2 * Copyright (c) 2015 Grzegorz Kostka (kostka.grzegorz@gmail.com)
3 * Copyright (c) 2015 Kaho Ng (ngkaho1234@gmail.com)
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
9 * - Redistributions of source code must retain the above copyright
10 * notice, this list of conditions and the following disclaimer.
11 * - Redistributions in binary form must reproduce the above copyright
12 * notice, this list of conditions and the following disclaimer in the
13 * documentation and/or other materials provided with the distribution.
14 * - The name of the author may not be used to endorse or promote products
15 * derived from this software without specific prior written permission.
17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
18 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
19 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
20 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
21 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
22 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
23 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
24 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
25 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
26 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
29 #include "ext4_config.h"
30 #include "ext4_blockdev.h"
32 #include "ext4_super.h"
33 #include "ext4_crc32c.h"
34 #include "ext4_balloc.h"
35 #include "ext4_debug.h"
42 #include "ext4_extent.h"
44 #if CONFIG_EXTENT_FULL
47 * used by extent splitting.
49 #define EXT4_EXT_MARK_UNWRIT1 0x02 /* mark first half unwritten */
50 #define EXT4_EXT_MARK_UNWRIT2 0x04 /* mark second half unwritten */
51 #define EXT4_EXT_DATA_VALID1 0x08 /* first half contains valid data */
52 #define EXT4_EXT_DATA_VALID2 0x10 /* second half contains valid data */
53 #define EXT4_EXT_NO_COMBINE 0x20 /* do not combine two extents */
55 static struct ext4_extent_tail *
56 find_ext4_extent_tail(struct ext4_extent_header *eh)
58 return (struct ext4_extent_tail *)(((char *)eh) +
59 EXT4_EXTENT_TAIL_OFFSET(eh));
62 static struct ext4_extent_header *ext_inode_hdr(struct ext4_inode *inode)
64 return (struct ext4_extent_header *)inode->blocks;
67 static struct ext4_extent_header *ext_block_hdr(struct ext4_block *block)
69 return (struct ext4_extent_header *)block->data;
72 static uint16_t ext_depth(struct ext4_inode *inode)
74 return to_le16(ext_inode_hdr(inode)->depth);
77 static uint16_t ext4_ext_get_actual_len(struct ext4_extent *ext)
79 return (to_le16(ext->block_count) <= EXT_INIT_MAX_LEN
80 ? to_le16(ext->block_count)
81 : (to_le16(ext->block_count) - EXT_INIT_MAX_LEN));
84 static void ext4_ext_mark_initialized(struct ext4_extent *ext)
86 ext->block_count = to_le16(ext4_ext_get_actual_len(ext));
89 static void ext4_ext_mark_unwritten(struct ext4_extent *ext)
91 ext->block_count |= to_le16(EXT_INIT_MAX_LEN);
94 static int ext4_ext_is_unwritten(struct ext4_extent *ext)
96 /* Extent with ee_len of 0x8000 is treated as an initialized extent */
97 return (to_le16(ext->block_count) > EXT_INIT_MAX_LEN);
102 * combine low and high parts of physical block number into ext4_fsblk_t
104 static ext4_fsblk_t ext4_ext_pblock(struct ext4_extent *ex)
108 block = to_le32(ex->start_lo);
109 block |= ((ext4_fsblk_t)to_le16(ex->start_hi) << 31) << 1;
115 * combine low and high parts of a leaf physical block number into ext4_fsblk_t
117 static ext4_fsblk_t ext4_idx_pblock(struct ext4_extent_index *ix)
121 block = to_le32(ix->leaf_lo);
122 block |= ((ext4_fsblk_t)to_le16(ix->leaf_hi) << 31) << 1;
127 * ext4_ext_store_pblock:
128 * stores a large physical block number into an extent struct,
129 * breaking it into parts
131 static void ext4_ext_store_pblock(struct ext4_extent *ex, ext4_fsblk_t pb)
133 ex->start_lo = to_le32((uint32_t)(pb & 0xffffffff));
134 ex->start_hi = to_le16((uint16_t)((pb >> 32)) & 0xffff);
138 * ext4_idx_store_pblock:
139 * stores a large physical block number into an index struct,
140 * breaking it into parts
142 static void ext4_idx_store_pblock(struct ext4_extent_index *ix, ext4_fsblk_t pb)
144 ix->leaf_lo = to_le32((uint32_t)(pb & 0xffffffff));
145 ix->leaf_hi = to_le16((uint16_t)((pb >> 32)) & 0xffff);
148 static int ext4_allocate_single_block(struct ext4_inode_ref *inode_ref,
150 ext4_fsblk_t *blockp)
152 return ext4_balloc_alloc_block(inode_ref, goal, blockp);
155 static ext4_fsblk_t ext4_new_meta_blocks(struct ext4_inode_ref *inode_ref,
157 uint32_t flags __unused,
158 uint32_t *count, int *errp)
160 ext4_fsblk_t block = 0;
162 *errp = ext4_allocate_single_block(inode_ref, goal, &block);
168 static void ext4_ext_free_blocks(struct ext4_inode_ref *inode_ref,
169 ext4_fsblk_t block, uint32_t count,
170 uint32_t flags __unused)
172 ext4_balloc_free_blocks(inode_ref, block, count);
175 static size_t ext4_ext_space_block(struct ext4_inode_ref *inode_ref)
178 uint32_t block_size = ext4_sb_get_block_size(&inode_ref->fs->sb);
180 size = (block_size - sizeof(struct ext4_extent_header)) /
181 sizeof(struct ext4_extent);
182 #ifdef AGGRESSIVE_TEST
189 static size_t ext4_ext_space_block_idx(struct ext4_inode_ref *inode_ref)
192 uint32_t block_size = ext4_sb_get_block_size(&inode_ref->fs->sb);
194 size = (block_size - sizeof(struct ext4_extent_header)) /
195 sizeof(struct ext4_extent_index);
196 #ifdef AGGRESSIVE_TEST
203 static size_t ext4_ext_space_root(struct ext4_inode_ref *inode_ref)
207 size = sizeof(inode_ref->inode->blocks);
208 size -= sizeof(struct ext4_extent_header);
209 size /= sizeof(struct ext4_extent);
210 #ifdef AGGRESSIVE_TEST
217 static size_t ext4_ext_space_root_idx(struct ext4_inode_ref *inode_ref)
221 size = sizeof(inode_ref->inode->blocks);
222 size -= sizeof(struct ext4_extent_header);
223 size /= sizeof(struct ext4_extent_index);
224 #ifdef AGGRESSIVE_TEST
231 static size_t ext4_ext_max_entries(struct ext4_inode_ref *inode_ref,
236 if (depth == ext_depth(inode_ref->inode)) {
238 max = ext4_ext_space_root(inode_ref);
240 max = ext4_ext_space_root_idx(inode_ref);
243 max = ext4_ext_space_block(inode_ref);
245 max = ext4_ext_space_block_idx(inode_ref);
251 static ext4_fsblk_t ext4_ext_find_goal(struct ext4_inode_ref *inode_ref,
252 struct ext4_extent_path *path,
256 uint32_t depth = path->depth;
257 struct ext4_extent *ex;
260 * Try to predict block placement assuming that we are
261 * filling in a file which will eventually be
262 * non-sparse --- i.e., in the case of libbfd writing
263 * an ELF object sections out-of-order but in a way
264 * the eventually results in a contiguous object or
265 * executable file, or some database extending a table
266 * space file. However, this is actually somewhat
267 * non-ideal if we are writing a sparse file such as
268 * qemu or KVM writing a raw image file that is going
269 * to stay fairly sparse, since it will end up
270 * fragmenting the file system's free space. Maybe we
271 * should have some hueristics or some way to allow
272 * userspace to pass a hint to file system,
273 * especially if the latter case turns out to be
276 ex = path[depth].extent;
278 ext4_fsblk_t ext_pblk = ext4_ext_pblock(ex);
279 ext4_lblk_t ext_block = to_le32(ex->first_block);
281 if (block > ext_block)
282 return ext_pblk + (block - ext_block);
284 return ext_pblk - (ext_block - block);
287 /* it looks like index is empty;
288 * try to find starting block from index itself */
289 if (path[depth].block.lb_id)
290 return path[depth].block.lb_id;
293 /* OK. use inode's group */
294 return ext4_fs_inode_to_goal_block(inode_ref);
298 * Allocation for a meta data block
300 static ext4_fsblk_t ext4_ext_new_meta_block(struct ext4_inode_ref *inode_ref,
301 struct ext4_extent_path *path,
302 struct ext4_extent *ex, int *err,
305 ext4_fsblk_t goal, newblock;
307 goal = ext4_ext_find_goal(inode_ref, path, to_le32(ex->first_block));
308 newblock = ext4_new_meta_blocks(inode_ref, goal, flags, NULL, err);
312 static uint32_t ext4_ext_block_csum(struct ext4_inode_ref *inode_ref,
313 struct ext4_extent_header *eh)
315 uint32_t checksum = 0;
316 struct ext4_sblock *sb = &inode_ref->fs->sb;
318 if (ext4_sb_feature_ro_com(sb, EXT4_FRO_COM_METADATA_CSUM)) {
319 uint32_t ino_index = to_le32(inode_ref->index);
321 to_le32(ext4_inode_get_generation(inode_ref->inode));
322 /* First calculate crc32 checksum against fs uuid */
323 checksum = ext4_crc32c(~0, sb->uuid, sizeof(sb->uuid));
324 /* Then calculate crc32 checksum against inode number
325 * and inode generation */
326 checksum = ext4_crc32c(checksum, &ino_index,
328 checksum = ext4_crc32c(checksum, &ino_gen,
330 /* Finally calculate crc32 checksum against
331 * the entire extent block up to the checksum field */
332 checksum = ext4_crc32c(checksum, eh,
333 EXT4_EXTENT_TAIL_OFFSET(eh));
338 static void ext4_extent_block_csum_set(struct ext4_inode_ref *inode_ref,
339 struct ext4_extent_header *eh)
341 struct ext4_extent_tail *tail;
343 tail = find_ext4_extent_tail(eh);
344 tail->et_checksum = to_le32(ext4_ext_block_csum(inode_ref, eh));
347 static int ext4_ext_dirty(struct ext4_inode_ref *inode_ref,
348 struct ext4_extent_path *path)
350 if (path->block.lb_id)
351 path->block.dirty = true;
353 inode_ref->dirty = true;
358 static void ext4_ext_drop_refs(struct ext4_inode_ref *inode_ref,
359 struct ext4_extent_path *path, bool keep_other)
370 for (i = 0; i <= depth; i++, path++) {
371 if (path->block.lb_id) {
372 if (path->block.dirty)
373 ext4_extent_block_csum_set(inode_ref,
376 ext4_block_set(inode_ref->fs->bdev, &path->block);
382 * Check that whether the basic information inside the extent header
385 static int ext4_ext_check(struct ext4_inode_ref *inode_ref,
386 struct ext4_extent_header *eh, uint16_t depth,
387 ext4_fsblk_t pblk __unused)
389 struct ext4_extent_tail *tail;
390 const char *error_msg;
393 if (to_le16(eh->magic) != EXT4_EXTENT_MAGIC) {
394 error_msg = "invalid magic";
397 if (to_le16(eh->depth) != depth) {
398 error_msg = "unexpected eh_depth";
401 if (eh->max_entries_count == 0) {
402 error_msg = "invalid eh_max";
405 if (to_le16(eh->entries_count) > to_le16(eh->max_entries_count)) {
406 error_msg = "invalid eh_entries";
410 tail = find_ext4_extent_tail(eh);
411 struct ext4_sblock *sb = &inode_ref->fs->sb;
412 if (ext4_sb_feature_ro_com(sb, EXT4_FRO_COM_METADATA_CSUM)) {
413 if (tail->et_checksum != to_le32(ext4_ext_block_csum(inode_ref, eh))) {
414 /* FIXME: Warning: extent checksum damaged? */
421 ext4_dbg(DEBUG_EXTENT, "Bad extents B+ tree block: %s. "
422 "Blocknr: %" PRId64 "\n",
427 static int read_extent_tree_block(struct ext4_inode_ref *inode_ref,
428 ext4_fsblk_t pblk, int32_t depth,
429 struct ext4_block *bh,
430 uint32_t flags __unused)
434 err = ext4_block_get(inode_ref->fs->bdev, bh, pblk);
438 err = ext4_ext_check(inode_ref, ext_block_hdr(bh), depth, pblk);
445 ext4_block_set(inode_ref->fs->bdev, bh);
451 * ext4_ext_binsearch_idx:
452 * binary search for the closest index of the given block
453 * the header must be checked before calling this
455 static void ext4_ext_binsearch_idx(struct ext4_extent_path *path,
458 struct ext4_extent_header *eh = path->header;
459 struct ext4_extent_index *r, *l, *m;
461 l = EXT_FIRST_INDEX(eh) + 1;
462 r = EXT_LAST_INDEX(eh);
465 if (block < to_le32(m->first_block))
475 * ext4_ext_binsearch:
476 * binary search for closest extent of the given block
477 * the header must be checked before calling this
479 static void ext4_ext_binsearch(struct ext4_extent_path *path, ext4_lblk_t block)
481 struct ext4_extent_header *eh = path->header;
482 struct ext4_extent *r, *l, *m;
484 if (eh->entries_count == 0) {
486 * this leaf is empty:
487 * we get such a leaf in split/add case
492 l = EXT_FIRST_EXTENT(eh) + 1;
493 r = EXT_LAST_EXTENT(eh);
497 if (block < to_le32(m->first_block))
503 path->extent = l - 1;
506 static int ext4_find_extent(struct ext4_inode_ref *inode_ref, ext4_lblk_t block,
507 struct ext4_extent_path **orig_path, uint32_t flags)
509 struct ext4_extent_header *eh;
510 struct ext4_block bh = EXT4_BLOCK_ZERO();
511 ext4_fsblk_t buf_block = 0;
512 struct ext4_extent_path *path = *orig_path;
513 int32_t depth, ppos = 0;
517 eh = ext_inode_hdr(inode_ref->inode);
518 depth = ext_depth(inode_ref->inode);
521 ext4_ext_drop_refs(inode_ref, path, 0);
522 if (depth > path[0].maxdepth) {
524 *orig_path = path = NULL;
528 int32_t path_depth = depth + 1;
529 /* account possible depth increase */
530 path = calloc(1, sizeof(struct ext4_extent_path) *
534 path[0].maxdepth = path_depth;
540 /* walk through the tree */
542 ext4_ext_binsearch_idx(path + ppos, block);
543 path[ppos].p_block = ext4_idx_pblock(path[ppos].index);
544 path[ppos].depth = i;
545 path[ppos].extent = NULL;
546 buf_block = path[ppos].p_block;
550 if (!path[ppos].block.lb_id ||
551 path[ppos].block.lb_id != buf_block) {
552 ret = read_extent_tree_block(inode_ref, buf_block, i,
558 ext4_block_set(inode_ref->fs->bdev, &bh);
563 eh = ext_block_hdr(&bh);
564 path[ppos].block = bh;
565 path[ppos].header = eh;
569 path[ppos].depth = i;
570 path[ppos].extent = NULL;
571 path[ppos].index = NULL;
574 ext4_ext_binsearch(path + ppos, block);
575 /* if not an empty leaf */
576 if (path[ppos].extent)
577 path[ppos].p_block = ext4_ext_pblock(path[ppos].extent);
585 ext4_ext_drop_refs(inode_ref, path, 0);
592 static void ext4_ext_init_header(struct ext4_inode_ref *inode_ref,
593 struct ext4_extent_header *eh, int32_t depth)
595 eh->entries_count = 0;
596 eh->max_entries_count = to_le16(ext4_ext_max_entries(inode_ref, depth));
597 eh->magic = to_le16(EXT4_EXTENT_MAGIC);
602 * Be cautious, the buffer_head returned is not yet mark dirtied. */
603 static int ext4_ext_split_node(struct ext4_inode_ref *inode_ref,
604 struct ext4_extent_path *path, int32_t at,
605 struct ext4_extent *newext,
606 ext4_fsblk_t *sibling, struct ext4_block *new_bh)
609 ext4_fsblk_t newblock;
610 struct ext4_block bh = EXT4_BLOCK_ZERO();
611 int32_t depth = ext_depth(inode_ref->inode);
613 ext4_assert(sibling);
615 /* FIXME: currently we split at the point after the current extent. */
616 newblock = ext4_ext_new_meta_block(inode_ref, path, newext, &ret, 0);
620 /* For write access.# */
621 ret = ext4_block_get(inode_ref->fs->bdev, &bh, newblock);
626 /* start copy from next extent */
627 ptrdiff_t m = EXT_MAX_EXTENT(path[at].header) - path[at].extent;
628 struct ext4_extent_header *neh;
629 neh = ext_block_hdr(&bh);
630 ext4_ext_init_header(inode_ref, neh, 0);
632 struct ext4_extent *ex;
633 ex = EXT_FIRST_EXTENT(neh);
634 memmove(ex, path[at].extent + 1,
635 sizeof(struct ext4_extent) * m);
637 to_le16(to_le16(neh->entries_count) + m);
638 path[at].header->entries_count = to_le16(
639 to_le16(path[at].header->entries_count) - m);
640 ret = ext4_ext_dirty(inode_ref, path + at);
645 ptrdiff_t m = EXT_MAX_INDEX(path[at].header) - path[at].index;
646 struct ext4_extent_header *neh;
647 neh = ext_block_hdr(&bh);
648 ext4_ext_init_header(inode_ref, neh, depth - at);
650 struct ext4_extent_index *ix;
651 ix = EXT_FIRST_INDEX(neh);
652 memmove(ix, path[at].index + 1,
653 sizeof(struct ext4_extent) * m);
655 to_le16(to_le16(neh->entries_count) + m);
656 path[at].header->entries_count = to_le16(
657 to_le16(path[at].header->entries_count) - m);
658 ret = ext4_ext_dirty(inode_ref, path + at);
666 ext4_block_set(inode_ref->fs->bdev, &bh);
669 ext4_ext_free_blocks(inode_ref, newblock, 1, 0);
678 static ext4_lblk_t ext4_ext_block_index(struct ext4_extent_header *eh)
681 return to_le32(EXT_FIRST_INDEX(eh)->first_block);
683 return to_le32(EXT_FIRST_EXTENT(eh)->first_block);
686 struct ext_split_trans {
688 struct ext4_extent_path path;
692 static int ext4_ext_insert_index(struct ext4_inode_ref *inode_ref,
693 struct ext4_extent_path *path,
695 struct ext4_extent *newext,
696 ext4_lblk_t insert_index,
697 ext4_fsblk_t insert_block,
698 struct ext_split_trans *spt,
701 struct ext4_extent_index *ix;
702 struct ext4_extent_path *curp = path + at;
703 struct ext4_block bh = EXT4_BLOCK_ZERO();
706 struct ext4_extent_header *eh;
710 if (curp->index && insert_index == to_le32(curp->index->first_block))
713 if (to_le16(curp->header->entries_count) ==
714 to_le16(curp->header->max_entries_count)) {
716 struct ext4_extent_header *neh;
717 err = ext4_ext_split_node(inode_ref, path, at, newext,
722 neh = ext_block_hdr(&bh);
723 if (insert_index > to_le32(curp->index->first_block)) {
724 /* Make decision which node should be used to
725 * insert the index.*/
726 if (to_le16(neh->entries_count) >
727 to_le16(curp->header->entries_count)) {
730 ix = EXT_LAST_INDEX(eh) + 1;
733 ix = EXT_FIRST_INDEX(eh);
738 ix = EXT_LAST_INDEX(eh);
747 if (curp->index == NULL) {
748 ix = EXT_FIRST_INDEX(eh);
750 } else if (insert_index > to_le32(curp->index->first_block)) {
752 ix = curp->index + 1;
759 len = EXT_LAST_INDEX(eh) - ix + 1;
760 ext4_assert(len >= 0);
762 memmove(ix + 1, ix, len * sizeof(struct ext4_extent_index));
764 if (ix > EXT_MAX_INDEX(eh)) {
769 ix->first_block = to_le32(insert_index);
770 ext4_idx_store_pblock(ix, insert_block);
771 eh->entries_count = to_le16(to_le16(eh->entries_count) + 1);
773 if (ix > EXT_LAST_INDEX(eh)) {
778 if (eh == curp->header)
779 err = ext4_ext_dirty(inode_ref, curp);
784 if (err != EOK || *need_grow) {
786 ext4_block_set(inode_ref->fs->bdev, &bh);
789 } else if (bh.lb_id) {
790 /* If we got a sibling leaf. */
791 ext4_extent_block_csum_set(inode_ref, ext_block_hdr(&bh));
794 spt->path.p_block = ext4_idx_pblock(ix);
795 spt->path.depth = to_le16(eh->depth);
796 spt->path.maxdepth = 0;
797 spt->path.extent = NULL;
798 spt->path.index = ix;
799 spt->path.header = eh;
800 spt->path.block = bh;
803 * If newext->ee_block can be included into the
806 if (to_le32(newext->first_block) >=
807 ext4_ext_block_index(ext_block_hdr(&bh)))
811 curp->p_block = ext4_idx_pblock(ix);
817 curp->p_block = ext4_idx_pblock(ix);
823 * ext4_ext_correct_indexes:
824 * if leaf gets modified and modified extent is first in the leaf,
825 * then we have to correct all indexes above.
827 static int ext4_ext_correct_indexes(struct ext4_inode_ref *inode_ref,
828 struct ext4_extent_path *path)
830 struct ext4_extent_header *eh;
831 int32_t depth = ext_depth(inode_ref->inode);
832 struct ext4_extent *ex;
837 eh = path[depth].header;
838 ex = path[depth].extent;
840 if (ex == NULL || eh == NULL) {
845 /* there is no tree at all */
849 if (ex != EXT_FIRST_EXTENT(eh)) {
850 /* we correct tree if first leaf got modified only */
855 * TODO: we need correction if border is smaller than current one
858 border = path[depth].extent->first_block;
859 path[k].index->first_block = border;
860 err = ext4_ext_dirty(inode_ref, path + k);
865 /* change all left-side indexes */
866 if (path[k + 1].index != EXT_FIRST_INDEX(path[k + 1].header))
868 path[k].index->first_block = border;
869 err = ext4_ext_dirty(inode_ref, path + k);
877 static bool ext4_ext_can_prepend(struct ext4_extent *ex1,
878 struct ext4_extent *ex2)
880 if (ext4_ext_pblock(ex2) + ext4_ext_get_actual_len(ex2) !=
881 ext4_ext_pblock(ex1))
884 #ifdef AGGRESSIVE_TEST
885 if (ext4_ext_get_actual_len(ex1) + ext4_ext_get_actual_len(ex2) > 4)
888 if (ext4_ext_is_unwritten(ex1)) {
889 if (ext4_ext_get_actual_len(ex1) +
890 ext4_ext_get_actual_len(ex2) >
891 EXT_UNWRITTEN_MAX_LEN)
893 } else if (ext4_ext_get_actual_len(ex1) + ext4_ext_get_actual_len(ex2) >
898 if (to_le32(ex2->first_block) + ext4_ext_get_actual_len(ex2) !=
899 to_le32(ex1->first_block))
905 static bool ext4_ext_can_append(struct ext4_extent *ex1,
906 struct ext4_extent *ex2)
908 if (ext4_ext_pblock(ex1) + ext4_ext_get_actual_len(ex1) !=
909 ext4_ext_pblock(ex2))
912 #ifdef AGGRESSIVE_TEST
913 if (ext4_ext_get_actual_len(ex1) + ext4_ext_get_actual_len(ex2) > 4)
916 if (ext4_ext_is_unwritten(ex1)) {
917 if (ext4_ext_get_actual_len(ex1) +
918 ext4_ext_get_actual_len(ex2) >
919 EXT_UNWRITTEN_MAX_LEN)
921 } else if (ext4_ext_get_actual_len(ex1) + ext4_ext_get_actual_len(ex2) >
926 if (to_le32(ex1->first_block) + ext4_ext_get_actual_len(ex1) !=
927 to_le32(ex2->first_block))
933 static int ext4_ext_insert_leaf(struct ext4_inode_ref *inode_ref,
934 struct ext4_extent_path *path,
936 struct ext4_extent *newext,
937 struct ext_split_trans *spt,
941 struct ext4_extent_path *curp = path + at;
942 struct ext4_extent *ex = curp->extent;
943 struct ext4_block bh = EXT4_BLOCK_ZERO();
947 struct ext4_extent_header *eh = NULL;
952 to_le32(newext->first_block) == to_le32(curp->extent->first_block))
955 if (!(flags & EXT4_EXT_NO_COMBINE)) {
956 if (curp->extent && ext4_ext_can_append(curp->extent, newext)) {
957 unwritten = ext4_ext_is_unwritten(curp->extent);
958 curp->extent->block_count =
959 to_le16(ext4_ext_get_actual_len(curp->extent) +
960 ext4_ext_get_actual_len(newext));
962 ext4_ext_mark_unwritten(curp->extent);
963 err = ext4_ext_dirty(inode_ref, curp);
968 ext4_ext_can_prepend(curp->extent, newext)) {
969 unwritten = ext4_ext_is_unwritten(curp->extent);
970 curp->extent->first_block = newext->first_block;
971 curp->extent->block_count =
972 to_le16(ext4_ext_get_actual_len(curp->extent) +
973 ext4_ext_get_actual_len(newext));
975 ext4_ext_mark_unwritten(curp->extent);
976 err = ext4_ext_dirty(inode_ref, curp);
981 if (to_le16(curp->header->entries_count) ==
982 to_le16(curp->header->max_entries_count)) {
984 struct ext4_extent_header *neh;
985 err = ext4_ext_split_node(inode_ref, path, at, newext,
990 neh = ext_block_hdr(&bh);
991 if (to_le32(newext->first_block) >
992 to_le32(curp->extent->first_block)) {
993 if (to_le16(neh->entries_count) >
994 to_le16(curp->header->entries_count)) {
997 ex = EXT_LAST_EXTENT(eh) + 1;
1000 ex = EXT_FIRST_EXTENT(eh);
1005 ex = EXT_LAST_EXTENT(eh);
1014 if (curp->extent == NULL) {
1015 ex = EXT_FIRST_EXTENT(eh);
1017 } else if (to_le32(newext->first_block) >
1018 to_le32(curp->extent->first_block)) {
1020 ex = curp->extent + 1;
1027 len = EXT_LAST_EXTENT(eh) - ex + 1;
1028 ext4_assert(len >= 0);
1030 memmove(ex + 1, ex, len * sizeof(struct ext4_extent));
1032 if (ex > EXT_MAX_EXTENT(eh)) {
1037 ex->first_block = newext->first_block;
1038 ex->block_count = newext->block_count;
1039 ext4_ext_store_pblock(ex, ext4_ext_pblock(newext));
1040 eh->entries_count = to_le16(to_le16(eh->entries_count) + 1);
1042 if (ex > EXT_LAST_EXTENT(eh)) {
1047 if (eh == curp->header) {
1048 err = ext4_ext_correct_indexes(inode_ref, path);
1051 err = ext4_ext_dirty(inode_ref, curp);
1056 if (err != EOK || *need_grow) {
1058 ext4_block_set(inode_ref->fs->bdev, &bh);
1061 } else if (bh.lb_id) {
1062 /* If we got a sibling leaf. */
1063 ext4_extent_block_csum_set(inode_ref, ext_block_hdr(&bh));
1066 spt->path.p_block = ext4_ext_pblock(ex);
1067 spt->path.depth = to_le16(eh->depth);
1068 spt->path.maxdepth = 0;
1069 spt->path.extent = ex;
1070 spt->path.index = NULL;
1071 spt->path.header = eh;
1072 spt->path.block = bh;
1075 * If newext->ee_block can be included into the
1078 if (to_le32(newext->first_block) >=
1079 ext4_ext_block_index(ext_block_hdr(&bh)))
1083 curp->p_block = ext4_ext_pblock(ex);
1089 curp->p_block = ext4_ext_pblock(ex);
1096 * ext4_ext_grow_indepth:
1097 * implements tree growing procedure:
1098 * - allocates new block
1099 * - moves top-level data (index block or leaf) into the new block
1100 * - initializes new top-level, creating index that points to the
1101 * just created block
1103 static int ext4_ext_grow_indepth(struct ext4_inode_ref *inode_ref,
1106 struct ext4_extent_header *neh;
1107 struct ext4_block bh = EXT4_BLOCK_ZERO();
1108 ext4_fsblk_t newblock, goal = 0;
1111 /* Try to prepend new index to old one */
1112 if (ext_depth(inode_ref->inode))
1113 goal = ext4_idx_pblock(
1114 EXT_FIRST_INDEX(ext_inode_hdr(inode_ref->inode)));
1116 goal = ext4_fs_inode_to_goal_block(inode_ref);
1118 newblock = ext4_new_meta_blocks(inode_ref, goal, flags, NULL, &err);
1123 err = ext4_block_get(inode_ref->fs->bdev, &bh, newblock);
1125 ext4_ext_free_blocks(inode_ref, newblock, 1, 0);
1129 /* move top-level index/leaf into new block */
1130 memmove(bh.data, inode_ref->inode->blocks,
1131 sizeof(inode_ref->inode->blocks));
1133 /* set size of new block */
1134 neh = ext_block_hdr(&bh);
1135 /* old root could have indexes or leaves
1136 * so calculate e_max right way */
1137 if (ext_depth(inode_ref->inode))
1138 neh->max_entries_count =
1139 to_le16(ext4_ext_space_block_idx(inode_ref));
1141 neh->max_entries_count =
1142 to_le16(ext4_ext_space_block(inode_ref));
1144 neh->magic = to_le16(EXT4_EXTENT_MAGIC);
1146 /* Update top-level index: num,max,pointer */
1147 neh = ext_inode_hdr(inode_ref->inode);
1148 neh->entries_count = to_le16(1);
1149 ext4_idx_store_pblock(EXT_FIRST_INDEX(neh), newblock);
1150 if (neh->depth == 0) {
1151 /* Root extent block becomes index block */
1152 neh->max_entries_count =
1153 to_le16(ext4_ext_space_root_idx(inode_ref));
1154 EXT_FIRST_INDEX(neh)
1155 ->first_block = EXT_FIRST_EXTENT(neh)->first_block;
1157 neh->depth = to_le16(to_le16(neh->depth) + 1);
1159 ext4_extent_block_csum_set(inode_ref, neh);
1161 inode_ref->dirty = true;
1162 ext4_block_set(inode_ref->fs->bdev, &bh);
1167 __unused static void print_path(struct ext4_extent_path *path)
1169 int32_t i = path->depth;
1174 ? (path->extent - EXT_FIRST_EXTENT(path->header))
1178 ? (path->index - EXT_FIRST_INDEX(path->header))
1183 ext4_dbg(DEBUG_EXTENT,
1184 "depth %" PRId32 ", p_block: %" PRIu64 ","
1185 "p_ext offset: %td, p_idx offset: %td\n",
1186 i, path->p_block, a, b);
1192 static void ext4_ext_replace_path(struct ext4_inode_ref *inode_ref,
1193 struct ext4_extent_path *path,
1194 struct ext_split_trans *spt,
1197 int32_t depth = ext_depth(inode_ref->inode);
1198 int32_t i = depth - level;
1199 ext4_ext_drop_refs(inode_ref, path + i, 1);
1200 path[i] = spt[level].path;
1203 static int ext4_ext_insert_extent(struct ext4_inode_ref *inode_ref,
1204 struct ext4_extent_path **ppath,
1205 struct ext4_extent *newext, uint32_t flags)
1207 int32_t i, depth, level;
1209 ext4_fsblk_t ptr = 0;
1210 bool need_grow = false;
1211 struct ext4_extent_path *path = *ppath;
1212 struct ext_split_trans *spt = NULL;
1213 struct ext_split_trans newblock;
1215 memset(&newblock, 0, sizeof(newblock));
1217 depth = ext_depth(inode_ref->inode);
1218 for (i = depth, level = 0; i >= 0; i--, level++)
1219 if (EXT_HAS_FREE_INDEX(path + i))
1223 spt = calloc(1, sizeof(struct ext_split_trans) * (level));
1231 depth = ext_depth(inode_ref->inode);
1235 ret = ext4_ext_insert_leaf(inode_ref, path, depth - i,
1236 newext, &newblock, flags,
1239 ret = ext4_ext_insert_index(
1240 inode_ref, path, depth - i, newext,
1241 ext4_ext_block_index(
1242 ext_block_hdr(&spt[i - 1].path.block)),
1243 spt[i - 1].ptr, &newblock,
1251 else if (spt && ptr && !ret) {
1252 /* Prepare for the next iteration after splitting. */
1257 } while (ptr != 0 && i <= depth);
1260 ret = ext4_ext_grow_indepth(inode_ref, 0);
1263 ret = ext4_find_extent(inode_ref, to_le32(newext->first_block),
1274 ext4_ext_drop_refs(inode_ref, path, 0);
1276 while (--level >= 0 && spt) {
1277 if (spt[level].ptr) {
1278 ext4_ext_free_blocks(inode_ref, spt[level].ptr,
1280 ext4_ext_drop_refs(inode_ref, &spt[level].path,
1285 while (--level >= 0 && spt) {
1286 if (spt[level].switch_to)
1287 ext4_ext_replace_path(inode_ref, path, spt,
1289 else if (spt[level].ptr)
1290 ext4_ext_drop_refs(inode_ref, &spt[level].path,
1300 static void ext4_ext_remove_blocks(struct ext4_inode_ref *inode_ref,
1301 struct ext4_extent *ex, ext4_lblk_t from,
1304 ext4_lblk_t len = to - from + 1;
1307 num = from - to_le32(ex->first_block);
1308 start = ext4_ext_pblock(ex) + num;
1309 ext4_dbg(DEBUG_EXTENT,
1310 "Freeing %" PRIu32 " at %" PRIu64 ", %" PRIu32 "\n", from,
1313 ext4_ext_free_blocks(inode_ref, start, len, 0);
1316 static int ext4_ext_remove_idx(struct ext4_inode_ref *inode_ref,
1317 struct ext4_extent_path *path, int32_t depth)
1323 /* free index block */
1324 leaf = ext4_idx_pblock(path[i].index);
1326 if (path[i].index != EXT_LAST_INDEX(path[i].header)) {
1327 ptrdiff_t len = EXT_LAST_INDEX(path[i].header) - path[i].index;
1328 memmove(path[i].index, path[i].index + 1,
1329 len * sizeof(struct ext4_extent_index));
1332 path[i].header->entries_count =
1333 to_le16(to_le16(path[i].header->entries_count) - 1);
1334 err = ext4_ext_dirty(inode_ref, path + i);
1338 ext4_dbg(DEBUG_EXTENT, "IDX: Freeing %" PRIu32 " at %" PRIu64 ", %d\n",
1339 to_le32(path[i].index->first_block), leaf, 1);
1340 ext4_ext_free_blocks(inode_ref, leaf, 1, 0);
1343 if (path[i].index != EXT_FIRST_INDEX(path[i].header))
1346 path[i - 1].index->first_block = path[i].index->first_block;
1347 err = ext4_ext_dirty(inode_ref, path + i - 1);
1356 static int ext4_ext_remove_leaf(struct ext4_inode_ref *inode_ref,
1357 struct ext4_extent_path *path, ext4_lblk_t from,
1361 int32_t depth = ext_depth(inode_ref->inode);
1362 struct ext4_extent *ex = path[depth].extent;
1363 struct ext4_extent *start_ex, *ex2 = NULL;
1364 struct ext4_extent_header *eh = path[depth].header;
1367 uint16_t new_entries;
1370 new_entries = to_le16(eh->entries_count);
1371 while (ex <= EXT_LAST_EXTENT(path[depth].header) &&
1372 to_le32(ex->first_block) <= to) {
1373 int32_t new_len = 0;
1375 ext4_lblk_t start, new_start;
1376 ext4_fsblk_t newblock;
1377 new_start = start = to_le32(ex->first_block);
1378 len = ext4_ext_get_actual_len(ex);
1379 newblock = ext4_ext_pblock(ex);
1381 len -= from - start;
1382 new_len = from - start;
1386 if (start + len - 1 > to) {
1387 len -= start + len - 1 - to;
1388 new_len = start + len - 1 - to;
1390 newblock += to + 1 - start;
1395 ext4_ext_remove_blocks(inode_ref, ex, start, start + len - 1);
1396 ex->first_block = to_le32(new_start);
1400 unwritten = ext4_ext_is_unwritten(ex);
1401 ex->block_count = to_le16(new_len);
1402 ext4_ext_store_pblock(ex, newblock);
1404 ext4_ext_mark_unwritten(ex);
1413 if (ex2 <= EXT_LAST_EXTENT(eh))
1414 memmove(start_ex, ex2, EXT_LAST_EXTENT(eh) - ex2 + 1);
1416 eh->entries_count = to_le16(new_entries);
1417 ext4_ext_dirty(inode_ref, path + depth);
1418 if (path[depth].extent == EXT_FIRST_EXTENT(eh) && eh->entries_count)
1419 err = ext4_ext_correct_indexes(inode_ref, path);
1421 /* if this leaf is free, then we should
1422 * remove it from index block above */
1423 if (err == EOK && eh->entries_count == 0 && path[depth].block.lb_id)
1424 err = ext4_ext_remove_idx(inode_ref, path, depth - 1);
1426 path[depth - 1].index++;
1431 static bool ext4_ext_more_to_rm(struct ext4_extent_path *path, ext4_lblk_t to)
1433 if (!to_le16(path->header->entries_count))
1436 if (path->index > EXT_LAST_INDEX(path->header))
1439 if (to_le32(path->index->first_block) > to)
1445 int ext4_extent_remove_space(struct ext4_inode_ref *inode_ref, ext4_lblk_t from,
1448 struct ext4_extent_path *path = NULL;
1450 int32_t depth = ext_depth(inode_ref->inode);
1453 ret = ext4_find_extent(inode_ref, from, &path, 0);
1457 if (!path[depth].extent) {
1462 bool in_range = IN_RANGE(from, to_le32(path[depth].extent->first_block),
1463 ext4_ext_get_actual_len(path[depth].extent));
1470 /* If we do remove_space inside the range of an extent */
1471 if ((to_le32(path[depth].extent->first_block) < from) &&
1472 (to < to_le32(path[depth].extent->first_block) +
1473 ext4_ext_get_actual_len(path[depth].extent) - 1)) {
1475 struct ext4_extent *ex = path[depth].extent, newex;
1476 int unwritten = ext4_ext_is_unwritten(ex);
1477 ext4_lblk_t ee_block = to_le32(ex->first_block);
1478 int32_t len = ext4_ext_get_actual_len(ex);
1479 ext4_fsblk_t newblock =
1480 to + 1 - ee_block + ext4_ext_pblock(ex);
1482 ex->block_count = to_le16(from - ee_block);
1484 ext4_ext_mark_unwritten(ex);
1486 ext4_ext_dirty(inode_ref, path + depth);
1488 newex.first_block = to_le32(to + 1);
1489 newex.block_count = to_le16(ee_block + len - 1 - to);
1490 ext4_ext_store_pblock(&newex, newblock);
1492 ext4_ext_mark_unwritten(&newex);
1494 ret = ext4_ext_insert_extent(inode_ref, &path, &newex, 0);
1501 struct ext4_extent_header *eh;
1502 struct ext4_extent *first_ex, *last_ex;
1503 ext4_lblk_t leaf_from, leaf_to;
1504 eh = path[i].header;
1505 ext4_assert(to_le16(eh->entries_count) > 0);
1506 first_ex = EXT_FIRST_EXTENT(eh);
1507 last_ex = EXT_LAST_EXTENT(eh);
1508 leaf_from = to_le32(first_ex->first_block);
1509 leaf_to = to_le32(last_ex->first_block) +
1510 ext4_ext_get_actual_len(last_ex) - 1;
1511 if (leaf_from < from)
1517 ext4_ext_remove_leaf(inode_ref, path, leaf_from,
1519 ext4_ext_drop_refs(inode_ref, path + i, 0);
1524 struct ext4_extent_header *eh;
1525 eh = path[i].header;
1526 if (ext4_ext_more_to_rm(path + i, to)) {
1527 struct ext4_block bh = EXT4_BLOCK_ZERO();
1528 if (path[i + 1].block.lb_id)
1529 ext4_ext_drop_refs(inode_ref, path + i + 1, 0);
1531 ret = read_extent_tree_block(inode_ref,
1532 ext4_idx_pblock(path[i].index),
1533 depth - i - 1, &bh, 0);
1538 ext4_idx_pblock(path[i].index);
1539 path[i + 1].block = bh;
1540 path[i + 1].header = ext_block_hdr(&bh);
1541 path[i + 1].depth = depth - i - 1;
1543 path[i + 1].extent = EXT_FIRST_EXTENT(
1544 path[i + 1].header);
1547 EXT_FIRST_INDEX(path[i + 1].header);
1552 if (!eh->entries_count)
1553 ret = ext4_ext_remove_idx(inode_ref, path,
1556 path[i - 1].index++;
1561 ext4_block_set(inode_ref->fs->bdev,
1570 /* TODO: flexible tree reduction should be here */
1571 if (path->header->entries_count == 0) {
1573 * truncate to zero freed all the tree,
1574 * so we need to correct eh_depth
1576 ext_inode_hdr(inode_ref->inode)->depth = 0;
1577 ext_inode_hdr(inode_ref->inode)->max_entries_count =
1578 to_le16(ext4_ext_space_root(inode_ref));
1579 ret = ext4_ext_dirty(inode_ref, path);
1583 ext4_ext_drop_refs(inode_ref, path, 0);
1589 static int ext4_ext_split_extent_at(struct ext4_inode_ref *inode_ref,
1590 struct ext4_extent_path **ppath,
1591 ext4_lblk_t split, uint32_t split_flag)
1593 struct ext4_extent *ex, newex;
1594 ext4_fsblk_t newblock;
1595 ext4_lblk_t ee_block;
1597 int32_t depth = ext_depth(inode_ref->inode);
1600 ex = (*ppath)[depth].extent;
1601 ee_block = to_le32(ex->first_block);
1602 ee_len = ext4_ext_get_actual_len(ex);
1603 newblock = split - ee_block + ext4_ext_pblock(ex);
1605 if (split == ee_block) {
1607 * case b: block @split is the block that the extent begins with
1608 * then we just change the state of the extent, and splitting
1611 if (split_flag & EXT4_EXT_MARK_UNWRIT2)
1612 ext4_ext_mark_unwritten(ex);
1614 ext4_ext_mark_initialized(ex);
1616 err = ext4_ext_dirty(inode_ref, *ppath + depth);
1620 ex->block_count = to_le16(split - ee_block);
1621 if (split_flag & EXT4_EXT_MARK_UNWRIT1)
1622 ext4_ext_mark_unwritten(ex);
1624 err = ext4_ext_dirty(inode_ref, *ppath + depth);
1628 newex.first_block = to_le32(split);
1629 newex.block_count = to_le16(ee_len - (split - ee_block));
1630 ext4_ext_store_pblock(&newex, newblock);
1631 if (split_flag & EXT4_EXT_MARK_UNWRIT2)
1632 ext4_ext_mark_unwritten(&newex);
1633 err = ext4_ext_insert_extent(inode_ref, ppath, &newex,
1634 EXT4_EXT_NO_COMBINE);
1636 goto restore_extent_len;
1641 ex->block_count = to_le16(ee_len);
1642 err = ext4_ext_dirty(inode_ref, *ppath + depth);
1646 static int ext4_ext_convert_to_initialized(struct ext4_inode_ref *inode_ref,
1647 struct ext4_extent_path **ppath,
1648 ext4_lblk_t split, uint32_t blocks)
1650 int32_t depth = ext_depth(inode_ref->inode), err = EOK;
1651 struct ext4_extent *ex = (*ppath)[depth].extent;
1653 ext4_assert(to_le32(ex->first_block) <= split);
1655 if (split + blocks ==
1656 to_le32(ex->first_block) + ext4_ext_get_actual_len(ex)) {
1657 /* split and initialize right part */
1658 err = ext4_ext_split_extent_at(inode_ref, ppath, split,
1659 EXT4_EXT_MARK_UNWRIT1);
1660 } else if (to_le32(ex->first_block) == split) {
1661 /* split and initialize left part */
1662 err = ext4_ext_split_extent_at(inode_ref, ppath, split + blocks,
1663 EXT4_EXT_MARK_UNWRIT2);
1665 /* split 1 extent to 3 and initialize the 2nd */
1666 err = ext4_ext_split_extent_at(inode_ref, ppath, split + blocks,
1667 EXT4_EXT_MARK_UNWRIT1 |
1668 EXT4_EXT_MARK_UNWRIT2);
1670 err = ext4_ext_split_extent_at(inode_ref, ppath, split,
1671 EXT4_EXT_MARK_UNWRIT1);
1678 static ext4_lblk_t ext4_ext_next_allocated_block(struct ext4_extent_path *path)
1682 depth = path->depth;
1684 if (depth == 0 && path->extent == NULL)
1685 return EXT_MAX_BLOCKS;
1687 while (depth >= 0) {
1688 if (depth == path->depth) {
1690 if (path[depth].extent &&
1691 path[depth].extent !=
1692 EXT_LAST_EXTENT(path[depth].header))
1694 path[depth].extent[1].first_block);
1697 if (path[depth].index !=
1698 EXT_LAST_INDEX(path[depth].header))
1700 path[depth].index[1].first_block);
1705 return EXT_MAX_BLOCKS;
1708 static int ext4_ext_zero_unwritten_range(struct ext4_inode_ref *inode_ref,
1710 uint32_t blocks_count)
1714 uint32_t block_size = ext4_sb_get_block_size(&inode_ref->fs->sb);
1715 for (i = 0; i < blocks_count; i++) {
1716 struct ext4_block bh = EXT4_BLOCK_ZERO();
1717 err = ext4_block_get(inode_ref->fs->bdev, &bh, block + i);
1721 memset(bh.data, 0, block_size);
1723 err = ext4_block_set(inode_ref->fs->bdev, &bh);
1730 int ext4_extent_get_blocks(struct ext4_inode_ref *inode_ref, ext4_fsblk_t iblock,
1731 uint32_t max_blocks, ext4_fsblk_t *result, bool create,
1732 uint32_t *blocks_count)
1734 struct ext4_extent_path *path = NULL;
1735 struct ext4_extent newex, *ex;
1739 uint32_t allocated = 0;
1740 ext4_fsblk_t next, newblock;
1748 /* find extent for this block */
1749 err = ext4_find_extent(inode_ref, iblock, &path, 0);
1755 depth = ext_depth(inode_ref->inode);
1758 * consistent leaf must not be empty
1759 * this situations is possible, though, _during_ tree modification
1760 * this is why assert can't be put in ext4_ext_find_extent()
1762 ex = path[depth].extent;
1764 ext4_lblk_t ee_block = to_le32(ex->first_block);
1765 ext4_fsblk_t ee_start = ext4_ext_pblock(ex);
1766 uint16_t ee_len = ext4_ext_get_actual_len(ex);
1767 /* if found exent covers block, simple return it */
1768 if (IN_RANGE(iblock, ee_block, ee_len)) {
1769 /* number of remain blocks in the extent */
1770 allocated = ee_len - (iblock - ee_block);
1772 if (!ext4_ext_is_unwritten(ex)) {
1773 newblock = iblock - ee_block + ee_start;
1782 uint32_t zero_range;
1783 zero_range = allocated;
1784 if (zero_range > max_blocks)
1785 zero_range = max_blocks;
1787 newblock = iblock - ee_block + ee_start;
1788 err = ext4_ext_zero_unwritten_range(inode_ref, newblock,
1793 err = ext4_ext_convert_to_initialized(inode_ref, &path,
1794 iblock, zero_range);
1803 * requested block isn't allocated yet
1804 * we couldn't try to create block if create flag is zero
1810 /* find next allocated block so that we know how many
1811 * blocks we can allocate without ovelapping next extent */
1812 next = ext4_ext_next_allocated_block(path);
1813 allocated = next - iblock;
1814 if (allocated > max_blocks)
1815 allocated = max_blocks;
1817 /* allocate new block */
1818 goal = ext4_ext_find_goal(inode_ref, path, iblock);
1819 newblock = ext4_new_meta_blocks(inode_ref, goal, 0, &allocated, &err);
1823 /* try to insert new extent into found leaf and return */
1824 newex.first_block = to_le32(iblock);
1825 ext4_ext_store_pblock(&newex, newblock);
1826 newex.block_count = to_le16(allocated);
1827 err = ext4_ext_insert_extent(inode_ref, &path, &newex, 0);
1829 /* free data blocks we just allocated */
1830 ext4_ext_free_blocks(inode_ref, ext4_ext_pblock(&newex),
1831 to_le16(newex.block_count), 0);
1835 /* previous routine could use block we allocated */
1836 newblock = ext4_ext_pblock(&newex);
1839 if (allocated > max_blocks)
1840 allocated = max_blocks;
1846 *blocks_count = allocated;
1850 ext4_ext_drop_refs(inode_ref, path, 0);