file-system description object
624 * This function returns the number of znodes to commit.
626 static int get_znodes_to_commit(struct ubifs_info
*c
)
628 struct ubifs_znode
*znode
, *cnext
;
631 c
->cnext
= find_first_dirty(c
->zroot
.znode
);
632 znode
= c
->enext
= c
->cnext
;
634 dbg_cmt("no znodes to commit");
639 ubifs_assert(!ubifs_zn_cow(znode
));
640 __set_bit(COW_ZNODE
, &znode
->flags
);
642 cnext
= find_next_dirty(znode
);
644 znode
->cnext
= c
->cnext
;
647 znode
->cnext
= cnext
;
651 dbg_cmt("committing %d znodes", cnt
);
652 ubifs_assert(cnt
== atomic_long_read(&c
->dirty_zn_cnt
));
657 * alloc_idx_lebs - allocate empty LEBs to be used to commit.
658 * @c: UBIFS file-system description object
659 * @cnt: number of znodes to commit
661 * This function returns %-ENOSPC if it cannot allocate a sufficient number of
662 * empty LEBs. %0 is returned on success, otherwise a negative error code
665 static int alloc_idx_lebs(struct ubifs_info
*c
, int cnt
)
667 int i
, leb_cnt
, lnum
;
671 leb_cnt
= get_leb_cnt(c
, cnt
);
672 dbg_cmt("need about %d empty LEBS for TNC commit", leb_cnt
);
675 c
->ilebs
= kmalloc(leb_cnt
* sizeof(int), GFP_NOFS
);
678 for (i
= 0; i
< leb_cnt
; i
++) {
679 lnum
= ubifs_find_free_leb_for_idx(c
);
682 c
->ilebs
[c
->ileb_cnt
++] = lnum
;
683 dbg_cmt("LEB %d", lnum
);
685 if (dbg_is_chk_index(c
) && !(prandom_u32() & 7))
691 * free_unused_idx_lebs - free unused LEBs that were allocated for the commit.
692 * @c: UBIFS file-system description object
694 * It is possible that we allocate more empty LEBs for the commit than we need.
695 * This functions frees the surplus.
697 * This function returns %0 on success and a negative error code on failure.
699 static int free_unused_idx_lebs(struct ubifs_info
*c
)
701 int i
, err
= 0, lnum
, er
;
703 for (i
= c
->ileb_nxt
; i
< c
->ileb_cnt
; i
++) {
705 dbg_cmt("LEB %d", lnum
);
706 er
= ubifs_change_one_lp(c
, lnum
, LPROPS_NC
, LPROPS_NC
, 0,
707 LPROPS_INDEX
| LPROPS_TAKEN
, 0);
715 * free_idx_lebs - free unused LEBs after commit end.
716 * @c: UBIFS file-system description object
718 * This function returns %0 on success and a negative error code on failure.
720 static int free_idx_lebs(struct ubifs_info
*c
)
724 err
= free_unused_idx_lebs(c
);
731 * ubifs_tnc_start_commit - start TNC commit.
732 * @c: UBIFS file-system description object
733 * @zroot: new index root position is returned here
735 * This function prepares the list of indexing nodes to commit and lays out
736 * their positions on flash. If there is not enough free space it uses the
737 * in-gap commit method. Returns zero in case of success and a negative error
738 * code in case of failure.
740 int ubifs_tnc_start_commit(struct ubifs_info
*c
, struct ubifs_zbranch
*zroot
)
744 mutex_lock(&c
->tnc_mutex
);
745 err
= dbg_check_tnc(c
, 1);
748 cnt
= get_znodes_to_commit(c
);
752 err
= alloc_idx_lebs(c
, cnt
);
757 err
= layout_commit(c
, no_space
, cnt
);
760 ubifs_assert(atomic_long_read(&c
->dirty_zn_cnt
) == 0);
761 err
= free_unused_idx_lebs(c
);
766 memcpy(zroot
, &c
->zroot
, sizeof(struct ubifs_zbranch
));
768 err
= ubifs_save_dirty_idx_lnums(c
);
772 spin_lock(&c
->space_lock
);
774 * Although we have not finished committing yet, update size of the
775 * committed index ('c->bi.old_idx_sz') and zero out the index growth
776 * budget. It is OK to do this now, because we've reserved all the
777 * space which is needed to commit the index, and it is save for the
778 * budgeting subsystem to assume the index is already committed,
779 * even though it is not.
781 ubifs_assert(c
->bi
.min_idx_lebs
== ubifs_calc_min_idx_lebs(c
));
782 c
->bi
.old_idx_sz
= c
->calc_idx_sz
;
783 c
->bi
.uncommitted_idx
= 0;
784 c
->bi
.min_idx_lebs
= ubifs_calc_min_idx_lebs(c
);
785 spin_unlock(&c
->space_lock
);
786 mutex_unlock(&c
->tnc_mutex
);
788 dbg_cmt("number of index LEBs %d", c
->lst
.idx_lebs
);
789 dbg_cmt("size of index %llu", c
->calc_idx_sz
);
795 mutex_unlock(&c
->tnc_mutex
);
800 * write_index - write index nodes.
801 * @c: UBIFS file-system description object
803 * This function writes the index nodes whose positions were laid out in the
804 * layout_in_empty_space function.
806 static int write_index(struct ubifs_info
*c
)
808 struct ubifs_idx_node
*idx
;
809 struct ubifs_znode
*znode
, *cnext
;
810 int i
, lnum
, offs
, len
, next_len
, buf_len
, buf_offs
, used
;
811 int avail
, wlen
, err
, lnum_pos
= 0, blen
, nxt_offs
;
818 * Always write index nodes to the index head so that index nodes and
819 * other types of nodes are never mixed in the same erase block.
821 lnum
= c
->ihead_lnum
;
822 buf_offs
= c
->ihead_offs
;
824 /* Allocate commit buffer */
825 buf_len
= ALIGN(c
->max_idx_node_sz
, c
->min_io_size
);
829 /* Ensure there is enough room for first write */
830 next_len
= ubifs_idx_node_sz(c
, cnext
->child_cnt
);
831 if (buf_offs
+ next_len
> c
->leb_size
) {
832 err
= ubifs_update_one_lp(c
, lnum
, LPROPS_NC
, 0, 0,
843 idx
= c
->cbuf
+ used
;
845 /* Make index node */
846 idx
->ch
.node_type
= UBIFS_IDX_NODE
;
847 idx
->child_cnt
= cpu_to_le16(znode
->child_cnt
);
848 idx
->level
= cpu_to_le16(znode
->level
);
849 for (i
= 0; i
< znode
->child_cnt
; i
++) {
850 struct ubifs_branch
*br
= ubifs_idx_branch(c
, idx
, i
);
851 struct ubifs_zbranch
*zbr
= &znode
->zbranch
[i
];
853 key_write_idx(c
, &zbr
->key
, &br
->key
);
854 br
->lnum
= cpu_to_le32(zbr
->lnum
);
855 br
->offs
= cpu_to_le32(zbr
->offs
);
856 br
->len
= cpu_to_le32(zbr
->len
);
857 if (!zbr
->lnum
|| !zbr
->len
) {
858 ubifs_err(c
, "bad ref in znode");
859 ubifs_dump_znode(c
, znode
);
861 ubifs_dump_znode(c
, zbr
->znode
);
864 len
= ubifs_idx_node_sz(c
, znode
->child_cnt
);
865 ubifs_prepare_node(c
, idx
, len
, 0);
867 /* Determine the index node position */
869 lnum
= c
->ilebs
[lnum_pos
++];
874 offs
= buf_offs
+ used
;
876 if (lnum
!= znode
->lnum
|| offs
!= znode
->offs
||
878 ubifs_err(c
, "inconsistent znode posn");
882 /* Grab some stuff from znode while we still can */
883 cnext
= znode
->cnext
;
885 ubifs_assert(ubifs_zn_dirty(znode
));
886 ubifs_assert(ubifs_zn_cow(znode
));
889 * It is important that other threads should see %DIRTY_ZNODE
890 * flag cleared before %COW_ZNODE. Specifically, it matters in
891 * the 'dirty_cow_znode()' function. This is the reason for the
892 * first barrier. Also, we want the bit changes to be seen to
893 * other threads ASAP, to avoid unnecesarry copying, which is
894 * the reason for the second barrier.
896 clear_bit(DIRTY_ZNODE
, &znode
->flags
);
897 smp_mb__before_atomic();
898 clear_bit(COW_ZNODE
, &znode
->flags
);
899 smp_mb__after_atomic();
902 * We have marked the znode as clean but have not updated the
903 * @c->clean_zn_cnt counter. If this znode becomes dirty again
904 * before 'free_obsolete_znodes()' is called, then
905 * @c->clean_zn_cnt will be decremented before it gets
906 * incremented (resulting in 2 decrements for the same znode).
907 * This means that @c->clean_zn_cnt may become negative for a
910 * Q: why we cannot increment @c->clean_zn_cnt?
911 * A: because we do not have the @c->tnc_mutex locked, and the
912 * following code would be racy and buggy:
914 * if (!ubifs_zn_obsolete(znode)) {
915 * atomic_long_inc(&c->clean_zn_cnt);
916 * atomic_long_inc(&ubifs_clean_zn_cnt);
919 * Thus, we just delay the @c->clean_zn_cnt update until we
920 * have the mutex locked.
923 /* Do not access znode from this point on */
925 /* Update buffer positions */
927 used
+= ALIGN(len
, 8);
928 avail
-= ALIGN(len
, 8);
931 * Calculate the next index node length to see if there is
934 if (cnext
== c
->cnext
)
937 next_len
= ubifs_idx_node_sz(c
, cnext
->child_cnt
);
939 nxt_offs
= buf_offs
+ used
+ next_len
;
940 if (next_len
&& nxt_offs
<= c
->leb_size
) {
946 wlen
= ALIGN(wlen
, 8);
947 blen
= ALIGN(wlen
, c
->min_io_size
);
948 ubifs_pad(c
, c
->cbuf
+ wlen
, blen
- wlen
);
951 /* The buffer is full or there are no more znodes to do */
952 err
= ubifs_leb_write(c
, lnum
, c
->cbuf
, buf_offs
, blen
);
957 if (nxt_offs
> c
->leb_size
) {
958 err
= ubifs_update_one_lp(c
, lnum
, LPROPS_NC
, 0,
967 avail
= buf_len
- used
;
968 memmove(c
->cbuf
, c
->cbuf
+ blen
, used
);
974 if (lnum
!= c
->dbg
->new_ihead_lnum
||
975 buf_offs
!= c
->dbg
->new_ihead_offs
) {
976 ubifs_err(c
, "inconsistent ihead");
980 c
->ihead_lnum
= lnum
;
981 c
->ihead_offs
= buf_offs
;
987 * free_obsolete_znodes - free obsolete znodes.
988 * @c: UBIFS file-system description object
990 * At the end of commit end, obsolete znodes are freed.
992 static void free_obsolete_znodes(struct ubifs_info
*c
)
994 struct ubifs_znode
*znode
, *cnext
;
999 cnext
= znode
->cnext
;
1000 if (ubifs_zn_obsolete(znode
))
1003 znode
->cnext
= NULL
;
1004 atomic_long_inc(&c
->clean_zn_cnt
);
1005 atomic_long_inc(&ubifs_clean_zn_cnt
);
1007 } while (cnext
!= c
->cnext
);
1011 * return_gap_lebs - return LEBs used by the in-gap commit method.
1012 * @c: UBIFS file-system description object
1014 * This function clears the "taken" flag for the LEBs which were used by the
1015 * "commit in-the-gaps" method.
1017 static int return_gap_lebs(struct ubifs_info
*c
)
1025 for (p
= c
->gap_lebs
; *p
!= -1; p
++) {
1026 err
= ubifs_change_one_lp(c
, *p
, LPROPS_NC
, LPROPS_NC
, 0,
1038 * ubifs_tnc_end_commit - update the TNC for commit end.
1039 * @c: UBIFS file-system description object
1041 * Write the dirty znodes.
1043 int ubifs_tnc_end_commit(struct ubifs_info
*c
)
1050 err
= return_gap_lebs(c
);
1054 err
= write_index(c
);
1058 mutex_lock(&c
->tnc_mutex
);
1060 dbg_cmt("TNC height is %d", c
->zroot
.znode
->level
+ 1);
1062 free_obsolete_znodes(c
);
1068 mutex_unlock(&c
->tnc_mutex
);