2 * Copyright (c) 2003-2006, Cluster File Systems, Inc, info@clusterfs.com
3 * Written by Alex Tomas <alex@clusterfs.com>
5 * Architecture independence:
6 * Copyright (c) 2005, Bull S.A.
7 * Written by Pierre Peiffer <pierre.peiffer@bull.net>
9 * This program is free software; you can redistribute it and/or modify
10 * it under the terms of the GNU General Public License version 2 as
11 * published by the Free Software Foundation.
13 * This program is distributed in the hope that it will be useful,
14 * but WITHOUT ANY WARRANTY; without even the implied warranty of
15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 * GNU General Public License for more details.
18 * You should have received a copy of the GNU General Public Licens
19 * along with this program; if not, write to the Free Software
20 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-
24 * Extents support for EXT4
27 * - ext4*_error() should be used in some situations
28 * - analyze all BUG()/BUG_ON(), use -EIO where appropriate
29 * - smart tree reduction
32 #include <linux/module.h>
34 #include <linux/time.h>
35 #include <linux/ext4_jbd2.h>
36 #include <linux/jbd2.h>
37 #include <linux/highuid.h>
38 #include <linux/pagemap.h>
39 #include <linux/quotaops.h>
40 #include <linux/string.h>
41 #include <linux/slab.h>
42 #include <linux/falloc.h>
43 #include <linux/ext4_fs_extents.h>
44 #include <asm/uaccess.h>
49 * combine low and high parts of physical block number into ext4_fsblk_t
51 static ext4_fsblk_t ext_pblock(struct ext4_extent *ex)
55 block = le32_to_cpu(ex->ee_start_lo);
56 block |= ((ext4_fsblk_t) le16_to_cpu(ex->ee_start_hi) << 31) << 1;
62 * combine low and high parts of a leaf physical block number into ext4_fsblk_t
64 ext4_fsblk_t idx_pblock(struct ext4_extent_idx *ix)
68 block = le32_to_cpu(ix->ei_leaf_lo);
69 block |= ((ext4_fsblk_t) le16_to_cpu(ix->ei_leaf_hi) << 31) << 1;
74 * ext4_ext_store_pblock:
75 * stores a large physical block number into an extent struct,
76 * breaking it into parts
78 void ext4_ext_store_pblock(struct ext4_extent *ex, ext4_fsblk_t pb)
80 ex->ee_start_lo = cpu_to_le32((unsigned long) (pb & 0xffffffff));
81 ex->ee_start_hi = cpu_to_le16((unsigned long) ((pb >> 31) >> 1) & 0xffff);
85 * ext4_idx_store_pblock:
86 * stores a large physical block number into an index struct,
87 * breaking it into parts
89 static void ext4_idx_store_pblock(struct ext4_extent_idx *ix, ext4_fsblk_t pb)
91 ix->ei_leaf_lo = cpu_to_le32((unsigned long) (pb & 0xffffffff));
92 ix->ei_leaf_hi = cpu_to_le16((unsigned long) ((pb >> 31) >> 1) & 0xffff);
95 static handle_t *ext4_ext_journal_restart(handle_t *handle, int needed)
99 if (handle->h_buffer_credits > needed)
101 if (!ext4_journal_extend(handle, needed))
103 err = ext4_journal_restart(handle, needed);
113 static int ext4_ext_get_access(handle_t *handle, struct inode *inode,
114 struct ext4_ext_path *path)
117 /* path points to block */
118 return ext4_journal_get_write_access(handle, path->p_bh);
120 /* path points to leaf/index in inode body */
121 /* we use in-core data, no need to protect them */
131 static int ext4_ext_dirty(handle_t *handle, struct inode *inode,
132 struct ext4_ext_path *path)
136 /* path points to block */
137 err = ext4_journal_dirty_metadata(handle, path->p_bh);
139 /* path points to leaf/index in inode body */
140 err = ext4_mark_inode_dirty(handle, inode);
145 static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode,
146 struct ext4_ext_path *path,
149 struct ext4_inode_info *ei = EXT4_I(inode);
150 ext4_fsblk_t bg_start;
151 ext4_fsblk_t last_block;
152 ext4_grpblk_t colour;
156 struct ext4_extent *ex;
157 depth = path->p_depth;
159 /* try to predict block placement */
160 ex = path[depth].p_ext;
162 return ext_pblock(ex)+(block-le32_to_cpu(ex->ee_block));
164 /* it looks like index is empty;
165 * try to find starting block from index itself */
166 if (path[depth].p_bh)
167 return path[depth].p_bh->b_blocknr;
170 /* OK. use inode's group */
171 bg_start = (ei->i_block_group * EXT4_BLOCKS_PER_GROUP(inode->i_sb)) +
172 le32_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_first_data_block);
173 last_block = ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es) - 1;
175 if (bg_start + EXT4_BLOCKS_PER_GROUP(inode->i_sb) <= last_block)
176 colour = (current->pid % 16) *
177 (EXT4_BLOCKS_PER_GROUP(inode->i_sb) / 16);
179 colour = (current->pid % 16) * ((last_block - bg_start) / 16);
180 return bg_start + colour + block;
184 ext4_ext_new_block(handle_t *handle, struct inode *inode,
185 struct ext4_ext_path *path,
186 struct ext4_extent *ex, int *err)
188 ext4_fsblk_t goal, newblock;
190 goal = ext4_ext_find_goal(inode, path, le32_to_cpu(ex->ee_block));
191 newblock = ext4_new_block(handle, inode, goal, err);
195 static int ext4_ext_space_block(struct inode *inode)
199 size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
200 / sizeof(struct ext4_extent);
201 #ifdef AGGRESSIVE_TEST
208 static int ext4_ext_space_block_idx(struct inode *inode)
212 size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
213 / sizeof(struct ext4_extent_idx);
214 #ifdef AGGRESSIVE_TEST
221 static int ext4_ext_space_root(struct inode *inode)
225 size = sizeof(EXT4_I(inode)->i_data);
226 size -= sizeof(struct ext4_extent_header);
227 size /= sizeof(struct ext4_extent);
228 #ifdef AGGRESSIVE_TEST
235 static int ext4_ext_space_root_idx(struct inode *inode)
239 size = sizeof(EXT4_I(inode)->i_data);
240 size -= sizeof(struct ext4_extent_header);
241 size /= sizeof(struct ext4_extent_idx);
242 #ifdef AGGRESSIVE_TEST
250 ext4_ext_max_entries(struct inode *inode, int depth)
254 if (depth == ext_depth(inode)) {
256 max = ext4_ext_space_root(inode);
258 max = ext4_ext_space_root_idx(inode);
261 max = ext4_ext_space_block(inode);
263 max = ext4_ext_space_block_idx(inode);
269 static int __ext4_ext_check_header(const char *function, struct inode *inode,
270 struct ext4_extent_header *eh,
273 const char *error_msg;
276 if (unlikely(eh->eh_magic != EXT4_EXT_MAGIC)) {
277 error_msg = "invalid magic";
280 if (unlikely(le16_to_cpu(eh->eh_depth) != depth)) {
281 error_msg = "unexpected eh_depth";
284 if (unlikely(eh->eh_max == 0)) {
285 error_msg = "invalid eh_max";
288 max = ext4_ext_max_entries(inode, depth);
289 if (unlikely(le16_to_cpu(eh->eh_max) > max)) {
290 error_msg = "too large eh_max";
293 if (unlikely(le16_to_cpu(eh->eh_entries) > le16_to_cpu(eh->eh_max))) {
294 error_msg = "invalid eh_entries";
300 ext4_error(inode->i_sb, function,
301 "bad header in inode #%lu: %s - magic %x, "
302 "entries %u, max %u(%u), depth %u(%u)",
303 inode->i_ino, error_msg, le16_to_cpu(eh->eh_magic),
304 le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max),
305 max, le16_to_cpu(eh->eh_depth), depth);
310 #define ext4_ext_check_header(inode, eh, depth) \
311 __ext4_ext_check_header(__FUNCTION__, inode, eh, depth)
314 static void ext4_ext_show_path(struct inode *inode, struct ext4_ext_path *path)
316 int k, l = path->p_depth;
319 for (k = 0; k <= l; k++, path++) {
321 ext_debug(" %d->%llu", le32_to_cpu(path->p_idx->ei_block),
322 idx_pblock(path->p_idx));
323 } else if (path->p_ext) {
324 ext_debug(" %d:%d:%llu ",
325 le32_to_cpu(path->p_ext->ee_block),
326 ext4_ext_get_actual_len(path->p_ext),
327 ext_pblock(path->p_ext));
334 static void ext4_ext_show_leaf(struct inode *inode, struct ext4_ext_path *path)
336 int depth = ext_depth(inode);
337 struct ext4_extent_header *eh;
338 struct ext4_extent *ex;
344 eh = path[depth].p_hdr;
345 ex = EXT_FIRST_EXTENT(eh);
347 for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ex++) {
348 ext_debug("%d:%d:%llu ", le32_to_cpu(ex->ee_block),
349 ext4_ext_get_actual_len(ex), ext_pblock(ex));
354 #define ext4_ext_show_path(inode,path)
355 #define ext4_ext_show_leaf(inode,path)
358 void ext4_ext_drop_refs(struct ext4_ext_path *path)
360 int depth = path->p_depth;
363 for (i = 0; i <= depth; i++, path++)
371 * ext4_ext_binsearch_idx:
372 * binary search for the closest index of the given block
373 * the header must be checked before calling this
376 ext4_ext_binsearch_idx(struct inode *inode,
377 struct ext4_ext_path *path, ext4_lblk_t block)
379 struct ext4_extent_header *eh = path->p_hdr;
380 struct ext4_extent_idx *r, *l, *m;
383 ext_debug("binsearch for %u(idx): ", block);
385 l = EXT_FIRST_INDEX(eh) + 1;
386 r = EXT_LAST_INDEX(eh);
389 if (block < le32_to_cpu(m->ei_block))
393 ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ei_block),
394 m, le32_to_cpu(m->ei_block),
395 r, le32_to_cpu(r->ei_block));
399 ext_debug(" -> %d->%lld ", le32_to_cpu(path->p_idx->ei_block),
400 idx_pblock(path->p_idx));
402 #ifdef CHECK_BINSEARCH
404 struct ext4_extent_idx *chix, *ix;
407 chix = ix = EXT_FIRST_INDEX(eh);
408 for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ix++) {
410 le32_to_cpu(ix->ei_block) <= le32_to_cpu(ix[-1].ei_block)) {
411 printk("k=%d, ix=0x%p, first=0x%p\n", k,
412 ix, EXT_FIRST_INDEX(eh));
414 le32_to_cpu(ix->ei_block),
415 le32_to_cpu(ix[-1].ei_block));
417 BUG_ON(k && le32_to_cpu(ix->ei_block)
418 <= le32_to_cpu(ix[-1].ei_block));
419 if (block < le32_to_cpu(ix->ei_block))
423 BUG_ON(chix != path->p_idx);
430 * ext4_ext_binsearch:
431 * binary search for closest extent of the given block
432 * the header must be checked before calling this
435 ext4_ext_binsearch(struct inode *inode,
436 struct ext4_ext_path *path, ext4_lblk_t block)
438 struct ext4_extent_header *eh = path->p_hdr;
439 struct ext4_extent *r, *l, *m;
441 if (eh->eh_entries == 0) {
443 * this leaf is empty:
444 * we get such a leaf in split/add case
449 ext_debug("binsearch for %u: ", block);
451 l = EXT_FIRST_EXTENT(eh) + 1;
452 r = EXT_LAST_EXTENT(eh);
456 if (block < le32_to_cpu(m->ee_block))
460 ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ee_block),
461 m, le32_to_cpu(m->ee_block),
462 r, le32_to_cpu(r->ee_block));
466 ext_debug(" -> %d:%llu:%d ",
467 le32_to_cpu(path->p_ext->ee_block),
468 ext_pblock(path->p_ext),
469 ext4_ext_get_actual_len(path->p_ext));
471 #ifdef CHECK_BINSEARCH
473 struct ext4_extent *chex, *ex;
476 chex = ex = EXT_FIRST_EXTENT(eh);
477 for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ex++) {
478 BUG_ON(k && le32_to_cpu(ex->ee_block)
479 <= le32_to_cpu(ex[-1].ee_block));
480 if (block < le32_to_cpu(ex->ee_block))
484 BUG_ON(chex != path->p_ext);
490 int ext4_ext_tree_init(handle_t *handle, struct inode *inode)
492 struct ext4_extent_header *eh;
494 eh = ext_inode_hdr(inode);
497 eh->eh_magic = EXT4_EXT_MAGIC;
498 eh->eh_max = cpu_to_le16(ext4_ext_space_root(inode));
499 ext4_mark_inode_dirty(handle, inode);
500 ext4_ext_invalidate_cache(inode);
504 struct ext4_ext_path *
505 ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block,
506 struct ext4_ext_path *path)
508 struct ext4_extent_header *eh;
509 struct buffer_head *bh;
510 short int depth, i, ppos = 0, alloc = 0;
512 eh = ext_inode_hdr(inode);
513 depth = ext_depth(inode);
514 if (ext4_ext_check_header(inode, eh, depth))
515 return ERR_PTR(-EIO);
518 /* account possible depth increase */
520 path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 2),
523 return ERR_PTR(-ENOMEM);
529 /* walk through the tree */
531 ext_debug("depth %d: num %d, max %d\n",
532 ppos, le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
534 ext4_ext_binsearch_idx(inode, path + ppos, block);
535 path[ppos].p_block = idx_pblock(path[ppos].p_idx);
536 path[ppos].p_depth = i;
537 path[ppos].p_ext = NULL;
539 bh = sb_bread(inode->i_sb, path[ppos].p_block);
543 eh = ext_block_hdr(bh);
545 BUG_ON(ppos > depth);
546 path[ppos].p_bh = bh;
547 path[ppos].p_hdr = eh;
550 if (ext4_ext_check_header(inode, eh, i))
554 path[ppos].p_depth = i;
555 path[ppos].p_hdr = eh;
556 path[ppos].p_ext = NULL;
557 path[ppos].p_idx = NULL;
560 ext4_ext_binsearch(inode, path + ppos, block);
562 ext4_ext_show_path(inode, path);
567 ext4_ext_drop_refs(path);
570 return ERR_PTR(-EIO);
574 * ext4_ext_insert_index:
575 * insert new index [@logical;@ptr] into the block at @curp;
576 * check where to insert: before @curp or after @curp
578 static int ext4_ext_insert_index(handle_t *handle, struct inode *inode,
579 struct ext4_ext_path *curp,
580 int logical, ext4_fsblk_t ptr)
582 struct ext4_extent_idx *ix;
585 err = ext4_ext_get_access(handle, inode, curp);
589 BUG_ON(logical == le32_to_cpu(curp->p_idx->ei_block));
590 len = EXT_MAX_INDEX(curp->p_hdr) - curp->p_idx;
591 if (logical > le32_to_cpu(curp->p_idx->ei_block)) {
593 if (curp->p_idx != EXT_LAST_INDEX(curp->p_hdr)) {
594 len = (len - 1) * sizeof(struct ext4_extent_idx);
595 len = len < 0 ? 0 : len;
596 ext_debug("insert new index %d after: %llu. "
597 "move %d from 0x%p to 0x%p\n",
599 (curp->p_idx + 1), (curp->p_idx + 2));
600 memmove(curp->p_idx + 2, curp->p_idx + 1, len);
602 ix = curp->p_idx + 1;
605 len = len * sizeof(struct ext4_extent_idx);
606 len = len < 0 ? 0 : len;
607 ext_debug("insert new index %d before: %llu. "
608 "move %d from 0x%p to 0x%p\n",
610 curp->p_idx, (curp->p_idx + 1));
611 memmove(curp->p_idx + 1, curp->p_idx, len);
615 ix->ei_block = cpu_to_le32(logical);
616 ext4_idx_store_pblock(ix, ptr);
617 curp->p_hdr->eh_entries = cpu_to_le16(le16_to_cpu(curp->p_hdr->eh_entries)+1);
619 BUG_ON(le16_to_cpu(curp->p_hdr->eh_entries)
620 > le16_to_cpu(curp->p_hdr->eh_max));
621 BUG_ON(ix > EXT_LAST_INDEX(curp->p_hdr));
623 err = ext4_ext_dirty(handle, inode, curp);
624 ext4_std_error(inode->i_sb, err);
631 * inserts new subtree into the path, using free index entry
633 * - allocates all needed blocks (new leaf and all intermediate index blocks)
634 * - makes decision where to split
635 * - moves remaining extents and index entries (right to the split point)
636 * into the newly allocated blocks
637 * - initializes subtree
639 static int ext4_ext_split(handle_t *handle, struct inode *inode,
640 struct ext4_ext_path *path,
641 struct ext4_extent *newext, int at)
643 struct buffer_head *bh = NULL;
644 int depth = ext_depth(inode);
645 struct ext4_extent_header *neh;
646 struct ext4_extent_idx *fidx;
647 struct ext4_extent *ex;
649 ext4_fsblk_t newblock, oldblock;
651 ext4_fsblk_t *ablocks = NULL; /* array of allocated blocks */
654 /* make decision: where to split? */
655 /* FIXME: now decision is simplest: at current extent */
657 /* if current leaf will be split, then we should use
658 * border from split point */
659 BUG_ON(path[depth].p_ext > EXT_MAX_EXTENT(path[depth].p_hdr));
660 if (path[depth].p_ext != EXT_MAX_EXTENT(path[depth].p_hdr)) {
661 border = path[depth].p_ext[1].ee_block;
662 ext_debug("leaf will be split."
663 " next leaf starts at %d\n",
664 le32_to_cpu(border));
666 border = newext->ee_block;
667 ext_debug("leaf will be added."
668 " next leaf starts at %d\n",
669 le32_to_cpu(border));
673 * If error occurs, then we break processing
674 * and mark filesystem read-only. index won't
675 * be inserted and tree will be in consistent
676 * state. Next mount will repair buffers too.
680 * Get array to track all allocated blocks.
681 * We need this to handle errors and free blocks
684 ablocks = kzalloc(sizeof(ext4_fsblk_t) * depth, GFP_NOFS);
688 /* allocate all needed blocks */
689 ext_debug("allocate %d blocks for indexes/leaf\n", depth - at);
690 for (a = 0; a < depth - at; a++) {
691 newblock = ext4_ext_new_block(handle, inode, path, newext, &err);
694 ablocks[a] = newblock;
697 /* initialize new leaf */
698 newblock = ablocks[--a];
699 BUG_ON(newblock == 0);
700 bh = sb_getblk(inode->i_sb, newblock);
707 err = ext4_journal_get_create_access(handle, bh);
711 neh = ext_block_hdr(bh);
713 neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode));
714 neh->eh_magic = EXT4_EXT_MAGIC;
716 ex = EXT_FIRST_EXTENT(neh);
718 /* move remainder of path[depth] to the new leaf */
719 BUG_ON(path[depth].p_hdr->eh_entries != path[depth].p_hdr->eh_max);
720 /* start copy from next extent */
721 /* TODO: we could do it by single memmove */
724 while (path[depth].p_ext <=
725 EXT_MAX_EXTENT(path[depth].p_hdr)) {
726 ext_debug("move %d:%llu:%d in new leaf %llu\n",
727 le32_to_cpu(path[depth].p_ext->ee_block),
728 ext_pblock(path[depth].p_ext),
729 ext4_ext_get_actual_len(path[depth].p_ext),
731 /*memmove(ex++, path[depth].p_ext++,
732 sizeof(struct ext4_extent));
738 memmove(ex, path[depth].p_ext-m, sizeof(struct ext4_extent)*m);
739 neh->eh_entries = cpu_to_le16(le16_to_cpu(neh->eh_entries)+m);
742 set_buffer_uptodate(bh);
745 err = ext4_journal_dirty_metadata(handle, bh);
751 /* correct old leaf */
753 err = ext4_ext_get_access(handle, inode, path + depth);
756 path[depth].p_hdr->eh_entries =
757 cpu_to_le16(le16_to_cpu(path[depth].p_hdr->eh_entries)-m);
758 err = ext4_ext_dirty(handle, inode, path + depth);
764 /* create intermediate indexes */
768 ext_debug("create %d intermediate indices\n", k);
769 /* insert new index into current index block */
770 /* current depth stored in i var */
774 newblock = ablocks[--a];
775 bh = sb_getblk(inode->i_sb, newblock);
782 err = ext4_journal_get_create_access(handle, bh);
786 neh = ext_block_hdr(bh);
787 neh->eh_entries = cpu_to_le16(1);
788 neh->eh_magic = EXT4_EXT_MAGIC;
789 neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode));
790 neh->eh_depth = cpu_to_le16(depth - i);
791 fidx = EXT_FIRST_INDEX(neh);
792 fidx->ei_block = border;
793 ext4_idx_store_pblock(fidx, oldblock);
795 ext_debug("int.index at %d (block %llu): %u -> %llu\n",
796 i, newblock, le32_to_cpu(border), oldblock);
801 ext_debug("cur 0x%p, last 0x%p\n", path[i].p_idx,
802 EXT_MAX_INDEX(path[i].p_hdr));
803 BUG_ON(EXT_MAX_INDEX(path[i].p_hdr) !=
804 EXT_LAST_INDEX(path[i].p_hdr));
805 while (path[i].p_idx <= EXT_MAX_INDEX(path[i].p_hdr)) {
806 ext_debug("%d: move %d:%llu in new index %llu\n", i,
807 le32_to_cpu(path[i].p_idx->ei_block),
808 idx_pblock(path[i].p_idx),
810 /*memmove(++fidx, path[i].p_idx++,
811 sizeof(struct ext4_extent_idx));
813 BUG_ON(neh->eh_entries > neh->eh_max);*/
818 memmove(++fidx, path[i].p_idx - m,
819 sizeof(struct ext4_extent_idx) * m);
821 cpu_to_le16(le16_to_cpu(neh->eh_entries) + m);
823 set_buffer_uptodate(bh);
826 err = ext4_journal_dirty_metadata(handle, bh);
832 /* correct old index */
834 err = ext4_ext_get_access(handle, inode, path + i);
837 path[i].p_hdr->eh_entries = cpu_to_le16(le16_to_cpu(path[i].p_hdr->eh_entries)-m);
838 err = ext4_ext_dirty(handle, inode, path + i);
846 /* insert new index */
847 err = ext4_ext_insert_index(handle, inode, path + at,
848 le32_to_cpu(border), newblock);
852 if (buffer_locked(bh))
858 /* free all allocated blocks in error case */
859 for (i = 0; i < depth; i++) {
862 ext4_free_blocks(handle, inode, ablocks[i], 1, 1);
871 * ext4_ext_grow_indepth:
872 * implements tree growing procedure:
873 * - allocates new block
874 * - moves top-level data (index block or leaf) into the new block
875 * - initializes new top-level, creating index that points to the
878 static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode,
879 struct ext4_ext_path *path,
880 struct ext4_extent *newext)
882 struct ext4_ext_path *curp = path;
883 struct ext4_extent_header *neh;
884 struct ext4_extent_idx *fidx;
885 struct buffer_head *bh;
886 ext4_fsblk_t newblock;
889 newblock = ext4_ext_new_block(handle, inode, path, newext, &err);
893 bh = sb_getblk(inode->i_sb, newblock);
896 ext4_std_error(inode->i_sb, err);
901 err = ext4_journal_get_create_access(handle, bh);
907 /* move top-level index/leaf into new block */
908 memmove(bh->b_data, curp->p_hdr, sizeof(EXT4_I(inode)->i_data));
910 /* set size of new block */
911 neh = ext_block_hdr(bh);
912 /* old root could have indexes or leaves
913 * so calculate e_max right way */
914 if (ext_depth(inode))
915 neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode));
917 neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode));
918 neh->eh_magic = EXT4_EXT_MAGIC;
919 set_buffer_uptodate(bh);
922 err = ext4_journal_dirty_metadata(handle, bh);
926 /* create index in new top-level index: num,max,pointer */
927 err = ext4_ext_get_access(handle, inode, curp);
931 curp->p_hdr->eh_magic = EXT4_EXT_MAGIC;
932 curp->p_hdr->eh_max = cpu_to_le16(ext4_ext_space_root_idx(inode));
933 curp->p_hdr->eh_entries = cpu_to_le16(1);
934 curp->p_idx = EXT_FIRST_INDEX(curp->p_hdr);
936 if (path[0].p_hdr->eh_depth)
937 curp->p_idx->ei_block =
938 EXT_FIRST_INDEX(path[0].p_hdr)->ei_block;
940 curp->p_idx->ei_block =
941 EXT_FIRST_EXTENT(path[0].p_hdr)->ee_block;
942 ext4_idx_store_pblock(curp->p_idx, newblock);
944 neh = ext_inode_hdr(inode);
945 fidx = EXT_FIRST_INDEX(neh);
946 ext_debug("new root: num %d(%d), lblock %d, ptr %llu\n",
947 le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max),
948 le32_to_cpu(fidx->ei_block), idx_pblock(fidx));
950 neh->eh_depth = cpu_to_le16(path->p_depth + 1);
951 err = ext4_ext_dirty(handle, inode, curp);
959 * ext4_ext_create_new_leaf:
960 * finds empty index and adds new leaf.
961 * if no free index is found, then it requests in-depth growing.
963 static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode,
964 struct ext4_ext_path *path,
965 struct ext4_extent *newext)
967 struct ext4_ext_path *curp;
968 int depth, i, err = 0;
971 i = depth = ext_depth(inode);
973 /* walk up to the tree and look for free index entry */
975 while (i > 0 && !EXT_HAS_FREE_INDEX(curp)) {
980 /* we use already allocated block for index block,
981 * so subsequent data blocks should be contiguous */
982 if (EXT_HAS_FREE_INDEX(curp)) {
983 /* if we found index with free entry, then use that
984 * entry: create all needed subtree and add new leaf */
985 err = ext4_ext_split(handle, inode, path, newext, i);
988 ext4_ext_drop_refs(path);
989 path = ext4_ext_find_extent(inode,
990 (ext4_lblk_t)le32_to_cpu(newext->ee_block),
995 /* tree is full, time to grow in depth */
996 err = ext4_ext_grow_indepth(handle, inode, path, newext);
1001 ext4_ext_drop_refs(path);
1002 path = ext4_ext_find_extent(inode,
1003 (ext4_lblk_t)le32_to_cpu(newext->ee_block),
1006 err = PTR_ERR(path);
1011 * only first (depth 0 -> 1) produces free space;
1012 * in all other cases we have to split the grown tree
1014 depth = ext_depth(inode);
1015 if (path[depth].p_hdr->eh_entries == path[depth].p_hdr->eh_max) {
1016 /* now we need to split */
1026 * search the closest allocated block to the left for *logical
1027 * and returns it at @logical + it's physical address at @phys
1028 * if *logical is the smallest allocated block, the function
1029 * returns 0 at @phys
1030 * return value contains 0 (success) or error code
1033 ext4_ext_search_left(struct inode *inode, struct ext4_ext_path *path,
1034 ext4_lblk_t *logical, ext4_fsblk_t *phys)
1036 struct ext4_extent_idx *ix;
1037 struct ext4_extent *ex;
1040 BUG_ON(path == NULL);
1041 depth = path->p_depth;
1044 if (depth == 0 && path->p_ext == NULL)
1047 /* usually extent in the path covers blocks smaller
1048 * then *logical, but it can be that extent is the
1049 * first one in the file */
1051 ex = path[depth].p_ext;
1052 ee_len = ext4_ext_get_actual_len(ex);
1053 if (*logical < le32_to_cpu(ex->ee_block)) {
1054 BUG_ON(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex);
1055 while (--depth >= 0) {
1056 ix = path[depth].p_idx;
1057 BUG_ON(ix != EXT_FIRST_INDEX(path[depth].p_hdr));
1062 BUG_ON(*logical < (le32_to_cpu(ex->ee_block) + ee_len));
1064 *logical = le32_to_cpu(ex->ee_block) + ee_len - 1;
1065 *phys = ext_pblock(ex) + ee_len - 1;
1070 * search the closest allocated block to the right for *logical
1071 * and returns it at @logical + it's physical address at @phys
1072 * if *logical is the smallest allocated block, the function
1073 * returns 0 at @phys
1074 * return value contains 0 (success) or error code
1077 ext4_ext_search_right(struct inode *inode, struct ext4_ext_path *path,
1078 ext4_lblk_t *logical, ext4_fsblk_t *phys)
1080 struct buffer_head *bh = NULL;
1081 struct ext4_extent_header *eh;
1082 struct ext4_extent_idx *ix;
1083 struct ext4_extent *ex;
1087 BUG_ON(path == NULL);
1088 depth = path->p_depth;
1091 if (depth == 0 && path->p_ext == NULL)
1094 /* usually extent in the path covers blocks smaller
1095 * then *logical, but it can be that extent is the
1096 * first one in the file */
1098 ex = path[depth].p_ext;
1099 ee_len = ext4_ext_get_actual_len(ex);
1100 if (*logical < le32_to_cpu(ex->ee_block)) {
1101 BUG_ON(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex);
1102 while (--depth >= 0) {
1103 ix = path[depth].p_idx;
1104 BUG_ON(ix != EXT_FIRST_INDEX(path[depth].p_hdr));
1106 *logical = le32_to_cpu(ex->ee_block);
1107 *phys = ext_pblock(ex);
1111 BUG_ON(*logical < (le32_to_cpu(ex->ee_block) + ee_len));
1113 if (ex != EXT_LAST_EXTENT(path[depth].p_hdr)) {
1114 /* next allocated block in this leaf */
1116 *logical = le32_to_cpu(ex->ee_block);
1117 *phys = ext_pblock(ex);
1121 /* go up and search for index to the right */
1122 while (--depth >= 0) {
1123 ix = path[depth].p_idx;
1124 if (ix != EXT_LAST_INDEX(path[depth].p_hdr))
1129 /* we've gone up to the root and
1130 * found no index to the right */
1134 /* we've found index to the right, let's
1135 * follow it and find the closest allocated
1136 * block to the right */
1138 block = idx_pblock(ix);
1139 while (++depth < path->p_depth) {
1140 bh = sb_bread(inode->i_sb, block);
1143 eh = ext_block_hdr(bh);
1144 if (ext4_ext_check_header(inode, eh, depth)) {
1148 ix = EXT_FIRST_INDEX(eh);
1149 block = idx_pblock(ix);
1153 bh = sb_bread(inode->i_sb, block);
1156 eh = ext_block_hdr(bh);
1157 if (ext4_ext_check_header(inode, eh, path->p_depth - depth)) {
1161 ex = EXT_FIRST_EXTENT(eh);
1162 *logical = le32_to_cpu(ex->ee_block);
1163 *phys = ext_pblock(ex);
1170 * ext4_ext_next_allocated_block:
1171 * returns allocated block in subsequent extent or EXT_MAX_BLOCK.
1172 * NOTE: it considers block number from index entry as
1173 * allocated block. Thus, index entries have to be consistent
1177 ext4_ext_next_allocated_block(struct ext4_ext_path *path)
1181 BUG_ON(path == NULL);
1182 depth = path->p_depth;
1184 if (depth == 0 && path->p_ext == NULL)
1185 return EXT_MAX_BLOCK;
1187 while (depth >= 0) {
1188 if (depth == path->p_depth) {
1190 if (path[depth].p_ext !=
1191 EXT_LAST_EXTENT(path[depth].p_hdr))
1192 return le32_to_cpu(path[depth].p_ext[1].ee_block);
1195 if (path[depth].p_idx !=
1196 EXT_LAST_INDEX(path[depth].p_hdr))
1197 return le32_to_cpu(path[depth].p_idx[1].ei_block);
1202 return EXT_MAX_BLOCK;
1206 * ext4_ext_next_leaf_block:
1207 * returns first allocated block from next leaf or EXT_MAX_BLOCK
1209 static ext4_lblk_t ext4_ext_next_leaf_block(struct inode *inode,
1210 struct ext4_ext_path *path)
1214 BUG_ON(path == NULL);
1215 depth = path->p_depth;
1217 /* zero-tree has no leaf blocks at all */
1219 return EXT_MAX_BLOCK;
1221 /* go to index block */
1224 while (depth >= 0) {
1225 if (path[depth].p_idx !=
1226 EXT_LAST_INDEX(path[depth].p_hdr))
1227 return (ext4_lblk_t)
1228 le32_to_cpu(path[depth].p_idx[1].ei_block);
1232 return EXT_MAX_BLOCK;
1236 * ext4_ext_correct_indexes:
1237 * if leaf gets modified and modified extent is first in the leaf,
1238 * then we have to correct all indexes above.
1239 * TODO: do we need to correct tree in all cases?
1241 static int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode,
1242 struct ext4_ext_path *path)
1244 struct ext4_extent_header *eh;
1245 int depth = ext_depth(inode);
1246 struct ext4_extent *ex;
1250 eh = path[depth].p_hdr;
1251 ex = path[depth].p_ext;
1256 /* there is no tree at all */
1260 if (ex != EXT_FIRST_EXTENT(eh)) {
1261 /* we correct tree if first leaf got modified only */
1266 * TODO: we need correction if border is smaller than current one
1269 border = path[depth].p_ext->ee_block;
1270 err = ext4_ext_get_access(handle, inode, path + k);
1273 path[k].p_idx->ei_block = border;
1274 err = ext4_ext_dirty(handle, inode, path + k);
1279 /* change all left-side indexes */
1280 if (path[k+1].p_idx != EXT_FIRST_INDEX(path[k+1].p_hdr))
1282 err = ext4_ext_get_access(handle, inode, path + k);
1285 path[k].p_idx->ei_block = border;
1286 err = ext4_ext_dirty(handle, inode, path + k);
1295 ext4_can_extents_be_merged(struct inode *inode, struct ext4_extent *ex1,
1296 struct ext4_extent *ex2)
1298 unsigned short ext1_ee_len, ext2_ee_len, max_len;
1301 * Make sure that either both extents are uninitialized, or
1304 if (ext4_ext_is_uninitialized(ex1) ^ ext4_ext_is_uninitialized(ex2))
1307 if (ext4_ext_is_uninitialized(ex1))
1308 max_len = EXT_UNINIT_MAX_LEN;
1310 max_len = EXT_INIT_MAX_LEN;
1312 ext1_ee_len = ext4_ext_get_actual_len(ex1);
1313 ext2_ee_len = ext4_ext_get_actual_len(ex2);
1315 if (le32_to_cpu(ex1->ee_block) + ext1_ee_len !=
1316 le32_to_cpu(ex2->ee_block))
1320 * To allow future support for preallocated extents to be added
1321 * as an RO_COMPAT feature, refuse to merge to extents if
1322 * this can result in the top bit of ee_len being set.
1324 if (ext1_ee_len + ext2_ee_len > max_len)
1326 #ifdef AGGRESSIVE_TEST
1327 if (ext1_ee_len >= 4)
1331 if (ext_pblock(ex1) + ext1_ee_len == ext_pblock(ex2))
1337 * This function tries to merge the "ex" extent to the next extent in the tree.
1338 * It always tries to merge towards right. If you want to merge towards
1339 * left, pass "ex - 1" as argument instead of "ex".
1340 * Returns 0 if the extents (ex and ex+1) were _not_ merged and returns
1341 * 1 if they got merged.
1343 int ext4_ext_try_to_merge(struct inode *inode,
1344 struct ext4_ext_path *path,
1345 struct ext4_extent *ex)
1347 struct ext4_extent_header *eh;
1348 unsigned int depth, len;
1350 int uninitialized = 0;
1352 depth = ext_depth(inode);
1353 BUG_ON(path[depth].p_hdr == NULL);
1354 eh = path[depth].p_hdr;
1356 while (ex < EXT_LAST_EXTENT(eh)) {
1357 if (!ext4_can_extents_be_merged(inode, ex, ex + 1))
1359 /* merge with next extent! */
1360 if (ext4_ext_is_uninitialized(ex))
1362 ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
1363 + ext4_ext_get_actual_len(ex + 1));
1365 ext4_ext_mark_uninitialized(ex);
1367 if (ex + 1 < EXT_LAST_EXTENT(eh)) {
1368 len = (EXT_LAST_EXTENT(eh) - ex - 1)
1369 * sizeof(struct ext4_extent);
1370 memmove(ex + 1, ex + 2, len);
1372 eh->eh_entries = cpu_to_le16(le16_to_cpu(eh->eh_entries) - 1);
1374 WARN_ON(eh->eh_entries == 0);
1375 if (!eh->eh_entries)
1376 ext4_error(inode->i_sb, "ext4_ext_try_to_merge",
1377 "inode#%lu, eh->eh_entries = 0!", inode->i_ino);
1384 * check if a portion of the "newext" extent overlaps with an
1387 * If there is an overlap discovered, it updates the length of the newext
1388 * such that there will be no overlap, and then returns 1.
1389 * If there is no overlap found, it returns 0.
1391 unsigned int ext4_ext_check_overlap(struct inode *inode,
1392 struct ext4_extent *newext,
1393 struct ext4_ext_path *path)
1396 unsigned int depth, len1;
1397 unsigned int ret = 0;
1399 b1 = le32_to_cpu(newext->ee_block);
1400 len1 = ext4_ext_get_actual_len(newext);
1401 depth = ext_depth(inode);
1402 if (!path[depth].p_ext)
1404 b2 = le32_to_cpu(path[depth].p_ext->ee_block);
1407 * get the next allocated block if the extent in the path
1408 * is before the requested block(s)
1411 b2 = ext4_ext_next_allocated_block(path);
1412 if (b2 == EXT_MAX_BLOCK)
1416 /* check for wrap through zero on extent logical start block*/
1417 if (b1 + len1 < b1) {
1418 len1 = EXT_MAX_BLOCK - b1;
1419 newext->ee_len = cpu_to_le16(len1);
1423 /* check for overlap */
1424 if (b1 + len1 > b2) {
1425 newext->ee_len = cpu_to_le16(b2 - b1);
1433 * ext4_ext_insert_extent:
1434 * tries to merge requsted extent into the existing extent or
1435 * inserts requested extent as new one into the tree,
1436 * creating new leaf in the no-space case.
1438 int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
1439 struct ext4_ext_path *path,
1440 struct ext4_extent *newext)
1442 struct ext4_extent_header * eh;
1443 struct ext4_extent *ex, *fex;
1444 struct ext4_extent *nearex; /* nearest extent */
1445 struct ext4_ext_path *npath = NULL;
1446 int depth, len, err;
1448 unsigned uninitialized = 0;
1450 BUG_ON(ext4_ext_get_actual_len(newext) == 0);
1451 depth = ext_depth(inode);
1452 ex = path[depth].p_ext;
1453 BUG_ON(path[depth].p_hdr == NULL);
1455 /* try to insert block into found extent and return */
1456 if (ex && ext4_can_extents_be_merged(inode, ex, newext)) {
1457 ext_debug("append %d block to %d:%d (from %llu)\n",
1458 ext4_ext_get_actual_len(newext),
1459 le32_to_cpu(ex->ee_block),
1460 ext4_ext_get_actual_len(ex), ext_pblock(ex));
1461 err = ext4_ext_get_access(handle, inode, path + depth);
1466 * ext4_can_extents_be_merged should have checked that either
1467 * both extents are uninitialized, or both aren't. Thus we
1468 * need to check only one of them here.
1470 if (ext4_ext_is_uninitialized(ex))
1472 ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
1473 + ext4_ext_get_actual_len(newext));
1475 ext4_ext_mark_uninitialized(ex);
1476 eh = path[depth].p_hdr;
1482 depth = ext_depth(inode);
1483 eh = path[depth].p_hdr;
1484 if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max))
1487 /* probably next leaf has space for us? */
1488 fex = EXT_LAST_EXTENT(eh);
1489 next = ext4_ext_next_leaf_block(inode, path);
1490 if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block)
1491 && next != EXT_MAX_BLOCK) {
1492 ext_debug("next leaf block - %d\n", next);
1493 BUG_ON(npath != NULL);
1494 npath = ext4_ext_find_extent(inode, next, NULL);
1496 return PTR_ERR(npath);
1497 BUG_ON(npath->p_depth != path->p_depth);
1498 eh = npath[depth].p_hdr;
1499 if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) {
1500 ext_debug("next leaf isnt full(%d)\n",
1501 le16_to_cpu(eh->eh_entries));
1505 ext_debug("next leaf has no free space(%d,%d)\n",
1506 le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
1510 * There is no free space in the found leaf.
1511 * We're gonna add a new leaf in the tree.
1513 err = ext4_ext_create_new_leaf(handle, inode, path, newext);
1516 depth = ext_depth(inode);
1517 eh = path[depth].p_hdr;
1520 nearex = path[depth].p_ext;
1522 err = ext4_ext_get_access(handle, inode, path + depth);
1527 /* there is no extent in this leaf, create first one */
1528 ext_debug("first extent in the leaf: %d:%llu:%d\n",
1529 le32_to_cpu(newext->ee_block),
1531 ext4_ext_get_actual_len(newext));
1532 path[depth].p_ext = EXT_FIRST_EXTENT(eh);
1533 } else if (le32_to_cpu(newext->ee_block)
1534 > le32_to_cpu(nearex->ee_block)) {
1535 /* BUG_ON(newext->ee_block == nearex->ee_block); */
1536 if (nearex != EXT_LAST_EXTENT(eh)) {
1537 len = EXT_MAX_EXTENT(eh) - nearex;
1538 len = (len - 1) * sizeof(struct ext4_extent);
1539 len = len < 0 ? 0 : len;
1540 ext_debug("insert %d:%llu:%d after: nearest 0x%p, "
1541 "move %d from 0x%p to 0x%p\n",
1542 le32_to_cpu(newext->ee_block),
1544 ext4_ext_get_actual_len(newext),
1545 nearex, len, nearex + 1, nearex + 2);
1546 memmove(nearex + 2, nearex + 1, len);
1548 path[depth].p_ext = nearex + 1;
1550 BUG_ON(newext->ee_block == nearex->ee_block);
1551 len = (EXT_MAX_EXTENT(eh) - nearex) * sizeof(struct ext4_extent);
1552 len = len < 0 ? 0 : len;
1553 ext_debug("insert %d:%llu:%d before: nearest 0x%p, "
1554 "move %d from 0x%p to 0x%p\n",
1555 le32_to_cpu(newext->ee_block),
1557 ext4_ext_get_actual_len(newext),
1558 nearex, len, nearex + 1, nearex + 2);
1559 memmove(nearex + 1, nearex, len);
1560 path[depth].p_ext = nearex;
1563 eh->eh_entries = cpu_to_le16(le16_to_cpu(eh->eh_entries)+1);
1564 nearex = path[depth].p_ext;
1565 nearex->ee_block = newext->ee_block;
1566 ext4_ext_store_pblock(nearex, ext_pblock(newext));
1567 nearex->ee_len = newext->ee_len;
1570 /* try to merge extents to the right */
1571 ext4_ext_try_to_merge(inode, path, nearex);
1573 /* try to merge extents to the left */
1575 /* time to correct all indexes above */
1576 err = ext4_ext_correct_indexes(handle, inode, path);
1580 err = ext4_ext_dirty(handle, inode, path + depth);
1584 ext4_ext_drop_refs(npath);
1587 ext4_ext_tree_changed(inode);
1588 ext4_ext_invalidate_cache(inode);
1593 ext4_ext_put_in_cache(struct inode *inode, ext4_lblk_t block,
1594 __u32 len, ext4_fsblk_t start, int type)
1596 struct ext4_ext_cache *cex;
1598 cex = &EXT4_I(inode)->i_cached_extent;
1599 cex->ec_type = type;
1600 cex->ec_block = block;
1602 cex->ec_start = start;
1606 * ext4_ext_put_gap_in_cache:
1607 * calculate boundaries of the gap that the requested block fits into
1608 * and cache this gap
1611 ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
1614 int depth = ext_depth(inode);
1617 struct ext4_extent *ex;
1619 ex = path[depth].p_ext;
1621 /* there is no extent yet, so gap is [0;-] */
1623 len = EXT_MAX_BLOCK;
1624 ext_debug("cache gap(whole file):");
1625 } else if (block < le32_to_cpu(ex->ee_block)) {
1627 len = le32_to_cpu(ex->ee_block) - block;
1628 ext_debug("cache gap(before): %u [%u:%u]",
1630 le32_to_cpu(ex->ee_block),
1631 ext4_ext_get_actual_len(ex));
1632 } else if (block >= le32_to_cpu(ex->ee_block)
1633 + ext4_ext_get_actual_len(ex)) {
1635 lblock = le32_to_cpu(ex->ee_block)
1636 + ext4_ext_get_actual_len(ex);
1638 next = ext4_ext_next_allocated_block(path);
1639 ext_debug("cache gap(after): [%u:%u] %u",
1640 le32_to_cpu(ex->ee_block),
1641 ext4_ext_get_actual_len(ex),
1643 BUG_ON(next == lblock);
1644 len = next - lblock;
1650 ext_debug(" -> %u:%lu\n", lblock, len);
1651 ext4_ext_put_in_cache(inode, lblock, len, 0, EXT4_EXT_CACHE_GAP);
1655 ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block,
1656 struct ext4_extent *ex)
1658 struct ext4_ext_cache *cex;
1660 cex = &EXT4_I(inode)->i_cached_extent;
1662 /* has cache valid data? */
1663 if (cex->ec_type == EXT4_EXT_CACHE_NO)
1664 return EXT4_EXT_CACHE_NO;
1666 BUG_ON(cex->ec_type != EXT4_EXT_CACHE_GAP &&
1667 cex->ec_type != EXT4_EXT_CACHE_EXTENT);
1668 if (block >= cex->ec_block && block < cex->ec_block + cex->ec_len) {
1669 ex->ee_block = cpu_to_le32(cex->ec_block);
1670 ext4_ext_store_pblock(ex, cex->ec_start);
1671 ex->ee_len = cpu_to_le16(cex->ec_len);
1672 ext_debug("%u cached by %u:%u:%llu\n",
1674 cex->ec_block, cex->ec_len, cex->ec_start);
1675 return cex->ec_type;
1679 return EXT4_EXT_CACHE_NO;
1684 * removes index from the index block.
1685 * It's used in truncate case only, thus all requests are for
1686 * last index in the block only.
1688 static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
1689 struct ext4_ext_path *path)
1691 struct buffer_head *bh;
1695 /* free index block */
1697 leaf = idx_pblock(path->p_idx);
1698 BUG_ON(path->p_hdr->eh_entries == 0);
1699 err = ext4_ext_get_access(handle, inode, path);
1702 path->p_hdr->eh_entries = cpu_to_le16(le16_to_cpu(path->p_hdr->eh_entries)-1);
1703 err = ext4_ext_dirty(handle, inode, path);
1706 ext_debug("index is empty, remove it, free block %llu\n", leaf);
1707 bh = sb_find_get_block(inode->i_sb, leaf);
1708 ext4_forget(handle, 1, inode, bh, leaf);
1709 ext4_free_blocks(handle, inode, leaf, 1, 1);
1714 * ext4_ext_calc_credits_for_insert:
1715 * This routine returns max. credits that the extent tree can consume.
1716 * It should be OK for low-performance paths like ->writepage()
1717 * To allow many writing processes to fit into a single transaction,
1718 * the caller should calculate credits under i_data_sem and
1719 * pass the actual path.
1721 int ext4_ext_calc_credits_for_insert(struct inode *inode,
1722 struct ext4_ext_path *path)
1727 /* probably there is space in leaf? */
1728 depth = ext_depth(inode);
1729 if (le16_to_cpu(path[depth].p_hdr->eh_entries)
1730 < le16_to_cpu(path[depth].p_hdr->eh_max))
1735 * given 32-bit logical block (4294967296 blocks), max. tree
1736 * can be 4 levels in depth -- 4 * 340^4 == 53453440000.
1737 * Let's also add one more level for imbalance.
1741 /* allocation of new data block(s) */
1745 * tree can be full, so it would need to grow in depth:
1746 * we need one credit to modify old root, credits for
1747 * new root will be added in split accounting
1752 * Index split can happen, we would need:
1753 * allocate intermediate indexes (bitmap + group)
1754 * + change two blocks at each level, but root (already included)
1756 needed += (depth * 2) + (depth * 2);
1758 /* any allocation modifies superblock */
1764 static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
1765 struct ext4_extent *ex,
1766 ext4_lblk_t from, ext4_lblk_t to)
1768 struct buffer_head *bh;
1769 unsigned short ee_len = ext4_ext_get_actual_len(ex);
1770 int i, metadata = 0;
1772 if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
1774 #ifdef EXTENTS_STATS
1776 struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
1777 spin_lock(&sbi->s_ext_stats_lock);
1778 sbi->s_ext_blocks += ee_len;
1779 sbi->s_ext_extents++;
1780 if (ee_len < sbi->s_ext_min)
1781 sbi->s_ext_min = ee_len;
1782 if (ee_len > sbi->s_ext_max)
1783 sbi->s_ext_max = ee_len;
1784 if (ext_depth(inode) > sbi->s_depth_max)
1785 sbi->s_depth_max = ext_depth(inode);
1786 spin_unlock(&sbi->s_ext_stats_lock);
1789 if (from >= le32_to_cpu(ex->ee_block)
1790 && to == le32_to_cpu(ex->ee_block) + ee_len - 1) {
1795 num = le32_to_cpu(ex->ee_block) + ee_len - from;
1796 start = ext_pblock(ex) + ee_len - num;
1797 ext_debug("free last %u blocks starting %llu\n", num, start);
1798 for (i = 0; i < num; i++) {
1799 bh = sb_find_get_block(inode->i_sb, start + i);
1800 ext4_forget(handle, 0, inode, bh, start + i);
1802 ext4_free_blocks(handle, inode, start, num, metadata);
1803 } else if (from == le32_to_cpu(ex->ee_block)
1804 && to <= le32_to_cpu(ex->ee_block) + ee_len - 1) {
1805 printk(KERN_INFO "strange request: removal %u-%u from %u:%u\n",
1806 from, to, le32_to_cpu(ex->ee_block), ee_len);
1808 printk(KERN_INFO "strange request: removal(2) "
1809 "%u-%u from %u:%u\n",
1810 from, to, le32_to_cpu(ex->ee_block), ee_len);
1816 ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
1817 struct ext4_ext_path *path, ext4_lblk_t start)
1819 int err = 0, correct_index = 0;
1820 int depth = ext_depth(inode), credits;
1821 struct ext4_extent_header *eh;
1822 ext4_lblk_t a, b, block;
1824 ext4_lblk_t ex_ee_block;
1825 unsigned short ex_ee_len;
1826 unsigned uninitialized = 0;
1827 struct ext4_extent *ex;
1829 /* the header must be checked already in ext4_ext_remove_space() */
1830 ext_debug("truncate since %u in leaf\n", start);
1831 if (!path[depth].p_hdr)
1832 path[depth].p_hdr = ext_block_hdr(path[depth].p_bh);
1833 eh = path[depth].p_hdr;
1836 /* find where to start removing */
1837 ex = EXT_LAST_EXTENT(eh);
1839 ex_ee_block = le32_to_cpu(ex->ee_block);
1840 if (ext4_ext_is_uninitialized(ex))
1842 ex_ee_len = ext4_ext_get_actual_len(ex);
1844 while (ex >= EXT_FIRST_EXTENT(eh) &&
1845 ex_ee_block + ex_ee_len > start) {
1846 ext_debug("remove ext %lu:%u\n", ex_ee_block, ex_ee_len);
1847 path[depth].p_ext = ex;
1849 a = ex_ee_block > start ? ex_ee_block : start;
1850 b = ex_ee_block + ex_ee_len - 1 < EXT_MAX_BLOCK ?
1851 ex_ee_block + ex_ee_len - 1 : EXT_MAX_BLOCK;
1853 ext_debug(" border %u:%u\n", a, b);
1855 if (a != ex_ee_block && b != ex_ee_block + ex_ee_len - 1) {
1859 } else if (a != ex_ee_block) {
1860 /* remove tail of the extent */
1861 block = ex_ee_block;
1863 } else if (b != ex_ee_block + ex_ee_len - 1) {
1864 /* remove head of the extent */
1867 /* there is no "make a hole" API yet */
1870 /* remove whole extent: excellent! */
1871 block = ex_ee_block;
1873 BUG_ON(a != ex_ee_block);
1874 BUG_ON(b != ex_ee_block + ex_ee_len - 1);
1877 /* at present, extent can't cross block group: */
1878 /* leaf + bitmap + group desc + sb + inode */
1880 if (ex == EXT_FIRST_EXTENT(eh)) {
1882 credits += (ext_depth(inode)) + 1;
1885 credits += 2 * EXT4_QUOTA_TRANS_BLOCKS(inode->i_sb);
1888 handle = ext4_ext_journal_restart(handle, credits);
1889 if (IS_ERR(handle)) {
1890 err = PTR_ERR(handle);
1894 err = ext4_ext_get_access(handle, inode, path + depth);
1898 err = ext4_remove_blocks(handle, inode, ex, a, b);
1903 /* this extent is removed; mark slot entirely unused */
1904 ext4_ext_store_pblock(ex, 0);
1905 eh->eh_entries = cpu_to_le16(le16_to_cpu(eh->eh_entries)-1);
1908 ex->ee_block = cpu_to_le32(block);
1909 ex->ee_len = cpu_to_le16(num);
1911 * Do not mark uninitialized if all the blocks in the
1912 * extent have been removed.
1914 if (uninitialized && num)
1915 ext4_ext_mark_uninitialized(ex);
1917 err = ext4_ext_dirty(handle, inode, path + depth);
1921 ext_debug("new extent: %u:%u:%llu\n", block, num,
1924 ex_ee_block = le32_to_cpu(ex->ee_block);
1925 ex_ee_len = ext4_ext_get_actual_len(ex);
1928 if (correct_index && eh->eh_entries)
1929 err = ext4_ext_correct_indexes(handle, inode, path);
1931 /* if this leaf is free, then we should
1932 * remove it from index block above */
1933 if (err == 0 && eh->eh_entries == 0 && path[depth].p_bh != NULL)
1934 err = ext4_ext_rm_idx(handle, inode, path + depth);
1941 * ext4_ext_more_to_rm:
1942 * returns 1 if current index has to be freed (even partial)
1945 ext4_ext_more_to_rm(struct ext4_ext_path *path)
1947 BUG_ON(path->p_idx == NULL);
1949 if (path->p_idx < EXT_FIRST_INDEX(path->p_hdr))
1953 * if truncate on deeper level happened, it wasn't partial,
1954 * so we have to consider current index for truncation
1956 if (le16_to_cpu(path->p_hdr->eh_entries) == path->p_block)
1961 static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start)
1963 struct super_block *sb = inode->i_sb;
1964 int depth = ext_depth(inode);
1965 struct ext4_ext_path *path;
1969 ext_debug("truncate since %u\n", start);
1971 /* probably first extent we're gonna free will be last in block */
1972 handle = ext4_journal_start(inode, depth + 1);
1974 return PTR_ERR(handle);
1976 ext4_ext_invalidate_cache(inode);
1979 * We start scanning from right side, freeing all the blocks
1980 * after i_size and walking into the tree depth-wise.
1982 path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1), GFP_KERNEL);
1984 ext4_journal_stop(handle);
1987 path[0].p_hdr = ext_inode_hdr(inode);
1988 if (ext4_ext_check_header(inode, path[0].p_hdr, depth)) {
1992 path[0].p_depth = depth;
1994 while (i >= 0 && err == 0) {
1996 /* this is leaf block */
1997 err = ext4_ext_rm_leaf(handle, inode, path, start);
1998 /* root level has p_bh == NULL, brelse() eats this */
1999 brelse(path[i].p_bh);
2000 path[i].p_bh = NULL;
2005 /* this is index block */
2006 if (!path[i].p_hdr) {
2007 ext_debug("initialize header\n");
2008 path[i].p_hdr = ext_block_hdr(path[i].p_bh);
2011 if (!path[i].p_idx) {
2012 /* this level hasn't been touched yet */
2013 path[i].p_idx = EXT_LAST_INDEX(path[i].p_hdr);
2014 path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries)+1;
2015 ext_debug("init index ptr: hdr 0x%p, num %d\n",
2017 le16_to_cpu(path[i].p_hdr->eh_entries));
2019 /* we were already here, see at next index */
2023 ext_debug("level %d - index, first 0x%p, cur 0x%p\n",
2024 i, EXT_FIRST_INDEX(path[i].p_hdr),
2026 if (ext4_ext_more_to_rm(path + i)) {
2027 struct buffer_head *bh;
2028 /* go to the next level */
2029 ext_debug("move to level %d (block %llu)\n",
2030 i + 1, idx_pblock(path[i].p_idx));
2031 memset(path + i + 1, 0, sizeof(*path));
2032 bh = sb_bread(sb, idx_pblock(path[i].p_idx));
2034 /* should we reset i_size? */
2038 if (WARN_ON(i + 1 > depth)) {
2042 if (ext4_ext_check_header(inode, ext_block_hdr(bh),
2047 path[i + 1].p_bh = bh;
2049 /* save actual number of indexes since this
2050 * number is changed at the next iteration */
2051 path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries);
2054 /* we finished processing this index, go up */
2055 if (path[i].p_hdr->eh_entries == 0 && i > 0) {
2056 /* index is empty, remove it;
2057 * handle must be already prepared by the
2058 * truncatei_leaf() */
2059 err = ext4_ext_rm_idx(handle, inode, path + i);
2061 /* root level has p_bh == NULL, brelse() eats this */
2062 brelse(path[i].p_bh);
2063 path[i].p_bh = NULL;
2065 ext_debug("return to level %d\n", i);
2069 /* TODO: flexible tree reduction should be here */
2070 if (path->p_hdr->eh_entries == 0) {
2072 * truncate to zero freed all the tree,
2073 * so we need to correct eh_depth
2075 err = ext4_ext_get_access(handle, inode, path);
2077 ext_inode_hdr(inode)->eh_depth = 0;
2078 ext_inode_hdr(inode)->eh_max =
2079 cpu_to_le16(ext4_ext_space_root(inode));
2080 err = ext4_ext_dirty(handle, inode, path);
2084 ext4_ext_tree_changed(inode);
2085 ext4_ext_drop_refs(path);
2087 ext4_journal_stop(handle);
2093 * called at mount time
2095 void ext4_ext_init(struct super_block *sb)
2098 * possible initialization would be here
2101 if (test_opt(sb, EXTENTS)) {
2102 printk("EXT4-fs: file extents enabled");
2103 #ifdef AGGRESSIVE_TEST
2104 printk(", aggressive tests");
2106 #ifdef CHECK_BINSEARCH
2107 printk(", check binsearch");
2109 #ifdef EXTENTS_STATS
2113 #ifdef EXTENTS_STATS
2114 spin_lock_init(&EXT4_SB(sb)->s_ext_stats_lock);
2115 EXT4_SB(sb)->s_ext_min = 1 << 30;
2116 EXT4_SB(sb)->s_ext_max = 0;
2122 * called at umount time
2124 void ext4_ext_release(struct super_block *sb)
2126 if (!test_opt(sb, EXTENTS))
2129 #ifdef EXTENTS_STATS
2130 if (EXT4_SB(sb)->s_ext_blocks && EXT4_SB(sb)->s_ext_extents) {
2131 struct ext4_sb_info *sbi = EXT4_SB(sb);
2132 printk(KERN_ERR "EXT4-fs: %lu blocks in %lu extents (%lu ave)\n",
2133 sbi->s_ext_blocks, sbi->s_ext_extents,
2134 sbi->s_ext_blocks / sbi->s_ext_extents);
2135 printk(KERN_ERR "EXT4-fs: extents: %lu min, %lu max, max depth %lu\n",
2136 sbi->s_ext_min, sbi->s_ext_max, sbi->s_depth_max);
2142 * This function is called by ext4_ext_get_blocks() if someone tries to write
2143 * to an uninitialized extent. It may result in splitting the uninitialized
2144 * extent into multiple extents (upto three - one initialized and two
2146 * There are three possibilities:
2147 * a> There is no split required: Entire extent should be initialized
2148 * b> Splits in two extents: Write is happening at either end of the extent
2149 * c> Splits in three extents: Somone is writing in middle of the extent
2151 static int ext4_ext_convert_to_initialized(handle_t *handle,
2152 struct inode *inode,
2153 struct ext4_ext_path *path,
2155 unsigned long max_blocks)
2157 struct ext4_extent *ex, newex;
2158 struct ext4_extent *ex1 = NULL;
2159 struct ext4_extent *ex2 = NULL;
2160 struct ext4_extent *ex3 = NULL;
2161 struct ext4_extent_header *eh;
2162 ext4_lblk_t ee_block;
2163 unsigned int allocated, ee_len, depth;
2164 ext4_fsblk_t newblock;
2168 depth = ext_depth(inode);
2169 eh = path[depth].p_hdr;
2170 ex = path[depth].p_ext;
2171 ee_block = le32_to_cpu(ex->ee_block);
2172 ee_len = ext4_ext_get_actual_len(ex);
2173 allocated = ee_len - (iblock - ee_block);
2174 newblock = iblock - ee_block + ext_pblock(ex);
2177 err = ext4_ext_get_access(handle, inode, path + depth);
2181 /* ex1: ee_block to iblock - 1 : uninitialized */
2182 if (iblock > ee_block) {
2184 ex1->ee_len = cpu_to_le16(iblock - ee_block);
2185 ext4_ext_mark_uninitialized(ex1);
2189 * for sanity, update the length of the ex2 extent before
2190 * we insert ex3, if ex1 is NULL. This is to avoid temporary
2191 * overlap of blocks.
2193 if (!ex1 && allocated > max_blocks)
2194 ex2->ee_len = cpu_to_le16(max_blocks);
2195 /* ex3: to ee_block + ee_len : uninitialised */
2196 if (allocated > max_blocks) {
2197 unsigned int newdepth;
2199 ex3->ee_block = cpu_to_le32(iblock + max_blocks);
2200 ext4_ext_store_pblock(ex3, newblock + max_blocks);
2201 ex3->ee_len = cpu_to_le16(allocated - max_blocks);
2202 ext4_ext_mark_uninitialized(ex3);
2203 err = ext4_ext_insert_extent(handle, inode, path, ex3);
2207 * The depth, and hence eh & ex might change
2208 * as part of the insert above.
2210 newdepth = ext_depth(inode);
2211 if (newdepth != depth) {
2213 ext4_ext_drop_refs(path);
2214 path = ext4_ext_find_extent(inode, iblock, path);
2216 err = PTR_ERR(path);
2219 eh = path[depth].p_hdr;
2220 ex = path[depth].p_ext;
2224 err = ext4_ext_get_access(handle, inode, path + depth);
2228 allocated = max_blocks;
2231 * If there was a change of depth as part of the
2232 * insertion of ex3 above, we need to update the length
2233 * of the ex1 extent again here
2235 if (ex1 && ex1 != ex) {
2237 ex1->ee_len = cpu_to_le16(iblock - ee_block);
2238 ext4_ext_mark_uninitialized(ex1);
2241 /* ex2: iblock to iblock + maxblocks-1 : initialised */
2242 ex2->ee_block = cpu_to_le32(iblock);
2243 ext4_ext_store_pblock(ex2, newblock);
2244 ex2->ee_len = cpu_to_le16(allocated);
2248 * New (initialized) extent starts from the first block
2249 * in the current extent. i.e., ex2 == ex
2250 * We have to see if it can be merged with the extent
2253 if (ex2 > EXT_FIRST_EXTENT(eh)) {
2255 * To merge left, pass "ex2 - 1" to try_to_merge(),
2256 * since it merges towards right _only_.
2258 ret = ext4_ext_try_to_merge(inode, path, ex2 - 1);
2260 err = ext4_ext_correct_indexes(handle, inode, path);
2263 depth = ext_depth(inode);
2268 * Try to Merge towards right. This might be required
2269 * only when the whole extent is being written to.
2270 * i.e. ex2 == ex and ex3 == NULL.
2273 ret = ext4_ext_try_to_merge(inode, path, ex2);
2275 err = ext4_ext_correct_indexes(handle, inode, path);
2280 /* Mark modified extent as dirty */
2281 err = ext4_ext_dirty(handle, inode, path + depth);
2284 err = ext4_ext_insert_extent(handle, inode, path, &newex);
2286 return err ? err : allocated;
2290 * Block allocation/map/preallocation routine for extents based files
2293 * Need to be called with
2294 * down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system block
2295 * (ie, create is zero). Otherwise down_write(&EXT4_I(inode)->i_data_sem)
2297 * return > 0, number of of blocks already mapped/allocated
2298 * if create == 0 and these are pre-allocated blocks
2299 * buffer head is unmapped
2300 * otherwise blocks are mapped
2302 * return = 0, if plain look up failed (blocks have not been allocated)
2303 * buffer head is unmapped
2305 * return < 0, error case.
2307 int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
2309 unsigned long max_blocks, struct buffer_head *bh_result,
2310 int create, int extend_disksize)
2312 struct ext4_ext_path *path = NULL;
2313 struct ext4_extent_header *eh;
2314 struct ext4_extent newex, *ex;
2315 ext4_fsblk_t goal, newblock;
2316 int err = 0, depth, ret;
2317 unsigned long allocated = 0;
2318 struct ext4_allocation_request ar;
2320 __clear_bit(BH_New, &bh_result->b_state);
2321 ext_debug("blocks %u/%lu requested for inode %u\n",
2322 iblock, max_blocks, inode->i_ino);
2324 /* check in cache */
2325 goal = ext4_ext_in_cache(inode, iblock, &newex);
2327 if (goal == EXT4_EXT_CACHE_GAP) {
2330 * block isn't allocated yet and
2331 * user doesn't want to allocate it
2335 /* we should allocate requested block */
2336 } else if (goal == EXT4_EXT_CACHE_EXTENT) {
2337 /* block is already allocated */
2339 - le32_to_cpu(newex.ee_block)
2340 + ext_pblock(&newex);
2341 /* number of remaining blocks in the extent */
2342 allocated = ext4_ext_get_actual_len(&newex) -
2343 (iblock - le32_to_cpu(newex.ee_block));
2350 /* find extent for this block */
2351 path = ext4_ext_find_extent(inode, iblock, NULL);
2353 err = PTR_ERR(path);
2358 depth = ext_depth(inode);
2361 * consistent leaf must not be empty;
2362 * this situation is possible, though, _during_ tree modification;
2363 * this is why assert can't be put in ext4_ext_find_extent()
2365 BUG_ON(path[depth].p_ext == NULL && depth != 0);
2366 eh = path[depth].p_hdr;
2368 ex = path[depth].p_ext;
2370 ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block);
2371 ext4_fsblk_t ee_start = ext_pblock(ex);
2372 unsigned short ee_len;
2375 * Uninitialized extents are treated as holes, except that
2376 * we split out initialized portions during a write.
2378 ee_len = ext4_ext_get_actual_len(ex);
2379 /* if found extent covers block, simply return it */
2380 if (iblock >= ee_block && iblock < ee_block + ee_len) {
2381 newblock = iblock - ee_block + ee_start;
2382 /* number of remaining blocks in the extent */
2383 allocated = ee_len - (iblock - ee_block);
2384 ext_debug("%u fit into %lu:%d -> %llu\n", iblock,
2385 ee_block, ee_len, newblock);
2387 /* Do not put uninitialized extent in the cache */
2388 if (!ext4_ext_is_uninitialized(ex)) {
2389 ext4_ext_put_in_cache(inode, ee_block,
2391 EXT4_EXT_CACHE_EXTENT);
2394 if (create == EXT4_CREATE_UNINITIALIZED_EXT)
2399 ret = ext4_ext_convert_to_initialized(handle, inode,
2412 * requested block isn't allocated yet;
2413 * we couldn't try to create block if create flag is zero
2417 * put just found gap into cache to speed up
2418 * subsequent requests
2420 ext4_ext_put_gap_in_cache(inode, path, iblock);
2424 * Okay, we need to do block allocation. Lazily initialize the block
2425 * allocation info here if necessary.
2427 if (S_ISREG(inode->i_mode) && (!EXT4_I(inode)->i_block_alloc_info))
2428 ext4_init_block_alloc_info(inode);
2430 /* find neighbour allocated blocks */
2432 err = ext4_ext_search_left(inode, path, &ar.lleft, &ar.pleft);
2436 err = ext4_ext_search_right(inode, path, &ar.lright, &ar.pright);
2441 * See if request is beyond maximum number of blocks we can have in
2442 * a single extent. For an initialized extent this limit is
2443 * EXT_INIT_MAX_LEN and for an uninitialized extent this limit is
2444 * EXT_UNINIT_MAX_LEN.
2446 if (max_blocks > EXT_INIT_MAX_LEN &&
2447 create != EXT4_CREATE_UNINITIALIZED_EXT)
2448 max_blocks = EXT_INIT_MAX_LEN;
2449 else if (max_blocks > EXT_UNINIT_MAX_LEN &&
2450 create == EXT4_CREATE_UNINITIALIZED_EXT)
2451 max_blocks = EXT_UNINIT_MAX_LEN;
2453 /* Check if we can really insert (iblock)::(iblock+max_blocks) extent */
2454 newex.ee_block = cpu_to_le32(iblock);
2455 newex.ee_len = cpu_to_le16(max_blocks);
2456 err = ext4_ext_check_overlap(inode, &newex, path);
2458 allocated = ext4_ext_get_actual_len(&newex);
2460 allocated = max_blocks;
2462 /* allocate new block */
2464 ar.goal = ext4_ext_find_goal(inode, path, iblock);
2465 ar.logical = iblock;
2467 if (S_ISREG(inode->i_mode))
2468 ar.flags = EXT4_MB_HINT_DATA;
2470 /* disable in-core preallocation for non-regular files */
2472 newblock = ext4_mb_new_blocks(handle, &ar, &err);
2475 ext_debug("allocate new block: goal %llu, found %llu/%lu\n",
2476 goal, newblock, allocated);
2478 /* try to insert new extent into found leaf and return */
2479 ext4_ext_store_pblock(&newex, newblock);
2480 newex.ee_len = cpu_to_le16(ar.len);
2481 if (create == EXT4_CREATE_UNINITIALIZED_EXT) /* Mark uninitialized */
2482 ext4_ext_mark_uninitialized(&newex);
2483 err = ext4_ext_insert_extent(handle, inode, path, &newex);
2485 /* free data blocks we just allocated */
2486 /* not a good idea to call discard here directly,
2487 * but otherwise we'd need to call it every free() */
2488 ext4_mb_discard_inode_preallocations(inode);
2489 ext4_free_blocks(handle, inode, ext_pblock(&newex),
2490 ext4_ext_get_actual_len(&newex), 0);
2494 if (extend_disksize && inode->i_size > EXT4_I(inode)->i_disksize)
2495 EXT4_I(inode)->i_disksize = inode->i_size;
2497 /* previous routine could use block we allocated */
2498 newblock = ext_pblock(&newex);
2499 allocated = ext4_ext_get_actual_len(&newex);
2501 __set_bit(BH_New, &bh_result->b_state);
2503 /* Cache only when it is _not_ an uninitialized extent */
2504 if (create != EXT4_CREATE_UNINITIALIZED_EXT)
2505 ext4_ext_put_in_cache(inode, iblock, allocated, newblock,
2506 EXT4_EXT_CACHE_EXTENT);
2508 if (allocated > max_blocks)
2509 allocated = max_blocks;
2510 ext4_ext_show_leaf(inode, path);
2511 __set_bit(BH_Mapped, &bh_result->b_state);
2512 bh_result->b_bdev = inode->i_sb->s_bdev;
2513 bh_result->b_blocknr = newblock;
2516 ext4_ext_drop_refs(path);
2519 return err ? err : allocated;
2522 void ext4_ext_truncate(struct inode * inode, struct page *page)
2524 struct address_space *mapping = inode->i_mapping;
2525 struct super_block *sb = inode->i_sb;
2526 ext4_lblk_t last_block;
2531 * probably first extent we're gonna free will be last in block
2533 err = ext4_writepage_trans_blocks(inode) + 3;
2534 handle = ext4_journal_start(inode, err);
2535 if (IS_ERR(handle)) {
2537 clear_highpage(page);
2538 flush_dcache_page(page);
2540 page_cache_release(page);
2546 ext4_block_truncate_page(handle, page, mapping, inode->i_size);
2548 down_write(&EXT4_I(inode)->i_data_sem);
2549 ext4_ext_invalidate_cache(inode);
2551 ext4_mb_discard_inode_preallocations(inode);
2554 * TODO: optimization is possible here.
2555 * Probably we need not scan at all,
2556 * because page truncation is enough.
2558 if (ext4_orphan_add(handle, inode))
2561 /* we have to know where to truncate from in crash case */
2562 EXT4_I(inode)->i_disksize = inode->i_size;
2563 ext4_mark_inode_dirty(handle, inode);
2565 last_block = (inode->i_size + sb->s_blocksize - 1)
2566 >> EXT4_BLOCK_SIZE_BITS(sb);
2567 err = ext4_ext_remove_space(inode, last_block);
2569 /* In a multi-transaction truncate, we only make the final
2570 * transaction synchronous.
2577 * If this was a simple ftruncate() and the file will remain alive,
2578 * then we need to clear up the orphan record which we created above.
2579 * However, if this was a real unlink then we were called by
2580 * ext4_delete_inode(), and we allow that function to clean up the
2581 * orphan info for us.
2584 ext4_orphan_del(handle, inode);
2586 up_write(&EXT4_I(inode)->i_data_sem);
2587 ext4_journal_stop(handle);
2591 * ext4_ext_writepage_trans_blocks:
2592 * calculate max number of blocks we could modify
2593 * in order to allocate new block for an inode
2595 int ext4_ext_writepage_trans_blocks(struct inode *inode, int num)
2599 needed = ext4_ext_calc_credits_for_insert(inode, NULL);
2601 /* caller wants to allocate num blocks, but note it includes sb */
2602 needed = needed * num - (num - 1);
2605 needed += 2 * EXT4_QUOTA_TRANS_BLOCKS(inode->i_sb);
2612 * preallocate space for a file. This implements ext4's fallocate inode
2613 * operation, which gets called from sys_fallocate system call.
2614 * For block-mapped files, posix_fallocate should fall back to the method
2615 * of writing zeroes to the required new blocks (the same behavior which is
2616 * expected for file systems which do not support fallocate() system call).
2618 long ext4_fallocate(struct inode *inode, int mode, loff_t offset, loff_t len)
2622 unsigned long max_blocks;
2623 ext4_fsblk_t nblocks = 0;
2627 struct buffer_head map_bh;
2628 unsigned int credits, blkbits = inode->i_blkbits;
2631 * currently supporting (pre)allocate mode for extent-based
2634 if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
2637 /* preallocation to directories is currently not supported */
2638 if (S_ISDIR(inode->i_mode))
2641 block = offset >> blkbits;
2642 max_blocks = (EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits)
2646 * credits to insert 1 extent into extent tree + buffers to be able to
2647 * modify 1 super block, 1 block bitmap and 1 group descriptor.
2649 credits = EXT4_DATA_TRANS_BLOCKS(inode->i_sb) + 3;
2650 mutex_lock(&inode->i_mutex);
2652 while (ret >= 0 && ret < max_blocks) {
2653 block = block + ret;
2654 max_blocks = max_blocks - ret;
2655 handle = ext4_journal_start(inode, credits);
2656 if (IS_ERR(handle)) {
2657 ret = PTR_ERR(handle);
2661 ret = ext4_get_blocks_wrap(handle, inode, block,
2662 max_blocks, &map_bh,
2663 EXT4_CREATE_UNINITIALIZED_EXT, 0);
2667 printk(KERN_ERR "%s: ext4_ext_get_blocks "
2668 "returned error inode#%lu, block=%u, "
2669 "max_blocks=%lu", __func__,
2670 inode->i_ino, block, max_blocks);
2672 ext4_mark_inode_dirty(handle, inode);
2673 ret2 = ext4_journal_stop(handle);
2677 /* check wrap through sign-bit/zero here */
2678 if ((block + ret) < 0 || (block + ret) < block) {
2680 ext4_mark_inode_dirty(handle, inode);
2681 ret2 = ext4_journal_stop(handle);
2684 if (buffer_new(&map_bh) && ((block + ret) >
2685 (EXT4_BLOCK_ALIGN(i_size_read(inode), blkbits)
2687 nblocks = nblocks + ret;
2690 /* Update ctime if new blocks get allocated */
2692 struct timespec now;
2694 now = current_fs_time(inode->i_sb);
2695 if (!timespec_equal(&inode->i_ctime, &now))
2696 inode->i_ctime = now;
2699 ext4_mark_inode_dirty(handle, inode);
2700 ret2 = ext4_journal_stop(handle);
2705 if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))
2709 * Time to update the file size.
2710 * Update only when preallocation was requested beyond the file size.
2712 if (!(mode & FALLOC_FL_KEEP_SIZE) &&
2713 (offset + len) > i_size_read(inode)) {
2716 * if no error, we assume preallocation succeeded
2719 i_size_write(inode, offset + len);
2720 EXT4_I(inode)->i_disksize = i_size_read(inode);
2721 } else if (ret < 0 && nblocks) {
2722 /* Handle partial allocation scenario */
2725 newsize = (nblocks << blkbits) + i_size_read(inode);
2726 i_size_write(inode, EXT4_BLOCK_ALIGN(newsize, blkbits));
2727 EXT4_I(inode)->i_disksize = i_size_read(inode);
2731 mutex_unlock(&inode->i_mutex);
2732 return ret > 0 ? ret2 : ret;